Handbook of
NON-INVASIVE METHODS and the
SKIN S E C O N D
E D I T I O N Edited by
Jorgen Serup Gregor B.E. Jemec Ga...
67 downloads
1615 Views
33MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Handbook of
NON-INVASIVE METHODS and the
SKIN S E C O N D
E D I T I O N Edited by
Jorgen Serup Gregor B.E. Jemec Gary L. Grove
Boca Raton London New York
A CRC title, part of the Taylor & Francis imprint, a member of the Taylor & Francis Group, the academic division of T&F Informa plc.
Handbook of
NON-INVASIVE METHODS and the
SKIN S E C O N D
E D I T I O N
Published in 2006 by CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2006 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group 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: 0-8493-1437-2 (Hardcover) International Standard Book Number-13: 978-0-8493-1437-7 (Hardcover) Library of Congress Card Number 2005045688 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 consequences 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 Handbook of non-invasive methods and the skin / edited by Jørgen Serup, Gregor B.E. Jemec, Gary L. Grove.--2nd ed. p. ; cm. Includes bibliographical references and index. ISBN 0-8493-1437-2 (alk. paper) 1. Skin--Diseases--Diagnosis. 2. Skin--Imaging. 3. Diagnosis, Noninvasive. 4. Bioengineering. I. Serup, Jørgen. II. Jemec, B.E. III. Grove, Gary L. (Gary Lee) [DNLM: 1. Skin Physiology. 2. Biomedical Engineering. 3. Skin Diseases--diagnosis. WR 102 H236 2005] RL105.H34 2005 616.5’075--dc22
2005045688
Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com Taylor & Francis Group is the Academic Division of Informa plc.
and the CRC Press Web site at http://www.crcpress.com
This handbook is dedicated to Professor Albert M. Kligman, who searched truth and gave so much to so many
Preface This second edition of the Handbook of Non-Invasive Methods and the Skin contains over 100 chapters on bioinstrumental examination of skin including classical reviews and 75 entirely new and updated chapters. The first edition published in 1995 included 89 chapters. In the meantime, much has happened in the field of imaging methods and computer-based techniques, and a number of advanced instruments were introduced. This development as well as the ongoing development of the classical instruments is covered in the second edition. Dr. Gary Grove has joined as co-editor of the second edition. The main purpose of the book is to review important techniques and present key information of practical importance as a guide both to the young and the senior researcher. It is aimed to be a useful guide in academia as well as product development. Taylor & Francis, along with Professor Howard I. Maibach as the main editor, also publishes a number of monographs on specific methods. Another important source of information is the Skin Research and Technology journal (http://blackwellpublishing.com/cservices), official publication of the International Society for Bioengineering and the Skin (ISBS), the International Society for Skin Imaging (ISSI) and the International Society for Digital Imaging of Skin (ISDIS). In October 2005, the ISBS and ISSI decided to merge and form the International Society of Biophysics and Imaging of the Skin (ISBS). This indexed journal publishes original research in the field. In the year 2000, Professsor Pierre Agache, Besancon published the handbook Physiologie de la Peau et Explorations Fonctionelles Cutanées with extensive reviews on bio-instrumental methods written by himself and many French authors. It is to our deep sorrow that Pierre Agache is no longer with us.
The second edition would not have been possible without the generous support and patience of the many contributors, who were invited as internationally recognized experts in the field. The editors wish to extend their gratitude to the contributors. Taylor & Francis has done tremendous work with the book. The support of Barbara Norwitz, Susan Fox-Greenberg and Erika Dery is acknowledged. The second edition is dedicated to Professor Albert M. Kligman of University of Pennsylvania, the Duhring Laboratory, as was the first edition. A whole generation of bio-instrumentalists have been inspired by his visions, sharpness, courage and dedication to truth, and many all over the world enjoy his generosity and personal friendship. Soon 90 and beyond aging, Albert remains at full speed. Jørgen Serup MD, DMSc, Chief-editor of the handbook Professor of Dermatology Linköping University, Sweden Bispebjerg Hospital, Copenhagen, Denmark Gregor BE Jemec MD, DMSc, Co-editor of the handbook Associate Professor of Dermatology University of Copenhagen Roskilde Hospital, Denmark Gary Grove PhD, Co-editor of the handbook Chief Scientist cyberDERM, Inc, Broomall, Pennsylvania
Contributors P. Åberg Karolinska Institutet Medical Engineering Novum Research Park Huddinge, Sweden Pierre G. Agache Department of Functional Dermatology University Hospital Besançon, France Peter J. Altmeyer (Deceased) Dermatological Clinic Ruhr University Bochum Bochum, Germany Peter Andersen Optics and Plasma Research Department Risø National Laboratory Roskilde, Denmark Lars Arndt-Nielsen Center for Sensory-Motor Interaction Laboratory for Experimental Pain Research Aalborg University Aalborg, Denmark Jorge E. Arrese Department of Dermatopathology University Hospital Sart Tilman Liège, Belgium Sitke Aygen Institut für Zentrale Analytik und Strukturanalyse Universität Witten/Herdecke Witten, Germany Lora Bankova Department of Dermatology Friedrich Schiller University Jena, Germany J.C. Barbenel Bioengineering Unit University of Strathclyde Glasgow, Scotland
André O. Barel Faculty of Physical Education and Physiotherapy Vrije Universiteit Brussel Brussels, Belgium Julian H. Barth Department of Chemical Pathlology General Infirmary at Leeds Leeds, United Kingdom T. Bauermann Institut für Zentrale Analytik und Strukturanalyse Universität Witten/Herdecke Witten, Germany Claus Bay Mathematical Statistical Department LEO Pharma Ballerup, Denmark Gianni Belcaro Irvine Laboratory for Cardiovascular Investigation and Research St. Mary’s Hospital Medical School London, United Kingdom Eva Benfeldt Department of Dermatology University of Copenhagen Bispebjerg Hospital Copenhagen, Denmark Enzo Berardesca Department of Dermatology Istituto Dermatologico di S. Maria e S. Gallicano Rome, Italy Andreas J. Bircher Department of Dermatology University Hospital Basel, Switzerland Ulrike Blume-Peytavi Department of Dermatology Hospital Charité Humboldt University Berlin, Germany
C.H. Chang Department of Dermatology College of Medicine Tzu-Chi Medical University Hualien, Taiwan
Shabtay Dikstein Unit of Cell Pharmacology School of Pharmacy The Hebrew University of Jerusalem Jerusalem, Israel
Steen Christiansen Technical University of Denmark Institute of Manufacturing Engineering Lyngby, Denmark
Peter Dykes Cutest Systems Ltd. Cardiff, United Kingdom
E. Claridge School of Computer Science University of Birmingham Birmingham, United Kingdom Peter Clarys Faculty of Physical Education and Physiotherapy Vrije Universiteit Brussel Brussels, Belgium Pierre Corcuff Laboratoires de Recherche de L’Oreal Aulnay Sous Bois, France
E. Anne Eady Department of Microbiology University of Leeds Leeds, United Kingdom C. Edwards Cardiff Biometrics Ltd. Cardiff, United Kingdom Howell G.M. Edwards Chemical and Forensic Sciences School of Pharmacy University of Bradford Bradford, United Kingdom
S. Cotton Astron Clinica The Mount Cambridge, United Kingdom
Jan Efsen Novo Nordisk A/S Bagsvaerd, Denmark
W. Courage Courage + Khazaka Electronic GmbH Köln, Germany
Mariko Egawa Bioengineering Research Labs Shiseido Co., Ltd. Yokohama, Japan
Pierre Creidi Laboratoire de Biologie et d’Ingénierie Cutanée Besançon, France W.J. Cunliffe Leeds Foundation for Dermatological Research Leeds General Infirmary Leeds, United Kingdom John Damia cyberDERM, Inc. Broomall, Pennsylvania David de Berker Bristol Dermatology Centre Bristol Royal Infirmary Bristol, United Kingdom Mitsuhiro Denda Shiseido Research Center Yokohama, Japan
Claudia El Gammal Dermatological Clinic Hospital Bethesda Freudenberg, Germany Stephan El Gammal Dermatological Clinic Hospital Bethesda Freudenberg, Germany Helmut Ermert Institute for High Frequency Techniques Ruhr University Bochum Bochum, Germany Jan Faergemann Department of Dermatology University of Gotenburg Sahlgren’s Hospital Gothenburg, Sweden
Nadia Farinelli Department of Dermatology University of Pavia Pavia, Italy Joachim Fluhr Department of Dermatology Friedrich Schiller University Jena, Germany Bo Forslind (Deceased) Department of Medicine Biochemistry and Biophysics Karolinska Institute Stockholm, Sweden Ann Fullerton LEO Pharma Ballerup, Denmark Bernard Gabard Egerkingen, Switzerland Johannes Gassmueller BioSkin Institut für Dermatologische Forschung und Entwicklung GmbH Hamburg, Germany Tijani Gharbi Laboratoire d’Optique P.M. Duffieux University of Franche-Comté Besançon, France Yolanda Gilaberte-Calzada San Jorge Hospital Huesca, Spain
Chee Leok Goh National Skin Centre Singapore Salvador González Dermatology Service Memorial Sloan-Kettering Cancer Center New York, New York and Department of Dermatology Massachusetts General Hospital Harvard Medical School Boston, Massachusetts and Clinica La Luz Madrid, Spain Costantino Grana Department of Computer Engineering University of Modena and Reggio Emilia Modena, Italy Lotte Groth LEO Pharma Ballerup, Denmark Gary Lee Grove cyberDERM, Inc. Broomall, Pennsylvania Mary Jo Grove cyberDERM, Inc. Broomall, Pennsylvania
Norm V. Gitis Center for Tribology, Inc. Campbell, California
P.N. Hall Department of Plastic Surgery Addenbrooke’s Hospital Cambridge, United Kingdom
Francesca Giusti Department of Dermatology University of Modena and Reggio Emilia Modena, Italy
Allan Halpern Dermatology Service Memorial Sloan-Kettering Cancer Center New York, New York
Monika Gniadecka Department of Dermatology Bispebjerg Hospital Copenhagen, Denmark
Hans Nørgaard Hansen Copenhagen, Denmark
Robert Gniadecki Department of Dermatoloy Bispebjerg Hospital Copenhagen, Denmark
C.W. Hargens Philadelphia, Pennsylvania Roland Hartwig Dermatological Practice Wuppertal, Germany
Stacy S. Hawkins Unilever Research, U.S. Edgewater, New Jersey
H. Irving Katz Minnesota Clinical Study Center Fridley, Minnesota
Andreas Herpens Research Bioengineering — Biophysics Beiersdorf AG Hamburg, Germany
Andrei Kecskés Schering AG Berlin, Germany
Jutta Hofmann proDERM Institute for Applied Dermatological Research Schenefeld/Hamburg, Germany K. Hoffmann Dermatologische Klinik der Ruhr Universität Bochum Bochum, Germany Elisabeth A. Holm Department of Dermatology Roskilde Hospital University of Copenhagen Roskilde, Denmark Takeshi Horio Department of Dermatology Kansai Medical University Osaka, Japan Sidney B. Hornby Neutrogena Corporation Los Angeles, California Philippe Humbert Laboratoire de Biologie et d’Ingénierie Cutanée Besançon, France Pedro Jaén-Olasold Dermatology Service Ramon y Cajal Hospital and Clinica La Luz Madrid, Spain Peter Jahn Schering AG, Diagnostika Koordination Berlin, Germany Gregor B.E. Jemec Department of Dermatology Roskilde Hospital University of Copenhagen Roskilde, Denmark
Jens Keiding LEO Pharma Ballerup, Denmark Nis Kentorp Department of Dermatology Bispebjerg Hospital Copenhagen, Denmark Albert M. Kligman Department of Dermatology University of Pennsylvania Philadelphia, Pennsylvania Jürgen Lademann Center of Experimental and Applied Cutaneous Physiology (CCP) Department of Dermatology University Hospital Charité Humboldt University Berlin, Germany Nicholas Lange Biometric and Field Studies Branch National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, Maryland C.H. Lee Department of Dermatology College of Medicine Kaohsiung Medical University Kaohsiung, Taiwan Jean-Luc Lévêque Laboratoires de Recherche de L’Oreal Aulnay Sous Bois, France Jane S. Lindholm Minnesota Clinical Study Center Fridley, Minnesota Sophie Mac-Mary Laboratoire de Biologie et d’Ingénierie Cutanée Besançon, France
Howard I. Maibach Department of Dermatology School of Medicine University of California San Francisco, California
Andrew N. Nicolaides Irvine Laboratory for Cardiovascular Investigation and Research St. Mary’s Hospital Medical School London, United Kingdom
Mette Midttun Department of Medical Physiology The Panum Institute University of Copenhagen Copenhagen, Denmark
Mikkel Noerreslet The Danish University of Pharmaceutical Sciences Copenhagen, Denmark
Jean Mignot Laboratoire de Métrologie des Interfaces Techniques Institut Universitaire de Technologie Besançon, France David L. Miller Bionet Incorporated Dallas, Texas Otto H. Mills, Jr. Hill Top Research, Inc. University of Medicine and Dentistry of New Jersey Robert Wood Johnson Medical School East Brunswick, New Jersey Susanne Møller Mathematical Statistical Department LEO Pharma Ballerup, Denmark M. Moncrieff Department of Plastic Surgery Addenbrooke’s Hospital Cambridge, United Kingdom P.S. Mortimer St. George’s and Royal Marsden Hospitals London, United Kingdom I. Nicander Department of Dermatology Huddinge University Hospital Huddinge, Sweden Carsten N. Nickelsen Hvidovre Hospital University of Copenhagen Hvidovre, Denmark
J. Nuutinen Delfin Technologies Ltd. Kuopio, Finland Ken-ichiro O’goshi Department of Dermatology Bispebjerg Hospital Copenhagen, Denmark Motoki Oguri Shiseido Research Center Yokohama-shi, Japan Chil Hwan Oh Department of Dermatology School of Medicine Korea University Seoul, Korea Hans Öhman Department of Dermato-Venereology University of Linköping Linköping, Sweden S. Ollmar Karolinska Institutet Medical Engineering Novum Research Park Huddinge, Sweden Constantin E. Orfanos Department of Dermatology University Medical Center Steglitz The Free University of Berlin Berlin, Germany Patricia García Ortiz Department of Dermatology University of Copenhagen Gentofte Hospital Hellerup, Denmark Alessandra Pagnoni Philadelphia, Pennsylvania
Giovanni Pellacani Department of Dermatology University of Modena and Reggio Emilia Modena, Italy Adeline Petitjean Laboratoire de Biologie et d’Ingénierie Cutanée Besançon, France Claudine Piérard-Franchimont Department of Dermatopathology University Hospital Sart Tilman Liège, Belgium Gérald E. Piérard Department of Dermatopathology University Hospital Sart Tilman Liège, Belgium J. Pinnagoda Ministry of Public Health Singapore Pascale Quatresooz Department of Dermatopathology University Hospital Sart Tilman Liège, Belgium Bernard Querleux Laboratoires de Recherche de L’Oréal Aulnay-sous-bois, France Milind Rajadhyaksha Dermatology Service Memorial Sloan-Kettering Cancer Center New York, New York E.F.J. Ring Medical Imaging Research Group School of Computing University of Glamorgan Pontypridd, United Kingdom Jeffrey S. Roth Department of Dermatology College of Physicians and Surgeons Columbia University New York, New York D. Hugh Rushton School of Pharmacy and Biomedical Sciences University of Portsmouth Portsmouth, United Kingdom
Iqbal Sadiq S.K.I.N., Inc. Conshohocken, Pennsylvania Jean-Marie Sainthillier Laboratoire de Biologie et d’Ingénierie Cutanée Besançon, France S. Schagen Research Bioengineering — Biophysics Beiersdorf AG Hamburg, Germany Harald Schatz Dermatologische Klinik der Ruhr Universität Bochum Bochum, Germany S. Scheede Research Bioengineering — Biophysics Beiersdorf AG Hamburg, Germany Richard K. Scher Department of Dermatology College of Physicians and Surgeons Columbia University New York, New York Stefania Seidenari Department of Dermatology University of Modena and Reggio Emilia Modena, Italy Per Sejrsen Department of Medical Physiology The Panum Institute University of Copenhagen Copenhagen, Denmark Jørgen Serup Department of Dermatology Linköping University Linköping, Sweden and Department of Dermatology Bispebjerg Hospital Copenhagen, Denmark Raja K. Sivamani Department of Dermatology School of Medicine University of California San Francisco, California
Tracy Stoudemayer S.K.I.N., Inc. Conshohocken, Pennsylvania Hachiro Tagami Department of Dermatology Tohoku University School of Medicine Sendai, Japan
Abel Torres Division of Dermatology Loma Linda University Hospital Loma Linda, California R.A. Tupker Department of Dermatology St. Antonius Hospital Nieuwegein, The Netherlands
Motoji Takahashi Shiseido Research Center Yokohama-shi, Japan and Bioengineering Research Labs Shiseido Co., Ltd. Yokohama, Japan
Andreas Tycho OCT Innovation ApS. Roskilde, Denmark
Hirotsugu Takiwaki Department of Dermatology The University of Tokushima School of Medicine Tokushima, Japan
Michael Vogt Institute for High Frequency Techniques Ruhr University Bochum Bochum, Germany
M. Tanaka Department of Bioengineering and Robotics Graduate School of Engineering Tohoku University Sendai, Japan
Karin Wårdell Department of Biomedical Engineering Linköping University Linköping, Sweden
J.P. Taylor Leeds Foundation for Dermatological Research Leeds General Infirmary Leeds, United Kingdom Roderick A. Thomas Snell International Tyn-Y-Coed Pontardulais, Swansea, United Kingdom Steven G. Thomas Optiscan Pty. Ltd. Notting Hill, Victoria, Australia Lars Thrane Optics and Plasma Research Department Risø National Laboratory Roskilde, Denmark Merete Thyme Quality Assurance Department Scantox (part of LAB Research International) LI. Skensved, Denmark
D. Van Neste Skinterface Tournai, Belgium
Wiete Westerhof Department of Dermatology Academic Medical Center University of Amsterdam Amsterdam, The Netherlands Martin A. Weinstock Dermatoepidemiology Unit VA Medical Center Roger Williams Medical Center and Brown University Providence, Rhode Island Elizabeth Grove Wickersheim cyberDERM, Inc. Broomall, Pennsylvania R. Randall Wickett College of Pharmacy University of Cincinnati Cincinnati, Ohio Klaus-Peter Wilhelm proDERM Institute for Applied Dermatological Research Schenefeld/Hamburg, Germany
Ximena Wortsman Servicio de Imagenologia Hospital del Profesor Santiago, Chile Gabriel Wu Department of Dermatology School of Medicine University of California San Francisco, California Hans Christian Wulf Department of Dermatology Bispebjerg Hospital University of Copenhagen Copenhagen, Denmark Toyonobu Yamashita Bioengineering Research Labs Shiseido Co., Ltd. Yokohama, Japan
H.S. Yu Department of Dermatology College of Medicine National Taiwan University Hospital Taipei, Taiwan H. Zahouani Laboratoire de Tribologie et Dynamique des Systems Ecully, France A. Zemtsov University Dermatology Center, P.C. Muncie, Indiana Charles Zerweck cyberDERM, Inc. Broomall, Pennsylvania Burton Zweiman University of Pennsylvania School of Medicine Philadelphia, Pennsylvania
Table of Contents SECTION I
General Introduction
Chapter 1 Personal Perspectives on Bioengineering and the Skin: The Successful Past and the Brilliant Future ...........................3 Albert M. Kligman Chapter 2 How to Choose and Use Non-Invasive Methods ...............................................................................................................9 Jørgen Serup Chapter 3 A Practical Guide to Resources on the Internet for the Skin Researcher .......................................................................15 Elizabeth Grove Wickersheim and Gary Lee Grove Chapter 4 The Skin Integument: Variation Relative to Sex, Age, Race, and Body Region ............................................................27 Nadia Farinelli and Enzo Berardesca Chapter 5 Seasonal Variations and Environmental Influences on the Skin ......................................................................................33 Chee Leok Goh Chapter 6 Non-Invasive Methods and Assessment of Skin Diseases ...............................................................................................37 Stefania Seidenari, Francesca Giusti, and Giovanni Pellacani Chapter 7 Standards for Good Clinical Practice (GCP)....................................................................................................................47 Merete Thyme Chapter 8 Statistical Analysis of Sensitivity, Specificity, and Predictive Value of a Diagnostic Test .............................................53 Nicholas Lange and Martin A. Weinstock Chapter 9 Sample Size Calculation ...................................................................................................................................................63 Claus Bay and Susanne Møller Chapter 10 Implementation of a Quality Management System in a Contract Laboratory Working with Non-Invasive Methods.......................................................................................................................................................67 Klaus-Peter Wilhelm and Jutta Hofmann
Chapter 11 Ethical Considerations.......................................................................................................................................................73 Mikkel Noerreslet and Gregor B.E. Jemec
SECTION II
Technique, Application, and Validation Skin Surface, Epidermal Structure, and Function
Clinical Photography, Surface Imaging Techniques, and Computerized Image Analysis Chapter 12 General Aspects in Medical/Clinical Photography...........................................................................................................81 Nis Kentorp Chapter 13 Use of Compact Digital Camera for Snap Photography..................................................................................................89 Ken-ichiro O’goshi Chapter 14 Computerized Image Analysis of Clinical Photos............................................................................................................95 Stacy S. Hawkins Chapter 15 Magnifying Lens — Non-Invasive Oil Immersion Examination of the Skin ...............................................................101 H. Irving Katz and Jane S. Lindholm Chapter 16 Dermatoscopy..................................................................................................................................................................109 Wiete Westerhof Chapter 17 Fiber-Optic Microscopy System for Skin Surface Imaging...........................................................................................125 Iqbal Sadiq and Tracy Stoudemayer Chapter 18 Automated Assessment of Pigment Distribution and Color Areas for Melanoma Diagnosis ......................................135 Stefania Seidenari, Giovanni Pellacani, and Costantino Grana Skin Surface Contour and Roughness Assessment Chapter 19 Skin Replication for Light and Scanning Electron Microscopy ....................................................................................147 Bo Forslind Chapter 20 Skin Surface Replica Image Analysis of Furrows and Wrinkles...................................................................................155 Pierre Corcuff and Jean-Luc Lévêque Chapter 21 Stylus Method for Skin Surface Contour Measurement ................................................................................................163 Johannes Gassmueller, Andrei Kecskés, and Peter Jahn
Chapter 22 Laser Profilometry...........................................................................................................................................................169 Jan Efsen, Steen Christiansen, Hans Nørgaard Hansen, and Jens Keiding Chapter 23 Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief ....................................................................179 Jean Mignot Chapter 24 The Morphological Tree of the Cutaneous Network of Lines.......................................................................................195 H. Zahouani and Philippe Humbert Chapter 25 Comparison of Methodologies for Evaluation of Skin Surface Contour and Wrinkles: Advantages and Limitations............................................................................................................................................205 Motoji Takahashi and Motoki Oguri Skin Surface Friction Chapter 26 Tribological Studies on Skin: Measurement of the Coefficient of Friction ..................................................................215 Raja K. Sivamani, Gabriel Wu, Howard I. Maibach, and Norm V. Gitis Chapter 27 Skin Friction Evaluation by Unidirectional Stress Using a Friction Tester ..................................................................225 Mariko Egawa and Motoji Takahashi Chapter 28 Haptic Finger...................................................................................................................................................................233 M. Tanaka Epidermis Structure Chapter 29 Cyanoacrylate Biopsy for Cytologic Evaluation of the Epidermis................................................................................239 Jorge E. Arrese, Pascale Quatresooz, Claudine Piérard-Franchimont, and Gérald E. Piérard Chapter 30 High-Resolution Sonography of the Epidermis In Vivo.................................................................................................245 Stephan El Gammal, Claudia El Gammal, Peter J. Altmeyer, Michael Vogt, and Helmut Ermert Chapter 31 Optical Coherence Tomography in Dermatology...........................................................................................................257 Peter Andersen, Lars Thrane, Andreas Tycho, and Gregor B.E. Jemec Chapter 32 In Vivo Reflectance Mode Confocal Microscopy in Clinical and Surgical Dermatology.............................................267 Salvador González, Yolanda Gilaberte-Calzada, Pedro Jaén-Olasold, Milind Rajadhyaksha, Abel Torres, and Allan Halpern
Chapter 33 In Vivo Reflectance Mode Confocal Laser Microscopy of Melanocytic Skin Lesions.................................................277 Giovanni Pellacani and Stefania Seidenari Chapter 34 In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers ...........................................................285 Steven G. Thomas Chapter 35 In Vivo Confocal Microscopy Application in Product Research and Development......................................................297 Toyonobu Yamashita and Motoji Takahashi Chapter 36 Nuclear Magnetic Resonance (NMR) Examination of the Epidermis In Vivo..............................................................307 Bernard Querleux Chapter 37 Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications .....................................315 E. Claridge, S. Cotton, M. Moncrieff, and P.N. Hall Epidermis Hydration Chapter 38 Epidermal Hydration: Measurement of High-Frequency Electrical Conductance ........................................................329 Hachiro Tagami Chapter 39 Measurement of Epidermal Capacitance ........................................................................................................................337 André O. Barel and Peter Clarys Chapter 40 Bioimpedance as a Non-Invasive Method for Measuring Changes in Skin..................................................................345 I. Nicander, P. Åberg, and S. Ollmar Chapter 41 Comparison of Commercial Electrical Measurement Instruments for Assessing the Hydration State of the Stratum Corneum.......................................................................................................................351 Bernard Gabard, Peter Clarys, and André O. Barel Desquamation Chapter 42 Methods to Determine Desquamation Rate....................................................................................................................361 C. Edwards Chapter 43 Application of Adhesive Techniques to Harvest Stratum Corneum Material ...............................................................371 David L. Miller
Chapter 44 Dry Skin and Scaling Evaluated by D-Squames and Image Analysis ..........................................................................375 Harald Schatz, Peter J. Altmeyer, and Albert M. Kligman Barrier Functions and Gradients Chapter 45 Measurement of Transepidermal Water Loss by Semiopen Systems ............................................................................383 R.A. Tupker and J. Pinnagoda Chapter 46 Measurement of Transepidermal Water Loss by Closed-Chamber Systems .................................................................393 J. Nuutinen Chapter 47 Measurement of Transcutaneous Oxygen Tension .........................................................................................................397 Hirotsugu Takiwaki Chapter 48 Measurement of Transcutaneous PCO2 ............................................................................................................................407 Carsten N. Nickelsen Chapter 49 Skin Surface pH: Mechanism, Measurement, Importance.............................................................................................411 Joachim Fluhr, Lora Bankova, and Shabtay Dikstein Chapter 50 The pH Gradient in the Epidermis Evaluated by Serial Tape Stripping .......................................................................421 Hans Öhman Chapter 51 Techniques for Visualization of Ionic Gradation in Human Epidermis.........................................................................429 Mitsuhiro Denda Chapter 52 Skin Chamber Techniques...............................................................................................................................................433 Burton Zweiman Chapter 53 Microdialysis Methodology for Sampling in the Skin...................................................................................................443 Lotte Groth, Patricia García Ortiz, and Eva Benfeldt Skin Surface Microflora Chapter 54 Sampling the Bacteria of the Skin..................................................................................................................................457 E. Anne Eady Chapter 55 Mapping the Fungi of the Skin.......................................................................................................................................467 Jan Faergemann
Dermis Structure and Function Dermis Structure Chapter 56 High-Frequency Ultrasound Examination of Skin: Introduction and Guide.................................................................473 Jørgen Serup, Jens Keiding, Ann Fullerton, Monika Gniadecka, and Robert Gniadecki Chapter 57 Ultrasound B-Mode Imaging and In Vivo Structure Analysis .......................................................................................493 Stefania Seidenari Chapter 58 Ultrasound Assessment of Dermal Water and Edema In Vivo .......................................................................................507 Monika Gniadecka Chapter 59 Ultrasound Assessment of Skin Aging ...........................................................................................................................511 Giovanni Pellacani, Francesca Giusti, and Stefania Seidenari Chapter 60 Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures .........................................................................515 Ximena Wortsman, Elisabeth A. Holm, and Gregor B.E. Jemec Chapter 61 Magnetic Resonance Spectroscopy of the Skin .............................................................................................................531 A. Zemtsov Chapter 62 Nuclear Magnetic Resonance Examination of Skin Disorders......................................................................................537 Stephan El Gammal, Roland Hartwig, Sitke Aygen, T. Bauermann, K. Hoffmann, and Peter J. Altmeyer Chapter 63 Raman Spectroscopy of Skin..........................................................................................................................................551 Howell G.M. Edwards Mechanical Properties Chapter 64 Identification of Langer’s Lines......................................................................................................................................565 J.C. Barbenel Chapter 65 Suction Chamber Method for Measuring Skin Mechanical Properties: The Dermaflex®.............................................571 Monika Gniadecka and Jørgen Serup Chapter 66 Suction Chamber Method for Measurement of Skin Mechanics: The Cutometer® ......................................................579 Ken-ichiro O’goshi
Chapter 67 Suction Chamber Method for Measurement of Skin Mechanics: The New Digital Version of the Cutometer ................................................................................................................................................583 André O. Barel, W. Courage, and Peter Clarys Chapter 68 Suction Chamber Method for Measurement of Skin Mechanics: The DermaLab........................................................593 Gary Lee Grove, John Damia, Mary Jo Grove, and Charles Zerweck Chapter 69 Twistometry Measurement of Skin Elasticity ................................................................................................................601 Pierre G. Agache Chapter 70 Levarometry.....................................................................................................................................................................613 Shabtay Dikstein and Joachim Fluhr Chapter 71 Indentometry....................................................................................................................................................................617 Shabtay Dikstein and Joachim Fluhr Chapter 72 The Gas-Bearing Electrodynamometer...........................................................................................................................621 C.W. Hargens Chapter 73 Ballistometry ...................................................................................................................................................................627 C.W. Hargens
The Cutaneous Vasculature Skin Color and Blood Vessels Chapter 74 Colorimetry......................................................................................................................................................................635 Wiete Westerhof Chapter 75 Quasi-L*a*b* Color Measurement from Digital Images...............................................................................................649 Hirotsugu Takiwaki Chapter 76 Practical Color Calibration for Dermatoscopic Images .................................................................................................653 Constantino Grana, Giovanni Pellacani, and Stefania Seidenari Chapter 77 Measurement of Erythema and Melanin Indices............................................................................................................665 Hirotsugu Takiwaki
Chapter 78 Dynamic Capillaroscopy .................................................................................................................................................673 H.S. Yu, C.H. Lee, and C.H. Chang Chapter 79 Capillaroscopy and Videocapillaroscopy Assessment of Skin Microcirculation: Dermatological and Cosmetic Approaches......................................................................................................................................................679 Philippe Humbert, Jean-Marie Sainthillier, Sophie Mac-Mary, Adeline Petitjean, Pierre Creidi, and Tijani Gharbi Blood Flow, Vasomotion, and Vascular Functions Chapter 80 Laser Doppler Measurement of Skin Blood Flux: Variation and Validation.................................................................691 Andreas J. Bircher Chapter 81 Examination of Periodic Fluctuations in Cutaneous Blood Flow..................................................................................697 Robert Gniadecki, Monika Gniadecka, and Jørgen Serup Chapter 82 Laser Doppler Flowmetry: Principles of Technology and Clinical Applications..........................................................709 Gianni Belcaro and Andrew N. Nicolaides Chapter 83 Laser Doppler Imaging of Skin ......................................................................................................................................717 Karin Wårdell Chapter 84 The Heat Wash-In and Heat Wash-Out Technique for Quantitative, Non-Invasive Measurement of Cutaneous Blood Flow Rate................................................................................................................723 Per Sejrsen and Mette Midttun Chapter 85 The 133Xenon Wash-Out Technique for Quantitative Measurement of Cutaneous and Subcutaneous Blood Flow Rates ....................................................................................................................................733 Per Sejrsen Chapter 86 Evaluation of Lymph Flow .............................................................................................................................................741 P.S. Mortimer Temperature and Thermoregulation Chapter 87 Sensors and Handheld Devices for Surface Measurement of Skin Temperature ..........................................................753 Roderick A. Thomas Chapter 88 Thermal Imaging of Skin Temperature ..........................................................................................................................769 E.F.J. Ring
Neural Supply Chapter 89 Assessment of Cutaneous Pain .......................................................................................................................................787 Lars Arendt-Nielsen
Sweat Gland Distribution and Function Chapter 90 Classical Techniques for the Localization of Sweat Glands..........................................................................................805 Peter Dykes Chapter 91 Micro-Sensor Mapping of Sudoral Activity and Skin Surface Hydration.....................................................................811 Jean-Luc Lévêque Chapter 92 Methods for the Collection of Eccrine Sweat ................................................................................................................817 Julian H. Barth Chapter 93 Methods for the Collection of Apocrine Sweat..............................................................................................................821 Julian H. Barth
Sebaceous Glands and Sebum Excretion Chapter 94 The Follicular Biopsy......................................................................................................................................................825 Otto H. Mills, Jr. Chapter 95 Measurement of Excreted Sebum Using Sebum-Absorbent Film and an Optical Reader: The TapeAnalyzer ...........................................................................................................................................................831 David L. Miller Chapter 96 Quantification of Sebum Output Using Sebum-Absorbent Tapes (Sebutapes®)............................................................835 Claudia El Gammal, Stephan El Gammal, Alessandra Pagnoni, and Albert M. Kligman Chapter 97 Optical Measurement of Sebum Excretion Using Opalescent Film Imprint: The Sebumeter® ....................................841 Ken-ichiro O’goshi Chapter 98 Gravimetric Technique for Measuring Sebum Excretion Rate (SER) ...........................................................................847 W.J. Cunliffe and J.P. Taylor Chapter 99 Fluorescence Photography of Sebaceous Follicles.........................................................................................................853 Andreas Herpens, S. Schagen, and S. Scheede
Chapter 100 Methods for Assessment of Follicular Transport in Ex Vivo and In Vivo......................................................................861 Jürgen Lademann
Hair, Physical Properties, and Growth Rate Chapter 101 Measurement of Hair Growth .........................................................................................................................................869 Julian H. Barth and D. Hugh Rushton Chapter 102 Microscopy of the Hair: The Trichogram.......................................................................................................................875 Ulrike Blume-Peytavi and Constantin E. Orfanos Chapter 103 Photographic and Computerized Techniques for Quantification of Hair Growth .........................................................883 D. Van Neste Chapter 104 Measurement of the Mechanical Strength of Hair .........................................................................................................895 R. Randall Wickett Chapter 105 Evaluating the Strength of Human Hair .........................................................................................................................903 Sidney B. Hornby
Nail Structure and Growth Chapter 106 Methods for Nail Assessment: An Overview .................................................................................................................911 David de Berker Chapter 107 Measurement of Longitudinal Nail Growth ...................................................................................................................919 Jeffrey S. Roth and Richard K. Scher Chapter 108 Measurement of Nail Thickness .....................................................................................................................................923 Gregor B.E. Jemec Chapter 109 Image Analysis of the Nail Surface................................................................................................................................925 Claudine Piérard-Franchimont and Gérald E. Piérard
SECTION III
Clinical Experimentation, Evaluation, and Quantification
Chapter 110 General Guidelines for Assessment of Skin Diseases....................................................................................................931 Elisabeth A. Holm and Gregor B.E. Jemec
Chapter 111 Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards .....................................................943 R.A. Tupker Chapter 112 Instrumental and Computer-Based Methods for Measurement of Surface Area Afflicted with Disease .....................957 Chil Hwan Oh Chapter 113 Instrumental Evaluation of Wheal-and-Flare Reactions.................................................................................................967 D. Van Neste Chapter 114 Instrumental Evaluation of Occluded Patch Test Reactions ..........................................................................................973 Stefania Seidenari, Francesca Giusti, and Giovanni Pellacani Chapter 115 Light Sources, Sunlight, and Radiation Dosimetry........................................................................................................981 Hans Christian Wulf Chapter 116 Phototesting: Phototoxicity and Photoallergy.................................................................................................................991 Takeshi Horio Index ...............................................................................................................................................................................997
Section I General Introduction
Perspectives on 1 Personal Bioengineering and the Skin: The Successful Past and the Brilliant Future Albert M. Kligman Department of Dermatology, University of Pennsylvania, Philadelphia, Pennsylvania
CONTENTS 1.1 Some Current Issues..................................................................................................................................................4 1.2 Perspectives on the Future.........................................................................................................................................5 1.3 Contact Dermatitis .....................................................................................................................................................6 1.4 Diagnosis of Predisease.............................................................................................................................................6 1.5 Photoaging .................................................................................................................................................................6 1.6 Epilogue .....................................................................................................................................................................6 References ...........................................................................................................................................................................7
The International Society for Bioengineering and the Skin, now 25 years of age, the brainchild of Ronald Marks and Harvey Blank, has passed through its adolescent phase in good health and is now poised to reap the fruits of adulthood. We now have a cornucopia of powerful instruments that can reveal the mysteries of skin in health and disease, all without touching the surface, a triumph of bioengineering creativity. Moreover, unlike traditional histological studies, which give a static picture at one point in time, ruining the site for further study, bioengineering techniques present a moving picture of sequential events in real time, a delight to watch and a gratifying experience for investigators whose daily work is generally pretty dull and dreary. These technical wonders were unthinkable at the end of my residency in 1950. The reigning doctrine at that time was that dermatology was a choice field of study because the skin was so accessible to the eye and to the finger. Everything was laid out before your eyes, unlike for internists and surgeons, who needed x-rays to see what was going on inside. I think that this entrenched belief in the diagnostic powers of touching and looking at the skin kept dermatology as a backwater, purely descriptive specialty for more than a hundred years. Dermatologists were the butt of many scornful jokes and dermatology was derided as a skin game. Medical graduates who chose dermatology as a career came from the bottom of the class.
In my essay on the invisible dermatoses, written 25 years ago, I opined that the most important early events in the pathogenesis of skin disease were largely invisible, hidden beneath the surface.1 Visible signs were a late stage in the disease process, obscuring what had gone on before, useful for classification but providing no insights regarding pathogenesis. Moreover, dermatologists were mistaken when they thought, as many still do, that clinical clearing of lesions is a reliable means to declare therapeutic success. The fact is that in chronic diseases such as psoriasis, the visibly cleared site remains abnormal for many months, evidenced by histologic study. Besides, the subjective estimate of say 50 to 75% clearing brings little satisfaction to patients who want to be free of lesions. Subjective estimates of improvement still plague us to this very day. The beauty of bioengineering and imaging techniques is that clinical changes are measurable and quantitative, based on objective techniques. Stopping treatment upon clearing is a sure invitation to recurrence, which in psoriasis occurs precisely at the “cleared” site. It is biological nonsense to think that the skin is a transparent window through which the pawing and peering dermatologist could ascertain the underlying changes. The professorate of my day loved the window imagery. I prophesied that the golden age of dermatology would begin when a totally blind student would not be barred from applying for a residency in dermatology.2 I now allow myself the conceit of foreseeing that new instruments would be 3
4
Handbook of Non-Invasive Methods and the Skin, Second Edition
developed that would far surpass the human eye for diagnosing and treating disease, as well as providing an increased understanding of how diseases evolved and resolved. The blind student can now “see” what is really important. I argued in my invisible dermatology thesis that occult, subclinical events were far more important in diagnosing and understanding disease than what was revealed to the naked eye. I gave examples of different disorders that illustrated how much dermatologists were missing by depending so much on the eye. To be sure, some masters in the early 20th century had an inkling of the invisible dermatology concept. For example, Gougerot observed that the normal-appearing skin of patients with leprosy showed granulomatous lesions histologically. I emphasized that the skin had a long memory of previous insults, for example, allergic contact dermatitis. I sensitized a volunteer to gold chloride who reacted strongly to a patch test. The inflamed site healed nicely. Then, when this person was given a gold salt orally for the treatment of rheumatoid arthritis, the patch test site flared up brilliantly. Another memorable experience happened to a colleague to whom we mistakenly gave 8 MEDs to a 1-cm circle. Now, every time she takes a hot bath, the site turns briefly red. The last flare was 11 years after the irradiation. The skin does not forget. Imaging techniques in both instances reveal that the cleared sites are not normal. The enterprise of developing novel, ingenious, noninvasive measuring and imaging devices has been a remarkable success story. The number and variety of these have been so great that Ronnie Marks has remarked that “one could well argue that it is time for a pause. We are confronted with a plethora of instrumentation which is simply overwhelming.” I take this to be the weary utterance of an aging dermatologist, who he himself has described as GOD (grumpy old dermatologist), struggling to stay afloat in the rushing stream of new inventions that stagger the mind. These are as thrilling as they are overwhelming. The creations of the bioengineering community are coming so fast that the instrument one buys today will be obsolete in 5 years. Why should we pause when each month seems to provide us with a novel technique to measure functional and structural changes that we could not have dreamed about 10 years ago. In his 1989 text, Lévêque was impressed by the availability of non-invasive devices, which he estimated to be about 20.3 In my introduction to the first edition of this volume I estimated that the inventory of new instruments was about 50. The number is now approaching 300 (Jørgen Serup, personal communication). Many of these are expensive, still experimental, and not commercially available. Most are highly specialized and are useful only to investigators with a specific focus, for example, in the diagnosis and treatment of malignancies. Still, there is no question that their power to provide important new images
and accurate measurements will compel investigators in all areas of the broad domain of skin to get on board this train, which is moving so swiftly to the future. It is now 10 years since the first publication of this classic text, Handbook of Non-Invasive Methods and the Skin. This new edition may be viewed as a celebration of the enormous advances that have taken place in one decade. A mere glance at the impressive list of authors of this text, comprising all the major players in the field, reflects the progress that has been made, for which the term fabulous is not an exaggeration.
1.1 SOME CURRENT ISSUES Lest one become inebriated by our dizzying success story, I bring attention briefly to some issues that are relevant to our growth and future status. Bioengineering has still not made a great impact on the practice of dermatology. At least in the U.S., very few departments possess more than one bioengineering instrument, and most not even that. This is regrettable since these non-invasive techniques have great potential for enhancing teaching, research, and patient care. We have still not penetrated the groves of academia where the current emphasis is on molecular biology, genomics, immunology, proteonomics, etc. These are lofty, important, fashionable subjects that are the favorites of national funding and granting agencies. No one denies that these are cutting-edge areas that deserve support. On the other hand, imaging and bioengineering fall into the same category. To my knowledge, the National Institutes of Health has not funded any study centered on this fast-growing field, which has enormous diagnostic potential. This is extremely shortsighted. It is regrettable that there has been virtually no effort to demonstrate to chairmen of researchoriented departments of dermatology the usefulness of bioengineering techniques for the practice of what is now called evidence-based medicine. We need to publish in more mainstream clinical journals and to find more places on national and international meetings of dermatologic societies. At recent meetings of the American Academy of Dermatology and the European Academy of Dermatology and Venereology, where at each venue about a thousand posters were exhibited, I found no more than two at either meeting that had bioengineering as a central theme, although some used such techniques in various clinical and experimental studies. We are not a service industry but a scientific enterprise that needs standing in its own right. We need to be less parochial as a society and to spend more time in educating the professors who mentor the young. At the same time, we must praise industry and its subsidiaries for investing in and developing the sophisticated technologies now in our hands. The skin care
Personal Perspectives on Bioengineering and the Skin: The Successful Past and the Brilliant Future
industry has shown both vision and financial courage in putting so much money and effort into developing new non-invasive models. It is also appropriate to pay tribute to the confederacy of engineers, material scientists, physicists, chemists, even mathematicians, that have come together to produce instruments that are marvels of ingenuity and precision. Who else but industry, often viewed as basely commercial, would have created such powerful tools as fluorescent confocal microscopy, optothermal coherent tomography, fringe projection imaging, capacitive pixelsensing technology, magnetic resonance imaging, ultrasonography, and many others. The latest inventions increasingly deal with dynamic physiological functions in addition to structure, for example, tracking in real time the complex events that take place during the inflammatory process.5,6 Non-invasive technology has the prospect of fulfilling the age-old dream of preventing rather than treating diseases. What we shall be able to do is constrained only by our imagination, not by technical limitations. The time is not far off when we will be tracking and visualizing structures at the molecular level, also establishing stratum corneum gradients of water and elements such as calcium, magnesium, and sodium. An aging population is now seeking help from cosmetic surgeons who are using a variety of antiaging surgical procedures, which include myriads of lasers, chemical peels, light therapies, microdermabrasion, etc., which can improve the appearance and quality of photoaged skin. Cosmetic dermatologists are beseiged by manufacturers of a huge array of these devices, whose efficacy is hawked to practitioners by intense advertising and marketing, with scanty science to back up the therapeutic benefits. The only way to differentiate among these procedures is by the use of noninvasive techniques. One will have to look hard to find how these devices compare in efficacy to each other for a given indication. Some cautions are in order since every good thing can be subverted toward unworthy and ignoble ends. A case in point is the burgeoning of hundreds of antiwrinkle topical creams and formulations now flooding the marketplace in a multi-billion dollar industry. The marketers of such products often exploit non-invasive technology to substantiate “scientifically” claims of efficacy, taking advantage of the fact that cosmetic products that promise to restore a youthful appearance are not regulated by the Food and Drug Agency, providing ample room for hype, gross exaggeration, and preposterous claims, served up to credulous consumers who are unable to distinguish fact from fiction. Physicians are often in the same predicament. For the unscrupulous manufacturer, non-invasive methodologies can be mobilized that can parade before the profession a host of “objective” data to support claims of efficacy. It is an easy matter to provide the “right” answers required by the sponsor by selecting the “right”
5
equipment. It is well to keep in mind that the bedrock reputation of bioengineering is based upon its credo of objectivity of providing unbiased information that is credible and truthful. Unfortunately, there are rampant exceptions to this state of detachment in the highly competitive antiaging marketplace. Anyone can select a number of instruments that seemingly, but falsely, support claims. It is an indispensable requirement to know what is being measured and whether it is truly relevant to the intended purpose. Most measurements are one-dimensional and focus on only one factor of a complex phenomenon in which many coexisting factors are interacting. One needs to know what the measurement actually means in order to interpret whether the result is good or bad. Not everything that is measurable is worth measuring. Many measurements, backed up by formidable statistics, are simply meaningless. One of my widely quoted maxims applies here: “A fool with a tool is still a fool.” One might add that a rogue with a probe is still a rogue. So much for the misuse of bioengineering methods to promote the current rash of products, some of which are marginally effective or even harmful. I am joined by JeanLuc Lévêque, who in the introduction to his 1989 text remarked that “an abundance of apparatuses with uncertain functions and measurements whose significance is often doubtful, debauched and contradictory will not serve the science of dermatology.”3 On the positive side are the publications created by members of our society aimed at establishing international guidelines for conducting non-invasive tests (which also happen to be nonaversive and patient friendly). These serve the extremely important function of standardizing procedures so that investigators in widely separated laboratories can obtain reproducible results and come to similar conclusions. These carefully constructed documents have been invaluable in precisely defining guidelines that ensure that contradictory or controversial results are not due to technical details. We know that subtle differences can have large effects. Panels of experienced investigators have now given us explicit guidelines for measuring transepidermal water loss, laser Doppler imaging, dermoscopy, stratum corneum hydration, tristimulus colorimetry, optical profilometry, and others in an ongoing international collaboration.
1.2 PERSPECTIVES ON THE FUTURE Ample tools are now in place with regard to clinical applications concerning the classification, diagnosis, pathogenesis, and treatment of acute and chronic dermatoses. Endless possibilities jump out before our eyes if we consider what can be done with confocal microscopy, which makes it possible to optically cut through the skin in horizontal sections 1 to 3 μ thick. Cells and organelles with different reflective properties can then be identified
6
Handbook of Non-Invasive Methods and the Skin, Second Edition
from the surface down to the papillary dermis. Corneocytes, keratinocytes, melanocytes, Langerhans cells, and dermal papillae come into view.4 It is a feast for the eye to see red blood cells coursing through capillaries accompanied by emigration of leukocytes in various inflammatory conditions. With fluorescent confocal microscopy it is possible to see terminal axons coursing through the epidermis, which will make it possible to show how nerves play a central role in what is now called neurogenic inflammatory processes.5 Non-invasive imaging of skin tumors is already extremely valuable in differentiating malignant melanomas. I cite the following possibilities from my own experience.
1.3 CONTACT DERMATITIS Patch testers find it exceedingly difficult to distinguish irritant from allergic reactions, especially when the responses are mild. Although allergic reactions come under the category of delayed hypersensitivity states, we have seen early microvascular events such as swelling of vascular endothelial cells and trafficking of leukocytes into the tissue as early as 1 hour after applying contact allergens such as nickel sulfate in sensitized patients. By contrast, in irritant reactions, such as those produced by sodium lauryl sulfate, we see very little dermal change after 2 to 3 hours except for swelling of corneocytes. Here, timing is of the essence since at the end of 24 hours, irritant and allergic reactions become indistinguishable. False positive irritant patch tests are a huge problem in the diagnoses of allergic contact dermatitis.
1.4 DIAGNOSIS OF PREDISEASE Chronic dermatoses do not spring up overnight, but may be preceded by occult changes, even years before the disorder becomes clinically visible. Being able to recognize the invisible early stages makes it possible to begin effective treatment long before the signs and symptoms of disease cause overt suffering. The ancient ideal of clinicians can then be realized, namely, to prevent disease rather than to start therapy after it has become clinically visible, by which time it may have caused irreversible damage. I cite two cases in which the strategy of prevention has succeeded. In high-risk young persons whose family history shows relatives with classic rosacea, but who show no signs of a red face, videomicroscopy with polarized light can reveal a network of invisible telangiectatic vessels, as early as age 10. Mild topical treatments starting then will likely greatly retard the later emergence of clinical rosacea, with its distressing psychosocial consequences.
In the case of high-risk prepubertal youngsters whose parents have experienced scarring acne in adolescence, microcomedones, the precursor of all later manifestations, can be demonstrated as early as age 8 in females using optical coherent tomography. Optical coherent tomography furnishes transverse images perpendicular to the surface, analogous to traditional histologic sectioning. One sees multiple follicles distended with horny material (microcomedones), completely invisible to the naked eye. Topical retinoids are very effective in eliminating these horn-filled follicles, preventing progression to the visible manifestations of acne, namely, comedones and papulo-pustules. Clinicians in all specialties are recognizing the great benefits of diagnosing predisease so that prevention can finally replace treatment; the latter is the paradigm physicians have followed for centuries.
1.5 PHOTOAGING By taking biopsies of the face throughout the entire life span of white persons, I showed four decades ago that there were striking alterations of the dermal matrix in young subjects, ages 10 to 15, with lovely, unblemished complexions. They all showed a moderate amount of branched, thickened, curled masses of abnormal elastic fibers (elastosis). Again, topical retinoids can reverse these recondite changes and, along with other protective measures against exposure to solar radiation, prevent the inevitable, and detested, dreadful signs of photoaging. It now turns out that we can estimate the degree of elastosis by using high-intensity blue light, in the 400- to 500-nm spectral range. Abnormally thickened and increased lastic fibers exhibit fluorescence, the amount of which can be captured and graded by appropriate imaging. We can then educate young people to take proper precautions against the destructive effects of heedless exposure to mid-day sunlight. Showing patients the difference between their chronological age and their photoage has greater impact on protective behaviors than warning against actinically induced cancers, such as squamous cell cancer and malignant melanoma.
1.6 EPILOGUE We shall know that non-invasive methodologies have finally come of age when volumes like this one can be found on the shelves of dermatologic libraries everywhere, along with the classic texts that are required reading for informed practitioners. My only criticism of this voluminous text, edited by the indefatiguable Jørgen Serup, is that it may contain much more information than you care to know about.
Personal Perspectives on Bioengineering and the Skin: The Successful Past and the Brilliant Future
The future of non-invasive technology is not only bright but positively radiant.
REFERENCES 1. AM Kligman. The invisible dermatoses. Arch Dermatol 127:1375, 1991. 2. AM Kligman. Blind man dermatology. J Soc Cosm Chem 17:505, 1966. 3. JL Lévêque. Cutaneous Investigations in Health and Disease. Marcel Dekker, New York, 1989.
7
4. M Rajadhyaksha, S Gonzalez, JM Zavislan, RR Anderson, and RH Webb. In vivo confocal scanning laser microscopy of human skin. II. Advances in instrumentation and comparison with histology. J Invest Dermatol 113:293, 1999. 5. LD Swindle, SG Thomas, M Freeman, and PN Delancy. View of normal human skin in vivo as observed using fluorescent fiber-optic confocal microscopic imaging. J Invest Dermatol 121:706, 2003. 6. E Ruocco, F Arganziano, G Pellacani, and I Seidenari. Non-invasive imaging of skin tumors. Am Soc Dermatol Surg 30:301, 2004.
to Choose and Use Non-Invasive 2 How Methods Jørgen Serup Department of Dermatology, Linköping University, Linköping, Sweden Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 2.1 Introduction................................................................................................................................................................9 2.2 Choice of Method and Instrument ............................................................................................................................9 2.3 The Legal or Authoritative Reference behind Study by Non-Invasive Methods ...................................................11 References .........................................................................................................................................................................12
2.1 INTRODUCTION Professor Albert Kligman is cited for the provocative statement “A fool with a tool is still a fool.” The complete tool in a study is much more than the instrument used for the measurement or scanning. It is also the study design, including the purpose and the declared endpoint, the sample, the statistical method, the legal reference, the guideline or operation procedure, the competence of the researcher and his allied, the ethical aspect and the resource, capacity and economy. It is not easy to be up to ideal standard in every aspect. It may actually be very difficult to avoid foolishness in research, and foolishness might even turn out to be innovative. Innovation and progress can be strangled by rigid research principles and burocracy. Use of methods can be so critical that even potential Nobel Prize winner projects are not started or brought to a conclusion. Projects range from the very first observation, which may be a result of serendipity, to the final confirmation and proof conducted under the strict requirements of an authoritative body. Nevertheless, academic use of noninvasive methods is to master precision and handle the signal-to-noise ratio in a sound, honest, and balanced way, and the need for precise and accurate measurement is paramount and not dependent on the project type and state, as Professor Kligman’s statement implicates. A fool is a fool, and we must be well prepared and careful with the methods and the projects.
2.2 CHOICE OF METHOD AND INSTRUMENT There is always an idea, a vision, or something broader behind the penciled endpoint in a protocol. The very first
question is simply to make a move in the right direction and enter the right door. We have very little tradition for qualitative research methods to outline the contour of a problem and find special facets or domains of major importance worthy for a detailed study by quantitative research methods, such as the non-invasive methods. Typically, we start with a somewhat occasional discussion among colleagues and sponsors, a discussion kept quite open to subjective preferences. However, we learned from evidence-based medicine that such personal opinion has variable value. Discussion about project perspective, idea, study endpoint, and choice of method can be improved and made more professional if some of the methods or principles developed in qualitative research are adopted. Especially, consumer-oriented studies can benefit from a prestudy assessment finding the right type of point to measure. Methods such as focus group interviews are established for such a purpose.1 The non-invasive methods are by nature quantitative and narrow. We follow the ground principle of natural sciences expressed by the French physiologist Claude Bernard, to measure and to make objects measurable. Today, we are digitalizing vigorously. The methods are narrow because on the sensor side, they are based on a single physical modality such as sound, light or ray, electricity, etc. Signal processing is with modern electronics highly complex, and every component of the process has its own linearity, cutoff, and upper limit. The resultant measured value is in comparison with the in vivo biological object, an artifact that only becomes valid through validation study, comparisons with reference methods, calibrations, etc. Computerized imaging is quantitative and involves filtering, manipulation, and measurement of small picture elements or pixels, but the biological sample imaged consists of surfaces, tissue, and cells not built up of 9
10
Handbook of Non-Invasive Methods and the Skin, Second Edition
digital squares, each of intensity 0 to 255. The image perceived with the naked eyes is a very complex and nonlinear comprehension, often purely qualitative and often individual or personal. Choice of method in a study to verify a hypothesis is a critical decision where the character of the endpoint and the usefulness and validation of the method are of special importance. In applied medicine and in product substantiation, the real endpoint may be identical to the protocol endpoint or broader in nature. It may be feeling of improvement of disease, impression of younger skin, or some other feature or change noticed with the eyes, felt with the finger, or otherwise perceived by a person. The ideal validation of a non-invasive method is only achieved if perception by lay individuals or special observers is established as the gold reference. In pure research, validation and reference to standard are, of course, much simpler, as exemplified by measurement of size and position of an object in the dermis by high-frequency ultrasound. The following 30-point prestudy checklist is proposed for systematic use when a study is initiated (for definitions regarding instrument performance and validation, see Table 2.1): TABLE 2.1 Instrument Validation, Terms and Definitions Accuracy: Degree of similarity between the value that is accepted either as a conventional true value (in-house or local standard) or as an accepted reference value (international standard) and the mean value found by performing the test procedure a certain number of times. Provides an indication of systemic error. Precision: Degree of similarity (degree of scatter) among a series of measurements obtained from multiple sampling of the same homogenous sample under prescribed conditions, expressed as repeatability and reproducibility. Repeatability: Expresses the situation under the same conditions, that is, same operator, same apparatus, short time interval, identical samples. Reproducibility: Expresses the situation under different conditions, that is, different laboratories, different samples, different operators, different days, different instruments from different manufacturers. Range: The interval between the upper and lower levels for which the procedure has been demonstrated as applicable with precision, accuracy, and linearity. Linearity: Ability of the procedure (within a given range) to obtain test results directly proportional to true values. Sensitivity: Capacity of the procedure to record small variations or differences within the defined range. Limit of detection: Lowest change above zero that is detectable. Limit of quantification: Lowest change above zero that can be quantitatively determined (not only detected) with defined precision and accuracy under the stated experimental conditions. Ruggedness: Evaluates the effects of small changes in the test procedure on measuring performance.
1. What is the study idea and what is the precise study endpoint supporting the idea? 2. Is the study endpoint truly quantitative in nature, narrow enough for specific study, and truly suited to support the idea? 3. Stratification of endpoints into primary, secondary, tertiary, etc.? 4. Shall one or more instruments be applied (monoinstrumental or multi-instrumental design)? 5. Function of the measurements and the instrument in the study: support, description, exclusion, comparison, validation during study, etc.? 6. Which structure or function is actually being measured? 7. Range, linearity, and expected change of variables during study? 8. When should measurements be performed? 9. Interperson, intraperson, and intralesion variation, and if possible, variability data from normal and healthy skin? 10. Influence of gender, age, and race? 11. Statistical evaluation of the design and the size of the sample studied? 12. Studies and literature validating the instruments applied? 13. If the target or measured area is small, do measurements need be repeated to overcome local site variation? 14. Existing in-house standards or recommendations, standard operating procedure (SOP)? 15. Guidelines and legal requirements, including ethical aspects? 16. Output from the instrument and source data — can these be handled and stored safely? 17. Is the laboratory room and are the laboratory conditions up to good standard, or are improvements needed? 18. Is a backup situation prepared if unexpected breakdown occurs of the device, laboratory, etc.? 19. Are ambient conditions such as temperature and humidity under control and expected to remain constant during the study period? 20. Needs for preconditioning of study subjects? 21. Are technicians well educated, trained, and well prepared for the specific study, and who is responsible for what? 22. Are various types of bias identified and, if possible, eliminated? 23. How is it ensured or monitored that the study develops as planned, and what are the requirements for constancy and the consequences of inconstancy?
How to Choose and Use Non-Invasive Methods
24. Calibration, maintenance, and control of instruments before, during, and at end of study? 25. Events and circumstances that exclude measurements from being performed or invalidate results? 26. How to conclude and report the study? 27. Timetable for the study — is it realistic and satisfactory? 28. Resources involved — are they available from start to end? 29. Is the study at an academic level where conclusion and interpretation are independent of economic interest, even if the study is conducted in an industrial environment? 30. Is the study documented and prepared for a special situation if some accusation about fraud would come up? The typical pitfalls are:
11
2.3 THE LEGAL OR AUTHORITATIVE REFERENCE BEHIND STUDY BY NONINVASIVE METHODS There are numerous references, some strictly technical, others more general (Figure 2.1). References may be institutional, national, regional, or global. References undergo constant change and update. It is up to the researchers to ensure that their study is in accordance with the relevant references at a given time. Good laboratory practices (GLPs) are standards mainly used in the pharmaceutical industry for preclinical safety testing in vitro and for experimental study of animals. Good clinical practices (GCPs) are standards used in the pharmaceutical industry for the good conduct of clinical trials from phase I to IV. The International Conference on Harmonization (ICH) GCP guideline (ICH-GCP) in 1996 introduced a tripartite GCP system recognized in the U.S., Europe, and Japan. GCP is particularly relevant for non-invasive measurement in humans and is covered in Chapter 7. However, GCP in a milder version was in 2001
1. Strategic error: poor design and study plan 2. Technological error: poor choice of variable and method 3. Technical error: measuring device not functioning, inconstant, or unstable 4. Performance error: imprecise or not competent use of the device 5. Inadequate measuring conditions: laboratory facilities not acceptable 6. Object-related error: poor selection and preconditioning of study subjects 7. Data error: wrong data acquisition, handling, and storing 8. Malconclusion: conclusion not really supported by data, or inconvenient conclusion suppressed 9. Explanatory mistake: inferior reporting and publication policy The soundness of research is a dimension in itself, difficult to describe and variable from project to project. Clarity and simplicity are at least two important technical elements. Honesty and goodwill are two other elements at another level. Wisdom and knowledge are not identical. A sound project has wisdom built in and, of course, a solid base of knowledge. Maybe soundness is illustrated best with the words of the English author Rudyard Kipling: “I keep six honest serving men. They taught me all I know. Their names are WHAT and WHY and WHEN and HOW and WHERE and WHO.”
FIGURE 2.1 Searching the literature. Cartoon by Storm P, 1944, Copenhagen, Denmark. (Published with permission of the Storm P Museum, Jens Bing, Copenhagen.)
12
Handbook of Non-Invasive Methods and the Skin, Second Edition
according to a European directive implemented in public institutions and hospitals for any clinical trial on patients. The strong emphasis on regulation of clinical studies for the last decades and the increased need for resource and economy to conduct studies has reduced the medical profession’s involvement and activity in clinical research initiated and conducted by the clinicians themselves, independent of sponsorship. Serendipity, as opener of major achievement and based on clinical observation, may soon belong to history. The development of standards and requirements has moved development away from innovation and in the direction of strictly scheduled confirmative research. Fortunately, many portable, accurate, easy-tooperate non-invasive devices were developed and commercialized, and the bioengineer is relatively privileged in comparison with other fields of clinical research. The Declaration of Helsinki of 1964 is the classical global standard for clinical study, with emphasis on ethical conduct of study in humans. This declaration introduced rules for information of study subjects, the informed consent. It became adopted in many countries and legally implemented. The original version and recent updates are found on the website of the World Medical Association (WMA). The device itself shall of course do no harm to the person studied. In the U.S., the Food and Drug Administration (FDA) has formulated rules for safety classification and marking of medicotechnical instruments. In Europe, the CE marking is a similar procedure. It is as a rule necessary that authorized technicians make a safety check of research instruments resulting in approval before the instrument is applied. Regarding instrument-related guidelines and standards, the manufacturer’s manual is a core document practically and legally. With or without reference to GLPs or GCPs, it is common that laboratories develop their own in-house standard operating procedure (SOP) or instruction.2 The Skin Research and Technology journal (Blackwell Publishing; blackwellpublishing.com), official journal of the societies active in the field, is fully indexed and the leading source of information about original academic publication on new non-invasive methods, their validation, and application. The active societies are the International Society for Bioengineering and the Skin (ISBS), the International Society for Skin Imaging (ISSI), and the International Society for Digital Imaging of Skin (ISDIS). These societies organize annual or biannual congresses in different parts of the world and work well together. The Standardization Group of the European Society for Contact Dermatitis in the journal Contact Dermatitis published a number of instrumental guidelines, i.e., on measurement of transepidermal water loss (TEWL),3 on laser Doppler measurement of blood flow,4 on measurement of skin color,5 on laser Doppler scanning of
FIGURE 2.2 Selection of monographs on non-invasive methods and the skin, including the first edition of this handbook, 1995, published by CRC Press, Boca Raton, FL. The method-specific monographs are referenced in the literature list (9–13).
cutaneous blood flow,5 and on assessment of irritant skin reactions elicited by the experimental standard irritant sodium lauryl sulfate (SLS).6 Authors of some of these standardization papers have contributed with chapters to this handbook of non-invasive methods. The European Group for Efficacy Measurements on Cosmetics and Other Topical Products (EEMCO) has produced a number of guidance papers and introductory reviews on use of non-invasive methods for efficacy documentation of cosmetics. These are found on the website under PubMed/EEMCO (see Chapter 3). An EEMCO guidance on standard scoring schemes for assessment of dry skin is not included in this index.8 Regarding evaluation of cosmetics, the European Cosmetic Toiletry and Perfumery Association, named COLIPA, published various recommendations of this organization on behalf of the industry. The COLIPA published a widely used standard on determination of sun protection factor (SPF) of sun blockers parallel to FDA and German industrial standards (DIN). The information landscape in the field of non-invasive methods and the skin thus includes very different sources of information, some very detailed, some broad and educational, some independent, and some partial. The method-specific books9–13 published by CRC Press (Figure 2.2), the present and updated second edition of The Handbook of Non-Invasive Methods, and the French handbook Physiologie de la peau et explorations fonctionelles cutanées,14 edited by Professor Pierre Agache in 2000, are all instruments for overview and introduction.
REFERENCES 1. Hill CE, Thompson BJ, Williams EN. A guide to conducting consensual qualitative research. The Counselling Psychologist 25:517–572, 1997.
How to Choose and Use Non-Invasive Methods
2. Serup J. Bioengineering and the skin: standardization. Clinics in Dermatology 13:293–297, 1995. 3. Pinnagoda J, Tupker RA, Agner T, Serup J. Guidelines for transepidermal water loss (TEWL) measurement. Contact Dermatitis 22:164–178, 1990. 4. Bircher A, de Boer EM, Agner T, Serup J. Guidelines for measurement of cutaneous blood flow by laser Doppler flowmetry. Contact Dermatitis 30:65–72, 1994. 5. Fullerton A, Fisher T, Lahti A, Wilhelm K-P, Takiwaki H, Serup J. Guidelines for measurement of skin colour and erythema. Contact Dermatitis 35:1–10, 1996. 6. Fullerton A, Stücker M, Wilhelm K-P, Wårdell K, Andersson C, Fischer T, Nilsson GE, Serup J. Guidelines for visualization of cutaneous blood flow by laser Doppler perfusion imaging. Contact Dermatitis 46:129–140, 2002. 7. Tupker RA, Willis C, Berardesca E, Lee CH, Fartasch M, Agner T, Serup J. Guidelines on sodium lauryl sulphate (SLS) exposure tests. Contact Dermatitis 37:78–81, 1997.
13
8. Serup J. EEMCO guidance for assessment of dry skin (xerosis) and ichthyosis. Skin Research Technology 1:109–114, 1995. 9. Berardesca E, Elsner P, Wilhelm K-P, Maibach HI, Eds. Bioengineering of the Skin: Methods and Instrumentation. CRC Press, Boca Raton, FL, 1995. 10. Berardesca E, Elsner P, Wilhelm K-P, Maibach HI, Eds. Bioengineering of the Skin: Cutaneous Blood Flow and Erythema. CRC Press, Boca Raton, FL, 1995. 11. Berardesca E, Elsner P, Wilhelm K-P, Maibach HI, Eds. Bioengineering of the Skin: Water and the Stratum Corneum. CRC Press, Boca Raton, FL, 1994. 12. Berardesca E, Elsner P, Wilhelm K-P, Maibach HI, Eds. Bioengineering of the Skin: Skin Surface Imaging and Analysis. CRC Press, Boca Raton, FL, 1997. 13. Berardesca E, Elsner P, Wilhelm K-P, Maibach HI, Eds. Bioengineering of the Skin: Skin Biomechanics. CRC Press, Boca Raton, FL, 2002. 14. Agache P, Ed. Physiologie de la peau et explorations fonctionelles cutanées. Editions Médicales Internationales, Cachan Cedex, France, 2000.
Practical Guide to Resources on the 3 AInternet for the Skin Researcher Elizabeth Grove Wickersheim and Gary Lee Grove cyberDERM, Inc., Broomall, Pennsylvania
CONTENTS 3.1 3.2 3.3 3.4 3.5 3.6 3.7
The Internet and Its History ....................................................................................................................................15 Using URLs to Search for Specific Web Pages......................................................................................................15 Using Selected Subject Directories to Search for Related Web Pages ..................................................................16 Online Literature Searches ......................................................................................................................................19 Online Tutorials and Educational Courses..............................................................................................................20 Search Engines ........................................................................................................................................................21 Closing Remarks......................................................................................................................................................24
3.1 THE INTERNET AND ITS HISTORY The Internet is an international network of computers that are linked up through various types of connections to exchange information. Through an Internet service provider, one can access the “Net” and a variety of related services, such as e-mail, web pages, online commerce, etc. Once you are connected, your computer can “talk” to any other computer anywhere in the world that is also connected. There are certain hardware and software requirements that need to be carefully considered, but these are beyond the scope of this chapter. For sure, a high-speed connection is desirable and a high-power processor will greatly facilitate dealing with images. The Internet is often thought to be of recent vintage, but it actually dates back to the Cold War in the 1960s, when the U.S. Department of Defense in response to Sputnik formed the Advanced Research Projects Agency (ARPA) as the world’s first decentralized computer network. In the next few years, other government agencies and various universities, such as UCLA, MIT, Stanford, and Harvard, flocked to join the Net. By the early 1970s the network had crossed the Atlantic to include English and Norwegian universities. The 1970s also saw the introduction of electronic mail, File Transfer Protocols, Telnet, and the Usenet newsgroup. The 1980s brought further improvements with one of the more important being Transmission Control Protocol/Internet Protocol (TCP/IP), which enables computers to talk the Net’s language. Thanks to intense media coverage from the 1990s onward, the Internet has become so well known that even
preschoolers can have an Internet savvy that will amaze their grandparents. The most successful part of the Internet is the World Wide Web (also known as WWW, W3, or simply the Web), a concept originally developed at CERN Laboratories in Geneva by particle physicists so they could share information throughout Europe. Over the years it has grown into a user-friendly point-and-click way of navigating through what now should be considered the world’s biggest library. The Web can be best defined as a distributed multimedia hypermedia system. Distributed refers to the various locations that one might find information among the computer systems around the world, multimedia means that the information can include text, graphics, sound clips, and video, while hypermedia means that it is possible to move to new information through highlighted phrases or images by clicking the mouse. There are many excellent resources that can be found on the Web. These include online journals, databases, textbooks, atlases, instrumental specifications, department and societal pages, etc. Since the content of the Web is constantly changing and expanding, this chapter will primarily deal with how to conduct a search of the Web rather than just presenting lists of useful web pages.
3.2 USING URLS TO SEARCH FOR SPECIFIC WEB PAGES What is a URL? A URL is a uniform resource locator, or in other words, the address for a specific web page. Think of it as a networked extension of the standard filename 15
16
Handbook of Non-Invasive Methods and the Skin, Second Edition
concept: not only can you point to a file in a directory, but now that file and that directory can exist on any machine on the network, can be served via any of several different methods, and might not even be something as simple as a file. To reach that page, you only need to carefully enter its URL into your browser’s location or address bar, which is usually located directly under the menu. When I begin searching for a specific organization’s website, I often start by trying to guess the central URL for that organization. With Netscape and Internet Explorer, leave off http://. Begin with the common WWW, add the name or acronym of the organization, and end with the appropriate domain ending. Some of the more common domain endings include: com for commercial edu for educational institutions org for other organizations gov for U.S. federal government mil for U.S. military net for Internet service providers and networks Sometimes this brings up the organization you want, other times it does not. For example, any skin researcher should want to visit the web page of the International Society for Bioengineering and the Skin. At the time this chapter was written, if you input www.ISBS.com into your browser’s location or address bar, you will find yourself on the web page for International Specialized Book Services. If you use the org domain ending or www.ISBS.org, you will find yourself on the web page for the International Society of Biomechanics in Sports. It is only by using www.I-S-B-S.org that you will be brought to the home page of the International Society for Bioengineering and the Skin (Figure 3.1). You can bookmark it and other frequently visited websites so that you will not have any problems in the future recalling their unique URLs. Another web page that any serious skin researcher should have bookmarked is that of the International Society for Skin Imaging (ISSI). This is another case in which just guessing the URL as www.ISSI.com does not work, as this is the home page of Integrated Silicon Solution, Inc. Instead you should use www.ISSI.de, where the domain ending refers to a national domain, which in this case is Deutschland, or Germany. This home page is shown in Figure 3.2. On both the ISBS and ISSI home pages you will find links not only to other pages within the site being hosted by that organization, but also to related pages at other sites strewn throughout the Web. The language of the Internet is Hypertext Markup Language (HTML), which allows the possibility of correlating a phrase or an image with a specific URL. By clicking on such links, one arrives at another site, which might link to even more related sites.
For example, if one goes to the home page for ISBS and clicks on JOURNAL within the left-side menu, you will be taken instantly to another page (Figure 3.3), which is still within that site. If one then clicks on the phrase Skin Research and Technology found in the text on the righthand side of the page, you will be transported to a page (Figure 3.4) that is outside the ISBS site and has the URL http://www.blackwellpublishing.com/journal.asp?ref= 0909-752X. Fortunately, you do not have to remember this URL, but only take advantage of the HTML textual or graphical links. If you are a member of either the ISBS or the ISSI, then you can access this journal online by logging on with your membership number and password. You may also check to see if this journal is available to y o u a s p a r t o f y o u r l o c a l l i b r a r y ’s o n l i n e subscriptions.
3.3 USING SELECTED SUBJECT DIRECTORIES TO SEARCH FOR RELATED WEB PAGES If the reader returns to the ISSI home page, we can introduce another simple search strategy, namely, selected subject directories. These are human-constructed lists that provide various links that are arranged and classified in a meaningful way. If one clicks on WWW LINKS on the lefthand side of the ISSI home page, you will find such a list of links that point to sites sorted by societies, institutions, events, journals dealing with skin imaging, medical news, and tools. Unless the webmaster is extremely zealous, some of these links may no longer work. When this occurs, try chopping off parts of the URL starting on the right-hand side and stopping at every slash (/). This essentially takes you back through the hierarchy of the site and may help you find related information at a new URL. Most webmasters appreciate learning about their “dead” links, and a good citizen of the World Wide Web should report them. As you travel over the Web you will invariably find many very useful sites that serve as gateways or portals because of their extensive selected subject directories or databases. In some cases a fee is charged, but fortunately, others are sponsored by pharmaceutical companies who offer free access to these databases. One of the better ones is www.DermNet.com (Figure 3.5), which is sponsored through an unrestricted educational grant from Dermik Laboratories. It is a great site to visit for several reasons. For one, it provides an extensive library of digital images of various skin diseases that is updated weekly and truly deserving of its moniker “The Dermatologists’ Image Resource.” It also contains an extensive LINKS page that points to most of the journals, organizations, medical departments, and educational materials that would be of interest to a dermatologist. Unfortunately, there is very
A Practical Guide to Resources on the Internet for the Skin Researcher
17
FIGURE 3.1 Home page of the International Society for Bioengineering and the Skin, with its URL being www.I-S-B-S.org.
FIGURE 3.2 Home page of the International Society for Skin Imaging, URL www.ISSI.de, as an example of a national domain, namely, Germany/Deutschland.
little about bioengineering methods for studying skin structure and function non-invasively, but this is a great place to start learning more about the pathological processes and underlying diseases that may be worth studying with our instrumental approach.
Another great site is www.Medscape.com. It requires users to register in order to access the site, but it does not involve any fees (Figure 3.6). You will encounter some advertisement blurbs as you navigate the site, but they are not at all intrusive.
18
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 3.3 Example of HTML link based on a printed phrase. Clicking on JOURNAL will transfer you to the web page for the Skin Research and Technology journal.
FIGURE 3.4 Home page of the Skin Research and Technology journal, which can be assessed online by subscribers with the proper password.
A Practical Guide to Resources on the Internet for the Skin Researcher
19
FIGURE 3.5 Home page of DermNet.com, which is an example of a selected subject directory that provides links to websites of interest to the clinical dermatologist.
Although it covers the full breadth of all the medical specialties (physicians, pharmacists, dentists, etc.), you can fine-tune it to deal more with skin topics by selecting Dermatology as your specialty home page (Figure 3.7). By doing so, you can easily browse recently posted content of interest to dermatologists, such as Food and Drug Administration (FDA) advisory meetings’ notes and conference coverage of all the major dermatological meetings, e.g., the AAD and the Society of Pediatric Dermatology, which are archived back to 2000. It also permits you to read selected articles from a limited number of journals, including the British Journal of Dermatology, Dermatology Online, SKINmed, and Wounds. Again, these articles are archived back for several years and easily accessed through a simple click of your mouse.
3.4 ONLINE LITERATURE SEARCHES Medscape.com also allows one to do a free literature search using Medline, which is the most widely used medical database. However, I find doing literature searches using PubMed, which is offered as a free service of the National Library of Medicine, often works better for me because it is a relational database that allows one to easily find other articles that are directly related to a specific abstract. It can be accessed at http://www.ncbi. nlm.nih.gov/entrez/query.fcgi (Figure 3.8). Time does not
permit a full discussion of how best to do these types of searches, but fortunately the reference librarians at the University of Florida Health Science Center Libraries have provided an excellent online tutorial on PubMed, which can be viewed at http://www.library.health.ufl.edu/ PubMed/PubMed2/ (Figure 3.9). It is extremely well done, and you will be able to easily learn how to use and navigate the many search features of PubMed. You can either go through the entire tutorial from top to bottom or choose from an index for specific topics. I have found their explanation of how to use Boolean operators, truncations, and phrase terms to be very useful for search engines in general. Also, PubMed has some excellent help pages of its own that can be found by clicking the Help and Tutorial links on the PubMed home page. Since the “hits” do differ, you should use both Medscape and PubMed to do a complete literature search. The returns can be both awesome and intimidating. For example, if you input “TEWL,” you will have more than 450 returns with PubMed, while the Medscape list is limited to 200 by design. With Medline, the list is in order of how relevant that paper is to the search criteria. With PubMed, they are displayed in reverse entrez date order: last in, first out. The entrez date is the date that a record was initially added to PubMed and should not be confused with the publication date, which is the date an article was published. Thus, it is not surprising that there is very little
20
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 3.6 Home page of Medscape.com, which allows one to do a free literature search using Medline.
overlap between the two lists, and this is why it is imperative that for a serious search you utilize both Medline and PubMeb.
3.5 ONLINE TUTORIALS AND EDUCATIONAL COURSES The University of Florida’s tutorial on PubMed is just one example of how one can easily learn over the Internet. Indeed, there is considerable interest in establishing “the worldwide classroom,” and many universities offer distance learning courses and remote tutoring via the Internet. So much so that in some cases, one can earn a degree without ever setting foot on campus. In addition to these curricular connections being sponsored by various colleges and universities, there are a number of commercially sponsored sites with high educational value. One of the best introductions to skin science is “The World of Skin Care” by Dr. John Gray, which c a n b e v i ew e d a t w w w. p g . c o m / s c i e n c e / s k i n care/Skin_tws_toc.htm. Although sponsored by P&G Skin Care Science Center, there are no commercial overtones.
Instead, what you will find is an excellent primer that provides a very clear-cut explanation of the more important aspects of skin structure and function that is 110 pages long. It is a must-read for anyone entering the field. Another very good treatise is the Medical Student Core Curriculum that appears as a graphical link on the home page for the American Academy of dermatology at www.AAD.org. This is a Web-based textbook for students interested in dermatology. It is more advanced than The World of Science and covers both basic science and clinical topics. Each chapter is up-to-date and written by experts who are thoroughly familiar with the topic. To the best of my knowledge, there are no educational courses or tutorials that deal with bioengineering and the skin directly. It is possible to locate many academic institutions, course descriptions, and faculty members with research interests in these areas, but no real educational materials per se. Results improve if we broaden our search to more basic subjects such as physics. One of the best examples is CUPOL, an acronym for the Clemson University Physics Online Laboratories, available at http:// phoenix.phys.clemson.edu/labs/cupol/. These labs include
A Practical Guide to Resources on the Internet for the Skin Researcher
21
FIGURE 3.7 Same Medscape site as in Figure 3.6, but now with Dermatology selected as the specialties home page.
six different university-level courses, which range from a survey of general physics to calculus-based waves, optics, and modern physics. All include tutorials, learning guides, and actual laboratory experiments that may be performed wherever the Internet is available. It is not designed to eliminate the traditional, hands-on laboratory experience, but rather complement it. A very interesting feature of these courses is the extensive use of video clips and input boxes, which makes you feel like you are actually having a real-time laboratory experience without really being in the lab. This site also has very well presented “Physics Lab Tutorials” under the heading of Suppliments (sic), which deal with oscilloscope basics, EXCEL spreadsheets, precision and accuracy, etc. It also has a neat stopwatch for timing the experiments and an interactive color chart for determining the value of resistors. Although there are many other examples that could be given, it is best not to do so. The Web is constantly changing and what was present at the time this was written may not exist now. Hopefully it has been upgraded or replaced by something even better. That means that it is extremely important you learn how to search the Web for the resources that you need.
3.6 SEARCH ENGINES We briefly introduced the notion of using search engines when we described doing literature searches using either Medscape or PubMed. To keep abreast of the ever-expanding Web, it is important that you learn more about search engines in general. Search engines attempt to find and index as many sites as possible that match certain search criteria. The capabilities of the various search engines vary greatly, as does the actual scope, size, and accuracy of the databases that they are exploring. Although search engines will include millions of pages in their databases, none of them come close to indexing the entire Web, much less the entire Internet. There are a number of search engines that have become very popular over the years, such as AltaVista, Ask Jeeves, MSN Search, and Yahoo. When asked to name a search engine, many mention Google. Not only was Google.com the first search engine to index more than a billion pages, but it also displays highly relevant search results faster than any other. Additionally, it has one of the largest databases, which includes many different file types, including Portable Document Format (PDF) and
22
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 3.8 Home page for PubMed website of the National Library of Medicine, which allows free literature searches of this database.
images. It is very easy to use, and I have found the relevancy of its hits to be excellent. To fully describe how a search engine uses special software robots called spiders to crawl through the Web and create indices to find all of the information on the hundreds of million web pages that exist is beyond the scope of this chapter. Again, the reader can find a very good description of search engines at http://computer.howstuffworks.com/search-engine1.htm. Indeed, the “howstuffworks” website is an excellent place to start to learn about how anything works. I especially like its coverage of electronic gadgets like digital cameras, cell phones, GPS, etc. It is an extremely good example of how well a selected subject directory can work and clearly should be bookmarked in anyone’s favorite list. Let us get back to search engines, especially Google.com (Figure 3.10). For most users, executing a search consists of simply typing in a number of descriptive terms and clicking a button. Google searches are not case
sensitive. All letters, regardless of how you enter them, are understood as lowercase. For example, searches for “Albert Kligman,” “albert kligman,” and “ALBERT KLIGMAN” all return the same results. Note that this example used a quoted string as a query, which forces an exact match of that phrase. If you just enter a series of search terms without any punctuation, you are asking Google to return all results that contained all of your search terms. There is no need to include the Boolean operator “AND,” as this is the default. Google also supports the Boolean operator “OR” to return pages that contain at least one of the search terms and the Boolean operator “NOT” (indicated by a minus sign before the term) to exclude pages that include that term. Google also has the capability to recognize certain numerical inputs as being unique search terms, with one of the more practical being the tracking number for a UPS or FedEx package. Just input the tracking code without any spaces and you
A Practical Guide to Resources on the Internet for the Skin Researcher
23
FIGURE 3.9 Online tutorial describing how best to use PubMed, prepared by the librarians at the University of Florida.
FIGURE 3.10 Basic search page for Google.com. Note that the language preference has been set to Elmer Fudd.
should be able to learn its current status in the respective system. It also recognizes and maps addresses and has a reverse phone number lookup function. In searching the Web, Google uses very sophisticated text-matching techniques to find pages that are both impor-
tant and relevant to your search in order of their Google PageRank. Google analyzes not only the candidate page, but also the pages linking into it to determine the value of the candidate page for your search. Google also prefers pages in which your query terms are coded as headings.
24
Handbook of Non-Invasive Methods and the Skin, Second Edition
The Google’s results page is full of information and includes links to all of the pages that match your query in order of their relevancy according to their PageRank algorithms. Depending upon the nature of your search terms, you may also have sponsored links, which are paid advertisements that will appear at the top and along the right-hand side of the results page. These commercial pages are separate from and do not influence the PageRank of the search results. Each listing provides one or more relevant excerpts (snippets) that show how your search terms are used in context on that page. In the excerpt, your search terms are displayed in bold text so that you can quickly determine if that result is from a page you want to visit. To view a page listed in your search results, click on the page title, the first line in each result. Any of your query words that appear in the title will be in boldface, and the title will be underlined, i.e., it is a link to the web page. Note that when you position your mouse pointer on the title, the URL for the web page will appear in your browser’s status bar, at the bottom of many browsers. To get the full benefit of Google, you must learn how to really use it and fully understand what it is really capable of doing. Such an in-depth treatment is well beyond the scope of this introductory chapter. Again, we can turn to the Web for help in learning these advanced techniques. Although there is considerable information and support within the Google site itself, I found that the “Google Guide” prepared by Nancy Blachman is well worth browsing through. The home page URL is www.googleguide.com and includes tutorials for both the novice and experienced user. It is neither affiliated nor endorsed by Goggle, but the underlying philosophy is clearly very positive about Google. It should also be mentioned because of the potential for Google to dominate the Web and privacy concerns over how it issues cookies and tracks individual searches, there are several watch guard groups, such as http://www.google-watch.org/, that maintain pages that place Google in a less favorable light. From the home page for www.Google.com (Figure 3.10), one can click on About Google to access the official Help and How to Search guide for Google. From the home page you can also bring up the Advanced Search menu (Figure 3.11). This form allows one to specify required words and exact phrases to include or exclude in the query. You can also specify the language of the web pages to return, file formats, and web page domains. In addition, you can specify a date range for when the web pages were updated and ranges of numbers that appeared in the web pages. You can even specify where in the web pages the search terms must appear, such as title, URL, body text, etc. This allows you to really fine-tune your search with very little effort. One of the more fascinating aspects of Google can be found under Language Tools. As mentioned before, it is
possible to limit your search to web pages written only in a specific language or geographic area. Once these pages are found, it is possible to translate those written in either French, German, Italian, Portuguese, or Spanish into English through the Google language tool. It is also possible to set the Google home page, messages, and buttons to read in your preferred native language. The language choices are both amazing and amusing. Currently, there are more than 100, which range from Afrikaans to Zulu and include all the major languages and several nontraditional ones as well. Indeed, if you are a Trekkie and want to use Klingon as your language of preference, it can easily be done. Same goes for Pig Latin or Elmer Fudd. Indeed, if you do not see your native language, you can help the Google translators create it. The Language Tools page also contains a very nice display of the national flags of all the local domains that contain a Google site. Although I occasionally amuse myself by seeing how the I’m Feeling Lucky button on Google’s home page appears in different languages, I do not use it very often. This button, instead of showing you a list of web pages, sends you immediately to the one result that may be most relevant to your query that is not a paid advertisement. Although the I’m Feeling Lucky button can save you some time, it is really a matter of only a few seconds with a high-speed connection, and I much prefer a broader search list with its snippets to select from.
3.7 CLOSING REMARKS A few years ago, just a passing knowledge of the Internet and the World Wide Web was enough for you to be considered a guru and land you a very nice IT job. Now in many fields it has become mandatory that you know how to surf the Web, and Google has become a verb familiar to many laypersons. Indeed, one of the problems confronting today’s physicians is that their patients can easily obtain information on their disease that the physician may not be aware of. Unfortunately, there are no authorities that guarantee the truthfulness of the information that can be found on the Net. Anyone with a computer and a connection can “publish” material without editorial oversight or peer review. Although as shown by the examples cited in this chapter there are many superbly written, wellillustrated, and highly informative resources, one can just as easily encounter pages with incorrectly labeled images and patently false statements. Often these sites are touting the benefits of some dubious skin care treatment. We will not be able to free the Internet of such commercial influences, but should strive to ensure that we are providing accurate and interesting information in the web pages that we do have control over. It is the responsibility of the user to evaluate information gathered from the Internet for its accuracy and worth.
A Practical Guide to Resources on the Internet for the Skin Researcher
25
FIGURE 3.11 Advanced search page for Google.com.
Some basic questions that need to be critically considered include but certainly are not limited to:
Currency: When was the information last updated? Are the links up-to-date?
Authorship: Who is the author? Is he or she a recognized authority? What experience does he or she have? What are his or her credentials? Affiliations: Is this a personal web page or is the material posted sanctioned by an organization or institution? Does that organization have a basis or a commercial conflict of interest? Content: Are there errors in spelling and grammar? Is the presentation logical? Are the facts and quotations properly cited and referenced in a bibliography?
In conclusion, we would like to emphasize that the Internet offers a wealth of information, much of which is neither monitored nor edited. Searching the Internet offers both significant challenges and amazing benefits. If you are just starting to surf the Internet, we feel that the best advice that we can give you is to just jump in and get started. The sooner you jump in, the sooner you will benefit from the many resources and learning opportunities offered through the Internet. We also hope that by sharing our basic strategies and philosophies, we have been able to enhance your Internet experiences.
Skin Integument: Variation 4 The Relative to Sex, Age, Race, and Body Region Nadia Farinelli Department of Dermatology, University of Pavia, Pavia, Italy
Enzo Berardesca Department of Dermatology, Instituto Dermatologico di S Maria e Gallicano, Rome, Italy
CONTENTS 4.1 4.2 4.3 4.4
Introduction..............................................................................................................................................................27 Sex............................................................................................................................................................................27 Age and Body Region .............................................................................................................................................27 Race..........................................................................................................................................................................29 4.4.1 Physiological Differences ............................................................................................................................29 4.4.2 Percutaneous Absorption .............................................................................................................................29 4.4.3 Biomechanical Properties, TEWL, and Susceptibility to Irritants .............................................................29 References .........................................................................................................................................................................30
4.1 INTRODUCTION The skin is not a uniform sheath covering the body, but a specialized organ with several functions changing from site to site. These regional variations are of great importance because they can influence skin behavior and thus susceptibility to disease. The major anatomical differences related to site involve stratum corneum thickness, distribution of appendages and melanocytes, variation in the structure of the dermoepidermal junction and of the dermis, and changes in blood supply.1 Anatomical changes often induce functional changes that can be quantified with combined non-invasive techniques that allow the assessment of skin function relative to sex, age, and race.2
4.2 SEX According to the HANES survey,3 skin pathology is consistently more prevalent among males than females. Most of this higher prevalence among males is accounted for by the higher prevalence of dermatophytoses and skin tumors. This difference, however, is felt to be related to differences in behavior (hygiene and occupation) and not in skin characteristics.
There is evidence of greater skin irritability in females than males, although this difference is not reported to be large.4,5 The skin irritability to sodium lauryl sulfate in males and females has been studied by measurement of skin water vapor loss:6 the study showed that mean values of unirritated skin in females were significantly lower than those in male volunteers. The differences seem due to a lower basal metabolic rate in females. There was no significant difference between the mean values of irritated skin of male and female volunteers, and the irritation index was significantly lower in males than females. The study concluded that female skin is more prone to irritation. In spite of this, some other reports confirm that there is no difference in reactivity between the sexes.5–7 However, variations in skin reactivity during the menstrual cycle seem to occur: changes in skin extensibility and increased proneness to develop strong irritant reactions are reported during the menstrual phase.8,9
4.3 AGE AND BODY REGION Skin aging, more or less a physiological event, is characterized by several biological and histopathological 27
28
Handbook of Non-Invasive Methods and the Skin, Second Edition
changes. Transepidermal water loss (TEWL) and skin hydration both decrease during the aging process, maintaining a directly proportional relationship. The decrease of TEWL during life is conspicuous after the age of 60.10 Several factors may be responsible. The increased size of corneocytes and the increased thickness of the stratum corneum due to the greater accumulation of corneocytes related to an impaired desquamation11 are factors that should be considered. Similarly, corneum hydration is decreased in elderly subjects.12,13 Reduction in moisture content is more noticeable in exposed areas, where damage is the predominant factor accentuating aging.13 The simultaneous decrease of TEWL and water content of the corneum is a distinct feature of elderly skin. It confirms the decreased corneum hydration without impairment of the barrier function. Accordingly, dry skin in the elderly may be differentiated from pathologically dry skin since in the latter the barrier function is defective. One of the sites where xerotic changes occur more frequently in aged skin is the extensor aspects of the lower legs. Tagami and coworkers,14 using an evaporimeter to measure TEWL and a skin surface hygrometer15 to evaluate the hydration state of the skin, compared hydration of the lower legs in young and aged subjects. They found that skin of aged subjects is not particularly dry, compared with that of young adults. A lowered TEWL in elderly subjects was confirmed by other groups.10 A comparative study between aged individuals and children further substantiated that skin surface in aged people is not necessarily dehydrated, indicating that aging of the skin itself does not induce any marked derangement in stratum corneum function. On the other hand, stratum corneum moisturization detected by an impedance technique13 revealed differences between chronically sun-exposed skin and unexposed skin in the same individuals. Electrical skin impedance varies greatly in the different sites investigated in relation to age. Levels in elderly subjects were higher than those in young controls in both exposed and unexposed sites, except for the palms. Statistically significant differences were recorded only in chronically exposed areas, such as forehead and neck. The aging process varies greatly from site to site and from individual to individual, explaining the controversial data of the literature. Furthermore, senile xerosis represents a special pathologic condition affecting only certain subjects. Stratum corneum obtained from patients with senile dry skin showed a reduced capacity for secondary bound water, which plays an important role in maintaining corneum flexibility and suppleness.14 The skin of children is more easily irritated than that of adults.4 The skin of elderly persons is also more reactive than that of younger adults. This is attributed to the dry skin factor, which ensues with age.5
Many differences between a senile and young epidermis have been described, but a consistent interpretation of the aging process has proved difficult. In part, this is because the epidermis varies from site to site.16 Young skin from the back,17 like that from the scalp and axilla,18 has deep and complex rete ridges, whereas that of the face18 has a fairly flat dermoepidermal junction. It is widely agreed that in areas where the junction is corrugated in youth, it becomes flattened by age.18–23 Similarly, there are differences in epidermal thickness even in young skin. On the face or on the dorsum of the hand, for example, it is considerably greater than on the arms, legs, or trunk.18 In many areas the whole epidermis becomes thinner with age and the cells become less evenly aligned on the basement membrane and less regular in size, shape, and staining properties.18,22–25 The thickness of the stratum corneum is obviously not uniform over the whole body surface. In particular, the differentiation of the stratum corneum of the palms and soles is unlike that of the rest of the skin.26 Hammarlund et al.27 reported regional variations in newborns in the rate of TEWL. Comparing the abdomen, forearm, and buttocks, they found twofold TEWL on the buttocks. The findings were partially confirmed by Osmark et al.28 using a Meeco analyzer to detect TEWL. The technique allowed more precise measurements, not biased by air turbulence and ambient relative humidity over the probe. They recorded similar TEWLs on these three sites, but found a lower hydration level (obtained by inducing skin occlusion with plastic film for 1 hour) on the abdomen, reporting a decreased water-holding capacity on this site. In adults too, regional variations of moisturization reflecting differences of thickness and function of the corneum occur. Tagami et al.29 reported drier skin on the extremities than on the trunk. This correlates with the fact that clinically dry skin tends to develop more frequently on the limbs during winter. The reduction in TEWL with an increased thickness of the stratum corneum is not as much as would be expected.5 In fact, in some regions, especially on the palms and soles, TEWL increases with the thickness of the stratum corneum. The increase in thickness of the stratum corneum seems to be compensated by a corresponding increase in its diffusivity,5 resulting in skin with a relatively uniform steady-state TEWL over many parts of the body.26 However, not all differences in thickness are compensated in this manner, and some variations in the TEWL with the regional skin sites do exist. This regional variation in TEWL and permeability cannot be explained on the basis of differences in the chemical nature of the keratin molecule.30 The regional variations in the total lipid concentrations of the stratum corneum, however, may be the most critical factor determining the regional variations in TEWL and permeability.31
The Skin Integument: Variation Relative to Sex, Age, Race, and Body Region
Thus, barrier efficiency is not uniform over the whole body surface.32 The scrotum has long been known to be particularly permeable.33 The face, forehead, and dorsa of the hands may also be more permeable to water than the trunk, arms, and legs. The palms are practically impermeable, except to water and most water-soluble molecules. This may be the major reason why contact dermatitis is seen less often on the palms than on the dorsa of the hands.32
4.4 RACE 4.4.1 PHYSIOLOGICAL DIFFERENCES Although stratum corneum thickness is equal in blacks and whites,34,35 more tape strippings are required to remove the stratum corneum in blacks. Weigand and associates36 reported this is due to an increased number of corneocyte layers in black skin. Moreover, the same study revealed a great interindividual variability in the black race, whereas data from white subjects were more homogeneous. A correlation between the number of stratum corneum layers and the degree of pigmentation has not yet been demonstrated. The increased number of cell layers in the stratum corneum and the increased resistance to stripping could be related to lipid molecules in the intercellular matrix that increase cell cohesion. Indeed, Rienertson and Wheatley,37 investigating the stratum corneum lipid content, found higher values in blacks. Weigand and coworkers36 confirmed this result. Other parameters investigated in different studies, such as skin electrical resistance,38 were consistent with these findings.39
4.4.2 PERCUTANEOUS ABSORPTION In vitro penetration of fluocinolone acetonide through skin samples obtained from amputated black and white legs revealed an increased permeability in whites.40 In vitro water permeation through human skin did not reveal the racial differences that had been reported by Bronaugh and coworkers.41 In vivo studies show different patterns of penetration depending on the molecules tested.42 Tritiated diflorasone diacetate does not have different pharmacokinetics in blacks and whites.43 Wedig and Maibach44 applied Clabeled dipyrithione in different vehicles to stripped and unstripped skin of black and white volunteers and found 34% less absorption in blacks. A significantly lower penetration in blacks (47%) was also noted when a cosmetic vehicle (1:12:22:25:39 sodium lauryl sulfate:propylene glycol:stearyl alcohol:white petrolatum:distilled water) was compared with methyl alcohol on the forehead and when the methyl alcohol vehicle was compared with the shampoo vehicle on the scalp. The penetration of intact
29
vs. stripped skin by either the cosmetic cream or the shampoo vehicle was not different. Racial differences in methylnicotinate-induced vasodilatation in human skin were studied by Guy and associates; they induced vasodilatation by applying the substance to the skin and monitored the response with laser Doppler velocimetry,45 reporting statistically indistinguishable differences among the groups in the time-topeak response, the area under the response–time curve, and the time from the response to 75% decay. Only the magnitude of the peak response revealed some significant differences, with increased levels in young white subjects. However, no important differences seem to exist between black and white skin when tested with this chemical model.46
4.4.3 BIOMECHANICAL PROPERTIES, TEWL, SUSCEPTIBILITY TO IRRITANTS
AND
These parameters have been measured in whites, Hispanics, and blacks to assess whether the melanin content could induce changes in skin biophysical properties.47 Differences appear in skin conductance but are more marked in biomechanical parameters: skin extensibility, skin elastic modulus, and skin recovery. These relative variations of the parameters on dorsal and ventral sites are different according to the races and highlight the influence of solar irradiation on skin and the role of melanin in maintaining it unaltered. Skin lipids may play a role in modulating the relationship between stratum corneum water content and TEWL, resulting in higher conductance values in blacks and Hispanics. Previously, equal baseline TEWL on the back was reported48,49 among whites, blacks, and Hispanics. Moreover, TEWL revealed a different pattern of reaction in whites after chemical exposure to sodium lauryl sulfate, with blacks and Hispanics developing stronger irritant reactions after exposure to 2% sodium lauryl sulfate. Skin color and race are an influential factor determining skin reactivity; black skin (Negroid) is the least susceptible to irritants.50 Darkly pigmented individuals from the Mediterranean region are also less susceptible than light-complexioned individuals. It is thus reasonable to assume that fair-skinned persons of Celtic origin (Scottish–Irish– Welsh) have the most susceptible skin.51 All races show significant differences between the volar and dorsal forearms.47 These results are in apparent contrast with TEWL recordings. Indeed, the higher the stratum corneum water content, the higher the TEWL that may be expected.52 The data may be explained on the basis of different intercellular cohesion, lipid composition, or hair distribution. A greater cell cohesion with a normal TEWL could result in increased skin water content. The racial variability should be taken into account in terms of different skin responses to topical and environmental
30
Handbook of Non-Invasive Methods and the Skin, Second Edition
agents. Race provides a useful tool to investigate and compare the effects of lifetime sun exposure. It is clearly evident that melanin protection prevents sun damage; differences between sun-exposed and sun-protected areas are not detectable in races with dark skin.
REFERENCES 1. Ebling, F.J.G., Eady, R.A.J., and Leigh, I.M., Anatomy and organization of human skin, in Textbook of Dermatology, Rook, A.J., Wilkinson, D.S., and Ebling, F.J.G., Eds., Blackwell Scientific Publications, Oxford, 1992, p. 49. 2. Leveque, J.L., Ed., Cutaneous Investigation in Health and Disease, Noninvasive Methods and Instrumentation, Marcel Dekker, New York, 1989. 3. Stern, R.S., The epidemiology of cutaneous disease, in Dermatology in General Medicine, Vol. 1, 3rd ed., Fitzpatrick, T.B., Eisen, A.Z., Wolff, K., Freedberg, I.M., and Austen, K.F., Eds., McGraw-Hill, New York, 1987, p. 7. 4. Kligman, A.M. and Wooding, W.M., A method for the measurement and evaluation of irritants on human skin, J. Invest. Dermatol., 49, 78, 1967. 5. Pinnagoda, J., Transepidermal Water Loss: Its Role in the Assessment of Susceptibility to the Development of Irritant Contact Dermatitis, Ph.D. thesis, London University, July 1990. 6. Goh, C.L. and Chia, S.E., Skin irritability to sodium lauryl sulphate — as measured by skin water vapour loss — by sex and race, Clin. Exp. Dermatol., 13, 16, 1988. 7. Coenraads, P.J., Lee, J., and Pinnagoda, J., Changes in water vapour loss from the skin of metal industry workers monitored during exposure to oils, Scand. J. Work Environ. Health, 12, 494, 1986. 8. Berardesca, E., Gabba, P., Farinelli, N., Borroni, G., and Rabbiosi, G., Skin extensibility time in women. Changes in relation to sex hormones, Acta Derm. Venereol., 69, 431, 1989. 9. Agner, T., Damm, P., and Skouby, S., Menstrual cycle and skin reactivity, J. Am. Acad. Dermatol., 24, 566, 1991. 10. Leveque, J.L., Corcuff, P., de Rigal, J., and Agache, P., In vivo studies of the evolution of physical properties of the human skin with age, Int. J. Dermatol., 23, 322, 1984. 11. Nicholls, S., King, C.S., and Marks, R., The Influence of Corneocytes Area on Stratum Corneum Function (abstract), paper presented at the ESDR Annual Meeting, Amsterdam, 1980. 12. Berardesca, E. and Maibach, H.I., Bioengineering and the patch test, Contact Derm., 18, 3, 1988. 13. Borroni, G., Berardesca, E., Bellosta, M., Bernardi, L., and Rabbiosi, G., Evidence for regional variations in water content of the stratum corneum in senile skin: an electrophysiologic assessment, Ital. Gen. Rev. Derm., 19, 91, 1982.
14. Tagami, H., in Cutaneous Aging, Kligman, A.M. and Takase, Eds., University of Tokyo Press, Tokyo, 1988, p. 99. 15. Tagami, H., Kanamura, Y., Inoue, K., Suehisa, S., Inoue, F., Iwatsuki, K., Yoshikuni, K., and Yamada, M., Water sorption-desorption test of the skin in vivo for functional assessment of the stratum corneum, J. Invest. Dermatol., 78, 425, 1982. 16. Graham-Brown, R.A.C. and Ebling, F.J.G., The ages of man and their dermatoses, in Textbook of Dermatology, Vol. 4, 5th ed., Rook, A.J., Wilkinson, D.S., and Ebling, F.J.G., Eds., Blackwell Scientific, Oxford, 1992, p. 2877. 17. Eller, J.J. and Eller, W.D., Oestrogenic ointments: cutaneous effects of topical application of natural oestrogens with report of three hundred and twenty-one biopsies, Arch. Dermatol. Syphilol., 59, 449, 1949. 18. Montagna, W., Morphology of the aging skin: the cutaneous appendages, in Advances in Biology of Skin, Vol. 6, Aging, Montagna, W., Ed., Pergamon Press, Oxford, 1965, p. 1. 19. Christophers, E. and Kligman, A.M., Percutaneous absorption in aged skin, in Advances in Biology of Skin, Vol. 6, Aging, Montagna, W., Ed., Pergamon Press, Oxford, 1965, p. 163. 20. Hill, W.R. and Montgomery, H., Regional changes and changes caused by age in the normal skin, J. Invest. Dermatol., 3, 321, 1940. 21. Lavker, R.M., Zheng, P., and Dong, G., Morphology of aged skin, Dermatol. Clin., 4, 379, 1986. 22. Montagna, W. and Carlisle, K., Structural changes in ageing skin, J. Invest. Dermatol., 73, 47, 1979. 23. Montagna, W. and Carlisle, K., Structural changes in ageing skin, Br. J. Dermatol., 122 (Suppl. 35), 61, 1990. 24. Gilchrest, B.A., Skin and Ageing Processes, CRC Press, Boca Raton, FL, 1984. 25. Lavker, R.M., Structural alterations in exposed and unexposed aged skin, J. Invest. Dermatol., 73, 59, 1979. 26. Scheuplein, R.J. and Blank, I.H., Permeability of the skin, Physiol. Rev., 51, 702, 1971. 27. Hammarlund, K., Nilsson, G., Oberg, A., and Sedin, G., Transepidermal water loss in newborn infants: relation to ambient humidity and site of measurement and estimation of total transepidermal water loss, Acta Paediatr. Scand., 68, 371, 1979. 28. Osmark, K., Wilson, D., and Maibach, H.I., In vivo transepidermal water loss and epidermal occlusive hydration in newborn infants: anatomical regional variations, Acta Derm. Venereol., 60, 403, 1980. 29. Tagami, H., Masatoshi, O., Iwatsuki, K., Kanamaru, Y., Yamada, M., and Ichijo, B., Evaluation of skin surface hydration in vivo by electrical measurement, J. Invest. Dermatol., 75, 500, 1980. 30. Tregear, R.T., The structures which limit the penetrability of the skin, J. Soc. Cosmet. Chem., 13, 145, 1962. 31. Elias, P.M., Cooper, E.R., Core, A., and Brown, B.E., Percutaneous transport in relation to stratum corneum structure and lipid composition, J. Invest. Dermatol., 76, 297, 1981.
The Skin Integument: Variation Relative to Sex, Age, Race, and Body Region
32. Baker, H., The skin as a barrier, in Textbook of Dermatology, Rook, A., Ed., Blackwell Scientific, Oxford, 1986, p. 355. 33. Smith, J.G., Jr., Fisher, R.W., and Blank, I.H., The epidermal barrier: a comparison between scrotal and abdominal skin, J. Invest. Dermatol., 36, 337, 1961. 34. Freeman, R.G., Cockerell, E.G., Armstrong, J., and Knox, J.M., Sunlight as a factor influencing the thickness of the epidermis, J. Invest. Dermatol., 39, 295, 1962. 35. Thomson, M.L., Relative efficiency of pigment and horny layer thickness in protecting the skin of Europeans and Africans against solar ultraviolet radiation, J. Physiol. (London), 127, 236, 1955. 36. Weigand, D.A., Haygood, C., and Gaylor, J.R., Cell layers and density of Negro and Caucasian stratum corneum, J. Invest. Dermatol., 62, 563, 1974. 37. Rienertson, R.P. and Wheatley, V.R., Studies on the chemical composition of human epidermal lipids, J. Invest. Dermatol., 32, 49, 1959. 38. Johnson, L.C. and Corah, N.L., Racial differences in skin resistance, Science, 139, 766, 1963. 39. Berardesca, E. and Maibach, H.I., Skin color and proclivity to irritation, in Exogenous Dermatoses, Menne, T. and Maibach, H., Eds., CRC Press, Boca Raton, FL 1990, p. 65. 40. Stoughton, R.B., Bioassay methods for measuring percutaneous absorption, in Pharmacology of the Skin, Montagna, W., Stoughton, R.B., and Van Scott, E.J., Eds., Appleton-Century-Crofts, New York, 1969, p. 542. 41. Bronaugh, R.L., Stewart, F.R., and Simon, M., Methods for in vitro percutaneous absorption studies. VII. Use of excised human skin, J. Pharm. Sci., 75, 1094, 1986. 42. Berardesca, E. and Maibach, H.I., Physical anthropology and skin: a model for exploring skin function, in Models in Dermatology 4, Maibach, H.I. and Lowe N., Eds., Karger, Basel, 1989, p. 202.
31
43. Wickema-Sinha, W.J., Shaw, S.R., and Weber, O.J., Percutaneous absorption and excretion of tritium-labelled diflorasone diacetate, a new topical corticosteroid in the rat, monkey and man, J. Invest. Dermatol., 7, 373, 1978. 44. Wedig, J.H. and Maibach, H.I., Percutaneous penetration of dipyrithione in man: effect of skin color (race), Am. Acad. Dermatol., 5, 433, 1981. 45. Guy, R.H., Tur, E., and Bierke, S., Are there age and racial differences to methylnicotinate-induced vasodilatation in human skin?, J. Am. Acad. Dermatol., 12, 1001, 1985. 46. Berardesca, E. and Maibach, H.I., Contact dermatitis in blacks, Dermatol. Clin., 6, 363, 1988. 47. Berardesca, E., de Rigal, J., Leveque, J.L., and Maibach, H.I., In vivo biophysical characterization of skin physiological differences in races, Dermatologica, 182, 89, 1991. 48. Berardesca, E. and Maibach, H.I., Racial differences in sodium lauryl sulphate induced cutaneous irritation: black and white, Contact Derm., 18, 65, 1988. 49. Berardesca, E. and Maibach, H.I., Sodium lauryl sulphate induced cutaneous irritation: comparison of white and Hispanic subjects, Contact Derm., 19, 136, 1988. 50. Kligman, A.M., Assessment of mild irritants, in Principles of Cosmetics for the Dermatologist, Frost, P. and Horwitz, S.N., Eds., C.V. Mosby, St. Louis, MO, 1982, p. 265. 51. Frosch, P.J. and Kligman, A.M., Recognition of a chemically vulnerable and delicate skin, in Principles of Cosmetics for the Dermatologist, Frost, P. and Horwitz, S.N., Eds., C.V. Mosby, St. Louis, MO, 1982, p. 287. 52. Rietschel, R.L., A method to evaluate skin moisturizers in vivo, J. Invest. Dermatol., 70, 152, 1978.
Variations and 5 Seasonal Environmental Influences on the Skin Chee Leok Goh National Skin Centre, Singapore
CONTENTS 5.1 5.2
Introduction..............................................................................................................................................................33 Effect of Seasonal Variation on Skin Functions .....................................................................................................33 5.2.1 Solar Radiation ............................................................................................................................................33 5.2.1.1 Immediate Effects of Solar Radiation .........................................................................................34 5.2.1.2 Long-Term Effects of Solar Radiation ........................................................................................34 5.2.2 Effect of Temperature on the Skin ..............................................................................................................34 5.2.3 Effect of Humidity on the Skin...................................................................................................................35 5.3 Other Environmental Factors Influencing the Skin ................................................................................................35 References .........................................................................................................................................................................35
5.1 INTRODUCTION The skin is subject to the influence of solar radiation, temperature, humidity, domestic contactants, occupational contactants, therapeutic agents, and a host of environmental agents. All of these environmental agents may have adverse effects on the skin. The characteristic structure of the human skin places it as an important interface between human beings and their physical, chemical, and biological environment.1 It is a primary organ of defense and adaptation. The skin is the largest organ of the human body. It is also the largest organ that is exposed to all elements of the external environment. It is a vulnerable target of environmental agents.2 As a target organ, the skin is capable of responding in a variety of pathologic patterns.1 It is also an important portal of entry for potentially hazardous agents. The cell structure and appendages of the skin provide it with defenses against environmental elements. It has protective elements against physical trauma, thermal stress, solar radiation, chemical agents, fluid loss, and antimicrobial agents. The stratum corneum provides a barrier against the various physical agents and biological agents, the sweat gland activities against thermal stress, and the pigmentary system, including the melanocytes and melanin, against ultraviolet radiation. This chapter discusses the effect of seasonal variation and environmental influences on the skin.
5.2 EFFECT OF SEASONAL VARIATION ON SKIN FUNCTIONS The climatic and physical environmental conditions in different latitudes of the world differ. In temperate latitudes, seasonal variations occur during different periods of the year. Such climatic differences have an influence on the integrity and functional capacity of the skin. Certain skin disorders are more prevalent in different countries because of climatic differences, and similarly, certain skin disorders tend to occur during different seasons of the year. The following environmental factors, which change with different seasons and which have an effect on the skin, will be discussed: 1. Solar radiation 2. Temperature 3. Humidity
5.2.1 SOLAR RADIATION Solar radiation, in particular ultraviolet light (UVL), has immediate and long-term effects on the skin. The flux of solar radiation varies during different seasons in temperate countries. Skin pigment provides protection against actinic or ultraviolet radiation, which causes sunburn (ultraviolet erythema — immediate effect) and fragmentation and destruction of elastic tissue fibers, UV-induced 33
34
Handbook of Non-Invasive Methods and the Skin, Second Edition
aging of the skin, actinic keratoses, skin cancer, and alteration of the immune function of the skin (long-term effects). 5.2.1.1 Immediate Effects of Solar Radiation Sunburn from UVL can be elicited in all human beings, but photosensitivity is inversely related to the degree of melanin pigmentation. UVL exposure results in the immediate and delayed dilation of blood vessels in the dermis, which is usually confined to the irradiated sites.3 UVB is the major cause of sunburn from sunlight. Cutaneous reactions from UVB are influenced by environmental conditions, season of the year, latitude and time of day, altitude, atmospheric pollution, and time of exposure and skin thickness and pigmentation, as well as other factors.4 UVB erythema becomes visible within 2 to 6 hours following irradiation, reaches a maximum at 24 to 36 hours, fades in 72 to 120 hours, and is followed in most individuals by increased skin pigmentation (tanning).4 Increased pigmentation provides protection against further damage from UVB, since the pigment is a remarkably effective absorber of UVB. Although UVA is 1000-fold less potent than UVB in causing erythema, its predominance in the solar spectrum reaching the Earth’s surface (10- to 100-fold more than UVB) may account for its toxic effect on the skin. Highintensity UVA light sources, which may emit as much as five times more UVA than sunlight, widely used for cosmetic tanning, have effects on the skin and contribute to the long-term effects of UVL on the skin. The effect of UVB and UVA on the skin has been studied in detail by Gilchrest et al.5,6 5.2.1.2 Long-Term Effects of Solar Radiation UVA, in addition to UVB, is believed to contribute to the long-term effects of chronic sun exposure, including premature skin aging, actinic keratosis, and skin carcinogenesis.7 Solar radiation is the principal cause of skin cancer in humans. The most important wavelength responsible is UVB (290 to 320 nm). Recent studies have documented that the environmental flux of UVL radiation from the sun is increasing, especially over the North and South Poles.16–18 This is contributed by the liberation into the atmosphere of tons of chlorofluorocarbons by human activities, which eventually removes the protective ozone shield in the stratosphere.19,20 UVB irradiation has been shown to induce immunosuppression.21 Irradiation with high-dose UVB results in systemic immunosuppression, while exposure to low-dose UVB produces local immunosuppression.22,23 There is substantial evidence to implicate the effect of UVL on the epidermal Langerhans cells as the cause of the change in immunosuppression.21 The effect of UVL on Langerhans
cells studies has also been demonstrated by studies that found lower Langerhans cells density in the non-sunexposed skin than in chronically sun-exposed skin.24 The effect of seasonal variation in UVL flux, in particular UVB, may have an influence on the immunological response of the skin to contact allergens. The afferent and efferent limbs of allergic contact dermatitis in experimental animals may be suppressed by irradiation with UVB.25 Bruze26 found fewer positive patch tests per tested patient in Sweden during the summer months of June, July, and August than the other months. Similarly, Veien et al.27 in Denmark also found significantly lower patch test reactivity during the same period when compared to other months. Epidemiological evidence has also shown the association of UVL exposure, which differs in different latitudes and different seasons and skin cancers. Studies have identified that in the white populations, there was an inverse relationship between latitude and the incidence of skin cancers.28 Skin cancers showed a rising incidence with increasing dose of UVL exposure at different latitudes in North America.29 A linear relationship was also observed on the incidence of no-melanoma skin cancers in countries of different proximity to the equator. A linear relationship was observed between the incidence of nonmelanoma skin cancers and the annual ultraviolet solar radiation.30
5.2.2 EFFECT
OF
TEMPERATURE
ON THE
SKIN
The skin is an important thermoregulation organ. The rate of blood flow and sweating controls body temperature. The body temperature is maintained at a very constant temperature with minimal variations. This is vital to the function of the various body organs. One of the body responses to an increase in environmental temperature is increased rate of blood flow through dilatation of dermal capillaries and stimulation of the sweat glands to increase secretion of sweat. Sweating allows the evaporation process to occur, leading to loss of skin surface heat. Increased sweating is associated with increased hydration of the stratum corneum. An increased hydration of stratum corneum will enhance the penetration of chemical agents on the skin. Excessive sweating has a clinical impact on contact dermatitis. Olumide et al.8 reported a high incidence in Nigerian workers of contact allergy to clothing dyes from work uniforms, caused by enhanced dissolution of dyes from clothing in a hot environment. Increased sweating secondary to high environmental temperature also tends to provoke workers to discard protective clothing, and therefore expose workers’ skin to irritants and allergens.9 Excessive sweating in intertriginous areas leads to skin maceration and dermatitis. It predisposes the moist skin to colonization of fungus and superficial fungal infections.9
Seasonal Variations and Environmental Influences on the Skin
Excessive sweating due to high ambient temperature may lead to sweat duct swelling, resulting in obstruction, leading to miliaria. If severe, heat exhaustion and heat stroke may occur.
5.2.3 EFFECT
OF
HUMIDITY
ON THE
SKIN
The effect of high ambient humidity on the skin is similar to that of high temperature. High ambient humidity prevents the skin surface sweat from evaporating and leads to an increased hydration of the stratum corneum. Low humidity has been reported to cause skin symptoms. Rycroft et al.10–12 described a phenomenon of “lowhumidity occupational dermatoses” that affected office workers. Affected workers presented with itch and urticaria on covered parts of the body and scaly eczema on the face, scalp, and ears. The cause was believed to be due to low relative humidity in the work environment. The symptoms in these workers improved when the relative humidity of the work environment was raised above 45%. Gaul and Underwood13 reported skin chapping on the hands, lips, and face and ichthyosis on lower legs and arms in subjects who were exposed to low environmental humidity. Similar findings were also reported by Chernosky14 in patients living in air-conditioned houses when the environmental humidity was lowered drastically. However, the effect of low humidity on skin symptoms has been disputed. Andersen et al.15 could not evoke similar symptoms in subjects exposed to 78 hours in dry air. Subjects could not accurately judge whether they were exposed to low (20%) or high (70%) air humidity when the temperatures were held constant.15
5.3 OTHER ENVIRONMENTAL FACTORS INFLUENCING THE SKIN Other factors influencing the skin include environmental contactants. The effects of skin contactants in different countries depend on several factors. They are influenced by the prevailing type of industry,31 availability of topical medicaments and prescribing habits of physicians,32 cultural and traditional habits of individuals in the country, and also the fauna of the country.9 The type of prevailing industrial activities in a country influences the prevalence of the type of contact dermatitis. For example, contact allergy to chromates from cement (representing about 6% of patients attending its patch test clinic) was prevalent in Singapore between 1980 and 1985 and declined (to less than 1% of patients attending its patch test clinic) after 1985, when the construction industry experienced a slump. Similarly, the prevalence of contact irritant dermatitis from solvents and cutting fluids, which are widely used in the electronics and metals industries, recorded an increase after 1985 as the two industrial activities took more prominence.34
35
Topical medicament allergy is common in developing countries where relatively cheap and more sensitizing medicaments are more widely used than in developed countries. In India contact allergy to nitrofurazone cream and in Singapore contact allergy to proflavine lotion were common causes of contact allergy to topical medicament, respectively.33,35 These allergens are not known to cause problems in developed countries. Preservatives allergy also varies in different regions of the world. Formaldehyde and formaldehyde releasers are common preservatives used in cosmetics and are the common cause of contact allergies from preservatives in Europe. However, contact allergies to formaldehyde and formaldehyde releasers are relatively uncommon in Singapore and probably in Southeast Asia. One reason could be the widespread use of Japanese cosmetics in Singapore, in which formaldehyde and formaldehyde releasers are not used as preservatives.36 Allergies from plants differ in different regions of the world; e.g., primin allergies from primulas, which are common in Europe, are uncommon in tropical countries. Rhus allergy is uncommon in Southeast Asian countries, whereas rengas allergy is very common in Southeast Asian countries.37 The characteristic warm and humid tropical climate is unique compared to the temperate climates. The tropical climate varies minimally throughout the year. The average ambient temperature of 30˚C and humidity of more than 70% are about the same throughout the year. Other types of skin disorders associated with such climate and peculiar to the tropics include: 1. Acne estivalis, a condition described in patients who developed acneiform eruption after spending time in the tropical climate. The exact mechanism is unknown, but heat is believed to play a role. Heat and high humidity have been known to affect the pilosebaceous unit.37,38 2. Miliaria, a common disorder in the tropics resulting from swelling of the keratin lining of the sweat ducts due to heat and high humidity. 3. Skin infections, high heat, and humidity in the tropics promote sweating and sweat retention, especially on skin folds, resulting in skin maceration. Secondary bacterial infection and fungal infections on macerated skin are common.
REFERENCES 1. Suskind, R.R., The environment and the skin, Environ. Health Perspect., 20, 27, 1977. 2. Suskind, R.R., Environment and the skin, Med. Clin. North Am., 74, 307, 1990.
36
Handbook of Non-Invasive Methods and the Skin, Second Edition
3. Farr, P.M. and Diffey, B.C., The vascular response of human skin to ultraviolet radiation, Photochem. Photobiol., 44, 501, 1986. 4. Harber, L.C. and Bickers, D.R., Eds., Photosensitivity Diseases: Principles of Diagnosis and Treatment, 2nd ed., B.C. Decker, Toronto, 1989, p. 112. 5. Gilchrest, B.A., Soter, N.A., Stoff, J.S., et al., The human sunburn reaction: histologic and biochemical studies, J. Am. Acad. Dermatol., 4, 411, 1981. 6. Gilchrest, B.A., Soter, N.A., Hawk, J.L.M., et al., Histologic changes associated with ultraviolet A-induced erythema in normal human skin, J. Am. Acad. Dermatol., 9, 213, 1983. 7. Staberg, B., Wulf, H.C., Klemp, P., et al., The carcinogenic effect of UVA irradiation, J. Invest. Dermatol., 81, 517, 1983. 8. Olumide, Y., Oleru, G.U., and Enu, C.C., Cutaneous implications of excessive heat in the work-place, Contact Derm., 9, 360, 1983. 9. Goh, C.L., Exogenous dermatoses in the tropics, in Exogenous Eczema, Menne, T. and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1990, p. 351. 10. Rycroft, R.J.G., Occupational dermatoses among office personnel, Occup. Med., 1, 323, 1986. 11. Rycroft, R.J.G., and Smith, W.D.L., Low humidity occupational dermatoses, Contact Derm., 6, 488, 1980. 12. White, I.R. and Rycroft, R.J.G., Low humidity occupational dermatoses: an epidemic, Contact Derm., 8, 287, 1982. 13. Gaul, L.E. and Underwood, G.B., Relation of dew point and barometric pressure to chapping of normal skin, J. Invest. Dermatol., 19, 9, 1952. 14. Chernosky, M.E., Pruritic skin disease and summer air conditioning, JAMA, 179, 1005, 1962. 15. Andersen, I.B., Lundqvist, G.R., Jensen, P.L., and Proctor, D., Human response to 78 hour exposure to dry air, Arch. Environ. Health, 29, 319, 1974. 16. Callis, L.B. and Natarajan, M., Ozone and nitrogen dioxide changes in the stratosphere during 1979–1984, Nature, 323, 772, 1986. 17. Solomon, S., Garcia, R.R., Rowland, F.S., and Wuebbles, D.J., On the depletion of the Antartic ozone, Nature, 321, 755, 1986. 18. Stolarski, R.S., Kreuger, A.J., Shcoeberl, M.R., McPeters, R.D., Newman, P.A., and Alper, J.C., Nimbus and satellite measurements of the springtime Antarctic ozone decrease, Nature, 322, 808, 1986. 19. Farman, J.C., Gardiner, B.G., and Shanklin, J.D., Large losses of total ozone in Antarctica reveal seasonal CLOx/Nox interactions, Nature, 315, 207, 1985. 20. McElroy, M.B., Salawitch, R.J., Wofsy, S.C., and Longan, J.A., Reductions of Antarctic ozone due to synergistic interaction of chlorine and bromine, Nature, 321, 759, 1986. 21. Crus, P.D. and Bergstresser, P.R., The low-dose model of UVB-induced immunosuppression, Photodermatology, 5, 151, 1988.
22. Bergstresser, P.R., Ultraviolet B radiation induces “local immunosuppression,” Curr. Prob. Dermatol., 15, 205, 1986. 23. Kripke, M.L. and Morison, W.L., Modulation of immune function by UV radiation, J. Invest. Dermatol., 85, 62s, 1985. 24. Czernielewski, J.M., Masouye, I., Pisani, A., Ferracin, J., Auvolat, D., and Ortonne, J.P., Effect of chronic sun exposure on human Langerhans cell densities, Photodermatology, 5, 116, 1988. 25. Sjovall, P. and Moller, H., The influence of locally administered ultraviolet light (UVB) on the allergic contact dermatitis in the mouse, Acta Derm. Venereol., 65, 465, 1985. 26. Bruze, M., Seasonal influence on routine patch test results, Contact Derm., 14, 184, 1986. 27. Veien, N.K., Hattel, T., and Laurberg, G., Is patch testing a less accurate tool during the summer months, Am. J. Contact Derm., 3, 35, 1992. 28. Scotto, J. and Fraumeni, J., Skin cancer (other than melanoma), in Cancer Epidemiology and Prevention, Schotterfeld, D. and Fraumeni, J., Eds., W.B. Saunders, Philadelphia, 1982, p. 996. 29. Russell Jones, R., Consequences for human health of stratospheric ozone depletion, in Ozone Depletion, Health and Environmental Consequences, Russel Jones, R. and Wigley, T., Eds., John Wiley & Sons, New York, 1989, p. 207. 30. Gordon, D. and Silverstone, H., Worldwide epidemiology of premalignant and malignant cutaneous lesions, in Cancer of the Skin, Andrade, R., Ed., W.B. Saunders, Philadelphia, 1976, p. 405. 31. Goh, C.L., Epidemiology of contact allergy in Singapore, Int. J. Dermatol., 27, 308, 1988. 32. Goh, C.L., Contact sensitivity to topical antimicrobials. 1. Epidemiology in Singapore, Contact Derm., 21, 46, 1989. 33. Goh, C.L., Contact sensitivity to topical medicaments, Int. J. Dermatol., 28, 25, 1989. 34. Goh, C.L., Occupational dermatitis in Singapore. Changing pattern: 1985–1989, Hifu (Skin Research), 33, 95, 1991. 35. Goh, C.L., Contact sensitivity to proflavine, Int. J. Dermatol., 25, 449, 1986. 36. Goh, C.L., Allergic contact dermatitis from cosmetics, J. Derm., 14, 248, 1987. 37. Goh, C.L., Occupational allergic contact dermatitis from Rengas wood, Contact Derm., 18, 300, 1988. 38. Lobitz, W.C. and Dobson, R.L., Miliaria, Arch. Environ. Health, 11, 460, 1965. 39. Taylor, J.S., The pilosebaceous unit, in Occupational and Industrial Dermatology, Maibach, H.I. and Gellin, G.A., Eds., Year Book Medical, Chicago, 1982, p. 125.
Methods and 6 Non-Invasive Assessment of Skin Diseases Stefania Seidenari, Francesca Giusti, and Giovanni Pellacani Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 6.1 6.2
Introduction..............................................................................................................................................................37 Instrumental Assessment of Psoriasis .....................................................................................................................38 6.2.1 Assessment of the Extent of the Disease (A in the PASI) .........................................................................38 6.2.2 Assessment of Erythema and Skin Blood Flow (E in the PASI)...............................................................38 6.2.3 Assessment of Induration and Desquamation (I and D in the PASI) ........................................................39 6.2.4 Recent Techniques .......................................................................................................................................39 6.2.5 Assessment of the Barrier in Psoriasis .......................................................................................................39 6.2.6 Assessment of Side Effects of Topical Drugs.............................................................................................40 6.3 Instrumental Assessment of Sclerotic Skin.............................................................................................................40 6.3.1 Localized Scleroderma ................................................................................................................................40 6.3.1.1 Skin Elasticity ..............................................................................................................................40 6.3.1.2 Ultrasound ....................................................................................................................................40 6.3.2 Systemic Sclerosis .......................................................................................................................................40 6.3.2.1 Skin Mechanical Properties .........................................................................................................41 6.3.2.2 Skin Thickness .............................................................................................................................41 6.3.2.3 Echogenicity Measurements ........................................................................................................41 6.3.2.4 Confocal Laser Scanning Microscopy.........................................................................................41 6.4 Instrumental Assessment of Atopic Dermatitis.......................................................................................................42 6.4.1 Transepidermal Water Loss in AD ..............................................................................................................42 6.4.2 Conductance and Capacitance as Parameters for Skin Hydration in AD ..................................................42 6.4.3 Reactivity to Irritants in AD Subjects .........................................................................................................43 6.4.4 Bioengineering Techniques and Topical Agents for AD ............................................................................43 References .........................................................................................................................................................................43
6.1 INTRODUCTION The severity of an inflammatory skin disease is a basic element both for guiding the diagnostic process and therapeutic choice and for making the prognosis. To get rid of avoidable mistakes, its evaluation should be as objective and reproducible as possible. Moreover, an estimate of the cost–benefit ratio of new treatments, in both economic and clinical terms, is increasingly required. Controlled clinical trials, where new therapies are evaluated in comparison to consolidated ones, represent a prerequisite for an objective assessment. Furthermore, a decisive loan in the definition and grading of skin alterations occurred with the introduction of standardized schemes to assess skin diseases. However, in spite of attempts to regulate clinical
judgment, there are still wide variations both in assessment rules and in the interpretation of their use, thus making intra- and interobserver variations unavoidable. In fact, even trained clinical observers’ opinions may greatly vary when attempting to repeatedly document the severity of the same clinical situation. Finally, since nearly all drugs do not lead to prompt and complete clearance of objective and subjective symptoms, but to a gradual reduction in clinical signs, there is a need for accurate measurements and for statistical analysis of the results to reach a valid conclusion concerning their real effectiveness. During the last decade several instrumental devices intended for in vivo measurement of skin anatomy and physiology have become commercially available. Because of their non-invasive approach, these techniques give out 37
38
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 6.1 Values Referring to a Psoriatic Plaque (Mean Values ± s.d. of 30 Lesions) Compared with Normal Skin (30 Control Subjects)
Normal skin Psoriatic plaque
TEWL
Erythema (a*)
Dermal Echogenicity (0–30 areas)
Epidermal Reflectivity (201–255 areas)
4.9 ± 1.5 21.1 ± 7.2
3.4 ± 1.1 14.4 ± 3.8
1629.5 ± 517.4 22780.1 ± 2460.3
409.13 ± 159.8 7822.81 ± 3042.88
information referring to different body sites at the same time of the investigation and follow-up examinations without influencing the natural course of the disease or the therapeutic effects. Moreover, the same skin sites can be simultaneously studied by different techniques, each assessing various aspects of the disease and of the healing process. For different diseases, biophysical parameters correlating with clinical data and suitable for the followup were identified. For disease monitoring, data referring to normal skin, to basal conditions of diseased skin, and to successive records should be compared.
6.2 INSTRUMENTAL ASSESSMENT OF PSORIASIS Psoriasis is a chronic disease characterized by hyperproliferation of the epidermis, inflammation and dilation, and elongation of capillaries. The severity of psoriasis has to be assessed both to establish treatment protocols and to evaluate the results of therapy. Psoriasis is also a reference disease for assessment of the potency and efficacy of topical corticosteroids. Clinical scoring systems, based on the gradation of erythema, scaling, and infiltration, are easy and quick to use, but are never objective and seldom reproducible. In spite of this, considering 122 studies evaluating the efficacy of psoriasis treatments performed from 2001 to 2004, bioengineering devices were employed by three authors alone.1–3 The most popular method for the assessment of psoriasis is the psoriasis area and severity index (PASI).4 It is calculated as follows: PASI = 0.1 (Eh + Ih + Dh)Ah + 0.3 (Et + It + Dt)At + 0.2 (Eu + Iu + Du)Au + 0.4 (El + Il + Dl)Al, where E = erythema, I = induration, D = desquamation, and A = area. In the different sites, head (h), trunk (t), upper extremities (u), and lower extremities (l), a numerical value is given to the extent of the lesions (1 < 10%, 2 = 10 to 29%, 3 = 30 to 49%, 4 = 50 to 69%, 5 = 70 to 89%, and 6 > 90%). E, I, and D are scored according to a 5-point scale: 0 = no lesions, 1 = slight, 2 = moderate, 3 = marked, and 4 = very marked. Even considering the most relevant clinical aspects of psoriasis severity, the subjective assessment of erythema, infiltration, desquamation, and extent of the psoriasis area unavoidably leads to intra- and interobserver variations in
the scores. Non-invasive techniques allow the objective and reproducible quantification of parameters of the PASI score system. Therefore, the evaluation of psoriasis should be based on both clinical and instrumental data. Table 6.1 shows typical values referring to a psoriatic plaque compared to those of normal skin.
6.2.1 ASSESSMENT OF THE EXTENT (A IN THE PASI)
OF THE
DISEASE
It has been shown that human eye assessment of psoriasis extent shows great variations and usually overestimates the skin surface area. Since error estimates have a significant effect on the PASI score, computer image analysis performed on pictures representing both compromised and healthy skin areas were introduced for the estimation of the involved skin area (A in the PASI). These techniques enable the evaluation of the affected area with great accuracy,5–7 even though they have the disadvantage of being time-consuming (photographing, processing of images) and technically demanding. Moreover, the cylindrical shape of the limbs produces shadows, sometimes leading to problems in calculating the surface of the involved skin areas. In view of the above, a simple horizontal averaging program enabling shade correction was recently developed to provide a quick and accurate assessment of disease extent.8
6.2.2 ASSESSMENT OF ERYTHEMA FLOW (E IN THE PASI)
AND
SKIN BLOOD
The erythematous component of psoriatic lesions was evaluated by colorimetry, spectrophotometry, and laser Doppler velocimetry.1,5,9–12 When desquamation is absent, instrumental values referring to erythema correlate well with the inflammatory component of the disease. On the contrary, in a psoriatic plaque surmounted by thick scales, erythema values approach those of healthy skin. When desquamation was scarce, both the E (Dermaspectrometer) and a* (Minolta Chromameter) parameters showed values that were twice those of normal skin, whereas they decreased as scales became thicker.9 By evaluating the activity of tacalcitol vs. placebo and betamethasone valerate on psoriatic plaques in two different patient groups, we observed a decrease in erythema parameters according
Non-Invasive Methods and Assessment of Skin Diseases
39
to the response to therapy in one group, whereas erythema increased, owing to the removal of the scaly layer, in the other. Final colorimetric values were similar to baseline ones and did not prove useful in the assessment of the response to treatment.12 In psoriasis changes in microvasculature of the upper dermis, with elongation and dilation of skin capillaries, are associated with increased blood flow. In psoriatic plaques prior to therapy, skin blood flow was significantly higher than that in uninvolved skin.13,14 During treatment skin blood flow progressively decreased, approaching that of uninvolved skin in 3 to 4 weeks. However, paradoxically, after descaling treatment, removal of scales can lead to an increase in blood flow owing to the proximity of deeper and larger vessels contributing to the signal, giving rise to increased values.
6.2.3 ASSESSMENT OF INDURATION AND DESQUAMATION (I AND D IN THE PASI) Induration corresponding to skin thickening can be measured by A- and B-scanning methods. By means of Ascanning procedures, Serup15 demonstrated an increase in skin thickness in psoriatic lesions with respect to healthy skin. This difference proved higher on the limbs than on the trunk. Thus, to study the efficacy of antipsoriatic treatments, it is advisable to select plaques localized to the limbs also because of fewer variations. High-frequency ultrasound providing high-resolution images of cutaneous structures has been employed as a method for assessing the severity of psoriasis plaques by numerous authors.10–12,16–18 Employing 20-MHz B-scanning, we confirmed that thickness measurement methods enable the evaluation of the disease’s progress, since this parameter gradually decreases according to clinical response to therapy.16 Moreover, we observed that psoriatic skin appears less echogenic than healthy skin, with a thick epidermal band, a hypoechogenic subepidermal band, corresponding to a combination of elongated papillae and inflammation, and numerous parallel acoustic shadows (Figure 6.1). Utilizing image analysis based on segmentation procedures, we also noticed a progressive reduction both in dermal echogenicity and in amplitude values corresponding to epidermal hyperreflectivity at skin sites treated with anthralin.16 In contrast with colorimetry, 20-MHz sonography enabled the distinction between the effects of tacalcitol, placebo, and betamethasone valerate in patients with psoriasis. 12 Colorimetry and ultrasound were also employed for the simultaneous assessment of the efficacy of different drugs in the same psoriatic patient in the psoriasis plaque test.11 Twenty-megahertz sonography is suitable for studying pathological changes and pharmacological effects in the dermis and the subcutaneous fat, but for investigating the epidermis, frequencies between 50 and 100 MHz are more
FIGURE 6.1 Echographic image of a psoriatic plaque. The entry echo is thickened and a hypoechogenic band is observable immediately below. Echogenicity of the dermis is decreased compared with unaffected skin (upper part of the image).
effective.18,19 By using 100-MHz ultrasound in psoriatic skin, El Gammal et al.19 found that the thickened horny layer appears echorich and its echodensity decreases after application of petrolatum. The acanthotic epidermis plus the dermis with the inflammatory infiltrate are represented as an echo-poor band, whose sonometric thickness proved to be correlated with histometric thickness.
6.2.4 RECENT TECHNIQUES Optical coherence tomography (OCT) was recently employed for the assessment of the response to therapy in psoriasis.20 OCT images of psoriatic plaques are characterized by a thickening of the entrance signal, which is composed of several parallel layers, with dilated signalfree cavities and significantly lower attenuation with respect to healthy skin. These changes revert to normal after treatment.
6.2.5 ASSESSMENT
OF THE
BARRIER
IN
PSORIASIS
Psoriatic lesions, which are generally thick and scaly, are known to produce high transepidermal water loss (TEWL) values compared to nonaffected skin. Thus, the altered stratum corneum in psoriasis has a defective water barrier function. Combined measurements of TEWL and stratum corneum hydration in psoriasis have shown an inverse relationship between water barrier and reservoir function; i.e., the higher the TEWL, the lower the skin hydration.21,22 However, these alterations are not directly correlated to the clinical situation and to response to therapy. Serup and Blichman21 reported that neither TEWL nor stratum
40
Handbook of Non-Invasive Methods and the Skin, Second Edition
corneum hydration was correlated to scaling, and they concluded that low conductance of psoriatic skin is probably a manifestation of abnormal keratinization.
6.2.6 ASSESSMENT DRUGS
OF
SIDE EFFECTS
OF
TOPICAL
Non-invasive bioengineering methods have also been employed to investigate the adverse effects of antipsoriatic drugs, such as calcipotriol23–25 and dithranol.26 Reactions to 48-hour patch tests with calcipotriol ointment resulted in redness, as evaluated by a Minolta Chromameter and laser Doppler flowmetry, whereas edema formation and skin thickening were observed in advanced reactions only by means of ultrasound.23 Clinically and instrumentally (TEWL, capacitance), Effendy et al.24 demonstrated that calcipotriol (0.005%) is less irritating than 1% sodium lauryl sulfate (SLS). We evaluated the irritant reaction induced by dithranol by means of visual assessment, colorimetry, and ultrasound and compared it to the 2% SLS response.26 In contrast with the SLS reaction, where a 24hour decrease in epidermal reflectivity was observable, responses to dithranol showed an accentuation of the hyperreflecting epidermal band at 24 to 96 hours.
6.3 INSTRUMENTAL ASSESSMENT OF SCLEROTIC SKIN 6.3.1 LOCALIZED SCLERODERMA Localized scleroderma (morphea) is a disorder of unknown cause characterized by localized dermal hardening. The early changes are superficial and deep perivascular and interstitial infiltrates of mixed inflammatory cells. As the infiltrates become progressively lymphoplasmacytic, collagen bundles in the dermis become thickened and packed together. In late lesions the infiltrates of lymphocytes and plasma cells, sometimes accompanied by eosinophils, wane and leave thickened bundles of collagen, located very close to one another and oriented parallel to the skin surface. Localized scleroderma has been studied by skin elasticity measurements; however, ultrasound measurements are particularly suitable for its assessment. 6.3.1.1 Skin Elasticity Employing a suction chamber method, Serup and Northeved27 observed that skin distensibility and stiffness decreased in every phase of the morphea plaque in contrast to thickening. In active plaques of morphea, skin thickening was found to be associated with decreased extensibility in comparison with symmetrical healthy control areas, whereas in regressing lesions, skin thickness did not differ from control areas, in spite of a persistent lower extensibility.28
6.3.1.2 Ultrasound A thickening of the skin at morphea plaques was observed using both A- and B-mode ultrasound devices.29–31 Using A-mode ultrasound, Serup29 showed increased thickening from plaques with slight alterations to plaques of clinically advanced scleroderma. The relative increase in thickness was higher in thinner skin, for example, on the limbs with respect to the trunk, and in patients with one or a few morphea plaques than in those with generalized morphea. Alterations of the dermis and hypodermis in sclerotic skin characterized by inflammation surrounding the vessels, edematous swelling of collagen bundles, and fibrosis not only correspond to modifications in skin thickness, but also bring about modifications in ultrasound reflections and echogenicity transformations referring to both intensity and echo distribution (Table 6.2). To describe a morphea plaque after image analysis, five different parameters can be used, namely, the absolute and relative extensions of isoechogenic areas and the number, size, and density of the objects (areas with definite shape and extension) composing the image.31 When assessment is performed with intensity bands covering the lower part of the amplitude scale, the image referring to sclerotic skin tissue appears relatively homogeneous with few, large objects within a thickened skin block, which occupy a more extensive image surface in comparison to normal skin images characterized by small and closely packed spots. On the contrary, images of localized scleroderma transformed by intermediate- to high-amplitude intervals appear with fewer objects of approximately the same size or smaller, which are less compressed with respect to healthy skin images. By employing a 13-MHz ultrasound device with a 60mm penetration depth for investigating morphea plaques, Cosnes et al.32 recently described undulations in the dermis, disorganization, loss of thickness and thickened hyperechoic bands in the hypodermis, and a characteristic dense image resembling a flattened yo-yo. A 92% sensitivity and a 100% specificity for localized scleroderma were found when at least four of these five signs were present.
6.3.2 SYSTEMIC SCLEROSIS Systemic sclerosis is a heterogeneous disease characterized by the overproduction of extracellular matrix by fibroblasts, damage of the endothelium of small vessels, and activation of the immune system, resulting in infiltration of the lower dermis and the upper subcutis. Widespread skin involvement is generally associated with internal organ involvement and bad prognosis.33 An objective evaluation of the extension of skin involvement is not only needed for prognostic purposes, but also to identify effective therapeutic interventions. The modified Rodnan total
Non-Invasive Methods and Assessment of Skin Diseases
41
TABLE 6.2 Numerical Values from Elaboration of Echographic Images of Sclerotic Skin and Normal Skin
Skin thickness (mm) Echogenicity values (0–30 areas)
PSS
HS
Morphea Plaque
Healthy Skin
1.5 ± 0.41a 12.500 ± 10.000 a
1.17 ± 0.13 6000 ± 4000
2.50 ± 0.92 b 14.000 ± 12.000 b
1.97 ± 0.67 3.300 ± 4.000
Note: PSS = mean values ± s.d. of sclerotic skin on the back of the hand referring to 18 subjects affected by systemic sclerosis. HS = mean values ± s.d. of healthy skin on the back of the hand in 20 healthy controls. Morphea plaque = mean values ± s.d. of 60 morphea plaques. Healthy skin = mean values ± s.d. of healthy controlateral skin in patients affected by morphea. a b
Significant with respect to healthy controls. Significant with respect to healthy skin.
skin thickness score, carried out by clinical palpation, is a commonly used outcome measure in trials of systemic sclerosis,34,35 but reproducibility represents a difficult task even for experienced clinicians. Bioengineering techniques provide a useful support to clinical judgment in this case too. 6.3.2.1 Skin Mechanical Properties Non-invasive suction devices were employed for the study of the elastic properties of the skin in systemic sclerosis.36,37 Investigating patients at a different stage of the disease (edematous and indurative phase), Dobrev37 observed that the edematous phase was characterized by significantly lower immediate distension, final distension, and higher viscoelastic-to-elastic ratio compared with the indurative phase, whereas low values in skin distensibility correlated with severe skin thickness or hidebound skin. 6.3.2.2 Skin Thickness The skin–phalanx distance on the digits and forearm skin thickness are increased in patients with acrosclerosis with respect to age-matched healthy controls.38–40 Ihn et al.41 demonstrated that skin thickening is also present at skin sites that are clinically uninvolved by the sclerotic process, for example, on the chest. Conversely, a skin thickening is not always observable at induration sites. As evaluated by 20-MHz B-scanning, forehead and cheek skin is thinner in systemic sclerosis than in healthy subjects.42 Therefore, results of different skin area measurements should be considered together when assessing the disease activity and as a measure of outcome, and each measure should be carefully estimated according to the affected skin site. Recently, Moore et al.43 described a 17-point dermal ultrasound scoring system, to be employed for evaluating skin thickness in systemic sclerosis, combining clinical and sonographic assessment. The method proved extremely reliable and was proposed as a useful measure
of outcome, showing acceptable intra- and interobserver variations. 6.3.2.3 Echogenicity Measurements Morphologic modifications of 20-MHz ultrasound images of the skin are observable in systemic sclerosis patients. When echographic images of the skin on the back of the hand, the forehead, and the cheek of patients with systemic sclerosis undergoing image analysis were compared to those referring to a population of healthy sex- and agematched subjects, marked differences in the echostructure of the tissue were observable.42 On the forehead and the cheek, the thinned skin appeared more echogenic with respect to the skin of healthy subjects, with smaller hyporeflecting objects and greater hyperreflecting areas, whereas the thickened skin on the back of the hand was less echogenic, with large hyporeflecting areas and small hyperreflecting objects. Numerical description of echographic images of patients with systemic sclerosis provides significantly different data from those of normal skin (Table 6.2). 6.3.2.4 Confocal Laser Scanning Microscopy In vivo confocal laser scanning microscopy is a new noninvasive microscopy technique with high resolution at a specific depth, allowing skin imaging at a cellular level in real time with a penetration depth of about 200 to 250 μm. Evaluating micromorphologic characteristics of the skin in patients with systemic sclerosis referring to melanization, epidermal hypotrophy, dislocation of capillaries, and collagen deposits in the papillary dermis, by confocal laser scanning microscopy, Sauermann et al.44 were able to identify histometric parameters that significantly differed with respect to healthy controls. Although at present expensive and time-consuming, this method may open new possibilities for the assessment and monitoring of systemic sclerosis.
42
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 6.3 Capacitance, Transepidermal Water Loss, pH, and Echogenicity Values in Healthy and Atopic Skin (Mean Values ± s.d. Referring to 200 AD Patients and 200 Healthy Subjects) Capacitance Healthy Subjects Cheek Volar forearm Antecubital fossa Back of the leg Affected skin areas
61 ± 53.7 ± 66.2 ± 53.6 ± —
9.7 9.1 7.6 8.1
TEWL
AD Patients 49 ± 13 56.3 ± 12.1 60.6 ± 12.9 51.6 ± 11.4 23–56
Healthy Subjects 6.59 4.06 5.82 5.00
± 2.21 ± 1.98 ± 2.43 ± 4.29 —
Echogenicity (0–30 areas)
pH AD Patients
8.61 ± 3.57 7.71 ± 4.13 10.09 ± 5.27 6.96 ± 3.81 8–55
Healthy Subjects 5.43 4.86 4.70 5.32
± 0.42 ± 0.45 ± 0.49 ± 0.48 —
AD Patients
Healthy Subjects
AD Patients
5.49 ± 0.57 5.23 ± 0.74 5.12 ± 0.73 5.55 ± 0.74 4.9–6.6
— 993 ± 345 — — —
— 1029 ± 581 — — 1500–4000
6.4 INSTRUMENTAL ASSESSMENT OF ATOPIC DERMATITIS
without clinical signs other than hand eczema in adult life, TEWL proved normal on the upper arm.65
For the evaluation of atopic dermatitis (AD), the introduction of standardized clinical scoring systems considering severity of skin lesions, extent of skin involvement, and subjective symptoms has represented a decisive breakthrough.45–47 These clinical methods should be integrated by the evaluation of the status of the barrier and the degree of inflammation at selected skin sites. Since epidemiological studies show that AD subjects have a high incidence of hand eczema induced by irritant substances, in both a domestic and an occupational environment, most experimental studies have concentrated on barrier assessment by conductance, capacitance,48–51 and TEWL measurements, referring to both baseline conditions52–54 and when the skin is challenged with model irritants.55–59 Table 6.3 shows instrumental data referring to AD patients.
6.4.2 CONDUCTANCE AND CAPACITANCE AS PARAMETERS FOR SKIN HYDRATION IN AD
6.4.1 TRANSEPIDERMAL WATER LOSS
IN
AD
Most authors report increased TEWL values in AD subjects, both adults and children, at eczematous and also at apparently unaffected skin areas.60–63 In a study performed on AD children and controls, we found significant alterations in TEWL, measured at different body sites on both involved and uninvolved skin.62,63 In patients with current eczema, TEWL values at healthy skin sites were higher than in patients without lesions,63 indicating that the presence of active eczematous lesions induces an impairment of the barrier at clinically healthy skin sites, and that TEWL values may vary according to the course of the disease. In AD patients, the skin barrier impairment appears reversible and the long-lasting absence of eczema makes water barrier restoration possible. Atopic individuals without active dermatitis for the past 2 years showed TEWL values that were similar to those of healthy volunteers.64 Also in patients with a past history of AD, but
Skin hydration in AD patients was first assessed by capacitance measurements by Werner.66 Lodén et al.67 observed lower capacitance values in atopics, especially with increasing degree of dryness. Reduced capacitance values were observed in both eczematous and uninvolved skin,48–50,62,63,66,67 but the reduction was particularly pronounced in severe AD.51 At healthy skin sites, these alterations were more marked in patients with active disease.63 Stratum corneum hydration depends on both the ability to bind and the ability to retain water.52 Investigating the hydration and water retention capacity of unaffected skin in patients with AD, Berardesca et al.50 reported that in atopic patients the stratum corneum water retention capacity, described by the skin surface water loss profile, was significantly reduced. Dynamic methods, like the sorption–desorption test (SDT) and the moisture accumulation test (MAT), were developed in order to study the horny layer hydration kinetics.54 When performed in children with AD, we observed that the stratum corneum of uninvolved atopic skin was less hydrated, but more easily hydratable, by water coming both from the deeper layers and from the environment, with respect to the skin of healthy subjects.54 On the contrary, the eczematous areas showed an increased avidity to retain water, but a reduced absorption capacity. In AD patients the barrier impairment coincides with marked alterations in the amount and composition of epidermal lipids.58,59,68,69 When investigating the relationship between different lipid classes and barrier impairment in 47 patients with AD,70 we found an inverse correlation between TEWL and ceramides and a direct correlation between the increase in free cholesterol and the reduction in ceramide 3 levels. Electrical impedence was reported
Non-Invasive Methods and Assessment of Skin Diseases
43
to be dependent on the lipid content of the stratum corneum. 71 Nicander and Ollmar 71 showed significant changes between baseline values of clinically normal atopic skin and healthy skin. Furthermore, impedence showed larger reactivity in AD patients after skin stripping and lipid extraction.
AD.79–84 In fact, certain moisturisers were shown to improve water barrier function, as reflected by TEWL, and skin susceptibility to irritants in atopic patients.80–82 Moreover, a significant relationship was noted between the reduction in TEWL and the clinical improvement of dryness.83 Lodén et al.83 performed an instrumental and clinical comparison of the effects of a cream containing 20% glycerine and a cream with 4% urea on the skin of AD patients, showing the superiority of one of the two preparations. For evaluating the effects of a ceramidedominant, physiologic lipid-based emollient for the treatment of AD children, SCORAD values proved to be less effective than TEWL, which was more sensitive for detecting subtle fluctuations in AD activity and for predicting potential relapse.84 Skin thickness ultrasound measurement showed that long-term tacrolimus ointment therapy in patients with AD is nonatrophogenic and reverses corticosteroid-induced skin atrophy.85 As evaluated by TEWL measurements, the same topical product was shown to enhance experimentally induced irritant contact dermatitis and not to accelerate healing of irritant contact dermatitis and UVB erythema.86
6.4.3 REACTIVITY
TO IRRITANTS IN
AD SUBJECTS
Both clinical and instrumental data demonstrate a cutaneous hyperreactivity in subjects with active AD, experimentally exposed to irritants, which is related to the degree of severity and the extension of the dermatitis. In an inactive phase of the disease, AD patients may or may not show enhanced skin reactions upon exposure to irritants with respect to nonatopics.64,72–75 Tupker et al.61 investigated skin irritability by repeated applications of different irritants and found increased TEWL values, both before and after exposure, in subjects with a history of AD with respect to subjects with a history of allergic contact dermatitis or controls. Agner60 challenged the skin of the flexor side of the upper arm with SLS for 24 hours and observed greater reactions in atopic patients than in controls, as assessed both clinically and instrumentally. Moreover, postexposure TEWL, correlating with baseline values, was significantly higher in atopics than in controls. After SLS challenge, we observed both an increase in TEWL and a decrease in capacitance, which were more marked in subjects with AD than in controls.76,77 When investigating skin reactions to 30 minutes of 0.5% SLS on the forearms in 20 healthy volunteers and on 34 subjects with localized eczema in a chronic phase, comprising 14 atopic patients and 20 individuals with contact dermatitis, the echographic assessment of SLS-exposed areas showed a significant decrease in epidermal reflectivity, indicating barrier function damage in atopic subjects, but not in contact dermatitis patients.78 Moreover, SLS pretreatment of nickel patch test sites induced an earlier and a more marked cutaneous damage in atopic nickel-sensitive patients with respect to nickel-sensitive nonatopics, followed by a more intense allergic response, indicating the role of barrier damage by irritants in the induction and elicitation of contact eczema in atopics.49 On the contrary, postexposure TEWL, capacitance, and echogenicity values did not differ between subjects with mucosal atopy and healthy volunteers.76
6.4.4 BIOENGINEERING TECHNIQUES AGENTS FOR AD
AND
TOPICAL
Objective monitoring of barrier impairment in AD is of considerable interest in studies evaluating the efficacy of anti-inflammatory drugs and moisturizing creams on
REFERENCES 1. Suh, D.H., Kwon, T.E., Kim, S.D., Park, S.B., Kwon, O.S., Eun, H.C., and Youn, J.I., Changes of skin blood flow and color on lesional and control sites during PUVA therapy for psoriasis, J. Am. Acad. Dermatol., 44, 987, 2001. 2. Stucker, M., Memmel, U., Hoffmann, M., Hartung, J., and Altmeyer, P., Vitamin B12 cream containing avocado oil in the therapy of plaque psoriasis, Dermatology, 203, 141, 2001. 3. Goh, C.L. and Khoo, L., Laser Doppler perfusion imaging (LDPI) and transepidermal water loss (TEWL) values in psoriatic lesions treated with narrow band UVB phototherapy. Dermal vascularity may be useful indicator of psoriatic activity, Ann. Acad. Med. Singapore, 33, 75, 2004. 4. Fredriksson, T. and Pettersson, U., Severe psoriasis-oral therapy with a new retinoid, Dermatologica, 157, 238, 1978. 5. Ormerod, A.D., Dwyer, C.M., and Weller, R., A comparison of subjective and objective measures of reduction of psoriasis with the use of ultrasound, reflectance colorimetry, computerized video image analysis, and nitric oxide production, J. Am. Acad. Dermatol., 37, 51, 1997. 6. Savolainen, L., Kontinen, J., Roning, J., and Oikarinen, A., Application of machine vision to assess involved surface in patients with psoriasis, Br. J. Dermatol., 137, 395, 1997.
44
Handbook of Non-Invasive Methods and the Skin, Second Edition
7. Savolainen, L., Kontinen, J., Alatalo, E., Roning, J., and Oikarinen, A., Comparison of actual psoriasis surface area and the psoriasis area and severity index by the human eye and machine vision methods in following the treatment of psoriasis, Acta Derm. Venereol., 78, 466, 1998. 8. Tanaka, M., Gaskell, S., Edwards, C., and Marks, R., Simple horizontal averaging programme enables shade correction for image analysis in psoriasis, Clin. Exp. Dermatol., 25, 323, 2000. 9. Takiwaki, H. and Serup, J., Measurement of color parameters of psoriatic plaques by narrow- and reflectance spectrophotometry and tristimulus colorimetry, Skin Pharmacol., 7, 145, 1994. 10. Hoffmann, M., Dirschka, T., Schwarze, H., el-Gammal, S., Matthes, U., Hoffmann, A., and Altmeyer, P., 20 MHz sonography, colorimetry and image analysis in the evaluation of psoriasis vulgaris, J. Dermatol. Sci., 9, 103, 1995. 11. Bangha, E. and Elsner, P., Evaluation of topical antipsoriatic treatment by chromametry, visiometry and 20 MHz ultrasound in the psoriasis plaque test, Skin Pharmacol., 9, 298, 1996. 12. Seidenari, S., Magni, R., and Giannetti, A., Assessment of the activity of tacalcitol on psoriatic plaques by means of colorimetry and high frequency ultrasound, Skin Pharmacol., 10, 40, 1997. 13. Staberg, B. and Klemp, P., Skin blood flow in psoriasis during Goeckerman or beach tar therapy, Acta Derm. Venereol., 64, 331, 1984. 14. Khan, A., Schall, L.M., Tur, E., Maibach, H.I., and Guy, R.J., Blood flow in psoriatic skin lesions: the effect of treatment, Br. J. Dermatol., 117, 193, 1987. 15. Serup, J., Non-invasive quantification of psoriasis plaques: measurement of skin thickness with 15 MHz pulsed ultrasound, Clin. Exp. Dermatol., 9, 502, 1984. 16. Di Nardo, A., Seidenari, S., and Giannetti, A., B-scanning evaluation with image analysis of psoriatic skin, Clin. Exp. Dermatol., 1, 121, 1992. 17. Gupta, A.K., Turnbull, D.H., and Harasiewics, K.A., The use of high-frequency ultrasound as a method of assessing the severity of a plaque of psoriasis, Arch. Dermatol., 132, 658, 1996. 18. Unholzer, A. and Korting, H.C., High-frequency ultrasound in the evaluation of pharmacological effects on the skin, Skin Pharmacol. Appl. Skin Physiol., 15, 71, 2002. 19. El Gammal, S., El Gammal, C., and Kaspar, K., Sonography of the skin at 100 MHz enables in vivo visualization of stratum corneum and viable epidermis in palmar skin and psoriatic plaques, J. Invest. Dermatol., 113, 821, 1999. 20. Welzel, J., Bruhns, M., and Wolff, H.H., Optical coherence otmography in contact dermatitis and psoriasis, Arch. Dermatol. Res., 295, 50, 2003. 21. Serup, J. and Blichman, C., Epidermal hydration of psoriasis plaques and the relation to scaling, Acta Derm. Venereol., 67, 357, 1987.
22. Berardesca, E., Fideli, D., Borroni, G., Rabbiosi, G., and Maibach, H.I., In vivo hydration and water retention capacity of the stratum corneum in clinically uninvolved skin in atopic and psoriatic patients, Acta Derm. Venereol., 70, 400, 1990. 23. Fullerton, A., Avnstorp, C., Agner, T., Dahl, J.C., Olsen, L.O., and Serup, J., Patch test study with calcipotriol ointment in different patient groups, including psoriatic patients with and without adverse dermatitis, Acta Derm. Venereol. (Stockh.), 76, 194, 1996. 24. Effendy, I., Kwangsukstith, C., Chiappe, M., and Maibach, H.I., Effects of calcipotriol on stratum corneum barrier function, hydration and cell renewal in humans, Br. J. Dermatol., 135, 545, 1996. 25. Fullerton, A., Benfeldt, E., Petersen, J.R., Jensen, S.B., and Serup, J., The calcipotriol dose-irritation relationship: 48 hour occlusive testing in healthy volunteers using Finn chambers, Br. J. Dermatol., 138, 259, 1998. 26. Schiavi, M.E., Belletti, B., and Seidenari, S., Ultrasound description and quantification of irritant reactions induced by dithranol at different concentrations. A comparison with visual assessment and colorimetric measurements, Contact Derm., 34, 272, 1996. 27. Serup, J. and Northeved, A., Skin elasticity in localized scleroderma (morphoea): introduction of a biaxial in vivo method, and the measurement of tensile distensibility, hysteresis and resilient distension of diseased and normal skin, J. Dermatol., 12, 318, 1985. 28. Kalis, B., De Rigai, l., Leonard, F., Léveque, l.L., Riche, O., Le Care, Y., and De Lacharriere, O., In vivo study of scleroderma by non-invasive techniques, Br. J. Dermatol., 122, 785, 1990. 29. Serup, J., Localized scleroderma (morphoea): thickness of sclerotic plaques as measured by 15 Mhz pulsed ultrasound, Acta Derm. Venereol. (Stockh.), 64, 214, 1984. 30. Hoffmann, K., Gerbaulet, U., El-Gammal, S., and Altmeyer, P., 20 MHz B-mode ultrasound in the monitoring of the course of localized scleroderma (morphea), Acta Derm. Venereol. Suppl. (Stockh.), 164, 3, 1991. 31. Seidenari, S., Conti, A., Pepe, P., and Giannetti, A., Quantitative description of echographic irnages of morphea plaques as assessed by computerized image analysis on 20 MHz B-scan recording, Acta Derm. Venereol. (Stockh.), 75, 442, 1995. 32. Cosnes, A., Anglade, M.C., Revuz, J., and Radier, C., Thirteen-megahertz ultrasound probe: its role in diagnosing localized scleroderma, Br. J. Dermatol., 148, 724, 2003. 33. Clements, P.J., Hurwitz, E.L., Wong, W.K, Seibold, J.R., Mayes, M., White, B., Wigley, F., Weisman, M., Barr, W., Moreland, L., Medsger, T.A., Jr., Steen, V.D., Martin, R.W., Collier, D., Weinstein, A., Lally, E., Varga, J., Weiner, S.R., Andrews, B., Abeles, M., and Furst, D.E., Skin thickness score as a predictor and correlate of outcome in systemic sclerosis: high-dose versus low-dose penicillamine trial, Arthritis Rheum., 43, 2445, 2000.
Non-Invasive Methods and Assessment of Skin Diseases
45
34. Rodnan, G.P., Lipinski, E., and Luksick, J., Skin thickness and collagen content in progressive systemic sclerosis and localized scleroderma, Arthritis Rheum., 22, 130, 1979. 35. Clements, P.J., Lachenbruch, P., Siebold, J., White, B., Weiner, S., Martin, R., Weinstein, A., Weisman, M., Mayes, M., and Collier, D., Inter- and intra-observer variability of total skin thickness score (mod Rodnan TSS) in systemic sclerosis, J. Rheumatol., 22, 1281, 1995. 36. Enomoto, D.N., Mekkes, J.R., Bossuyt, P.M., Hoekzena, R., and Bos, J.D., Quantification of cutaneous sclerosis with a skin elasticity meter in patients with generalized scleroderma, Am. Acad. Dermatol., 35, 381, 1996. 37. Dobrev, H.P., In vivo study of skin mechanical properties in patients with systemic sclerosis, J. Am. Acad. Dermatol., 40, 436, 1999. 38. Serup, J., Quantification of acrosclerosis: measurement of skin thickness and skin-phalanx distance in females with 15 MHz pulsed ultrasound, Acta Derm. Venereol. (Stockh.), 64, 35, 1984. 39. Akesson, A., Forsberg, L., Hederstrom, E., and Wollheim, F., Ultrasound examination of skin thickness in patients with progressive systemic sclerosis (scleroderma), Acta Radiol. Diagn., 27, 91, 1986. 40. Myers, S.L., Cohen, J.S., Sheets, P.W., and Bies, J.R., B-mode ultrasound evaluation of skin thickness in progressive systemic sclerosis, J. Rheumatol., 13, 577, 1986. 41. Ihn, H., Shimozuma, M., Fujimoto, M., Sato, S., Kikuchi, K., Igarashi, A., Soma, Y., Tamaki, K., and Takehara, K., Ultrasound measurement of skin thickness in systemic sclerosis, Br. J. Rheumatol., 34, 535, 1995. 42. Seidenari, S., Belletti, B., and Conti, A., A quantitative description of echographic images of sclerotic skin in patients with systemic sclerosis, assessed by computerized image analysis on 20 MHz B-scan recordings, Acta Derm. Venereol. (Stockh.), 76, 361, 1996. 43. Moore, T.L., Lunt, M., McManus, B., Anderson, M.E., and Herrick, A.L., Seventeen-point dermal ultrasound scoring system: a reliable measure of skin thickness in patients with systemic sclerosis, Rheumatology (Oxford), 42, 1559, 2003. 44. Sauermann, K., Gambichler, T., Jaspers, S., Radenhausen, M., Rapp, S., Reich, S., Altmeyer, P., Clemann, S., Teichmann, S., Ennen, J., and Hoffmann, K., Histometric data obtained by in vivo confocal laser scanning microscopy in patients with systemic sclerosis, BMC Dermatol., 2, 1, 2002. 45. Kunz, B., Oranje, A.P., Labreze, L., Stalder, J.F., Ring, J., and Taieb, A., Clinical validation and guidelines for the SCORAD index: consensus report of the European Task Force on Atopic Dermatitis, Dermatology, 195, 10, 1997. 46. Sprikkelman, A.B., Tupker, R.A., Burgerhof, H., Schouten, J.P., Brand, P.L., Heymans, H.S., and van Aalderen, W.M., Severity scoring of atopic dermatitis: a comparison of three scoring systems, Allergy, 52, 944, 1997.
47. Barbier, N., Paul, C., Luger, T., Allen, R., De Prost, Y., Papp, K., Eichenfield, L.F., Cherill, R., and Hanifin, J., Validation of the eczema area and severity index for atopic dermatitis in a cohort of 1550 patients from the pimecrolimus cream 1% randomized controlled clinical trials programme, Br. J. Dermatol., 150, 96, 2004. 48. Gollhausen, R., The phenomenon of irritable skin in atopic eczema, in Handbook of Atopic Eczema, Ruzicka, T., Ring, J., and Przybilla, B., Eds., SpringerVerlag, Berlin, 1991, p. 306. 49. Seidenari, S., Reactivity to nickel sulfate at sodium lauryl sulfate pre-treated sites is higher in atopics: an echographic evaluation by means of image analysis performed on 20 MHz B-scan recordings, Acta Derm. Venereol. (Stockh.), 74, 245, 1994. 50. Berardesca, E., Fideli, D., Borroni, G., Rabbiosi, G., and Maibach, H.I., In vivo hydration and water-retention capacity of stratum corneum in clinically uninvolved skin in atopic and psoriatic patients, Acta Derm. Venereol. (Stockh.), 70, 400, 1990. 51. Tanaka, M., Okada, M., Zhen, Y.X., Inamura, N., Kitano, T., Shirai, S., Sakamoto, K., Inamura, T., and Tagami, H., Decreased hydration state of the stratum corneum and reduced amino acid content of the skin surface in patients with seasonal allergic rhinitis, Br. J. Dermatol., 139, 618, 1998. 52. Tagami, H., Kanamaru, Y., and Inoue, K., Water sorption-desorption test of the skin in vivo for functional assessment of the stratum corneum, J. Invest. Dermatol., 78, 425, 1982. 53. Werner, Y., Lindberg, M., and Forslind, B., The waterbinding capacity of stratum corneum in dry non-eczematous skin of atopic eczema, Acta Derm. Venereol. (Stockh.), 62, 334, 1981. 54. Pellacani, G. and Seidenari, S., Water sorption-desorption test and moisture accumulation test for functional assessment of atopic skin in children, Acta Derm. Venereol. (Stockh.), 81, 100, 2001. 55. Imokawa, G., Akasaki, S., Minematsu, Y., and Kawai, M., Importance of intercellular lipids in water-retention properties of the stratum corneum: induction and recovery study of surfactant dry skin, Arch. Dermatol. Res., 281, 45, 1989. 56. Elias, P.M. and Menon, G.K, Structural and lipid biochemical correlates of the epidermal permeability barrier, Adv. Lipid. Res., 24, 1, 1991. 57. Grubauer, G., Elias, P.M., and Feingold, K.R., Transepidermal water loss: the signal for recovery of barrier structure and function, J. Lipid. Res., 30, 323, 1989. 58. Melnik, B., Hollmann, J., Hofmann, U., Yuh, M.S., and Plewig, G., Lipid composition of outer stratum corneum and nails in atopic and control subjects, Arch. Dermatol. Res., 282, 549, 1990. 59. Yamamoto, A., Serizawa, S., Ito, M., and Sato, Y., Stratum corneum lipid abnormalities in atopic dermatitis, Arch. Dermatol. Res., 283, 219, 1991. 60. Agner, T., Susceptibility of atopic dermatitis patients to irritant dermatitis caused by sodium lauryl sulphate, Acta Derm. Venereol. (Stockh.), 71, 296, 1991.
46
Handbook of Non-Invasive Methods and the Skin, Second Edition
61. Tupker, R.A., Pinnagoda, J., Coenraads, P.J., and Nater, J.P., Susceptibility to irritants: role of barrier function, skin dryness and history of atopic dermatitis, Br. J. Dermatol., 123, 199, 1990. 62. Seidenari, S. and Giusti, G., Objective assessment of the skin of children affected by atopic dermatitis: a study on pH, capacitance and TEWL in eczematous and clinically uninvolved skin, Acta Derm. Venereol. (Stockh.), 75, 429, 1995. 63. Giusti, G. and Seidenari, S., La barriera cutanea nei bambini con dermatite atopica: valutazione strumentale in 200 pazienti e 45 controlli, Riv. Ital. Pediatr., 24, 954, 1998. 64. Löffler, H. and Effendy, I., Skin susceptibility of atopic individuals, Contact Derm., 40, 239, 1999. 65. Agner, T., Skin susceptibility in uninvolved skin of hand eczema patients and healthy controls, Br. J. Dermatol., 125, 140, 1991. 66. Werner, Y., The water content of the stratum corneum in patients with atopic dermatitis. Measurement with the Corneometer CM 420, Acta Derm. Venereol. (Stockh.), 66, 281, 1986. 67. Lodén, M., Olsson, H., Axéll, T., and Werner Linde, Y., Friction, capacitance and transepidermal water loss (TEWL) in dry atopic and normal skin, Br. J. Dermatol., 126, 137, 1992. 68. Imokawa, G., Abe, A., Jin, K., Higaki, Y., Kawashima, M., and Hidano, A., Decreased level of ceramides in stratum corneum of atopic dermatitis: an etiologic factor in atopic dry skin? J. Invest. Dermatol., 96, 523, 1991. 69. Schäfer, L. and Kragballe, K., Abnormalities in epidermal lipid metabolism in patients with atopic dermatitis, J. Invest. Dermatol., 96, 10, 1991. 70. Di Nardo, A., Wertz, P., Giannetti, A., and Seidenari, S., Ceramide and cholesterol composition of the skin of patients with atopic dermatitis, Acta Derm. Venereol. (Stockh.), 78, 27, 1998. 71. Nicander, I. and Ollmar S., Clinically normal atopic skin vs. non-atopic skin as seen through electrical impedance, Skin Res. Technol., 10, 178, 2004. 72. Stolz, R., Hinnen, U., and Elsner, P., An evaluation of the relationship between ‘atopic skin’ and skin irritability in metalworkers trainees, Contact Derm., 36, 281, 1997. 73. Basketter, D.A., Miettinen, J., and Lahti, A., Acute irritant reactivity to sodium lauryl sulfate in atopics and nonatopics, Contact Derm., 38, 253, 1998. 74. Hannuksela, A. and Hannuksela, M., Irritant effects of a detergent in wash, chamber and repeated open application tests, Contact Derm., 34, 134, 1996.
75. Van der Valk, P.G.M., Nater, J.P., and Bleumink, E., Vulnerability of the skin to surfactants in different groups of eczema patients and controls as measured by water vapour loss, Clin. Exp. Dermatol., 10, 98, 1985. 76. Seidenari, S., Belletti, B., and Schiavi, M.E., Skin reactivity to sodium lauryl sulfate in patients with respiratory atopy, J. Am. Acad. Dermatol., 35, 47, 1996. 77. Seidenari, S., Skin sensitivity, interindividual factors: atopy, in The irritant Contact Dermatitis Syndrome, Van der Valk, P.G. and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1996, p. 267. 78. Seidenari, S. and Di Nardo, A., B-scanning evaluation of irritant reactions with binary transformation and image analysis, Acta Derm. Venereol. (Stockh.), 175, 9, 1992. 79. Aalto-Korte, K., Improvement of skin barrier function during treatment of atopic dermatitis, J. Am. Acad. Dermatol., 33, 969, 1995. 80. Lodén, M., Barrier recovery and influence of irritant stimuli in skin treated with a moisturising cream, Contact Derm., 36, 256, 1997. 81. Held, E., Sveinsdottir, A., and Agner, T., Effect of longterm use of a moisturiser on skin hydration, barrier function and susceptibility to irritants, Acta Derm. Venereol. (Stockh.), 79, 49, 1999. 82. Lodén, M., Andersson, A.C., and Lindberg, M., Improvement in skin barrier function in patients with atopic dermatitis after treatment with a moisturizing cream (Canoderm®), Br. J. Dermatol., 140, 264, 1999. 83. Lodén, M., Andersson, A.C., Andersson, C., Frödin, T., Öman, H., and Lindberg, M., Instrumental and dermatological evaluation of the effect of glycerine and urea on dry skin in atopic dermatitis, Skin Res. Technol., 7, 209, 2001. 84. Chamlin, S.L., Kao, J., Frieden, I.J., Sheu, M.Y., Fowler, A.J., Fluhr, J.W., Williams, M.L., and Elias, P.M, Ceramide-dominant barrier repair lipids alleviate childhood atopic dermatitis: changes in barrier function provide a sensitive indicator of disease activity, J. Am. Acad. Dermatol., 47, 198, 2002. 85. Kyllönen, H., Remitz, A., Mandelin, J.M., Elg, P., and Reitamo, S., Effects of a 1-year intermittent treatment with topical tacrolimus monotherapy on skin collagen synthesis in patients with atopic dermatitis, Br. J. Dermatol., 150, 1174, 2004. 86. Fuchs, M., Schliemann-Willers, S., Heinemann, C., and Elsner, P., Tacrolimus enhances irritation in a 5-day human irritancy in vivo method, Contact Derm., 46, 290, 2002.
for Good Clinical Practice 7 Standards (GCP) Merete Thyme Quality Assurance Department, Scantox (part of LAB Research International), LI. Skensved, Denmark
CONTENTS 7.1 7.2 7.3 7.4 7.5 7.6 7.7
Introduction..............................................................................................................................................................47 Declaration of Helsinki ...........................................................................................................................................47 Background of Good Clinical Practice ...................................................................................................................47 Regulatory Requirements ........................................................................................................................................48 ICH-GCP Principles ................................................................................................................................................48 Essential Documents................................................................................................................................................49 Roles and Responsibilities.......................................................................................................................................49 7.7.1 Sponsor Responsibilities..............................................................................................................................49 7.7.2 Monitor Responsibilities..............................................................................................................................49 7.7.3 Investigator Responsibilities........................................................................................................................50 References .........................................................................................................................................................................52
7.1 INTRODUCTION Various national legislation places the responsibility for establishing the safety and efficacy of drugs on the manufacturer of the regulated product. The authorities are responsible for reviewing the test results and determining whether the product’s safety and efficacy can be demonstrated. The marketing of the product is permitted only when the agencies are satisfied that safety and efficacy have been established adequately.
7.2 DECLARATION OF HELSINKI The purpose of biomedical research involving human subjects is to improve diagnostic, therapeutic, and prophylactic procedures and the understanding of the etiology and pathogenesis of diseases. The World Medical Association has prepared recommendations as a guide to every physician in biomedical research involving human subjects. In 1964, the 18th World Medical Assembly adopted the Declaration of Helsinki. The declaration sets forth 12 basic principles for the protection of human subjects in both clinical and nonclinical research. The Declaration of Helsinki has been amended regularly during the years, with the latest revision in Edinburgh, Scotland, in October 2000. It must be stressed that the standards are only a
guidance to physicians all over the world. Physicians are not relieved from criminal, civil, and ethical responsibilities under the law of their own countries.
7.3 BACKGROUND OF GOOD CLINICAL PRACTICE Since the beginning of the 1960s the requirements concerning the conduct and documentation of clinical research have been increased heavily. The first inspection of clinical studies to ascertain the standards of conduct and record keeping took place by the Food and Drug Administration (FDA) of the U.S. in the early 1960s. During the early years of regulatory inspections a significant number of malpractices were discovered, which led to a number of legal prosecutions. Since the implementation of the FDA Clinical Inspection Programme in 1977, other authorities around the world issued their own guidelines concerning good clinical practice. With the requirement that clinical trials conducted in one country should be widely accepted internationally for registration purposes, the good clinical practice rules were developed internationally in order to agree on a common quality standard to be followed when conducting clinical trials. 47
48
Handbook of Non-Invasive Methods and the Skin, Second Edition
Good clinical practice is based on the philosophy that quality standards and principles should be harmonized between countries. This would ensure optimal use and protection of the subjects to be included in the clinical trials, thereby restricting the number of necessary subjects. It would not be necessary to repeat the trials in different countries, and the data would be reliable by following the same documentation practice. Last but not least, a goal was that the time for registration of the drug would be minimized. The definition of good clinical practice (GCP) as given by the GCP principles is “a standard for the design, conduct, performance, monitoring, auditing, recording, analyses, and reporting of clinical trials that provides assurance that the data and reported results are credible and accurate, and that the rights, integrity, and confidentiality of trial subjects are protected.” Compliance with GCP ensures that the rights, safety, and well-being of trial subjects are protected and that the clinical trial data are credible. GCP principles are not a term of science improvement; the purpose of GCP is to ensure a correct basis for drug registration.
7.4 REGULATORY REQUIREMENTS Based on the need for international harmonization of GCP, an International Conference on Harmonization (ICH) was held in 1991. The objective was to provide a unified standard of the European Union (EU), Japan, and the U.S. to facilitate the mutual acceptance of clinical data by the regulatory authorities in these jurisdictions. Regulatory authorities and the pharmaceutical industry from the EU, the U.S., and Japan participated in the conference. This resulted in the preparation of the ICHGCP guideline (ICH-GCP), which was finalized in 1996. Since 1997 it has been required that clinical trials that should be used as documentation to the authorities in connection with registration applications in the EU, the U.S., and Japan should be conducted in compliance with the ICHGCP. An exemption was investigator-initiated clinical trials that were not within the scope of this requirement. In May 2004 a new EU Clinical Trials Directive became effective. The directive applies to all phases of clinical trials and to academic as well as commercially driven studies. This means that investigator-initiated clinical trials are within the future scope of the requirements.
7.5 ICH-GCP PRINCIPLES The ICH-GCP describes in detail all perspectives related to:
• • • • • •
Protection of subjects (ethics committees, patient information, and informed consent) Roles and responsibilities of the involved parties (investigator, sponsor, monitor) Study design (randomization, blinding, statistical analysis) Quality assurance Data management Archiving of data
The ICH-GCP consists of the following stated 13 general principles, which give a good impression of the purpose of the guideline: 1. Clinical trials should be conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki, and that are consistent with GCP and the applicable regulatory requirement(s). 2. Before a trial is initiated, foreseeable risks and inconveniences should be weighed against the anticipated benefit for the individual trial subject and society. A trial should be initiated and continued only if the anticipated benefits justify the risks. 3. The rights, safety, and well-being of the trial subjects are the most important considerations and should prevail over interests of science and society. 4. The available non-clinical and clinical information on an investigational product should be adequate to support the proposed clinical trial. 5. Clinical trials should be scientifically sound and described in a clear, detailed protocol. 6. A trial should be conducted in compliance with the protocol that has received prior institutional review board (IRB)/independent ethics committee (IEC) approval/favorable opinion. 7. The medical care given to, and medical decisions made on behalf of, subjects should always be the responsibility of a qualified physician or, when appropriate, a qualified dentist. 8. Each individual involved in conducting a trial should be qualified by education, training, and experience to perform his or her respective task(s). 9. Freely given informed consent should be obtained from every subject prior to clinical trial participation. 10. All clinical trial information should be recorded, handled, and stored in a way that allows its accurate reporting, interpretation, and verification. 11. The confidentiality of records that could identify subjects should be protected, respecting the
Standards for Good Clinical Practice (GCP)
privacy and confidentiality rules in accordance with the applicable regulatory requirement(s). 12. Investigational products should be manufactured, handled, and stored in accordance with applicable good manufacturing practice (GMP). They should be used in accordance with the approved protocol. 13. Systems with procedures that ensure the quality of every aspect of the trial should be implemented.
7.6 ESSENTIAL DOCUMENTS The documents that serve to demonstrate the compliance of the investigator, sponsor, and monitor with GCP and with all applicable regulatory requirements have been defined as essential documents by ICH-GCP. These documents permit individually and collectively evaluation of the conduct of a clinical trial and the quality of the data produced. The minimum list of essential documents to be generated and maintained in a clinical trial has been specified in the guideline. A description is given of the purpose of each document, and whether it should be filed in either the investigator or sponsor files, or both. It is required that trial master files are established at the beginning of a trial, both at the investigator site and at the sponsor’s office. A final closeout of a trial can only be done when it has been confirmed that all necessary documents are available in the appropriate files. The essential documents have been grouped in three sections according to the stage of the trial during which they will normally be generated. The three stages are: 1. Before the clinical phase of the trial commences 2. During the clinical conduct of the trial 3. After completion or termination of the trial Upon request from the monitor, auditor, ethics committee, or competent authorities, the investigator/institution should make available for direct access all requested trial-related records.
7.7 ROLES AND RESPONSIBILITIES GCP defines roles and responsibilities of the different parties involved in the conduct of clinical trials. Besides the trial subjects there are regulatory authorities, ethics committees, data protection agencies, sponsors, monitors, investigators, and study staff at the investigator site, as well as at the sponsor site. The following sections give a summary of the roles and responsibilities to be undertaken by the sponsor, the monitor, and the investigator, who, apart from the subjects, are the main parties in a clinical trial.
49
7.7.1 SPONSOR RESPONSIBILITIES The sponsor is most often represented by the industry. The sponsor selects the investigators and should ensure that these have appropriate documented qualifications within the therapeutic area to be investigated. The sponsor is responsible for informing properly about the investigational medicinal products to be investigated. It is required that an investigator’s brochure is made available to the participating investigators prior to study start. An investigator’s brochure is a compilation of nonclinical and clinical data regarding the investigational medicinal product and includes all available safety data, e.g., nonclinical study results, reported adverse events. It is mandatory that written approval of the protocol has been obtained from regulatory authorities and ethics committees prior to initiation of a trial. It is the responsibility of the sponsor to ensure that all required approvals have been obtained. The sponsor must ensure that the investigational medicinal product has been manufactured according to good manufacturing practice (GMP), which is the quality assurance system applicable to drug manufacturing. Detailed accountability should be maintained for all investigational medicinal products from the production at the pharmaceutical company to the use by the individual subjects within the trial. At the end of the trial there should be a documented chain of custody from manufacturing to destruction of unused drug.
7.7.2 MONITOR RESPONSIBILITIES The monitor is a person appointed by the sponsor with the purpose of performing monitoring visits at the investigator site. Usually the monitor is an employee within the clinical research department at the sponsor site. The monitor should oversee the process of a clinical trial in order to verify that the rights and well-being of human subjects are protected, the reported trial data are accurate, complete, and verifiable from source documents, and the conduct of the trial is in compliance with the applicable regulatory requirements. The monitor should also verify that any adverse events or serious adverse events have been properly recorded, reported, and followed up. When performing data verification, the monitor needs direct access to the source data. This presumes that written informed consent has been obtained from each subject included in the trial. Data verification on all patient data is a requirement, and if a subject refuses to give access to the medical records, the subject cannot participate in the trial.
50
Handbook of Non-Invasive Methods and the Skin, Second Edition
7.7.3 INVESTIGATOR RESPONSIBILITIES The ICH-GCP defines several responsibilities to be undertaken by an investigator. The following gives an overview of the responsibilities and tasks to be performed by an investigator involved in a clinical trial: Qualification: The investigator should maintain curriculum vitae, documenting appropriate education, training, and experience, including knowledge in GCP, prior to involvement in a clinical trial. Investigational medicinal product (IMP): The investigator should be thoroughly familiar with the IMP as described in the investigators brochure provided by the sponsor. Furthermore, the investigator is responsible for the accountability of the IMP at the respective site. The investigator should maintain documentation concerning the IMP delivered at the site, the inventory at the site, the use by the individual subjects, and the return to the sponsor of unused products. It must be possible to reconcile all investigational medicinal products received from the sponsor. The investigator should ensure that the IMP is used in accordance with the approved protocol, and it should be documented that doses provided to the subjects were in accordance with the protocol. The IMP should be stored as specified by the sponsor and in accordance with applicable regulatory requirements with documented temperature monitoring during the storage period. Protocol compliance: The investigator should conduct the trial in compliance with the protocol, GCP, and applicable national legislation. To confirm the agreement, the investigator should sign the protocol, or an alternative contract. An investigator should never implement any deviations or changes to the protocol without prior written agreement with the sponsor and documented approval from the ethics committee. It should be stressed that if it becomes necessary to eliminate hazards to a trial subject, implementation of the change should be performed immediately at the discretion of the investigator. However, as soon as possible, the implemented deviation or change, the reason for it, and, if appropriate, a proposed protocol amendment should be submitted to the ethics committee, the sponsor, and the competent authorities, if required by national legislation. Monitoring: The investigator should permit monitoring and auditing by the sponsor and competent authorities and should be aware that the investigator is expected to be available for answering questions during the visits.
Delegation of duties: The investigator should maintain a list of persons to whom study-related duties have been delegated. It is the responsibility of the investigator to adequately inform any involved study staff at the investigator site about the study protocol, the investigational medicinal product, and the delegated duties. Resources: The investigator should demonstrate potential for recruiting the required number of subjects and have sufficient time to properly conduct and complete the trial within the agreed time period. Furthermore, there should be an adequate number of qualified personnel and facilities to conduct the trial safely and properly. Medical care: A qualified physician, who is an investigator or subinvestigator for the trial, must be responsible for all trial-related medical decisions. During and following a subject’s participation in a trial the investigator should ensure that adequate medical care is provided to the subject for any adverse events. The investigator must inform the subject when medical care is needed for intercurrent illness of which the investigator becomes aware. If a subject wishes to withdraw from the trial, the investigator should make reasonable efforts to ascertain the reasons — while fully respecting the subject’s rights. Ethics committees: Before initiation of a clinical trial the investigator must ensure that a written dated approval letter has been obtained from the ethics committee. The ethics committee approval must cover the trial period, the informed consent forms to be used in the trial, the subject recruitment procedures, and any other information to be provided to the subjects regarding the study. The investigator should provide a copy of the investigators brochure to the ethics committee and ensure submission of any updates regarding documentation relevant for the ethics committee during conduct of the trial. Randomization and blinding: When including subjects into a clinical trial, the investigator should follow the randomization procedure of the trial. The code is only to be broken in accordance with the protocol. If the trial is blinded, unblinding of the study may occur due to serious adverse events or by accident. In these cases, the investigator should promptly document and explain the unblinding to the sponsor. Informed consent of trial subjects: A critical task is to ensure that the subjects have been properly informed verbally as well as in writing and that informed consent has been obtained before inclusion in a trial. The informed consent form and any other trial-related written information to be
Standards for Good Clinical Practice (GCP)
provided to the subjects should be approved by the ethics committee prior to subject inclusion. The document should follow the guidelines of the hospital (if any) and should be revised whenever important new information becomes available that may be relevant to the consent of the subjects. All revised versions must be approved by the ethics committee prior to implementation. Information should be given in a timely manner should new information become available. The informed consent procedure as well as specification of the information to be provided to the subjects is detailed in the ICH-GCP. Records and reports: Records should be accurate, complete, legible, and timely pertinent to the data reported to the sponsor in the case report forms (CRFs) and other required documents. The data reported on the CRFs should be derived from source documents (e.g., medical records, x-rays, ECG) and should be consistent with the source documents, and all discrepancies should be explained. Any corrections to the entered data should be dated and initialed, explained, and should not obscure the original entry, whether the entry is written or electronic. Record retention requirements: The ICH-GCP requirement is that “essential documents should be retained until at least 2 years after the last marketing application in an ICH region and until there is no pending or contemplated marketing application in an ICH region or at least 2 years have elapsed since the formal documentation of clinical development of the investigational medicinal product.” This often means in practice “forever.” It is advisable to clarify the archiving procedures with the sponsor prior to involvement in a clinical trial in order for the investigator to become aware of the expectations concerning archiving at the investigator site. Surprises often arise when the investigator realizes that the hospital does not have sufficient archiving facilities or that hospital procedures require discarding of source documents after a defined period. Precautions should be taken in collaboration with the sponsor to ensure proper retaining of the source documents for the period required by ICH-GCP. Safety reporting: It is required that all serious adverse events (SAEs) are reported immediately to the sponsor except for those SAEs that the protocol or other document identifies as not needing immediate reporting. The immediate reports should be followed promptly by detailed written reports. The investigator should be aware of applicable regulatory requirements related to
51
reporting of unexpected SAEs to the competent authorities and the ethics committee. In addition to this, adverse events or laboratory abnormalities as identified in the protocol as critical to the safety evaluations should be reported to the sponsor within the periods specified in the protocol. In case of reported deaths, the investigator should supply the sponsor and the ethics committee with any additional requested information, e.g., autopsy reports, terminal medical reports. Premature termination or suspension of a clinical trial: If a trial is suspended or prematurely terminated for any reason, the investigator should promptly inform the trial subjects and should ensure appropriate therapy and follow-up where required. Depending on who decided to terminate the trial and for which reason, there are requirements to the reporting procedures. This is detailed in the ICH-GCP. Final reports: Upon completion of the trial, the investigator, where applicable, should inform the institution; the investigator/institution should provide the ethics committee with a summary of the trial outcome and the regulatory authorities with any reports required. The role of an investigator in a clinical trial is timeconsuming, often more than expected. Although some of the investigator duties as defined by GCP may be delegated to other staff members at the investigator site, the overall responsibility for the data at the site resides with the investigator. An investigator should be able to demonstrate active involvement in a trial; otherwise, the data may be refused by the authorities. One goal of GCP was to prevent or at least reduce the occurrence of fraud. It is unknown whether this has been obtained. The number of reported fraud cases is known, but the number of unreported and unknown cases of fraud is hard to discover. Although the time from development of a new drug until marketing has been prolonged during the last 30 years due to the increased requirements concerning the amount and quality of the documentation to be used for registration purposes, it is nice to realize that the number of subjects to be included in clinical trials sufficient to reach conclusive results has been reduced. The attitude of the authorities is generally that “if it is not documented, it never happened.” If a clinical trial is conducted in compliance with GCP, it should be possible many years hence to look at the records of the work and determine easily why, how, and by whom the work was done, who was in control, what equipment and methods were used, the results obtained, any problems that were encountered, and how they were overcome. It should
52
Handbook of Non-Invasive Methods and the Skin, Second Edition
be remembered that GCP is not simply a list of restrictive rules and regulations; much is common sense. It is important that the quality concept is integrated into every stage of the laborious clinical drug development process. It must be linked to every task undertaken and be included from the very first step of a project. Quality can never be introduced retrospectively.
REFERENCES 1. World Medical Association, Declaration of Helsinki. Recommendations Guiding Physicians in Biomedical Research Involving Human Subjects. Amended by the 52nd WMA General Assembly, Edinburgh, Scotland, October 2000. Available from: http:/www.wma.net/e/ policy/b3.htm. 2. CPMP/ICH/135/95 Step 5, Note for Guidance on Good Clinical Practice (CPMP adopted July 1996). Available from http://www.eudra.org/emea.html (juni1995).
Analysis of Sensitivity, 8 Statistical Specificity, and Predictive Value of a Diagnostic Test Nicholas Lange Biometric and Field Studies Branch, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland
Martin A. Weinstock Dermatoepidemiology Unit, VA Medical Center, Roger Williams Medical Center, and Brown University, Providence, Rhode Island
CONTENTS 8.1 8.2 8.3
Introduction..............................................................................................................................................................53 Basic Concepts, Definitions, and Methods .............................................................................................................54 Limitations of Sensitivity, Specificity, and Predictive Value..................................................................................55 8.3.1 The Gold Standard.......................................................................................................................................55 8.3.2 The Clinical Context....................................................................................................................................56 8.3.3 The Artificial Dichotomy.............................................................................................................................56 8.4 Receiver Operating Characteristic (ROC) Curves ..................................................................................................56 8.4.1 Definition of the ROC Curve ......................................................................................................................56 8.4.2 Area under the ROC Curve .........................................................................................................................57 8.4.3 Correction of the ROC Curve for Verification Bias ...................................................................................58 8.4.4 Regression Methods.....................................................................................................................................60 8.5 Recommendations....................................................................................................................................................61 8.5.1 Considerations Other than Validity in Test Evaluation...............................................................................61 8.5.2 Specific Recommendations for Future Reports ..........................................................................................61 References ........................................................................................................................................................................61
8.1 INTRODUCTION In a perfect world, all of our diagnostic tests would be perfectly accurate — perfectly sensitive and specific — and 100% predictive of the disorder at issue. However, whereas perfect tests are all alike in that regard, every imperfect test is imperfect in its own way. Some will miss many cases, yet make few false diagnoses; others may miss few cases, yet falsely diagnose many. This chapter reviews appropriate methods for measuring that accuracy of tests. The example we use to illustrate these principles is the use of the aspartate aminotransferase (AST) test to diagnose hepatic fibrosis in patients receiving methotrexate therapy. Methotrexate is a very effective treatment
for severe recalcitrant psoriasis. However, its long-term use is limited by the occurrence of hepatic fibrosis, which can lead to cirrhosis of the liver and death. To avoid the adverse risk, a test is needed for the early stages of hepatic fibrosis so that the methotrexate can be stopped and clinical sequelae avoided. The present recommendation for monitoring includes periodic biopsies of the liver to determine the presence or absence of hepatic fibrosis. However, the biopsies themselves can have complications, so there is a need for a less invasive, safer procedure for determining whether hepatic fibrosis has developed. One such test is the AST test, a determination of the levels of aspartate aminotransferase in the blood. High levels suggest injury to the liver. 53
54
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 8.1 The Hepatic Fibrosis Example Hepatic Fibrosis? Elevated AST Level? Yes No Total
Yes
No
Total
15 8 23
3 24 27
18 32 50
TABLE 8.2 The Generic 2 × 2 Table True State Diagnostic Test Result Positive (d) – Negative (d) Total
Diseased (D)
Disease-Free – (D)
Total
a c a+c
b d b+d
a+b c+d n
The liver biopsy is viewed as the most accurate test for hepatic fibrosis. The AST test, either alone or in combination with other factors, is the alternative test for hepatic fibrosis used for the examples in this chapter. For simplicity, the numbers used in our examples are fictional. The reader is referred to an article by O’Conner and colleagues1 for actual observational data pertaining to the issues presented here.
8.2 BASIC CONCEPTS, DEFINITIONS, AND METHODS We begin by defining some basic terms and concepts. Many of these ideas are motivated by the generic 2 × 2 table. Table 8.1 gives an example of such a table, showing results for 50 fictitious patients cross-classified by their AST level and the presence of hepatic fibrosis. In order to understand more fully the properties of such a test, it is useful to abstract the clinical situation. Table 8.2 gives the generic form of the empirical cross-classification. In general, we use lowercase characters to denote observed quantities and events, and uppercase or Greek characters to denote theoretical quantities. For instance, p(·) is an observed probability or relative frequency, an empirical estimate of the theoretical probability P(·). In addition, the – symbol d denotes a positive test result, d a negative test – result, D a truly diseased state, and D a truly disease-free state. The observed prevalence of the disease is defined as p(D) = (a + c)/n, 0.46 or 46% in our example. Thus, the observed prevalence is a marginal probability, for this measure sums over the rows of the table, ignoring the test
result. If it can be assumed that the study population is representative of all such patient populations, then the observed quantities can serve as valid, accurate estimates of the corresponding true, theoretical quantities. In representative samples, the observed prevalence can be interpreted as an estimate of the probability that a randomly selected individual will have the disease, i.e., p(D) = P(D) on the average. If, on the other hand, the patient population under study cannot be assumed to be representative of all such patients, p(D) is a biased prevalence estimate, i.e., p(D) ≠ P(D) on the average. Selection bias would be present, for instance, if patients were either included or excluded according to criteria not accounted for in the cross-classification. In the following, however, until Section 8.3.1, we assume that the patients tested comprise a representative sample of all such patients. The observed sensitivity of the diagnostic test is defined as p(d|D) = a/(a + c), 0.65 in our example, the observed proportion of true positives among the diseased.* Similarly, the observed –– specificity of the diagnostic test is defined as p(d|D) = d/(b + d), 0.89 in our example, the observed proportion of true negatives.** Sensitivity and specificity are conditional probabilities as they include only those patients who are truly diseased or truly disease-free in their denominators, respectively. The overall accuracy of the diagnostic test is the sum of these two components weighted by the observed probabilities of the conditioning events, i.e., –– – p(d|D)p(D) + p(d|D)p(D). Inspection of the preceding conditional probabilities yields some useful relationships. A diagnostic test that is sensitive but not specific will correctly identify a large proportion of truly positive cases at the cost of labeling a large proportion of disease-free individuals as diseased. – In such a case, the proportion p(d|D) = b/(b + d) of false positives will be large. Conversely, a diagnostic test that is specific but not sensitive will correctly identify a large proportion of truly negative cases at the cost of not labeling a large proportion of diseased individuals as diseased; – the proportion p(d|D) = b/(b + d) of false negatives will be large. An extreme and unrealistic diagnostic test that declares every individual as diseased will not miss a single case and thus be completely sensitive, i.e., p(d|D) = 1, yet have a specificity of zero; similarly, a test that declares every individual as disease-free would be completely spe–– cific, i.e., p(d|D) = 1, yet have a sensitivity of zero. In the other extreme and more important case, a diagnostic – – test for which P(d|D) and P(d|D) are both zero, or both assumed to be zero, is called a gold standard for the disease in question. When using a diagnostic test that does * Read p(d|D) as “the observed proportion of patients testing positive given (‘|’) that they are truly diseased”, and similarly for other conditional probability statements. ** Note that the symbol d here denotes the number of disease-free patients that have a negative test result, not to be confused with the event d, a positive test result.
Statistical Analysis of Sensitivity, Specificity, and Predictive Value of a Diagnostic Test
not qualify as a gold standard, trade-offs are required between increasing sensitivity at the cost of decreasing specificity, and vice versa, in order to develop a test with optimal properties; see Section 8.4 on receiver operating characteristic curves for more detail on this point. The predictive value of a positive test result, or simply the positive predictive value, is defined as p(D|d) = a/(a + b). Similarly, the predictive value of a negative test – – result is defined as p(D|d) = d/(c + d). In our example, the positive predictive value is 0.83 and the negative predictive value is 0.75. Note the reversal of conditioning: given that the diagnostic test is positive, the positive predictive value is the proportion of truly positive cases, and similarly for negative test results and true negatives. A simple form of Bayes’ theorem relates the two types of conditional probabilities:
p (D d) =
p (d D) ⋅ p (D)
( ) ( )
p ( d D) ⋅ p (D) + p d D ⋅ p D
–– and similarly for p(D|d). This identity is verified trivially by inspection of Table 8.1 and Table 8.2.* In words, Bayes’ theorem applied here shows that positive predictive value =
edge of the test result. Equation 8.1 shows that this updating of prior information is given explicitly as posterior ∝ (likelihood) · (prior) with likelihood p(D|d) and constant of proportionality the denominator of Equation 8.1. This relationship can also be expressed in terms of odds ratios. Indeed, an odds ratio version of Bayes’ theorem is p (D d)
=
p ( d D) p ( D) ⋅ p dD p D
( ) ( ) ( )
p Dd
– with prior odds p(D)/p(D), 0.85 in our example, and like– lihood ratio p(d|D)/p(d|D), 5.9 in our example, and hence a posterior odds of 5.0.
8.3 LIMITATIONS OF SENSITIVITY, SPECIFICITY, AND PREDICTIVE VALUE In this section, we discuss the role of a definitive reference test, the clinical context, and the need to consider more than a simple normal/abnormal dichotomy.
8.3.1 THE GOLD STANDARD
( sensitivity ) ⋅ ( prevalence ) ( sensitivity ) ⋅ ( prevalence ) + (1 = specificity ) ⋅ (1–prevalence )
Equations 8.1 and 8.2 show exactly how the predictive value of a test depends on prevalence. Sensitivity and specificity, on the other hand, are conditional measures and thus do not depend on prevalence. The positive predictive value, p(D|d), is also called a posterior probability, as it is the probability estimate after the test result is known. It is thus an updated version of the prior probability p(D), the prevalence estimate, which is the a priori probability of disease prior to any knowl* Equation 8.1 is only the simplest, discrete form of Bayes’ theorem, however. When there are more than two possible true states (i.e., not – only D and D), the sum in Equation 8.1 gets more lengthy, until, in the limit, the state D is replaced by a parameter θ and the sum replaced by an integral over all possible parameter values, so that
p (θ d) =
55
p ( d θ) ⋅ p (θ)
∫
p ( d θ) ⋅ p (θ)
⋅ ∝ p ( d θ)
Although we do not develop this idea further here, the preceding relationship suggests Bayesian inference for the types of problems discussed in this chapter; interested readers should see the accessible article by Breslow2 on the subject.
Table 8.1 presupposes that we know who really has hepatic fibrosis, as Table 8.2 presupposes that for each person, we know whether the disease is actually present. The standard for these determinations is commonly named the gold standard or definitive reference test; i.e., the test that we assume is perfectly sensitive and specific. For hepatic fibrosis, the gold standard is generally the liver biopsy. The requirement for a gold standard raises several problems. Typically, there is some difficulty with the gold standard that motivates the search for a sensitive and specific alternative. This difficulty may also impede study of the proposed diagnostic test, such as the AST, and the gold standard in the same large group of patients. With the liver biopsy, the difficulties include a small but nonzero risk of mortality or serious morbidity as well as discomfort and expense. The gold standard itself may be an imperfect indicator of the disease being studied. This imperfection (inaccuracy) will decrease the observed sensitivity and specificity of the diagnostic test if the gold standard’s inaccuracy is independent of the diagnostic test’s result. However, inaccuracies in the gold standard may have opposite effects under other circumstances. Consider again our example of the AST test for hepatic fibrosis. If the pathologist who interpreted the liver biopsy was aware of the AST test when interpreting the histopathologic specimen, in
56
Handbook of Non-Invasive Methods and the Skin, Second Edition
equivocal cases he or she may have overdiagnosed hepatic fibrosis in the presence of an abnormal AST, and underdiagnosed the disease when the AST was normal, therefore increasing artificially the measured sensitivity and specificity of the AST test. A second type of bias would occur if, for instance, mild asymptomatic hepatitis caused both an increase in AST levels and a systematic overdiagnosis of hepatic fibrosis, and therefore an artifactual increase in measured sensitivity and specificity. Finally, it may be the case that the gold standard cannot be applied to all individuals subject to the diagnostic test due to logistical, cost, or ethical considerations, or due to comorbidity or risk of complications. If the result of the diagnostic test is used to determine which patients are subject to the gold standard evaluation, sensitivity and specificity estimates may be biased substantially; see Section 8.4.3. A gold standard may be unavailable, in which case measurements of validity, including sensitivity and specificity, become problematic. Nevertheless, validity may be tentatively assessed by measuring a surrogate endpoint known to be associated with the diagnostic test. In addition, test–retest reliability and correlation with other diagnostic tests for the disorder support validity.
normal. This approach has the advantage of producing simple, easy-to-understand results, yet it is clearly not as informative as the actual result reported on a continuous scale.
8.4 RECEIVER OPERATING CHARACTERISTIC (ROC) CURVES Changing the critical value of a test almost invariably changes both its sensitivity and specificity. It is useful, therefore, to understand the consequences of choosing particular critical values in more detail and to use this understanding to help make optimal choices. To address the artificial dichotomy problem, a receiver operating characteristic (ROC) curve for a diagnostic test is often developed.*
8.4.1 DEFINITION
OF THE
ROC CURVE
8.3.3 THE ARTIFICIAL DICHOTOMY
The ROC curve displays the range of sensitivities and specificities that are possible for a corresponding range of choices for the critical value. Let us assume that an AST value greater than a critical value z is labeled as a positive test result and that an AST value less than or equal to z is labeled as a negative test result. Table 8.3 shows the range of AST values found in our example, providing classifications of finer resolution than that shown in Table 8.1. If the total number of possible test results is equal to t(t = 13 in our example), then each row j defines a different cross-classification, a different 2 × 2 table, for each j = 1, …., t. The shaded cells in the fifth row of Table 8.3 (j = 5) define Table 8.1. Using the subscript j to indicate this 2 × 2 table, we thus find that a5 = 15, b5 = 3, c5 = 23 – 15 = 8, and d5 = 27 – 3 = 24, so that p5(d|D) = 15/(15 + 8) = 0.65, as noted previously. The test becomes less conservative as the critical AST value decreases, since a lower AST value is then required for the result to be reported as positive. In other words, sensitivity increases and specificity decreases as one scans the third and fifth columns from top to bottom. The ROC curve for these data is shown in Figure 8.1. This curve is a plot, for j = 1, …, t, of the observed sensitivities pj(d|D) = aj/(aj + cj) on the vertical axis against –– – 1 minus the observed specificities, 1 – pj (d|D) = pj (d|D) = bj /(bj + dj), on the horizontal axis. Scanning the rows in Table 8.3 from top to bottom corresponds to starting from the lower-left diagonal point (0, 0) in Figure 8.1 and
The terms sensitivity and specificity presume that a test result is simply normal (negative) or abnormal (positive). Typically, however, the test result is measured on a continuous scale and a critical value chosen to dichotomize the result. For instance, results of the AST test are reported initially in Système International units. Results above the critical value are labeled as abnormal, and results below,
* ROC curves were first developed in signal detection theory (Peterson et al.3). The “operating” end of this system (in this case, the true disease state) determines the signal sent to the “receiver” (the test result), together with noise; the task is then to determine what signal was actually sent. Each point on the curve describes the receiver’s criteria for distinguishing between signal and noise, and is called an operating position on the curve. A classic text on ROC techniques is that by Green and Swets;4 for medical applications, see, for instance, Metz.5
8.3.2 THE CLINICAL CONTEXT Sensitivity and specificity are used widely in part because these measures are independent of the prevalence of the disorder, as mentioned previously. However, this independence does not imply that sensitivity and specificity are constant. The diagnostic test may have different sensitivities and in different stages of the disease, or in different forms of the disease, and these differences may vary geographically. Similarly, a variety of illnesses, treatments, or medications may affect the performance of the test. The test’s validity may also depend on age, gender, socioeconomic status, ethnic background, and other demographic characteristics. Hence, when interpreting published measures of test validity, careful attention must be given to the clinical context in which such measures were determined. Test validity in a highly referred patient population may differ from that in a primary care setting; spatial and temporal factors may also affect validity. Replication of validity measures under diverse conditions is therefore quite helpful to establish the generalizability of results.
Statistical Analysis of Sensitivity, Specificity, and Predictive Value of a Diagnostic Test
57
TABLE 8.3 Data Table Used to Construct the Empirical ROC Curve Diseased Row (j) 1 2 3 4 5 6 7 8 9 10 11 12 13 Total
Critical Value of the AST Test Result, z >38 38 37 36 35 34 33 32 31 30 29 28 <28
Subjects With AST = z 9 2 0 4 3 2 1 0 1 1 0 0 0 23
Subjects With AST = z
34
0.6
8.4.2 AREA 0.4
0.2
0.0 0.0
0 0 2 1 0 1 3 2 1 4 9 2 2 27
Subjects With AST ≤ z 27 27 27 25 24 24 23 20 18 17 13 4 2
different tests. Clearly, a diagnostic test with desirable properties will be one whose ROC curve lies in the upperleft-hand region of the plot, where sensitivity and specificity are maximized jointly. A completely worthless test, no better than a coin toss, will have an ROC curve that lies exactly on the diagonal between the lower-left- and upper-right-hand corners.
35 Sensitivity
Subjects With AST > z 0 9 11 11 15 18 20 21 21 22 23 23 23
1.0
0.8
Disease-Free
0.2
0.4 0.6 1-specificity
0.8
1.0
FIGURE 8.1 The empirical ROC curve for the hepatic fibrosis example; from Table 8.3.
proceeding to the upper-right diagonal point (1, 1) in the plot. As indicated, a critical value of AST equal to 34 U/l shows an increase in sensitivity at no cost to specificity when compared with those values attained using a critical value of 35 U/l for this test. ROC curves help to remove the arbitrariness of clinical decision making by allowing one to investigate and control the critical value of a test to optimize decisions. ROC curves also facilitate the comparisons among
UNDER THE
ROC CURVE
Single-number summaries of ROC curves are useful for judging the validity of the test in question and also for making statistical comparisons of two or more ROC curves for competing diagnostic tests. Computing the area under an ROC curve is one way to reduce it to a single quantity. The area under an ROC curve (AUC), 0.8, for instance, can be interpreted as the probability (80%) that a randomly selected case from the diseased population will have a response to the diagnostic test worse (i.e., more abnormal, more positive) than that of a randomly selected individual from the disease-free population. For a completely uninformative test, AUC = 0.5; for a perfectly accurate test, AUC = 1.0. Estimates of the AUC can be obtained by parametric and nonparametric methods. A parametric method to compute the AUC (Dorfman and Alf7) assumes that the test results for the disease-free population are distributed as standard Gaussian (normal) with mean 0 and variance 1, and that the test results for the diseased population are distributed also as Gaussian with mean m and variance s2. Under these assumptions it has been shown that
58
Handbook of Non-Invasive Methods and the Skin, Second Edition
⎛ a AUC1 = Φ ⎜ ⎝ 1 + b2 with a =
⎞ ⎟, ⎠
μ 1 and b = σ σ
where Φ(·) is the standard Gaussian distribution function. If one makes the additional simplifying assumption that the points of the empirical ROC curve lie on the smooth curve defined by the Gaussian distribution functions, as they nearly do in our example, estimates of the parameters μ and σ2 can be obtained by routine methods; otherwise, iterative maximization routines are required. The simplifying assumption in our case yields estimates of the mean and variance of the diseased group, which are simply the sample mean and sample variance of the distribution of test responses in the diseased group after these responses have been centered and scaled by the mean and the variance of the responses in the disease-free group. Under that simplifying assumption, our example yields estimates of 2.028 and 0.971 and give an estimate AUC1 = Φ(1.455) = 0.9272. However, the maximum likelihood approach is to be preferred when the empirical ROC curve is not smooth. Maximum likelihood estimates, obtained from the program ROCFIT, courtesy of C. Metz, are 2.283 and 0.929 and give an estimate AUCML = Φ(1.443) = 0.9255. One can use the trapezoidal rule to obtain a simple nonparametric estimate of the AUC. If the ROC curve has been evaluated at t points, then the AUC can be decomposed into t – 1 trapezoidal strips (t = 13 in our example, 12 trapezoids). The area of each trapezoid is the product of the length of its base and its average height; the AUC is the sum of these areas. That is, for j = 1, …, t, the –– horizontal coordinates are xj = 1 – pj (d|D), i.e., the observed proportions of false positives at the jth level of the AST test, and the vertical coordinates are yj = pj(d|D), i.e., the observed proportions of true positives at the jth level of the AST test. The area under the ROC curve by the trapezoidal approximation is thus t –1
AUC2 =
∑(x j=1
j +1
– xj )
y j +1 + y j 2
The trapezoidal approximation to the area under the ROC curve shown in Figure 8.1 is AUC2 = 0.9147. A trapezoidal approximation such as AUC2 is smaller than the area under any smooth concave curve connecting the observed points, such as that assumed for AUC1 and AUCML. Also, the nonparametric estimate AUC2 is more sensitive to changes in the ordinates xj. It should be noted that the trapezoidal approximation reflects the fact that the results have been binned: there are ties among diseased and disease-free individuals at
identical AST values. Without such binning, the ROC curve has a staircase appearance, taking discrete jumps connected by horizontal and vertical line segments throughout. The nonparametric estimate of the AUC is thus underestimated by the binned trapezoidal approximation (see, for instance, Zweig and Campbell6). The magnitude of the underestimation depends, of course, on the number of ties and is an issue worthy of consideration when the results of a diagnostic test have been lumped into categories at a coarser resolution (e.g., definitely abnormal, probably abnormal, equivocal, probably normal, definitely normal) than those reported originally. Hanley and McNeil8 note the equivalence of the distribution of AUC2 to that of the Wilcoxon–Mann–Whitney statistic9: both measure the probability of correctly ranking a pair of disease-free and diseased patients among all such pairs in the study population. This equivalence is important, for it enables one to estimate the variance of AUC2 explicitly. Specifically, suppose that there are nD diseased and nD– disease-free patients. Let the symbol a = AUC2 denote the area under the ROC curve as approximated by the trapezoidal rule. Denote by q1 the estimated probability that two randomly selected and truly diseased patients are ranked below one randomly selected and truly disease-free patient. In addition, let q2 denote the estimated probability that one randomly selected and truly diseased patient is ranked below two randomly selected and truly disease-free patients. Then, q1 = a/(2 – a), q2 = 2a2/(1 + a), and the estimated variance of AUC2 is var [ AUC2 ] = var [ A ] =
(
1 nD nD
)
(
)
⋅ ⎡⎣ A (1 – A ) + ( nD – 1) q1 – A 2 + ( nD – 1) q2 – A 2 ⎤⎦
In our example, A = 0.9147, as stated previously. Application of the preceding approach to our example yields q1 = 0.8428 and q2 = 0.8740. Hence, the estimate of the area under the ROC curve has an estimated standard error of SE [ A ] = var [ AUC2 ] = 0.0019 = 0.0436 . A test of the null hypothesis that the true AUC is only 0.5 yields a z-score of z = (0.9147 – 0.5)/0.0436 = 9.512 using a standard Gaussian reference distribution and clearly rejects this null hypothesis.
8.4.3 CORRECTION OF THE ROC CURVE VERIFICATION BIAS
FOR
Note that ROC curve analyses may need to be corrected for verification bias in certain cases. Verification bias arises when the subjects used to assess the properties of the test have been selected in a nonrandom manner (Begg and Greenes,10 Gray et al.,11 Begg12). Proportions
Statistical Analysis of Sensitivity, Specificity, and Predictive Value of a Diagnostic Test
TABLE 8.4 The Cross-Classification of the Larger Population of Subjects from which the Verified Sample Shown in Table 8.1 was Drawn Hepatic Fibrosis? Elevated AST Level? Yes No Total
Yes
No
Total
15 16 31
3 48 51
18 64 32
of subjects selected for test validation from a larger population of subjects may differ in some systematic, nonrandom manner across the levels of test results. In general, if the selection bias favors inclusion of more subjects with high abnormal test results, then the reported sensitivity of the test is inflated artificially. Conversely, if selection bias favors inclusion of more subjects with low normal test results, then, in general, the reported specificity of the test would be inflated artificially. If selection bias favors higher proportions of subjects on both extremes of the test results, then both sensitivity and specificity are generally inflated. Verification bias can be corrected if the sampling fractions across the levels of test results are available. In our example, suppose that one knows the total number of available subjects at each level of the AST test results, only a fraction of whom have been selected for test verification. Such a possibility is shown in Table 8.4, giving the cross-classifications of a larger fictional population of subjects from which Table 8.1 was constructed. Table 8.4 is Table 8.1 with the second-row entries increased by a factor of 2. The subjects comprising Table 8.1 are a nonrandom sample from Table 8.4: those testing positive are sure to be included in the verification sample, whereas subjects with negative test results have only a 50:50 inclusion probability. Sensitivity has been inflated artificially. (Note, however, that predictive value remains unchanged; both positive and negative predictive values are not altered when a single row is multiplied by a constant.) Correction of the ROC curve analysis for verification bias employs Bayes’ theorem, as follows. Let the symbol s denote a selection indicator for positive and negative subjects, so that p(s|d) = 18/18 = 1.00 for those testing positive — sure – inclusion — and p(s|d) = 32/64 = 0.50 for those testing negative. By definition, one does not know the true disease status for those subjects not selected for verification. One knows only that fraction testing positive out of the total number of subjects, i.e., p(d) = 18/82 or 22%, and, of the selected subjects testing positive, what fraction were found to be truly diseased, i.e., p(D|d, s) = 15/18 or 83%, as given previously in Table 8.1. Similar calculations apply for those testing negative and selected. One makes
59
the additional assumption, plausible under the null hypothesis of no association, that true disease state and selection are conditionally independent given the test result. In other words, assume that P(D, s|d) = P(D|d) · P(s|d), implying P(D|d) = P(D|d, s), and similarly for – P(D|d, s). Then, by Bayes’ theorem, the observed sensitivity corrected for verification bias is
p * (d D) =
p ( D d, s ) ⋅ p ( d )
(
) ()
p ( D d, s ) ⋅ p ( d ) + p D d, s ⋅ p d
⎛ 15 ⎞ ⎛ 18 ⎞ ⎜⎝ ⎟⎠ ⋅ ⎜⎝ ⎟⎠ 18 82 = ⎛ 15 ⎞ ⎛ 18 ⎞ ⎛ 8 ⎞ ⎛ 64 ⎞ ⎜⎝ ⎟⎠ ⋅ ⎜⎝ ⎟⎠ + ⎜⎝ ⎟⎠ ⋅ ⎜⎝ ⎟⎠ 18 82 32 82 = 0.484 matching the sensitivity indicated in the complete data in Table 8.4. Similar calculations yield a corrected specificity of 0.941, also matching the specificity indicated in Table 8.4. Table 8.5 is a separate example of verification bias at a higher resolution, at a finer level of detail. Table 8.3 has been extended in Table 8.5 to include two additional columns that give the total numbers of subjects, twice the number shown in Table 8.1, and fractions selected at each level of the test results. (Results given in Table 8.1 are again indicated by shading.) The fractions of selected subjects with high AST scores are consistently greater than the fractions of selected subjects with mid-level and low AST scores: the ROC curve analysis is thus again susceptible to verification bias. To see such biases in finer detail, we need a little more notation. Denote by qk(·) that fraction of patients out of the total who have AST result at level k, and let qk(·|·, s) denote that fraction of patients giving positive or negative test results out of those included. For instance, at k = 4, qk(d) = 7/100, qk(D|D, s) – – – = 4/(4 + 1), q4(D|d, s) = 1/(4 + 1), and, again, q4(d) = 7/100. (This latter fraction is equal to q4(d): the fourth row in Table 8.5 represents the same fraction of available subjects regardless of their disease states. For each k, the – fractions qk(d) and qk(d) are identical.) Observed sensi* tivities p j (d|D) corrected for verification bias are thus j
∑ q ( D d, s ) ⋅ q ( d) k
p (d D) = ∗ j
k
k =1 t
∑ q ( D d, s ) ⋅ q ( d) k
k =1
k
,
j = 1,..., t
60
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 8.5 Augmented Data Table Used to Correct the ROC Curve for Verification Bias Verification Sample Critical Value of the AST Test Result, z
Row (j) 1 2 3 4 5 6 7 8 9 10 11 12 13 Total
Diseased Subjects With AST = z
>38 38 37 36 35 34 33 32 31 30 29 28 <28
Disease-Free
Subjects With AST > z
9 2 0 4 3 2 1 0 1 1 0 0 0 23
0 9 11 11 15 18 20 21 21 22 23 23 23
0 0 2 1 0 1 3 2 1 4 9 2 2 27
Total Patient Population
Subjects With AST ≤ z 27 27 27 25 24 24 23 20 18 17 13 4 2
Available
Fraction Selected
9 2 2 6 4 4 6 4 5 14 28 7 9 100
1.00 1.00 1.00 0.83 0.75 0.75 0.67 0.50 0.40 0.36 0.32 0.29 0.22 0.50
35 and 34 U/l demonstrate inflated sensitivities of the uncorrected results at the cost of deflated specificities.
1.0
0.8
8.4.4 REGRESSION METHODS
34 34 35
Sensitivity
Subjects With AST = z
0.6 35
0.4
Corrected Uncorrected
0.2
0.0 0.0
0.2
0.4
0.6
0.8
1.0
1-specificity
FIGURE 8.2 The corrected and uncorrected ROC curves for the hepatic fibrosis example; from Table 8.5.
yielding corrected vertical coordinates for the ROC curve. Similar weighted partial sums over disease-free patients given observed negative test results and the inclusion of these patients in the verified sample yield corrected hori–– zontal coordinates, p*j (d|D), j = 1, …, t. Figure 8.2 shows the corrected and uncorrected ROC curves deriving from Table 8.3 and Table 8.5. Although the corrected ROC curve appears better overall, the indicated AST values of
In addition to the ROC curve methods discussed here, there are several multivariable methods that should be mentioned, including ordinal regression methods (see McCullagh13) and logistic regression methods (see, for instance, Hosmer and Lemeshow14 and Hunink and Begg15). The purpose of these methods is to calculate a single number that will predict the presence or absence of the disorder better than the diagnostic test alone. In order to do so, this single number is a function of the test result as well as other factors that are correlated with disease state, such as gender, age, and other intrinsic and extrinsic effects mentioned in Section 8.3.2. The multivariable function that produces this single number is termed a prediction rule. The prediction rule can be subjected to ROC analysis, and its AUC compared with that of the original diagnostic test alone. For instance, if gender, age, etc., are correlated with the outcome of the diagnostic test, then logistic regression methods allow for the inclusion of covariates x in a multivariable model that improves predictive performance. Estimated coefficients from the fit of such a model are combined with covariate values linearly to yield a prediction rule with superior performance, i.e., higher values of p(D|d, x) than those attainable through ROC analyses that do not accommodate such factors.
Statistical Analysis of Sensitivity, Specificity, and Predictive Value of a Diagnostic Test
8.5 RECOMMENDATIONS We conclude the chapter with a brief discussion of additional elements of test evaluation and our specific recommendations for future reports.
8.5.1 CONSIDERATIONS OTHER TEST EVALUATION
THAN
VALIDITY
IN
Several considerations pertain to the assessment of tests beyond quantification of test validity. If the reasons for the observed inaccuracies are discovered, then it may be possible to enhance validity or to identify settings in which validity is greatest and poorest. It is important to assess both the reliability and validity of the test, since poor reliability may be an important source of inaccurate individual test results. It is also important to determine test performance outside of the research setting and under the conditions used by others. Issues related to associated costs and health risks must also be considered. Results of cost–benefit and cost-effectiveness analyses may be crucial to the test’s adoption in settings other than one’s own. The test itself may have an impact on the clinical setting in which it is used; determination of appropriate circumstances for the test must be made. Devising a better test may not be worthwhile if test results will not affect significant therapeutic decisions. Finally, the consequences of inaccurate test results must also be considered. If a diagnostic test is less costly yet less accurate, then the consequences of errors that may follow (e.g., suffering, disability, and death) may outweigh the savings in monetary cost. In some circumstances, a formal decision analysis that takes into account all of the preceding issues may be worthwhile.
8.5.2 SPECIFIC RECOMMENDATIONS REPORTS
FOR
FUTURE
The following 12 criteria may be useful in evaluation of future study reports that claim to validate diagnostic tests (from Weinstock16): 1. The test and other potential predictors of the disorder, the disorder to be diagnosed, and the gold standard for diagnosis of the disorder are defined with sufficient clarity and detail that an independent replication of the study may be conducted. 2. The population for which the test was validated is described. This includes the spectrum of disorder among those affected, the diagnoses among those not affected, demographic and medical data, selection criteria for the test and evaluation with the gold standard, and any relevant referral patterns.
61
3. Statistical methods are described, applied appropriately, and cited in the literature clearly. 4. The test is interpreted blindly with respect to the gold standard diagnosis and also to the proposed test. In some circumstances, it may be reasonable to assess whether attempts to blind the observers were successful. 5. The reliabilities of the test and of the gold standard are estimated and reported. 6. The sensitivity, specificity, and predictive values of the test are calculated, and where appropriate, the dependence of these characteristics on other medical or demographic factors is estimated. 7. The ROC curve for the test is presented and the area under the ROC curve is also calculated and reported, if the test results are reported on an ordinal or interval scale. 8. The relation of the test to other predictors of the disorder, including results of other tests, is assessed. The incremental value of the test is determined and prediction rules are considered. 9. If a prediction rule is suggested, it is defined precisely and its validity is evaluated and reported. If feasible, its validity is determined in a group not used to derive the prediction rule, or other statistical techniques are used to estimate its validity in such groups. The derivation of the prediction rule is described clearly, including variables considered and inclusion criteria. 10. Consideration is given to the generalizability and possible sources of bias. 11. Consideration is given to mechanisms that may account for the observed inaccuracies. 12. Consideration is given to the impact of the test in practice, i.e., effect on treatment decisions, consequences of inaccuracies, testing complications, costs, and benefits. As diagnostic testing becomes more sophisticated, more costly, and scrutinized more intensively, proper attention to quantification of validity becomes crucial to test adoption, dissemination, and appropriate use.
REFERENCES 1. O’Conner, G.T., Olmstead, E.M., Zug, K., Baughman, R.D., Beck, J.R., Dunn, J.L., and Lewandowski, J.F., Detection of hepatotoxicity associated with methotrexate therapy for psoriasis, Arch. Dermatol., 125, 1209, 1989. 2. Breslow, N., Biostatistics and Bayes (with comments), Stat. Sci., 86, 557, 1990.
62
Handbook of Non-Invasive Methods and the Skin, Second Edition
3. Peterson, W.W., Birdsall, T.G., and Fox, W.C., The theory of signal detection, Trans. IRE Prof. Group Inf. Theory, PGIT-4, 171, 1954. 4. Green, D.M. and Swets, J.A., Signal Detection Theory and Psychophysics, revised edition, Krieger, Huntington, NY, 1974. 5. Metz, C.E., Basic principles of ROC analysis, Semin. Nucl. Med., 8, 283, 1978. 6. Zweig, M.H. and Campbell, G., Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine, Clin. Chem., 39, 561, 1993. 7. Dorfman, D.D. and Alf, E., Maximum likelihood estimation of parameters of signal detection theory and determination of confidence intervals: rating method data, J. Math. Psychol., 6, 487, 1969. 8. Hanley, J.A. and McNeil, B.J., The meaning and use of the area under a receiver operating characteristic (ROC) curve, Radiology, 143, 29, 1982. 9. Colton, T., Statistics in Medicine, Little, Brown & Company, Boston, MA, 1974.
10. Begg, C.B. and Greenes, R.A., Assessment of diagnostic tests when disease verification is subject to selection bias, Biometrics, 39, 206, 1983. 11. Gray, R., Begg, C.B., and Greenes, R.A., Construction of receiver operating characteristic curves when disease verification is subject to selection bias, Med. Dec. Making, 4, 151, 1984. 12. Begg, C.B., Biases in the assessment of diagnostic tests, Stat. Med., 6, 411, 1987. 13. McCullagh, P., Regression models for ordinal data (with discussion), J. R. Stat. Soc. Ser. B, 42, 109, 1980. 14. Hosmer, D.W. and Lemeshow, S., Applied Logistic Regression, John Wiley & Sons, New York, 1989. 15. Hunink, M.G. and Begg, C.B., Diamond’s correction method: a real gem or just cubic zirconium, Med. Dec. Making, 11, 201, 1991. 16. Weinstock, M.A., Validation of a diagnostic test, Arch. Dermatol., 125, 1260, 1989.
9 Sample Size Calculation Claus Bay and Susanne Møller Mathematical Statistical Department, LEO Pharma, Ballerup, Denmark
CONTENTS 9.1 9.2 9.3
Introduction..............................................................................................................................................................63 Sample Size and Power ...........................................................................................................................................64 t-Test for Continuous Data ......................................................................................................................................64 9.3.1 Example with Independent Data .................................................................................................................64 9.3.2 Example with Paired Data...........................................................................................................................65 9.4 Chi-Square Test for Dichotomous Data ..................................................................................................................65 9.4.1 Example with Independent Data .................................................................................................................65 9.5 Discussion and Recommendations ..........................................................................................................................65 References .........................................................................................................................................................................66
9.1 INTRODUCTION Planning a scientific experiment consists of deciding on a number of equally important components: main purpose, design, variables to measure, methods of measurements, and hypotheses to test. The statistical planning is an integral part of this process. The statistical methods and tests of the specified hypotheses should be considered at an early stage in the process. An important design feature is the number of observations on a particular response variable that the experiment should produce. The consequence of not considering this in relation to the questions the experimenter wants the study to answer may very well be that the size of the study is inadequate and therefore turns out to be inconclusive. The sample size must be motivated by statistical considerations concerning difference to detect and power of statistical test (e.g., as stated in the EEC guidelines for Good Clinical Practice for Trials on Medical Products [GCP]1 in Section 9.4.3). The idea underlying the calculation of sample size of an experiment is to be able to detect, with a suitably high probability, an important difference between experimental units (e.g., treatment groups in a clinical trial), if such a difference exists. Thus, the experimenter should decide on the size of the difference he would not like to overlook (often called the minimal clinical relevant difference) and how sure he would like to be on the decision made from the study, i.e., the probability of finding a difference if it really exists (= the power) and the probability of finding a difference
when no true difference exists (= level of significance). Also, it should be decided how to measure the difference. When a number of variables are measured, it is preferable to name one variable or derived parameter to be of primary interest and consider the rest of secondary importance. The sample size is then determined based on this primary variable. If this procedure is considered infeasible, the sample size can be determined for all those variables considered of primary interest, and the study is dimensioned according to the largest of the calculated sample sizes, in this way ensuring at least the specified power in respect of all variables. The following sections will give a simple introduction to how sample size calculations can be done for most experiments and examples to illustrate the methods that will be given. Only two-sided tests are considered, since one-sided tests are rarely used in practice. It is emphasized that the formulas are only approximate, but for the vast majority of cases the gain in accuracy by applying the exact methods is negligible and need only to be considered for small sample sizes of less than about 20 for continuous data. It is beyond the scope of this chapter to go into a detailed discussion of the mathematical statistical theory from which the following formulas are derived. The interested reader is referred to Desu and Raghavarao2 for a thorough introduction to sample size methodology. Altman3 gives an excellent introduction to statistical methods in general for medical research, and in Cox and Hinkley4 the more theoretical aspects of test theory can be found. The special case of bioequivalence studies in general is dealt with in Chow and Liu.5 63
64
Handbook of Non-Invasive Methods and the Skin, Second Edition
9.2 SAMPLE SIZE AND POWER To be able to understand the calculation of sample size, it is important to understand the concept of power of a statistical test. Consider the situation where the means μ1 and μ2 of two normal distributed statistics are compared by a U-test. The test statistic |U| is the numerical value of the difference between the means divided by its standard error, distributed as N(0, 1). The conventional null hypothesis is then H0, μ1 – μ2 = 0, against the alternative hypothesis HA, μ1 – μ2 ≠ 0. The p value of the test is the probability of having observed the actual data (or more extreme data) when H0 is true, i.e., true means are equal, P = P(|U| > u1–α/2| μ1 = μ2), where u1–α/2 is the 1 – α/2 fractile of the standard normal distribution function (e.g., a = 0.05, u1–α/2 = 1.96, and P(|U| > 1.96|μ1 = μ2) = 0.05). The cutoff point for statistical significance, the significance level, denoted by a, is equal to the probability of rejecting the null hypothesis when it is true, i.e., the probability of obtaining a false positive result. This value is also referred to as a type I error. From the specification of the alternative hypothesis it appears that a whole range of values of μ1 and μ2 are contained in the set defined by HA. For every pair of values, the probability of accepting H0 when it is false is β = P(|U| < u1–α/2| μ1 – μ2 ≠ 0), i.e., the probability of a false negative result can be calculated. This value is called a type II error. The power function f(α, δ) = P(|U| > u1–α/2| μ1 – μ2 = δ) is monotonously increasing with increasing value of δ. Note that the power function is equal to α for δ = 0. This follows from f(α, 0) = P(|U| > u1–α/2| H0) = α. For a specific alternative H1 defined by μ1 – μ2 = δ0, the type II error, β, is equal to 1 – f(α,δ0). The quantity 1 – β is called the power of the test. When β is high, the power 1 – β is low, and it is unlikely that a true difference will be detected, i.e., yielding a statistically significant result of the statistical test, as the probability of failing to reject H0 is high even when H0 is false. This is obviously not desirable for the experimenter, so the aim is to have high power. Let us now examine the general situation where the test statistic X is normal distributed as N(0, Σ 20 ) under the null hypothesis, H0, and normal distributed as N(δ, Σ12 ) under the alternative hypothesis, H1, where δ < 0 or δ > 0. Often Σ 20 and Σ12 will depend on N: Σ 20 = σ 20 /N and Σ12 = σ12 /N. With the significance level a and power 1 – β, Lachin6 shows that the sample size in this case is obtained by ⎡ u1–α /2 σ 0 + u1–β σ1 ⎤ N ( α, β, δ ) = ⎢ ⎥ δ ⎣ ⎦
2
TABLE 9.1 Commonly Used u-Fractiles Significance Level, α
u1-α/2
Type II, Error, β
u1–β
1% 5% 10%
2.58 1.96 1.65
20% 10% 5%
0.84 1.28 1.65
In the following it will be shown how this simple expression can be used in the case of Student’s t-test for equality of means of normal distributed variables with unknown variance and for chi-square tests for proportions. The corresponding u-values for commonly used values of a and b appear in Table 9.1 (e.g., α = 5%, u1–α/2 = 1.96; β = 20%, u1–β = 0.84). With these and the above formula, the sample size can be determined for most experiments.
9.3 t-TEST FOR CONTINUOUS DATA Let (xi)i=1,…,N and (yi)i=1,…,N be sets of mutually independent observations from normal distributions with means m1 and m2, respectively, and variance σ2. The usual t-test statistic is T = ( x – y ) s 2 N where x and y are the averages of the xis and yis, respectively, and s is the pooled estimate of the standard deviation σ from the two samples. The difference x – y is normal distributed as N(0, 2σ2/N) under H0 and normal distributed as N(δ0, 2σ2/N) under H1. In this case the sample size of each group necessary to have power 1 – β to detect a difference equal to or greater than δ0 is ⎡ ( u1–α /2 + u1–β ) s ⎤ N = 2⎢ ⎥ δ0 ⎢⎣ ⎥⎦
2
The variances are equal under H0 and H1, and σ0 and σ1 are substituted by the estimate s. Because the true variance is unknown and replaced by an estimate of the variance, the formula is only approximate since T is t distributed. The exact value of N is obtained by substituting the u-fractiles with the corresponding t-fractiles. N is then found by iteration.2 For paired data, s is the estimate of the intrasubject standard deviation.
9.3.1 EXAMPLE
WITH INDEPENDENT
DATA
Consider a clinical study in psoriatic patients. Two topical treatments are compared in a parallel group design. The response variable is the percentage reduction in psoriatic area and severity index (PASI; Frederikson and Petterson7) from start of treatment to end of treatment. The betweenpatient standard deviation of the percentage change in PASI is assumed to be 35%-points. The clinician wants
Sample Size Calculation
65
with 80% probability (power) to detect a true difference between treatments of 10%-points. The sample size calculation gives 2
⎡ (1.96 + 0.84 ) 35 ⎤ N = 2⎢ ⎥ = 192 patients in eaach group 10 ⎦ ⎣
N= ⎡ u1–α /2 2 π 0 (1 – π 0 ) + u1–β π1 (1 – π1 ) + π 2 (1 – π 2 ) ⎤ ⎥ ⎢ π1 – π 2 ⎥⎦ ⎢⎣
2
Extensive tabulations of the sample size using the exact distribution can be found in Fleiss.8
9.3.2 EXAMPLE
WITH
PAIRED DATA
Consider an experiment in psoriatic patients where two topical treatments are compared in a right/left design. The response variable is the change in skin thickness measured by ultrasound scanning from start of treatment to end of treatment. The intrapatient standard deviation of the change in skin thickness is assumed to be 0.25 mm. The clinician wants with 80% probability to detect a true difference between treatments of 0.1 mm. The total sample size is then 2
⎡ (1.96 + 0.84 ) 0.25 ⎤ N = 2⎢ ⎥ = 98 patients 0.10 ⎣ ⎦
9.4 CHI-SQUARE TEST FOR DICHOTOMOUS DATA Let (xi)i=1,…,N and (yi)i=1,…,N be sets of mutually independent observations from binomial distributions with means (probabilities) π1 and π2, respectively. These parameters are estimated by the observed frequencies p1 and p2. The usual chi-square test statistic is used to test the null hypothesis that π1 = π2 = π0. The general formula for sample size can be applied by noting that the square root of a chi-square distributed variable with one degree of freedom is normally distributed. In this case the standard error of the difference p1 – p2 depends on the values of the proportions. The standard deviations of p1 – p2 under H0 and H1 are estimated by σ 0 = 2 π 0 (1 – π 0 ) where π 0 = ( π1 + π 2 ) 2 σ1 = π1 (1 – π1 ) + π 2 (1 – π 2 ) Using these expressions, the sample size for each group necessary to have power 1 – β against the alternative given by (π1, π2) is
9.4.1 EXAMPLE
WITH INDEPENDENT
DATA
Consider a clinical study in psoriatic patients. Two topical treatments are compared in a parallel group design. The response variable is the proportion of patients who achieve a marked improved or cleared status at the end of treatment according to the investigator’s overall assessment of treatment response. It is assumed that the average proportion of response according to this criteria is about 60%. The clinician wants with 80% probability to detect a true difference between treatments of 20%-points corresponding to π1 = 0.5 and π2 = 0.7. The sample size calculation gives N =
⎡ 1.96 ⎢ ⎣
2 0.6 (1 – 0.6 ) + 0.84 0.7 (1 – 0.7 ) + 0.5 (1 – 0.5 ) ⎤ 0.7 – 0.5
⎥ ⎦
2
= 93
patients in each group.
9.5 DISCUSSION AND RECOMMENDATIONS We have seen from the previous examples that the calculation of sample size is technically very simple and can be done with the use of a pocket calculator. The difficulties lie in specifying the different elements of the calculations. Many statisticians have experienced that making an experimenter decide on the power and the difference to detect is very difficult — thus the emphasis in the previous section to explain the concept of power of a statistical test. The fixing of power at 80 or 90% is, of course, in many cases rather arbitrary, but it is recommended to carry out experiments with at least 80% power of the statistical test, the reason being that the power can be thought of as the receiver’s (experimenter’s) risk of getting a defective item (inconclusive result). The decision about the difference to detect will depend on the purpose of the experiment. Suppose a novel drug to treat patients with a disease where no effective treatments are available, then even a small effect may be of clinical relevance. Conversely, in patients suffering from
66
Handbook of Non-Invasive Methods and the Skin, Second Edition
a disease with several efficacious treatments, a new drug may only be of interest if it shows a much better effect than placebo in a placebo-controlled trial. In any case, the difference that one can detect with a realistic number of observations depends, as we have seen, on the standard deviation of the response variable. How should the standard deviation be determined? The ideal situation is that it is well established from documented experiments with the same response variable, associated measuring technique, and comparable subject populations. However, this is often not the case. One solution is to perform a small pilot study to obtain an initial estimate of the standard deviation. If this is infeasible, it is the author’s experience that, as a rule of thumb, the biological between-individual coefficient of variation (CV) of many parameters is about 40%. Another problem that often arises is that in published data only the interindividual standard deviation is reported. This is not of much help in an experiment where differences in responses are measured within individuals, since the variance of the difference between related (paired) observations is equal to 2σ2 – 2τ, where σ is the interindividual standard deviation and τ is the (unknown) covariance. Often the covariance will be at least half the size of the interindividual variance, so in many cases, an upper limit for the relevant intrasubject variance σ 2int ra is s2, the between-individuals variance. Thus, the estimate of
the interindividual variance can, in many cases, be used as an estimate of the unknown intraindividual variance.
REFERENCES 1. CPMP Working Party on Efficacy of Medical Products, Good Clinical Practice for Trials on Medical Products in the European Community, Note for Guidance, Commission of the European Communities, 1990. 2. Desu, M.M. and Raghavarao, D., Sample Size Methodology, Academic Press, San Diego, 1990. 3. Altman, D.G., Practical Statistics for Medical Research, Chapman & Hall, London, 1991. 4. Cox, D.R. and Hinkley, D.V., Theoretical Statistics, Chapman & Hall, London, 1974. 5. Chow, S.-C. and Liu, J.-P., Design and Analysis of Bioavailability and Bioequivalence Studies, Marcel Dekker, New York, 1992. 6. Lachin, J.M., Introduction to sample size determination and power analysis for clinical trials, Controlled Clin. Trials, 2, 93, 1981. 7. Frederikson, T. and Petterson, U., Severe psoriasis: oral therapy with a new retinoid, Dermatologica, 157, 238, 1978. 8. Fleiss, J.L., Statistical Methods for Rates and Proportions, 2nd ed., John Wiley & Sons, New York, 1981.
of a Quality 10 Implementation Management System in a Contract Laboratory Working with Non-Invasive Methods Klaus-Peter Wilhelm and Jutta Hofmann proDERM Institute for Applied Dermatological Research, Schenefeld/Hamburg, Germany
CONTENTS 10.1 Introduction..............................................................................................................................................................67 10.2 What Is Quality?......................................................................................................................................................67 10.3 Quality Management and the Principles of ISO 9001:2000 ..................................................................................68 10.4 Quality in the Skin Bioengineering Laboratory .....................................................................................................69 10.5 Performance Assessment in the Skin Bioengineering Laboratory .........................................................................72 References .........................................................................................................................................................................72
10.1 INTRODUCTION
10.2 WHAT IS QUALITY?
The concept of quality has changed extensively during recent decades. The traditional thinking of quality management, which comes from the manufacturing industry, was primarily result oriented and focused on testing the output. More recent quality concepts have taken into account that “quality begins in the head” and that early analysis, planning, and arrangement of the development and production process helps to prevent subsequent errors.1,2 Prevention rather than cure is the adage, and it is applicable to all industries, including the service sector. In a climate of increasingly worldwide competition, quality management assumingly will play a key role to implement innovations rapidly and economically and to respond to the requirements of customers with appropriate stateof-the-art products and services.3 This chapter intends to illustrate the implementation of a quality management system in a dermatological contract research institute. It deals with the idea of what quality is and outlines briefly the process-based ISO 9001:2000 quality management philosophy. Additionally, the factors influencing quality in a laboratory are described and their significance highlighted. The last section provides a concept about how performance of business may be measured and evaluated using the International Organization for Standardization (ISO) standard.
It appears to be a simple question, but the answer is not as simple as expected. In the everyday context, like beauty, everybody may have their own idea of what quality is. At its simplest, quality can be defined by attributes related to the product or by its fitness for use. These categories of quality have measurable characteristics; their rating leads to the evaluation of the products’ value, and subsequently to the conclusion that one product is better or worse than another product in terms of conformity with the demands at stake. The argument not to buy might be that, for example, a car is not fast enough or that one test for mad cow disease is more sensitive than another. The complexity of what quality is nowadays is presented in the standardized ISO definition: quality is the “degree to which a set of inherent characteristics fulfills requirements.”4 Generally speaking, quality occurs when all those features of a product or service that are required by the customer are met. What does this imply for a skin bioengineering laboratory? Often the sponsor will not explicitly mention basic requirements, but silently assume their fulfillment.3 For example, a sponsor will expect as a matter of fact that a clinical study will fulfill the good clinical practice standard as well as applicable laws and statutory regulations based on these laws in the country in which the 67
68
Handbook of Non-Invasive Methods and the Skin, Second Edition
investigation takes place or that sound scientific practices will be applied. Accessory requirements such as timelines or format of report or presentation of results may be less clear. However, these characteristics offer the sponsor the opportunity to satisfy its need for individuality and the laboratory the opportunity to make its service stand out from those of the competitors.3 For a good reputation, quality is fundamental, and those who have it will find all doors open to them. If the sponsor returns and the report does not, it can, in our experience, safely be assumed that quality in terms of customer satisfaction has been achieved.
10.3 QUALITY MANAGEMENT AND THE PRINCIPLES OF ISO 9001:2000 Changing the interpretation of quality has changed the concepts of quality management as well. In the beginning, in the 1930s, there was the concept of quality control. Products were manufactured in the production line, and at the end of the line people inspected the finished products whether they complied with distinct requirements or not. If products were defective, they were thrown away or reworked. If they adhered to the specification, they were shipped to the customer. This strategy made no contribution at all to avoid or eliminate the origins of the fault. The next more effective step in the 1950s was the quality assurance approach. “Quality built in” was the motto; in-process quality control, project management, and independent audits were part of the concept. Error rates could be lowered; quality in terms of conformance with quality requirements was ensured. Competition and the “financial lid” have prompted quality initiatives extending beyond those traditional goals. Providing better quality at equivalent cost or equivalent quality at lower cost is one aim of today’s quality management. A graphical overview of the history of quality management is given in Figure 10.1. How can this purpose be achieved? As already indicated, the perception of quality depends on the customer
and on the product or service delivered. Additionally, the company’s activities to create these products or services have to be taken into account. Hence, a system to manage quality cannot be objective in the sense that every time everybody can identically reproduce the system of a company in any given situation. Each quality management system is an individual solution defined and recorded in the handbook. However, the philosophy is that the requirements for quality management systems are generic, that requirements for what a company must do to manage quality can be formulated.5 Those essential features are clarified in the ISO 9000 series of standards, which were first published in 1987, 40 years after the foundation of ISO. In December 2000, the revised ISO 9001:2000 standard was issued. Currently ISO 9000 is the most widely used series of standards for quality management worldwide.3 The standard is written without bias to manufacturing or service, but emphasizes acknowledged key factors for the management of quality.5 Key factors and some ground rules for the implementation of the ISO approach are shown in Table 10.1. Hence, quality is a management task; process management is regarded as a complement to line management and project management. In addition to the top-down approach, competent and motivated staff who accept this initiative and put it from the bottom up into practice are needed. What does managing by process mean? One of the keys to successful process management is the separation of the functional view of the organization (What has to be done?) from the operational view (How it is done?).3 A process should be stable despite organization and technology changes, whereas a procedure may be regarded as unstable, subject to change through continuous improvement. A process is generally characterized by utilizing resources to transform inputs to outputs.4 In the view that the activities of the company are a series of cooperating processes that are dependent upon each other in order to achieve quality, inputs of processes frequently are outputs from upstream processes. 3 Within a company the
Quality assurance
Quality control Verification that the requirements of quality are fulfilled Control of the final product
Quality management
System to ensure conformance with quality requirements
Corporate responsibility, design and management of the whole creation of value chain
Control at critical points during the process, modern project management and auditing to guarantee quality
Management responsibility and involvement of staff at all levels in planning and control, continual improvement
1950
FIGURE 10.1 Quality management. Brief overview of the history.
2000
Time
Implementation of a Quality Management System in a Contract Laboratory Working with Non-Invasive Methods 69
TABLE 10.1 Management of Quality: Key Factors and Ground Rules for the Implementation of the ISO Approach Key Factor
Ground Rules
Management responsibility
Demonstrate that executive management takes the responsibility for the adequacy of the quality system Develop clear beliefs and objectives and show commitment Encourage effective participation of personnel and trust
Resource management
Plan the workload of the personnel, set up role-based training and ongoing mentoring, and use reward and recognition schemes for motivation Utilize a standardized set of equipment, well-thought-out scheduling of its maintenance, quality control, and use Establish a functional facility design and set up a schedule for carrying out studies Estimate costs and use budget planning
Process approach
Define the activities necessary to obtain the desired result Manage activities and related resources as a process View and manage the series of cooperating processes as a system Keep in mind simple day-to-day management
Customer focus
Understand current and future customer needs Fulfill customer requirements and make efforts to exceed customer expectations
Performance measurement
Establish a fair way to set objectives by considering factors such as training, administrative duties, and time frames in evaluations Define pass/fail criteria or rating on organizational objectives
Evaluation and improvement of all appropriate areas of the business
Base effective decisions on the analysis of data and information
processes can be divided at least into two categories: core processes and support processes. Core processes, often called value chain processes, directly contribute to the manufacturing of the product or the service, whereas support processes ensure that the core processes can run smoothly. Typical support processes are the human resource process, the document and data control process, or the finance process. Vital for a successful process management is to clearly state appropriate objectives that can be accomplished and that allow the drawing of reliable conclusions about the current quality situation in the company.3 The companywide quality policy is one quality management measure; in addition, individual process objectives can be specified. This approach of ISO 9001:2000 is consistent with the well-known plan–do–check–act cycle, or Deming cycle, as it is often called with respect to the contribution of Edwards Deming, an American statistician, to Japan’s quality improvement efforts after World War II.6 Deming taught about problem solving and teamwork, concepts that were new to statistical quality control of his time. He recommended that business processes should be placed under a feedback loop in order to identify and change those parts of the process that needed to be improved. The diagram in Figure 10.2 illustrates the continuous process to tackle improvement in the context of business process management.4,5
Generally speaking, ISO 9001:2000 certification means “we know what we want, we say what we do, we do what we say, we prove it, and improve it.”
10.4 QUALITY IN THE SKIN BIOENGINEERING LABORATORY Skin bioengineering studies, like other clinical studies, have to respond to quality requirements of different parties. There are the rights and interests of the panelists, the interests of the sponsor and those of the consumer, and last but not least the interests of the research institute. In our view, long-term sustained success for both the sponsor and the laboratory can only be created when the legitimate interests of all these parties are respected. In order to accomplish this goal, a clear organizational structure and comprehensive quality procedures are a must. And of course, the system has to be written down, readily accessible, and user-friendly. Our approach is based on a hierarchical set of different documents, as shown in Figure 10.3. On top there is our quality policy. As a management tool this policy is basically our outward statement of commitment to respect the sponsor, the different involved parties, and our employees, and outlines briefly what we do to provide a service of high quality. It finds its practical and measurable expression, for example, in our quality
70
Handbook of Non-Invasive Methods and the Skin, Second Edition
Decide on changes and take action to continually improve process performance.
PLAN
ACT
CHECK
Monitor and measure the processes and products against quality requirements quality policy, and objectives. Report the results to decision maker.
Establish objectives and processes necessary to deliver results in accordance with customer's quality requirements, the company’s quality policy and individual process objectives.
DO
Implement the processes.
FIGURE 10.2 Plan–do–check–act cycle. Business processes should be analyzed and measured in order to identify sources of variations that cause deviations from quality requirements and to tackle improvement.
Intention and direction of the company as expressed by top management Description of the quality management system (general requirements and rules) Document that states how an activity or procedure is to be done (general instruction, not study specific)
Document that gives specific instruction for individual work steps
Quality policy
Quality manual (= QMH part A)
Standard operating procedure (SOP) (= QMH part B)
Working instruction (AA)
FIGURE 10.3 Default documents. A hierarchical set of documents is used to fully describe the quality management system. The level of detail increases from top to bottom.
objectives, which are newly defined every year after our management review. The second level is our so-called quality management handbook (QMH). It describes our organizational structure, the series of our cooperating processes, and what has to be done to handle the different factors influencing the quality of our work properly. Generally speaking, our internal standards are defined, and all elements demanded by the ISO standard and the requirements of good clinical practice are addressed.
The third and biggest part is the standard operating procedures (SOPs), also called QMH Part B. The SOPs describe how we do what we do, and therefore permit the reconstruction of events that lead to the generation of a set of data. SOPs are the further translation of guidelines and legal requirements, scientific standards, and internal standards into the real-life situation of the laboratory. SOPs are required to ensure standardization of procedures. As general instructions for organizational-, method-, and equipment-related procedures, they are not study specific.
Implementation of a Quality Management System in a Contract Laboratory Working with Non-Invasive Methods 71
The fourth level is the working instructions. These are specific instructions for individual work steps; for example, they describe how to operate our data bank to generate and print out appendices for our reports. To write down what to do and how to do it alone is clearly not enough. All documents should be accessible to staff at all times for reference and (self-)training. Auditors, internal or external, will look for evidence of training in individuals’ training records and in the knowledge of the procedures demonstrated during the interview and review of data and documentation. In our laboratory the documents are accessible via the intranet. Authorized paper copies are handed out for the equipment procedures only, and for training purposes, so-called information exemplars are circulated as paper copies. The document management system also includes regular and timely routine reviews and provision for premature review and revision when required. A security system is in place to prevent unauthorized modification and circulation of the documents, as well as to recall and archive out-of-date versions. Even with the most effectively run and comprehensive documented quality management system, of course, the final implementation is dependent on the knowledge and commitment of the personnel themselves. During our quality system audits, staff at all levels are interviewed and given the opportunity to demonstrate these qualities. It is important to emphasize that an exact recollection of precise wording is not expected, but rather an impression of understanding the quality philosophy and principles, and the ability to access detailed instructions rapidly. Furthermore, it is vital to have not only well-trained staff, but also motivated staff. This in turn demands strategies to support motivation. How important the social and technical environment of the workplace is for the personnel is often underestimated. High levels of absenteeism due to sickness, as well as high levels of staff turnover, indicate problems in the laboratory. If one keeps in mind that replacement of personnel often leads to queries and raises the potential that mistakes occur, it becomes clear how important the human factor is. Yearly interviews with every staff member are a crucial tool in our quality management system. The actual situation is discussed, and individual objectives and training programs for the next period are fixed. Finally, it is important to note that staff who recognize weak points must have the permission to define and implement improvement measures. In this context, for example, involving staff in the process of SOP writing and review can improve the daily routine at work and the acceptance of SOPs themselves. The laboratory equipment evidently directly affects the integrity of the measurement; its proper function is of prime consideration. Hence, provisions must be made for appropriate preventive maintenance. But how can reliable data and the proper function be ensured? In this context
the contribution and advantage of standards become visible. For example, the gold standard for all measurements of length was for a long time the primary meter in Paris. Today there are more modern physical methods to define the length, but the primary meter in Paris is responsible that all over the world we have the same understanding of how long a meter really is. There is a strict system to ensure this. If we are talking about numbers that our bioinstrumentation for skin provides, we often lack this kind of gold standard, but fortunately enough we do have other methods to validate new instrumentation and increase the trust in our measurements. Changing instruments requires careful measures. For example, the absolute values of skin hydration determined by capacitance measurements using different devices are not the same, even though they are based on the same measuring principles. A study comparing both the Corneometer 820 and the new version, the Corneometer 825, showed us that the data obtained with the new device were significantly lower than those of the old one.7 This led to some confusion of the user of the Corneometer, and it really took some time until it was clarified, although it was a very simple effect. In order to overcome these restrictions, sophisticated calibration methods can be used that should cover all crucial instrumental functions influencing the measurement. For the Corneometer, for example, three controls are vital: the check for equivalence of the measuring heads, for the spring’s force, and for the triggering mechanism.7 But there is not one but many laboratories, and the question is: How can the values of these different laboratories be compared? In this case, external quality assurance measures should be established. For the skin bioengineering laboratory the possibility exists to participate in interlaboratory studies. How important such comparisons are can be illustrated by the following example.8 Four laboratories measured the redness of their volunteers with Chromameters, and because they had different volunteers, there was some differences between the laboratories, as expected. However, there was one laboratory in the round-robin study that totally fell out of range. What was wrong? The Chromameter of this laboratory was not functioning right; it needed to be repaired. But the laboratory did not know about this defect until it participated in the interlaboratory comparison. Now it has implemented procedures for checking the instrument and hopefully will get earlier signs of a malfunctioning of the instrument. Nevertheless, participation in interlaboratory studies continues to be very important. Another question arises: If you have two different instruments that are built for the same objective,9 are they identical or equivalent? If there is an obvious and significant correlation between the measurements of two instruments, this correlation provides a first hint and a good overview. But in order to really check whether the
72
Handbook of Non-Invasive Methods and the Skin, Second Edition
results of two instruments are “identical,” equivalence testing with a similar statistical approach as used in bioavailability/pharmacokinetic studies should be conducted, to provide more definite answers. In the context of the skin bioengineering laboratory materials, for example, calibration standards for the pH meter, water for dilutions, or sodium dodecyl sulfate, which is often used as positive control in patch tests, also have to be considered.10 The influence that these materials can have on the quality is not always obvious. For example, a standard with poor performance can never produce a reliable result; a chemical with the expiry date long gone can be of poor quality and its normal function can no longer safely be assumed. Simple actions like comparing the new standard with the one in use or the strict consideration of expiry dates can help to prevent these defects. Additionally, a sensible selection of suppliers helps to minimize risks. In essence, the quality management system must encompass, coordinate, and manage the so-called 5 M’s — man, machine, methods, materials, and marginal conditions — in order to achieve quality.
10.5 PERFORMANCE ASSESSMENT IN THE SKIN BIOENGINEERING LABORATORY The implementation of a quality management system is not an end in itself, but is closely linked to the target of business success. To address the success of our system, different quality objectives were established. One focus has been to reduce the frequency of work returned internally between different divisions, for example, data management, reporting, and quality assurance, in order to enhance efficiency. By interviewing staff at all levels, weak points such as uncertain responsibilities, imprecise distribution and authorization of documents, or unawareness of activities that are linked closely to one another were identified and the work flow reorganized. The positive impact was evident: for studies following standard protocols, for example, the rate of return could be noticeably lowered between all personnel involved. In line with the common saying that “It is much more expensive to find a new customer than it is to service a current one” and that satisfied customers will spread good word about the company, customer satisfaction is an important quality objective. Information, for example, on the number of reclamations, the frequency of repeat clients, and the growth of the customer pool, is used to
monitor if the sponsor is satisfied. These data can be gathered without direct interaction with the sponsor, with tools that include a database of customer information and with periodic reviews of these data looking for trends. Fortunately, during the last year the number of reclamations of our reports was low, the number of customer retention, with over 80%, was definitely high, and our customer pool is continuously growing. Last but not least, the sponsor is interested in a reliable relationship with the company, and he develops it with the people who work there. Hence, a pool of highly satisfied and, in return, highly loyal customers is vital for bottomline business success along with experienced and motivated staff. Using quality objectives as a strategic tool helps to assess and improve the quality situation in the laboratory, and to stay in business successfully.
REFERENCES 1. Kamiske, G.F. and Malorny, C., Total quality management: Ein bestechendes Führungsmodell mit hohen Anforderungen und großen Chancen, Z. Führung Organ., 61, 274, 1992. 2. Feigenbaum, A.V., Total Quality Control, 3rd ed., McGraw-Hill, London, 1961. 3. Pfeifer, T., Quality Management, 3rd ed., Hanser Verlag, Munich, 2002. 4. DIN EN ISO 9000, Quality Management Systems: Fundamentals and Vocabulary, 2000. 5. DIN EN ISO 9001, Quality Management Systems: Requirements, 2000. 6. Deming, W.E., Out of the Crisis, MIT Center for Advanced Engineering Study, Cambridge, MA, 1982. 7. Wilhelm, K.-P., Possible pitfalls in hydration measurements, in Skin Bioengineering Techniques and Applications in Dermatology and Cosmetology, Elsner, P., Barel, A.O., Berardesca, E., Gabard, B., and Serup, J., Eds., Basel, Karger, 1998, p. 223. 8. Fullerton, A., Lahti, A., Wilhelm, K.-P., Perrenoud, D., and Serup, J., Interlaboratory comparison and validity study of the Minolta Chromameters CR-200 and CR300, Skin Res. Technol., 2, 126, 1996. 9. Serup, J. and Agner, T., Colorimetric quantification of erythema: a comparison of two colorimeters (Large Micro Color and Minolta Chromameter CR-200) with a clinical scoring scheme and laser-Doppler flowmetry, Clin. Exp. Dermatol., 15, 267, 1990. 10. Agner, T., Serup, J., Handlos, V., and Batsberg, W., Different skin irritation abilities of different qualities of sodium lauryl sulphate, Contact Derm., 21, 184, 1989.
11 Ethical Considerations Mikkel Noerreslet The Danish University of Pharmaceutical Sciences, Copenhagen, Denmark
Gregor B.E. Jemec Department of Dermatology, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark
CONTENTS 11.1 Introduction..............................................................................................................................................................73 11.2 Morality and Ethics .................................................................................................................................................73 11.3 Medical Ethics and Bioethics..................................................................................................................................73 11.4 Bioethics and Principalism......................................................................................................................................74 11.5 Ethical Guidelines in Research ...............................................................................................................................74 11.6 Informed Consent ....................................................................................................................................................74 11.7 Non-Invasive Skin Methods ....................................................................................................................................76 11.8 Conclusion ...............................................................................................................................................................76 References .........................................................................................................................................................................76
11.1 INTRODUCTION Ethical considerations within medicine have gained much attention during the last 50 years, and the importance of integrating these aspects in medical practice or research is today well recognized worldwide. By dealing with the general ethical considerations, and more specifically with the considerations related to the interpersonal relationship between the physician (the researcher) and the patient (the human subject), the aim of this chapter is to create an awareness of these aspects. This is so that such considerations are integrated into the design of a research project using non-invasive skin methods.
dards governing society — such as respecting others, avoiding doing harm, etc., as they are introduced to us by others. When interacting with others, the historical aspect of morality comes forth — in other words, we learn the correct moral behavior, just like children are taught how to behave correctly by their parents. Therefore, in addition to being a social construction, morality is also a historical construction.5 To introduce a third variable in the construction of morality, the culture in which we live will also shape morality. What is considered ethically correct behavior in one culture might not be ethically correct in another culture. These aspects, which underline the dynamic character of morality/ethics, therefore sharpen the necessity of a continuous attention to ethics.
11.2 MORALITY AND ETHICS When dealing with ethics it is clear that morality is a central concept, since morality refers to the traditions of beliefs about what is right and what is wrong human conduct. As mentioned by DeVries and Subedi,1 morality is socially constructed and can therefore fundamentally not be a personal policy or code — individuals as such do not exclusively create their morality by making their own rules. If this line of thought is accepted, the core parts of morality therefore already exist before an individual accepts it. We thus learn about moral codices and expectations through a process of interaction with others in society, and we come to understand the normative stan-
11.3 MEDICAL ETHICS AND BIOETHICS Within the area of medicine the philosophical reflections of what is right and wrong human conduct (morality) are an essential part of medical practice and research. Considering the position of medicine and its influence in society, this appears appropriate, as it relates to medicine’s interaction with every stage of human life and death. To ensure accordance with societal norms and uniformity among physicians and other health care professionals, the morality/ethics of medical practice and research have be explicated in rules or codes of conduct. 73
74
Handbook of Non-Invasive Methods and the Skin, Second Edition
The need for this practice has been recognized for approximately 2500 years. In ancient Greece, philosophers were preoccupied with ethical considerations of life and death, and hence the act of medicine. By swearing the Hippocratic Oath (400 B.C.), physicians of ancient Greece were obliged to act in conformity with the rules of the medical profession, and to current best practice for the benefit of the patient. In those days medical ethics was preoccupied with professional responsibilities and privileges — an aspect that has characterized medical ethics throughout thousands of years. With ideas conceived in the wake of the Nuremberg trials, which took place in 1947, the traditional physician centeredness was replaced by a set of principles, which focused more on the research subjects — the patient.4 These principles were known as the Nuremberg Code. With the Nuremberg Code the ethical tradition today known as bioethics entered the scene. Contrary to the traditions in medical ethics, where physicians exclusively discussed norms governing their own conduct, theologians and moral philosophers added to the ethical discussions in bioethics. This cross-disciplinary aspect ensured a number of different traditions and new viewpoints on ethical/moral issues. At first, this new approach to ethics within medicine had a hard time establishing itself, and it was not until the crises of medicine in the 1960s and 1970s that bioethics gained momentum and influence in health policy. This shift was partly brought on by a change in society’s conceptualization of medicine and its professions and technologies. Today medical ethics has largely been succeeded by bioethics, although the two are often used as synonyms, despite their distinctions.
11.4 BIOETHICS AND PRINCIPALISM In 1977, Tom Beauchamp and James Childress6 published a set of bioethical principles that later came to be known as principalism. The four principles, which were meant to provide a starting point for moral judgment and policy evaluations within health care, were respect for autonomy, nonmaleficence, beneficence, and justice. Acknowledging the fact that simple principles do not de facto contain sufficient content to decide ethical issues alone, and acknowledging the problem in deciding which principle is to rule, it has been emphasized that information regarding the issue should always be consulted before deciding.6 Among the four bioethical principles, respect for autonomy has gradually been established as the core principle in most Western countries, especially in America. In its most extreme form, the status of respect of autonomy is exemplified by the patient’s right to end his or her own life-sustaining treatment, despite physicians’ recommendations. Similarly, patients have the right to end
participation in any medical experiments/research projects. Tracing the development back in time, it is clear that the respect of autonomy is rooted in the liberal moral and political tradition of the importance of individual freedom and choice. As stated by Tom Beauchamp and LeRoy Walters,5 autonomy refers “to personal self-governance: personal rule of the self by adequate understanding while remaining free from controlling interference by others and from personal limitations that prevent choice” (p. 19). A patient is thus autonomous when he or she is free from external constrain and when he or she has sufficient information and understanding of this to make a choice. Another explanation for the status of the governing principle might also be due to the aspect that the principle is easy to codify and implement.2 However, it should always be remembered that the rise of patient autonomy as the general bioethical principle does not imply that the other principles, nonmaleficence, beneficence, and justice, are without relevance. Beneficence, for instance, might arise as the foremost principle in special situations, e.g., in an emergency situation where the patient is unable to express his or her own will/choices (unable to give informed consent) and where no surrogate decision maker is present, e.g., in a car crash. Physicians must in such situations act in accordance with the principle of beneficence and decide the “right” way to tackle an ethical situation from a professional point of view.
11.5 ETHICAL GUIDELINES IN RESEARCH When doing medical research that involves human subjects, certain guidelines must be complied with in most Western countries. Internationally, research projects are fundamentally governed by the Declaration of Helsinki, which the local ethical committees of the member states and as minimum comply with in their guidelines. The Declaration of Helsinki was adopted by the members of the World Medical Association in 1964 and has repeatedly been amended and ratified — the latest being in 2003.
11.6 INFORMED CONSENT In bioethics, informed consent forms one of the cornerstones, on a theoretical as well as on a practical level. This is also the case in the Helsinki Declaration, and is very evident in paragraph 22: In any research on human beings, each potential subject must be adequately informed of the aims, methods, sources of funding, any possible conflicts of interest, institutional affiliations of the researcher, the anticipated benefits and potential risks of the study and the discomfort it may entail. The subject should be informed of the right to abstain from participation in the study or to withdraw
Ethical Considerations
consent to participate at any time without reprisal. After ensuring that the subject has understood the information, the physician should then obtain the subject’s freely-given informed consent, preferably in writing. If the consent cannot be obtained in writing, the non-written consent must be formally documented and witnessed.7
The informed consent is usually given in writing, but it can also be given in other ways — as long as this is “formally documented and witnessed.” The patient and the physician are directed to keep a copy of the documentation for a certain period, so that it is available in case of any possible future disagreement. Signing the informed consent document is a formal manifestation by the patient that full information has been received and understood — and that the patient wishes to participate in the research. It is by no means an irrevocable acceptance of participation, and the patient can therefore decide to refrain from the research project at any time. If the patient wishes to withdraw, this decision must not directly have any negative influence on the patient’s current or future treatment. If the patient’s withdrawal has a negative effect, e.g., if the patient withdraws in the middle of a treatment program that should not be ended abruptly, the patient must be informed about the consequences of doing so, and a solution to the problem must be sought — a solution that embraces the wishes of the patient. It is therefore a clear violation of the patient’s autonomy if the physician pressures a patient who wants to withdraw from the research project. The basic requirements for the informed consent are that it must be: 1. 2. 3. 4.
Informative Understandable Voluntary Competent
That the consent is informative means that it must contain a full package of information. It is therefore not enough to present the patient with selected information, since this will limit the patient’s basis for making a free choice. For example, when informing the patient about risks related to participation in the research project, it is important that information on all known risks (side effects, etc.), fatal as well as minor, is given and placed in context. If the actual likelihood of the risks is known, such information should of course also be given. That the consent is understandable implies that the information provided by the physicians is understood by the patient. In the attempt to ensure this, the physician must address and recognize the capacities, perspectives, choices, and actions of the patient. In other words, the physician has to accept the personal values and beliefs, which function as the foundation for the behavior of the
75
patient. It is therefore clear that dialogue between the physician and the patient is of utmost importance, i.e., whether the patient has understood the information or not, and how the patient has understood the information. Structuring the process of informing the patient in an appropriate way will usually help this task — by drawing up a manuscript, listing the minimum information that should be given, the physician (or others involved in the research project) can keep track of the process and ensure that the necessary information is given to all patients participating in the project. When informing the patient orally, the physician must keep certain aspects of communication in mind. The information must be delivered using common and simple words, with as few medical and technical terms as possible — it must be understandable for the layperson. The art of setting oneself in the position of the receiver is a difficult task, but nonetheless a skill to strive for, since it will help the physician to show empathy with the patient and to get a feel of the patients situation when delivering information. Last but not least, it is important to allow the patient to ask questions in the process. This requires that the physician knows what information to give and is competent within the area of research. If the oral presentation of information is supplemented by written information, the language of the text must be clear, somewhat simple, brief, accurate, complete, and humane. Illustrations, pictures, diagrams, etc., will usually help to support the information given — as long as they are not misleading or taken out of a completely different context. To test whether the information is understood by the patient, it is advisable to conduct a pilot study. For the pilot study to be as representative as possible, it is important that the test persons share the same characteristics as the research population. So if the research is to test the transdermal water loss (TDWL) on patients with psoriasis, it is a good idea to test the delivering of information on a few persons with psoriasis. By encouraging these test persons to evaluate the information and the process of informing in an unstructured way, by letting them give descriptions in their own words, important aspects that might otherwise have been taken for granted or missed in the design of the research project can now be gathered.3 That the informed consent is voluntary of course implies that the patient gives his or her informed consent free from coercion. If the consent is not voluntary, this must be seen as an invasion of the patient’s autonomy. The physician must therefore not put pressure on the patient by, for instance, telling the patient that a refusal will have a negative influence on future treatment. If the patient is assisted by a third person, and pressured by this person, the physician must also stress that it is the patient’s choice to participate or not.
76
Handbook of Non-Invasive Methods and the Skin, Second Edition
The last requirement is related to the fact that the person giving his or her informed consent is competent to do so. In some instances the patient is not capable of understanding or dealing with the information provided. If the patient is disorientated, depressed, or otherwise mentally and psychologically unstable, the patient will not be competent to give informed consent. In such instances there are certain requirements related to the process of informed consent, which should be fulfilled. In other instances the patient can be a child or another person not recognized as a self-determining autonomous person. In these situations certain guidelines will also have to be followed. On the basis of this, it is worth noticing that obtaining informed consent is a dynamic process — not a definitive commitment by the patient to participate in the research forever. Information about the research procedures therefore do not end with the patient giving his or her informed consent. Rather, information should be given continuously before, during, and after the research procedures. It is furthermore important that informed consent is given deliberately and explicitly by the patient, and is not presumed consent resting on the physician’s presumption.
11.7 NON-INVASIVE SKIN METHODS From the medical point of view, non-invasive skin methods are often regarded as a harmless, effective, quick, and cheap way to gather information about the patient’s skin condition, especially if compared with the invasive skin methods. If it is assumed that the patients share this view, this might very well cause problems before, during, and after the research project. When dealing with non-invasive skin methods, it is therefore important that: •
•
Risks related to the use of the methods are assessed and communicated to the patients before they give their consent The procedures of use of the methods are explained to the patients before the informed consent is given and before the methods are actually used on the patients in the clinic
When dealing with risks, it should be remembered that all research on patients involves some degree of risk, even if the risk involved is only related to the patients’ transportation to and from the research site. The planning of the research must therefore correspond to these risks. Additional risks occur if substances are applied to the skin; e.g., cosmetic research involves independent risks of irritant or allergic reactions. Although the border between
voluntary use of such substances in everyday private use and use in scientific research may be blurred, the fact that it occurs in an experimental setting puts a special onus on the researcher to have considered risk explicitly. To comply with the governing bioethical principle, it is therefore important that the physician invest time and effort in understanding the patient perspective throughout every stage of the project — and question his or her own assumptions regarding the methods used. In addition to giving information on all relevant aspects of the research, the physician should dialogue with the patient, so that the patient as well as the physician is clear on what the research project and the patient’s participation is about.
11.8 CONCLUSION Ethics is a social construction (as well as a historical and cultural one). This implies that ethics is a dynamic concept that can and will change over time in accordance with the development of society. Ethical procedures will therefore, to a certain extent, reflect society’s attitudes toward what is right and wrong. Physicians planning to do research should therefore always consult their national/local ethical committee before conducting research on patients, to ensure that they comply with the current national guidelines. In addition to this, and in relation to the use of non-invasive skin methods, physicians should acknowledge that patients might not share their view of the excellence and safety of these methods, and should therefore engage in a dialogue with patients before, during, and after the research, to ensure a common understanding.
REFERENCES 1. DeVries R. and Subedi J., Eds., Bioethics and Society: Constructing the Ethical Enterprise, Prentice Hall, Englewood Cliffs, NJ, 1998. 2. Wolpe P.R., The triumph of autonomy in American bioethics: a sociological view, in DeVries R. and Subedi J., Eds., Bioethics and Society: Constructing the Ethical Enterprise, Prentice Hall, Englewood Cliffs, NJ, chap. 3, pp. 38–59. 3. Donovan J., et al., Improving design and conduct of randomised trials by embedding them in qualitative research: ProtectT (prostate testing for cancer and treatment) study, British Medical Journal, 325: 766–770, 2002. 4. Shuster E., The Nuremberg Code: Hippocratic ethics and human rights, Lancet, 351: 974–977, 1998. 5. Beauchamp T.L. and Walters L.R., Contemporary Issues in Bioethics, 5th ed., Wadsworth Publishing Company, 1999.
Ethical Considerations
6. Beauchamp T.L. and Childress J.F., Principles of Biomedical Ethics, 4th ed., Oxford University Press, New York, 1994.
77
7. World Medical Association, Declaration of Helsinki: Ethical Principles for Medical Research Involving Human Subjects, revised version, 2002, http://www. wma.net/e/policy/pdf/17c.pdf, downloaded July 6, 2004.
Section II Technique, Application, and Validation
Skin Surface, Epidermal Structure, and Function Clinical Photography, Surface Imaging Techniques, and Computerized Image Analysis
Aspects in Medical/Clinical 12 General Photography Nis Kentorp Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 12.1 Ethical Guidelines.................................................................................................................................................82 12.2 Getting Consent ....................................................................................................................................................82 12.3 Keeping Track of Clinical Photos ........................................................................................................................82 12.4 Scaling...................................................................................................................................................................82 12.5 Perspective ............................................................................................................................................................82 12.6 Depth of Field.......................................................................................................................................................82 12.7 Lighting.................................................................................................................................................................83 12.8 Room Space ..........................................................................................................................................................85 12.9 Background ...........................................................................................................................................................85 12.10 Patient Positions....................................................................................................................................................86 12.11 Movement .............................................................................................................................................................87 12.12 Ring Flash 8 ........................................................................................................................................................87 12.13 Optimal Exposure .................................................................................................................................................87 12.14 Cameras, Film, and Processing ............................................................................................................................87 12.15 Specialist Photography .........................................................................................................................................87 12.16 Output and Viewing ..............................................................................................................................................88 12.17 Guidelines for Publication ....................................................................................................................................88 12.18 Standardization......................................................................................................................................................88 12.19 Conclusion ............................................................................................................................................................88 References .........................................................................................................................................................................88
Clinical photography is an essential part of medical research, diagnostics, and documentation. Knowing how to take clear and consistent pre- and postoperative photos is one of the most precise ways to document a patient’s progress, demonstrate to prospective patients who are having the same procedure what results may look like, and create a visual record of your work. High-quality clinical photos are also important for use at scientific meetings, in medical journals, and for other educational purposes. Photographs can tell us a great deal about a patient’s condition at an instant in time, and serial photographs taken over a period can tell us much more of the story, about the progress of disease or response to treatment. So why is it that the clinical photographs we see so often are a disappointment? They are too dark, too light, dark shadows, lack of sharpness, color variations, important detail obscured, untidy, indeterminate scale, etc.
Undoubtedly, fully trained clinical photographers take the best medical photographs in professional studios with the full range of lighting and equipment available. But such a facility is not always available, and certainly not at all hours of the day and night or in every location in which they see patients. The clinical photographers are not concerned with photographic tricks, such as are seen in some advertising for anything from hair tonics to plastic surgery. Playing with angles of view, makeup and clothing, perspective tricks, soft focus, and lighting can exaggerate the benefits of treatments on offer. Images can be manipulated at the processing stage, and this is both easier to do and harder to detect in digital images. There is no place in clinical recording or publication for use of such photographic manipulation to misrepresent outcomes, and in some cases it would be illegal. Original films and written 81
82
Handbook of Non-Invasive Methods and the Skin, Second Edition
records should be retained, and for digital images an audit trail should be maintained. All institutions that deal with patients should require photographers that have been specially trained. This should ensure the best way to take the type of pictures that the doctors need. In general, the photographic output depends on what the doctor has asked for — the photographer will tell the patient what is necessary in order to obtain the required pictures. It is important that the photographs demonstrate the condition as clearly as possible. In order to do this, it may be necessary to tie or pin up the patient’s hair or to remove jewelry. In some cases any makeup should be removed. Sometimes the condition the photographer has been asked to record will mean that any clothing that would appear in the picture is removed. Occasionally it is necessary to use special instruments (such as nasal retractors) to help get a clear picture. If this is required in a specific case, the photographer will explain the instruments to the patient before the photographs are taken. The photographer should be informed of any concerns or reservations there might be about the photography, and be sure that the patient is given all the information that is needed.
12.1 ETHICAL GUIDELINES The photographer must show respect for the patient, and is the basis for the professional ethical responsibility concerning race, sexual orientation, gender, religion, political observance, and national or social origin. The photographer is responsible in legal and ethical aspects, and all information and material should be treated with the greatest confidentiality.
12.2 GETTING CONSENT The patients should sign a photo consent form. Ideally, it should be done even if it is not planned to share the photos in a journal or on a website. Taking this precaution may make it easier should you change your mind later on. If the patient is easily recognized, the patient consent forms are required if the clinical photos are to be published in a journal, for educational purposes, or on a website. This form is regarded as legally binding and very important, and as such, it must be as specific as possible.
12.3 KEEPING TRACK OF CLINICAL PHOTOS For expediency and to remember the patient’s name, age, and the procedure performed, write down all relevant information before taking the photographs, along with the
date on a piece of paper. This will help you keep track and may avoid delays and misinformation later on.
12.4 SCALING The best way to standardize scale is to use a lens that is marked with reproduction ratios on the barrel. The lens is set to a specific magnification ratio, and focusing is achieved by moving the whole camera back and forward. Some lenses, such as the Nikon Micro-Nikkor lenses, are factory marked for 35-mm cameras, but for digital photography, using shorter focal lengths, standardize on either the marked distances or a certain ratio by photographing a ruler and marking the lens barrel with tape or thin paint lines. With simpler cameras, this approach might be difficult or impossible, but standardization can still be achieved by maintaining the same distance from the patient. If the camera has a zoom lens, the amount of zoom should be fixed so that the image size remains the same at that distance. Some experimentation might be needed, and cameras will differ in their ability to achieve this predictably. The addition of a linear scale gives a comparative method of assessing the size of a lesion, although caution should be applied to any attempt to take measurements from photographs: some distortion is inevitable, and it must be remembered that a single photograph is a twodimensional representation of a three-dimensional subject. In photographs for forensic or medicolegal purposes, a two-dimensional right-angle scale is recommended.
12.5 PERSPECTIVE The perception of depth that perspective gives to a clinical photograph is an important attribute. The concept of diminishing image size with distance is familiar, but perspective also affects the appearance of objects and people as viewpoint changes: come too close and features can appear distorted, but take the photograph from too great a distance and the perception of depth is lost. That is why not only the image size, but also the distance from which a photograph is taken, should be standardized. For 35-mm film, a lens of around 100-mm focal length allows distortion-free images in most clinical situations, but a shorter focal length of around 50 to 60 mm will be needed for full-length photographs, or when working in a confined space.
12.6 DEPTH OF FIELD Depth of field is the amount of a subject that appears sharp in front of, and behind, the principal plane of focus. The smaller the lens aperture, the greater the depth of field, and this is particularly critical in close-up work where
General Aspects in Medical/Clinical Photography
83
FIGURE 12.1 A platform covered with cloth in the same color as the background. Used in order to get the right mid-point angle of the lower extremities.
there is considerable depth, such as when photographing the teeth. Relative aperture size is denoted on a lens by f numbers — the higher the number, the greater the depth of field, but the smaller the amount of light passed through the lens. Generally, clinical photography needs the maximum depth of field, so the flash should be bright enough to allow it to use the smallest possible aperture. The 35mm cameras should aim for at least f/16 and up to f/32 for close-up work where depth is important. Using a faster film will improve depth of field, but at the expense of some image quality.
12.7 LIGHTING The human eyes and brain have an extraordinary ability to adjust to different lighting conditions — we work quite happily in daylight, or under tungsten and fluorescent lights. Color temperature is measured in Kelvin (K) and pr. definition is the unit of thermodynamic temperature, the fraction 1/273.16 of the thermodynamic temperature of the triple point of water, which also, in fact, is the absolute temperature scale. (The conversion from Celsius to Kelvin is as follows: K = ˚C + 273). Normal daylight will be measured to approximately 5500 Kelvin — in blue sky and in shadow more. Tungsten light is approximately 3400 Kelvin. Halogen bulbs are stable in color temperature, but conventional light bulbs will be more red/yellowish down to 2900 Kelvin in daily/monthly use. Electronic flash produces color temperature similar to daylight.
FIGURE 12.2 Photograph taken with a daylight film in tungsten lighting (examination light). This shows a strong orange/yellow cast, too much contrast, and a small depth of field due to a small f number.
Cameras and films are less adaptable, particularly in mixed light, so photographs taken with a daylight film in fluorescent lighting will have a green cast, while those taken in tungsten lighting will have a strong orange/yellow cast. Indoor lighting is not bright enough, and the quality of the light is never appropriate for clinical photography. Electronic flash is ideal for providing a bright light source that is of such short duration that it is capable of “freezing” any movement. It is powerful enough to negate the effects of normal room lighting, although if you work in a room with strong sunlight coming in through the windows, you should close your blinds or curtains. Similarly, powerful examination lights or operating lights should be dimmed while taking photographs with flash. The existing light, e.g., strong light in the examine room, can be controlled by a higher shutter speed. If an existing — tungsten — light is used in analog photography, a film for tungsten light should be used, alternatively a filter (blue) for compensation. Again, this will
84
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 12.4 Electronic flash above the camera lens is used and shows shadow, shape, and texture. FIGURE 12.3 Examination light combined with electronic flash. The nonflash light is controlled with faster shutter speed.
affect the speed of the film. In digital photography this is not a problem, but the white balance has to be set. This is a normal procedure in professional digital photography. The above issues emphasize that the most secure method to produce acceptable photos is to use a flash. Modern compact cameras have a flash built into them, usually with automatic exposure control, making it difficult to produce a badly exposed photograph. This is fine for a quick record, but the lighting is too close to the lens axis to provide sufficient modeling effect to demonstrate shape and texture. It will also produce red eye — the red reflexes you see in people’s eyes in photographs taken with compact cameras. (Red-eye reduction modes use a preflash, which causes a delay in the shutter firing, so they are not ideal for moving subjects, such as lively children.) A still photograph does not have the advantage one has in a live examination of moving oneself or the patient
to look at a subject from any angle. Shape and texture can be revealed by off-axis lighting from a portable flashgun held in the hand that is not holding the camera, or attached to a bracket. Ensure that it is pointing directly at the subject to avoid the light falling off toward the edges. Lighting should as nearly as possible replicate what we are used to seeing: light from the sun, or indoor lighting from the ceiling, throwing shadows downward. Lighting should therefore come from above, and not below, in relation to the anatomical position in which we pose the patient. Getting this wrong can have strange effects. For the sake of consistency, always use the same camera, flash, lens, film, lighting, and patient position. It is axiomatic that the only variable among photographs taken to show change over time should be in the patient. Everything else should stay the same — viewpoint, positioning, lighting, color, magnification, perspective, contrast, and background. The relative importance of these properties of a photograph might vary: color might not always be important to an orthopedic surgeon, although to a
General Aspects in Medical/Clinical Photography
85
FIGURE 12.5 Ring flash is used and provides virtually shadowless lighting, with the flash tube wrapped around the camera lens.
dermatologist it might be the most salient feature of a condition. However, the principles of standardization should apply to any set of two or more photographs taken at different times. In practice, it is extremely difficult to standardize absolutely so many variables — the photographs might be taken by different people, in different rooms, using different cameras, lenses, or films, under different lighting, and from a different distance or angle to the patient. Slight variations in the film processing or digital treatment are highly likely. Different manufacturing batches of film — or in digital photography, the manufacture of the charge coupled device (CCD) chip — will have variations in sensitivity and response to color. Clearly, standardization requires a certain amount of planning, a systematic approach, adherence to protocols, and attention to detail.
12.8 ROOM SPACE Set aside a place solely for clinical photography to maximize privacy and patient comfort. Try to have the space situated near to the exam room for convenience.
FIGURE 12.6 Professional studio flash used in a studio gives maximum control of lighting and large depth of field.
12.9 BACKGROUND Try to have a seamless background, using a rich blue and green. Backgrounds with too much color can reflect onto, or throw a cast of the complementary color into, the subject, so the background should be dimmed blue/green or neutral white/gray. If you routinely photograph patients in a clinic or study, it is worth fixing up a plain, neutral background sheet or using the more or less sterile cloth or tissue, normally in “hospital” color. Professional medical photographers often prefer to use a black background, which requires several carefully placed lights to ensure that the edges and hair of the patient are not lost in the photograph. When photographing patients in a ward, it is best to place them against a plain white, gray, or dimmed blue/green color sheet. Use sticky tape, velcro, or bulldog clips to suspend the sheet and smooth it out as evenly as possible.
86
Handbook of Non-Invasive Methods and the Skin, Second Edition
12.10 PATIENT POSITIONS The patients should use the same positions, angles, and poses for both pre- and postoperative photos. Keep image size consistent in your pre- and postoperative shots. Occasionally, it might be useful to include contextual information in a photograph, such as showing the patient sitting up in bed, or attached to equipment. Similarly, if the photograph is to show dermatitis caused by an elasticity bra strap, it might be useful to include in the photograph the item of clothing responsible. In most cases, however, it is only the clinical appearance of the patient that is interesting, so all other distractions and possible influences to judgment should be excluded from the image. Use a chair with an adjustable back to pose the seated patient and exclude any parts of the furniture from the frame. Remove jewelry and makeup as far as possible. Allow patients time to replace them in private afterwards, and provide a mirror for that purpose. Ideally, all clothing should be removed from the field of view. This clearly requires some sensitivity and tact, as few people are used to exposing themselves to a camera. The patient can remove only the parts of clothing appropriate for each picture, rather than leaving them exposed any longer than is necessary. Photographs of patients hitching up their clothing can appear less dignified, less clinical, and probably cause the patient no less discomfort than if they removed those items completely. Modesty garments can be worn if it is not necessary to show the genitalia; disposable white underwear should be available, or use small sheets held up with tape or bulldog clips. Any extraneous items of clothing appearing at the edge of the frame should be cropped out afterwards. Try and provide photographic garments (underwear and paper gowns) for patients so that they can feel comfortable before taking photos and so they can move from the exam room to the clinical photography area. Discuss the pre- and postoperative photos. Avoid having the clinical photography feature be a surprise to patients. Let them know in advance (this should be part of your initial consultation) that photos are an essential part of medical treatment, confidential, and provide a good visual of the before and after treatment and surgery. If the patient is assisted, for example, a child held by a parent, care should be taken to avoid including that person in the photograph. While photographing a child on a mother’s lap, ask the mother to sit sideways, so that she is not seen behind. Keep helping hands out of frame, or be as discreet as possible. In close-up photographs, where a hand is seen retracting eyelids, lips, etc., ask the person who is doing the retraction to wear an examination glove. The distraction of a patient’s or parent’s dirty fingernail in the field of view can ruin an otherwise excellent closeup view. Hair should be swept back from the face, using
a hair band or hair clips. Hair clips can also be used to expose lesions on the scalp to the best advantage. While photographing the lateral aspect of the head, long hair should be swept over the opposite shoulder, or held up in a bun. Much confusion can arise from photographs — particularly close-ups — in which the patient’s position, or the orientation of a body part, is ambiguous. If we are going to produce a serial record of a patient, the positioning should be consistent. The convention used by most medical photographers is, wherever possible, to photograph the patient in the anatomical position. All clinical photographs should be viewed with reference to this position — the top of the photograph should always be nearest the top of the head. This works for most views, but some parts of the anatomy are best viewed from other angles, or in different positions. The arms are best photographed in extension, and are normally photographed in a horizontal position. However, the palms should still face forward, so the back of the forearm should be photographed with the patient facing away from the photographer. Close-up views can easily be misorientated, so they should normally be accompanied by an establishing view to show their precise anatomical position. A full-length photograph of a patient is one of the most difficult to achieve satisfactorily. Fortunately, it is not often necessary or particularly useful, except to show abnormal stature, posture, or body shape. The distribution of skin lesions is better shown with separate photographs from waist up, waist down, both front and back, with the arms photographed separately. For some conditions it is preferable to show the patient weight bearing. If the patient can stand unaided, he or she should do so in the anatomical position, as described above. While photographing the standing patient, remember that from your head position the patient’s feet are a long way away compared to his or her head. This can cause both perspective distortion, with the patient appearing to have tiny feet and legs, and a dramatic falloff in exposure between the head and feet. You should position the camera level with the patient’s mid-point — about waist level. For pictures taken of the lower extremities from the knees to the feet, a platform covered with cloth in the same color as the background is useful in order to get the right mid-point angle. It is difficult to produce satisfactory photographs of patients, other than infants and small children, in the prone or supine position. Such photographs require a greater working distance than can be achieved with the patient lying in bed — even if you stand on a chair or stepladder. A short working distance means using an extreme wideangle lens, which will result in unacceptable distortion. Babies, before they can sit or stand, can be photographed on a physiotherapy mat, or several layers of blanket, covered in a white sheet on the floor. Not only is this safer (they have nowhere to fall), but from a standing
General Aspects in Medical/Clinical Photography
position you can achieve a full-length view without distortion. You will need the assistance of a parent or helper to position the child for lateral views, and to gently extend the legs to be shown in full length.
12.11 MOVEMENT If the photo requires movement, the photographer should be the one to move with the camera and not the patient. Have the patient stand on a mark (you can have an X marked with black tape) and stand still.
12.12 RING FLASH 8 There is a common misconception that all medical photographs should be taken with a ring flash. The ring flash is an important tool for photographing cavities, where the shadow of a directional flash would obscure important detail. Thus, it is useful for dental photography or for deep wounds in surgical photography. A ring flash provides virtually shadowless lighting, with the flash tube wrapped around the camera lens. This type of lighting is very flat and reduces modeling. It also causes large circular reflections, which are particularly noticeable on wet surfaces such as the eye. For this type of work, a handheld flash is preferable: a powerful light source providing modeling and a single, small reflex.
12.13 OPTIMAL EXPOSURE Photographs that are too dark or too light are a bitter disappointment, and clinical photography presents greater challenges than almost any other type of photography. Clinical detail is easily lost in washed-out pale skin or underexposed dark skin. The extremes of light and dark in specialties such as dental or operative photography can disrupt the most sophisticated light metering systems. It is really worthwhile to do some tests for your camera–flash–film combination with different subject matter. Predetermined exposures that can be manually set for any magnification ratios are more reliable than automatic exposure meters, because automatic metering can be influenced by the background to cause the area of interest to be incorrectly exposed.
12.14 CAMERAS, FILM, AND PROCESSING In choosing a camera for clinical photography, the two main choices are between digital and conventional film and between compact and single-lens reflex (SLR) cameras. Your choice will depend largely upon the budget, and just how portable you need the camera to be. One of the biggest attractions of digital cameras is their immediacy.
87
After taking a photograph you can check to see whether you have a usable image, and you can download photographs onto a computer within minutes. There are hundreds of digital cameras in the market, presenting a confusing range of choices. If you want a compact camera that fits neatly in your pocket or medical bag, you will have to make compromises in the control you have over lighting, image size, and working distance. Greater control can be achieved with an SLR, which will allow you to change lenses and attach different types of flash. However, digital cameras of this type are prohibitively expensive and are used mainly by professional photographers. The alternative is to use a conventional camera and have your slides or negatives scanned. Many processing laboratories will process and scan film as a routine service, giving you a CD of all your images, sometimes in several different resolutions. If not shooting with a digital camera, try to always use Kodak, Fuji, or another recognized brand name in film. The film should be ISO 100 to 200. Remember to always use a flash.
12.15 SPECIALIST PHOTOGRAPHY Specialist clinical photography requires more specialized equipment and techniques. For dental photography, a range of mirrors and retractors are essential. Ophthalmic photography uses sophisticated slit lamp and retinal cameras, with which investigative photographs can be taken, such as fluorescent angiograms. Endoscopic photography uses cameras designed to attach to the instrument. Most instrument manufacturers supply, and can advise on, proprietary recording systems. Orthopedic patients might present handling problems, requiring special chairs and other equipment. In dermatology, some conditions can be photographed in the invisible spectrum, using infrared, ultraviolet, or fluorescence techniques. Some conditions can be better illustrated using special photographic techniques, such as infrared and ultraviolet (direct/fluorescence) photography, transillumination, dermaphotography, or photomicrography. Please take extra care to preserve patient confidentiality when handling patient recordings of all types, and prevent patient images from being seen or used by anyone other than the appropriate professionals. Patients’ wishes and written consent must be complied with at all times. Clinical photographs, video, and other images should all be considered to be a part of the clinical case notes. Photographs should be stored and presented appropriately for their use, and images for publication should be prepared according to the instructions to authors. Digital images for publication should be sized appropriately for the final reproduction size.
88
Handbook of Non-Invasive Methods and the Skin, Second Edition
12.16 OUTPUT AND VIEWING Consideration should be given to the conditions for viewing images. Clinical photographs might be seen as prints in a patient’s case notes, as digital images on a computer screen, or, more rarely these days, as projection slides. Prints should be viewed in good lighting — preferably daylight or white light fluorescent illumination (5500 K). Ask your laboratory to print photographs on glossy, rather than matte or textured paper. Prolonged exposure to ultraviolet rays will cause fading and color change, so keep prints stored out of direct light, preferably between sheets of acid-free paper. Slides are better for group viewing and, with consistent processing in a good laboratory, can give quite consistent results. Store them in archival-quality polypropylene filing sheets to avoid chemical damage. Computer viewing and projection of digital files can present problems of both quality and consistency. Use a good-quality monitor positioned away from brightly lit windows or colored surrounds. The most important factor to control in viewing is consistency. Use a good professional laboratory for films and do not leave films in the camera for months before processing — both exposed and unexposed film deteriorate over time. Keep your films in a cool place. Store large batches of unexposed film in a refrigerator, but remove film at least an hour before use to allow it to return to room temperature. Digital files for projection should be saved as highquality JPEG (.jpg) files, but most publishers prefer TIFF (.tif) files. The file size will depend upon the reproduction size, and many publishers will state exactly the file size to submit. It is normally safe to save them at a resolution of 300 dpi (dots per inch) at the final reproduction size. Check recent issues of the journal to see what is the maximum size at which images are normally reproduced (usually defined by the column width).
12.17 GUIDELINES FOR PUBLICATION If possible, avoid cropping and cutting the image. This is because most of the photos will be altered according to the specifications of the particular organization you send them to. If you are dealing with analog images, retain the
negatives or slides for yourself. Most organizations will ask for the actual photos; keep them for your own records.
12.18 STANDARDIZATION Standard representation is essential in clinical photography, to allow objective repeatability of views for comparison between different dates, or among different patients. Many clinical photographs are taken at fixed magnifications, and the views and magnification scales are recorded for each patient visit. Lighting, background, viewpoint, perspective, film, and processing should all be highly standardized and controlled. Some specialties have standard sets of views for particular groups of patients. These include dental/maxillofacial surgery, craniofacial surgery, facial plastic surgery, and intersex clinic.
12.19 CONCLUSION Familiarity with equipment and adherence to simple protocols can make all the difference between success and failure in clinical photography. A systematic approach is essential. This should extend beyond the photography itself to the handling and storage of photographs. Considerable attention should also be paid to legal and ethical issues before undertaking any clinical photography.
REFERENCES 1. Harvard Medical School, Educational articles, Harvard Medical School, New England Regional Research Center, Boston, MA, 2003. 2. UCLH Trust and Medical School, Guideline for Clinical Photography, UCLH Trust and Medical School, January 23, 2004. 3. Wentz MG, Clinical photography simplified: developing a personal set of views, 15: 211–214, 1995. 4. Ainslie G, Reilly J, The use of linear scales in the photography of skin lesions, J Audiov Media Med, 26: 15–22, 2003. 5. Aesthetic Surgery Journal, Paul Bernstein, public education, media relations, managing editor, Paul Kushner, senior director, marketing. 6. Institute of Laryngology and Otology, 2003.
of Compact Digital Camera for 13 Use Snap Photography Ken-ichiro O’goshi Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 13.1 Introduction and Background ..................................................................................................................................89 13.2 Advantages of Using a Compact Digital Camera...................................................................................................89 13.2.1 Immediacy....................................................................................................................................................89 13.2.2 Storage in Image Bank ................................................................................................................................89 13.2.3 Lasting Quality ............................................................................................................................................89 13.2.4 Image Analysis ............................................................................................................................................89 13.3 Choice of Camera ....................................................................................................................................................91 13.3.1 Easy Operation ............................................................................................................................................91 13.3.2 Fast Start-up, Shooting, and Playback ........................................................................................................91 13.3.3 Fine Picture..................................................................................................................................................91 13.3.4 Zoom ............................................................................................................................................................91 13.3.5 Mobility .......................................................................................................................................................92 13.3.6 Easy Connection to Image Banks and Tools ..............................................................................................92 13.4 Process of Saving Digital Images ...........................................................................................................................92 13.5 Teledermatology ......................................................................................................................................................93 References .........................................................................................................................................................................94
13.1 INTRODUCTION AND BACKGROUND A single-lens reflex camera (Figure 13.1 and Figure 13.2) operating with photographic films is available in almost every dermatology clinic. However, compact digital cameras (Figure 13.3 and Figure 13.4) are now popular and available at a reasonable price. These both offer convenience, and images may even be taken as 1-cm close-ups. There is no need to buy films and wait for development since digital images are available within seconds. Images may be accepted, deleted, saved, and printed out directly using a color printer with or without a personal Macintosh or Windows computer (PC).
13.2 ADVANTAGES OF USING A COMPACT DIGITAL CAMERA 13.2.1 IMMEDIACY You can check what you shot and discuss the diagnoses or treatment for some hard-to-cure cases with colleagues immediately or at a conference.
13.2.2 STORAGE
IN IMAGE
BANK
Recorded images are conveniently stored in a tiny memory card (Table 13.1) or on a hard disk of your PC. You can manage recorded images that you want to recall and search them easily using an appropriate software. Storage programs are image formats that are in widespread use on websites, for example, PNG (Portable Network Graphics), JPEG (Joint Photographic Experts Group), or GIF (Graphic Interchange Format).
13.2.3 LASTING QUALITY There is no change of image over time, and replicates from the original can be produced at any time.
13.2.4 IMAGE ANALYSIS Size and number of eruptions can be measured with appropriate. NIH (National Institutes of Health) Image is a public domain image processing and analysis program. It is very useful, but only for the Macintosh. NIH Image can be used to measure area, mean, centroid, perimeter, etc., 89
90
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 13.1 Example of a single-lens reflex camera for dermatological photography.
FIGURE 13.3 Example of a compact digital camera (rear view). FIGURE 13.2 Example of a compact digital camera (front view).
of user-defined regions of interest for research. It also performs automated particle analysis and provides tools
for measuring path lengths and angles. Results can be printed out, exported to text files, or copied to the clipboard. Furthermore, you can compare different skin conditions during the treatment when you open the files on the PC monitor at the same time.
Use of Compact Digital Camera for Snap Photography
91
without waiting for a long time to save the previous image. This will help in photography of moving objects and children.
13.3.3 FINE PICTURE
FIGURE 13.4 Card slots of a color printer. The SD/MS PRO slot is used for an SD Card™ (by SanDisk) and a Memory Stick™ or Memory Stick PRO™ (by SONY). The xD/SM slot is used for an xD-Picture Card™ (by Olympus and Fujifilm) and a SmartMedia™ (by Toshiba). The CF slot is used for a CompactFlash™ (by SanDisk).
13.3 CHOICE OF CAMERA The camera market is highly competitive and offers many brands. It is not necessary to choose an expensive singlelens reflex camera for daily clinical uses. Below are some points in choosing a good camera at a store.
13.3.1 EASY OPERATION The high-quality megapixel panel should be user-friendly and easy to operate, even under direct shiny sunlight.
A 2 or 3 megapixels camera has sufficient resolution for routine clinical photos. For graphic monitors, the screen resolution signifies the number of dots (pixels) on the entire screen. For example, a 640 × 480 pixel screen is capable of displaying 640 distinct dots on each of 480 lines, or about 300,000 pixels. This means it can print out 10,000 dots per square inch. A 3 CCD (charge coupled device) facility can make recorded images very fine and clear as well as support the printout. A CCD is a volatile memory whose semiconductors are connected so that the output of one serves as the input of the next. Its advantage is small size and speed to access. Adapted multipoint autofocus gives priority to the images’ composition and prevents off-center captured images where the subjects are out of focus. It automatically adjusts the lens system relative to the distance measured between the camera and several points in the images. The autofocus function operates with an ultrasound emitter and sensor, and the measured distance and sharpness are dependent on physical evenness of the object surface. If the camera has a builtin flash, a high-intensity white light-emitting diode (LED), you can capture dark and fine images with correct/appropriate luminance even when a close-up subject is in a lowlight condition.
13.3.4 ZOOM 13.3.2 FAST START-UP, SHOOTING,
AND
PLAYBACK
Within a second after turning the camera on, you should be able to start taking pictures. With a between-shot interval of around 2 seconds, you can record the image quickly
The cameras have not only optical lenses, but also digital zoom, which provides a maximum zooming up to about 10∞. On the other hand, if you want to capture the closeup eruptions, you can get as close as even 1 cm from the
TABLE 13.1 Recent Memory Cards for Recording of Images Memory Card Memory Stick
Standard support by Windows. Can handle from black and white to full color (16,777,216 colors); basically, can save noncompressed images; optionally, can compress 16- and 256-color forms
SD Card
Invented by SanDisk Co., Matsushita Co., and Toshiba Co.; not only possible to record images, but also suitable to record music for portable digital audio player
SmartMedia
Proposed by Toshiba Co.; small as a postage stamp; widespread use for PDA (personal digital assistant) or digital cameras; adapter is necessary to connect to computer
xD-Picture Card
Invented by Olympus Co. and Fujifilm Co. in 2002; smallest size in the world; adapter is necessary to connect to computer
CompactFlash
Proposed by SanDisk Co.; possible to connect directly to your laptop computer through an exclusive adapter; this connection system is much cheaper than any other media
92
Handbook of Non-Invasive Methods and the Skin, Second Edition
13.4 PROCESS OF SAVING DIGITAL IMAGES
FIGURE 13.5 The tulip mark of supermacro mode used for close-up photography. You can even capture 1-cm close-up images with the supermacro function.
object and capture the fine images with the setting of the supermacro function (Figure 13.5).
13.3.5 MOBILITY The smaller the camera is, the more frequently you can practice with it. If every young doctor has a reasonable camera, he or she can record at anytime and consult with senior doctors.
13.3.6 EASY CONNECTION TOOLS
TO IMAGE
BANKS
AND
When buying a digital camera, a CD-ROM (compact disc read-only memory) with software to save the images on a PC, manage them simply, and even edit them in various ways is often included. You can save the images in some recording ways like JPEG, TIFF (Tagged Image File Format), BMP (bit map), PICT, etc. (Table 13.2). There are some cables in the kit to connect the camera to a PC, a printer, or a TV. Furthermore, if you have a PictBridgecompatible printer, you can print pictures directly via the supplied universal serial bus (USB) cable by connecting the camera. This allows PC-less printing.
After buying your favorite camera, read the manual book and charge the battery for the camera first. Put the memory card inside the camera, turn it on to shooting position, and push the button to zoom in or out. Set the date and time with checking in the finder. For practice, capture any image you want and delete any image you do not want to save. Choose your setting to take pictures. Select resolution of image (pixels). If you save images as a part of the patient’s record, 2 to 4 megapixels is acceptable. Memory cards have limitations regarding numbers of photos depending on the pixels of individual images (Table 13.3). When taking close-up pictures, select the macro or supermacro setting (Figure 13.5). Generally, you do not have to use the full-auto shooting mode (Figure 13.6), except for dark objects, for example, nevi on the scalp surrounded by many hairs. Even when the object is dark in color, you can take pictures with full-auto mode with half-push capturing, and the camera can automatically do the lighting without flash. In the clinic, whenever you think it is worth saving the image, you must identify the patient. If possible, you can capture the patient’s label on the front page of his or her medical record before you take a picture of the eruptions. Before you file the images, first install the software included in the package. Connect the camera directly to your PC with the supplied USB cable, or insert a memory card directly into your PC or into the memory card slot of a printer that has a function to scan it. Using editorial software, choose “Scan to PC,” name each image, and select files where you want to keep the images. On the hard disk of your PC, make two directories, one for patient image and another for ID, and save the data in separate files. You can easily identify and search
TABLE 13.2 Common Image Formats Image Format BMP
Standard support to Windows; can handle colors from black and white to full color (16,777,216 colors); basically, can save noncompressed images; optionally, can compress 16- and 256-color forms
JPEG
Compression rate is around 1/10 to 1/100; good for compression of natural photos, but not good for computer graphics; patented by Forgent Co. in July 2002
PICT TIFF
Standard support to Mac OS of Apple Co.; can save both vector data and bit map data; can save full-color images and compress data Invented by Aldus Co. and Microsoft Co.; helpful for storage of one image in various formats; pixels, color numbers, encoding procedures in the same file
PDF
Invented by Adobe Systems Co.; possible to recall almost the same layout, fonts, decorative letters, and photos as an original text
Use of Compact Digital Camera for Snap Photography
TABLE 13.3 Number of Photos Stored on a Memory Card Depends on the Capacity of Memory Card and the Selected Resolution Memory Card Capacity Resolution Setting
16 MB
64 MB
512 MB
8M 6M 4M 2M WEB
8 10 16 33 150
35 45 70 143 645
284 363 568 1150 5200
Note: M = megapixel; MB = megabyte; WEB = the amount of mode needed to record the image to present on the website or attach to e-mails.
AUTO
93
you need to fix color tones of the image, you can correct with tone curve chosen from image adjustment of the tool. When you make a new presentation with Microsoft PowerPoint®, open a new file and set the background chosen from the menu format. Then, using Adobe Photoshop, crop or select all of the image (Ctrl + A buttons), select the image, copy it (Ctrl + C buttons), and paste it (Ctrl + V buttons) on the page of Microsoft PowerPoint. Write some explanations and save it with a title. It is really helpful to e-mail with images attached. It is an easy and fast way to discuss cases or studies via the Internet. As you send e-mails, open e-mail software such as Microsoft Outlook Express® or free mail software like Hotmail of Microsoft, click the attachment button, select the image, choosing from your recorded file, attach it within the limitation of its size due to the software, and send it with some comments at the same time. Commonly, the size limit is around 2 megabytes in an ordinary free e-mail software.
13.5 TELEDERMATOLOGY
FIGURE 13.6 The AUTO mark of full-auto setting.
not only any image you want to recall, without entering the finding mode, but also the patient’s medical record, xrays, and histopathology if included. If necessary, make backups, which means other copies of the same images, in the CD or digital versatile disc (DVD) or external hard disk for storage. CD is cheaper. An external hard disk is preferable regarding speed and reasonable in price as well. Using image processing software (Table 13.4) like Adobe Photoshop®, you can crop the area with your favorite marquee tool, for example, a rectangular one. If
Telemedicine is the practice of medicine at a distance using high-tech devices such as PC, digital camera, dermoscopy, microscopy, scanner, color printer, and telecommunication system. Teledermatology refers to the practice or clinical field in dermatology using telemedicine for diagnosis and treatment over long distance. This field is rapidly growing since many patients living in distant locations need immediate help and up-to-date medical service. Teledermatology has been used only since 1996.1 Telemedicine reports2–4 included collaborations on dermoscopy and dermatopathology.5–7 However, the direct relationship between doctor and patient remains very important. Telemedicine can be nothing but optional to a medical consultation.
TABLE 13.4 Common Software for Processing and Display of Digital Images Software
Company
Classifying
Adobe Systems
Photo retouch software
To process, crop, or fix the color tone of the images
Acrobat
Adobe Systems
Application software for PDF files
To process or crop the images
PowerPoint
Microsoft
Presentation software
To make presentation slides with text and images
Outlook Express
Microsoft
Messaging software
To e-mail with text and images
Lotus Notes
Lotus Development
Groupware
To e-mail with text and images
Photoshop
Using It
94
Handbook of Non-Invasive Methods and the Skin, Second Edition
REFERENCES 1. Wootton R. Telemedicine: a cautious welcome. Br Med J 313: 1375–1377, 1996. 2. Eedy DJ, et al. Teledermatology: a review. Br Med J 144: 696–707, 2001. 3. Goldberg DJ, et al. Digital photography, confidentiality, and teledermatology. Arch Dermatol 140: 477–478, 2004. 4. Oakley AMM, et al. Retrospective review of teledermatology in the Waikato, 1997–2002. Australasian J Dermatol 45: 23–28, 2004.
5. Shapiro M, et al. Comparison of skin biopsy trigger decisions in 49 patients with pigmented lesions and skin neoplasmas. Arch Dernatol 140: 525–528, 2004. 6. Collins K, et al. Patient satisfaction with teledermatology: quantitative results from a randomized controlled trial. J Telemed Telecare 10: 29–33, 2004. 7. Collins K, et al. General practitioners’ perceptions of asynchronous telemedicine in a randomized controlled trial. J Telemed Telecare 10: 94–98, 2004.
Image Analysis of 14 Computerized Clinical Photos Stacy S. Hawkins Unilever Research U.S., Edgewater, New Jersey
CONTENTS 14.1 Introduction..............................................................................................................................................................95 14.2 Advanced Two-Dimensional Facial Prototyping Methods .....................................................................................96 14.2.1 Facial Averaging Methods: Shape, Color, and Texture ..............................................................................96 14.2.2 Characterizing Appearance Differences in Age Groups .............................................................................96 14.2.3 Characterizing Product-Induced Changes ...................................................................................................96 14.2.4 Predictive Transforms ..................................................................................................................................97 14.3 Three-Dimensional Image Analysis ........................................................................................................................97 14.3.1 Three-Dimensional Scanning/Acquisition Systems ....................................................................................97 14.3.2 Cyberware® Scanning ..................................................................................................................................98 14.3.3 Objective Ratings from Facial Scans ..........................................................................................................98 14.3.3.1 Facial Scanning for Measuring Rapid Improvement in Texture .................................................98 14.3.3.2 Facial Scanning for Measuring Improvement in Facial Sagging................................................99 14.4 Discussion ................................................................................................................................................................99 References .........................................................................................................................................................................99
14.1 INTRODUCTION One of the key considerations for a clinical face care image analysis system is to provide accurate and objective tracking of improvement to skin health and condition with intrinsic aging and repair of extrinsic factors due to photodamage. Clinical photography and high-resolution three-dimensional volumetric measurements over a relatively small area (skin surface replicas followed by laser profilometry, or in vivo profilometry systems) have been used to objectively measure these attributes.1–8 Digital photography provides a two-dimensional representation, from which different attributes may be quantified, for example, the area covered by wrinkles in the crow’s-feet, the number of wrinkles, and wrinkle length and width (Figure 14.1). Clinical photography and archiving systems have been developed for teledermatology and semiautomated classification and diagnosis applications.9–11 One of the drawbacks to developing automated image analysis algorithms from two-dimensional photographs, however, is that color/pigmentation changes, shine, and a slight change in positioning or lighting conditions may alter the performance of the image acquisition and processing phases.
FIGURE 14.1 Sample clinical images of before and after product application of effective antiaging formulation. Several features have changed in the after image, including improved overall texture, fine lines, and appearance of sagging.
Objective and sensitive image analysis is therefore dependent on optimization of the image acquisitions systems used: • •
Calibrated, high-resolution photos Optimizing views and lighting for the study
95
96
Handbook of Non-Invasive Methods and the Skin, Second Edition
Further, changes in appearance of skin that occur with health, aging, and photodamage are multidimensional and may require global facial assessment. From a two-dimensional perspective image, many parameters that relate to these multidimensional appearance scales may be derived, including shape, color, texture, frequency, and severity of attribute. From three-dimensional range images, parameters that relate to these multidimensional appearance scales include absolute depth, shape, and volume of facial features.
14.2 ADVANCED TWO-DIMENSIONAL FACIAL PROTOTYPING METHODS 14.2.1 FACIAL AVERAGING METHODS: SHAPE, COLOR, AND TEXTURE
FIGURE 14.2 Method for calculating an average face to represent skin health of all subjects at any given week in the study.
Facial averaging, or morphing, has evolved as a powerful tool for characterizing healthy skin and understanding key drivers of attractive appearance.12–16 Previous research by Perrett et al.17–21 has demonstrated the potential of facial averaging and caricaturing for the development of aging, photodamage, and healthy skin measurement models. These facial prototypes could then be used to evaluate improvement from baseline in skin along multidimensional attributes with skin care formulations. The advantages of this technique over other image processing algorithms are that there are no a priori assumptions on areas or features that improve, and only the important features/sources of variation that are representative of the whole panel will appear in a facial average. Facial prototyping and modeling can therefore drive the development of more relevant objective measures in two and three dimensions. Recent advances in facial averaging have provided more accurate modeling capabilities for representing texture. Wavelet analysis has been used to study the variation of images at multiple scales (coarse to fine) and orientations. The addition of wavelets to the computation of facial averages and caricatures was described by Tiddeman et al.22 to preserve a more accurate statistical model of texture. Facial average computation for shape, color, and texture is a four-step process (Figure 14.2):22
Previous research has shown that by including the average shape, color, and texture to generate facial models, the resulting facial average will be characterized by assessors as belonging to the appropriate age and severity of photodamage groups.23
1. Semiautomatic calculation of over 200 facial features detected in the photographs 2. Calculation of the average facial shape 3. Morphing individual images into the average shape and blending together 4. Computing the average textural features (facial rhytides, mottled hyperpigmentation, pores) and superimposing these features onto the image from step 3
Morning — mild self-foaming wash, clarifying toner, mild cream moisturizer with SPF 15 Evening — mild cleansing pillow, clarifying toner, mild cream moisturizer with no SPF
14.2.2 CHARACTERIZING APPEARANCE DIFFERENCES IN AGE GROUPS Calibrated digital images of female subjects (ages 20 to 68) were captured from a front view. Facial averages were computed for subjects in different decade groups — 20s, 30s, 40s, 50s, and 60s — with 8 to 12 subjects in each age group. The difference in age groups represents a composite of both intrinsic and extrinsic aging factors (Figure 14.3). Additionally, composite facial models for 19 healthy appearance scales have been developed for attributes, including severity of pores, lines and wrinkles, healthy glow/color, and irritation.21
14.2.3 CHARACTERIZING PRODUCT-INDUCED CHANGES Twenty-eight female subjects, ages 18 to 48, with normal, healthy skin, signed informed consent forms to participate in this skin care regimen study.24 This regimen, recommended by the investigative dermatologist for all subjects with normal, healthy skin, was as follows:
Digital photographs were taken of the subjects at baseline and after 1 and 2 weeks of product application. The photographs of the subjects at each evaluation were averaged using the facial averaging technique described in
Computerized Image Analysis of Clinical Photos
97
appearance. Overall, the skin appeared brighter as well as more moisturized (Figure 14.4).
14.2.4 PREDICTIVE TRANSFORMS Once appropriate models for a particular attribute have been developed using facial averaging, varying changes in skin condition can be morphed onto an individual image. The average overall improvement for a panel of subjects after 12 weeks using an antiaging cream may be transformed to a different individual subject to show predicted improvement with an antiaging skin care regime (Figure 14.5).20
14.3 THREE-DIMENSIONAL IMAGE ANALYSIS FIGURE 14.3 Facial averages for subjects in their 20s, 30s, 40s, 50s, and 60s.
14.3.1 THREE-DIMENSIONAL SCANNING/ACQUISITION SYSTEMS Three-dimensional scanners have been commercially available for over 15 years and have been used for anthropometry, animation, and industrial design. The principle of operation for the different systems includes triangulation with laser sources, triangulation with pattern projection (structured light systems), interferometry and Moiré fringe pattern interpretation, stereo photogrammetry, phase measuring profilometry, and interpreting depth by shape-from-shading cues. With all of these systems, there exists a trade-off between resolution and total field of view; however, recent advances in digital cameras and computer graphics rendering techniques allows for detailed capture and quantification of shape, color, and texture. With tabletop systems (scanning occurs along one
FIGURE 14.4 Before and after averages of subjects with mostimproved skin condition after 2 weeks of a cleansing, toning, and moisturizing regimen.
Section 14.2.1, to evaluate how the skin changed over time. Digital pictures of each subject’s face were “morphed” together to generate one composite image representing all study participants. Over the course of the 2week study, this image evolved as subjects were reevaluated. In this way, multidimensional improvement to skin health representing all subjects was computed and translated into one image. After reviewing the average image/facial morphs, the investigative dermatologist saw the greatest change in the cheeks, including less visible pores and smoother skin appearance. The forehead also showed significantly improved evenness of skin tone and a smoother skin
FIGURE 14.5 Transform based on before and after images of subjects using an effective antiaging formulation on a different subject’s face.
98
Handbook of Non-Invasive Methods and the Skin, Second Edition
plane), there is some falloff in depth resolution for global face views.
14.3.2 C
YBERWARE®
SCANNING
The Cyberware® (Cyberware, Monterey, CA) threedimensional scanning technology has been used in the film industry for film special effects and computer gaming applications to create virtual scenes,25,26 for reconstructive and cosmetic surgery,27–29 and in the garment industry to ensure exact-fitting clothing and in designing ergonomically friendly products.30 The Cyberware three-dimensional facial scanner shines a safe, low-intensity laser (780 nm) on a subject to create a lighted profile. A high-quality video sensor captures the profile from two viewpoints. This stereophotogrammetric view then allows the computation of the three-dimensional geometry/range data on the face. The system can digitize thousands of these profiles (scan lines) in a few seconds. Simultaneously, a second video sensor acquires color information (color camera at an angle normal to the range line being captured). A self-contained light source illuminates the surface with cold white light. The color video camera is fitted with a filter that blocks infrared light, and the range sensor is fitted with a filter that blocks visible light. This enables both color digitizing and geometry digitizing to occur simultaneously. Cyberware facial scans provide a virtual photograph model of the face. For viewing and processing the threedimensional scans, a grid that represents a very coarse texture map of the skin may be used to rotate and move to the proper zoom level; resulting images may then be rendered as a three-dimensional virtual image with or without a color overlay (Figure 14.6). It was of interest to test the utility of the measurement of facial landmarks on these surface volume scans for quantifying texture improvement with cosmetic antiaging formulations.
FIGURE 14.6 Sample three-dimensional scan views using Cyberware facial scanner. Left: Preview grid. Middle: Texture data alone. Right: Texture data with color information overlaid result in the appearance of a smoother surface than when texture is viewed alone.
14.3.3 OBJECTIVE RATINGS
FROM
FACIAL SCANS
14.3.3.1 Facial Scanning for Measuring Rapid Improvement in Texture Facial scanning measurements were taken in a randomized, split-face, double-blind, paired comparison, homeuse test of an effective antiaging topical cosmetic formula containing glycolic acid vs. its vehicle.20 The subject population was comprised of 18 healthy female subjects, between the ages of 45 and 70 years, who provided informed consent. This institutional review board (IRB)approved study was conducted over an 8-week period. The following assessments were made at weeks 0, 4, and 8: •
•
•
Visual assessment of photodamage on the face (0 to 9 scale for fine lines and wrinkles on the eye and cheek, discrete hyperpigmentation, and overall photodamage) Self-assessment of photodamage on the face (0 to 4 scale forced-choice directed difference for overall appearance, softness, smoothness, dryness, healthy appearance, even skin color, elastic appearance, firm or loose/sagging skin, fine lines, and wrinkles) Facial scans
Although previous facial clinical studies have confirmed antiaging benefits associated with the glycolic acid cosmetic formulation, there were no clinically measurable differences in visual assessment vs. the vehicle cream in this study. However, several self-perception attributes were significantly different between treatments, favoring the glycolic acid cosmetic formulation: overall appearance, softer skin, healthier appearance, less noticeable wrinkles, and firmer skin after 4 weeks, and less loose/sagging skin after 8 weeks. The facial scanner measurements showed significant improvement of the glycolic acid cosmetic formulation over the vehicle by week 4. In the crow’s-feet area, the glycolic acid cosmetic formulation delivered a significant reduction in wrinkle depth (37% reduction) after 8 weeks, while the vehicle did not significantly reduce this parameter. In addition, the glycolic acid cosmetic formulation significantly reduced wrinkle length (longest line of crow’s-feet area; Figure 14.7) by week 4, and continuing at week 8 (16% reduction). The glycolic acid cosmetic formulation improved wrinkle length significantly greater than the vehicle (p < 0.05). Therefore, facial scanning provided added sensitivity over visual assessment for discriminating effects of a cosmetic antiaging ingredient on lines/wrinkles as early as 4 weeks, where there was no significant difference in visual assessments between treatments even by week 8, with a small panel size.
Computerized Image Analysis of Clinical Photos
99
and week 25 (p < 0.01), with the exception of the vehicle cream at week 11.
14.4 DISCUSSION
FIGURE 14.7 Facial scanning results (mean reduction in length of crow’s-feet) from an 8-week study using a glycolic acid formulation vs. its vehicle cream.
14.3.3.2 Facial Scanning for Measuring Improvement in Facial Sagging In a separate 25-week monadic application facial study, facial scanning measurements were taken to assess improvement to the appearance of texture/creasing on the cheek, including the nasolabial fold (crease from the tip of the nose to the corner of the mouth; see Figure 14.8 for measurement area).20 Measurements were taken at baseline and after 11 and 25 weeks following application of two cosmetic antiaging formulations (formulation 1 and 2) relative to a vehicle control. Both treatments provided significantly greater improvement in volume around the nasolabial fold than the vehicle treatment after 11 weeks. By week 25, volume around the nasolabial fold area increased in formulation 1 by 15%, formulation 2 by 11%, and vehicle by 7%. All treatments had significantly improved from baseline condition at week 11 (p < 0.1)
FIGURE 14.8 Facial scanning results (mean improvement in appearance of lines around the nasolabial area due to sagging) from a 25-week study using two different antiaging formulations vs. their vehicle cream.
Drivers of facial skin appearance scales are multidimensional, and the apparent skin condition of one attribute may be influenced by another or a combination of other attributes. Therefore, it was of interest to develop analysis and visualization techniques to more accurately characterize healthy and photoaging skin conditions. Twodimensional facial averaging or prototyping has been previously used in facial recognition, perception, beauty, and animation applications, and may be successfully applied to quantify appearance attributes. With two-dimensional facial averaging, many attributes may be accurately quantified, including shape, color, texture, frequency, and severity. Recent advances in digital camera technology and three-dimensional scanning systems provide more accurate, sensitive, and objective measurement opportunities of absolute depth and volume for characterizing skin conditions.
REFERENCES 1. Stiller, M.J., Bartolone, J., Stern, R., et al., Topical 8% glycolic acid and 8% L-lactic acid creams for the treatment of photodamaged skin. A double-blind vehiclecontrolled clinical trial, Arch Dermatol, 132, 631, 1996. 2. Friedman, P.M., Skover, G.R., Payonk, G., Kauvar, A.N., and Geronemus, R.G., 3D in-vivo optical skin imaging for topographical quantitative assessment of non-ablative laser technology, Dermatol Surg, 28(3), 199, 2002. 3. Beitner, H., Randomized, placebo-controlled, double blind study on the clinical efficacy of a cream containing 5% alpha-lipoic acid related to photoageing of facial skin, Br J Dermatol, 149(4), 841, 2003. 4. Nardin, P., Nita, D., and Mignot, J., Automation of a series of cutaneous topography measurements from silicon rubber replicas, Skin Res Technol, 8(2), 112, 2002. 5. Piche, E., Hafner, H.M., Hoffmann, J., and Junger, M., FOITS (fast optical in vivo topometry of human skin): new approaches to 3-D surface structures of human skin, Biomed Tech, 45(11), 317, 2000. 6. Fischer, T.W., Wigger-Alberti, W., and Elsner, P., Direct and non-direct measurement techniques for analysis of skin surface topography, Skin Pharmacol Appl Skin Physiol, 12(1-2), 1, 1999. 7. Hawkins, S.S., Wright, S.L., Barrows, J., Bartolone, J., and Weinkauf, R.L., Objective Methods to Evaluate Improvement in Photodamaged Facial Skin, paper presented at the 12th International Symposium on Bioengineering and Skin, Boston, June 25–27, 1998.
100
Handbook of Non-Invasive Methods and the Skin, Second Edition
8. Meyers, C.L., Hoyberg, K., Velez, S., Wright, S.L., and Hawkins, S.S., Quantification of skin texture improvement following washing with ultra-mild liquid cleansers, poster presented at the International Society for Skin Imaging, Washington, DC, March 2001. 9. Miyamoto, K., Takiwaki, H., Hillebrand, G.G., and Arase, S., Utilization of a high-resolution digital imaging system for the objective and quantitative assessment of hyperpigmented spots on the face, Skin Res Technol, 8(2), 73, 2002. 10. Perednia, D.A., Gaines, J.A., and Butruille, T.W., Comparison of the clinical informativeness of photographs and digital imaging media with multiple-choice receiver operating characteristic analysis, Arch Dermatol, 131(3), 292, 1995. 11. Vidmar, D.A., Cruess, D., Hsieh, P., Dolecek, Q., Pak, H., Gwynn, M., Maggio, K., Montemorano, A., Powers, J., Richards, D., Sperling, L., Wong, H., and Yeager, J., The effect of decreasing digital image resolution on teledermatology diagnosis, Telemed J, 5(4), 375, 1999. 12. Perrett, D.I., Burt, D.M., Penton-Voak, I.S., Lee, K.J., Rowland, D.A., and Edwards, R., Symmetry and human facial attractiveness, Evol Hum Behav, 20, 295, 1999. 13. Perrett, D.I., May, K., and Yoshikawa, S., Attractive characteristics of female faces: preference for non-average shape, Nature, 368, 239, 1994. 14. Bruce, V., Ness, H., Hancock, P.J.B., Newman, C., and Rarity, J., Four heads are better than one: combining face composites yields improvements in face likeness, J Appl Psychol, 87(5), 894, 2002. 15. Lee, K.J., and Perrett, D.I., Manipulation of colour and shape information and its consequence upon recognition and best-likeness judgments, Perception, 29(11), 1291, 2000. 16. Calder, A.J., Rowland, D., Young, A.W., Nimmo-Smith, I., Keane, J., and Perrett, D.I., Caricaturing facial expressions, Cognition, 76(2),105, 2000. 17. Perrett, D.I., Penton-Voak, I.S., Little, A.C., Tiddeman, B.P., Burt, D.M., Schmidt, N., Oxley, R., Kinloch, N., and Barrett, L., Facial attractiveness judgements reflect learning of parental age characteristics, Proc R Soc Lond B Biol Sci, 269(1494), 873, 2002.
18. Burt, D.M. and Perrett, D.I., Perceptual asymmetries in judgements of facial attractiveness, age, gender, speech and expression, Neuropsychologia, 35(5), 685, 1997. 19. Burt, D.M. and Perrett, D.I., Perception of age in adult Caucasian male faces: computer graphic manipulation of shape and colour information, Proc R Soc Lond B Biol Sci, 259(1355), 137, 1995. 20. Hawkins, S.S., Perrett, D.I., Tiddeman, B., et al., Novel approaches in texture measurement for cosmetic antiaging evaluation, in Proceedings of the 22nd IFSCC Congress, Edinburgh, Scotland, September 2002, p. 317. 21. Hawkins, S.S., Perrett, D.I., Burt, D.M., Rowland, D.A., and Murahata, R.I., Prototypes of facial attributes developed through image averaging techniques, Int J Cos Sci, 21, 159, 1999. 22. Tiddeman, B., Burt, D.M., and Perrett, D.I. Prototyping and transforming facial textures for perception research, IEEE Comput Graphics Applications, 21, 42, 2001. 23. Hawkins, S.S., Andrew, J., Weinkauf, R.L., Tiddeman, B.P., Payne, K.R., Burt, D.M., Perrett, D.I., and Murahata, R.I. Quantification of Facial Texture by Image Averaging Techniques, paper presented at the International Society for Skin Imaging, Washington, DC, March 2001. 24. Hawkins, S.S., Subramanyan, K., Liu, D., and Bryk, M., Dermatologic Ther, 17(1 Suppl), in press, 2004. 25. New Cyberware Software Makes 3D Scans Ideal for Animation, Virtual Reality, Cyberware press release, Monterey, CA, July 28, 1992. 26. Cyberware Scanner Used to Digitize Michelangelo’s David, Cyberware press release, Monterey, CA, July 14, 2000. 27. O’Grady, K.F. and Antonyshyn, O.M., Facial asymmetry: three-dimensional analysis using laser surface scanning, Plast Reconstr Surg, 104(4), 928, 1999. 28. Bush, K. and Antonyshyn, O., Three-dimensional facial anthropometry using a laser surface scanner: validation of the technique, Plast Reconstr Surg, 98(2), 226, 1996. 29. Hubbs, L., Ed. Taking a Closer Look at Post-Surgical Skin Changes, Skin Aging, October, 100, 2003. 30. The World’s First Whole Body Scanners Bring True Human Forms to Computer Graphics, Cyberware press release, Monterey, CA, May 11, 1995.
Lens — Non-Invasive Oil 15 Magnifying Immersion Examination of the Skin H. Irving Katz and Jane S. Lindholm Minnesota Clinical Study Center, Fridley, Minnesota
CONTENTS 15.1 Introduction............................................................................................................................................................101 15.2 Objective ................................................................................................................................................................102 15.3 Methodological Principals .....................................................................................................................................102 15.4 Sources of Error.....................................................................................................................................................105 15.5 Correlation with Other Methods ...........................................................................................................................106 15.6 Recommendations..................................................................................................................................................106 References .......................................................................................................................................................................106
15.1 INTRODUCTION Visualization of the skin is perhaps the most important part of the dermatologic examination. The observation of a cutaneous sign reflects a pathologic change in the skin due to a derangement within the epidermis, dermis, and/or subcutaneous tissue. Recognition of cutaneous findings allows a definitive or differential diagnosis of a dermatological disorder. Using naked-eye inspection the experienced observer detects many visible gross morphologic features, such as size, shape, color, or type of lesion. However, some subtle or ambiguous diagnostic features require further amplification of the finding. The use of skin surface magnification techniques assists an observer in visualizing such enigmatic findings. Skin surface magnification or surface microscopy is not a new method to examine changes occurring on or within the skin. Hinselmann, Goldman, Gilje, and O’Leary, along with many others since, have reported the usefulness of skin surface microscopy methods for a variety of dermatologic conditions.1–12 In addition the past decade surface microscopy has been to study and differentiate benign and malignant pigmented melanocytic neoplasms.13–19 Extremes of skin surface magnification that are available range from just slightly more than life size to hundreds of times the size of the actual image viewed, as summarized in Table 15.1 Although high-power resolution is obtainable, such applications are not within the reach of most practitioners. Therefore, only low-power methods
TABLE 15.1 Skin Surface Microscopy Techniques Low-power surface magnification (generally <10×) Simply magnifying lens Magnifying loupes Otoscopes Ophthalmoscopes Commercial skin scopes Medium-power magnification (generally <40×) Monocular surface microscope Colposcope Surgical microscope High-power magnification (up to 1000×) Video microscopy systems
of skin surface magnification, such as simple lens systems, are within the scope of practice of most observers. Low-power skin surface magnification is considered less than 10× magnification for the purposes of this chapter. We have found that a good-quality single lens, handheld 8× magnifier with adequate incident illumination is a practical technique, albeit not the most accomplished alternative, to assist the astute observer in the recognition of subtle morphologic features. We also have available a Wild M650 binocular surgical stereomicroscope with a 100- to 250-mm lens system and automatic 35-mm camera, which was used to obtain the photographs in this chapter (Wild Heerbrugg, Ltd., Heerbrugg, Switzerland).20 Further enhancement of the observer’s ability to detect obscure morphologic findings occurs by pretreatment of 101
102
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 15.2 Selected Clinical Findings Amplified by Enhanced Skin Surface Magnification Terminal Hair Shaft Alterations 1. Varying lengths of hairs (trichotillomania, dystrophic hair shafts, androgenetic alopecia) 2. Varying diameter of hairs (androgenetic alopecia) 3. Black dot hairs (tinea capitis) 4. Fractured hairs (dystrophic hair shafts, trichotillomania, tinea capitis) 5. Dystrophic hairs (dystrophic hair shafts) 6. Casts (inflammatory dermatoses) 7. Presence of foreign material (pediculosis, trichomycosis axillaris) 8. Exclamation point hairs (alopecia areata) Upper Epidermis (Stratum Corneum) Alterations 1. Altered dermatoglyphics (neoplasia, atrophy, forms of sclerosis, scar formation) 2. Distorted or missing follicular orifices (neoplasia, scar, forms of sclerosis) 3. Follicular plugging (discoid lupus erythematosus, lichen sclerosus) 4. Surface irregularities (icthyoses, porokeratoses, actinic keratoses, basal cell nevus syndrome, pitted keratolysis) 5. Presence of foreign material (scabies, chromomycosis) 6. Scale formation (papulosquamous diseases) 7. Crust formation (inflammatory dermatoses, neoplasia, trauma) 8. Presence of blood (inflammatory dermatoses, neoplasia, trauma) 9. Black dots (nevocellular neoplasms, North American blastomycosis, tinea nigra palmaris) Deeper Epidermis (Granular, Spinous, and Basal Cell Layers) 1. Reticulated white opacification (lichen planus, lichenoid dermatoses) 2. Diffuse white opacification (lichenoid dermatoses, lupus erythematosus) 3. Brown, tan, and dark discolorations (nevocellular neoplasms, seborrheic keratoses, photodamaged skin, lentigines, angiokeratomas) 4. Pigment network (nevocellular neoplasms) 5. Presence of foreign material (tattoos, perforating dermatoses) 6. Denundation (trauma, neoplasia, inflammatory dermatoses) 7. Appendageal pores (comedones, chromohydrosis, trichostasis, folliculoma) Dermal Alterations 1. Distortion of capillary loop patterns (psoriasis vulgaris, inflammatory dermatoses, lupus erythematosus, scleroderma) 2. Visualization of the normally occult subpapillary vascular plexus (preatrophy) 3. Telangeictases (atrophy and atrophogenic skin conditions, neoplasms) 4. Brown, gray, blue, and other colorations (nevocellular neoplasms, foreign materials, hemangiomata) 5. Blanching erythema (inflammatory dermatoses) 6. Nonblanching erythema (petechial dermatoses, hemorrhage) 7. Presence of foreign material (foreign bodies, tattoos) 8. Ulcer formation (neoplasms, vascular abnormalities, trauma) 9. Altered dermal structure proliferations (neoplasms)
the normally opaque stratum corneum with mineral oil.2 Such enhanced low-power skin surface magnification allows non-invasive visualization of changes within the epidermis and superficial dermis. This chapter will describe the methodology and practical use of low-power skin surface magnification using a hand lens through an oil-glass interface for benign dermatoses. It is suitable for use by practicing clinicians for patients with a variety of dermatological disorders.
to help delineate subtle morphological features occurring either on or within the superficial layers of the skin. Recognition of a subtle cutaneous morphologic feature can suggest a single condition or focused group of conditions, as summarized in Table 15.2. A nonpathognomonic finding integrated with the patient’s history, gross examination, or other established dermatological examinations may possess diagnostic significance.
15.3 METHODOLOGICAL PRINCIPALS 15.2 OBJECTIVE The major objective of non-invasive enhance low-power skin surface magnification during the skin examination is
The simplest form of low-power skin surface magnification occurs when the observer moves closer to what he or she is trying to visualize. An individual’s near-vision
Magnifying Lens — Non-Invasive Oil Immersion Examination of the Skin
visual acuity diminishes after reaching the fourth or fifth decade of life (physiological presbyopia). Therefore, further amplification of visual acuity is required as the mature observer reaches the age when bifocal lenses are necessary to see an object. Bifocal lenses increase the resolution of the eyes by slightly magnifying the size of the object within its field of view. Bifocals represent the simplest form of low-power skin surface magnification. We have found that an 8× lens magnifier (manufactured by Coil, 200 Bath Road, Berkshire, 5L14DW, England) is a practical option for low-power magnification in a clinical setting. It is possible using 8× magnification with an oil-glass interface and adequate lighting to discern occult detail of potential diagnostic significance. When used without an oil-glass interface, low-power skin surface magnification reveals only surface detail because the stratum corneum is opaque when viewed with incident light. The irregular surface of the stratum corneum reflects incident light in multiple directions rather than transmitting the light to the deeper layers of the epidermis and dermis. Enhanced cutaneous visualization of the skin is achieved by placing a drop of mineral oil and a glass cover slip on the skin surface during low power skin surface magnification. Mineral oil acts as a conforming fluid that has a refractive index similar to that of the stratum corneum. It tends to smooth out the surface and prevent light from scattering. Thus, the incident light penetrates the epidermis, thereby allowing the skin to appear more translucent. The mineral oil renders the normally opaque epidermis translucent. The glass cover slip reduces reflective glare and stabilizes the oil in the visual field. The oil-glass interface provides the opportunity to view structures or findings located within the epidermis, at the dermal-epidermal junction, and in the papillary dermis. When used with an oil-glass interface, skin surface magnification with adequate illumination provides a suitable basis for an improved method to observed subtle changes. An external non-glare white light, preferably a portable cold light fiber optic source, held about 6 in. away from the skin, provides suitable illumination. A commercially available hand-held skin scope, somewhat similar in principle to an otoscope, with a built-in light source and a magnifying lens, is discussed in Chapter 16. Normal skin demonstrates a varied but definitive dermatoglyphic surface pattern at 8× magnification. The dermatoglyphic pattern is made up of intersecting grooves that yield triangular, rhomboidal, and other geometric shapes that make up the architecture of the skin surface, as shown in Figure 15.1. In addition, depending upon the anatomic location, other normal skin surface characteristics, such follicular orifices, hairs, and sweat pores are visualized. In tanned skin delicate brown-colored lines form a grid-like uniform pigment network background.
103
FIGURE 15.1 The dermatoglyphic pattern of normal skin surface with a faint pigment network at 8× through an oil-glass interface.
FIGURE 15.2 Relatively normal-appearing fingertip with an illdefined white area at 8× without an oil-glass interface.
The fine reticulated brown-color outline corresponds to the paucity of epidermal melanin at the tips of the dermal papillae and the apparent relative excess pigmentation due to a layering of melanin-containing basal cells of the rete ridges at the epidermal-dermal junction when viewed from the surface.13–17,19 Alteration of the pigment network may occur if there is an inflammatory or neoplastic disorder involving the epidermal-dermal junction. With practice enhanced low-power skin surface magnification examination will maximize the recognition of subtle morphologic detail. An example of such a subtle finding that was difficult to visualize with both naked-eye examination and with low-power skin surface magnification without an oil-glass interface is demonstrated in Figure 15.2. In contrast, a white dermal foreign body opacification is easily seen after application of a drop of mineral oil and 8× magnification, as demonstrated in Figure 15.3. Both Figures 15.2 and 15.3 were taken at the same magnification! The only difference was that the mineral oil rendered the stratum corneum translucent in Figure 15.3, thereby allowing the incident light to penetrate to a deeper level and reflect the finding. With the exception of certain topographic and textural changes, such as surface dryness and scale formation, the oil-glass interface enhances
104
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 15.3 Small white opacification at site of a foreign body in the same finger depicted in Figure 15.2. However, it is viewed at 8× through an oil-glass interface.
FIGURE 15.5 Follicular plugging from a plaque of lichen sclerosus at 8× through an oil-glass interface.
visualization of alterations occurring either on or within the superficial layers of the skin. The cutaneous morphologic signs observed using enhanced skin surface magnification are the same as those that are sought during the routine naked-eye dermatological examination. The finding may just be amplified and easier for the clinician to discern! Selected clinical findings observed with enhanced low-power skin surface magnification are summarized in Table 15.2. The findings are grouped according to their respective epidermal or dermal location. Examples of clinically significant changes occurring either on or within the stratum corneum include hair abnormalities, follicular plugging, and burrow formation. Exclamation point hairs in a patient with alopecia areata above the skin surface are seen in Figure 15.4. In addition, hair shaft abnormalities such as the trichostasis, spinulosa, tinea capitis, trichorrhexis nodosa, and others may be revealed using this technique. Follicular plugging in lichen sclerosus is demonstrated in Figure 15.5. A burrow occurring within the stratum corneum of a patient with scabies is demonstrated in Figure 15.6. Higher magnification is required to visualized the actual mite. Further examples of stratum corneum alterations that may be seen include surface depressions (basal cell nevus syndrome,
FIGURE 15.6 Burrows in a patient with scabies at 8× through an oil-glass interface.
FIGURE 15.4 Exclamation point hairs in a patient with alopecia areata at 8× through an oil-glass interface.
FIGURE 15.7 Reticulated white opaque streaking from a patient with lichen planus at 8× through an oil-glass interface.
porokeratosis of Mibelli, pitted keratolysis), discolorations (tinea nigra palmaris, chromomycosis), and contents of appendageal pores (comedones, chromhydrosis, trichofolliculoma). Alterations within the epidermis are also demonstrated with enhanced low-power skin surface magnification. Reticulated delicate white opacification streaking (Wickham’s striae) occurring within a lesion of lichen planus is shown in Figure 15.7. Other inflammatory dermatoses, such as various forms of dermatitis and psoriasis
Magnifying Lens — Non-Invasive Oil Immersion Examination of the Skin
FIGURE 15.8 Well-demarcated erythema and patterned epidermal thinning containing coiled capillary loops and bleeding points in a plaque of psoriasis vulgaris at 8× through an oil-glass interface.
vulgaris, may demonstrate distinguishing features. Dermatitis may show superficial crust with dried blood and acanthosis with dilation of vertically and horizontally oriented blood vessels. Psoriasis vulgaris has scale, an alternating pattern of acanthosis, and thinning of the epidermis along dilated vertically oriented coiled capillary loops which penetrate the thinned portion of the checkerboard patterned acanthotic epidermis. In addition, delicate pin point bleeding from the thinned parts of the epidermis may be noted (Auspitz’s sign), as seen in Figure 15.8. Enhanced low-power skin surface magnification is useful to detect occult alterations in the superficial cutaneous vasculature as seen in atrophy of the skin. Telangeictases are a prominent finding in forms of cutaneous atrophy shown in Figure 15.9. An early sign of iatrogenic topical steroid atrophy is the observation of the normally obscure horizontally oriented subpapillary vascular plexus that has been termed preatrophy.20 Preatrophy is due to steroid-induced shrinkage of the papillary dermis allowing the underlying structures, such as the subpapillary vascular plexus, to be nearer to the skin surface. Preatrophy change is shown with the oil-glass interface in Figure 15.10.
105
FIGURE 15.10 Horizontally oriented network of anastomosing delicate vascular channels in a patient with iatrogenic topical steroid cutaneous preatrophy 8× through an oil-glass interface.
FIGURE 15.11 Mottled brown pigment change in a person with photodamaged forearm skin at 8× through an oil-glass interface.
The presence of pigment within the epidermis or dermis is easily visualized with enhanced skin magnification. Alterations of the normal delicate pigment network can occur when there is a pathological process involving the dermal epidermal junction.19 Mottled pigmentation in a person with extensive photodamaged skin is demonstrated in Figure 15.11. In addition, pigment changes are found in a variety of inflammatory dermatoses and neoplastic disorders. Skin surface microscopy changes found in pigmented neoplasms are discussed in the next chapter. Higher-power resolution may be necessary in order to distinguish benign from malignant nevocellular neoplasms. Extraneous discolorations depending on their age and/or depth, such as tattoo pigment, incontinence of melanin, hemoglobin, hemosiderin, along with other forms of foreign material, may be recognizable with experience.
15.4 SOURCES OF ERROR
FIGURE 15.9 Telangeictases in a patient with overt cutaneous atrophy at 8× through an oil-glass interface.
Low-power skin surface magnification techniques can provide important diagnostic information. Some subtle findings may be diagnostic for cutaneous conditions, whereas others simply may be suggestive. Recognition and interpretation of such a finding(s) will depend on the
106
Handbook of Non-Invasive Methods and the Skin, Second Edition
15.5 CORRELATION WITH OTHER METHODS
FIGURE 15.12 Surface dryness, adherent keratotic debris, and altered surface relief in a patient with X-linked ichthyosis at 8× without an oil-glass interface. See Figure 15.13.
The findings using low-power skin surface magnification augment the routine dermatological examination. Lowpower skin surface magnification does not supplant the gross examination of the skin.19 Rather, low-power skin surface magnification is fine-tuning of observable morphological skills that the dermatologist or other practitioner uses in everyday practice. Subtle cutaneous findings observed with low-power skin surface magnification of a benign dermatosis can be pathognomonic. However, further confirmation by clinical history, gross dermatological examination, a skin scraping, or a biopsy is necessary for a finding that is not pathognomonic. The combined use of traditional dermatological techniques along with lowpower skin surface magnification fosters the recognition of clinical findings that strengthen the diagnostic acumen of a practitioner.
15.6 RECOMMENDATIONS
FIGURE 15.13 Same as Figure 15.12, except at 8× through an oil-glass interface. The dryness and surface relief are obscured by the oil-glass interface.
competence of the observer. However, as with any type of clinical examination, a nonpathognomonic sign or a suspicious sign of malignancy should be further investigated with a biopsy. Subtle low-power skin surface magnification findings may be found in one or more dermatological conditions. A learning curve may be necessary for a morphologist to become familiar with the use of lowpower skin surface magnification. Visualization of certain surface changes, such as dryness, scale formation, and shadowing may be obscured when an oil-glass interface is used, as demonstrated in Figures 15.12 and 15.13. Such surface topographical details is best appreciated without an oil-glass interface. Pigment network alterations, minute abnormal changes within capillary loops and other vascular channels may require medium or higher-power magnification to be fully appreciated. Unless a highly specific low-power skin surface magnification morphological finding is found, further examination and clinical pathological correlation are necessary for it to be diagnostically meaningful.
Low-power skin surface magnification (8×) through an oil-glass interface is a useful technique that augments rather than supplants the routine naked-eye dermatological examination. It is non-invasive and easy to perform. The findings can assist the observer in making a diagnosis or developing differential diagnoses. Enhanced low-power skin surface magnification is a method of cutaneous examination that has a place in the dermatologist’s diagnostic armamentarium.
REFERENCES 1. Hinselmann, H., Die Bedeutung der Kolposkopie fur den Dermatologen, Dermatol. Wochenschr., 96, 533, 1933. 2. Goldman, L., Some investigative studies of pigmented nevi with cutaneous microscopy, J. Invest. Dermatol., 16, 407, 1951. 3. Gilje, O., O’Leary, P.A., and Baldes, E.J., Capillary microscopic examination of skin diseases, Arch. Dermatol., 68, 136, 1958. 4. Goldman, L., Clinical studies in microscopy of the skin at moderate magnification, Arch. Dermatol., 75, 345, 1957. 5. Goldman, L., A simple portable skin microscope for surface microscopy, Arch. Dermatol., 78, 246, 1958. 6. Goldman, L., Direct microscopy of skin in vivo as a diagnostic aid and research tool, J. Dermatol. Surg. Oncol., 6, 744, 1980.
Magnifying Lens — Non-Invasive Oil Immersion Examination of the Skin
7. Goldman, L. and Younker, W., Studies in microscopy of the surface of the skin., J. Invest. Dermatol., 9, 11, 1947. 8. Gilje, O., Capillary microscopy in clinical dermatology, Bibl. Anat., 1, 203, 1961. 9. Cunliffe, W.J., Forster, R.A., and Williams, M., A surface microscope for clinical and laboratory use, Br. J. Dermatol., 90, 619, 1974. 10. Davis, M.J. and Lorincz, A.L., An improved technique for capillary microscopy of the skin, J. Invest. Dermatol., 28, 283, 1957. 11. Epstein, E., Magnifiers in dermatology: a personal survey, J. Am. Acad. Dermatol., 13, 687, 1985. 12. Tring, F.C. and Murgatroyd, L.B., Surface microtopography of normal human skin, Arch. Dermatol., 109, 223, 1974. 13. Soyer, H.P., Smolle, J., Hodl, S., Pacernegg, H., and Kerl, H., Surface microscopy — a new approach for the diagnosis of cutaneous pigmented tumors, Am. J. Dermatopathol., 11, 1, 1989. 14. Fritsch, P. and Pechlaner, R., Differentiation of benign from malignant melanocytic lesions using incident light microscopy, in Pathology of Malignant Melanoma, Ackerman, A.B., Ed., Masson, New York, 1981, 301. 15. Pehamberger, H., Steiner, A., and Wolff, K., In vivo epiluminescence microscopy of pigmented skin lesions. I. Pattern analysis of pigmented skin lesions, J. Am. Acad. Dermatol., 17, 571, 1987.
107
16. Steiner, A., Pehamberger, H., and Wolff, K., In vivo epiluminescence microscopy of pigmented skin lesions. II. Diagnosis of small pigmented lesions and early detection of malignant melanoma, J. Am. Acad. Dermatol., 17, 584, 1987. 17. Fritsch, P. and Pechlaner, R., The pigment network: a new tool for the diagnosis of pigmented lesions, J. Invest. Dermatol., 74, 458, 1980. 18. Knoth, W., Boepple, D., and Lang, W.H., Differentialdiagnostiche Untersuchungen Met dem Dermatoskop bei ausgewahlten Erkrankungen, Hautarzt, 30, 7, 1979. 19. Bahmer, F.A., Fritsch, P., Kreusch, J., Pehamberger, H., Rohrer, C., Schindera, I., Smolle, J., Soyer, H.P., and Stolz, W., Terminology in surface microscopy, J. Am. Acad. Dermatol., 23, 1159, 1990. 20. Katz, H.I., Prawer, S.E., Mooney, J.J., and Samson, C.R., Preatrophy: covert sign of thinned skin, J. Am. Acad. Dermatol., 20, 731, 1989. 21. Lehmann, P., Zheng, P., Lavker, R.M., and Kligman, A.M., Corticosteroid atrophy in human skin. A study by light, scanning, and transmission electron microscopy. J. Invest. Dermatol., 81, 169, 1983.
16 Dermatoscopy Wiete Westerhof Department of Dermatology, Academic Medical Center, University of Amsterdam, The Netherlands
CONTENTS 16.1 Introduction............................................................................................................................................................109 16.2 Object.....................................................................................................................................................................109 16.3 Instrumentation and Technical Methods ...............................................................................................................110 16.3.1 Instruments.................................................................................................................................................110 16.3.2 Infrared Photography and Dermatoscopy .................................................................................................112 16.3.3 Ultraviolet Photography and Dermatoscopy .............................................................................................113 16.3.4 Surface Microscopy8–11 ..............................................................................................................................114 16.4 Examples of Dermatological Entities Studied with a Dermatoscope ..................................................................114 16.4.1 Benign Conditions .....................................................................................................................................114 16.4.2 Alterations in Cutaneous Vessels in Various Diseases..............................................................................116 16.4.3 Malignant Conditions ................................................................................................................................116 16.5 Validation of the Method.......................................................................................................................................121 16.6 Archiving and Follow-Up......................................................................................................................................122 References .......................................................................................................................................................................122
16.1 INTRODUCTION Macroscopic morphological examination of skin disorders is, besides history taking, general physical examination, and laboratory investigation, one of the pilots of the diagnostic methods in dermatology. An extension of the morphological examination is the invasive technique of biopsy taking for histopathology, histochemistry, immune-histochemistry, and autoradiographic methods. Dermatoscopy, which is a non-invasive micromorphological method, can be considered as a trait d’union between microscopic morphological investigations and invasive histological techniques. From the historical point of view epiluminescence microscopy has been used for two different applications: capillary microscopy and dermatoscopy. The capillary microscopy of the nail fold and the nail bed will be dealt with in another chapter. Dermatoscopy is a readily available diagnostic method, allowing us to appreciate the fine details not normally seen with the naked eye. It furnishes us with the overall detail, which is not the case in low-power scanning electron microscopy, the latter being, moreover, an elaborate and very expensive technique. The unexpected details seen with the dermatoscope often give us answers regarding the pathophysiology of various skin diseases, as well as the urge to pose questions requiring further
investigation. As such, the dermatoscope can be applied as an objective tool in research, e.g., follow up of ultraviolet (UV)-erythema reaction, evaluation of sweat gland function, skin texture measurements, etc. With the existing techniques of UV, infrared (IR), and fluorescence photography, dermatoscopy is used to apply these methods to the fine detail of the skin. The possibilities of the technique are described in the following paragraphs. Dermatoscopic photography provides the clinical instructor with a new valuable teaching aid that not only enhances what is seen clinically, but also creates a surprisingly pleasing visual image of color, shadows, texture, and form.
16.2 OBJECT The dermatoscope allows the examination of skin lesions at different magnifications. Based on these images differential diagnosis is possible. So far, relatively little has been published about the use of the dermatoscope in the diagnosis and follow up of general dermatological lesions. Several skin diseases constitute, even for experienced dermatologists, a diagnostic problem. This applies in individual cases to macroscopic detail as well as to the microscopic aspects. Often the diagnosis is based on the combination of the microscopic clinical symptoms and the 109
110
Handbook of Non-Invasive Methods and the Skin, Second Edition
3
4
6
8
10
2
1
20
25
27
5
28
7
9
11
26
30
29
31
21 32
22
12
33
34
23 14
13
35 36
24 16
19
15
17
37
38
18 39
FIGURE 16.1 Map of the regional differences in skin texture.
microscopic details. However, there are limits to this approach and other avenues have to be explored. In dermatology normally a magnifying lens is used to help in the diagnosis of microscopic details. The magnification is maximally three times and, therefore, the link with the microscopic detail does not exist. Basically the dermatoscope can be used for five groups of skin lesions: 1. 2. 3. 4.
Lesions having a vascular background Lesions having a difference in color pattern Lesions having a change in the skin texture Lesions having abnormalities in the distribution of skin adnexae 5. Lesions with a malignant background Before knowing what is abnormal it is necessary to study the normal features of the skin. In this context one has to remember that this depends on the age, sex, the site of the skin investigated, and the environmental conditions in which the investigation with the dermatoscope is carried
out. Figure 16.1 gives an overview of the different locations of the body, where the skin detail is characteristically different from neighboring areas. It is clear from this table that “normal” skin does not exist. Furthermore, the skin of a baby cannot be compared with the skin of a 90-yearold person. In order to be able to make a diagnosis based on dermatoscopy, it is not possible to go by primary efflorescences as is used in dermatological practice. So new criteria describing skin changes at the microscopic level have to be developed.
16.3 INSTRUMENTATION AND TECHNICAL METHODS 16.3.1 INSTRUMENTS In this chapter different microscopic techniques are described, which make use of similar optical instruments. These are dermatoscope, capillary microscope,
Dermatoscopy
111
FIGURE 16.2 The Zeiss dermatoscope.
epi-luminescence microscope, and surface microscope. Their optical properties are in essence equal to an operation or dissecting microscope. Sometimes the names of these instruments are exchanged in the text. Some of the most current instruments are described here. Many more are available on the market. However, it is not possible within the framework of this chapter to list them all. •
•
Zeiss dermatoscope (Zeiss, Germany):1 The dermatoscope is supplied with a bendable arm fixed to the wall (Figure 16.2). For the examination of the patient it is not difficult to position the dermatoscope due to the flexibility of the arm. Some dermatoscopes are mounted on a movable tripod. For the handy manipulation of the tool the supporting arm is supplied with a counterweight, so that the dermatoscope remains in the working position. The Zeiss dermatoscope has a binocular and a monocular tube and one arm of the tube can also be fitted with a photocamera or film camera. The magnification system (motorized zoom) is stepless from 0.4 to 40 times. An objective is used with a focus of 175 mm. The ocular enlarges 10 times. The dermatoscope has a field cross section of 22.5 mm the standard equipment can be extended with different additional options. The binocular tube can be supplied with a microscope body with a magnification step times 3. Without problems also film cameras or a video camera can be coupled to the optical system. With a beemsplitter the visual image can be photographed when a camera is positioned on an additional tube. The stereoscopical observation by at least two doctors is an advantage. Wild M650 (Wild Heerbrugg AG, Switzerland): Epiluminescence microscopy (ELM) is a term often used in relation to the investigation of pigmented lesions but is in essence similar to dermatoscope. ELM can be performed with a Wild M650 binocular surface microscope
FIGURE 16.3 The Delta 10 dermatoscope (Heine).
•
equipped with objectives of 91 mm working distances. Magnifications obtainable are ×6, ×10, ×16, ×25, and ×40. All pigmented skin lesions are first examined for surface structure. They are then covered with immersion oil and a glass slide that is applied with slight pressure; this renders the epidermis translucent and allows study of the dermoepidermal junction zone. Photographs of the pigmented skin lesions can be taken with an Olympus CM10 automatic camera, mounted on a side arm of the microscope. Delta 10 dermatoscope (Heine Optotechnik, Germany): With the Delta 10 dermatoscope after Braun-Falco, Billeck and Stolz, Heine Optotechnik developed an instrument2 which, because of its small weight and easy handling, facilitates a quick analysis of pigmented skin lesions. The Delta 10 dermatoscope (Figure 16.3) is provided with an achromatic lens. It makes a tenfold magnification possible of skin changes, when the skin is treated with emulsion oil. The illumination of the object is under an angle of 20° with a halogen lamp inside the dermatoscope and powered by a battery. With the Delta 10 only small areas can be studied, like the interdigital spaces, the nails, the eye corners, the genital, anal, and retro-auricular regions, as only an 8-mm diameter contact lens is illuminated by the halogen lamp via a special fiber optical system. Photographs cannot be made with this system and stereoscopic vision is also an omission of this instrument. Heine has developed a photographic system in which a reflex camera is supplied with a Dermaphot which is a special option for photographing contact dermatoscopical images (Figure 16.4). It has a built-in electron flash system. This Dermatophot might help to overcome the
112
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 16.4 The Dermaphot (Heine).
•
shortcomings of the Delta 10 dermatoscope. Although the contact photography is not as easy as the technique used with the Zeiss dermatoscope it is possible to make good pictures. Microscan System (Fort, United Kingdom): The Fort UK Microscan MS 2500 (Figure 16.5) is a video microscope ideal for dermatological applications. The microscope consists of a video processing unit with an integral highintensity cold light source. The miniature color camera has a built-in fiber optic ring light to which interchangeable objective lenses also having fiber optic ring lights are attached. A wide range of detachable lenses are available both in contact and noncontact formats with magnifications from 5 to 1000 times. The ring lights ensure even illumination of the subject under examination. The miniature camera with its lens has been designed for hand-held operation. The contact objective lenses have been specifically designed so that the contact face is smooth, rounded and comfortable for the patient under examination. When using the noncontact lenses the camera/lens assembly may be mounted into a special microscope stand which is also supplied by FORT UK. The cold light source is 150-W quartz halogen with electronic variable control of the lamp’s intensity and features an exclusive, easy to operate filter wheel that has four neutral density-filters for coarse setting of the illumination intensity and has red, orange, green, and blue colored filters. The use of colored light on the subject under examination can enhance features of interest. Also available are a microprocessor for image acquisition and analysis, a monitor, a mouse fitted to the PC for use as a pointing device, and image analysis software. The software is menu driven using a point-and-click-style user interface. The software provides facilities for
FIGURE 16.5 The Fort UK Microscan MS2500.
storage and retrievel of records consisting of an image and associated notes and comments. The Fort Microscan outputs PAL composite video and can therefore be used with any picture monitor, monochrome or color. The Microscan is fully compatible with all currently available video printers for permanent hard copies and composite video tape recorders.
16.3.2 INFRARED PHOTOGRAPHY DERMATOSCOPY
AND
The skin and superficial tissues reflect most of the infrared falling on and penetrating a short distance into the body, whereas the blood in the veins absorbs much of the infrared. This provides a tone separation. In an infrared color transparency the veins appear blue as well as dark, because they lay under a scattering layer.3 This technique introduces differences in color as well as tone (Figure 16.6). The pattern could be traced visually, but is much more apparent and detailed in the infrared photograph since it has become darkened. In general photographic practice it will be found that infrared radiation between 700 and 900 nm can penetrate the skin to a depth of about 3 mm. The translucency of tissues and particle size govern delineation and tone value in the reflected component. The venous circulation near the surface of the body is superficial and flows through larger vessels than its arterial counterpart. Venectasies, essential teleangiectasia, vascularization of
Dermatoscopy
113
infrared translation color for melanin varies from a light reddish brown to a dark brown — quite similar to that observed visually except that it is always more red. Skin not containing appreciable melanin appears as a pale, cold white. With the basic translation colors for the normal pigments determined, several of the pigmentary disturbances encountered in human skin were explored. Argyria varied in skin coloration from deep brown to blue. In the patient with argyria, the greyness observed in the skin was intensified in the infrared record. But there was little added from the diagnostic point of view.
16.3.3 ULTRAVIOLET PHOTOGRAPHY DERMATOSCOPY
FIGURE 16.6 Pattern of dermal blood vessels of the arm using a Kodak high-speed infrared film 2481.
tumors, and dermal pigmentation in the skin, are detectable earlier by infrared/black and white photography than by visual observation.4 Incidentally, the more prominent superficial veins are recordable in black skin, which makes this type of photography of value in mapping venous patterns in dark-skinned people. The absorption of melanin must be mainly in the visible region, because melanin granules on photomicrographic slides transmit actinic infrared. This is evidenced by the fact that melanotic tissue sections, and the black melanophores in frog skin, record red in infrared color transparencies made by transmission photomicrography, demonstrating translucency to infrared.3,5 It has generally been assumed, rather without base, that black skin photographs are lighttoned by infrared because melanin reflects infrared. However, the findings obtained by infrared color photography indicate that this is not so. The mechanism must be that of a high infrared transmittance to the underlying, or intermixed, tissue. These reflect the infrared back through the almost “transparent” melanin granules. Thus, as far as infrared photography of the skin is affected, the melanin particles might as well be a thin superficial layer of transmittant ground glass, with high absorption in the visible range of the spectrum, but little absorption in the infrared. The false-color renditions of several skin lesions are described and discussed in Gibson.3 In this paper, he demonstrated differentiation between a pigmented nevus of the fingernail and a splinter hemorrhage in the nail bed. The
AND
A source of ultraviolet light from which all visible rays have been excluded by a Wood’s (nickel oxide) filter is an important investigative tool in the diagnosis and treatment of many dermatoses. Only the most important of the many uses to which ingenuity and imagination have contributed are mentioned here. Wood’s Light6 is a powerful UVradiation source. The various dermatoses can also be photographed by dermatoscope. A flash light is used for the illumination, which can be filtered with colored filters, but can also be used to flash with parts of the ultraviolet spectrum. This enables us to visualize certain disease patterns, e.g., a disease in which fluorescence, evoked by ultraviolet illumination, is becoming visible. Special UV films are used to develop the image. Examples of these are different fungus infections which give green fluorescence. Certain bacterial conditions like erythrasma give a coral-red fluorescence. A red fluorescence arises in the blister fluid of patients with porphyria cutania tarda. Epidermal pigmentary disorders, such as vitiligo lesions (Figure 16.7) or café au lait spots, are better demonstrated with the help of ultraviolet illumination. Detection of fluorescing dyes causing contact dermatitis can be demonstrated specifically. The remnants of mineral oils present in the pores of hair follicles persist even after washing. The presence of a black-light is therefore of importance to localize and to determine the size of a condition before photographing the condition. In the second place it is possible to combine the ultraviolet light with a fluorescent dye with fluorescein. This substance can be injected intravenously as a sodium salt. This is practiced, for example, in ophthalmology, which could also be applied in certain conditions of the skin. Under ultraviolet illumination the vasculature and the diseased skin can be studied with the dermatoscope when the passage of the fluorescing dye is maximal.7 Characteristic changes can occur, for example, in chronic wounds, scabies, and in the case of malignant tumors. For the determination of the progress of relatively fast movements it would be handy to use a film camera instead of a photocamera. This is the socalled time-lapse cinematography.
114
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 16.7 UV photograph (Kodak Panatomic-X film 135) enhances the contrast between pigmented and nonpigmented skin. Borders might reveal interesting dermatoscopic detail.
FIGURE 16.8 Yeast infection of the skin demonstrated by surface microscopy (courtesy of Dr. G.E. Pierard).
16.3.4 SURFACE MICROSCOPY8–11 Skin surface biopsy consists of sampling the superficial layers of the stratum corneum. There are two variants. One type is performed with a cyanoacrylate adhesive and a sheet of transparent plastic. The other one relies on the use of commercially available small adhesive discs. The method is non-invasive and painless. It may be sequentially repeated at the same site. The indications are multiple. They primarily concern the evaluation of xeroses and the diagnosis of squamous inflammatory dermatitis, superficial infections (Figure 16.8), and parasitoses, as well as pigmented neoplasms (Figure 16.9). Experimental studies concerning the biology of the dermatoscopy and the pharmacokinetics of various drugs may also be undertaken.
16.4 EXAMPLES OF DERMATOLOGICAL ENTITIES STUDIED WITH A DERMATOSCOPE 16.4.1 BENIGN CONDITIONS Skin texture — The system of grading cutaneous microphotographs described by Beagley and Gibson12 relates to changes in skin surface texture, which in normal skin is
FIGURE 16.9 Melanoma cell in stratum corneum (surface biopsy) supporting the diagnosis by a non-invasive method (courtesy of Dr. G.E. Pierard).
composed of a series of transverse and diagonal primary lines, which intersect to form quadrilaterals and triangles. Within these primary figures are sets of smaller secondary lines which often meet in the center of the figure, forming a star configuration (Figure 16.10). The six-step grading system devised by Beagley and Gibson, and based on alterations in these skin surface characteristics thought to reflect actinic damage, is summarized in Table 16.1. Facial skin wrinkling or “crow’s feet” are more common and more severe in cigarette smokers than in nonsmokers.13 “Crow’s feet” (wrinkles in the lateral periorbital area) have been thought also to be caused by sun
Dermatoscopy
115
FIGURE 16.10 Primary, second, and tertiary skin lines.
TABLE 16.1 The Beagley-Gibson System of Grading Cutaneous Microtopographs Taken From the Dorsum of the Hand Grade
Features
1
Primary lines are all of the same depth. Secondary lines are all clearly visible, are nearly the same depth as the primaries, and often meet to form an apex of triangles (star formation).
2
Some flattening and loss of clarity of the secondary lines. Star formations are still present, but often one or more of the secondary lines making up the configuration are unclear.
3
Unevenness of the primary lines. Noticeable flattening of the secondaries with little or no star formation.
4
Macroscopic deterioration in texture. Coarse, deep primary lines. Distortion and loss of secondary lines.
5
Noticeable flat skin between the primary lines. Few or no secondary lines.
6
Large deep and widely spaced primary lines. No secondary lines.
exposure, the aging process, and other factors, such as massive weight reduction.14 Daniell13 inspected live subjects and reviewed photographs taken from the “crow’s foot” area of their faces; the latter is called paraocular photography (POP). Dermatoglyphics — Dermatoglyphics is a term applied to both the configurations of ridged skin, and the subject that deals with it. Characteristic ridge patterns of whorls and loops are found on the volar skin which are unique for any individual (Figure 16.11). The systematic classification of ridge patterns, as a means of personal identification or for use in studies of inheritance, requires image analysis as is possible with the Microscan System (Fort). Dermatoglyphic features may indicate an increased tendency to develop a particular condition and can be an aid to diagnosis in dermatology. For example, findings
FIGURE 16.11 Dermatoglyphics.
include, in alopecia areata, a decreased incidence of ulnar loops in the second left digit in both sexes, and in psoriasis an increased incidence of whorls in the fourth digit, more marked in the right hand.15,16 Ectodermal dysplasia — Hypoplasia of appendages occurs in the skin and mucus membranes. In the anidrotic form the skin is dry and there may be palmoplantar keratoses. The most important feature is heat intolerance due to deficient sweat glands. This can be demonstrated in patients by rubbing the finger over a slightly rough surface, thereby provoking heat. In normal persons the sweat drops that appear due to profuse sweating at the orifices of the sweat ducts on top of the ridges can easily be seen with the dermatoscope. In patients with ectodermal dysplasia this feature is absent or highly decreased (Figure 16.12A and B). The decrease in number of sweat glands can also be demonstrated in female family members having the trait of this X-linked recessive inheritable disease (personal observation). Parapsoriasis en plaque — The clinical evaluation of parapsoriasis en plaque with the dermatoscope gives an even so-called quiet image. The structure of the skin surface shows a regular pattern. The scales are discrete, not elevated or bizarre. The colors are evenly distributed in all the microscopic fields. Seborrheic eczema — Seborrheic eczema has a variable pattern.1 The image shows a certain succulence with yellow scales and an irregular surface. Also the colors are contrasting. In the first stages of mycosis fungoides the images arising from the dermatoscope resemble those of seborrheic dermatitis. Also here differentiating parapsoriasis en plaque from other eczemas is difficult. Morphea — The morphea-like lichen sclerosis at atrophicus (LSEA) is in comparison to the small plaque type of circumscribed scleroderma especially remarkable with regard to the distinct follicular keratosis. This is not a new phenomenon; however, in the investigations of Knoth et al.1 the regularity of the crater-like or plaquelike surface of the hyperkeratosis was very prominent in LSEA. In case of morphea the surface pattern is more uniform and smooth. In lichen sclerosis et atrophicus of the glans of the penis or on the prepuce the follicular
116
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 16.12 (A) Sweat droplets appear at the openings of the sweat ducts on the ridges, as seen in a normal person.
FIGURE 16.12 (B) Absence of dermatoglyphics in a patient with ectodermal dysplasia; no sweat secretion.
pattern is normally absent, which was unknown before the dermatoscopical investigation. Lichen ruber planus — The various developmental stage of lichen ruber planus are especially remarkable in the early stage. The primary and secondary skin folds disappear. The papules are grouped, discrete, and slightly swollen. The development of hyperkeratosis and the Wickham striae comes later. Corpora aliena — It is easily investigated with a dermatoscope, whether the particle is superficial or deep in case of corpora aliena in the face as a result of explosion of gun powder, which is an important implication with regard to the choice of treatment, for example excision or dermabrasion.
16.4.2 ALTERATIONS IN CUTANEOUS VESSELS VARIOUS DISEASES
IN
Besides its usefulness in studying the anatomy and function of the superficial cutaneous vessels, capillary microscopy shows some promise in being used to differentiate skin diseases. At present, however, its use as a diagnostic aid is limited by the lack of complete data on the various changes in the superficial circulation of the skin in disease.17,18
Atopic skin — Capillary microscopy is particularly useful for studying vascular changes in atopic dermatitis in response to pharmacologic or physiologic stimuli. There is an enhanced vasoconstrictor tendency in atopic skin, which is demonstrable by the ease with which white dermographism is produced on stroking, by a more rapid rate of cooling in a cold environment and a slower rate of cooling in a cold examination and a slower rate of warming in a warm examinations, by exaggerated response to cold pressor tests, and finally by the delayed blanch phenomenon. In atopic skin there is a paradoxical reaction, which consists of the appearance of a white halo around the wheal formed by introducing dilute acetylcholine into the skin. This delayed blanch phenomenon begins between 3 and 5 min after the injection of acetylcholine and has been interpreted as vascoconstriction of the skin vessels in response to a vasodilator drug. This reaction was studied by dermatoscopy in a series of atopic patients18 and direct observations were made of the capillary bed, after the introduction of acetylcholine. The capillaries at the periphery of the wheal remained dilated while the area appeared grossly blanched. Introduction of hyaluronidase into the test site before the injection of acetylcholine eliminated the blanching, and it appears that the whiteness seen grossly is due to the edema forming as a result of vasodilation of the capillaries in response to a vasodilator substance. Atopic dermatitis — The capillaries in the involved skin of patients with atopic dermatitis are somewhat dilated. There is a regular distribution of the capillaries (which are all open) but acanthotic changes in the epidermis obscure the subpapillary plexus. The capillaries themselves are frequently obscured as a result of edema in the papillae and epidermis.18 The arrangement of the vessels in atopic skin is somewhat reminiscent of the regular arrangement seen in psoriatic plaques; however, there is little tortuosity of the capillaries. The vascular bed in the uninvolved skin of atopic patients appears normal. The capillary changes seen in areas of localized neurodermatitis are not distinguishable from those of atopic dermatitis. Psoriasis — Certain consistent alterations occur in the capillary bed in psoriatic lesion. Gilje17 has described the regularly arranged distribution of capillaries in psoriatic plaques which, when viewed through the scales on the surface of the lesion, appear as puff balls. The endcapillaries in this disease are extremely tortuous and coiled. The subpapillary plexus is obscured, probably because of the acanthosis present. The capillaries are longer than normal but dilatation is not a prominent feature of the capillary changes in psoriasis. This is in contrast to the usual impression one gains from reviewing histological sections of psoriatic lesions. The apparent dilatation of the capillary vessels is probably an artifact which results from sectioning through a coiled vessel and
Dermatoscopy
unavoidably getting a tangential cut through at least some vessels. There is no increase in the number of capillaries. Flow through these capillaries appears normal and the capillaries can be seen to pulsate. The diameter of the venous and arterial limbs of these capillaries are approximately equal; this being another departure from normal. In early psoriatic lesions the capillaries are also abnormal in that the end-vessels are somewhat tortuous, but not nearly to the degree seen in old lesions. The “normal” skin of patients with psoriasis also shows some abnormally tortuous capillaries. This suggests that more attention should be directed to the role of the capillaries which may be involved in some way in producing the manifestations of psoriasis. The tortuosity and increased length of the capillaries in a psoriatic lesion represent a considerable increase in endothelial surface. This correlates well with the elevated oxygen consumption in psoriatic lesions. However, tortuosity of a capillary is not a finding limited to psoriasis; it has been found with some consistency in the nail folds of subjects who have neurasthenia and no evidence of skin disease. Psoriasis pustulosa — In psoriasis pustulosa there is a similar picture, but it varies somewhat from the preceding picture of psoriasis. There is a ringlike arrangement of the capillaries around a prominent round or oval structure. Eczema — The capillary microscopic picture in eczema varies a good deal from the clinical picture. There is a more irregular arrangement of the capillaries and these do not resemble rolled-up balls as in psoriasis. The tips of the capillary loops are always much smaller, and they may vary in size within the same field of vision. The group-like arrangement of the dilated capillaries corresponds to vesicles or papules. The dermal ridges are pronounced here. The subpapillary plexus is neither visible here nor in the pictures of psoriasis. Lichen ruber planus — Lichen ruber planus shows an entirely different picture. The papule shows an irregular lighter central field, like a plateau, surrounded by a border of straight and slender, slightly twisted capillary loops, which slant inward and upward towards the central lighter region. In this condition there is a marked departure from the normal in the end-capillaries and in the subpapillary plexus. Some of the vessels are obliterated while others are deformed and dilated. There is a regular arrangement of the capillaries. This is seen not only in lichen planus but also in psoriasis and in other papular dermatoses. The violaceous hue, seen in lichen planus papules, is attributable to the engorgement of the capillaries and, to a greater extent, to the engorged venules in the subpapillary plexus seen through a hyperkeratotic stratum corneum. When the hyperkeratotic layer is removed, the lesions appear a dusky red. Lupus erythematosus — In lupus erythematosus the most striking feature of the capillary microscopic picture is a round or oval formation which represents the follicular
117
hyperkeratosis and the short, thick, usually horizontal loops or arcs of vessels. Some of these are arranged around the cornified plugs, others are distributed irregularly throughout the field of vision. There are almost no ordinary capillary loops here. There is almost a complete loss of the capillaries in lesions of chronic discoid lupus erythematosus. The vessels in the subpapillary plexus are deformed and dilated. A patient with lesions of several years’ duration on the face, with no lesions in other areas, had atrophic changes grossly in the skin over the dorsum of the hands and forearms, sites which had never been involved with discoid lesions. This patient had no clinical or laboratory evidence of constitutional disease. The vascular beds on the dorsum of the hands and forearms were decidedly abnormal: there was an absence of end-capillaries and a dilated tortuous subpapillary plexus. The vascular bed on the volar surface of the forearm appeared normal. Gilje et al.19 found abnormal capillaries in the nail folds of patients with discoid or disseminated lupus erythematosus. Ichthyosis — The capillary microscopic picture in ichthyosis is entirely different from the foregoing diseases. The picture in ichthyosis usually shows capillaries which are slender, long and narrow with a narrow crest and with no particular differentiation of the arterial and venous branches. In some areas the capillaries branch out from the subpapillary plexus like plants in a bed or like the spray in a fountain. In some of the thin capillary loops it is possible to follow the blood stream, which has the appearance of beads on a string. Hereditary hemorrhagic telangiectasia20 — Examination of the grossly visible ectasias shows them to consist of markedly coiled, widely dilated superficial vascular channels. The flow of blood through these angiomatous vessels is slow. It is difficult to visualize the inflow and outflow vessels to the coiled mass; therefore, it is not known whether the mass consisted of one or more vessels. The normal skin of two patients with hereditary hemorrhagic telangiectasia also show unusual vascular structures. These consisted of a circular vascular channel, about 0.3 mm in diameter, enveloping a second circular channel. The rate of flow of blood in these vessels was slow. Approximately two such structures per square millimeter could be seen. There were no end-capillaries in these areas and the circular channels may have been the precursors of the markedly coiled vessels in the involved skin of telangiectasia. These vessels superficially resembled the schematic representations of arteriovenous anastomoses. They were smaller than normal glomus bodies and were more superficially situated in the skin. Senile skin and ectatic vessels — The vascular bed in senile skin shows certain distinguishing characteristics. The end-capillaries appear to be shorter than those in young subjects and the subpapillary plexuses appear more prominent. Since we do not yet have a method for
118
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 16.13 Ectatic tortuous blood vessels in senile skin.
FIGURE 16.14 (B) Shakespeare type C, nevus flammeus on cheek, male 17 years old.
accurately measuring the lengths of capillaries, because of their perpendicular orientation in the skin, the observation that the capillaries are shortened must remain only an impression. Shortening of the end-capillaries in aging skin correlates with the flattening of the dermal papillae. Thinning of the overlying structures accounts for the apparent prominence of the subpapillary vessels in aging skin. The ectatic and tortuous cutaneous vessels that characterize many diseases are dilated subpapillary venules, and possibly arterioles, but not capillaries (Figure 16.13). In areas of the skin in which ectatic vessels are found, a greatly reduced number of end-capillaries appear. The epidermis and dermal papillae above these ectatic vessels are usually atrophic. In skin diseases that are characterized by eventual atrophy of the skin, such as discoid lupus erythematosus, poikiloderma, or chronic radiodermatitis, there is usually complete destruction of the papillary capillaries and a dilatation of the subpapillary vessels. The capillaries seem to be more susceptible to destruction in disease processes than are the subpapillary plexuses. Telangiectasias are better visualized when the horny layer is removed with formic acid.21 Nevus flammeus — In a study performed in the Academic Medical Center of the University of Amsterdam the dermatoscopic picture of nevus flammeus is investigated before argon laser treatment (Figure 16.14).
FIGURE 16.14 (C) Shakespeare type D, nevus flammeus on cheek, female 15 years old. (Courtesy of Prof. Dr. Ir, M.J.C. van Gemert, Dr. C.M.A.M. van der Horst, and Dr. S.D. Strackee.)
FIGURE 16.14 (A) Shakespeare type B, nevus flammeus on chin, female 2 years old.
The different appearances of capillaries in port wine stains of increasing severity are described in Table 16.2 (after Shakespeare and Carruth).22 Chronic ulcer — The capillary blood supply to chronic venous ulcers can be visualized by making use of dermatoscopy and even better in combination with intravenous perfusion of a fluorescent dye (Figure 16.15).7,23
16.4.3 MALIGNANT CONDITIONS Before describing malignant conditions it is necessary to delineate a benign lesion, e.g., a pigmented melanocytic nevus. On dermatoscopic investigation a spider weblike superficial pigmentary pattern is observed (Figure 16.16). The color is brown to black and uniformly distributed. The greatest density of pigment is central and fading to the periphery. There are no dilated and abnormal capillaries seen. Consequently, there is no erythema. Simple visual inspection is not always sufficient for the clinical diagnosis of melanocytic skin tumors.24 This is particularly relevant for the different diagnosis between dysplastic nevi and an early melanoma, as well as for the differentiation between melanomas and tumors of nonmelanocytic origin. For this differentiation the epiluminescence microscope is being used with increasing frequency. Pehamberger et al.25,26 and others27–31 have
Dermatoscopy
119
TABLE 16.2 Transcutaneous Microscopic Classification of PortWine Stains by Blood Vessel Type Type A
Capillaries of normal appearance, increased in number, are seen. The capillaries show the “pinpoint” ends visible in normal skin, and are analogous to the “constricted type”.
Type B
Capillaries have increased diameters but otherwise a normal appearance. They have a “pin-head” appearance, are very prominent, and are analogous to the “intermediate type”.
Type C
Capillaries with thickened ends, resembling “ring doughnuts”, are seen. The ends of the capillaries are flattened and turned to one side. These capillaries are very prominent, and are analogous to the “intermediate type”.
Type D
Ectatic capillaries with ballooning, show “microcavernous hemangiomata”. Thickening and ballooning of the capillary wall along the ascending and descending limbs is present. These capillaries are analogous to the “dilated type”.
Type E
Very extended capillaries of reduced diameter and complex helical convulsions are seen. These are unusual, are generally seen on the limbs only, and have no analogue in Ohmori and Huang’s description.
Type F
Formation of interconnection between capillaries and vessels in the subpapillary dermal plexus with formation of a vascular network. Capillary “tufts” arising from small feeder vessels may be seen. These are analogous to the “deep-located type”.
previously reported that epiluminescent microscopy (ELM) definitely improves the diagnostic accuracy of pigmented skin lesions. Goldman,32 MacKie,28 Fritsch and Pechlaner,27 and many others2,25,29 have published articles about the differential diagnostic possibilities of the epiluminescence microscope. In these publications differential diagnostic microscopic criteria of a widespread nomenclature were described and evaluated. The work group Analytical Morphological of the Arbeitsgemeinschaft Dermatologische Forschung organized a consensus meeting in 1989 in Hamburg about the nomenclature used in epiluminescence microscopy. With uniform and original description it will be easier to communicate between investigators and this will facilitate a better understanding of the method. The technique is usually alluded to as epiluminescence microscopy or surface microscopy. In practice immersion oil is applied to the pigment lesion and covered with a cover slip. The oil makes the stratum corneum translucent, allowing a better observation of the underlying pigmented structures and patterns. Under epiluminescence illumination the lesion is then investigated
FIGURE 16.15 (A) Intravenous perfusion of fluorescent dye visualizing capillaries in border of granulating venous ulcer. (B) Same as (A) after transcutaneous electrical nerve stimulation. Observe that the newly formed capillaries are leaking. (Courtesy of Dr. M.J.H.M. Jacobs.)
FIGURE 16.16 Dermatoscopic picture of junctional nevus.
with a stereomicroscope with magnifying range of 6 to 50 times. The relations between the histological substrate and the diagnostic meaning of single epiluminescence microscopical criteria are summarized in Table 16.3. Other criteria of diagnostic importance which express the special arrangement of some of the properties inside a lesion are peripheral or central accentuation, regular or irregular distribution.
120
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 16.3 Diagnostic Criteria in Epiluminescence Microscopy (Proposal of the Consensus Meeting of the Work Group On Analytical Morphology, 1989, Hamburg) ELM Criterion Normal pigment network Discrete pigment network Prominent pigment network Regular pigment network Irregular pigment network Wide pigment network Narrow pigment network Broad pigment network Delicate pigment network Irregular extensions, pseudopods Radial streaming Brown globules “Black dots” Whitish veil, “milky way” White scar-like depigmented areas Greyish-blue areas Hypopigmentation Reticular depigmentation Milia-like cysts Comedo-like openings Telangiectasias Reddish-blue areas Maple leaf-like areas
Histological Substrate Pigmented retelists Slightly hypopigmented retelists Hyperpigmented retelists Regularly distributed retelists Irregularly distributed retelists Widely spaced retelists Tightly spaced retelists Broad retelists Narrow retelists Confluent pigmented junctional nets in the periphery Radial distribution of pigmented junctional nest Superficial pigmented nests in the upper dermis Aggregate pigmented melanocytes in the horny layer Compact orthokeratosis and hypergranulosis Decrease of melanin and fibrosis Superficial fibrosis melanophages Decrease of melanin Very wide retelists Intraepidermal horn pearls in epidermis under the skin surface Intraepidermal horn pearls in epidermis Dilated blood vessels in the upper dermis Telangiectasia in the upper dermis Pigmented epithelial nests
It should be emphasized that a diagnosis obtained with epiluminescence microscopy should never be based on a single criterion. In this respect it must be clear that under the term “diagnostic meaning” only frequently occurring features are mentioned. The demonstration of a single feature is not proof for the diagnosis and the absence of criteria is not exclusive of a diagnosis either. An exact epiluminescence microscopic diagnosis can only be made in combination with careful simultaneous investigation of other criteria present. With the epiluminescence microscope a difficult problem is constituted by the over protection of skin levels so that it is difficult to recognize the structures and the colors, and it is also often impossible to associate a certain phenomenon with a corresponding specific histological feature. To support a diagnosis it is therefore important to include the histological as well as the epiluminescence microscopic diagnostic criteria in combination, rather than in isolation. The description of microscopic criteria by epiluminescence, for example, bizarre patterns and other multicomponent compositions, is often very complex and rich in variations, and not as exact as histopathology.
Diagnostic Meaning Normal pigmented skin Likely to be a benign melanocytic lesion Likely to be a malignant melanocytic lesion Benign melanocytic lesion Dysplastic nevus malignant melanoma Melanoma, e.g., melanoma in situ Benign melanocytic lesion Melanoma, v.a. melanoma in situ Benign melanocytic lesion Malignant melanoma Malignant melanoma pigment spindle-shaped cells Regular dermal nevus: irregular malignant melanoma Malignant melanoma Malignant melanoma Regressive malignant melanoma Regressive malignant melanoma In many melanocytic lesions Spitz-nevus Verruca seborrhoica papillomatous dermal nevus Verruca seborrhoica papillomatous dermal nevus Basal cell carcinoma Hemangioma Pigmented basal cell carcinoma
In the study of Steiner et al.40 it was shown that better diagnostic accuracy could be achieved for pigmented Spitz nevi. Criteria have now been established but still have to be tested more extensively. This is important for small pigmented lesions that cannot be diagnosed by clinical criteria alone and in particular for the differentiation of pigmented Spitz nevi from malignant melanoma, dysplastic nevi, or common nevi. By applying ELM pattern analysis to Spitz nevi, we were able to increase the diagnostic accuracy from 56 to 93% (Table 16.4). This is a distinct improvement, particularly because some of these Spitz nevi had clinically been considered to be malignant melanoma. ELM correctly diagnosed three of five lesions as Spitz nevi. In two lesions the ELM pigment pattern was suggestive of malignant melanoma but histopathologically these lesions proved to be Spitz nevi. Schultz41 developed a score system for dysplastic nevi, derived from a group of patients with junctional and dermal melanocytic nevi, which be compared with dysplastic nevi (Table 16.5). The technology underlying epiluminescence microscopy or dermatoscopy is based on the principle that oil reduces reflection from the skin surface, allowing light
Dermatoscopy
121
TABLE 16.4 Epiluminescence Microscopy Appearance of Pigmented Spitz Nevus and Melanoma
General appearance Pigment network Brown globules Black dots Depigmentation Border
Spitz nevi
Melanoma
Monomorphous, starburst lesion, coffee bean-like appearance Prominent, regular, stops abruptly at periphery or thins out, “negative” pigment network Different size, regular throughout the lesion, rim of large brown globules at periphery of lesion Center or throughout lesion, regular distribution Bizarre, retiform in center No pseudopods, no radial streaming
Polymorphous, multiple pattern Prominent irregular, stops abruptly at periphery or thins out, no “negative” pigment network Different size, haphazardly spaced, no rim
TABLE 16.5 Epiluminescence Microscopic Scoring Protocol for Dysplastic Nevi
1. 2. 3. 4. 5. 6. 7. 8. 9. 10.
Criterium
Points
Black dot on blue background Area with a regular distribution of capillaries Pseudospod-like border pattern Regressive remodeling of tissue with melanophages at the periphery Abrupt disappearance of pigment of trabeculae Dendrite-shaped blue-gray trabeculae Bizarre network Multicomponent pattern Gray color Gray-blue globuli in center of papillae
10 10 10
7 6 5 5 2
Evaluation of score: 10–15 suspect for dysplastic nervus >15 highly suspect for dysplastic nevus.
rays to penetrate more deeply. In general benign lesions exhibit consistent patterns of coloration and shape and have defined borders under epiluminescence microscopy, whereas malignant lesions tend to have irregular, less homogeneous patterns of development and color. With epiluminescence microscopy clear distinction can be made between melanocytic and pigmented nonmelanocytic lesions as well as between malignant and benign melanocytic lesions. Cutaneous pigmented lesions, when small and intensely pigmented, may provide difficulties in diagnosis. They may represent a lentigo, a vascular lesion such as angiokeratoma, a minute seborrheic keratosis, a pigmented basal cell carcinoma, or even foreign pigment such as a tattoo. With the aid of epiluminescence microscopy, these lesions can be differentiated. It was found that this technique was useful and even small darkly pigmented papules measuring 1 to 3 mm in diameter of intracutaneous metastatic melanoma could be recognized as such. The cutaneous secondaries showed asymmetric brown-black globules and lacked the irregular pigment
Often only at periphery, irregular distribution Irregular, often in periphery Often pseudopods, often radial streaming
network, radial streaming, pigmented pseudopods, or black dots associated with primary melanoma. However, the epiluminescence microscopic appearance of these melanoma metastases is unlikely to be specific and their appearance would probably be shared by some forms of pigmented dermal nevi. Nonetheless, in the context of newly developing lesions, particularly when they are asymmetric, the finding can be helpful.42 Epiluminescence microscopy provides a practical and valuable diagnostic toll for the recognition of pigmented tumors of the skin.38,39
16.5 VALIDATION OF THE METHOD The dermatoscope is very useful in routine investigations of different diseases. The excellent stereoscopic image of the different observed skin areas are very realistically experienced by the investigator. The use of a simple instrument, for example as described by Braun-Falco et al.,2 is interesting but has draw backs because of the monocular vision and the poor illumination sources giving rise to very poor images. Epiluminescence microscopy has proved 85% accuracy compared with 60% accuracy for clinical evaluation. In a study by Curley et al.43 of 116 pigmented lesions, only 50% were diagnosed correctly by experienced dermatologists. To aid in diagnosis, Doppler sonography,44 three-dimensional nevoscopy,45 computerized digital imaging processing46 for primary melanomas, and immunoscintigraphy with radioactive-labeled monoclonal antibodies47 for metastatic disease have been developed. However, most of these methods are in the investigational stage, require experienced personnel, are not readily available, and are impractical for general use. With defined morphologic criteria, skin surface microscopy was shown to improve the diagnostic accuracy of most pigmented lesions significantly.25,26 In recent reviews25,26,48 the surface microscopic morphology of most pigmented lesions has been defined and illustrated, but the appearance of micropapular intracutaneous metastatic melanoma was not included.
122
Handbook of Non-Invasive Methods and the Skin, Second Edition
16.6 ARCHIVING AND FOLLOW-UP A new development is the use of the miniature video still camera with fiber optic illumination and optical lenses designed by various companies. They bring out the full structure of skin features, from intracutaneous examination to three-dimensional assessment of the surface of the skin. Full details are imaged quickly and easily to assure the highest level of diagnostic capability. The microscopic system comprises of a solid-state CCD-camera with illumination and camera controls housed in the compact supply box. The dermascope is compatible with a wide range of video recording and printing systems, overcoming problems associated with recording data over time. Hard copy prints of video-recorded information allow for progressive detailed study and analysis or presentation to teaching or research groups. The size and growth of skin melanomas, for example, can be progressed using the optional measurements scale located in the probe. Absolute measurements over time are easily recorded, unaffected by changing magnification and with no requirement for calibration. Optional accessories include the high- and low-magnification adaptors, which extend the capability of the microscopic system for more and fast investigation or where larger areas require referencing for achieving or recording purposes. A measurement graticule (1 cm in 1 mm divisions) easily locates on to the contact probe and is always in focus with the specimen.
REFERENCES 1. Knoth, W., Boepple, D., and Lang, W.H., Differential diagnostische Untersuchungen mit dem Dermatoskop bei ausgewahlten Erkrankungen, Hautarzt, 30, 7, 1979. 2. Braun-Falco, O., Stolz, W., Bilek, P., Merkle, T., Lanthaler, M., Das Dermatoskop: Eine Vereinfachung der Auflichtmikroscopie von pigmentierten Hautveränderungen, Hautzarzt, 41, 131, 1990. 3. Gibson, H.L., Further data on the use of infrared color film, J. Biol. Photogr. Assoc., 33, 155, 1965. 4. Aldis, A.S., and Marshall, R.J., Metastatic melanoma, detection by infrared recording, Med. Biol. Illus., 13, 2, 1963. 5. Gibson, H.L., Medical infrared color photography, II, Visual Med., 2, 43, 1967. 6. Rook, A., Wilkenson, D.S., and Champion, R.H., Principles of diagnosis, in Textbook of Dermatology, 4th ed., Rook et al., Eds., Blackwell Scientific, London, 1987, 61. 7. Jacobs, M.J.H.M., Breslau, P.J., Slaaf, D.W., and Reneman, R.S., Nomenclature of Raynaud’s phenomenon: a capillary microscopic and hemorheologic study, Surgery, 101, 136, 1987.
8. Marks, R. and Dawber, R.P.R., Skin surface biopsy: an improved technique for the examination of the horny layer, Br. J. Dermatol., 84, 117, 1971. 9. Cunliffe, W.J., Forster, R.A., and Williams, M., A surface microscope for clinical and laboratory use, Br. J. Dermatol., 90, 619, 1974. 10. Cohen, P.R., Examination of the male genitalia with an ophthalmoscope: a rapid and simple approach to the detection of penile, venereal warts, J. Am. Acad. Dermatol., 20, 521, 1989. 11. Pierard, G., Pierard-Franchiomont, C., and Dowlati, A., The skin surface biopsy in clinical and experimental dermatology, Rev. Eur. Dermatol., 4, 455, 1992. 12. Beagley, J. and Gibson, I.M., Changes in Skin Condition in Relation to Degree of Exposure to Ultra Violet Light, School of Biology, Western Australian Institute of Technology, Perth, 1980. 13. Daniell, H.W., Smoker’s wrinkles. A study in the epidemiology of ‘crow’s feet’, Ann. Intern. Med., 75, 873, 1971. 14. Holman, C.D.J., Armstrong, B.K., Evens, P.R., Lumsden, G.J., Dallimore, K.J., Meehan, C.J., Beagley, J., and Gibson, I.M., Relationship of solar keratosis and history of skin cancer to objective measures of actinic skin damage, Br. J. Dermatol., 110, 129, 1984. 15. Schamann, B. and Alter, M., Dermatoglyphics in Medical Disorders, Springer-Verlag, New York, 1976. 16. Verbov, J.L., Dermatoglyphic and other findings in alopecia areata and psoriasis, Br. J. Clin. Pract., 22, 257, 1968. 17. Gilje, O., Capillary microscopy in the differential diagnosis of skin diseases, Acta Derm. Venereal, 33, 303, 1953. 18. Davis, M.H. and Lawler, J.C., Capillary microscopy in normal and diseased human skin, Biol. Skin, 2, 1961. 19. Gilje, O., Kierland, R., and Baldes, E.J., Capillary microscopy in diagnosis of dermatologic diseases, J. Invest. Dermatol., 22, 199, 1954. 20. Ryan, T.J. and Wells, R.S., Hereditary benign telangiectasia, Trans. Rep. St. John’s Hosp. Derm. Soc. Lond., 57, 148, 1971. 21. Lehmann, P. and Kligman, A.M., In vivo removal of horny layer with formic acid, Br. J. Dermatol., 109, 313, 1983. 22. Shakespeare, P.G. and Carruth, J.A.S., Investigating the structure of the port-wine stain by transcutaneous microscopy, Lasers Med. Sci., 1, 107, 1985. 23. Jacobs, M.J.H.M., Jörning, P.J.G., Beckers, R.C.Y., Ubbing, D.T., Van Kleef, M., Slaaf, D.W., and Reneman, R.S., Foot salvage and improvement of microvascular blood flow as a result of epidural spinal cord electrical stimulation, J. Vasc. Surg., 12, 354, 1990. 24. Bahmer, F.A., Fritsch, P., Kreusch, J., Pehamberger, H., Rohrer, C., Schindera, I., Smolle, J., Soyer, W.P., and Stolz, W., Diagnostische Kriterien in der Auflichtmickroskopie, Hautarzt, 41, 513, 1990. 25. Pehamberger, H., Steiner, A., and Wolff, K., In vivo epiluminescence microscopy of pigmented skin lesions. I. Pattern analysis of pigmented skin lesions, J. Am. Acad. Dermatol., 17, 571, 1987.
Dermatoscopy
26. Steiner, A., Pehamberger, H., and Wolff, K., In vivo epiluminescence microscopy of pigmented skin lesions. II. Diagnosis of small pigmented lesions and early detection of malignant melanoma, J. Am. Acad. Dermatol., 17, 584, 1987. 27. Fritsch, P. and Pechlaner, R., Differentiation of benign from malignant melanocytic lesions using incident light microscopy, in Pathology of Malignant Melanoma, Ackerman, A.B., Ed., Masson, New York, 1981, 301. 28. MacKie, R., An aid to the preoperative assessment of pigmented lesions of the skin, Br. J. Dermatol., 85, 232, 1971. 29. Bahmer, F.A. and Rohrer, C., Rapid and simple macrophotography of the skin, Br. J. Dermatol., 114, 135, 1986, 30. Soyer, H.P., Smolle, H., Hoedl, S. et al., Surface microscopy: a new approach to the diagnosis of cutaneous pigmented tumours, Am. J. Dermatopathol., 11, 1, 1989. 31. Stolz, W., Bilek, P., and Langthalen, M., Skin surface microscopy, Lancet, 2, 864, 1989. 32. Goldman, L., Some investigative studies of pigmented nevi with cutaneous microscopy, J. Invest. Dermatol., 16, 407, 1951. 33. Haas, N., Ernst, T.M., and Stüttgen, G., Makrofotografie im transmittierenden Licht. Ein Beitrag zur horizontalen Strukturanalyse pigmentierter Hauttumoren, Z. Hautkr., 59, 985, 1984. 34. Haas, N. and Ernst, T.M., Makrophotographische Korrelate zur Histologie bei Precursor-Nävi und SSm in Anlehnung an das Schema von McGovern, Z. Hautkr., 61, 1535, 1986. 35. Hughes, B.R., Black, D., Srivastava, A., Dalziel, K., and Marks, R., Comparison of techniques for the non-invasive assessment of skin tumours, Clin. Exp. Dermatol., 12, 108, 1986. 36. Kreusch, J. and Rassner, G., Struktur-analyse melanozytischer Pigmentmale durch Auflichtmikroskopie, Hautarzt, 41, 27, 1990. 37. Soltani, K., Use of the ophthalmoscope in the clinical diagnosis of cutaneous pigmented lesions, J. Am. Acad. Dermatol., 17, 521, 1989.
123
38. Soyer, H.P., Smolle, J., Kerl, H., and Stettner, H., Early diagnosis of malignant melanoma using surface microscopy, Lancet, II, 8562, 1987. 39. Soyer, H.P., Smolle, J. Kresbach, H., Hoedl, S., Glavanovitz, P., Pachernegg, H., and Kerl, H., Zur Auflichtmikroskopie 410 von Pigmenttumoren der Haut, Hautarzt, 39, 223, 1988. 40. Steiner, A., Pehamberger, H., Binder, M., and Wolff, K., Pigmented Spitz nevi: improvement of the diagnostic accuracy by epiluminescence microscopy, J. Am. Acad. Dermatol., 27, 697, 1992. 41. Schultz, H., Auflichmicroscopischer score zur differential-diagnose dysplastischer Nävi, Hautarzt, 43, 487, 1992. 42. Pang, B.K. and Kossard, S., Surface microscopy in the micropapular cutaneous metastatic melanoma, J. Am. Acad. Dermatol., 5, 775, 1992. 43. Curley, R., Marsden, R., and Fallowfield, M., Diagnostic accuracy in the clinical evaluation of melanocytic lesions, Br. J. Dermatol., 119(Suppl. 33), 345, 1988. 44. Srivastava, A., Hughes, L., Woodcock, J., et al., Vascularity in cutaneous melanoma detected by Doppler sonography and histology, Br. J. Cancer, 59, 89, 1989. 45. Dhawan, A., Early detection of cutaneous malignant melanoma by three-dimensional nevoscopy, Comput. Methods Progr. Biomed., 21, 59, 1985. 46. Murray, A., Neill, S., Harland, C. et al., A new method for the quantification of physical change in unstable pigmented lesions using digital image processing techniques, Br. J. Dermatol., 199(Suppl. 33), 45, 1988. 47. Mechl, Z. and Bauer, J., The detection of metastases, in Cutaneous Melanoma Biology and Management, Cascinelli, N., Santinami, M., and Veronesi, U., Eds., Masson, Milan, 1990, 218. 48. Bahmer, F., Fritsch, P., Kreusch, J. et al., Terminology in surface microscopy, J. Am. Acad. Dermatol., 23, 1159, 1990.
Microscopy System for 17 Fiber-Optic Skin Surface Imaging Iqbal Sadiq and Tracy Stoudemayer S.K.I.N., Inc., Conshohocken, Pennsylvania
CONTENTS 17.1 Introduction ..........................................................................................................................................................125 17.2 Video Imaging .......................................................................................................................................................126 17.3 Optics .....................................................................................................................................................................126 17.3.1 Working Distance ......................................................................................................................................126 17.3.2 Light-Gathering Power ..............................................................................................................................126 17.3.3 Depth of Field............................................................................................................................................127 17.4 Fiber-Optics Cable.................................................................................................................................................127 17.5 Light Source ..........................................................................................................................................................127 17.6 Theory of Surface and Subsurface Imaging .........................................................................................................127 17.7 Hi-Scope System ...................................................................................................................................................128 17.7.1 Hi-Scope Components ...............................................................................................................................128 17.7.1.1 Optical Head...............................................................................................................................128 17.7.1.2 Control Unit................................................................................................................................128 17.7.1.3 Image Grabber Board.................................................................................................................128 17.7.1.4 Image Capture Software.............................................................................................................128 17.7.1.5 Vertical vs. Horizontal Illumination...........................................................................................128 17.7.1.6 Polarized Light Imaging.............................................................................................................129 17.7.2 Application of Hi-Scope in Skin Imaging ................................................................................................129 17.7.2.1 Imaging Skin with Vertical and Horizontal Illumination ..........................................................129 17.7.2.2 Photodamage ..............................................................................................................................129 17.7.2.3 Rosacea.......................................................................................................................................129 17.7.2.4 Acne............................................................................................................................................129 17.7.2.5 Scaliness .....................................................................................................................................130 17.7.2.6 Wound Healing...........................................................................................................................130 17.7.2.7 Hairs ...........................................................................................................................................130 17.8 Good Videomicroscopy Practices..........................................................................................................................133 17.9 Future .....................................................................................................................................................................133 References .......................................................................................................................................................................133
17.1 INTRODUCTION A fiber-optic microscopy system generally refers to videomicroscopes where optical illumination is provided through a fiber-optic cable and live video images are obtained by a charge coupled device (CCD) camera. The live images are displayed on a television (TV) monitor. This video signal can also be recorded directly to a video recorder. Advances in computer technology have made it possible to display digitized video-rate images on the
computer screen and enhance them in real time by selecting proper brightness, contrast, saturation, hue, and sharpness. From this live display, still images are captured and stored in the computer. The videomicroscopes of recent years are generally self-contained, small-size, handheld devices with flexible fiber-optic and electrical cables. The microscope head is brought to the object or lesion, compared to the standard laboratory microscope, where the specimen is brought to the microscope. The modular design with changeable 125
126
Handbook of Non-Invasive Methods and the Skin, Second Edition
lenses offers a wide range of magnifications. With the zoom lens, magnification is controlled while watching the live display. Inside the microscope head assembly, the light is guided to illuminate the surface either vertically or horizontally (i.e., from a narrow or wide angle to the surface). Polarized light attachment is also available. Videomicroscope is well suited for in vivo skin imaging. First, microscopic features invisible to the naked eye are revealed by videomicroscope at higher magnifications. Second, surface and subsurface structures are clearly viewed by manipulating the illumination optics. Third, still images as well as video can be recorded. Advanced image analysis and measurements can be performed using various software. Image transfer over the Internet is becoming increasingly popular, especially with reference to teledermatology. Digital images of skin lesions as well as histology slides can be sent from a remote location to a dermatology center for expert analysis and diagnosis. The videomicroscope system can add live video or streaming image clips to be sent over the Internet. One can zoom in and out of a lesion and survey the area in and around the lesion. The above attributes of a videomicroscope system, along with digital image processing, have made it an essential tool in the field of skin research and dermatology.1,2 Videomicroscopy has been successfully used in tracking acne lesions during treatment,3 assessing the sweat gland activity in the sole,4 and studying the efficacy of a product in protecting against particulate pollutants.5 Videomicroscopes can be adapted to perform enhanced subsurface imaging (dermatoscopy) by using oil or gel as optical coupling. Using this technique, malignant melanocytic lesions,6 microvasculature of basal cell carcinoma, and actinic keratosis7 have been studied successfully. Laser treatments of port-wine stain (capillary vascular malformation)8,9 have also been evaluated.
17.2 VIDEO IMAGING In an effort to transmit live images over a distance, Zworykin10 in 1935 developed a vacuum tube camera that he called iconoscope. A mosaic of photosensitive picture elements (pixels) was deposited on a plate that was scanned in a raster pattern by an electron beam inside a glass vacuum tube. Electrical signal, which was proportional to the amount of light on each pixel, was collected over time, line by line. This analog signal was transmitted over the wires. This concept was later developed into TV systems. In the American TV system (National Television Systems Committee [NTSC]) 525 lines are scanned 30 times a second, while in European TV (Phase-Alternating Line [PAL] system) 625 lines are scanned 25 times per second. Advances in solid-state physics led to the development of solid-state cameras. These are lightweight and compact devices containing a
flat, rectangular chip that does not need vacuum. The most common cameras are composed of CCDs in which a large array of photodiode picture elements are deposited on a rectangular silicon chip. At each pixel point electrical charge accumulates, proportional to the amount of light. The electrical charges are read line by line, resulting in an analog signal that is transmitted over the wire. This video signal can be viewed directly on a TV monitor or can be digitized and sent to a computer. A large variety of digitizing boards, sometimes called image grabbers, are available. The CCDs come in a range of physical sizes, conventionally named in inches, but actual dimension is given in millimeters. Some common sizes are 1/3 inch (4.8 × 3.6 mm), 1/2 inch (6.4 × 4.8 mm), 2/3 inch (8.8 × 6.6 mm), and 1 inch (12.8 × 9.6 mm). The pixel density varies as well. A detailed technical description of videomicroscopes is given by Inoue and Spring.11 To capture a reasonably good-quality image of skin, the CCD pixel size should be 768 × 480 or higher. The digitized image on the computer screen should be 640 × 480 pixel or better. For video-rate imaging, the pixel size is limited by the speed and efficiency of the image processing electronics currently available.
17.3 OPTICS The handheld videomicroscope is usually modular with changeable lenses and light sources. Magnifications from less than 1× to 1000× can be used. However, in practice, objectives above 600× are difficult to use because of motion artifacts, when skin imaging is done in vivo. The quality of image and ease of operation depend on working distance, light-gathering power, and depth of field.
17.3.1 WORKING DISTANCE In a videomicroscope the working distance, i.e., the distance between the objective lens and the specimen, is relatively large, ~5 cm or more (except for high magnifications), compared to the standard laboratory microscopes where this distance is less than 1 cm. One needs this space to arrange light sources, illuminating the specimen from various directions.
17.3.2 LIGHT-GATHERING POWER The light-gathering power of an objective lens depends on its numerical aperture (N.A.), which in general terms is defined as N.A. = n Sin u where n is the refractive index of the medium between the lens and the object and u is the half angle of the cone of light entering the lens (angular aperture). This is shown
Fiber-Optic Microscopy System for Skin Surface Imaging
127
n2 u
n1
Optical axis
Optical fibre
Front lens
FIGURE 17.1 Angular aperture of the front lens.
in Figure 17.1. In the case of videomicroscopes the medium between the objective lens and skin is air (n = 1), and therefore N.A. is the ratio of the diameter of the front lens to the focal length.
17.3.3 DEPTH
OF
i
FIELD
Depth of field refers to the range of focal distances in a field of view. In videomicroscopy relatively large depths of field are used. This is useful for imaging skin since the surface is not flat.
17.4 FIBER-OPTICS CABLE A fiber-optic cable (optical waveguide) is essentially composed of a bundle of a large number of fine fibers of either glass, quartz, or plastic, with the ends embedded into molds of various shapes. These can be circle to circle, circle to rectangle, circle to line, or circle to ring. Light transport inside a single fiber is due to total internal reflection from the fiber surface. Inside the fiber, light striking the fiber surface at any angle greater than the critical angle from the normal to the surface will be reflected. This critical angle depends on the ratio of the refractive indices of the fiber material to its surrounding medium. The critical angle is described by Sin ic = n2/n1 where ic is the critical angle, n1 is the index of refraction of the fiber, and n2 is the index of refraction of the surrounding medium (Figure 17.2). Light entering the fiber travels in helical paths and emerges from the other end in a diverging cone with relatively small loss. However, significant loss of light can occur in a number of ways, e.g., defects in the fiber material, rough regions on the surface producing scattering centers, and light entering the cable at a wide angle. A good-quality fiber-optic cable would have fibers with very smooth surface, optically clear material, and highly
FIGURE 17.2 The principle of total internal reflection inside a fiber. A light ray striking the inner surface of the fiber at an angle i is reflected without loss if angle i is greater than the critical angle ic.
polished and flat ends, perpendicular to the axis of the fiber.
17.5 LIGHT SOURCE A variety of light sources are available, e.g., tungstenhalogen, metal-halide, and xenon short-arc lamps. The most common light source is a tungsten-halogen lamp, which is usually embedded in a conical reflector. One end of the fiber-optics cable is mounted in front of this light source; the other end is attached to the microscope head. In some setups, other optical components such as focusing lens, shutter, and iris diaphragm are inserted between the tungsten-halogen lamp and the fiber-optic cable.
17.6 THEORY OF SURFACE AND SUBSURFACE IMAGING Skin is a translucent material that partially reflects light from the surface (specular reflection) and partially diffuses light and then reflects it (diffuse reflection). Reflected light shows an enhanced view of the surface structure; diffusereflected light shows subsurface erythemas and pigmentation. Vertical illumination is used to preferably show surface reflections, revealing the topography of skin, follicular lesions, acne, etc. The horizontal lighting reveals the features of deeper levels such as pigment, erythema, capillary loops, etc. The theory of imaging of deeper levels by horizontal lighting is different from cross-polarized light. Since for specular reflection the angle of incidence is equal to the angle of reflection, light striking a surface at a narrow angle is mainly reflected at a narrow angle; hence, most of the light reflected off the shining surface leaves the field of view. Only the portion diffused into the skin is collected by the imaging lens. Therefore, the oily shine of the skin surface is reduced, giving way to visualization of deeper levels. It may be noted that horizontal light also strongly illuminates the surface scales and hairs, marring the subsurface view. In the case of polarized light imaging, parallel polarization enhances the surface structure while cross-polarization removes the surface shine (specular reflection) and shows the subsurface features.
128
Handbook of Non-Invasive Methods and the Skin, Second Edition
17.7.1 HI-SCOPE COMPONENTS
Basic layout of the videomicroscope
17.7.1.1 Optical Head
Video probe
Fibre-optic cable
Camera cable
Light source Controller and image processor
Video display
S-video
S-video Image grabber Computer monitor
PCI Monitor cable
Computer
FIGURE 17.3 Block diagram of the Hi-scope system.
Camera cable
Fibreoptic cable
Vertical Lighting
Camera cable
Fibreoptic cable
Horizontal lighting
The Hi-scope head has a 1/2 inch CCD, with a physical size of 6.4 × 4.8 mm and the effective image size of 682 horizontal × 492 vertical pixels. It has a zoom lens with a selectable range of 20× to 100× magnification, as well as a selection of single objective lenses with magnification values of 20×, 50×, 80×, 100×, 160×, 250×, 400×, 800×, 1000×, and 1600×. For imaging large areas of skin, a macrozoom lens is also available with magnifications above 1× to much lower than 1×. 17.7.1.2 Control Unit The main control unit contains the camera electronic controls and the light source. The optical head is connected to this unit by the camera cable and fiber-optics cable. The amount of light delivered to the optical system can be adjusted from the control unit in two ways: (1) by adjusting the distance of the fiber-optic cable from the light source by turning the ring at the entry point of the fiberoptic cable or (2) by changing the lamp power supply by turning the lamp power knob. It may be noted that turning down the lamp power also changes the color temperature, which must then be corrected either electronically or by the software. The camera controls available on the main control unit are white balance, light level, and video level. 17.7.1.3 Image Grabber Board
FIGURE 17.4 Vertical vs. horizontal lighting system.
17.7 HI-SCOPE SYSTEM We will describe in detail a videomicroscope, Hi-scope model KH-2200 (Hirox Co., Ltd., Tokyo, Japan), which is commonly used in dermatological research laboratories. Hi-scope provides exquisite details of the skin’s surface. Manipulation of the lighting allows enhancement of microstructures such as scales, telangiectasia, dyspigmentations, erythema, fine lines and wrinkles, etc. The basic layout of the Hi-scope system is shown in Figure 17.3. The videomicroscope head is connected to the control box by electrical and fiber-optic cables. The video signal is transferred to a TV monitor through an Svideo connector. Another S-video connector takes this signal to an image grabber board (Flashpoint by Integral Technologies, Inc., Indianapolis, IN), installed inside the computer on a PCI socket. A program, Fpg32, was used to display the image on the computer screen.
The image grabber board Flashpoint 128 with 4 megabytes of video memory was used. This board is also a video display driver for the computer monitor. The original video driver of the computer was completely removed. The board received the video signal through an S-video connector. The features include choice of 24/16/8 bit video digitization, with pixel resolutions of 640 × 480 for NTSC and 760 × 570 for PAL system. 17.7.1.4 Image Capture Software The image displayed on the computer monitor was controlled by the software Fpg32. When an S-video connector is used, adjustment of brightness, contrast, saturation, and hue can be done in real time. When a BNC cable is used to collect composite video signal, sharpness can also be controlled. During a series of image captures, these parameters are kept constant. 17.7.1.5 Vertical vs. Horizontal Illumination A prominent feature of the Hi-scope system is that the direction of illumination can be changed from fully vertical to fully horizontal using a special attachment. Here vertical direction means illumination from a large angle
Fiber-Optic Microscopy System for Skin Surface Imaging
to the skin surface (~85˚) and horizontal direction means narrow angle to the surface (~15˚). The vertical illumination is used to see the surface features like fine lines and wrinkles, edematous lesions, vesicles, sweat droplets, etc. The horizontal illumination shows the subsurface features like capillaries and pigmented lesions. The horizontal illumination also enhances the visualization of scales. The scales become a secondary source of light due to strong scattering and show in the image as bright structures. View of vellus hairs on the skin surface is also enhanced.
129
Vertical vs horizontal illumination A
17.7.1.6 Polarized Light Imaging Another attachment available with the Hi-scope system allows cross- and parallel polarized images to be captured. The parallel polarized image shows surface features like fine lines, wrinkles, and acne lesions. In the cross-polarized mode subsurface features like pigmented lesions, capillaries, etc., are clearly shown. The difference between the cross-polarized imaging and horizontal light imaging is that in horizontal light imaging, appearance of scales and hairs impairs the subsurface view. In the cross-polarized imaging the effect of scales and hairs is minimal.
1 mm
B
17.7.2 APPLICATION OF HI-SCOPE IN SKIN IMAGING The Hi-scope system is increasingly being used in skin research and dermatology. Images are captured and stored before and after treatments to assess the changes that occurred. Shapes, sizes, and density of lesions are measured using image analysis software like Optimas, SIS, and Image Pro Plus. We present here some of the applications of the Hi-scope in skin research. 17.7.2.1 Imaging Skin with Vertical and Horizontal Illumination An example of vertical vs. horizontal light imaging is shown in Figure 17.5. Using the zoom lens with 40× magnification setting, vertical illumination shows a shiny surface (A). The same site shows erythema when horizontally illuminated (B). At a higher magnification (400×) individual capillaries are visible with horizontal illumination (Figure 17.6). 17.7.2.2 Photodamage Lifetime, cumulative exposures to sun result in dyspigmented, wrinkled, rough, and dull skin. We have compared dorsal forearms to volar forearms, with the general assumption that the dorsal surface is exposed to sun more than the volar surface. Images of the dorsal and volar forearms of a 62-year-old, grossly sun-damaged, female subject were obtained by Hi-scope zoom lens set at 60×.
1 mm
FIGURE 17.5 Oily facial skin shows surface features when vertically illuminated (A) and subsurface erythema when horizontally illuminated (B). Also note the enhanced view of hairs and follicular casts with horizontal illumination. Zoom lens at 40×.
The difference between dorsal and volar surface is shown in Figure 17.7. 17.7.2.3 Rosacea Facial images of a rosacea patient compared to a normal subject are shown in Figure 17.8. These images were obtained using the macro-z lens with an approximate magnification of 1× and illumination with an external fiberoptic ring light. The rosacea patient shows a large number of telangiectatic blood vessels compared to the normal skin subject. At 40× magnification (Figure 17.9) the shapes and sizes of the blood vessels are clearly observed. 17.7.2.4 Acne Hi-scope is an excellent tool to study acne, particularly the smaller lesions. Evolution and resolution of specific
130
Handbook of Non-Invasive Methods and the Skin, Second Edition
Imaging of capillary loops A A
100 μm 1 mm
B B
100 μm
FIGURE 17.6 Vertically (A) vs. horizontally (B) illuminated images of forehead at a higher magnification (400×). Capillary loops are clearly visible in the horizontally illuminated image. In live video, blood flow can also be observed.
1 mm
lesions can be observed for several weeks. Figure 17.10 shows the resolution of a pustular acne lesion in 8 days. Vertical lighting was used with 20× magnification. Imaging of comedones with polarized light is shown in Figure 17.11. At 40× magnification, the parallel polarized image shows the surface features of the comedones while the cross-polarized image shows the pigment distribution.
FIGURE 17.7 The volar surface (B) is slightly scaly. The primary glyphic lines, which are deeper and wider, are well defined. The secondary furrows that intersect with the primary lines are very clear. Compared to this, the dorsal forearm (A) shows a disrupted surface with absence of glyphic lines, probably obscured by overlying thick scales, a common feature of photodamaged skin. Hyperpigmented patches are also typical of photodamaged skin.
17.7.2.5 Scaliness
example of an experimental wound, produced by a 3-mmdiameter circular biopsy punch. The images were recorded 9 and 16 days postbiopsy, showing a dramatic reduction in size during this phase of wound healing (Figure 17.13).
Scaliness is enhanced in Hi-scope images when horizontal illumination is used. Dry skin treatments can be studied very well. Figure 17.12 shows dramatic reduction in scaliness on a dry leg after 2 weeks of Aquaphor application. 17.7.2.6 Wound Healing A very good use of the videomicroscope is in wound healing assays. A single wound can be followed for several weeks, recording the day-by-day changes. The shape and size of the wound and formation of crust can be assessed using the image analysis software. We present here an
17.7.2.7 Hairs Videomicroscopy is extensively used in hair research. The hair surface, distribution/density, and growth can be studied. Also, effects of hair products can be evaluated. Figure 17.14 shows the angle of shaving cut on the beard hairs. The image was captured by a 400× lens with horizontal illumination.
Fiber-Optic Microscopy System for Skin Surface Imaging
131
Normal vs Rosacea cheeks
5 mm
5 mm Rosacea
Normal
FIGURE 17.8 The cheek of the rosacea patient (right) shows redness and a large number of telangiectatic blood vessels compared to the cheek of a normal subject (left). Macro-z lens at 1× with ring light. Telangiectatic blood vessels
I mm
I mm Rosacea
Normal
FIGURE 17.9 The blood vessel network in the rosacea patient is clearly shown at 40× magnification (right). The normal skin subject does not show any large blood vessel (left). Imaging by zoom lens at 40× with horizontal illumination also shows vellus hairs and follicular casts. Resolution of acne lesion Day 1
Day 8
5 mm
5 mm
FIGURE 17.10 Pustular acne lesion (left) cleared in 8 days (right). The dotted circle shows the remnants of the lesion. Vertical lighting with 20×.
132
Handbook of Non-Invasive Methods and the Skin, Second Edition
Parallel vs cross polarized imaging of comedones Parallel polarized Cross polarized
1 mm
1 mm
FIGURE 17.11 A relatively large, open comedone surrounded by smaller, closed comedones. Parallel polarized light shows enhanced surface features, while cross-polarized light shows distribution of pigment. Polarized light with 40×.
Dry leg treatment Pre-
1 mm
Post
1 mm
FIGURE 17.12 Two weeks of Aquaphor treatment clears the scales on a dry leg. Horizontal lighting with 40×.
Wound healing
I mm
I mm
FIGURE 17.13 Reduction in the size of wound from 9 days (A) to 16 days (B) postwound creation.
Fiber-Optic Microscopy System for Skin Surface Imaging
133
the cross-polarized lighting system should be used instead of horizontal lighting.
17.9 FUTURE
100 μm
FIGURE 17.14 Angle of the shaving cut shown on the beard hairs. Imaging with 400× lens and horizontal illumination.
17.8 GOOD VIDEOMICROSCOPY PRACTICES The Hi-scope system can be successfully used in skin research and dermatology if the right lens and illumination system are chosen and care is taken to avoid pitfalls. Before capturing a series of images on the computer, using the software Fpg32, a test image should be obtained and adjustment of brightness, contrast, saturation, and hue should be performed. These values should be recorded for future use during a similar imaging session. More often than not, the default setup results in poor image quality. It is a good idea to capture images of an optical grid with lenses of different magnifications and store them in the computer memory for the purpose of spatial calibration. In our laboratory we have used a high-quality optical grid made of glass with a unit dimension of 0.25 mm. For lower magnifications we have used a scale with a unit dimension of 2 mm. Skin surface should be free from topical products during imaging. Surface imaging can be done by both a vertical lighting system and a parallel polarized lighting system. The amount of light is slightly less in the polarized system, particularly with higher magnifications. Imaging specific lesions, e.g., acne or blister wound, one must select appropriate magnification. Often it is a good idea to capture a series of images with decreasing magnifications. This will give information on the area surrounding the specific lesion. The image quality can be marred by surface scales and hairs when using a horizontal lighting system for subsurface imaging of erythemas and pigmentation. The scales and hairs scatter light, strongly degrading the quality of the image. When a lot of scales or hairs are present,
Recent advances in digital imaging, video systems, optics, electronics, and computer technology have made it possible to enhance the capabilities of videomicroscopes. Quartz optics and quartz fibers can be used for ultraviolet light imaging as well as ultraviolet light-induced fluorescence imaging. Higher-speed and more efficient image grabber boards can increase the resolution and quality of the captured image. Recently the high-definition television (HDTV) system, with 1080 lines of video signal, is beginning to appear in the market as television sets and video recorders. Compared to the NTSC video system (525 lines), the HDTV system should produce much higher resolution images. However, we have to wait for the image grabber boards to be developed for this. Another area of recent advancement is the digital video systems (digital video recorders). Two standards, developed by Motion Picture Expert Group (MPEG), are already in use: MPEG2 and MPEG4. These systems can someday be utilized in videomicroscopy to view remarkably high quality images. Digital imaging at the rate of 15 frames per second have already been developed, e.g., Hi-scope model KH3000. This produces much higher resolution images of 2.11 megapixel (1600 × 1200 pixel) size. Recent advancement in the field of digital photography, which produces still images, has led to the development of imaging chips of 6 to 12 megapixels. The challenge for the future is to develop high-speed electronic circuitry to process these images at speeds higher than 30 frames/second.
REFERENCES 1. Sadiq, I., Kolbe, L., Pagnoni, A., Rizova, E., Stoudemayer, T., and Kligman, A.M., High resolution videomicroscopy to record ultra-fine in-vivo changes in experimental and clinical dermatology, in AAD 56th Annual Meeting, February 27–March 4, 1998, p. 287. 2. Stolz, W., Schiffner, R., Pillet, L., Vogt, T., Harms, H., Schindewolf, T., Landthaler, M., and Abmayr, W., Improvement of monitoring of melanocytic skin lesions with the use of a computerized acquisition and surveillance unit with a skin surface microscopic television camera, J. Am. Acad. Dermatol., 35, 202–207, 1996. 3. Rizova, E., Pagnoni, A., Stoudemayer, T., Poncet, M., and Kligman, A.M., Polarized light photography and videomicroscopy greatly enhance the capability of estimating the therapeutic response to a topical retinoid (adapalene) in acne vulgaris, Cutis, 68(4 Suppl.) 25–33, 2001.
134
Handbook of Non-Invasive Methods and the Skin, Second Edition
4. Nishiyama, T., Sugenoya, J., Matsumoto, T., Iwase, S., and Mano, T., Irregular activation of individual sweat glands in human sole observed by a videomicroscopy, Autonomic Neurosci Basic Clin, 88, 117–126, 2001. 5. Armand-Stussi, I., Basocak, V., Pauly, G., and McCaulley, J., An interesting source of moringa oleifera: active ingredients for skin and hair care, Personal Care, May 2003. 6. Braun, R.P., Meier, M.L., Pelloni, F., Ramelet, A.A., Schilling, M., Tapernoux, B., Thürlimann, W., Saurat, J.-H., and Krischer, J., Teledermatoscopy in Switzerland: a preliminary evaluation, J. Am. Acad Dermatol., 42, 770–775, 2000.
7. Newel, B., Bedlow, A.J., Cliff, S., Drysdale, S.B., Stanton, A.W.B., and Mortimer, P.S., Comparison of the microvasculature of basal cell carcinoma and actinic keratosis using intravital microscopy and immunohistochemistry, Br. J. Dermatol, 149, 105–110, 2003. 8. Motley, R.J., Arch. Dermatol., 133, 921–922, 1997. 9. Sivarajan, V. and Mackay, I.R., The depth measuring videomicroscope (DMV): a non-invasive tool for the assessment of capillary vascular malformations, Lasers Surg. Med., 34, 193–197, 2004. 10. Zworykin, V.K., The iconoscope: a modern version of the electric eye, Proc. IRE, 22, 16–32, 1934. 11. Inoue, S. and Spring, K.R., Video Microscopy: The Fundamentals, 2nd ed., Plenum Press, New York, 1997.
Assessment of Pigment 18 Automated Distribution and Color Areas for Melanoma Diagnosis Stefania Seidenari and Giovanni Pellacani Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
Costantino Grana Department of Computer Engineering, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 18.1 Introduction............................................................................................................................................................135 18.2 Clinical Assessment of Colors in Dermoscopic Images.......................................................................................135 18.3 Computer Assessment............................................................................................................................................136 18.3.1 Computer Assessment of Dark Areas .......................................................................................................136 18.3.2 Automated Assessment of Color Type and Number in MMs and Nevi ..................................................138 18.3.3 Automated Description of Color Area Features .......................................................................................140 18.3.4 Automated Assessment of Colors in Atypical Nevi .................................................................................140 18.3.5 Colors in Image Blocks.............................................................................................................................141 18.4 Conclusion .............................................................................................................................................................142 References .......................................................................................................................................................................142
18.1 INTRODUCTION Reports assessing accuracy in the clinical diagnosis of malignant melanoma (MM) show sensitivity values of 65 to 80%, depending on the dermatologist’s experience.1,2 Surface microscopy or dermoscopy, providing magnified images with access to subsurface structures, achieves a 10 to 27% higher sensitivity than clinical diagnosis by the naked eye.3 Thus, dermoscopic images, especially when enabling inspection of their digital version, provide an important input to clinical diagnosis. As a further step toward diagnostic performance, a number of diagnostic checklists have been proposed, which specify dermoscopic visual features associated with malignant lesions, including asymmetry, irregularity of border outline, color variegation, and presence of specific patterns.4–9 However, the interpretation of dermoscopic images, especially of the visual features, is often inconsistent both within and among observers.10,11 Computer-based image interpretation, offering objectivity and repeatability, seems to be an attractive proposition.
Many computer-based techniques apply standard measures representing variations of the ABCD rules to quantify visual features and then subject the derived measurements to statistical analysis to identify numerical characteristics specific for MM.12–28
18.2 CLINICAL ASSESSMENT OF COLORS IN DERMOSCOPIC IMAGES The assessment of colors is essential for MM diagnosis, both for pattern analysis on dermoscopic images and when employing semiquantitative methods. A blue, red, or white color component is more often present in MMs with respect to nevi.10 Therefore, the detection of these colors within an image may have great diagnostic importance. Mnemonics such as the ABCD of dermoscopy, which has been advocated as an aid to diagnosis, comprise the color count in the final score.4 In this rule the presence or absence of red, blue-gray, white, dark brown, light brown, and black in melanocytic lesions gives rise to a score varying between 0.5 and 3 (on a total possible score 135
136
Handbook of Non-Invasive Methods and the Skin, Second Edition
ranging from 1 to 8.9). Blue and gray are considered together in the ABCD rule, whereas according to Menzies, they are considered separately for the color count and together for the assessment of the pepper-like pattern.5 Moreover, according to Menzies’ method, the presence of a single color (black, gray, blue, dark brown, tan, and red) is regarded as a negative feature, enabling the exclusion of the diagnosis of MM, whereas the presence of five or six of the latter colors (white is not scored as a color) represents a positive feature.5 The 7-point checklist considers the bluish white color for the assessment of the veil, blue for the description of pepper-like granules and white associated with scar-like depigmentation.6 In the modified ABC point list of dermoscopy by Blum et al.,8 the presence of three or more colors contributes to 20% of the final score for an MM. MacKie et al.7 observed that the most powerful identifying feature of lesions subsequently shown on pathological examination to be MM is the presence of three or more colors seen in the lesion on dermoscopy. In fact, malignant lesions frequently show more than three colors, whereas in nevi, three or less colors are usually observed.10 Greene et al.29 stated that the appearance of new black pigmentation in a dysplastic nevus was the best predictor of the presence of early MM. Recently, we clinically assessed color type and distribution in melanocytic lesion (ML) images to describe color characteristics in dermoscopic atypical nevi and to analyze the importance of color parameters for the diagnostic process leading to discrimination between benign and malignant lesions, diagnosis, and decision to excise an ML.30 Overall, 603 images referring to MLs were retrospectively subdivided into four groups according to diagnosis performed exclusively by dermoscopy by two trained dermatologists; 288 of these images corresponded to lesions with a clearly benign aspect. The second group included 101 lesions that were considered benign, but eligible for a follow-up examination. These two groups together comprised the category of lesions with no indication to excise on a dermoscopic basis (nonexcision category). The third group comprised 93 lesions with atypical morphological features, whereas the last group included 121 lesions with the dermoscopic diagnosis of MM. The two latter groups together formed the excision category. For each image the presence of black, dark brown, light brown, gray-blue, red, and white was clinically assessed on the dermoscopic digital images and the number of colors counted. Moreover, the color distribution symmetry was evaluated by considering the presence of at least one asymmetrically distributed color in the lesion along at least one axis. In this study we showed that the mean number of colors per lesion increased progressively from the first group (clearly benign lesions, no follow-up), which showed a mean number of colors of 2.94, to group 4,
(dermoscopic MMs) showing 4.07 colors per lesion. Nonexcision lesions showed 2.99 colors per lesion, whereas excision lesions showed a significantly higher number of colors (3.61). The mean number of colors per lesion was 4.15 for MMs and 2.99 for nevi. Atypical nevi did not show a higher number of colors with respect to clearly benign lesions; however, the black and blue-gray components were more frequently found, whereas in dermoscopic MMs, the black, red, white, and blue-gray components were significantly more frequent than in atypical nevi. Furthermore, with respect to clearly benign lesions, atypical ones, eligible for excision according to dermoscopic criteria, more frequently showed an asymmetric distribution of at least one color along at least one axis of the lesion. In conclusion, even if absolute values (color type and number) are strictly dependent on acquisition technique and image type, we confirmed that clinical assessment of color parameters on dermoscopic images represents an important element for the diagnosis of MM and atypical nevus, and strongly influences the decision to excise and the management of an ML.
18.3 COMPUTER ASSESSMENT 18.3.1 COMPUTER ASSESSMENT
OF
DARK AREAS
In contrast with common nevi, which generally show a homogeneous and regularly distributed pigmentation, brown to black pigment areas irregular in shape or asymmetrically distributed are frequently observable in MMs.5,31–35 Different parameters, based on the measurement of brightness values, have been employed by image analysis programs for the quantification of the overall darkness of the lesion.21,22,24 The concept of darkness is not absolute, but is related to human perception, which can be influenced by the overall pigmentation of the lesion as well as by the color of the skin. In order to explore the influence of the evaluation of dark areas in the diagnostic judgment of melanocytic lesions, we compared two automatic methods for their identification in videomicroscopic ML images.36 The first one enables the identification of absolute dark areas, defined as areas that are darker than the skin. To this purpose, the mean brightness of the skin is first evaluated as a reference level. Subsequently, the ratio between the brightness of each lesion pixel and the skin’s mean brightness is assessed. If this value is lower than a given threshold, the pixel is considered dark. Since the method is based on absolute thresholds, a lesion may or may not have an absolute dark area. Relative dark areas identify the darkest lesion area with respect to the overall brightness of the lesion, regardless of the intensity of its pigmentation. In order to identify the darkest area inside the lesion, the color histogram is
Automated Assessment of Pigment Distribution and Color Areas for Melanoma Diagnosis
137
TABLE 18.1 Mean and Standard Deviation of the Parameters Calculated for Absolute and Relative Dark Areas on 339 ML Images Comprising 113 Melanomas and 226 Melanocytic Nevi Absolute Dark Area Parameters
MMs
AREA DIST-BAR MID REG-MID EXT REG-EXT SYM-MIN
0.327 0.101 0.342 0.176 0.084 0.061 0.565
± ± ± ± ± ± ±
Relative Dark Area
NEVI
0.261 0.084 0.298 0.115 0.125 0.066 0.296
0.246 0.057 0.248 0.099 0.034 0.020 0.667
± ± ± ± ± ± ±
0.224 0.047 0.284 0.084 0.080 0.032 0.255
MMs 0.262 0.093 0.248 0.204 0.021 0.027 0.593
± ± ± ± ± ± ±
0.045 0.057 0.081 0.081 0.036 0.037 0.218
Parameter Abbreviation
Range
Parameter Description
AREA
(0–1)
DIST-BAR
(0–1)
Number of pixels belonging to the region divided by the area of the lesion (proportion of the dark area in respect to the lesion area) Distance between the color region centroid and the lesion centroid, divided by the major axis length
MID, EXT
(0–1)
REG-MID REG-EXT
(0–1)
SYM-MIN
(0–1)
They correspond to the mean number of pixels per sector for each zone, describing the dark area involvement inside the middle (MID) and external (EXT) zones, respectively They correspond to the standard deviation of the number of pixels per sector for each zone, describing the regularity of dark area distribution inside the middle (MID) and external (EXT) zones, respectively Three values, one for each zone, are first computed separately for major and minor axes as the absolute difference between the number of points located on each side; in order to obtain a single descriptor for each axis, the symmetry values of three zones are summed together and normalized by the region area; SYM-MIN represents the lowest symmetry values along major or minor axes thus extracted
divided into four zones following an iterative process, which begins by dividing the gray-level histogram along the median value and proceeds selecting the largest (as number of gray levels) of the two halves. The process is repeated until the number of required zones is obtained. The zone corresponding to the lowest gray levels is considered the relative dark area and is present in every single lesion. Subsequently, in order to calculate numerical descriptors of the aspect and distribution of the identified areas, the lesion is divided into three zones (external, middle, and internal) and into eight sectors per zone. Dark area extension, color distribution balance, color distribution
NEVI 0.253 0.046 0.224 0.143 0.007 0.008 0.764
± ± ± ± ± ± ±
0.023 0.028 0.073 0.055 0.011 0.014 0.129
Parameter Meaning Dark area extension Color distribution balance Dark area distribution
Regularity of the dark area distribution
Symmetry of the dark area distribution
density, dark area distribution, regularity, and symmetry of dark area distribution, and percentage of lesions presenting the dark area are then calculated. When tested on a set of ML images to evaluate the influence of these descriptors in distinguishing between MMs and nevi, the extracted features showed that absolute dark areas were more frequently found in MMs with respect to melanocytic nevi, and were more unevenly and irregularly distributed, often involving the external zone. On the other hand, relative dark areas appeared more asymmetrically distributed and less aggregated in MMs with respect to melanocytic nevi (Table 18.1).
138
Handbook of Non-Invasive Methods and the Skin, Second Edition
18.3.2 AUTOMATED ASSESSMENT OF COLOR TYPE AND NUMBER IN MMS AND NEVI As regards automatic classification of MLs and computer diagnosis of melanoma by means of colors, most methods for image analysis have described colors in lesion images by color descriptors, which are statistical parameters calculated from different color channels, like average value and standard deviation of the RGB or HSV color channels. Binder et al.22 considered the number and range of different colors in a reduced color model, which represented one of the most important variables for automatic classification. Red, green, and blue average decile and quartile values were employed by us and were included in the equation for discriminant analysis classification.21,23 The image analysis program employed by Andreassi et al.24 evaluates red, green, and blue averages, quartile and decile values, and color islands, including extension and imbalance of so-called peripheral dark regions — dark green, green dominant, blue-gray — and transition areas. In a recent paper by Kahofer et al.,37 mean and standard deviation, skewness, kurtosis, minimum, and maximum were calculated in the intensity image and in red, green, and blue images for each element. Umbaugh et al.38 proposed a color variance parameter in the L*a*b* color space and detected color variegation using an analytical color technique. Day and Barbour39 employed a feature measuring the average color difference between the skin and the lesion (relative chromaticity green). This feature was previously used by Ercal et al.14 Cotton and Claridge40 employed an optical model of the skin to interpret the colors occurring in a lesion. They found that all normal skin colors lie on a two-dimensional surface patch within a three-dimensional color space. Ganster et al.26 adopted an original approach. Besides minimum, maximum, average, and variance of the intensity and hue channels, they considered 15 significant colors obtained by the median cut color quantization algorithm of Heckbert41 and calculated the percentage of the skin lesion containing absolute shades of reddish, bluish, grayish, and blackish areas, and the number of those color shades present within the skin lesion in dermatoscopy images.26 MM detection has also been performed based on probability analysis for three classes of colors (benign, MM, and other colors) found from relative color histograms.42 In fact, certain colors are associated with MLs, whereas occasional patches of red, white, and blue are more specific for malignant lesions. A data-driven relative color histogram analysis technique was investigated for determining colors characteristic for MMs, i.e., MM colors. For that study relative colors were used to compensate for skin lesion images that were digitized from slides. Subtracting the average surrounding skin color from the lesion color, to correct for variations in image acquisition and
skin pigmentation, relative colors were obtained. Color histogram analysis using relative colors was then performed to identify MM colors. Subsequently, the number of pixels within a lesion that were labeled as MM colors was divided by the lesion area, and the percentage of MM color feature was computed for individual skin lesions. The color histogram analysis technique was subsequently extended to evaluate colors in different regions of the skin lesion.43 Moreover, the color clustering ratio, defined as the ratio of the total MM color eight-connected neighbors of MM-colored pixels within the skin lesion to the total number of eight-connected neighbors for all MMcolored pixels within the lesion, was assessed. The authors concluded that, whereas the latter parameter has a low diagnostic power, the region closest to the skin lesion boundary contains the greatest color discrimination information for lesion screening. Most of the statistical parameters employed for color description were chosen primarily for computational convenience; however, they do not model the methods employed by the human brain for the clinical diagnostic procedure, where areas that share similar colors (representing a set of pixels consisting of a mixture of RGB components) are considered together as homogeneous color areas and simply called red, blue-gray, white, dark brown, light brown, black, etc. We therefore developed an automatic method for assessment of colors in melanocytic lesion images mimicking the human perception of lesion colors, based on the identification of a color palette comprising the more representative colors perceived by the human eye in a melanocytic lesion.44 Our palette comprised six color groups (black, dark brown, light brown, gray-blue, red, and white) and was created interactively employing ML images, unequivocally showing black, dark brown, light brown, red, white, and blue-gray color components (Figure 18.1). Square color patches of arbitrary size were selected manually on different regions of interest pertaining to different images, and an average RGB color was extracted by the program. To avoid the selection of too many color patches, a tolerance factor was used based on the distance of RGB values. Thus, colors that were similar to previously selected ones were merged to the others. The final palette obtained with our database consisted of 98 color patches. The color patches corresponding to the same color (as perceived by the human observer) were collected to form a color group. On a visual basis, 15 were attributed to black, 10 to dark brown, 28 to light brown, 9 to red, 12 to white, and 24 to blue. The number of patches selected for each color group corresponds to the minimum number of color shades permitting a sufficiently accurate description of the color. This palette was then employed to detect color regions from the images according to a nearest-neighbor approach.45 Each pixel of the image was assigned to the
Automated Assessment of Pigment Distribution and Color Areas for Melanoma Diagnosis
139
(a)
Black Dark brown Light brown Red White Blue-grey
(b)
FIGURE 18.1 Example of color area identification in a melanoma: (a) the 20-fold videomicroscopic image; (b) the corresponding image with highlighted color areas. Between the two images is the color palette employed for color group attribution. The black color comprises 15 patches of black hues; the dark brown, 10 patches; the light brown, 28 patches; the red, 9 patches; the white, 12 patches; and the blue-gray, 24 patches. The colored square at the beginning of the line corresponds to the false color attributed to the pixels belonging to that color group.
color patch that minimized its Euclidean distance in the RGB color space (Figure 18.1). After assigning all pixels to their corresponding patches, those belonging to the same group were merged together to form the region corresponding to that particular color. Threshold values for areas of color regions were introduced, in order to avoid the inclusion of color areas, which were too small and lacking clinical relevance, in the calculation. Subsequently, colors were assessed by the computer program
on 331 melanocytic lesion images composing our image database, and results were compared to the evaluation of lesion colors performed by the clinician. By this method, we were able to objectively determine that black, blue-gray, and white were more frequently found in MMs and that the number of colors in MMs is higher than in nevi. When we compared the automatic color evaluation to the one performed by dermatologists, we observed that over 70% of the clinical decisions were
140
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 18.2 Black, White, and Blue-Gray Color Areas in Melanomas and Nevi as Assessed by Computer Color Descriptors Black Parameters AREA DIST-C SPRE
PERC-INT PERC-EXT % of presence
White
Blue-Gray
Parameter Meaning
MMs
NEVI
MMs
NEVI
MMs
NEVI
Number of pixels composing the color area divided by the number of pixels corresponding to the lesion area Distance between the color region centroid and the lesion centroid, divided by the length of the major axis First invariant obtained by the sum of the second-order moments with respect to the lesion centroid, divided by the squared area Number of pixels of the color region in the internal zone divided by the total number of pixels of that color area Number of pixels of the color region in the external zone divided by the total number of pixels of that color area Percentage of presence in melanomas or nevi
0.330
0.174
0.020
0.012
0.015
0.006
0.082
0.054
0.135
0.082
0.079
0.062
0.544
0.596
4.371
4.534
8.125
10.349
0.883
0.967
0.480
0.269
0.669
0.712
0.116
0.032
0.520
0.731
0.301
0.288
83.3%
61.9%
26.5%
13.8%
45.1%
25.7%
replicated. The highest correlation coefficients were observed for black, white, and blue-gray, which in our images represented the most important colors for diagnosis, indicating that this automatic procedure correlates well with the human evaluation method.
18.3.3 AUTOMATED DESCRIPTION FEATURES
OF
COLOR AREA
However, not only do the type and number of colors between benign and malignant ML images vary, but so does the color distribution, which strongly influences the evaluation of the asymmetry of the lesion. In a subsequent study we used an automatic procedure that breaks down the image into color areas according to the clinical examination process, also supplying a description of their extension and distribution with parameters correlated to the clinical concepts of regularity and homogeneity.46 After calculation of the lesion border,27 a numerical description of color region area, distance from the centroid, spread, color area distribution in the internal and external parts of the lesion, and asymmetries is provided. In order to evaluate the efficacy of these descriptors in distinguishing between benign and malignant ML images, and identify the most relevant color parameters for MM diagnosis, the descriptors were tested on a population comprising MMs and nevi. Significant differences in color parameters were observed for each color group, showing that color areas are more unevenly distributed in MMs than nevi. By means of the discriminant analysis method, the extension of black, white, and blue-gray areas and the location of these colors at the periphery of the lesion, together with their asymmetric distribution, were identified as the most relevant descriptors for the differentiation between benign and malignant lesions (Table 18.2).
18.3.4 AUTOMATED ASSESSMENT ATYPICAL NEVI
OF
COLORS
IN
Atypical nevi share some dermoscopic features with early MM, and a correct diagnosis cannot always be established on a clinical or dermoscopic basis. In these cases, however, the choice between no therapeutic intervention and excision is crucial, and computer image elaboration could represent a fundamental support to clinical diagnosis. In the study by Salopek et al.,47 the presence of seven colors, their predominance, and the total number of colors seen in images referring to MMs and atypical nevi (AN) were assessed. The authors observed that the presence of red, blue, gray, and white was a statistically significant predictor of MM, as was the presence of four or more colors. Blue and white were highly specific features, whereas more than three colors was the only sensitive parameter for early MM. According to Nilles et al.,48 100% of MMs exhibit more than three colors, whereas only 87% of AN do so. AN were classified according to distribution and intensity of their pigmentation by Hoffmann-Wellenhof et al.49 Central and peripheral hyper- or hypopigmentation, arranged in a localized and multifocal distribution, were considered. The authors drew attention to the peripherally hyperpigmented type, mimicking an in situ MM. Color changes were considered substantial modifications, indicating a possible transformation into an MM by Kittler et al.,50 who employed digital ELM for the follow-up of patients with multiple AN. To automatically assess colors in images referring to atypical nevi, and to compare data with those referring to clearly benign nevi and MMs, the above-described image analysis program was employed on videomicroscopic images referring to 76 AN, 288 clearly benign nevi, and 95 MMs.51 Thus, a mathematical description of the aspect
Automated Assessment of Pigment Distribution and Color Areas for Melanoma Diagnosis
FIGURE 18.2 Reconstruction of colors (pseudocolors) in (a) a clearly benign nevus, (b) an atypical nevus, and (c) a melanoma, according to numerical description of color areas by the computer.
and distribution of different color areas in AN was provided. Some colors, such as black, white, and blue-gray, were more frequently found in AN than in clearly benign nevi, but less frequently than in MMs. Moreover, color area features significantly differed between the three populations. If we consider AN as possible MM precursors, we could imagine a continuous transformation process leading from the aspect of a clearly benign nevus to the one of an MM (Figure 18.2). Starting from the ML with an obvious benign dermoscopic aspect, where dark color areas (dark brown and black) small in size are symmetrically distributed in or around the center, we can observe a progressive migration of these color regions toward the periphery, where they become larger and more unbalanced and asymmetric. The distribution of red progressively changes from a clearly benign lesion to an MM. This is expressed by an advancing imbalance of red areas, which appear more asymmetric in AN and MMs. In the latter population the red color component is more localized, possibly corresponding to areas of increased vascularization. The extent and distribution of white progressively change from clearly benign lesions to MMs, showing an intermediate aspect in AN. While in clearly benign lesions the white color component is scarce and is part of the nevus texture, being distributed all over the image, intermingled with other colors and arranged in a symmetric distribution, in MMs white areas become larger, less balanced and less symmetric, more compact, and located in the inner part of the lesion, possibly corresponding to regression areas. The blue-gray color component progressively increases from clearly benign lesions to MMs, and becomes unbalanced, asymmetric, and shifted to the external part of the lesion in MMs, indicating the presence of asymmetrically distributed gray-blue areas in malignant lesions.
18.3.5 COLORS
IN IMAGE
BLOCKS
Another possible approach for the evaluation of colors in melanocytic lesion images consists of subdividing the image into color blocks and assessing color variations between the different blocks.52 This generates parameters for the description of pigment distribution.
141
FIGURE 18.3 Image analysis process for the assessment of structural asymmetry in melanocytic lesion images. The digital image (a, e) is subdivided by a grid (b, f). The average color inside each color block is computed, considering only pixels belonging to the lesion (c, g). Excluded blocks (when less than 25% is comprised inside the lesion border) are shown in white (d, h).
To this purpose, we employed a three-step procedure. First, color details in the image were eliminated by means of its simplification into color blocks. After selecting the level of detail, the image was subdivided by a grid (Figure 18.3). Then, the average color inside each color block was computed, considering only pixels belonging to the lesion (Figure 18.3c and g). Image blocks were considered valid only if more than 25% of their area was formed by lesion points, whereas they were excluded if less than 25% was comprised inside the lesion border (excluded blocks are shown in white in Figure 18.3d and h). The third algorithm phase assesses the color difference between blocks using the Euclidean distance in the RGB color space and extracts measures of color distribution within the lesion image employing mean, variance, and maximum color differences. This procedure was employed both on images composed by 91 ∞ 96 pixels per block and on images composed by 48 ∞ 49 pixels per block, equivalent to a resizing of the image to 1 and 2%, and corresponding to a mean number of valid blocks per image of 9.45 and 31.63, respectively. This process can produce a very high number of comparisons between pixels, since it needs n(n – 1)/2 comparisons, where n is the number of valid pixels in the grid; this implies that up to 595 and 8385 distances, respectively, were computed for large lesions. Thus, we obtained a set of three parameters (mean distance, variance, and maximum distance) for each detail level (1 and 2%), corresponding to six measurements per lesion. When we applied these measurements to a set comprising 95 MMs, 76 AN, and 288 clearly benign lesions, significant differences in pigment distribution parameters (mean RGB distance, variance, and maximum distance) between the three ML populations were observed, permitting a good discrimination of MMs, which were characterized by color variegation and a complex architecture. The pigment distribution assessment method described in this study translates structural differences into RGB color ones. Details such as globules, dots, regression structures, areas of increased vascularization, and
142
Handbook of Non-Invasive Methods and the Skin, Second Edition
gray-blue areas contribute to RGB values within the image block they are included in. The distribution of these structures is then assessed by comparing RGB values pertaining to different image blocks. A great difference between RGB values means that the lesion’s structure is nonhomogeneous and its architecture is complex. This corresponds to structural asymmetry as assessed by the clinician when employing semiquantitative methods.
18.4 CONCLUSION Reproducibility of color assessment is generally high, pointing at the high diagnostic relevance of color parameters. In a recent consensus study on dermoscopy via the Internet, intraobserver agreement showed a k value of 0.64 for the number of colors in the ABCD rule and of 1 for the presence of a single color according to Menzies.5 When assessing the informativeness of compressed videomicroscopic images, differences between kappa values of uncompressed and compressed images referring to intraobserver agreement on colors were lower than for other features, indicating that reproducibility of color evaluation is high, even in lower-quality images, and that colors are less affected by compression than morphological details in the image.11 This underlines the potential of color assessment-based diagnosis in an era where the importance of teledermatology is ever increasing. Of course, caution must be taken in selecting color features. The absolute color measure is strictly dependent on the acquisition technique and the imaging system; i.e., RGB is a device-dependent color space.17,53 Efforts have to be encouraged for improving standardization of image colors.54
REFERENCES 1. Miller, M., Ackermann, A.B., How accurate are dermatologists in the diagnosis of melanoma? Degree of accuracy and implicants, Arch Dermatol, 128, 559, 1992. 2. Wolf, I.H., Smolle, J., Soyer, H.P., Kerl, H., Sensitivity in the clinical diagnosis of malignant melanoma, Melanoma Res, 8, 425, 1998. 3. Mayer, J., Systematic review of the diagnostic accuracy of dermatoscopy in detecting malignant melanoma, Med J Aust, 167, 206, 1997. 4. Nachbar, F., Stolz, W., Merkle, T., et al., The ABCD rule of dermatoscopy, J Am Acad Dermatol, 30, 551, 1994. 5. Menzies, S.W., Ingvar, C., McCarthy, W.H., A sensitivity and specificity analysis of the surface microscopy features of invasive melanoma, Melanoma Res, 6, 55, 1996.
6. Argenziano, G., Fabbrocini, G., Carli, P., et al., Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions. Comparison of the ABCD rule of dermatoscopy and a new 7-point checklist based on pattern analysis, Arch Dermatol, 134, 1563, 1998. 7. MacKie, R.M., Fleming, C., McMahon, A.D., Jarret, P., The use of the dermatoscope to identify early melanoma using the three-colour test, Br J Dermatol, 146, 481, 2002. 8. Blum, A., Rassner, G., Garbe, C., Modified ABC-point list of dermoscopy: a simplified and highly accurate dermoscopic algorithm for the diagnosis of cutaneous melanocytic lesions, J Am Acad Dermatol, 48, 672, 2003. 9. Soyer, H.P., Argenziano, G., Zalaudek, I., et al., Threepoint checklist of dermoscopy. A new screening method for early detection of melanoma, Dermatology, 208, 27, 2004. 10. Argenziano, G., Soyer, H.P., Cimenti, S., et al., Dermoscopy of pigmented skin lesions: results of a consensus meeting via the Internet, J Am Acad Dermatol, 48, 679, 2003. 11. Seidenari, S., Pellacani, G., Righi, E., Di Nardo, A., Is JPEG-compression of videomicroscopic images compatible with telediagnosis? Comparison between diagnostic performance and pattern recognition on uncompressed TIFF images and JPEG compressed ones, Telem J e-Health, 10, 294, 2004. 12. Cascinelli, N., Ferrario, M., Bufalino, R., et al., Results obtained by using a computerized image analysis system designed as an aid to diagnosis of cutaneous melanoma, Melanoma Res, 2, 163, 1992. 13. Umbaugh, S.E., Moss, R.H., Stoecker, W.V., Hance G.A., Automatic color segmentation algorithms: with application to skin tumor feature identification, IEEE Eng Med Biol, 9, 75, 1993. 14. Ercal, F., Chawla, A., Stoecker, W.V., et al., Neural network diagnosis of malignant melanoma from color images, IEEE Trans, 9, 837, 1994. 15. Schindewolf, T., Schiffner, R., Stolz, W., et al., Evaluation of different image acquisition techniques for a computer vision system in the diagnosis of malignant melanoma, J Am Acad Dermatol, 31, 33, 1994. 16. Green, A.C., Martin, N.G., Pfitzner, J., et al., Computer image analysis in the diagnosis of melanoma, J Am Acad Dermatol, 31, 958, 1994. 17. Seidenari, S., Burroni, M., Dell’Eva, G., et al., Computerized evaluation of pigmented skin lesion images recorded by a videomicroscope: comparison between polarizing mode observation and oil/slide mode observation, Skin Res Technol, 1, 187, 1995. 18. Hall, P.N., Claridge, E., Morris Smith, J.D., Computer screening for early detection of melanoma: is there a future? Br J Dermatol, 132, 325, 1995. 19. Aitken, J.F., Pfitzner, J., Battistutta, D., et al., Reliability of computer image analysis of pigmented skin lesions of Australian adolescents, Cancer, 78, 252, 1996.
Automated Assessment of Pigment Distribution and Color Areas for Melanoma Diagnosis
20. Gutkowicz-Krusin, D., Elbaum, M., Szwaykowski, P., Kopf, A.W., Can early malignant melanoma be differentiated from atypical melanocytic nevus by in vivo techniques? Part II. Automatic machine vision classification, Skin Res Technol, 3, 15, 1997. 21. Seidenari, S., Pellacani, G., Pepe, P., Digital videomicroscopy improves diagnostic accuracy for melanoma, J Am Acad Dermatol, 39, 175, 1998. 22. Binder, M., Kittler, H., Seeber, A., et al., Epiluminescence microscopy-based classification of pigmented skin lesions using computerized image analysis and an artificial neural network, Melanoma Res, 8, 261, 1998. 23. Seidenari, S., Pellacani, G., Giannetti, A., Digital videomicroscopy and image analysis with automatic classification for detection of thin melanomas, Melanoma Res, 9, 163, 1999. 24. Andreassi, L., Perotti, R., Rubegni, P., et al., Digital dermoscopy analysis for the differentiation of atypical nevi and early melanoma, Arch Dermatol, 135, 1459, 1999. 25. Pellacani, G., Martini, M., Seidenari, S., Digital videomicroscopy with image analysis and automatic classification as an aid for diagnosis of Spitz nevus, Skin Res Technol, 5, 266, 1999. 26. Ganster, H., Pinz, A., Rohrer, R., Wilding, E., Binder, M., Kittler, H., Automated melanoma recognition, IEEE Trans Med Imag, 20, 233, 2001. 27. Grana, C., Pellacani, G., Cucchiara, R., Seidenari, S., A new algorithm for border description of polarized light surface microscopic images of pigmented skin lesions, IEEE Trans Med Imaging, 22, 959, 2003. 28. Cucchiara, R., Grana, C., Seidenari, S., Pellacani, G., Exploiting color and topological features for region segmentation with recursive fuzzy c-means. Mach Graphics Vision, 11, 169, 2002. 29. Greene, M.H., Clark, W.H., Tucker, M.A., et al., High risk of malignant melanoma in melanoma-prone families with dysplastic nevi, Ann Intern Med, 102, 458, 1985. 30. Seidenari, S., Pellacani, G., Martella, A., Acquired melanocytic lesions and the decision to excise: the role of colour variegation and distribution as assessed by dermoscopy, Dermatol Surg, 31, 184, 2005. 31. Pehamberger, H., Steiner, A., Wolff, K., In vivo epiluminescence microscopy of pigmented skin lesions. I. Pattern analysis of pigmented skin lesions, J Am Acad Dermatol, 17, 571, 1987. 32. Kenet, R.O., Kang, S., Kenet, B.J., et al., Clinical diagnosis of pigmented lesions using digital epiluminescence microscopy, Arch Dermatol, 129, 157, 1993. 33. Steiner, A., Binder, M., Schemper, et al., Statistical evaluation of epiluminescence microscopy criteria for melanocytic pigmented skin lesions, J Am Acad Dermatol, 29, 581, 1993. 34. Nilles, M., Boedeker, R.H., Schill, W.B., Surface microscopy of naevi and melanomas. Clues to melanoma, Br J Dermatol, 130, 349, 1994.
143
35. Soyer, H.P., Smolle, J., Leitinger, G., Rieger, E., Kerl, H., Diagnostic reliability of dermoscopic criteria for detecting malignant melanoma, Dermatology, 190, 25, 1995. 36. Pellacani, G., Grana, C., Cucchiara, R., Seidenari, S., Automated extraction and description of dark areas in surface microscopy melanocytic lesion images, Dermatology, 208, 21, 2004. 37. Kahofer, P., Hoffmann-Wellenhof, R., Smolle, J., Tissue counter analysis of dermatoscopic images of melanocytic skin tumors: preliminary findings, Melanoma Res, 12, 71, 2002. 38. Umbaugh, S.E., Moss, R.H., Stoecker, W.V., Hance, G.A., Automatic color segmentation algorithms: with application to skin tumor feature identification, IEEE Eng Med Biol, 9, 75, 1993. 39. Day G.R., Barbour, R.H., Automated skin lesion screening: a new approach, Melanoma Res, 11, 31, 2001. 40. Cotton, S.D., Claridge, E., Developing a predictive model of human skin coloring. Proc SPIE Med Imaging Phys Med Imaging, 2708, 814, 1996. 41. Heckbert, P. Color image quantization for frame buffer display, Comput Graphics, 16, 297, 1982. 42. Faziloglou, Y., Stanley, R.J., Moss, R.H., et al., Colour histogram analysis for melanoma discrimination in clinical images, Skin Res Technol, 9, 147, 2003. 43. Chen, J., Stanley, R.J., Moss, R.H., Van Stoecker, W., Colour analysis of skin lesion regions for melanoma discrimination in clinical images, Skin Res Technol, 9, 94, 2003. 44. Seidenari, S., Pellacani, G., Grana, C., Computer description of colours in dermoscopic melanocytic lesion images reproducing clinical assessment, Br J Dermatol, 149, 523, 2003. 45. Fukunaga, K., Introduction to Pattern Recognition, 2nd ed. Academic Press, New York, 1990. 46. Pellacani, G., Grana, C., Seidenari, S., Automated description of colours on surface microscopic images of melanocytic lesions, Melanoma Res, 14, 125, 2004. 47. Salopek, T.G., Kopf, A.W., Stefanato, C.M., et al., Differentiation of atypical moles (dysplastic nevi) from early melanomas by dermoscopy, Dermatol Clinics, 19, 337, 2001. 48. Nilles, M., Boedeker, R.H., Schill, W.B., Surface microscopy of naevi and melanoma: clues to melanoma, Br J Dermatol, 130, 349 1994. 49. Hoffmann-Wellenhof, R., Blum, A., Wolf, I., et al., Dermoscopic classification of atypical melanocytic nevi (Clark nevi), Arch Dermatol, 137, 1575, 2001. 50. Kittler, H., Pehamberger, H., Wolff, K., Binder, M., Follow-up of melanocytic skin lesions with digital epiluminescence microscopy: patterns of modifications observed in early melanoma, atypical nevi, and common nevi, J Am Acad Dermatol, 43, 467, 2000. 51. Seidenari, S., Pellacani, G., Grana, C., Colors in atypical nevi: a computer description reproducing clinical assessment, Skin Res Technol, 11, 36, 2005.
144
Handbook of Non-Invasive Methods and the Skin, Second Edition
52. Seidenari, S., Pellacani, G., Grana, C., Pigment distribution in melanocytic lesion images: a digital parameter to be employed for computer-aided diagnosis, Skin Res Technol, in press. 53. Pellacani, G., Seidenari, S., Comparison between morphological parameters in pigmented skin lesion images acquired by means of epiluminescence surface microscopy and polarized-light videomicroscopy, Clin Dermatol, 20, 222, 2002.
54. Grana, C., Pellacani, G., Seidenari, S., Practical color calibration for dermatoscopic images, Skin Res Technol, in press.
Skin Surface Contour and Roughness Assessment
Replication for Light and 19 Skin Scanning Electron Microscopy Bo Forslind Department of Medicine Biochemistry and Biophysics, Karolinska Institute, Stockholm, Sweden Motto: Every honest researcher I know admits he is just a professional amateur. He is doing whatever he is doing for the first time. That makes him an amateur! He has sense enough to know that he is going to have a lot of trouble, so that makes him a professional! —Anonymous
CONTENTS 19.1 Introduction............................................................................................................................................................147 19.2 Carbon and Metal/Carbon Replication of Dry Solid Surfaces.............................................................................148 19.2.1 Methodological Principle ..........................................................................................................................148 19.2.2 Sources of Errors .......................................................................................................................................148 19.2.3 Correlation with Other Methods ...............................................................................................................148 19.3 Plastic Impression Technique ................................................................................................................................149 19.3.1 Methodological Principle ..........................................................................................................................149 19.3.2 Sources of Errors .......................................................................................................................................149 19.3.3 Correlation with Other Methods ...............................................................................................................149 19.4 Silicone Elastomer Replication .............................................................................................................................149 19.4.1 Methodological Principle ..........................................................................................................................149 19.4.2 Sources of Errors .......................................................................................................................................150 19.4.3 Recommendations......................................................................................................................................150 19.5 Present Status of Replication Techniques in Dermatology ..................................................................................152 Acknowledgment.............................................................................................................................................................153 References .......................................................................................................................................................................153
19.1 INTRODUCTION Replication is actually an old method within the realm of electron microscopy. Before the introduction of the scanning electron microscope (SEM), surface analysis was routinely performed by using multiple-stage replication techniques. A general feature of these techniques was the fact that they were all destructive, in the sense that the replicated material was lost in the preparation procedure. Hence, its applicability to investigation of the skin was limited to situations where skin (or in more general terms, integument) samples were taken by biopsy. Recent reviews show that it is only to a minor extent that replication techniques have been used in clinical and experimental dermatology.2,10 The advantages of this non-
invasive technique, therefore, may still reveal unexplored areas of application. If analysis of the surface structure of skin and nails can be performed without the need for biopsy it will be possible to study processes of inflammation and/or infection sequentially, as well as the effect of topical drugs and direct physical wear and tear at the same site. Sampling of topographical information for analysis of stratum corneum have been achieved through various techniques, including tape stripping,16 plastic impression techniques,1 skin surface biopsy,4 besides the subject of this chapter. The plastic impression technique and, to a certain extent, the tape stripping technique (if it is to contain more than a few scattered stratum disjunctum cells), have the drawback that they require a thorough defattening of the skin surface. The surface biopsy strips off a thin surface 147
148
Handbook of Non-Invasive Methods and the Skin, Second Edition
layer of the stratum corneum, but the surface that can be investigated in the SEM (or light microscope) represents the rift zone between the part of the stratum corneum in situ and the sample. This is then a surface that has not been directly exposed to the environment but represents the stratum corneum at a depth of one or more cell layers. Hence, the details of the surface structures such as villous projections4 will be more pronounced than those of the actual surface of the stratum corneum. In principle the replica of a surface can be a negative (or direct) one or a positive, essentially two-step, replica. Since the skin surface of a living subject contains moisture a number of replication media that require a dry surface are disqualified. Materials designed for replication in dental work have the required property of adhering tightly even to a wet surface and are therefore interesting candidates for skin replication molds. The silicon plastics have proven to allow detailed recording of surface characteristics of skin and nails, and have found extensive use in medical and biological scanning microscopy.2,3,6,10–14 One of the first papers to report the use of silicone rubber plastics to produce the negative mold for hard plastic replicas was given by Sampson11 for use on plant material and insects. Although intended for light microscopic use the technique was soon adopted for SEM analysis of surfaces. This chapter will, on a selective basis, review the surface replication technique as it has evolved in electron microscopy. We will not aim at providing the reader with a complete catalogue of materials but rather put forward some general and practical hints for the particular dermatological applications. The emphasis is put on simple and fast measures, suitable for use in the consultation office, which will produce replicas that can be analyzed at magnifications up to at least 2000×. Some practical aspects of producing and checking the result of a replication are forwarded. Readers with interests in high-resolution work are referred to current textbooks on electron microscopic preparation techniques and to the comprehensive reviews by Pfefferkorn and Boyd9 and Pameijer,7,8 which cover the literature up to 1978.
19.2 CARBON AND METAL/CARBON REPLICATION OF DRY SOLID SURFACES 19.2.1 METHODOLOGICAL PRINCIPLE This technique was developed for transmission electron microscopy (TEM) of surfaces before the SEM was available. It is mandatory that the object is completely free of water and any other substances that are volatile in the vacuum system used for evaporative coating the specimen.
For biological tissues this means that replication must be preceded by fixation, usually chemical fixation, but cryomethods are also possible. Subsequently a thorough dehydration of the specimen is required. This can be done according to the common protocol used for preparation of biological tissue for TEM. Alternatively critical point drying may be used. The surface to be replicated should be cleaned by an appropriate method, e.g., ultrasonication, organic solvents, water, etc., and subsequently dried. The replication requires six steps (Figure 19.1): 1. The clean surface is covered with a thin carbon or metal (Cr, Pd/Pt, Pt, Au) film at an angle (often 6 to 10°) by evaporation in vacuum. 2. The evaporated film is stabilized by application of a thin plastic film, e.g., formvar. 3. The object is removed mechanically (which often disrupts the replica) or preferably by chemical dissolution. 4. The (negative) replica is stabilized by a thick layer of carbon evaporated onto the replica in vacuum. 5. The plastic film is removed by the appropriate solvent for the plastic in question. 6. The positive carbon replica in now transferred to a conventional electron microscopic grid and viewed in the TEM. This technique has a resolution that is satisfactory at least down to 20 nm (200 Å) on metal surfaces. For biological specimens the resolution is somewhat less and very much dependent on the nature of the specimen surface and the preparation. A similar process, omitting the plastic intermediate stage, is used in the production of freeze-fracture replication, which has been successfully used in membrane research.
19.2.2 SOURCES
OF
ERRORS
It is to be noted that all fixation and drying procedures induce various degrees of surface structure artefacts due to linear or volume changes, shrinkage, etc.
19.2.3 CORRELATION
WITH
OTHER METHODS
The main drawback of the technique is the loss of the original sample and the fragmentation of the replica on transferring to the electron microscopic grid, something that is virtually impossible to avoid. Thus, the technique is best when high resolution is required and therefore preferentially a TEM technique.
Skin Replication for Light and Scanning Electron Microscopy
149
focused on low-resolution details of the skin, such as the cutaneous patterns of furrows that form patterns characteristic of a certain area of the integument. The negative imprint is obtained simply by spreading the liquid plastic over the area to be sampled and allowing it to cure before mechanical removal. Household glues based on plastics in organic solvents that evaporate quickly can produce acceptable results at the low resolution required.
19.3.2 SOURCES
OF
ERRORS
The preparation of the skin surface to be replicated, including shaving and degreasing with organic solvents, is likely to produce artefactual changes of fine surface structures. The gross features of the skin area replicated, i.e., the patterns of furrows and wrinkles, reproduce well even at low magnifications, e.g., >10×.
19.3.3 CORRELATION
WITH
OTHER METHODS
There is some loss of fine details at the high magnifications attainable in the SEM.
19.4 SILICONE ELASTOMER REPLICATION 19.4.1 METHODOLOGICAL PRINCIPLE The negative replica — The silicone-elastomer replication has its basis in the products developed for use in clinical dentistry.3,14 There are five main requirements on the plastic used for producing the negative mold.
FIGURE 19.1 The steps in making a metal/carbon replica of surfaces. The clean surface (a) is covered with a thin carbon or metal (Cr, Pd/Pt, Pt, Au) film at an angle (often 6 to 10°) by evaporation in vacuum (b). The evaporated film is stabilized by application of a thin plastic film, e.g., formvar (c) (hatched area). The object is removed mechanically (which often disrupts the replica) or preferably by chemical dissolution (d). The (negative) replica is stabilized by a thick layer of carbon evaporated onto the replica (e) in vacuum. The plastic film is removed by the appropriate solvent for the plastic in question. The positive carbon replica is now transferred to a conventional electron microscopic grid (f) and viewed in the TEM.
19.3 PLASTIC IMPRESSION TECHNIQUE 19.3.1 METHODOLOGICAL PRINCIPLE Before introduction of the SEM, but even more recently, replicas for light microscopy have been made from plastic compositions that required a cleaned and dry, hairless area for the replication.1,6 Consequently interest has been
1. The silicone plastic should have a low viscosity to adhere closely even to the fine details of the surface 2. It should adhere well even to wet surfaces 3. After a fast, and complete, polymerization it should be released from the original specimen without leaving any material behind. 4. It should possess an elastic memory to allow a complete return to the original status even when withdrawn from undercuts. 5. The polymerization process should not produce heat, i.e., involve an exothermic reaction which may change surface properties of the object, and cause the discomfort of the subject. The positive replica — The plastic used for producing the positive replica should cure at room temperature with as little release of heat as possible to prevent deformation of the negative mold. In the SEM micrographs accompanying this chapter (Figure 19.2) molds were made from Provil-L® (Bayer Dental D-5090 Leverkusen, Germany), which is characterized as a low-viscosity, type I silicone meeting the
150
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 19.2 Replicas of a normal skin surface (volar aspect of lower arm) obtained by the silicon elastomer two-step method. (a) Skin surface after approximately 15 min of exposure to damp cloth saturated with distilled water, 230×.
FIGURE 19.2 (d) 212×.
19.3c). It is then allowed to set for 3 min before gentle removal from the (skin) surface (Figure 19.3d). The negative replica is subsequently covered with an Araldite® plastic (CIBA-Geigy) (Figure 19.3e) which cures within 3 to 5 h depending on the volume applied. Alternatively we have used a methacrylate designed for whole-mount embedding of insects, etc., which has longer curing time. The surface of the plastic, positive replica is subsequently made conductive by gold sputtering (Figure 19.3f and g).
19.4.2 SOURCES
FIGURE 19.2 (b) 503×. There are no obvious villiform projections on the surface that has been exposed to the environment.
OF
ERRORS
The negative mold — A large negative imprint of the skin surface (i.e., >1 × 1 cm) tends to bend when loosened from the original surface and this large curvature remains when the positive replica is made (Figure 19.3f). Due to high total absorption of energy in the electron beam, a larger-than-the-stub specimen tends to be unstable in the beam, i.e., be subject to drift during viewing the SEM. The positive replica — When making the positive replica the amount of accelerator may be crucial to the final results. If the curing process occurs at too fast a rate, gas bubbles will accumulate at the replica surface.
19.4.3 RECOMMENDATIONS
FIGURE 19.2 (c) Corresponding area sampled dry on the following day, 101×.
requirements of ISO 4823, type (e) 3, category A (adhesion-induced polymerization). The silicone plastic is thoroughly mixed with an equal volume of catalyst and immediately applied to the surface to be replicated (Figure
The negative replica — The mixing of silicon base and curer is a critical stage in making a replication. The two components should be thoroughly mixed but agitation should not be so vigorous as to produce air bubbles. The drawbacks of manual mixing can be virtually eliminated when using a dual vessel ejector (Bayer Cartridge delivery dispensing gun) (Figure 19.3c). When making a negative replica for SEM care should be taken to cover a surface area no greater than the SEM specimen holder (the “stub”) to avoid the unnecessary heating that follows from having a large specimen surface. When the negative replica has been removed after setting, its surface can be inspected under a light microscope at about 40× magnification to
Skin Replication for Light and Scanning Electron Microscopy
151
FIGURE 19.3 Silicon elastomer replication of an integument surface. (a) The skin surface is briefly cleaned by a quick rinse in cool tap water and (b) blotted dry. (c) The elastomer is applied and allowed to cure for about 3 min. (d) The negative mold produced is gently removed. (e) The mold is covered with a plastic to produce a positive replica. (f) After gold-sputtering the positive replica (left) and the negative mold can be inspected for surface defects by light microscopy.
artefacts. Sometimes an improvement of resolution of details by the silicone is obtain through a quick cool water rinsing which undoubtedly removes water-soluble surface material. At low magnification (e.g., 40×) no swelling is apparent from this process. Rinsing may, however, introduce artefacts in lesional skin. It is then preferable to remove loose surface material by making two or three
impressions from the same surface rather than making the lesion subject to tap water rinse. The sequentially obtained molds can be checked against each other for artifacts. The negative imprint, the mold, is not suited for direct study in the SEM because it will melt and evaporate when hit by the electron beam. However, it can be used directly for light microscopic and photographic observations at
152
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 19.3 (g) Even large objects can be faithfully replicated. Left: fingertip of digit V of the author. (h) A stub machined to have a groove with undercut will prevent drift in large objects during SEM study.
low and moderate magnifications.8 If this is the object, rather than a study in the SEM, large areas can be replicated, e.g., 2 × 2 cm (Figure 19.3g). The positive replica — The making of a successful positive replica involves a good choice of plastic. We have used Araldite® (Ciba-Geigy) in a 1:1 mixture with accelerator. The manufacturer’s advice on mixing proportions for plastic and accelerator should be tested for each batch of plastic as it may vary during aging of these materials. It is our experience that it is not unusual that the amount of accelerator should be somewhat reduced to get a curing rate that does not produce heat and solvent gas bubbles. However, it is easy to get an incomplete curing that results in a sticky surface which deforms on removal from the negative mold. An alternative way of reducing the risk of gas bubbles at the interface between the negative mold and the plastic is to moisten the surface of the silicon imprint with the solvent of the plastic (e.g., acetone for Araldite®) immediately before pouring the plastic onto the negative template. It is advised that positive casts are inspected for the presence of gas bubbles in the surface structures under a preparation microscope after gold sputtering. If bubbles are present they usually attain a size that allows them to be seen at a magnification of 40. As an
alternative to Araldite® we have used a methacrylate designed for embedding of large objects such as insects, e.g., a beetle. This methacrylate, which takes more than 24 h to cure even in thin sheets, tends to be very brittle. It reproduces the surface details well in our experience. Pfister and Neukirchner10 used a polystyrol granulate dissolved in toluol for the positive replica. The non-cured plastic has a syrup-like consistency. To avoid air bubbles in small crevices of the negative replica the authors “moisten” it with the solvent, toluol, before applying the plastic. The hardening time of this polystyrol plastic is comparatively long, approximately 24 h. The authors claim that magnifications up to 5000× are attainable with this technique. Gold sputtering — The plastic material of the positive replica is an insolator. Gold sputtering of the surface provides a conductive film that distributes charges to ground potential but also contributes as a heat sink. The sputtering should be performed so as to get a continuous contact between the replica surface and the specimen stub. This is most easily achieved if the stub surface is cleared of the specimen at small point. When large objects are used, e.g., a replica of a fingertip with a nail, it is advantageous if the stub can be molded into the positive plastic replica. This can be achieved by making a groove with undercut (Figure 19.3h) with a milling cutter. Alternatively a cavity with undercut in the stub surface can be made using a dental drill. Through these means drift is virtually completely eliminated.
19.5 PRESENT STATUS OF REPLICATION TECHNIQUES IN DERMATOLOGY Most areas of the human integument in health1,2,6,12,15 have been described using replication techniques. In addition pathological conditions, including lesions of psoriasis,13 superficial actinic porokerastosis, as well as more unusual conditions like Gorlin’s syndrome,6 have been documented. It is noteworthy that data presented in the literature on topographic data collected by replication (and corresponding) techniques on skin and its appendices in general merely have a descriptive character and provide little, if any, functional interpretation of the findings. Cosmetic industries have long utilized SEM studies of the effect of cosmetic formulations on the skin surface and the integument appendices, but details of this information have not been publicly available and cannot be scientifically evaluated. It is obvious that topographical methods of investigating the skin surface represent an interesting and potentially fruitful area of dermatological research. Combination with morphometric systems, image analysis systems, or other physical measurement systems5 will allow quantitative analysis of changes in the surface structures as a result of the progress of a disease or a
Skin Replication for Light and Scanning Electron Microscopy
treatment of a disease. In addition to such applications a more extensive use of the excellent replication materials presently available will no doubt increase our knowledge of the dynamics of skin function in health and disease.
ACKNOWLEDGMENT I am indebted to Mr. Eva Lundewall of the Swedish branch of Phillips Industrial Electronics AB for producing the SEM micrographs and to Ms. Margareta Andersson for photography and lay-out of illustrations. The golden rule is that there are no golden rules. George Bernhard Shaw
REFERENCES 1. Chinn, H.D. and Dobson, R.L., The topographic anatomy of human skin, Arch. Dermatol., 89, 155, 1964. 2. Forslind, B., Clinical applications of scanning electron microscopy and X-ray microanalysis in dermatology, Scanning Electron Microsc., 1, 183, 1984. 3. Jokstad, A. and Mjör, L.A., Assessment of marginal degradation of restorations on impressions, Acta Odontol. Scand., 49, 15, 1991. 4. Marks, R. and Dawber, R.P.R., Skin surface biopsy: an improved replica for the examination of the horny layer, Br. J. Derm., 84, 117, 1971. 5. Marks, R. and Pearse, A.D., Surfometry. A method of evaluating the internal structure of the stratum corneum, Br. J. Dermatol., 92, 651, 1975.
153
6. Nayler, J.R., Applications of the skin surface replica technique in dermatology, J. Audiovis. Media Med., 21, 1984. 7. Pameijer, C.H., Replica techniques for scanning electron microscopy — a review, Scanning Electron Microsc., II, 831, 1978. 8. Pameijer, C.H., Replication techniques with new dental impression materials in combination with different negative impression materials, Scanning Electron Microsc., II, 571, 1979. 9. Pfefferkorn, G. and Boyde, A., Review of replica techniques for scanning electron microscopy, Scanning Electron Microsc., 1, 75, 322, 1974. 10. Pfister, T.C. and Neukirchner, A., Raster-elektronenmikroskopischer Untersuchungen am kranken Nagel mittels Abdruck-Verfahren, Fortschr. Med., 98, 1465, 1980. 11. Sampson, J., A method for replicating dry or moist surfaces for examination by light microscopy, Nature, 191, 932, 1961. 12. Tring, F.C. and Murgatroyd, L.B., Surface microtopography of normal human skin, Arch. Dermatol., 109, 223, 1974. 13. Tring, F.C. and Murgatroyd, L.B., Psoriasis — changes in surface microtopography, Arch. Dermatol., 111, 476, 1975. 14. Walsh, T.F., Waimsley, A.D., and Carrotte, P.V., Scanning electron microscopic investigation of changes in the dentogingival area during experimental gingivitis, J. Clin. Periodontol., 18, 20, 1991. 15. Wagner, G. and Goltz, R.W., Human cutaneous topography. A new photographic technique: observation on normal skin, Cutis, 23, 830, 1979. 16. Wolf, J., Die innere Struktur der Zellen des Stratum desquamans der Menschlichen Epidermis, Z. Mikr. Anat. Forsch., 46, 170, 1939.
Surface Replica Image Analysis of 20 Skin Furrows and Wrinkles Pierre Corcuff and Jean-Luc Lévêque Laboratoires de Recherche de L’OREAL, Aulnay Sous Bois, France
CONTENTS 20.1 Introduction............................................................................................................................................................155 20.2 Objective ................................................................................................................................................................155 20.3 Basic Methodology................................................................................................................................................156 20.3.1 Skin Surface Replica .................................................................................................................................156 20.3.2 Shadowing Method ....................................................................................................................................156 20.3.3 Image Analysis ..........................................................................................................................................156 20.3.3.1 Image Analyzer ..........................................................................................................................156 20.3.3.2 Analysis of the Image ................................................................................................................157 20.3.3.3 Measurement Parameters ...........................................................................................................157 20.3.3.4 Automation .................................................................................................................................158 20.3.4 Choice of Lighting Angle..........................................................................................................................159 20.4 Sources of Error.....................................................................................................................................................159 20.4.1 Replica Artifacts ........................................................................................................................................159 20.4.2 Analysis of the Image................................................................................................................................160 20.5 Correlation with Other Methods ...........................................................................................................................160 20.6 Recommendations..................................................................................................................................................161 References .......................................................................................................................................................................161
20.1 INTRODUCTION Among the various non-invasive methods described in this handbook, the description and measurement of the geometric organization of the skin relief involve the same goal, i.e., the study of underlying biological and physiological phenomena based on their impact on the surface. The skin surface “messages” result mainly from the organization of the dermis and its collagen and elastin networks, but the state of the epidermis and stratum corneum can also play a role. The problem is more complex than the simple measurement of an excretion (sebum, sweat, transepidermal water loss), as the structure is three-dimensional and needs a minimum of geometric parameters; it is therefore difficult to make a simple description and interpretation. A method developed in the 1970s to measure the roughness of metallic surfaces — mechanical profilometry — has since been adapted by several authors1–3 to study cutaneous topography. Profilometry is still widely used today, as the replacement of the physical probe by a laser beam has
overcome two major obstacles: contact between the instrument and the surface, and the time required for measurement. Given the early drawbacks of this method, we proposed in 1981 a technique based on image analysis of a skin replica, which was rapid and well adapted to the anisotropy of the skin relief, but which was prohibitively expensive.4 With the fall in the cost of electronic components and computers, image analysis has now become more accessible, and an increasing number of teams have adopted this method, which, as we shall see later in this chapter, requires impeccable technique and discipline.
20.2 OBJECTIVE At the beginning of the 1980s, the only method available was thus mechanical profilometry. It required a supple negative replica to be made, followed by a counter-replica made of hard resin (which resisted deformation by the probe). The main criticisms were the accumulation of artifacts by successive replication, the directional traces 155
156
Handbook of Non-Invasive Methods and the Skin, Second Edition
that were poorly adapted to the anisotropy of the skin surface, and the fact that the roughness parameters were difficult to interpret in terms of skin relief. Image analysis was used by few teams at that time. We discovered that it was capable of scanning the first replica in several directions and that the geometric parameters provided by this technique (orientation, number, and depth) were easy to interpret. Finally, contact between the instrument and the study surface was avoided. The objective was to describe the organization of the primary lines according to Hashimoto’s classification5 by using a minimum of parameters. When this objective had been reached, the method was extended to the study of wrinkles of the crow’s-foot area. In order to obtain objective and reliable results, we opted for an entirely automated method, which disallowed all interactive intervention.
Positive cast
Negative replica
20.3 BASIC METHODOLOGY The principle of image analysis can be likened to an air passenger watching the shadows crossing the Alps as the sun traverses the winter sky. Each mountain has a dark side and a bright side; some valleys are illuminated in the morning, others in the afternoon. The brightness increases the contrast between the areas of shadow and light, just like topologic details.
20.3.1 SKIN SURFACE REPLICA Applied to skin relief, this method carried the following constraints. The mold had to be white (as snow), matte, and opaque. Among the brands studied at the time, the silicon resin SILFLO from Flexico (England) fit the bill.6 It guaranteed reliable reproduction, the absence of further deformation, and documented artifacts.7 The fact that the samples could be stored for about 2 years was an added advantage. The negative replica is obtained in the following way. Two or three drops of Bayer catalyzer are added to 1 g of SILFLO resin and rapidly mixed in a cup with a spatula. The paste is then immediately applied to the study surface, which has first been delimited with an adhesive paper ring. The resin hardens within 2 or 3 min and the replica is removed gently by lifting from the tongue of the ring.
20.3.2 SHADOWING METHOD The cutaneous topography is such that the observer is obliged to shadow negative replicas rather than positive replicas, as is clearly shown in Figure 20.1. The sun in the above analogy is replaced by an optical fiber system providing illumination at a precisely defined angle relative to the plane of observation. For evident technical reasons, it is simpler to rotate the sample than the lighting system
FIGURE 20.1 Shadows generated at the surface of negative and positive replicas by grazing lighting. On the negative imprint the width of the shadow is large enough to allow a correct estimation of the peak height.
to simulate the movement of the sun. The negative replica with its ring is inserted into a metallic device covered with nonreflective black felt. This ensures that the sample is perfectly flat; it also means that the sample can be held perfectly horizontal and centered with regard to the light source and video camera.
20.3.3 IMAGE ANALYSIS 20.3.3.1 Image Analyzer The basic principle of the image analyzer is the segmentation of an image according to shades of gray, and this is perfect for selecting areas of shadow created by the oblique light impinging on the sample (skin furrows). In 1979, we selected the Quantimet 720 manufactured by Cambridge Instruments, as it had a number of technical features particularly adapted to this application and could be automated. We now use the Quantimet 970, which still has the essential elements of its predecessor; in this way, we can compare the results obtained in 1981 with those obtained in 1993. One important feature of this generation of system is the image format: 720 lines, 880 points per line, 605,000 pixels. To avoid introducing a bias into the measurements during sample rotation, the field of measurement must be circular, and this brings us down to 300,000 pixels (which is still a considerable amount). By way of comparison, a standard image formed of 512 × 512 pixels only provides a circular field of measurement of 180,000 pixels. Table 20.1 gives comparative data for the various measurement methods we have used. With the
Skin Surface Replica Image Analysis of Furrows and Wrinkles
TABLE 20.1 Comparison of Methods Used for Three-Dimensional Analysis of Skin Surface Replica According to the Size of the Explored Surface, the Number of Points, Depth Resolution, and Recording Time
Instrument Quantimet 970 Standard I.A. Mechanical Profilometer Optical (laser) Profilometer Confocal microscope (TSM)
(a)
(b)
Surface Area (mm2) 100 50 25 50 1
Picture Depth Points Resolution Time (no.) (μm) (min) 300,000 180,000 62,500 250,000 250,000
8 8 1 3 1
5 2 90 10 10
(c)
(d)
FIGURE 20.2 The selection of shadows by the image analyzer. In this series of pictures, white features correspond to binary images of shadows segmented at threshold level 45 (A) the principal orientation of furrows is perpendicular to the lighting; (B) binary image resulting from the vertical opening of A. (C) the principal orientation of furrows is not perpendicular to the lighting; (D) the vertical opening almost suppressed the nonperpendicular shadows.
Quantimet, one can analyze 1 cm2 of skin with a resolution of 10 mm in the horizontal plane and 8 mm in the vertical plane. One question that arises is whether more weight should be given to the field of measurement or to the resolution. Our experience shows that the area measured should be greater than 0.5 cm2 and that a resolution of 10 mm is adequate to study the primary lines. The most important factors are the sampling procedure and the reproducibility of the measurements. Obviously, the study
157
of secondary furrows needs those higher resolutions provided by profilometry. But software should be capable of analyzing separately the various classes of skin furrows. 20.3.3.2 Analysis of the Image The selection of the shadows is the most important phase of the operation. The 6-bit analog-to-digital (A/D) converter provides a scale of 64 shades of gray (0 = black, 63 = white). Figure 20.2A and C show the results obtained with a threshold at level 45, i.e., selection of all the levels between 0 and 45. The position of the light source relative to the video scan determines the significance of the measurement parameters: the light comes from the right of the screen. Subsequent operations consist of extracting only the shadows formed by peaks perpendicular to the light source, i.e., in a vertical direction. To do so, the binary image is opened by 15 pixels (erosion + dilation) with a linear, vertical structuring element (Figure 20.2B and D). This mathematical morphologic operation8 has two consequences: to eliminate small events (noise, round objects, bubbles) and to emphasize the anisotropy of the skin furrows (suppression of oblique and horizontal shadows, joining of vertical shadow segments). 20.3.3.3 Measurement Parameters The two parameters measured in binary images are known as field parameters. They correspond, in fact, to the very first stereological parameters available on image analyzers when the latter were incapable of identifying objects. The area fraction AA is the percentage surface area occupied by shadows in the field of measurement. The intercept I is the horizontal projection of these shadows. Given the particular arrangement of the replica–lighting–camera ensemble, the I lines correspond exactly to the crests of the skin furrows, and thus represent the total length of furrows in the field of analysis (Figure 20.3). On the basis of these two stereological parameters and the angle of incident light a, I can be used to estimate the number of lines per square centimeter (or per linear centimeter), while AA, I, and a can be used to estimate their mean depth. These parameters are measured at each step of the sample rotation (9˚ steps through 360˚, giving 41 series of measurements). If the values of I are plotted in polar coordinates according to the angle of rotation, one obtains the rose of directions illustrated in Figure 20.4. This graph contains local maxima, with a 180˚ symmetry, which corresponds to the main orientations of the primary lines. The orientation of each network of parallel furrows can then be deduced relative to the reference axis (the tongue of the sample); the density of lines and their mean depth can also be deduced for each main axis of furrows. All these
158
Handbook of Non-Invasive Methods and the Skin, Second Edition
Lighting
k1 = 2.10 −4 ⋅ π ⋅ n ⋅ p k1 =
Axis 1
1 − k12
⎛ k + k2 ⎞ If k1 > 1 then c = ln ⎜ 1 ⎝ k1 − k2 ⎟⎠ AA
I Axis 2
⎛k ⎞ If k1 < 1 then c = π − 2 arctan ⎜ 1 ⎟ ⎝ k2 ⎠ 2 ⋅ k1 c + π π ⋅ k2
Finally, E = L
α
FIGURE 20.3 Parameters selected by the image analyzer. AA is the fraction area of shadows, I is the intercept line drawing the crest line of negative furrows, α is the lighting angle.
If E1 and E2 are the coefficients for each of the two axes mutually forming an angle b, CDSS = E1 [1 + (E2 – 1) sinb]. The importance of this parameter has been shown in studies of chronological aging,9,10 the effect of ultraviolet irradiation,11 and the skin deformation process.12 20.3.3.4 Automation
Microrelief: 4427
2
Full automation of the image analysis procedure has been the subject of a previous publication.13 A new “robot” has recently been designed to increase the analytical capacity without augmenting the time required for a full cycle. Eighty replicas instead of 40 can be analyzed in an 8-hour period. Figure 20.5 shows details of the automated
a
1
b
c
FIGURE 20.4 A rose of direction. Arrow 1 gives the orientation of the first axis of furrows and arrow 2 the second axis.
operations, i.e., the analysis of the orientations and parameters of each network, are carried out with a computer program written in Pascal. A supplementary parameter has been forwarded to characterize the relief by means of a single value. It is known as the coefficient of developed skin surface (CDSS), which represents the tissue reserve or deformation reservoir of the skin. On the basis of the mathematical cycloid arch model,4 a coefficient E is calculated for each furrow axis. A simplified formula is provided below, which is valid when the field of analysis is close to 1 cm2 (±10%). Consider n the number of furrows per centimeter squared and p their mean depth in microns determined for a main axis:
d
e
f
FIGURE 20.5 The “robot”. (a) Objective lens, (b) optical fiber lighting, (c) the circular tray accepts 80 samples, (d) a replica in position for analysis, (e) magnetic arm, (f) autofocus motor.
Skin Surface Replica Image Analysis of Furrows and Wrinkles
The lighting angle plays a crucial role in determining the nature of the shadows that will be analyzed: the lower the light source, the greater the detail. In this way, the lighting acts as a high-pass filter. In the case of a fine and regular microrelief, such as that observed in a child, an angle of 17 or 20˚ will enable the observer to measure a large number of shallow furrows (Figure 20.6A), while an angle of 38 or 45˚ will select the few deeper furrows. In this type of topography, line density is the most sensitive parameter. In the case of aged skin (Figure 20.6B), from which the fine lines have disappeared, the lighting angle has little influence on the number of furrows, but more influence on the mean depth. With a very low angled light source, two risks arise. If the sample is not perfectly flat, large areas of shadow can be generated, but this artifact is generally easy to identify in the measurements. Very high peaks can mask lower peaks located in their shade, and this bias is, on the contrary, difficult to detect. A compromise solution has been based on the value of the coefficient E1 as a function of the lighting angle (Figure 20.7). This coefficient passes through a maximum that depends on the surface topography. For the skin microrelief, the maximum is from 17 to 26˚; we use the higher value to avoid the above-mentioned risks. The optimum value when studying crow’sfeet is 38˚.
20.4 SOURCES OF ERROR 20.4.1 REPLICA ARTIFACTS The artifacts created during the production of the skin surface replica arise from the preparation of the volunteer and the technician’s experience. It is essential that the subject remain immobile during the polymerization phase. To this end, the room should be calm, with dim lighting and a temperature of 20˚C; in addition, the subject should be comfortable and given time to relax and adapt to the surroundings. These considerations are particularly important when taking replicas of the crow’s-foot area; in this case, the volunteer is placed in the lateral decubitus position, eyes closed and face relaxed.
30
μm
20
50
10
25
Depth
LIGHTING ANGLE
Young skin
0
0 17 20
26
32
38
45
Incident light angle (degrees) μm 150
Aged skin
30
125
20
100
10
75
0
Depth
OF
Line density per cm of skin
20.3.4 CHOICE
40
Line density per cm of skin
apparatus. A circular tray presents each sample, which is translated by a magnetic arm for positioning under the camera. The replica is then raised by the autofocus motor of the Quantimet and rotated by a stepwise motor. At the end of the analysis, the replica is lowered, placed in its socket, and the next sample is presented for analysis. The time previously required to control the position of each sample is thus saved.
159
50 38 17 20 26 32 Incident light angle (degrees)
45
FIGURE 20.6 Density of lines and mean depth of furrows plotted vs. the lighting angle: (A) young skin replica of the volar forearm, (B) aged skin replica of the volar forearm.
The conditions necessary for obtaining negative replicas must be followed to the letter. The main sources of artifacts are as follows. Bubbles of sweat can form holes at the intersection of primary lines if the room is too warm or the subject stressed or emotional (Figure 20.8). Areas lacking any relief are due to inadequate mixing of the resin with the hardener (Figure 20.9). One of the most frequent causes of artifacts is polymerization of the resin before it is applied to the skin. To ensure that application has been carried out correctly, the underside of the replica should be inspected: if the surface is smooth and shiny, the application is correct; if the surface is irregular, matte, and wrinkled, the resin had already started to harden before application (Figure 20.10).
160
Handbook of Non-Invasive Methods and the Skin, Second Edition
E1 1.25
1.20 c 1.15 a
b
b
FIGURE 20.9 Replica artefacts: (a) an “academic” replica, (b) flat areas (arrows) obtained with a nonhomogenous mixture.
1.10 a 1.05
α°
1.00 17 20
26
32
38
45
Incident light angle (degrees)
FIGURE 20.7 Evolution of the cycloid arch coefficient E1 (see text) according to the lighting angle: (a) young skin of Figure 20.6A, (b) aged skin of Figure 20.6B, (c) crow’s-feet wrinkles.
a
b
FIGURE 20.10 Replica artefacts: (a) the smooth verso of an “academic” replica, (b) aspect of a replica performed during the polymerization of the resin.
Interfering lights and changes in the ambient lighting influence the reproducibility of the measurements, and it is best to work in total darkness. Incorrect focusing is also a source of error, and autofocusing is thus a major advantage. As a rule, all these types of error can be overcome by regular analysis of a standard replica during a series of measurements.
FIGURE 20.8 SEM picture of a skin surface replica showing droplets at the intersect of primary furrows.
20.4.2 ANALYSIS
OF THE IMAGE
A frequent source of error is that the replica is not perfectly flat. Analysis through 360˚ permits this type of error to be identified simply, since the values for each 180˚ segment should be symmetrical. Normally, the rigid metal sample support enables most flatness errors to be corrected. All other errors are due to the image analyzer. The most frequent are time shifts and instability of the light source or electronic circuits of the camera. The uniformity of the incident light is also a critical factor: in general, there is a light gradient that has to be corrected by the image analyzer (background substraction).
20.5 CORRELATION WITH OTHER METHODS There have been few publications comparing mechanical profilometry and image analysis by shadowing. In 1985, we reported a comparison of the effects of antiwrinkle product on the face, using mechanical profilometry for the wrinkles of the forehead and image analysis for the crow’s-foot wrinkles.14 The percentage reductions were 27% for the profile area and 30% for the CDSS. The depth of the deeper lines (>50 mm) fell by 23% in the profilometric method and 16% in the image analysis technique. Schmidt et al.15 compared the two methods on the same replicas taken from the crow’s-foot area. They found a good correlation between the depth given by image analysis and the height of the peaks given by profilometry,
Skin Surface Replica Image Analysis of Furrows and Wrinkles
and also between the CDSS and peak surfaces. The best correlation was obtained with peak heights between 50 and 100 mm, i.e., the optimal domain in image analysis with a lighting angle of 38˚. Hayashi et al.16 recently demonstrated that the fractions of shadow corrected for the lighting angle (RWA parameter) and the maximum depth of the wrinkles (V) were independent of the lighting angle. V was identical to the height given by a micrometer. Schrader and Bielfeldt17 compared data from image analysis, mechanical profilometry, Corneometer®, and methylene blue staining of the skin during cosmetic treatments. Linear regression study led to weak but significant correlations between the methods. Some authors have preferred a profilometric approach to a threshold method in image analysis: the optical profilometry traces curves corresponding to gray levels along a scanning line in the video image. The profiles are analyzed with the same parameters as those used to analyze a mechanical trace, and good correlations between the two types of profile have been reported.18 Grove and Grove,19 also using optical profilometry by image analysis, showed that there was a good correlation between the roughness parameters (Ra, Rz) and Daniell’s visual classification during treatment of the crow’s-foot area with retinoids.
20.6 RECOMMENDATIONS The study of the skin microtopography or facial lines by means of image analysis, whatever the technique used, is more a problem of sampling and reproducibility than one of sensitivity. It is thus best to use approaches that analyze as large an area as possible, and a large number of samples. The time required for analysis is consequently important, and automation clearly has a role to play. In the special case of crow’s-foot wrinkles, the number of significant events (lines) is relatively small. Studies of changes in these lines during treatment must thus be based on strictly identical areas of analysis (before and after treatment). The image analyzer is a powerful tool in this setting, since it enables the study areas to be superimposed. As we have seen, the production of the replica is a crucial step that requires the utmost care. This is an essential consideration because, despite the fact that the replica method permits samples to be studied some time after their collection, it is essential that the replicas truly reflect the state of the skin at the time they are made. Replicas are durable and easy to store, and are also readily transportable. It is thus astonishing that so few multicenter studies comparing the different available methods have been published. A consensus on the effects of antiaging treatments could be arrived at by this approach, which would finally convince the international scientific community of their efficacy.
161
REFERENCES 1. Cook, T.H., Profilometry of skin. A useful tool for the substantiation of cosmetic efficacy, J. Soc. Cosmet. Chem., 31, 339, 1980. 2. Makki, S., Mignot, J., and Zahouani, H., Statistical analysis and three dimensional representation of human skin surface, J. Soc. Cosmet. Chem., 35, 311, 1984. 3. Hoppe, U. and Sauermann, G., Quantitative analysis of the skin surface by means of digital signal processing, J. Soc. Cosmet. Chem., 36, 105, 1985. 4. Corcuff, P., de Rigal, J., and Lévêque, J.L., Image analysis of the cutaneous microrelief, Bioeng. Skin Newslett., 4, 16, 1982. 5. Hashimoto, K., New methods for surface ultrastructure. Comparative studies of scanning electron microscopy, transmission electron microscopy and replica method, Int. J. Dermatol., 13, 357, 1974. 6. Makki, S., Barbenel, J.C., and Agache, P., A quantitative method for the assessment of the microtopography of human skin, Acta Derm. Venereol., 59, 285, 1979. 7. Gordon, K.D., Pitting and bubbling artefacts in surface replicas made with silicone elastomers, J. Microsc., 134, 183, 1984. 8. Serra, J., in Image Analysis and Mathematical Morphology, Vol. 2, Theoretical Advances, Serra, J., Ed., Academic Press, London, 1988. 9. Corcuff, P., de Rigal, J., Makki, S., Lévêque, J.L., and Agache, P., Skin relief and aging, J. Soc. Cosmet. Chem., 34, 177, 1983. 10. Corcuff, P., Lévêque, J.L., Grove, G.L., and Kligman, A.M., The impact of aging on the microrelief of periorbital and leg skin, J. Soc. Cosmet. Chem., 82, 145, 1987. 11. Corcuff, P., François, A.M., Lévêque, J.L., and Porte, G., Microrelief changes in chronically sun-exposed human skin, Photodermatology, 5, 92, 1988. 12. Corcuff, P., de Lacharrière, O., and Lévêque, J.L., Extension induced changes in the microrelief of the human volar forearm: variation with age, J. Gerontol. Med. Sci., 46, 223, 1991. 13. Corcuff, P., Chatenay, F., and Lévêque, J.L., A fully automated system to study skin surface patterns, Int. J. Cosmet. Sci., 6, 167, 1984. 14. Corcuff, P., Chatenay, F., and Brun, A., Evaluation of anti-wrinkle effects on humans, Int. J. Cosmet. Sci., 7, 117, 1985. 15. Schmidt, C., Camus, C., Candiu, H., Her, C., Soudant, E., and Bazin, R., Correlation d’une technique d’analyse d’images et d’une méthode profilométrique dans l’étude des rides de la patte d’oie, Int. J. Int. Sci., 9, 21, 1987. 16. Hayashi, S., Matsuki, T., Matsue, K., Arai, S., Fukuda, Y., and Yoneya, T., Changes in facial wrinkles by aging and application of cosmetics, in Proceedings of the IFSCC Congress, Yokohama, Japan, 1992, p. 733. 17. Schrader, K. and Bielfeldt, S., Comparative studies of skin roughness measurements by image analysis and several in vivo skin testing methods, J. Soc. Cosmet. Chem., 42, 385, 1991.
162
Handbook of Non-Invasive Methods and the Skin, Second Edition
18. Marshall, R.J. and Marks, R., Assessment of skin surface by scanning densitometry of macrophotographs, Clin. Exp. Dermatol., 8, 121, 1983.
19. Grove, G. and Grove, M.J., Effects of topical retinoids on photoaged skin as measured by optical profilometry, in Methods in Enzymology, Packer, L., Ed., Academic Press, New York, 1990, p. 360.
Method for Skin Surface 21 Stylus Contour Measurement Johannes Gassmueller BioSkin Institut für Dermatologische Forschung und Entwicklung GmbH, Hamburg, Germany
Andrei Kecskés Schering AG, Dermatologie/Humanpharmakologie, Berlin, Germany
Peter Jahn Schering AG, Diagnostika Koordination, Berlin, Germany
CONTENTS 21.1 Introduction............................................................................................................................................................163 21.2 Skin Surface Measurement....................................................................................................................................164 21.3 Methodological Principle ......................................................................................................................................164 21.3.1 The Replica................................................................................................................................................164 21.3.2 The Stylus Method ....................................................................................................................................164 21.3.2.1 Technical Equipment..................................................................................................................164 21.3.2.2 Profile Characterization..............................................................................................................165 21.3.2.3 A Recent Development: The Touchless Acoustic Stylus ..........................................................165 21.4 Sources of Error.....................................................................................................................................................166 21.4.1 The Test Area.............................................................................................................................................166 21.4.2 Taking the Replica.....................................................................................................................................166 21.4.3 Handling the Equipment............................................................................................................................166 21.5 Correlation with Other Methods ...........................................................................................................................166 21.5.1 Laser Profilometry .....................................................................................................................................166 21.5.2 Image Analysis ..........................................................................................................................................167 21.6 Recommendations..................................................................................................................................................167 Acknowledgment.............................................................................................................................................................168 References .......................................................................................................................................................................168
21.1 INTRODUCTION Measuring surface texture requires a precise understanding of what is meant by surface. A surface is the boundary between two media. In medical terms, the surface of the skin is the boundary between the individual and the physical environment with all its diverse influences on the function and structure of the skin. In his handbook for surface texture analysis, Mummery1 gives a very plastic description of what we have to deal with: “When specifying a surface profile, the examiner must be aware, that there are as many surface profiles as there are landscapes. The Himalayas and the Black Forest in Germany are both
mountain ranges. This is where the similarity between them ends. The two mountain ranges differ not only in magnitude (height of the peaks) but also in their form (the shape of the peaks and valleys). Describing a surface is as complex as describing a mountain range.” The landscape of normal or impaired skin is determined by the cutaneous architecture. The arrangement and interlocking of adjacent keratinocytes, the epidermal rete ridges, the dermal papillary structure, and the cutaneous appendages all contribute to the surface form. Both internal and external processes such as aging, dehydration, hydration (cosmetic products), or atrophy 163
164
Handbook of Non-Invasive Methods and the Skin, Second Edition
(corticosteroids) continuously remodel the cutaneous landscape. Profilometry allows the objective measurement of the above-mentioned effects on skin surface and is of particular interest for the evaluation of the pharmaceutical or cosmetic action on the skin.
21.2 SKIN SURFACE MEASUREMENT The ideal approach for measuring the skin surface would be a touchless in vivo method. However, up to now no reliable and reproducible method has been reported that approaches the standard of indirect methods that rely on a negative replica of the skin. The main reason for the failure of direct measurements is the relative movement of the anisotropic skin surface while scanning the profile (pulsation of minor arteries, moving, trembling of subjects). Therefore, measuring the profile or surface texture of a replica remains the most common approach to describe surface characteristics related to the skin’s surface geometry.
21.3 METHODOLOGICAL PRINCIPLE In our first attempts in 1980 to measure the influence of topical corticosteroids on skin surface texture, we adapted the stylus method. Established instruments for profilometry, like the Hommel Tester (Hommelwerke GmbH, Schwenningen, Germany), were originally developed to measure tool traces that occur when working on a metallic surface. Later the method was modified for use in almost all areas of microgeometry. Accepted standardized parameters like roughness, waviness, total profile, spacing, and others serve to quantify surface texture. Therefore, surface texture analysis has grown into a field of its own with increasing importance not only for metalwork, but also for many other research and production branches. This is especially true for experimental dermatology.
FIGURE 21.1 Measuring set-up using a column-mounted linear traverse unit.
21.3.2 THE STYLUS METHOD A diamond stylus is traversed across the replica surface. An electrical signal equivalent to the vertical displacement of the stylus is amplified and converted into a digital signal. The digital information is analyzed by computer according to selected parameters for roughness (Figure 21.1). 21.3.2.1 Technical Equipment
Several materials have been used to obtain reliable impressions of the skin’s surface. After having tested a number of materials, Makki et al.2 found silicone rubber, a dental impression material, to be most suitable for this purpose. After evaluating several materials ourselves, we found Provil® L C.D.* silicone rubber impression material to be the material of choice. Since the Hommel Tester automatically calculates an inverse profile of the measured negative impression, a second impression of the negative primary cast is not necessary. This eliminates possible errors introduced by secondary positive casting.
Pick-up: The pick-up has two functions: it supports the stylus and acts as a transducer converting the vertical movement of the stylus on the surface to an electrical signal. For dermatological purposes, the pick-up consists of a diamond stylus that is cone shaped with a 5-mm tip radius and a 60˚ tip angle. To protect the surface of the replica (to avoid “ploughing” through the ridges), a sliding Teflon** skid (Figure 21.2) with a round-shaped tip is mounted beside the stylus. This results in a more even and constant scanning of the surface. Furthermore, the use of a skid eliminates waviness due to mechanical filtering of the profile. Transducer: The most common transducer used for dermatological purposes is a half-bridge inductive transducer. An inductive transducer is well suited for texture surface measurement because of its linearity and insensitivity to the surroundings (ambient temperature and humidity). Its compact size allows for packaging in a small housing. The resolution makes the measurement of displacements as small as 0.001 mm possible.
* Registered trademark of Bayer Dental Werke AG, Leverkusen, Germany.
** Registered trademark of E.I. du Pont de Nemours and Company, Inc., Wilmington, Delaware.
21.3.1 THE REPLICA
Stylus Method for Skin Surface Contour Measurement
165
Ra =
FIGURE 21.2 Special pick-up for use with a flexible replica.
Traverse unit: The traverse unit furnishes the relative movement between the replica and the pickup. The replica (Figure 21.1) is moved under the stationary pick-up mounted on a flexible pick-up holder. Computer: Instrument control, data analysis of the digitized signal, and data output are all computer controlled. Instruments are easily kept up-to-date by updating the software as standards change. 21.3.2.2 Profile Characterization Ra, average roughness: Ra can be called the grandfather of all roughness parameters and is still young enough to do a good job. Numeric values in micrometers (μm) are obtained. It is commonly employed because of the ease of calculation when using simple analog devices. Although several other standard parameters have been applied to quantify the profile of the skin, we still prefer the mean roughness value Ra. The different parameters for roughness are discussed in detail by Cook.3 From Figure 21.3 it can be seen that the quantity Ra is the average distance from the profile to the mean line over the traverse length of assessment. Ra is determined by the formula
x
FIGURE 21.3 Mean roughness value Ra.
y dx
(21.1)
0
lm is the traverse (scan) length and |y| is the absolute value of the location of the profile relative to the mean profile height (x-axis in Figure 21.3). Ra is a standardized roughness parameter according to relevant German and international industrial standards. Filtering: A profile filter can be compared to a sieve. If a pile of rocks and stones is put through a sieve, it will be separated into two piles. One pile consists of rocks unable to pass through the sieve, while the other is gravel able to pass through. The sieve hole size defines what is called rock and what is called gravel.1 The filtering of surface profiles follows the same rules. A filter with a defined cutoff length divides roughness (gravel) from waviness (stones). The cutoff length is analogous to the hole size of the sieve. Filtering does not change the original profile, but the way of looking at it. Statistical analysis: In addition to the average roughness Ra, the surface can be described by descriptive statistics such as variance, skewness, kurtosis, autocovariance, and autocorrelation functions, as well as Fourier analysis.1 In order to go a step further and perform a three-dimensional analysis, it is necessary to take a number of parallel measurements. This may be of interest for specific questions concerning nonhomogenic surfaces like the skin. In future mathematical models describing regularities and irregularities will probably play an important role in the description of three-dimensional images.
A very recent development that promises to revolutionize the standard stylus principle is the new touchless acoustic pick-up with a resolution of 10 nm, reflecting a precise image of the measured surface without mechanical alteration (NANOSWING, Hommelwerke GmbH, Schwenningen, Germany). The new pick-up is fully compatible with the standard equipment.4
Ra lm
∫
lm
21.3.2.3 A Recent Development: The Touchless Acoustic Stylus
y
lm = traverse length
1 lm
Principle: By means of an electronic sensor a standardized diamond tip is moved over the surface at a constant elevation. The vertical movement exactly corresponds to the surface profile. Function: In order to maintain the diamond tip at a constant elevation, the distance to the surface has to be measured continuously. To achieve this,
166
Handbook of Non-Invasive Methods and the Skin, Second Edition
oscillations with very small but constant amplitude are induced in the tip itself. The very low contact pressure of only about 0.2 mN does not result in a measurable impression on the elastic replica material. The power necessary to maintain the oscillation is measured by a precise electronic device completely integrated into the pickup. The closer the oscillating diamond tip gets to the surface, the more power is consumed as a result of the air friction between the tip and the surface. The power consumption of the oscillation serves as the input value for an electronic regulating circuit. The output is connected to an electric positioner for the diamond tip. Since the power consumption of the oscillation serves as a measure of the distance between the tip and the surface, maintenance of constant power results in a constant distance to the surface.
21.4 SOURCES OF ERROR When employing profilometry in experimental dermatology there are a number of pitfalls that can lead to wrong conclusions: (1) an adequate test site must be chosen, (2) the replica must be made, and (3) the equipment must be operated properly.
with an inner marking that leaves an impression on the hardened replica is necessary for orientation.
21.4.3 HANDLING
THE
EQUIPMENT
For comparable results the replica always has to be scanned in the same direction. The direction should be determined by the course of the tension lines. It is advisable to measure the replicas within a comparable time interval. The weight of the pick-up on the connecting surface always has to be the same (between 1 and 2 g) and should be controlled with a balance before the start of every measurement series. Precise measurements require care and a stable environment. The regular use of a calibrated roughness standard ensures proper instrument function. Vibration is one of the main sources of measurement error and should be avoided as far as possible. In most buildings walking through the room while measuring causes distinct vibration. Protection from air currents is a matter of course (open windows). With the standard stylus repeated measurements on the same replica should not be performed along the same line, but have to be carried out at a sufficient distance from the previous measurement.
21.5 CORRELATION WITH OTHER METHODS
21.4.1 THE TEST AREA
21.5.1 LASER PROFILOMETRY
A suitable test site for measuring influences of topical treatments should have an even surface with a regular structure and few or no hair follicles. The volar side of the forearm meets most of the requirements. Subjects with too many hair follicles, scars, tattoos, visible dehydration (detergents), extensive sun exposure, and pathological skin conditions (scaling) have to be excluded except under special circumstances.
Laser profilometry5 is a computer-assisted structural analysis of the skin surface that uses laser beams for touchless measurements with a very high resolution. A three-dimensional profile of the surface can be stored digitally. Different parameters of roughness can be determined. Additional mathematical and statistical procedures, such as Fourier transformation and autocorrelation function, complete the analysis. The touchless laser beam is said to allow more precise imaging of the peaks and valleys of a replica independent of its elastic properties (no bending of the peaks through contact) than the conventional stylus method. However, one must be aware of the limitations of the method that are caused by the geometrical and optical properties of the object of interest. One aspect is the critical inclination (10 to 15˚) of a profile, where the laser system tends to overestimate the real depth of the profile. Furthermore, a high optical contrast (dark to bright areas) of the scanned surface may lead to a misinterpretation of the profile. Under unfavorable conditions a mistake of several hundred percent may result because the laser registers the optical contrast as well as the geometrical profile. The profile features produced by the optical contrast are indiscernable from the geometrical profile. Further errors well known to the experienced investigator result from the entry of the beam into the replica, porous surfaces, or structures with optic imaging properties.
21.4.2 TAKING
THE
REPLICA
For precise casting the replica material should be of low viscosity, fast hardening, and elastic without shrinking. The material should not produce heat while hardening. Large residues should not remain on the skin after removing the replica. The most commonly used material is silicone rubber impression material for dental purposes. With Provil® L C.D. all the above demands are covered. Other materials are listed by Cook.3 Provil L C.D. is a two-component material supplied in a ready-to-use cartridge, eliminating the necessity of further mixing. If a two-component material is used that has to be mixed by hand, it takes time and practice to mix the two components rapidly enough but at the same time without air bubbles. If the substance is already too hard, the replica will not be sufficient. For most purposes a mold
Stylus Method for Skin Surface Contour Measurement
Reg-papier RP 50
Homelwerke
Reg-papier RP 50
167
Reg-papier RP 50
Homelwerke
Reg-papier RP 50
FIGURE 21.4 Skin replica before topical therapy with a corticosteroid ointment of medium strength.
FIGURE 21.4 (Continued). Skin replica 3 weeks after topical therapy with a corticosteroid ointment of medium strength.
These inevitable but not always predictable errors should be kept in mind when referring to the very high resolution of laser profilometry. In addition, to take advantage of the highest possible resolution over large areas of the replica surface, laser profilometry is very time-consuming (requiring hours per replica). It should also be mentioned that the equipment for the laser method is considerably more expensive than that for the stylus method (both standard or acoustic pick-up). Despite the technical advantages of the laser method compared to the conventional stylus method, it remains to be seen whether this method is actually superior in practice and equal to the new touchless acoustic principle applied to the stylus method.
21.6 RECOMMENDATIONS
21.5.2 IMAGE ANALYSIS The basic principle underlying image analysis is the measurement of shadows generated by incident lighting at the surface of a replica. Main target parameters are the number and mean depth of wrinkles. A major problem with this method is uneven lighting that may result from unlevel replicas that are true to the skin structure.
A well-standardized procedure to assess the influence of different corticosteroids on the epidermal macropattern is the Duhring chamber test6 in combination with profilometry. The loss of the detailed structure of the epidermis can be quantified (Figure 21.4). No residual cream should be present on the test site. In general, it is advisable to stop treatment 24 hours before any measurement. Skin replicas made of silicone rubber impression material have the disadvantage that their elastic properties may lead to incorrect measurement of the “mountain peaks.” On the other hand, the material allows for repeated measurements by taking several replicas from the same area without substantial alteration of the skin surface. Cyanoacrylate replicas give a stable imprint of the skin surface but alter the test site by marked stripping when removing the replica. One should keep an eye on the hydration of the skin because it influences the profile considerably. This is especially important when testing substances for their influence on aging skin. A reduced roughness may be due to temporarily enhanced hydration and not to reduced depth of the wrinkles because of changes in the elastic fibers.
168
Handbook of Non-Invasive Methods and the Skin, Second Edition
Whichever roughness parameter is chosen by the experienced examiner, he or she must be aware of the method’s limitations. It must always be kept in mind that while new technologies and parameters for skin surface texture analysis may reveal statistically significant differences between two skin conditions, the most important authority to validate significant data is clinical efficacy.
ACKNOWLEDGMENT We thank B. Hughes for help with the English translation.
REFERENCES 1. Mummery, L., Surface Texture Analysis: The Handbook, Hommelwerke GmbH, VS-Muehlhausen, Germany, 1992.
2. Makki, S., Barbenel, J.C., and Agache, P., A quantitative method for the assessment of the microphotography of human skin, Acta Derm. Venereol., 59, 285, 1979. 3. Cook, T.H., Profilometry of skin: a useful tool for the substantiation of cosmetic efficacy, J. Soc. Cosmet. Chem., 31, 339, 1980. 4. Personal communication from Volk, R., Hommelwerke GmbH, Schwenningen, Germany. 5. Saur, R., Schramm, U., Steinhoff, R., and Wolff, H.H., Strukturanalyse der Hautoberfläche durch computergestützte Laser-Profilometrie, Hautarzt, 42, 499, 1991. 6. Frosch, P.J., Kligman, A.M., and Wendt, H., The Duhring chamber test for assaying corticosteroid atrophy in humans, in Percutaneous Absorption of Steroids, Mauvais-Jarvis, P., Vickers, C.F.H., and Wepierre, J., Eds., Academic Press, London, 1980, p. 185.
22 Laser Profilometry Jan Efsen Novo Nordisk A/S, Bagsvaerd, Denmark
Steen Christiansen Technical University of Denmark, Institute of Manufacturing Engineering, Lyngby, Denmark
Hans Nørgaard Hansen Copenhagen, Denmark
Jens Keiding LEO Pharma, Ballerup, Denmark
CONTENTS 22.1 Introduction............................................................................................................................................................169 22.2 Object.....................................................................................................................................................................170 22.3 Methodological Principle ......................................................................................................................................170 22.3.1 Preparation of Object: Making a Replica .................................................................................................170 22.3.2 Control of the Optical Profilometer ..........................................................................................................170 22.3.3 Calibration Methods ..................................................................................................................................171 22.3.3.1 Optical Sensor ............................................................................................................................171 22.3.3.2 Air-Bearing Table .......................................................................................................................171 22.3.3.3 Software......................................................................................................................................172 22.3.3.4 Standard Roughness Specimens.................................................................................................172 22.3.3.5 Frequency Response Analysis....................................................................................................172 22.3.4 Summary of Calibration Methods.............................................................................................................172 22.3.5 Comparison of the Optical Profilometer with Mechanical Stylus Instruments .......................................172 22.3.5.1 Measuring and Data Collection with the Profilometer..............................................................173 22.3.6 Characterization of Surfaces with Stratified Structure .............................................................................173 22.3.6.1 Preprocessing the Parameter Calculations .................................................................................173 22.3.6.2 Algorithm ...................................................................................................................................174 22.3.6.3 Other Parameters ........................................................................................................................175 22.3.6.4 Using Three-Dimensional Parameters for Characterizing Skin Replicas .................................175 22.3.6.5 What Has Quantitative Analysis of Skin Structure Been Used For?........................................177 22.4 Recommendations..................................................................................................................................................177 References .......................................................................................................................................................................177
22.1 INTRODUCTION The skin is a most versatile organ. It functions simultaneously as a protective first-line defense for the body and an area of chemical communication between the body and the external world. The inside part of the skin has been studied by a series of techniques. Histological, biophysical, and biochemical
studies have led to a detailed understanding of the internal structure of the skin, i.e., organization of cell layers, fibers, and chemical constituents in different parts of the skin. More systematic studies of the outside part of the skin have been carried on since the 1930s. In the beginning the technique was used merely to describe features on the surface of the skin or a skin replica that could be observed in a microscope. 169
170
Handbook of Non-Invasive Methods and the Skin, Second Edition
More quantitative studies of skin structure started about 30 years ago and have included both two-dimensional and, more recently, three-dimensional studies. These methods can be subdivided into two categories:
Normally skin replicas, i.e., impressions that are used as reproductions of the skin relief, are used as study objects in profilometric studies. This chapter will give a short introduction to profilometry with an emphasis on the methodology of optical profilometry. The performance of a commercially available instrument is discussed, as is the skin structure data that can be obtained using this technique and how the dermatologist can make use of these data.
UBM Laserbeam
ire
ct
io n
X-direction
Air bearing table
Yd
1. Structure mapping based on analysis of an image of the surface. These systems include image analysis of videoscans, as in the Magiscan image analysis system,1 and the fully automated system developed by Corcuff et al.2 in which the shadows cast by incident light on a replica may be used to determine primary and secondary direction and depth of furrows and also determine skin surface area. 2. Topographical mapping of the structure based on scanning of surface height (two- or threedimensional scanning). This category includes profilometers — mechanical or optical.
Optical sensor (fixed)
FIGURE 22.1 Optical profilometer system. 12
14
Analogue output 13
4 2 1
3
8 5
10
22.2 OBJECT
6 7
The purpose of this chapter is to present the operation of an optical profilometer and how it may be calibrated. We also introduce the use of software for a three-dimensional surface description and give a practical example of the use of this system.
9
22.3 METHODOLOGICAL PRINCIPLE
7 6
Sensor 11
FIGURE 22.2 Autofocus principle.
must be carefully controlled in order to avoid air bubbles getting stuck in the replica.
22.3.1 PREPARATION OF OBJECT: MAKING A REPLICA The making of a replica is in principle very easy. In our routine we first attach an adhesive ring to the skin area to be studied. A silicone rubber (Silflo® Flexico) is mixed with a catalyst, and the mixture is distributed as a thin layer covering the central opening of the ring and is allowed to harden for 3 to 5 minutes. The ring with the replica attached is gently removed, and finally the replica is cut into an appropriate size with a cutting device. In principle, this method leads to an accurate replica of the skin surface structure.3 In practice, however, there are often problems. The skin under study may need pretreatment to remove scales, hair, etc. Due to the anatomy, it may be difficult to obtain a replica of an appropriate structure and size. The process of mixing rubber and catalyst
22.3.2 CONTROL
OF THE
OPTICAL PROFILOMETER
The optical profilometer (Microfocus 1080) used in this investigation was provided by UBM Messtechnik (Ettlingen, Germany). The profilometer consists of four main parts: an optical sensor, an air-bearing table, a control unit, and a computer. The optical sensor is fixed, and the object whose surface is to be measured is placed on the airbearing table. By means of the control unit the table can be translated in two perpendicular directions, an X-direction and a Y-direction, as illustrated in Figure 22.1. It is possible to move the table 150 mm in both directions. Operation of this system is based upon the autofocus principle. An illustration of the system is shown in Figure 22.2 and described in UBM Messtechnik.6 Infrared light
Laser Profilometry
(wavelength of 780 nm) emitted from a laser diode (1) is focused onto a small spot (diameter of 1 mm) by a system of lenses (8 and 9). The light reflected from the object surface (11) is directed back into the sensor and is imaged as a pair of spots onto an arrangement of photodiodes (5). This is done in such a manner that both diodes are illuminated equally only when the objective lens (9) is precisely in its focal distance from the surface. If the distance to the object changes, the focus point is shifted too, and the illumination of the photodiodes becomes unequal. This unequal illumination of the photodiodes generates a focus error signal. A control circuit monitors the error signal and moves the objective lens accordingly. This is the autofocus principle. The movement of the lens is accomplished by a coil (6) and magnet (7) arrangement. The vertical movement of the objective lens is registered by a light barrier measurement system (10) and corresponds to variations in the height of the object surface. The standoff distance between object surface and optical sensor is approximately 2 mm. The optical sensor has two vertical measurement ranges: ±50 and ±500 μm. It is possible to obtain a vertical resolution of approximately 6 nm in the range ±50 μm and 60 nm in the range ±500 μm. The surface of the object may be viewed during the measurements through a window (4) in the sensor by using a microscope (13) and a CCD camera (14). All operations of the system are controlled by a computer.
22.3.3 CALIBRATION METHODS Calibration and control of the optical profilometer is necessary to obtain reliable results. Our calibration procedure was divided into the following five parts: • • • • •
Control of optical sensor Control of air-bearing table Control of software Calibration against standardized roughness specimens Determination of the frequency response of the optical profilometer
The tests were carried out at the Technical University of Denmark. The tests are described in detail in Efsen and Hansen.4 22.3.3.1 Optical Sensor The linearity of the sensor was tested with a sine bar. This is a metal bar mounted on two gauge blocks to obtain a well-defined angle (Figure 22.3). The sine bar was measured with the optical profilometer and the linearity determined from these results. The test was carried out for vertical measurement ranges ±50 and ±500 μm.
171
Optical sensor
Optical sensor
Sine bar Gauge block
Gauge block
Air bearing table
FIGURE 22.3 Sine bar.
Optical sensor
Optical sensor
Max. angle
Roundness standard specimen
FIGURE 22.4 Roundness standard specimen.
The ability of the optical sensor to measure inclined surfaces was tested with a roundness standard specimen. This is a glass ball with an almost perfect roundness. This specimen was measured with the optical profilometer as indicated in Figure 22.4, and the maximum allowable slope of the surface determined. A supplementary test of the sensor covered an investigation of which surfaces the sensor was able to detect. Furthermore, the control circuit of the sensor was tested by changing control parameters in the software. 22.3.3.2 Air-Bearing Table The flatness of the movement of the air-bearing table was tested with an optical flat. The optical flat was measured in two areas: 1 × 1 mm2 and 100 × 100 mm2. The small area represents a typical three-dimensional area for investigation in mechanical engineering, and the large area covers almost the entire possible movement of the airbearing table. The positioning accuracy and repeatability in both the X- and the Y-axis direction was tested with laser interferometry. Furthermore, the perpendicularity of the two axes was tested with an angle plate.
172
Handbook of Non-Invasive Methods and the Skin, Second Edition
Calibration standard specimen ISO 5436 type A 6
Profile height (μm)
TABLE 22.1 Suitable Calibration Methods
Overshoot
4 2
Object
0
Optical sensor
−2 −4
Air-bearing table
−6 Overshoot
−8 −10
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
2
Profile length (mm)
FIGURE 22.5 Measuring a calibration standard specimen (DS/ISO 5436 type A) with the optical profilometer.
Software
Standard roughness specimens
Property
Test Specimen
Linearity Maximum allowable slope Flatness Positioning accuracy of axis Repeatability Perpendicularity of axis Parameter calculation Filter test
Sine bar Roundness standard specimen Optical flat Laserinterferometry
Background noise
Static amplificationa
Frequency response analysis a
Ability to measure oscillating surfaces
Laserinterferometry Angle plate Theoretic sine profile Signal analyzer, DIN 4777 Optical flat
Calibration standard specimen (DS/ISO 5436, type A) Determination of frequency response
Overshoot
22.3.3.5 Frequency Response Analysis FIGURE 22.6 PTB parameter specimen.
22.3.3.3 Software The software was tested with a theoretic sine profile. The results were compared to already existing and tested programs and calculations performed after the standard DS/ISO 4287/1.7 The filter characteristics were determined by means of a signal analyzer (Brüel & Kjær, type 2032). A filter test as described in DIN 47778 was carried out. 22.3.3.4 Standard Roughness Specimens The optical profilometer was calibrated with standard roughness specimens designed for mechanical stylus instruments. An optical flat was used to determine the background noise in the system during a measurement. A single grooved calibration standard specimen (DS/ISO 5436, type A9) was used to determine the static amplification of the profilometer (Figure 22.5). These measurements revealed overshoot of the sensor when measuring steep edges. A parameter specimen developed by Physikalisch Technische Bundesanstalt (PTB) (Germany)5 was used to test the ability to measure real surfaces (Figure 22.6). The PTB parameter specimens were not found suitable for calibration of the optical profilometer.
By the term frequency response we mean the steady-state response of the optical sensor to a sinusoidal input. This analysis reveals the ability of the sensor to detect vertically oscillating surfaces. When measuring surfaces with lateral waviness, the frequency response of the sensor determines which wavelength components it is possible to detect. The analysis showed that the vertical range of a surface influences the response of the optical sensor.
22.3.4 SUMMARY
OF
CALIBRATION METHODS
The tests found suitable for calibration of the optical profilometer are listed in Table 22.1.
22.3.5 COMPARISON OF THE OPTICAL PROFILOMETER WITH MECHANICAL STYLUS INSTRUMENTS The performance of the optical profilometer was compared with that of mechanical stylus instruments (Rank Taylor Hobson Talysurf 5 and Rank Taylor Hobson Surtronic 3P). The comparison consisted of measuring a wide range of processed surfaces, i.e., turned, milled, and lapped surfaces. Also, the well-defined surface of a PTB parameter specimen was measured. The standardized parameters calculated from profiles obtained by the optical profilometer were on all surfaces studied higher than the
Parameter values (optical profilometer)
Laser Profilometry
173
Optical profilometer versus mechanical stylus instrument
Optical system, f 1 mm
100 Mechanical system, r > 2 mm
10 1 0.1 0.01 0.001 0.001
0.01 1 10 0.1 Parameter values (mechanical stylus) Rt (μm)
100
FIGURE 22.8 Comparison between optical stylus (laser beam) and a typical mechanical stylus.
Ra (μm)
FIGURE 22.7 Measuring industrial surfaces with both an optical profilometer and a mechanical stylus instrument.
corresponding parameters from a mechanical stylus instrument. This is illustrated in Figure 22.7, where the Ra and Rt values are plotted. Values obtained from the mechanical stylus instrument are plotted along the X-axis, and values obtained from the optical profilometer are plotted along the Y-axis. If both instruments would have given the same parameter values, all the points should have been placed on the marked line. It is obvious that parameter values obtained with the optical profilometer are higher than those obtained with the mechanical stylus instrument. This can be explained by the fact that the 1-mm laser spot detects more and deeper valleys in the surface than a mechanical stylus instrument (radius of typically >2 mm), as shown in Figure 22.8. Furthermore, a mechanical stylus works like a mechanical filter (Figure 22.9). Finally, steep edges represent a problem, as illustrated by the overshoot phenomenon shown in Figure 22.5. On a system calibrated as described, it is possible to obtain a repeatability of 5 to 10% when measuring skin replicas. 22.3.5.1 Measuring and Data Collection with the Profilometer The operation of the profilometer is controlled by an extensive software that is an integrated part of the equipment. It makes automatic scanning of a series of objects possible. Scanning takes place with the object positioned on the air-bearing table. Several objects may be scanned in one measuring sequence. In a setup table it is possible to predefine up to 99 positions on the table, and to each position a specific scan area may be defined. The information on position and area may be saved in the memory of the computer. When used together with information on speed and intensity of scanning, it allows the user to set up automatic scan procedures for all these operations. It is possible also to automize the data analysis. An extended list of standard two-dimensional roughness parameters
FIGURE 22.9 The stylus working as a mechanical filter.
and a few three-dimensional parameters is included with the software. Power spectra and autocorrelation analysis in two and three dimensions and two-dimensional material ratio curves are some of the other parameters included. Software routines for three-dimensional material ratio parameters are, however, not included. In the following we will show how such parameters can be calculated and how they may be used for description of skin surface structure.
22.3.6 CHARACTERIZATION OF SURFACES STRATIFIED STRUCTURE
WITH
This is a very common problem in engineering surfaces, since different layers in a surface have different functional properties. If the top part of a surface does not look like the valley part, the surface cannot be characterized with commonly used parameters, such as Ra and Rz, as these parameters cannot distinguish up and down in the profile. This is also a problem in skin structure studies, as these parameters cannot distinguish a groove from a ridge. Therefore, other parameters that can distinguish stratified structures are needed. 22.3.6.1 Preprocessing the Parameter Calculations A complete description of the skin surface is very complex since the replica contains many waves caused by the specific test conditions, such as the location where the
174
Handbook of Non-Invasive Methods and the Skin, Second Edition
replica has been taken and the replica preparation method. Grooves, ridges, and holes must be classified as microstructure and must be separated from the general form (the macroform), e.g., the form of the leg, the arm, or the face, before a reliable characterization of the microstructure alone can be made. Generally the true measured geometrical form — known as the raw set of data — may be classified into three categories: • • •
Microstructure, e.g., small holes, marks, and cracks Macrostructure, e.g., a cylinder, plane, or conus An arbitrary level, depending on the vertical range in which the measurement has taken place
It is difficult to distinguish between these categories since no exact limit between micro and macro has been defined yet. Different methods, e.g., fast Fourier transformation and digital cutoff filtering, have been tried,10 and one of the most suitable ways to remove the macroform seems to be subtraction of the Nth-order least squares polynomial fit from the raw set of data, as shown in this formula: R(x, y) = Z(x, y) – F(x, y) where R(x, y) is the residual surface, Z(x, y) is the original datapoint, and F(x, y) is the least squares Nth-order polynomial fit. Different methods for definition of a reference plane have also been examined, e.g., the arithmetic mean plane and the least squares plane. The least squares plane method has been evaluated to be the best since this method is more robust than the arithmetic method. All parameters should be calculated with respect to this reference plane. 22.3.6.2 Algorithm When the form element has been removed and the reference plane is defined, the parameters can be computed after the following precept:11 1. The material ratio curve is computed as follows: a. The surface is truncated into 4096 levels. b. The relative volume of material in each level is calculated, leading to the material distribution. c. The material distribution is accumulated over the 4096 levels leading to the material ratio curve, which shows how much material is found above a specified surface level (this is illustrated in Figure 22.10 and Figure 22.11).
Heightdistribution
Y
l 3%
10%
FIGURE 22.10 Illustration of height distribution curve. Y
Material ratio curve
l 0%
100%
FIGURE 22.11 Calculation of the three-dimensional material ratio curve.
2. The 40% secant is moved along the materialbearing curve until the 40% secant with least slope is found. These points are marked with A and B. 3. A line is projected through the A and B to the intersection with 0 and 100% material ratio. These points are labeled C and D, as shown in Figure 22.12. The vertical distance between C and D is the core surface depth (Sk). Sr1 and Sr2 are the material ratios at the top and the bottom above and below the core surface. 4. The areas A1 and A2 are computed. Spk is defined as the height in the triangle that has area A1 and baseline Sr1. The Svk is defined as the height in the triangle that has area A2 and baseline Sr2. Sk — the core surface depth — measures the height of the core material portion. It depicts the flat-test part of the material ratio curve, i.e., the region with the greatest increase in material. A small Sk value generally indicates that the skin surface is very smooth, since the material volume is very large. This is seen, e.g., on the skin surface of a baby where the skin is virtually devoid of wrinkles. Spk — the reduced surface peak height — denotes the height of the surface peak projecting beyond the core surface. A low Spk value indicates that the surface does not include many ridges and may be free of extreme peaks. Svk — the reduced surface valley depth — denotes the proportion of surface valleys extending into the material below the core surface. It provides useful information on grooves and holes in the surface. Generally it can be said
Laser Profilometry
175
B
A 40% y
40%
γ Spk
β
Sk
α
sekant
Svk 0
Sr1
Peak area
50 Bearing area
Sr2
100
Valley area
FIGURE 22.12 Calculation of the three-dimensional material ratio curve parameters.
that large Svk values indicate that the surface contains a large amount of wrinkles. Basically, this set of parameters divides the surface into three categories: top, core, valleys. One parameter is related to each category, and therefore, the total set of parameters is required for a suitable characterization of surfaces with stratified structure. 22.3.6.3 Other Parameters Other useful parameters for characterization of skin surfaces can be the arithmetic mean value Sa, the difference between maximum and minimum height St and skewness Ssk. Digitally the Sa is calculated as follows: 1 Sa = MN
MN
∑Z
i
i=1
where M is the number of points per profile, N is the number of profiles, and Zi is the numerical distance from least squares mean plane to the ith residual surface height. Sa is a mean value and gives an overall impression of the roughness properties of the object. St is the maximum vertical distance between highest peak and deepest valley. This parameter gives information about extreme conditions, e.g., extremely deep holes. Ssk is the skewness of the material distribution curve. This parameter can be used to describe the shape of the surface height distribution. It is given by the formula 1 Ssk = MNSq3
MN
∑Z i=1
3 i
where M is the number of points per profile, N is the number of profiles, Zi is the distance from least squares mean plane to the ith residual surface height, and Sq is the root mean square deviation of the surface. For an asymmetric distribution of the surface heights, the skewness may be negative if the distribution has a longer tail at the lower side of the mean plane, which means that the surface consists primarily of valleys and holes. This is in contrast to positive skewness, which indicates a surface including many peaks and ridges. If skewness is zero, no primary trend can be seen. These three parameters have to be handled with care and conclusions based on these parameters alone must be avoided. The definition of Sa includes the numerical values, which means that this parameter cannot see the difference between up and down. St is very sensitive for single-point values. One extreme value that may be caused by a vibration during the measurement can result in misleading conclusions. It can be seen from the formula for skewness, Ssk, that this parameter is dimensionless, since it is normalized by Sq. Therefore, skewness can never give information on absolute properties of a surface; only relative properties can be characterized. 22.3.6.4 Using Three-Dimensional Parameters for Characterizing Skin Replicas Sixteen skin replicas were investigated. An area of 5.6 × 5.6 mm2 was measured with the optical profilometer and analyzed by means of the three-dimensional parameters described above. In this investigation only the parameters Sa, St, Svk, and Sk were calculated.
176
Handbook of Non-Invasive Methods and the Skin, Second Edition
PI/DTH
MM KH-1
REPLIKA JE/HNH
24. 05. 93 LEKH1LM
150 mm
−200 mm 200 mm
5.60 mm 20 p/mm
z
5.60 mm: 20 p/mm
y
0°
x
LEO
FIGURE 22.13 Example of a profile plot from category I.
PI/DTH
MM 04-I
REPLIKA JE/HNH
24. 05. 93 LEO41LM
150 mm
−200 mm 200 mm
5.60 mm 20 p/mm
5.60 mm: 20 p/mm
z 0°
y
x
LEO
FIGURE 22.14 Example of a profile plot from category IIb.
The surface plots obtained from the optical profilometer were divided into two categories by means of visual inspection: Category I: Characteristic waviness with or without craters Category II: No characteristic waviness but possibly craters
Category II could be further subdivided into these two subcategories: Category IIa: No characteristic waviness but craters Category IIb: No characteristic waviness and no craters
Laser Profilometry
177
22.4 RECOMMENDATIONS TABLE 22.2 Categorization of Skin Replica Based on ThreeDimensional Surface Parameters
Category
Description
Suitable 3D Parameters (μm)
Value
I
Waviness, with or without craters
Sa St Sk Svk
Sa ≈ 30 210 < St < 300 80 < Sk < 120 42 < Svk < 49
IIa
No waviness, with craters
Sa St Sk Svk
Sa ≈ 20 230 < St < 280 60 < Sk < 100 30 < Svk < 40
IIb
No waviness, without craters
Sa St Sk Svk
Sa ≈ 20 170 < St < 195 50 < Sk < 70 28 < Svk < 33
Figure 22.13 shows an example of a profile plot from category I, and Figure 22.14 shows a replica from category IIb. Table 22.2 shows the combination of parameters and parameter values found in this investigation. In this material it was possible based on a calculation of the parameters Sa, St, Svk, and Sk to place each replica into one of these three categories. All four parameter values were needed to place the replica in the correct category. 22.3.6.5 What Has Quantitative Analysis of Skin Structure Been Used For? Image analysis studies have been used by Corcuff and Lévêque12 to describe development of skin structure as a function of age. They have shown that the regularity of the skin relief decreases with age, associated with the disappearance of secondary lines. They also showed that the aging process is a combination of physical age and external factors acting on the skin surface. Grove et al.,13 among others, have used optical profilometry in clinical studies on the effect of retinoid treatment on wrinkles and have shown that roughness parameters Ra and Rz correlate with clinical gradings based on clinical ratings.13 Zahouni et al.14 have used profilometric studies to calculate volume and area of leg ulcers and have used the data for evaluation of wound-healing treatments. Image analysis of melanoma has recently been shown to be of help in the clinical diagnosis of pigmented lesions,15 and a profilometric study has shown that it is possible to distinguish between melanomas and nevocellular nevi from a profilometric analysis of replicas taken from these areas.
In this chapter emphasis has been put on controlling and standardizing the operation of the optical scanner. However, it is equally important to have a standardized method of developing the material for study — the replica. The process of producing the replica and the quality of the replica should be controlled. Sometimes the skin may need pretreatment prior to taking the replica, in order to remove scales or hair or clean the skin to remove residuals from skin surface treatments. With respect to the use of the profilometer, it is necessary to specify the size of the area measured, the speed of the scanner, and the scanning intensity (i.e., points/millimeter). In addition to these specifications, which are included in the standard menu prior to starting the measurement, a series of hardware specifications are available. The actual set values should also be specified in order to make reproducible measurements and for lab-to-lab comparisons. In the test of the profilometer a test of the software was included. Only a small part of the extensive software was tested, however. An important part is the use of filtering methods in the determination of roughness parameters. If a digital filtering algorithm is used for removal of form error instead of a polynomial fit, it is important to select the correct filter algorithm. A phase-correct filter (M-filter), not the common RC2 filter (which is not phase correct), should be used. A wrong choice of filter may introduce peaks where no such peaks can be identified in reality. Also, the cutoff wavelength should be carefully chosen. In general, the cutoff wavelength should be 0.2* the evaluation length.
REFERENCES 1. Grove, G.L. and Grove, M.J., Objective methods for assessing skin surface topography noninvasively, in Cutaneous Investigation in Health and Disease, Lévêque, J.-L., Ed., Marcel Dekker, New York, 1989, chap. 1. 2. Corcuff, P., Chatenay, F., and Lévêque, J.L., A fully automated system to study skin surface patterns, Int. J. Cosmet. Sci., 6, 167, 1984. 3. Cook, T.H., Profilometry of skin: a useful tool for the substantiation of cosmetic efficacy, J. Soc. Cosmet. Chem., 31, 339, 1980. 4. Efsen, J. and Hansen, H.N., Optisk ruhedsmåling, Ed. MM.93.34, Institute of Manufacturing Engineering, Technical University of Denmark, Copenhagen, Denmark, 1993. 5. Deutscher Kalibrierdienst, Physikalisch Technische Bundesanstalt, Calibration of Stylus Instruments, Guideline DKD-R4-2, Braunschweig, Germany, 1991.
178
Handbook of Non-Invasive Methods and the Skin, Second Edition
6. UBM Messtechnik, Microfocus Measuring System Manual, Ettlingen, Germany, 1992. 7. DS/ISO 4287/1, Surface Roughness. Terminology, Part 1, Surface and Its Parametres, 1st ed., Dansk Standardiseringsråd, København, 1986. 8. DIN 4777, Teil 1, Oberflächenmesstechnik. Profilfilter zur Anwendung in elektrischen Tastschnittgeräten, Phasen-korrekte Filter, Berlin, 1988. 9. DS/ISO 5436, Calibration Specimens: Stylus Instruments: Types, Calibration and Use of Specimens, 1st ed., Dansk Standardiseringsråd, København, 1987. 10. Stout, K.J., Sullivan, P.J., Dong, W.P., Manisah, E., Lou, N., Mathia, T., and Zahouani, H., The Development Methods for Characterisation of Roughness in 3 Dimensions, Vols. 1 and 2, BCR report, EC contract 3374/1/0/170/90/2, Centre for Metrology, University of Birmingham, Birmingham and L’Ecole Centrale de Lyon, Lyon, France, 1993. 11. Christiansen, S., Function-Related 3-Dimensional Definition of Surface Microtopography, Ph.D thesis, Technical University of Denmark, Institute of Manufacturing Engineering.
12. Corcuff, P. and Lévêque, J.-L., Age-related changes in skin microrelief measured by image analysis, in Aging Skin, Lévêque, J.-L. and Agache, P.G., Eds., Marcel Dekker, New York, 1993, chap. 13. 13. Grove, G.L., Grove, M.J., Leyden, J.J., Lufrano, L., Schwab, B., Perry, B.H., and Thorne, G., Skin replica analysis of photodamaged skin after therapy with tretinoin emollient cream, J. Am. Acad. Dermatol., 25(2), 231, 1991. 14. Zahouni, H., Assoul, M., Janod, P., and Mignot, J., Theoretical and experimental study of wound healing: application to leg ulcers, Med. Biol. Eng. Comput., 30, 234, 1992. 15. Busche, H., Connemann, B.J., Kreusch, J., and Wolff, H.H., Surface Topography in the Diagnosis of Malignant Melanoma, poster session, 3rd Conference of the International Society for Ultrasound and the Skin, Elsinore, Denmark, 1993.
Evaluation of Skin 23 Three-Dimensional Surface: Micro- and Macrorelief Jean Mignot Laboratoire de Métrologie des Interfaces Techniques, Institut Universitaire de Technologie, Besançon, France
CONTENTS 23.1 Introduction............................................................................................................................................................179 23.2 Aim of the Study ...................................................................................................................................................180 23.3 Improvements in the Apparatus.............................................................................................................................180 23.3.1 Focusing System........................................................................................................................................180 23.3.2 Triangulation System.................................................................................................................................181 23.4 Quantification of Surfaces: Microrelief ................................................................................................................182 23.4.1 Parameters Obtained from the Extension of Classical and Standard Roughness Parameters to Three Dimensions...............................................................................................................182 23.4.2 Statistical Analysis.....................................................................................................................................183 23.4.3 Textural Analysis of the Skin Surface.......................................................................................................184 23.4.3.1 Directional Quantification of Furrows .......................................................................................185 23.5 Quantification of Surfaces: Macrorelief................................................................................................................187 23.5.1 Quantification of Wrinkles ........................................................................................................................187 23.5.2 Quantification of Wounds..........................................................................................................................188 23.5.2.1 Performance and Results............................................................................................................189 23.5.2.2 Theoretical and Experimental Evolution of Healing.................................................................191 23.6 Conclusion .............................................................................................................................................................191 References .......................................................................................................................................................................192
23.1 INTRODUCTION Skin surface topography has been a matter of interest for dermatologists for 50 years. The first studies were carried out by Cummins and Midlo,1 and these were followed by many others by Wolf,2 Tring and Murgatroyd,3 and Marks and Saylan.4 However, all these studies have been bidimensional, which means that the surface itself has not been studied, but one or several profiles of the surface in one or several directions have been. This simplification was necessary because of the capacities of the equipment available: no measurement device was able to analyze a sufficient number of data related to the surface analyzed within a short time. Until recent years, these profilometric studies were made with equipment designed for measuring functional surfaces of parts of machines in the mechanical industry. The measuring part of these types of apparatus is a
mechanical sensor or stylus in direct contact with the surface to be analyzed.5,6 The detector of this sensor (Figure 23.1) is usually a thin diamond tip (radius of curvature from 5 to 10 μm) displaced at a constant speed on the surface (0.5 to 1 mm/sec) according to a direction determined by the operator.
Analog signal v = f(z)
Fixed coil
z Surface
Fixed coil x
Axis
Skid Profilometer tip
FIGURE 23.1 Principle of stylus profilometer. 179
180
Handbook of Non-Invasive Methods and the Skin, Second Edition
This type of device has several main drawbacks: •
•
•
It must not be in direct contact with the skin, because the pressure of the spheric head creates a major deformation of the skin surface.7 This is why soft replicas of the surface are made,8,9 from which solid replicas are obtained, using in most cases a polymerized material.10 The scanning speed is limited in two-dimensional profilometry systems because of the mechanical contact with the analyzed surface, and this results in slow data acquisition. The results obtained with two-dimensional analysis (or roughness parameters) show the topography of the surface in one direction only. A top view examination of the skin surface (Figure 23.2), or the study of the changes of the various frequential components of the surface (two-dimensional Fourier transform) in all directions, are sufficient to show the anisotropy of this surface (Figure 23.3).
These main drawbacks have two consequences: • •
Traditional surface measurement devices are not adequate. Because of its anisotropy, the analysis of the skin surface with profilometers is not reliable.
23.2 AIM OF THE STUDY Because of the disadvantages of profilometry, it seems necessary to develop new devices that quickly acquire data and analyze skin topography, in three dimensions, and with a satisfactory speed. Nowadays, only noncontact sensors have the capacity to meet these requirements. Furthermore, the anisotropy of skin surface demands a threedimensional treatment of data. μm 220 200 180 160 140 120 100 80 60 40 20 0
2.5 mm
2.5 mm
FIGURE 23.2 Cutaneous surface, top view.
FIGURE 23.3 Two-dimensional Fourier transform of the cutaneous surface in a 25-year-old male subject.
23.3 IMPROVEMENTS IN THE APPARATUS The study of the topography of a surface is usually carried out by scanning the surface with parallel profiles.11 In the case of the scanning of a surface of 512 ↔ 512 measurement points separated by a distance of 10 μm, data acquisition would take more than 2 hours if a mechanical sensor were used under standard speed conditions (0.5 mm/sec).12 This length of time is unacceptable for assembly line monitoring. The recent development of laser diodes and their miniaturization has given rise to the production of several devices based on two principles: focusing of the laser beam and optical triangulation.
23.3.1 FOCUSING SYSTEM The principle of this system, used in industry, is explained in Figure 23.4: the beam from the laser diode (λ ≈650 to 800 nm) is focused on a point of the surface to be analyzed. This surface is displaced under the sensor at a constant speed, and any variation in height of each measured point enlarges the beam. This unfocusing induces a decrease of the energy received by the detector, which sends a signal to the linear motor that positions the optical system, thus regulating the convergence, and refocuses the beam. At each point of variation in the relief of the surface, a refocusing of the beam is necessary. The measurements, therefore, of the displacement of the optical system are equivalent to the variations of the relief. This system is of interest because there is no contact with the analyzed surface; rapid measurements are theoretically possible. However, the mechanical displacement of the optical system, at each refocusing, requires time, and this waste of time diminishes the benefit that the principle of the system might bring. It does, however, have a very good vertical definition that reaches 0.1 μm.
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
181
Laser diode
Laser diode
Pre-amplifier
Prism
P.S. detector
P.S. detector
Pre-amplifier
Detector Collimator O
O
O
O
Lens L1
Focusing lens
Lens L2
Lens L2
Surface Surface
FIGURE 23.5 Optical profilometer: triangulation principle.
FIGURE 23.4 Optical focus profilometer.
For the measurement of skin relief made on silicon rubber negative replicas,13 average measurement speeds of 0.5 mm/sec in profilometry (two dimensions) and 0.3 mm/sec in three dimensions have been obtained. This system has another disadvantage: if it meets sudden changes in height on the relief, it has problems finding a new focusing point. These changes are often due to defects (bubbles) in the silicon rubber used to make the replicas.
6 μm
23.3.2 TRIANGULATION SYSTEM This kind of system, used in research laboratories only, is based on the classical principle of optical triangulation (Figure 23.5) utilized in noncoherent light. From a point of light projected on the surface being analyzed, a corresponding image is obtained on the surface of the detector: any variation of the image of the light on the detector corresponds to a change in the height. By using a laser diode (λ = 850 nm) and photodetectors with semiconductors (position-sensing detector [PSD]), the position of the image obtained on the surface of the PSD can be measured to the micrometer, thus giving a real sensitivity of about 3 μm. Figure 26.6 shows the image obtained with a roughness standard of a maximum amplitude of 6 μm, showing the present limitations of this type of system. Figure 26.7 and Figure 26.8 compare the results obtained with two sensors of different types: one with mechanical contact, the other with a triangulation system. The results show the validity of the latter system, even for the measurement of very short displacements. The main advantage of the triangulation sensor is its speed, which is limited only by the speed of precision translators. An average speed of 10 mm/sec and accelerations of 8 mm/sec2 when starting each measurement line can be reached. With this sensor, the acquisition time of an image of 5.12 × 5.12 mm with one measurement point every 10 μm is only 5 minutes.
2.5 mm 2.5 mm
FIGURE 23.6 Roughness standard (6 μm amplitude).
μm 180 160 140 120 100 80 60 40 20 0
FIGURE 23.7 Mechanical detector.
182
Handbook of Non-Invasive Methods and the Skin, Second Edition
Since the height z = f(x, y) is known, the quantification of the surface is obtained using different methods.
μm 160
23.4.1 PARAMETERS OBTAINED FROM THE EXTENSION OF CLASSICAL AND STANDARD ROUGHNESS PARAMETERS TO THREE DIMENSIONS
140 120
Since standardization of three-dimensional measurement of surfaces has not yet been established, extending the known two-dimensional parameters is the obvious procedure to follow. A tri-dimensional parameter can correspond to a bi-dimensional parameter; for example, one of the most common two-dimensional parameters is Ra (arithmetic mean deviation of a profile), which gives the arithmetic mean of the absolute values of the profile departure, within a sampling length, i.e.:
100 80 60 40 20 0
FIGURE 23.8 Optical detector.
Ra =
Another point of interest regarding this sensor is its wide measurement capability. It can measure a few tens of micrometers as well as variations in levels of up to 8 mm. Examples of the measurements of wounds showing wide surfaces and deep holes are given below.
1 N
N
∑ z(i)
where N is the number of experimental points. The same definition can be extended to surfaces: 1 SR a = N1N 2
23.4 QUANTIFICATION OF SURFACES: MICRORELIEF
N1
N2
∑ ∑ z ( i,j) i=1
μm 60 40 20 0 1
2
3 0
0 −10 −20 −30 −40 −50 −60 −70 −80 −90 μm
FIGURE 23.9 Skin profile and Abbott curve.
20
(23.2)
j=1
A similar extension can be obtained for all the parameters quantifying heights. The bearing curve of a profile is of special interest, as it is a fundamental parameter in the study of friction and wear problems. The graph in Figure 23.9 illustrates the
The image of a surface analyzed by a sensor, whatever its principle, is represented by the law z = f(x, y), known for a discrete number of points (usually 512 × 512 or 256 × 256 points), with a vertical range between 12 and 14 bits.
0
(23.1)
i=1
40
4 60
80
5 100
6
mm
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
183
−58.2 μm −107 μm z
y x
FIGURE 23.11 Autocovariance function of a sine surface. 25.86%
44.31%
29.83%
FIGURE 23.10 Surface of plateaus at different levels.
relationship between the values of the profile bearing length ratio to the profile section level. It shows the projected length at a determined height. By applying this principle to studies on the surface of the skin, the surface of plateaus, and consequently their average size (Figure 23.10), are obtained. z
23.4.2 STATISTICAL ANALYSIS Another way to study the skin surface is by using statistical analysis that quantifies its anisotropy. It is possible to associate the autocovariance C(α, β) to the height z(x, y) of each point of the surface:
C ( α, β ) =
1 N1N 2
N1
N2
∑ ∑ z ( x,y ) z ( x+iΔx,y+jΔy) i=0
j= 0
with α = iΔx
x
FIGURE 23.12 Autocovariance C(α, β) of a cutaneous surface in a 9-year-old child.
ied wavelengths that constitute the surface. Figure 23.13A and B illustrate the spectral density of the skin surfaces of two subjects of different ages. When G(kx, ky) is known, it is possible to find the anisotropy of surfaces.14 The spectral moments described as
β = jΔy
mpq =
(23.3) This value indicates the correlation between any point of the surface and another point distant from iΔx and jΔy, with Ox and Oy being the experimental steps in the reference directions x and y, and i and j the integers of any value. The function C(α, β) shows possible periodicities; examples are provided showing the results obtained from a strictly regular surface (Figure 23.11) or from the skin surface (Figure 23.12). Using C(α, β), the bi-dimensional Fourier transform G ( kx,ky ) =
∑ ∑ C (α, β) exp– 2iπ (αkx + βky) α
β
(23.4) is the spectral density of the surface: it represents the stored energy of any part of the surface and shows the amplitude distribution of the different components of var-
y
∑ ∑ kx ky G ( kx,ky) p
kx
p
(23.5)
ky
and particularly the moments of second order m20, m02, and (zero order) m00, give the anisotropy coefficient γ2 (according to Longuet-Higgins14):
γ2 =
⎧ m2 min = 1 ⎡⎢( m20 + m 02 )+ ( m20 + m 02 )2 + 4 m112 ⎤⎥ m 2 min ⎪ 2⎣ ⎦ with ⎨ 1⎡ 2 2 ⎤ m 2 max m 2 max = ⎢( m 20 + m 02 )+ ( m 20 + m 02 ) + 4 m11 ⎥ ⎪⎩ 2⎣ ⎦
(23.6) The main directions, i.e., the two directions on the surface in which the variations of m2 are extreme, are obtained from the values of m2 min and m2 max. The examples in Figure 23.14 illustrate the correlation between the visual aspect of the surface and the value of its coefficient γ2 of anisotropy. In the case of a purely isotropic surface, γ2 → 1, and in that of a purely anisotropic surface, γ2 → 0. The
184
Handbook of Non-Invasive Methods and the Skin, Second Edition
z
y z
x
y x (a)
z y z
x
y
FIGURE 23.13 (A) Spectral density G(kx, ky) function in a 9year-old male subject; (B) spectral density G(kx, ky) in an 86year-old male subject.
values indicated in Figure 23.14 show that the influence of aging is well described by variations of this parameter.
23.4.3 TEXTURAL ANALYSIS
OF THE
SKIN SURFACE
The particular structure of the skin, made up of plateaus crossed by valleys, can be quantified by special operators applied to the image in its entirety. The techniques developed for image analysis make the quantification of furrows possible. Using a method proposed by Peuker and Douglas,15 cutaneous furrows can be described by using techniques initially developed for earth relief study.
x (b)
FIGURE 23.14 (A) The skin surface of a 9-year-old male subject; (B) the skin surface of an 86-year-old male subject.
Any closed area, including a part of a furrow, shows particular variations of the local relief z = f(x, y), characterized by two valleys (Figure 23.15). Thus, the height variations (zi – zM) will show four changes of sign for any point M belonging to a cutaneous furrow, and the deepest furrows will be distinguished from the smallest by the value of the amplitude of the difference. Since any point selected belongs to a furrow (Figure 23.16), the density of furrows can be calculated, as well as their vertical distribution compared to the vertical distribution of the whole surface.
0
μm 70 0
Zi-Z M 0
0.5
1
mm
FIGURE 23.15 Detection of particular points (furrows): analysis of the height variations along a closed contour.
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
185
FIGURE 23.16 Recognition of the points in furrows.
The ability to separate points on furrows from points belonging to the whole surface obviously induces greater sensitivity. For example, antiwrinkle cosmetic products are developed to act mainly on deep furrows. An investigation of the whole surface would therefore drown the particular information needed in unnecessary experimental points. Furthermore, a selective study of the furrows of medium or large size (secondary or main furrows) can be obtained using discriminating thresholds. Another method used in classifying cutaneous furrows into two families is the Fourier analysis or frequential analysis of the surface. The Fourier transform F(u, v) of the surface z(x, y) is given by discrete variables x = k and y′ = k′: F ( nΔu, mΔv) =
∑ ∑ z ( k,k′ ) exp – 2iπ ( nk+mk′ ) N k
k′
(23.7) where u = nDu and v = mDv in the frequency space. One of the interesting features of this breakdown is that the entire surface can be reconstructed from only one part of the initial spectrum, i.e., after performing a lowpass filtering. The initial spectrum is composed of two parts: one with long wavelengths only (characteristic of plateaus separated by main furrows), the other with short wavelengths (characteristic of secondary furrows crossing the plateaus partially or completely), i.e.: k1
Z ( nΔu,mΔv) =
k2
∑ ∑ z ( k,k′ ) exp – 2iπ ( nk+mk′ ) N+ 0
0
∞
∞
∑ ∑ z ( k,k′ ) exp – 2iπ ( nk+mk′ ) N k1
k2
(23.8) A first approximation of the values of limits k1, k2 of wavelengths is given by calculating the average distance
between plateaus when the cutaneous surface is cut by its mean plane. With this value, an approximate surface is reconstructed using the inverse transform on the part of the spectrum showing the low frequencies only: Z LF ( k,k ′ ) =
∑ ∑ Z ( nΔu,mΔv) exp + 2iπ ( nk+mk′ ) n
m
(23.9) In Figure 23.17, such a reconstruction applied to a skin profile can be seen: (a) the initial profile, (b) the corresponding total spectrum, (c) the profile reconstructed on the wavelength band [k1, ∞], and (d) the low-frequency part obtained from determining the mean distance Sm between plateaus (the limit k1 = distance parameter Sm). The minima of profile (c) show the position of the main furrows. The main furrows, detected by one or the other of the above-mentioned methods, are then subtracted from all the detected furrows, leaving only secondary furrows. Both families of furrows can then be quantified separately (Figure 23.18), resulting in a good description of the skin surface relief. 23.4.3.1 Directional Quantification of Furrows Several methods of determining the linear forms of an image are available: one of them is the method proposed by Groch,16 which first scans the image under study in two perpendicular directions so that the forms can be detected as precisely as possible, and then uses thinning and compression algorithms (Lu and Wang,17 Holt et al.18) to reduce the calculation time. Such a method, used by Awajan et al.19 in the detection of cutaneous furrows, provides good general results but poor directional sensitivity. As proposed by Rosenfeld et al.21 and Duda and Hart,22 application of the Hough20 transform to the extraction of linear forms is a more precise method. Hough uses a change of space: two points on a straight line of the
186
Handbook of Non-Invasive Methods and the Skin, Second Edition
29.61 15.62 1.63 −12.36 −26.34 −40.33 0
1
2
3.1
4.1
5.1 (mm)
6.1
7.2
8.2
9.2
10.2
84
98
112
126
140
6.1
7.2
8.2
9.2
10.2
84
98
112
126
140
(z)
(a) 65 52 39 26 13 0 0
14
28
42
56
70 (f ) (b)
27.83 16.51 5.19 −6.14 −17.46 −20.78 0
1
2
3.1
4.1
5.1 (mm)
(z)
(c) 65 52 39 26 13 0 0
14
28
42
56
70 (f ) (d)
FIGURE 23.17 (a) Initial profile, (b) Fourier spectrum, (c) reconstructed profile from low frequencies, (d) Fourier spectrum of low frequencies.
image correspond to a point (r, υ) in the parameter space. In the space of parameters (r, υ), the resulting curve corresponds to each straight line intersecting a point xi yi of the image space: r = f ( θ, x i y i ) = x i cos θ + y i sin θ
(23.10)
Therefore, a sine curve in the parameter space corresponds to each point (xi yi) of the image space; the alignment of points on the image space will be expressed by a group of sine curves intersecting at the same point in the space of the parameters (Figure 23.19). Several preliminary operations are necessary before applying this transform to the investigation of cutaneous furrows. The reference image must be simplified so that
the points at the bottom of the furrows appear. As there are furrows of varied widths, it is necessary to develop a thinning technique for these linear forms so that an image can be obtained, which reproduces the original distribution of furrows and is composed of two different levels only: the first one classifying all the points belonging to furrows (level 1) and the second one presenting all the external points (level 0). The analysis of furrows is conducted as follows: • • •
Determination of cutaneous furrows Thinning of furrows Quantification
Furrows are localized by applying a local operator, which identifies a furrow on a point of the image. Several
(μm)
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
187
then the points likely to be eliminated will be
69.94 55.95 41.97 27.98 13.99 0
2 ≤ A (P ) ≤ 7 0
1
2
3
4
5.1
6.1
7.1
8.1
5.1 6.1 (mm)
7.1
8.1
9.2
The condition A(P) > 2 is necessary to retain the extremities of the skeleton, and the condition A(P) 7 states that the current point belongs to the edge. All the points detected by the first scanning, thus belonging to the edges of the furrows, are analyzed by a second scanning, and those satisfying the following relation are eliminated:
10.2
(a) 69.94 55.95 41.97 27.98 13.99 0 0
1
2
3
4
9.2
(23.12)
10.2
| ΣAi – ΣBi | = 3
(b)
with Ai and Bi defined in Figure 23.21. Examples of such a configuration are shown in Figure 23.22. The skeleton obtained after the application of this equation is not always of unit thickness, especially at the intersection of several furrows. To obtain a skeleton with a unit thickness, the point under discussion is eliminated as soon as one of the configurations shown in Figure 23.23 is found. The result of the application of these different operators, shown in Figure 23.24, is compared to the original cutaneous surface. From this result, the Hough transform can be seen to express the directional distribution of furrows (density and directions).
FIGURE 23.18 (a) Main furrows, (b) secondary furrows. O
X
Y r
300
2 points
200
3 points
100
23.5 QUANTIFICATION OF SURFACES: MACRORELIEF
0 −100 −200 0°
45°
90°
135°
180°
Up to this point, we have studied only cutaneous microrelief. Two other sorts of skin surface characteristics, wrinkles and wounds, can be studied. They demand other methods of measurement.
θ
FIGURE 23.19 Line detection using the Hough transform.
operators can be used, particularly those proposed by Awajan,23 Cocquerez,24 Groch,16 Jimenez and Navalon,26 and Montavert.27 Awajan uses a specific operator covering eight neighboring points, P1 to P8 (Figure 23.20): if A ( P ) =
∑ Pi
23.5.1 QUANTIFICATION
OF
WRINKLES
The quantification of wrinkles is possible by using the silicon rubber replicas described earlier, and a sensor with a sufficient vertical range, under the condition that a suitable method be utilized. Usually the aim of this kind of study is to test the efficiency of an antiaging product, and the investigation is therefore based on the comparison of
(23.11)
i
P5
P6
P7
1
0
1
0
0
1
1
0
1
1
0
1
1
1
1
P4
P
P8
0
1
0
0
1
1
0
1
1
1
1
1
1
1
1
P3
P2
P1
0
1
0
1
0
1
1
0
1
1
0
1
1
0
1
FIGURE 23.20 (a) Neighboring pixels, (b) masks used in the thinning algorithm (point elimination).
188
A1
Handbook of Non-Invasive Methods and the Skin, Second Edition
A2
A3
P B1
B2
B3
A1
A2
B1
B1
P
A3
B2
B2
B3
P
B3
A1
B2
B1
A2
B3
P
A1
A3
A2
A3
FIGURE 23.21 Configurations for pixel elimination.
1
1
1
−
1
1
0
−
1
0
0
−
−
1
−
0
1
1
0
1
1
0
1
1
0
0
0
0
0
−
0
−
1
−
1
1
1
1
−
1
−
0
−
0
0
0
0
0
1
1
0
1
1
0
1
1
0
−
1
−
−
0
0
1
−
0
1
1
−
1
1
1
FIGURE 23.22 Examples of application.
0
0
0
1
1
1
0
0
0
0
1
1
1
0
1
1
0
0
1
0
0
1
0
0
1
1
0
0
FIGURE 23.23 Masks used to obtain skeletons with a width of one unit.
To meet the first condition, replicas are made at t0 and t1 on similar parts of the body, the size of each replica being larger than the surface to be studied. Then the replica is scanned by a three-dimensional apparatus, either mechanical or optical, the scanned area also being larger than the wrinkle(s) to be examined. Both analyzed surfaces (Figure 23.25) are visualized: an area is selected on one of them and reproduced on the other one, either by a visual position control or using the intercorrelation of both samples, which consists of bringing to a maximum the product of correlation: C ( α, n ) =
∑ ∑ z ( x ,y ) z ( x + α, y + α ) 1
i
i
i
i
where z1 is the height of a point on the first image with coordinates xi yi and z2 is the height of a point on the second image. This product reaches a maximum when both images are correlated perfectly. Once selected and carefully positioned, the areas are quantified as follows. Each profile is obtained by cutting the analyzed surface by a plane perpendicular to the mean plane of the surface. The real volume of the wrinkle is then the sum of all the elementary volumes. An elementary volume is located between two parallel neighboring planes, the surface of the valley, and the two successive straight lines connecting the extreme points of each profile comprising the upper surface (Figure 23.26). The maximum depth h, on average, of the whole surface is calculated:
23.5.2 QUANTIFICATION
two types of results: the evolution of a wrinkle without treatment (the reference) and the evolution of another wrinkle after treatment. The results of the analysis depend on the different evolutions of both surfaces during the treatment (times t0 and t1). To be reliable, this comparison study must be conducted on similar surfaces (same size and necessarily similar position) and be quantified by representative parameters.
2
(23.13)
1 h= N
FIGURE 23.24 Skeleton of the area shown in Figure 23.16.
i
α
N
∑h
i
(23.14)
i=1
OF
WOUNDS
The geometric characteristics of large wounds can be measured with one optical system because of its vertical range capacity and its wide precision translators. A typical example is a leg ulcer, which can present differences in levels of several millimeters on a surface of a few square centimeters. The healing progress of such wounds is monitored by making silicon rubber replicas of the surface at various times. It is useful to make such surface replicas for two main reasons: •
Keeping surface prints is important because they can be used later to check measurements and even to try new methods. These replicas can be stored over a long time.
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
Perimeter: 12.1 mm Surface: 8.07 mm2 Mean height: 123 μm Mean depth: 352 μm Volume: 992 mm2 × μm
189
Perimeter: 12.1 mm Surface: 8.13 mm2 Mean height: 76.5 μm Mean depth: 213 μm Volume: 622 mm2 × μm
FIGURE 23.25 Measurement of topographical parameters made on the same area, before and after treatment.
the wound is absolutely painless. It can reproduce the tiniest details: amplitudes of about a micrometer can be detected, which is sufficient for this kind of investigation. The white color of this material is accepted by the optical triangulation captor. In addition, it takes only 3 to 5 minutes to make a replica, even a large one. hi
23.5.2.1 Performance and Results To analyze the surface of a wound (Figure 23.27), a scan of this surface is made line by line. The measuring step along a line and the distance between two neighboring lines are practically multiples of 10 μm. The result of this scanning is a table of N × L points, since the scanned area is generally rectangular. The acquisition time depends on the size of the surface and the motor used (step-by-step motors); the step motor can be of 10, 50, or 100 μm, producing measuring speeds of up to several tens of millimeters per second, the maximum frequency being 6000 steps per second. The maximum scale of the vertical measuring range is 8 mm; it can be increased by using a large-scale system.28,29
FIGURE 23.26 Evaluation of the volume of a part of a furrow.
•
Only one measuring system is used, even if the prints come from other sources, thus allowing multicentric studies. Risk of error is reduced with just one system only.
Materials like SILFLO®* have been widely tested in dental surgery and in dermatology. This material is completely safe and causes no wound reaction; application to * Registered trademark of Flexico Developments Ltd.
7.5 mm
42 mm
FIGURE 23.27 Ulcer replica scanning.
56 mm
190
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 23.28 Determination of the boundary of the ulcer.
Leg
axis
FIGURE 23.29 Elimination of the local form of the leg.
Perimeter: 639 mm Surface: 297 cm2 Mean height: 817 μm Volume: 24265 mm3
FIGURE 23.30 Geometric parameters of a leg ulcer.
Once the surface z(x, y) is analyzed, the geometrical parameters are obtained in the following manner: flattening the surface by the least squared plane and determining the edge of the wound, following it from a top view of the surface, with heights appearing as false colors (Figure 23.28). The perimeter of the boundary and the surface of the wound are the first parameters to be used, directly plotting the edges of the wound with transparent paper.30 This method needs the operator’s judgment, and it can be replaced either by following up the contour on a screen with a mouse or by an automatic determination of neighboring points.31 Three-dimensional representation of wounds has been tried by photostereogrammetry,32 but this method demands heavier apparatus and is not as accurate as the direct measurement of a relief by triangulation. The exact plot of coordinates x, y, z of any point on the surface plays a part in the evaluation of the volume of the wound, which is a valuable parameter for the study of the healing process. However, the volume is significant only after elimination of the local form of the leg. As shown in Figure 23.29, this elimination is made by scanning the surface of the wound parallel to the longitudinal axis of the leg, and at a distance always greater than the real size of the wound. With this method it is possible to obtain, at the beginning of each profile, part of the healthy skin, which will be the reference. For each profile, a segment of the straight line joins the reference parts and gives the general direction of the leg where there is no wound. The segments as a whole form a surface that reproduces the general shape of the leg without any wound; thus, the real wound that will be analyzed is represented by the volume between the surface measured (included inside the boundary) and the “initial” surface of the leg. Figure 23.30A shows the initial stage of an ulcer and Figure 23.30B the evolution of its boundary during the time of treatment, and the corresponding values of
Perimeter: 516 mm Surface: 185 cm2 Mean height: 658 μm Volume: 14239 mm3
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
191
A1 B1
C1
A1 t
C2
t + dt
A2
r1 B1
r2 B3 B4
A3 r3 A4 r4
FIGURE 23.31 Contour of the ulcer at moments t and t + dt.
parameters: perimeter, surface, and volume. Taking into account the volume of the wound avoids imprecise results in which the periphery and surface are likely to evolve in an unknown direction, whereas the volume decreases in accordance with the treatment duration.
FIGURE 23.32 Theoretical ulcerous leg evolution as a function of time.
23.5.2.2 Theoretical and Experimental Evolution of Healing In view of this ability to trace the boundary of the wound, interesting future developments can be considered, such as theoretical research on its evolution, and thus the prognosis of its evolution. Interesting theoretical research was carried out by Amiez33 using geometric criteria combined with the results of a clinical investigation. She studied the displacement of the contour of an ulcer over a specific period. If C1 and C2 are two contours at two very close moments t and t + dt (Figure 23.31), contour C2 is deduced from C1 by successive progressions defined by quantities r1, r2, r3 … of points A1, A2, A3 … of the contour at t. Any element such as Ai Ai + 1 scans the elementary surface defined by the quadrilateral Ai, Ai + 1, Bi + 1, Bi. With this method, the whole boundary can be split into several quadrilaterals. According to the clinical observations made by Agache,34 the change of the contour depends on its local curvature: an arc of the contour with a large curvature radius will change more quickly than an element with a smaller radius, so that the surface dSi scanned by the element di + 1 is given by dSi = Kdt di + 1, where K is a constant. Amiez links di + 1 to quantities ri, ri + 1, for the N points of the boundary, thus developing a system with N equations, with N unknown. Therefore, the resolution of the system is possible, and with the knowledge of the boundary at an instant t, knowledge of the contour at a further instant t + dt is produced. This new contour is then the basis for a new calculation, giving the solution at the next stage. The successive healing stages of the contour of the wound can be foreseen with this iterative method. Figure 23.32 gives an example of the theoretical evolution of the real boundary of an ulcer applying the above method; Figure 23.33 shows the results of the theoretical calculation as well as the experimental results (broken lines) applied to a real boundary.
FIGURE 23.33 Ulcerous leg contour evolution. Theoretical contour, continuous lines; experimental contour, broken lines.
23.6 CONCLUSION The study of the microrelief of the surface of the skin is of interest to dermatologists, surgeons, and manufacturers of cosmetic products. The skin relief connected to the dermis and epidermis is dependent on numerous parameters that play a role in its evolution. Therefore, precise knowledge of skin relief can provide important information regarding the effects that certain illnesses, aging, and radiation have on human beings. The capabilities of skin microrelief study have been greatly improved because of the existence of three dimensions in topographical studies. Because of the use of both the real space z = f(x, y) and the frequency (or wavelength) space, a large field of application has been opened up, where furrows can be separated into two families: one family that is connected to the dermis and one that is connected only to the stratum corneum. Parameters specific to the cutaneous surface can take the place of traditional two-dimensional parameters, which were borrowed from the methods used in measuring industrial surfaces. The transfer to three dimensions opens up the field of study of texture, because the direction of the furrows can be detected and quantified, showing in particular the effects of the local mechanical deformation and aging. The macrorelief of the skin is composed of furrows of great amplitude, wrinkles, and wounds.
192
Handbook of Non-Invasive Methods and the Skin, Second Edition
The study of wrinkles can be performed with much greater precision in three dimensions because the isolated detection of one and the same area at different stages of a treatment becomes possible. The quantification of geometric parameters such as surface, volume, and depth of the same part of a wrinkle offers a reliable way to test the efficiency of antiwrinkle products. Wounds evolve, like wrinkles, according to different parameters. The healing process can be measured in an objective and quantitative manner by following the evolution of the geometric parameters: perimeter, surface, and volume of the wound. These are detected by making replicas that are then utilized for measuring the relief. Analyzed areas of large size demand the use of noncontact sensors, which are the only devices capable of working in three dimensions at satisfactory speeds. The samples discussed throughout this chapter show the interest in and necessity for three-dimensional measurements of the geometric characteristics associated with the micro- and macrorelief of the skin, as well as all the potential possibilities of texture analysis.
REFERENCES 1. Cummins, H. and Midlo, C., Palmar and plantar epidermal ridge configuration in European-Americans, Am. J. Phys. Anthropol., 9, 471, 1926. 2. Wolf, J., Das oberflächenrelief der menschlichen Haut [Skin surface relief in man], Z. Mikr. Anat. Forsch., 47, 351, 1940. 3. Tring, F.C. and Murgatroyd, L.B., Surface microtopography of normal skin, Arch. Dermatol., 109, 223, 1974. 4. Marks, R. and Saylan, T., The surface structure of the stratum corneum, Acta Derm. Venereol., 52, 119, 1972. 5. Thomas, T.R., Recent advances in the measurement and analysis of surface microgeometry, Wear, 33, 205, 1975. 6. Snaith, B., Edmonds, M.J., and Probert, S.D., Use of a profilometer for surface mapping, Precision Eng., 141, 87, 1981. 7. Xie, Y., in Quantification de la topographie des surfaces, no. 246, Thesis, University of Besançon, France, February 1992, chap 1. 8. Sampson, J., A method of replicating dry or moist surfaces for examination by light microscopy, Nature, 191, 932, 1961. 9. Sarkany, I., Method for studying the microtopography of the skin, Br. J. Dermatol., 74, 254, 1962. 10. Cook, T.H., Craft, T.J., Brunelle, R.L., Norris, F., and Griffin, W.A., Quantification of the skin’s topography by skin profilometry, Int. J. Cosmet. Sci., 4, 195, 1982. 11. Williamson, J.P., The microtopography of surfaces, Proc. Inst. Mech. Eng., 182, 21, 1968. 12. British Standard 1134, 1972, revised 1988. 13. Makki, S., Barbenel, J.C., and Agache, P., A quantitative method for the assessment of microtopography of human skin, Acta Derm. Venereol., 59, 285, 1979.
14. Longuet-Higgins, M.S., The statistical analysis of a random moving surface, Philos. Trans. R. Soc. London Ser. A, 249, 966, 321, 1957. 15. Peuker, T.K. and Douglas, D.H., Detection of surface points by local parallel processing of discrete terrain evaluation data, Comput. Graphics Imaging Process., 4, 373, 1975. 16. Groch, W.D., Extraction of line shaped objects from aerial images using a special operator to analyse the profiles of functions, Comput. Graphics Imaging Process., 18, 347, 1982. 17. Lu, H.E. and Wang, P.S.P., An improved fast parallel thinning algorithm for digital pattern, in IEEE Computer Society Conference on Computer Vision and Pattern Recognition, San Francisco, June 19–23, 1985, p. 364. 18. Holt, C.M., Stuart, A., Cunt, M., and Perrot, R.H., An improved parallel thinning algorithm, Commun. ACM, 30, 156, 1987. 19. Awajan, A., Rondot, D., and Mignot, J., Quick method of measuring the furrows distribution on skin surface replicas, Med. Biol. Eng. Comput., 27, 379, 1989. 20. Hough, P.V.C., Method and Means for Recognizing Complex Patterns, U.S. Patent 3,069,654, December 18, 1962. 21. Rosenfeld, A., Thurston, M., and Lee, Y.H., Edge and curve detection for visual scene analysis, IEEE Trans. Comput., C2D, 562, 1971. 22. Duda, R.O. and Hart, P.E., Use of the Hough transformation to detect lines and curves in pictures, Commun. Assoc. Comput. Mech., 15, 11, 1972. 23. Awajan, A., Detection et analyse des structures linéaires d’une image. Applications biomédicales et industrielles, no. 78, Thesis, University of Besançon, France, 1988. 24. Cocquerez, J.P., Analyse d’images aériennes: extraction de primitives rectilignes et antiparallèles, Ph.D. thesis, Paris Sud (Orsay) University, France, 1984. 25. Cocquerez, J.P. and Devars, J., Détection de contours dans les images aériennes: nouveaux opérateurs, Traitement Signal, 2, 45, 1985. 26. Jimenez, J. and Navalon, J., A thinning algorithm based on contours, Comput. Vision Graphics Imaging Process., 99, 186, 1987. 27. Montavert, A., Obtention d’une ligne médiane par connexion de l’axe médian, in 5th Congress of AFCETINRIA, Grenoble, France, November 27–29, 1985, p. 777. 28. Chuard, M., Mignot, J., Nardin, P., and Rondot, D., Range expansion and automation of a classical profilometer, J. Manuf. Syst., 6, 223, 1987. 29. Zahidi, M., Assoul, M., Bellaton, B., and Mignot, J., A fast 2D/3D optical profilometer for wide range topographical measurement, Wear, in press. 30. Carrel, A. and Hartmann, A., Cicatrisation of wounds. The relation between the size of a wound and its rate of cicatrisation, J. Exp. Med., 24, 429, 1916. 31. Zahouani, H., Assoul, M., Janod, P., and Mignot, J., Theoretical and experimental study of wound healing: application to leg ulcers, Med. Biol. Eng. Comput., 30, 234, 1992.
Three-Dimensional Evaluation of Skin Surface: Micro- and Macrorelief
32. Eriksson, G., Eklund, A.E., Torlegard, K., and Dauphin, E., Evaluation of leg wear treatment with stereophotogrammetry, Br. J. Dermatol., 101, 123, 1979.
193
33. Amiez, G., Cicatrisation des ulcères, paper presented at the National Meeting of Numerical Analysis, Port Barcarès, France, 1988. 34. Agache.
Morphological Tree of the 24 The Cutaneous Network of Lines H. Zahouani Laboratoire de Tribologie et Dynamique des Systèmes, Ecully, France
Ph. Humbert Laboratoire de Biologie et d’Ingénierie Cutanée, Besançon, France
CONTENTS 24.1 Introduction............................................................................................................................................................195 24.2 Material and Method .............................................................................................................................................196 24.3 Fourier Transform of Skin Lines Network and Frequency Range.......................................................................196 24.3.1 Spatial Frequency Range of Lines Network .............................................................................................197 24.3.2 Fourier Spectrum Representation ..............................................................................................................197 24.3.3 Skin Lines Orientation in the Fourier Space ............................................................................................198 24.3.4 Spectral Rose of Skin Lines Anisotropy...................................................................................................198 24.4 Lines Identification by Anisotropic Spectral Filtering .........................................................................................198 24.4.1 Directional Extraction of the Wrinkle.......................................................................................................199 24.4.2 Directional Extraction of Lines Family ....................................................................................................200 24.5 Determination of the Three-Dimensional Tree Skin Spectrum............................................................................200 24.5.1 Statistical Analysis of Skin Lines Morphology ........................................................................................200 24.5.2 Morphological Tree of Skin Network of Lines ........................................................................................201 24.6 Conclusion .............................................................................................................................................................203 References .......................................................................................................................................................................203
24.1 INTRODUCTION The skin surface shows a specific topography depending on the anatomical site, age and sex. In general, the skin morphology presents a deterministic network of lines, who by its organization expresses all the multidirectional tensions of elastic fibers and the collagen beams. Microlines, primary lines, fine wrinkles and wrinkles represent, in fact, the special organization of collagen bundles and elastic fibers in the superficial dermis, and there is a relationship between the morphology of skin lines and elastic network [1,2]. Different functions can be attributed to the lines network. The first function is the retention and drainage canals of the sebum and sweat. They collect preferentially and retain for a long time the substances applied to the skin: they are thus preferential sites of percutaneous absorption.
This reservoir function allows the applied topical products to be stored on the skin surface and then eventually to diffuse in its different layers. The second function is mechanic; during aging the depth, width, density, and orientation of skin lines change. Some lines become more marked; they evolve progressively in marked anisotropy connected to the decrease of the elasticity of the collagen fibers. The resistance of the skin to traction prevails in a direction, different according to areas from the body, which follows the principal lines (Langer Lines). Langer Lines were established on the corpse starting from the oval form taken by a wound carried out with a round punch, and they correspond to the large axis of the oval. An excised skin narrows more and presents a minimum of extensibility in this direction. It thus acts of an anisotropy of the spontaneous tension of the skin, with distinguishing well from an additional tension of this one induced by a stretching of muscular or visceral origin. Hashimoto [3] gave a precise classification of the line network scales: the 195
196
Handbook of Non-Invasive Methods and the Skin, Second Edition
primary lines are clearly marked and are between 20 and 100 μm deep. The secondary lines are more discrete and correspond to a depth of 5–40 μm from the diagonals to the primary lines. Tertiary and quaternary lines cannot be seen visually. The tertiary lines correspond to the corneocyte border (about 0.5 μm) and the quaternary lines correspond to each corneocyte morphology (about 0.05 μm). In this work we have developed a new approach that is able to classify the skin network of lines with respect to the depth, width, and orientation. The method involves last microscope or optical fringe projection measurements [4,5] and image processing. Fourier transform and associated image processing techniques can also assist automated identification and processing of the features of interest. The global anisotropy between different skin line components has been represented in a polar histogram in the form of a spectral rose of the frequencies. The multiscale spectrum of the lines network can be represented as a three-dimensional tree that shows the hierarchic organization of the skin morphology vs. the scale of lines, wavelength, and orientation: f (ρ, λ, θ). This morphological spectrum is highly useful in finding the correlation between the scale of lines and aging or cosmetic treatment.
24.2 MATERIAL AND METHOD The last 30 years have seen a great step forward in the field of three-dimensional microscopy, especially in the analysis of the surface topography. The methodology was greatly improved when its principle was developed so as to make it possible to reproduce the surface topography in a three-dimensional space. This technological achievement allows scientists to reconstruct three-dimensional images from a nanometric scale and to reconstruct threedimensional images from a nanometric scale to a macroscopic relief. The vertical and lateral resolution and the maximum vertical range of the principal three-dimensional profilers used to assess the skin relief are summarized in the Table 24.1. Depending on the range of skin relief and zone, we analyze in this work the skin relief by laser defocusing system or fringe projection method [4,5]. The defocusing laser microscope uses an optical measuring head, which is a servomotor operating as a transmitter-receiver. A laser diode (transmitter) sends out a beam of light to the surface and the reflected signal is sent back to a set of four photodiodes (receiver). A servomotor drives a lens in order to get a maximal reflected light intensity on the photodiodes. The diameter of the spot on a flat surface is 1 μm; this diameter varies when it is transmitted to a rough surface. The diameter variation of the spot in the course of the measurement triggers an automatic localization control by vertical shifting of the lens. The lens is part of a vertical movement system, which records the localization height
TABLE 24.1 Principle Three-Dimensional Profilers Measurement System Tactile profiler Fringe projection Laser triangulation Laser defocusing microscope Confocal microscopy
Vertical Resolution 0.1 5–10 1 0.3
μm μm μm μm
0.003 μm
Lateral Resolution 2–5 20–50 50 1
Vertical Range of Measurement
μm μm μm μm
6000 μm 1000 μm 10 mm 1000 μm
1 μm
500 μm
that forms the topographical signal. The fringe projection method involves an optical measuring procedure, which uses a combination of gray code and phase-shift techniques to generate three-dimensional information. It is possible to record absolute spatial coordinates of all object points in the image area recorded with a high degree of accuracy in less than 1 sec. The measuring system consists of a projection unit and a CCD camera, which are fixed at the “triangulation angle.” With the gray code method, gratings with a rectangular brightness distribution and differing number of lines are projected consecutively. The number of lines is doubled with each projection run, which unambiguously defines the stripe order for each image point. When using the phase-shift technique, only one grating with sinus-shaped intensity distribution is projected multiple times with varying phase positions.
24.3 FOURIER TRANSFORM OF SKIN LINES NETWORK AND FREQUENCY RANGE Fourier analysis is a powerful tool used to condense information represented in the spatial (or time) into the frequency domain, and to enhance the information to detail individual frequency or wavelength values. It reveals the absolute and relative contributions of different wavelength components to the mean square height of surfaces. In addition, it is more convenient to define and to separate the so-called roughness and waviness components by twodimensional spectral analysis. In Fourier analysis the basis function is the exponential function, although Fourier series are frequently written in alternative form using the trigonometric functions sine and cosine. In this work the skin topography components have been analysed by two-dimensional Fourier transform. The determination of the real and imaginary parts of Fourier transform permits the computation of the amplitude, direction, and wavelength of each spatial frequency [6,7]. The complex Fourier transform computed from scanned data is given by the relation
The Morphological Tree of the Cutaneous Network of Lines
0
N/2
197
N–1
–N/2 –N/2
0
f (x, y) → F(vu, vv)
N/2
High frequencies
0
N/2
High frequencies
0
Low frequencies N/2
N–1
(a)
(b)
FIGURE 24.1 Spatial frequencies repartition.
F ( vu , vv ) =
1 1 N M
N −1 M −1
∑∑ f (x , y ) j
vlu =
k
J =0 k =0
(24.1)
1 1 1 = = = 0.1953 cycles mm –1 N Δ x M Δ y 512 × 0.01 (24.4)
exp − 2 πi ( vu , x j , vv , yk ) equivalent to a wavelength of λl = 1/0.1953 = 5.12 mm which represents the spectrum of surface topography defined in the finite bandwidth by the low and high spatial frequencies limit, respectively:
vlu =
u v cycles mm −1 , vlv = NΔ x MΔ y
vhu =
1 1 cycles mm −1 , vhv = 2Δ y 2Δ x
(
(
)
and
) (24.2)
(Δx, Δy are the sampling steps) the subscripts u and v indicate the wavelength and orientation of a sine wave in the original data, where N and M are the number of data points along x and y axes.
24.3.1 SPATIAL FREQUENCY RANGE NETWORK
OF
LINES
The bandwidth of the frequency range is given by the relation (24.3). For example, if N = M = 512 as the choice for the maximum sample size. If the sample intervals are fixed at Δx, = Δy = 10 μm, this sample size makes the sample dimension 5.12 mm × 5.12 mm. Hence the wavelength of the longest component of skin furrows that can be identified in the sample is approximately 5.12 mm. The frequency range of the analysis can be readily calculated from the expression (3). The high frequency limit
vhu =
1 1 1 = = = 50 cycles mm –1 2 Δ x 2 Δ y 2 × 0.01
24.3.2 FOURIER SPECTRUM REPRESENTATION Two-dimensional data recorded containing N × M points with an origin located at j = k = 0 will give rise to N × M array or spectral coefficients under the two-dimensional Fourier transformation. The spectral coefficients will form the pattern of the skin relief. The origin will be located in a position that corresponds to the origin of the data array. The frequency of these spectral coefficients increases along directions that correspond to the positive directions of the x and y axes of the array up to the Nyquist frequencies (N/2 and M/2). The coefficients are then repeated in inverse order. The representation of the Figure 24.1a has no physical significance. To overcome this, the quadrants of the spectrum can be rearranged as illustrated in Figure 24.1b. This operation moves the origin to the center of the array, and the pattern formed by the rearranged quadrants is that which would be generated by an optical Fourier transform system or diffraction (Fraunhoffer) type analysis [8]. In practice the exchange of quadrants is effected by a mathematical operation rather than by physical rearrangement of the transform coefficients. Gonzalez and Wintz (1977) show that multiplying f (xj, yk) by (–1)j+K is equivalent to shifting the origin to (N/2, M/2) as required. So the low frequencies of skin relief spectrum F(νx, νy), are located in the center of the spectrum, by using the frequential translation property [6,7]: TF[exp(–2πi(vu0x + vv0y)f(x,y)] = F(vu – vu0, vv – vv0) (24.5) for
(24.3) vu 0 = vv0 = equivalent to a wavelength of λh = 1/0.02 = 0.02 mm or 20 μm. Similarly, the low frequency limit νlu is given by
we have
N 2
(24.6)
198
Handbook of Non-Invasive Methods and the Skin, Second Edition
μm
60.4 51.3 42.3 33.2 24.2 15.1 6.0 –3.0 –12.1 mm –21.1 –30.2 –39.2 –48.3 –57.4 –66.4 –75.5 –84.5
3
2
1
0
0
1
2 mm 2D View of skin lines topography
3 Spectrum
FIGURE 24.2 Example of Fourier spectrum skin lines.
N N⎞ x y ⎛ TF ⎡( -1) ( -1) Z ( x,y ) ⎤ = F ⎜ vu − , vv – ⎟ ⎣ ⎦ ⎝ 2 2⎠ (24.7) Figure 24.2 shows the Fourier spectrum of the volar forearm skin lines.
24.3.3 SKIN LINES ORIENTATION SPACE
IN THE
FOURIER
The complex coefficients F(νu, νv) of a two-dimensional spectrum can be used to evaluate the amplitude and power spectra of the topography. The magnitude of these coefficients has the normal interpretation. The frequency domain coordinates, however, have special significance. The subscript (u, v) indicate the wavelength and orientation of a sine wave in the original data. The axes u and v are perpendicular in Fourier space, so the frequency of a particular coefficient in the transform is given by
(
1 2 2 v
v = vu2 + v
)
(24.8)
whiles the orientation of the wave, relative to the u axis, is given by ⎛ u⎞ θ = tan −1 ⎜ ⎟ ⎝ v⎠
24.3.4 SPECTRAL ROSE
OF
(24.9)
SKIN LINES ANISOTROPY
The amplitude of skin relief spectrum is given by: A(νu, νv) = 2|F(νu, νv)*F(νu, νv)|
(24.10)
where F(νu, νv)* is the complex conjugate of F(νu, νv). The amplitude spectrum can be expressed with polar coordinates A(ρ,θ). The various topographical components with respect to the wave direction identified in the polar spectrum can be directionally represented in the polar diagram in the form of a spectral rose [8,9]: RS = (θ), (0 ≤ θ ≤ π)
(24.11)
generated with the computation of the direction of each spatial frequency, identified with respect to the x axis in the original sample [6]. The geometrical form of the spectral rose represents the global anisotropy and permits the identification of the direction of each line’s family. Figure 24.3 shows the spectral roses with respect to the age of the skin morphology.
24.4 LINES IDENTIFICATION BY ANISOTROPIC SPECTRAL FILTERING Traditionally the anisotropy of surface topography has been studied by different anisotropy indexes, defined initially by L. Higgins [11] and adapted to surface roughness by Nayak [12]. The major part of the anisotropy indexes have been defined in the sense of the variation of the surface gradient vector. The method developed in this work quantified the anisotropy of skin relief by computing the direction of each component in the spectrum. The basic idea of this method is to introduce the notion of the anisotropy between form, waviness, and microrelief using the direction as an indicator parameter of each component, which allows exact extraction of the spatial frequencies of each morphological family. These two methods of anisotropic extraction have been developed to extract the low or high frequencies vs. the anisotropy of the relief morphology of skin.
The Morphological Tree of the Cutaneous Network of Lines
μm
5
21 years 4
mm
3
2
33 years 1
0 0
1
2
3
4
μm
5
62 51 40 29 18 7 –4 –15 –26 –37 –48 –59 –70 –81 –92 –103 –114
5
4
3 mm
62 51 39 27 16 4 –7 –19 –30 –42 –53 –65 –77 –88 –100 –111 –123
199
2
1
0
mm
μm
54 years
4
3
Spectral rose
2
85 years
1 Skin Morphology 0
1
2
mm
3
4
μm
5
122 100 78 56 34 12 –10 –32 –54 –76 –98 –120 –141 –163 –185 –207 –229
2
3
4
5
3
4
5
5
4
mm
5
0
1
mm
mm
101 85 68 52 36 19 3 –13 –30 –46 –62 –79 –95 –111 –127 –144 –160
0
3
2
1
0
0
1
2 mm
FIGURE 24.3 Spectral roses with respect to the age of skin lines morphology.
5 4.5 4 3.5 3 2.5 2 1.5 1 0.5
1 1.5 0 0.5 mm 0
2 2.5
3 3.5
4 4.5
4.5 4 3.5 3 2.5
5 4.5 4 3.5 3 2.5 2 1.5
5
1 0.5
1 1.5 0 0.5 mm 0
Original morphology
2 2.5
3 3.5
4 4.5
5
2 1.5
4.5
1 0.5 0 0 mm
0.5
1
1.5
2
2.5
3
3.5
4
The micro relief
Extraction of the wrinkle
FIGURE 24.4 Extraction of the wrinkle.
24.4.1 DIRECTIONAL EXTRACTION
OF THE
WRINKLE
Two fundamental sampling templates are introduced to extract the specific morphology in the frequency domain [6]. The first template is the parallel line template, which involves the computation of the total power observed along a series of superimposed parallel lines across the Fourier spectrum. If we consider the case of the wrinkles frequencies oriented in the x or y axis of the spectrum. The extraction of the frequencies can be realized by the
inverse Fourier transform of the frequencies oriented in x or y direction as following: + vx
Z ( x, y )Wrinkles = TF −1
∑ ⎡⎣ F ( x , v )⎤⎦ x
y
θ⊥ u or // u
v = − vx
(24.12) Figure 24.4 shows the separation of the principal wrinkle and the furrows oriented differently.
200
Handbook of Non-Invasive Methods and the Skin, Second Edition
Following Langer’s investigations of mechanical properties of the integument, recent in vivo studies confirmed and assessed the anisotropy of the skin network. The anisotropy is generally characterized by different mathematical or imaging analysis methods. These approaches consider the anisotropy as a nonhomogenous directional distribution D(θ) of skin lines, represented in an x,y plane without any information about volumetric anisotropy. Unfortunately this method doesn’t represent the reality of the skin network distribution, which can be represented by three anisotropic families:
90°
0°
(a)
FIGURE 24.5a Directional sampling.
24.4.2 DIRECTIONAL EXTRACTION
OF
LINES FAMILY
The second method is the radial sampling template; the wedge sampling filter integrates spectral energy in a number of given angular directions (Figure 24.5a):
Z(x,y) = αF(micro-relief, θ1) + βG(Langer-Lines, θ2) + γ H(wrinkles, θ3) (24.14) α, β, and γ are the amplitudes skin relief, which depend on sex, age, zone, and pathology. θ1, θ2, θ3 are dependent on the depth scale and wavelength of skin morphology.
θ j + Δθ
Z ( x, y )texture ( x, y ) Δθ = TF
−1
∑ ⎡⎣ F ( v , v )⎤⎦ x
y
θj
(24.13)
24.5.1 STATISTICAL ANALYSIS MORPHOLOGY
As shown in Figure 24.5b, this Fourier decomposition is adapted to the separation of the multidirectional volarforearm skin lines.
μm
2 mm
Δθ = 0° → 90° Skin morphology top view
0 0
1
mm
2
3
3
2 μm
60.4 51.3 42.3 33.2 24.2 15.1 6.0 –3.0 –12.1 –21.1 –30.2 –39.2 –48.3 –57.4 –66.4 –75.5 –84.5
3
μm
2
79.4 70.6 61.9 53.1 44.4 35.6 26.8 18.1 9.3 0.6 –8.2 –16.9 –25.7 –34.4 –43.2 –51.9 –60.7
mm
30.2 26.0 21.9 17.8 13.6 9.5 5.3 1.2 –2.9 –7.1 –11.2 –15.4 –19.5 –23.7 –27.8 –31.9 –36.1
θ ≈ 0°
1
1
0
0
1
mm
2
3
θ ≈ 90° 1
Δθ = 90° → 180° 0
0
1
mm
2
μm
3
55.1 46.8 38.5 30.3 22.0 13.7 5.4 –2.9 –11.1 –19.4 –27.7 –36.0 –44.2 –52.5 –60.8 –69.1 –77.4
3
2
1
0 0
1
0
1
mm
2
3
3
2
1
0
2 mm
(b)
FIGURE 24.5b Directional spectral extraction of skin lines.
SKIN LINES
The nature of the skin relief shows that each local motif of lines network can be represented by three parameters:
3
mm
μm
35.3 30.1 24.9 19.8 14.6 9.4 4.3 –0.9 –6.1 –11.2 –16.4 –21.5 –26.7 –31.9 –37.0 –42.2 –47.4
OF
mm
270°
24.5 DETERMINATION OF THE THREEDIMENSIONAL TREE SKIN SPECTRUM
Frequencies of lines familiy
mm
Directional Sampling 180°
3
The Morphological Tree of the Cutaneous Network of Lines
3.0%
6%
2.5%
5%
2.0%
4%
1.5%
3%
1.0%
2%
0.5%
1%
0.0%
0%
201
Morphological Rose
90
110
70 50
130
ρ
0
170 λs
24
116
30
150
720
10
180
0%
λsmean = 186 μm
Zmean = 33.49 μm
1%
2%
3%
0
FIGURE 24.6 Statistical distribution of lines morphology. 45.5
120
90
120
60
90
90
120
60
60
36.7 28.0
150
150
30
150
30
30
19.2 10.5 1.7 –7.0
0% 0.25 0.5 0.75
–1.2 → 14.4 μm
14.4 → 29.9 μm
90
90
120
–15.8 –24.5
0
0% 0.25 0.5 0.75
120
60
150
0% 0.25 0.5 0.75
0
29.9 → 45.5 μm 90
60
150
30
0
120 30
60
150
30
–33.2 –42.0
0% 0.25 0.5 0.75
–50.7
–47.8 → –32.3 μm
–59.5
90 120
0
0% 0.25 0.5 0.75
0
–32.3 → –16.7 μm 120
60
90
0
0% 0.25 0.5 0.75
–16.7 → –1.2 μm 90
120
60
60
–68.2 –77.0
150
150
30
150
30
30
–85.7 μm
–94.5
0% 0.25 0.5 0.75
0
0% 0.25 0.5 0.75
–78.9 → –63.4 μm
–94.5 → –78.9 μm
0
0% 0.25 0.5 0.75
0
–63.4 → –47.8 μm
FIGURE 24.7 Anisotropy of furrows vs. the depth of the relief.
Z(x,y) = f(ρ,λ,θ)
(24.15)
ρ the amplitude, λ the wavelength, and θ the direction. The local elements ρij, λij, and θij are quantified after the spectral sampling of different furrows family of the skin relief. The density of distribution of each parameter ρij, λij, and θij can be quantified as shown in Figure 24.6.
24.5.2 MORPHOLOGICAL TREE OF LINES [13–15]
OF
SKIN NETWORK
The morphological rose of skin lines network is generated by computing the density of lines oriented in a direction θ identified with respect to the x axis in the original sample. The geometrical form of the morphological rose
represents the local and global lines anisotropy and permits to quantify the phase of surface topography. The knowledge of the local heights of each line’s motif ρ permits the representation of the morphological rose vs. the line depth: RM(θ) = f(Δρ)
(24.16)
This original representation of the anisotropy vs. the level of lines family can be used to study the volumetric aspect of anisotropic skin network. Figure 24.7 shows an example of the anisotropy of lines vs. the depth of the relief after the spectral sampling of the lines in different directions.
202
Handbook of Non-Invasive Methods and the Skin, Second Edition
ρ
Sarah tree 3D view
? ?
? ?
μm
? ?
3D view
45.5 28.0 10.5 –7.0 –24.5 –42.0 –59.5 –77.0 –94.5
? ?
50.8 33.9 17.0 0.1 –16.8 –33.8 –50.7 –67.6 –84.5
θij, λij
? ? ?
θ = 0°, ϕ = 15°
90
180
180
0
270
Zij(x, y)
μm
→ (ρij, λij, θij)
270
FIGURE 24.8 Morphological tree of skin lines.
μm
93 75 57 39 21 3 –15 –33 –51 –69 –87 –105 –123 –141 –159 –177 –195
4
25 years 3
110
90
70 50
mm
130
2
150
30
170
1
10
180
0%
1%
2% 3%
0
0 0
1
2 mm
3
4
88 63 37 12 –13 –39 –64 180 –90 –115 μm 270
90 0
84 years 3 110
mm
μm
98 85 71 58 45 31 18 5 –8 –22 –35 –48 –62 –75 –88 –102 –115
130
2
50
10
170 180
0
70
30
150
1
90
0
1
2 mm
0% 1% 2% 3% 4% 5% 6% 7% 8% 9%
3
0
85 56 27 –1 –30 –58 –87 –116 –144 μm
180
90
270
0
(a)
FIGURE 24.9a Assessment of skin aging.
This method has been extended to the totality of the skin lines and represented in a three-dimensional tree. Figure 24.8 represents the tree of a volar forearm network of lines of a young women, from the summit of the plateau to the deepest valley of lines in the form of a threedimensional tree. The z axis of the tree represents the
height of the furrows; the density of lines orientations is represented for each scale of the skin relief family. An illustration of this method applied to the assessment of the aging and cosmetic application is given in Figure 24.9a, b, and c [13–16].
The Morphological Tree of the Cutaneous Network of Lines
203
–60 –40 –20 0 20 40 60 Z μm
100 Before
2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0
–60 –40 –20 0 20 40 60 Z μm
2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 mm
80 60 Altitude μm
40 0 –20 –40 –60 –80 –100 0
90
180
0
1 2 3 Density %
4
0
1 2 3 Density %
4
100
6 months after treatment
80
2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 2.4 2.2 2 1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 mm
20
60 Altitude μm
40 20 0 –20 –40 –60 –80 –100 0
90
180
(b)
FIGURE 24.9b Reorganization of the secondary lines under the effect of vitamin C [16].
24.6 CONCLUSION Since the 1980s, the study of the aging process or the effect of a cosmetic by the analysis of the morphology of the skin lines has used quantifying techniques developed for mechanical engineering. As mentioned above, the morphology of the cutaneous relief is not to be compared to the surface of an iron sheet. The skin physiology seems mainly mechanical, merely following the dimensions of a contracted and elastic medium dermis. The main lines are formed in the superficial dermis where they remain visible after separation from the epidermis, whereas the epidermis loses every trace of them. However, the secondary lines marked on the plateau of the relief can be of epidermic origin. For this reason an original analysis technique was developed that is well adapted to the physiological phenomena. This new approach is similar to a mathematical microscope that can reveal the skin at different magnifications and shows the results with a morphological tree. This novel method displays the hierarchy of the many families of the skin lines, taking into account their depth, width,
and direction. With this new technique, it is now possible to quantify objectively and accurately the morphological changes of the lines during the aging process or after application of a dermocosmetic. The precision of the analysis evidences the influence of a cosmetic on the branches of the tree by distinguishing the smoothing, restructuring, or thinning of the lines and wrinkles. The current technological advances will spread the use of this technique in skin pathologies, probably supplemented usefully by videomicroscopy.
REFERENCES 1. C.E. Pierard, C. Franchioment, C.H.M. Lapierre, “Le vieillissement, son expression au niveau de la microanatomie et des propriétés physiques de la peau.” Int. J. Cosmet. 2:209, 1980. 2. C. Plewing, “Regional differences of cell sizes in the human stratum corinum. Effect of sex and age.” J. Invest. Dermatol., 73:67, 1979. 3. L. Hashimoto, “New methods for surface ultrastructure.” Int. J. Dermatol. 13:357–381, 1974.
204
Handbook of Non-Invasive Methods and the Skin, Second Edition
1 Mean depth = 387 μm
Before
30 28
26 24
22 20
18 16
14 12
10 8 6
4 2 4 2 0 0 mm
6
Altitude mm
0.5
22 20 18 16 14 12 10 8
0 –0.5 –1
180
270
–1.5 0
Mean depth = 182 μm
1
2 3 4 5 Density %
24 22 20 18 16 14 12 10 8
After 28 26 24
22 20
Altitude mm
0.5
18 16 14 12
10 8 6 4 2 0
0 –0.5 –1
6 4 2
180
0
270
–1.5 0 1 2 3 4 5 6 Density %
(c)
FIGURE 24.9c Wrinkle treatment. 4. S. Jaspers, H. Hopermann, U. Hoppe, R. Lunderstadt, J. Ennen, “Rapid in vivo measurement of topography of human skin by active image triangulation using a digital micromirror device.” Skin Res. Technol., 15:195–207, 1999. 5. K.J. Stout, P.J. Sullivan, W.P. Dong, E. Mainsah, N. Luo, T.G. Mathia, H. Zahouani. “The development of methods for the characterisation of roughness in three dimensions.” Commission of the European Communities, Brussels, (ISBN 7044 13132), 1993. 6. H. Zahouani, “Spectral and 3D motifs identification of anisotropic topographical components. Analysis and filtering of anisotropic patterns by morphological rose approach.” Int. J. Mach. Tools Manuf., 38:615–623, 1998. 7. I. Sherrington and E.H. Smith, “Fourier models of the surface topography of engineering components.” Surface Topography, Vol. N° 1, 1988, pp. 11–25. 8. G. George Lendaris, Gordon L. Stanley, “Diffraction Pattern Sampling for Automatic Pattern Recognition,” Proc. of the IEEE, Vol. 58 N° 2, February 1970, pp. 198–216. 9. H. Zahouani, PhD thesis, Besançon, 1989. 10. H. Zahouani, V. Jardret, T.G. Mathia, “Morphological characterisation of rough material,” Surface Modification Technologies, Vol. 3, Edited by T.S. Sudarshan and M. Jeandin, The Institute of Materials, Vol. 3, pp. 135–147, 1995.
11. M.S.L. Higgins, “Statistical properties of an isotropic random surface,” Philosophical Transactions of the Royal Society, Vol. 250, Series A, 1957, pp. 157–174. 12. P.R. Nayak, “Some aspects of surface roughness measurement,” Wear, Vol. 26, 1973, pp. 165–174. 13. H. Zahouani, R. Vargiolu, Ph. Humbert, “3D Morphological tree representation of the skin relief. A new approach of skin imaging characterisation,” International Federation of the Societies of Cosmetic Chemists. Paper N° 30, pp. 69–80. Cannes, France, September 14–18, 1998. 14. H. Zahouani, S-H. Lee, R. Vargiolu, “The Multi-Scale Mathematical Microscopy of Surface Roughness. Incidence in Tribology.” Lubrication at the Frontier. Elsevier Science B.V., pp 379–390, 1999. 15. H. Zahouani, M. Assoul, R. Vargiolu, T. Mathia, “The morphological tree transform of surface motifs. Incidence in tribology.” Int. J. Mach. Tools Manufact., 41: 1961–1979, 2001. 16. Ph. Humbert, M. Haftek, P. Creidi, C. Lapiere, A. Richard, A. Rougier, H. Zahouani. “Topical ascorbic acid on photoaged skin. Clinical, topographical and ultrastructural evaluation: double-blind study vs. placebo.” Exp. Dermatol., 12:237–244, 2003.
of Methodologies for 25 Comparison Evaluation of Skin Surface Contour and Wrinkles: Advantages and Limitations Motoji Takahashi and Motoki Oguri Shiseido Research Center, Yokohama-shi, Japan
CONTENTS 25.1 Introduction............................................................................................................................................................205 25.2 General Description of the Instruments to Measure Skin Surface Morphology .................................................206 25.2.1 Skin Surface Replicas and Image Analysis ..............................................................................................206 25.2.1.1 Skin Surface Contour Measurement1 .........................................................................................206 25.2.1.2 Wrinkle Measurement by the Shadowing Method2...................................................................207 25.2.2 Optical Profilometry Using Skin Surface Replicas and Three-Dimensional Analysis ............................208 25.2.3 In Vivo Measurement (PRIMOS7 and FOITS8) ........................................................................................208 25.2.4 Parameters for Skin Surface Contour in Three-Dimensional Analysis....................................................208 25.2.5 Parameters for Wrinkles in Three-Dimensional Analysis ........................................................................209 25.3 Comparison of Image Analysis and Optical Profilometry....................................................................................209 25.3.1 Skin Surface Contour ................................................................................................................................209 25.3.2 Wrinkles.....................................................................................................................................................209 25.4 Comparison of Replica and In Vivo Measurement ...............................................................................................211 25.5 Advantages and Limitations of Each Method ......................................................................................................211 25.6 Conclusion .............................................................................................................................................................212 References .......................................................................................................................................................................212
25.1 INTRODUCTION In the past, different methods and instruments have been developed to measure skin surface contour or wrinkles. The most common technique is the replica method, where a silicon imprint of the skin is measured with mechanical tactile or optical devices or by image analysis instead of the living skin. Recently, in vivo measurement without taking replicas, which means direct measurement on the skin of the living person, has been developed. Image analysis of a replica has been widely used because of the low cost of instruments and simple manipulation, though it can scarcely measure the exact peak height of skin surface
topography. Optical profilometory using replicas can get precisely three-dimensional data in sufficient speed to compare with tactile data. On the other hand, in vivo measurement has an advantage to escape from experimental artifacts when taking a replica and making possible repeated measurement at the same area on the skin. However, this method has a disadvantage that the results are not a little affected by pulsatory motion or by body movement. This chapter describes a comparison of image analysis of skin surface replica, optical profilometry using a replica, and the in vivo method from the standpoints of accuracy, reproducibility, handling, and so on.
205
206
Handbook of Non-Invasive Methods and the Skin, Second Edition
25.2 GENERAL DESCRIPTION OF THE INSTRUMENTS TO MEASURE SKIN SURFACE MORPHOLOGY 25.2.1 SKIN SURFACE REPLICAS AND IMAGE ANALYSIS 25.2.1.1 Skin Surface Contour Measurement
1
In measurement of skin furrows using negative skin replicas and image analysis, the replicas are illuminated from three directions with 120° between each pair of directions in a light-tight box to reduce light other than that from the controlled light source. In skin surface contour measurement it is necessary to illuminate the replica evenly by light because the pattern of skin furrows running in every direction is very crucial, unlike in wrinkle measurement. The images of the replicas are taken by digital camera. The gray level (levels of brightness) at each pixel is obtained, and the binary images are prepared by classifying each pixel with a gray level above and below a certain level (slice level) as black or white. The binary image is reduced and linearized to extract the characteristics of skin furrows. The thinned image is obtained by substituting a broken string of black dots in the horizontal direction for its central point, and the linearized image is obtained by substituting fine lines in the thinned image for vertical, horizontal, and 45° diagonal lines (Figure 25.1). Various parameters are defined as shown in
TABLE 25.1 Parameters Used for Image Analysis of Skin Surface Replica1 Index KSD
RMAX
VC1
VC2
AVSD16
ALL
KOSU
LEN
NUM WD (a)
(c)
(b)
(d)
FIGURE 25.1 Extraction of characteristics of skin surface pattern from negative replica image. (a) Camera image of skin relief; (b) binary image; (c) thinned image; (d) linearized image.1
Definition Standard deviation of gray level value at each pixel in camera image Ratio between the maximum and minimum values of average run length of black dots in four directions (vertical, horizontal and 45° diagonal), each in binary image Variation coefficient of number of black dots of each 13 × 13 mesh composing binary image Variation coefficient of number of white dots of each of 13 × 13 mesh composing binary image Standard deviation of number of black dots in each of 4 × 4 meshes composing binary image Total number of black dots in binary image Number of meshes with a proportion of black dots above 45% in binary image divided into 13 × 13 meshes Average length of a straight line between the points of intersection in linearized image Number of straight lines in linearized image Ratio of dots between binary and thinned image
Meaning Average of skin roughness Anisotropy of skin furrows
Anisotropy of skin furrows
Anisotropy of skin ridges
Size of hair follicles
Proportion of hair follicles and skin furrows Number of larger hair follicles
Average length of skin furrows
Number of skin furrows Average width of skin furrows
Table 25.1 and calculated for these images to extract their geometric characteristics. Using KSD (average of skin roughness: standard deviation of gray level at each pixel in the image) and VC1 (anisotropy of skin furrows: variation coefficient of number of black dots in each 13 × 13 mesh composing the binary image), age-related changes in the skin surface contour at cheek sites for 295 Japanese females ranging from 3 to 65 years of age were examined.1 The typical images for each generation are shown in Figure 25.2. These parameters can describe the change in skin surface pattern with aging (Figure 25.3 and Figure 25.4); VC1, especially, could detect subtle age-related changes in skin surface pattern. VC1 and KSD can also detect the
Comparison of Methodologies for Evaluation of Skin Surface Contour and Wrinkles: Advantages and Limitations 207
9Y
15Y
25Y
33Y
42Y
55Y
62Y
FIGURE 25.2 Change in skin surface pattern at cheek site with aging.1
8.0
0.6 n.s.
0.5
n.s. ∗∗
∗∗
6.5
∗∗
∗∗
0.3
∗∗: p < 0.01 (compared with 3–9Y) Mean ± S.E.
0.2 ∗∗: p < 0.01 (compared with 3–9Y)
6.0 5.5
0
10
20
30 40 Age (years)
50
60
∗∗
∗∗
0.4
7.0
VC 1
KSD
7.5
n.s.
∗∗
∗∗
Mean ± S.E.
n.s.
0.1 70
0
0
10
20
30 40 Age (years)
50
60
70
FIGURE 25.3 Change in surface roughness at cheek site with aging.1
FIGURE 25.4 Change in anisotropy of skin surface furrows at cheek site with aging.1
change in skin surface contour caused by skin hydration cream applied to a cheek with dry skin. After daily treatment with cream for 4 weeks, KSD increased from 8.83 to 11.0 and VC1 decreased from 0.45 to 0.39, as shown in Figure 25.5. Both parameters are very useful in the efficacy test of cosmetics to examine skin surface contour changes.
perfectly flat. The peak height and area of shadows created by the oblique light impinging on the replica are analyzed to examine the characteristics of wrinkles. The most important operation is the extracting of shadows recognized as wrinkles, because the peak heights of shadows depend on the threshold of gray level to be used. Recently, the calibration method using a standard scale for wrinkle measurement was proposed for obtaining exact peak height.3 However, the shadowing method has a few intrinsic disadvantages: small wrinkles are hidden by adjacent large ones, and the peak height of wrinkles changes depending on sample (replica) tilt. Many parameters are calculated in wrinkle analysis; the following ones are generally used in the shadowing method: the fraction of
25.2.1.2
Wrinkle Measurement by the Shadowing Method2
A negative replica is illuminated by oblique light at a precisely defined angle relative to the plane of observation in the wrinkle measurement. The replica must be kept
208
Handbook of Non-Invasive Methods and the Skin, Second Edition
based on a touch-free fringe projection, and PRIMOS, based on active triangulation in conjunction with the phase-shift technique. Both make it possible to identify and readjust previously recorded skin areas at video rate, and can measure skin surface configuration of the human body in three dimensions. Because of avoiding the need to make replicas, they are free from unavoidable loss of information connected with the replica technique.
25.2.4 PARAMETERS FOR SKIN SURFACE CONTOUR IN THREE-DIMENSIONAL ANALYSIS
Before VC 1: 0.45 KSD: 8.83
There are many two-dimensional parameters to be used in the field of surface metrology, such as Ra (roughness average), Rz (10-point height), Ry (maxium height of the profile), and Sm (mean spacing of profile irregularities).9 In order to describe the skin surface contour, these parameters are extended to surfaces, i.e., three dimensions, and defined as SRa, SRz, SRy, and SSm, respectively. 1 l f ( x ) dx : Arithmetical mean deviation l 0 of profile. f(x) is the height of the assessed profile.
Ra = After VC 1: 0.39 KSD: 11.0 (42, female, cheek)
1 l 2 l1 f ( x, y ) dxdy : Ra extended l 1l 2 0 0 to three dimensions. f(x,y) is the height of the assessed profile.
SRa (1) =
FIGURE 25.5 Change in skin surface microtopography after application of hydration cream for 4 weeks.1
shadows in a defined area (WA), the average peak height of all wrinkles (WH), and the highest peak in wrinkles (HP).
25.2.2 OPTICAL PROFILOMETRY USING SKIN SURFACE REPLICAS AND THREE-DIMENSIONAL ANALYSIS Currently, instead of a mechanical sensor or stylus in direct contact with the surface to be analyzed (mechanical profilometry), a noncontact optical sensor is frequently used (optical profilometry), whose principles are focusing of the laser beam4 (confocal scanning laser microscopic method) and optical triangulation (slit projection method5 or pattern projection method6). These methods have an advantage in that they can quickly acquire data and analyze skin topography in three dimensions.
25.2.3 IN VIVO MEASUREMENT (PRIMOS7 FOITS8)
∫
∫ ∫
Ra 0 + Ra 45 + Ra 90 + Ra 135 :Ra0,Ra45, Ra90, 4 and Ra135 are defined as Ra in 0, 45, 90, and 135° directional lines, respectively (Figure 25.6).
SRa (2 ) =
90° 135° 45°
0°
AND
Nowadays, nontouching systems for a rapid in vivo measurement of skin topography are developed and commercially available. They are FOITS, whose technique is
FIGURE 25.6 Ra, Rz, Ry, and Sm are calculated in four different directions (0, 45, 90, and 135°) for extension to three dimension. Three-dimensional parameters (SRa, SRz, SRy, and SSm) are obtained by averaging the values in four directions, respectively.
Comparison of Methodologies for Evaluation of Skin Surface Contour and Wrinkles: Advantages and Limitations 209
5
Rz =
5
∑
ypi +
i =1
∑y
vi
i =1
: Average of the height of 5 the five highest peaks (ypi) plus the five deepest valleys (yvi) over the evaluation length.
Rz 0 + Rz 45 + Rz 90 + Rz135 :Rz0, Rz45, Rz90, and 4 Rz135 are defined as Rz in 0, 45, 90, and 135° directional lines, respectively. Ry = (Maximum of |ypi |) + (maximum of |yvi |): the maximum peak to lowest valley vertical distance over the evaluation length. SRz =
Ry 0 + Ry 45 + Ry 90 + Ry 135 :Ry 0 , Ry 45 , Ry 90 , 4 and Ry135 are defined as Ry in 0, 45, 90, and 135° directional lines, respectively.
TABLE 25.2 Correlation Coefficient with Two-Dimensional (VC1 and KSD) and Three-Dimensional (SRa, SRz, SRy, and SSm) Parameters VC1 VC1 KSD SRa(1) SRa(2) SRz SRy SSm
KSD
SRa(1) SRa(2)
SRz
SRy
SSm
1 –0.499 –0.215 –0.209 –0.254 –0.318 0.207 –0.499 1 0.651 0.647 0.746 0.773 –0.018 –0.215 0.651 1 0.964 0.892 0.904 0.324 –0.209 0.647 0.964 1 0.929 0.936 0.275 –0.254 0.746 0.892 0.929 1 0.961 0.240 –0.318 0.773 0.904 0.936 0.961 1 0.089 0.207 –0.018 0.324 0.275 0.240 0.089 1
Note: N = 58.
SRy =
1 Sm = N
N
∑ Si : The mean spacing between peaks. i =1
A peak must cross above the mean line and then back below it. Si is the width of each peak. Sm 0 + Sm 45 + Sm 90 + Sm 135 : Sm0,Sm45, 4 Sm90, and Sm135 are defined as Sm in 0, 45, 90, and 135° directional lines, respectively.
SSm =
25.2.5 PARAMETERS FOR WRINKLES DIMENSIONAL ANALYSIS
IN
THREE-
The total volume of all wrinkles in a defined area (WV) and the profile length ratio (Lr)5 are used for quantification of wrinkles other than WA, WH, and HP, described in Section 25.2.1.2. Usually the aim of this kind of study is to test the efficacy of antiaging products, and the investigation is therefore based on the comparison of two types of results: the evolution of a wrinkle before treatment (the reference) and the evolution of another wrinkle after treatment. To be reliable, this comparison study must be conducted on the same skin surface (same size and same position).
25.3 COMPARISON OF IMAGE ANALYSIS AND OPTICAL PROFILOMETRY 25.3.1 SKIN SURFACE CONTOUR Correlation coefficients with the parameters used in image analysis (KSD and VC1) and those in optical profilometry (SRa, SRz, SRy, and SSm) are shown in Table 25.2.
Because three-dimensional parameters obtained by profilometry are originally defined as those showing surface roughness of the industrial materials, KSD, which is the skin surface roughness parameter, correlates well with SRa, SRz, and SRy. On the other hand, VC1, which shows anisotropy of skin furrows, does not correlate with these three-dimensional parameters. From this result KSD obtained by image analysis seems to be a good substitutive parameter for SRa, SRz, and SRy in three-dimensional analysis. Considering that VC1 can detect subtle agerelated changes in anisotropy of skin furrows and has no relationship with these three-dimensional parameters generally used in industry, VC1 is a unique parameter to describe skin surface contour.
25.3.2 WRINKLES The relationship between image analysis and optical profilometry in WA and WH is shown in Figure 25.7. There is a good correspondence between them in WA, but not in WH, though the correlation coefficient is a little higher. As WH is an average peak height of all wrinkles, the discrepancy between image analysis and profilometry depends on the difference of methodology. Typical cases of discrepancy between them in peak height of wrinkles are shown in Figure 25.8 and Figure 25.9. Figure 25.8 shows an example that a peak height obtained by image analysis is lower than that obtained by optical profilomtry, as the average height of roughness profile following the analyzed wrinkle is higher than the reference line. On the other hand, Figure 25.9 is an opposite case, where the peak height is higher in image analysis than in profilometry. It is considered that two-dimensional parameters like WA (the fraction of wrinkles in a defined area) have a good correspondence of image analysis to profilometry, but not three-dimensional parameters like WH.
210
Handbook of Non-Invasive Methods and the Skin, Second Edition
WH (mm)
30
0.3
Image analysis
Image analysis
WA (%)
20
0.2
0.1
10
r = 0.731, n = 31
r = 0.639, n = 31 0
20 10 Optical profilometory (a)
0
30
(%)
0
0
0.1 0.2 Optical profilometory (b)
0.3 (mm)
FIGURE 25.7 Relationship between image analysis by the shadowing method and optical profilimetry in wrinkle parameters WA (fraction of wrinkle area) and WH (average peak height of wrinkles).
10
10
600
300
500 200
8
100
100 0
7 0
2
4
6
8
0 −100
10 6
−100 −200
μm
200
mm 5
−300
Y-direction (mm)
μm
300
9
9
8 0
2
4
6
8
10
7
mm 6
−200 −300
5
4 4 3
Y-direction (mm)
400
25° 3
25° 2
2 1 1 0 500 0 −500 μm
FIGURE 25.8 Example showing that peak height of wrinkles obtained by image analysis is lower than that by profilometry.
0 500 0 −500 μm
FIGURE 25.9 Example showing that peak height of wrinkles obtained by image analysis is higher than that by profilometry.
Comparison of Methodologies for Evaluation of Skin Surface Contour and Wrinkles: Advantages and Limitations 211
Optical profilometry by confocal scanning laser microscope (HD 100D, Laser tech. Corp., Japan )
In vivo measurement by PRIMOS (reversed)
FIGURE 25.10 Comparison of in vivo measurement and optical profilometer with replica from the same skin area on the ventral side of the forearm. The result of the in vivo measurement is reversed compared with that by the profilometer. The replica method with optical profilometer shows finer and sharper skin surface pattern. The measuring area is 10 × 10 mm. OpenGL10 representation mode.
25.4 COMPARISON OF REPLICA AND IN VIVO MEASUREMENT According to a previous paper,7 the difference between in vivo and replica measurement is noticeable. For example, the two-dimensional roughness parameter Rq (root mean square) of the in vivo results measured by PRIMOS was 10 to 30% higher than the corresponding value measured by the replica obtained from the same part of the human body.7 Authors also reported that the measured texture of replica at the back of the hand was blurred, whereas the in vivo result showed finer and sharper structures. From these results they concluded that in vivo measurement is superior to the replica method. We also measured the skin surface contour of the ventral forearm and crow’s-feet at exactly the same area using replica and in vivo (PRIMOS) methods. Results are shown in Figure 25.10 with OpenGL
representation mode and somewhat different from the previous result.7 In our result profile, the in vivo measurement has more noises than that of the replica method, especially in forearm measurement, where the texture is finer and shallower than eye wrinkles. It is considered that pulsatory motion or body movement still affects the result of in vivo measurement.
25.5 ADVANTAGES AND LIMITATIONS OF EACH METHOD Advantages and disadvantages of image analysis of skin replicas, three-dimensional analysis of replicas with optical profilometry, and in vivo measurement are summarized in Table 25.3. In the replica technique the impression materials must penetrate into all ends of furrows and
TABLE 25.3 Comparison of Methodologies for Evaluation of Skin Surface Configuration Method
Instruments
Advantages
Disadvantages
Image analysis of replica
Combination of light source and digital camera
Optical profilometry
Optical focus profilometer Confocal scanning laser microscope Triangulation optical profilometer FOITS PRIMOS
Low cost of instrument Simple manipulation Spatially analyze orientation of skin texture High accuracy Fast analysis Reproducible Not necessary to take replica Fast analysis Repeated measurement at the same area
Artifact when making replica Not obtain exact peak height Lose lower peaks behind higher peak Artifact when making replica
In vivo measurement
Affected by body movement
212
Handbook of Non-Invasive Methods and the Skin, Second Edition
wrinkles to reproduce all details of skin topography and a short hardening time is demanded. Therefore, the skill of taking a replica influences experimental results. On the other hand, in in vivo measurement making a replica is not necessary, but data acquisition must be completed in a very short time to avoid the affect of self-movement of the living human body.
25.6 CONCLUSION The methodology for measurement of skin surface topography has been developed ranging from image analysis of replicas to the in vivo technique. Each method has advantages and limitations. Among them, in vivo measurement is quite promising, because it avoids the skin hydration effect induced by replica impression materials and the morphological changes in skin surface induced by hydration. However, this technique is slightly influenced by pulsatory motion or self-movement of the human body. Further improvement of the instrument is needed. Technical development for collecting data within 1/10 of a second would be a pressing matter to precisely analyze skin surface contour in direct measurement without taking a replica.
REFERENCES 1. Takahashi, M., Image analysis of skin surface contour, Acta Derm. Venereol. (Stockh.), Suppl.185, 9–14, 1994.
2. Hayashi, S., Matsuki, T., Matsue, K., Arai, S., Fukuda, Y., and Yoneya, T., Changes of facial wrinkles by aging, sunlight exposure and application of cosmetics, J. Soc. Cosmet. Chem. Japan, 27, 355, 1993. 3. Tamura, K., Okano, Y., Okamura, H., and Masaki, H., Quantitative method of wrinkle depth using standard scale, J. Soc. Cosmet. Chem. Japan, 35, 50, 2001. 4. Mignot, J., Three-dimensional evaluation of skin surface: micro- and macrorelief, in Handbook of Noninvasive Methods and the Skin, Serup, J. and Jemec, G.B.E., Eds., CRC Press, Boca Raton, FL, 1995, p. 107. 5. Takasu, E., Umeya, Y., and Horii, I., Development of wrinkle analyzing system using three-dimensional curved shape measurement and the changes of Japanese women’s facial wrinkles with aging, J. Soc. Cosmet. Chem. Japan, 29, 394, 1996. 6. Jaspers, S., Hoperman, H., Sauermann, G., Hoppe, U., Lunderstadt, R., and Ennen, J., Rapid in vivo measurement of the topography of human skin by active image triangulation using a digital micromirror device, Skin Res. Technol., 5, 195, 1999. 7. Hof, C. and Hopermann, H., Comparison of Replicaand In Vivo-Measurement of the Microtopography of Human Skin, research report of the Institute of Automation of the University of the Federal Armed Forces, Hamburg, Germany, 2000. 8. Rohr, M., Brandt, M., and Schrader, A., Skin surface: claim support by FOITS, SOFW J., 126, 2, 2000. 9. ISO 4287:1997, Surface texture: Profile method (Surface Metrology Guide, http://www.predev.com/smg/ parameters.htm). 10. OpenGL — The Industry’s Foundation for High Performance Graphics. http://www.opengl.org/.
Skin Surface Friction
Studies on Skin: 26 Tribological Measurement of the Coefficient of Friction* Raja K. Sivamani, Gabriel Wu, and Howard I. Maibach Department of Dermatology, School of Medicine, University of California, San Francisco, California
Norm V. Gitis Center for Tribology, Inc., Campbell, California
CONTENTS 26.1 Introduction............................................................................................................................................................215 26.1.1 Experimental Designs................................................................................................................................216 26.1.2 Hydration ...................................................................................................................................................217 26.1.3 Lubricants/Emollients/Moisturizers...........................................................................................................217 26.1.4 Probes.........................................................................................................................................................217 26.1.5 Normal Load..............................................................................................................................................217 26.2 Skin Friction Coefficient Values............................................................................................................................218 26.2.1 Hydration ...................................................................................................................................................218 26.2.2 Lubricants/Emollients/Moisturizers...........................................................................................................220 26.2.2.1 Talcum Powder...........................................................................................................................220 26.2.2.2 Lubricant Oils.............................................................................................................................220 26.2.2.3 Emollients and Moisturizers ......................................................................................................220 26.2.3 Probes.........................................................................................................................................................221 26.2.4 Anatomic Region, Age, Gender, and Race ...............................................................................................222 26.3 Conclusion .............................................................................................................................................................222 References .......................................................................................................................................................................223
26.1 INTRODUCTION Mechanically, friction allows us to keep from slipping as we step out of the shower, hold Styrofoam cups of coffee in the morning, or turn the steering wheel in our cars. Because the skin is a surface itself, it is convenient to analyze and describe it in terms of a surface phenomenon like friction; friction studies on skin provide valuable insight into how the skin interacts with other surfaces. It also provides information about the skin under various conditions (e.g., age and gender) and under various chemical treatments (e.g., lotions and moisturizers). Studying the friction of skin supplements other mechanical tests. Friction studies can be conducted with
noninvasive methods and give a measure of the skin’s health — skin hydration, for example: Naylor1 showed that moistened skin has an elevated friction response, and El-Shimi2 demonstrated that drier skin has a lowered friction response. Friction provides a quantitative measurement to assess skin. The friction parameter generally measured is the coefficient of friction. To measure the friction coefficient, a surface is brought into contact with another and moved relative to it. When the two surfaces are brought into contact, the perpendicular force is defined as the normal force (N). The friction force (F) is the force that opposes relative movement between the two surfaces. From Amonton’s law, the coefficient of friction (μ) is
* Modified with permission from Skin Research and Technology, 9, 227–234, 2003.
215
216
Handbook of Non-Invasive Methods and the Skin, Second Edition
defined as the ratio of the friction force to the normal force: μ = F/N The friction coefficient can be measured in two ways: the static friction coefficient (μs) and the dynamic or kinetic friction coefficient (μk). The static friction coefficient is defined as the ratio of the force required to initiate relative movement and the normal force between the surfaces; the dynamic or kinetic friction coefficient is defined as the ratio of the friction force to the normal force when the two surfaces are moving relative to each other. For simplicity, much of the research has focused on the dynamic friction coefficients wherein the two surfaces move at a relative constant velocity. Most of the friction studies on skin have dealt with the dynamic friction coefficient, and the subscript k is usually dropped. This overview references the dynamic coefficient of friction unless otherwise noted. According to Amonton’s law, the dynamic friction coefficient remains unchanged regardless of the probe velocity or applied normal load in making the measurement. Amonton’s law holds true in the case of solids with limited elastic properties. Although Naylor1 concluded Amonton’s law was true, later studies by El-Shimi,2 Comaish and Bottoms,3 and Koudine et al.4 have found that skin deviates from Amonton’s law, since the friction coefficient increased when the normal load was decreased. El-Shimi2 and Comaish and Bottoms3 reasoned that the rise in friction coefficient resulted from the viscoelastic nature of the skin, allowing for a nonlinear deformation of the skin with increasing load.
26.1.1 EXPERIMENTAL DESIGNS Various experimental designs have been devised to measure the friction on skin. They focus on measuring friction by pressing a probe onto the skin with a known normal force, and then detecting the skin’s frictional resistance to movement of the probe. The designs fall into two categories: 1. A probe moved across the skin in a linear fashion 2. A rotating probe in contact with the skin surface In the linear designs, the probe movement is accomplished in several ways. Comaish and Bottoms3 utilized one of the simplest linear designs: they moved the probe across the skin by attaching it to a pan of weights by means of a pulley. Weights are placed in the pan such that the probe slides over the skin at a constant velocity. This allows for the calculation of the dynamic friction coefficient by dividing the total weight in the pan by the normal load on the probe.
FIGURE 26.1 UMT test setup for volar forearm. The forearm is strapped into place with a holder to immobilize the forearm. The copper probe is then brought down to carry out the friction and electrical measurement. (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
More sophisticated linear designs followed the simple design used by Comaish and Bottoms,3 providing motorized unidirectional movement of the probe or the use of a reciprocating motor to move the probe back and forth. In both designs the motorization affords greater control in maintaining the velocity of the probe. Strain gauges measure the friction force as the probe moves along the skin surface. Figure 26.1 shows a biomedical tribometer friction measurement device where the normal load and the probe speed are computer controlled. The second design category measures friction with a rotating wheel pressed onto the surface of the skin with a known normal force. Highley et al.5 measured the frictional resistance by determining the angular recoil of the instrument as the wheel contacted the skin. They measured this angular recoil by recording the proportion of light that hit a dual-element photocell. An electrical signal was then generated in proportion to the frictional resistance. Comaish et al.6 developed a portable, handheld device (Newcastle friction meter) that relied on a torsion spring to measure the skin’s frictional resistance. The devices are surveyed in Table 26.1. An important part of designing a friction measurement apparatus is choosing the probe size, shape, and material.
Tribological Studies on Skin: Measurement of the Coefficient of Friction
217
TABLE 26.1 μ) of Untreated Normal Probe and Apparatus Used to Measure the Dynamic Friction Coefficient (μ Skin In Vivo Author Naylor1 El-Shimi2 Comaish and Bottoms3 Koudine et al.4 Highley et al.5 Prall7 Cua et al.8 Johnson et al.9 Asserin et al.10 Elsner et al.11 Sivamani et al.17 Sivamani et al.24
Probe Size and Shape
Probe Material
Motion of Test Apparatus
Maintenance of Normal Load
8-mm-diameter sphere 12-mm-diameter hemisphere 15-mm-diameter annular ring Hemisphere lens
Polyethylene Stainless steel (rough), stainless steel (smooth) Teflon, nylon, polyethylene, wool Glass
Linear, reciprocating Rotational
Static weights Static weights
Linear
Static weights
Linear
Disc Disc 15-mm-diameter disc 8-mm (radius of curvature) lens 3-mm-diameter sphere 15-mm-diameter disc 10-mm-diameter sphere
Nylon Glass Teflon Glass
Rotational Rotational Rotational Linear, reciprocating
Static weights; balance beam Spring load Spring load Spring load Static weights
Ruby Teflon Stainless steel
Linear Rotational Linear
Copper
Linear
13-mm-diameter cylinder
Because friction is an interaction between two surfaces, the probe geometry and material will affect the values calculated for the friction coefficient of the other surface. Several shapes and material have been used, as outlined in Table 26.1. Also, results will be more accurate when the probe’s normal force is maintained at a constant value or continuously monitored; previous methods used to maintain the normal force include spring mechanisms or static weights to weigh down the probe (Table 26.1). These parameters are revisited critically later in this article. Much effort has been spent in understanding how skin friction changes with differing biological conditions and upon the application of various products to the skin surface. These studies have been of interest to various industries that manufacture products meant as skin topical agents because friction measurements can provide clues regarding the effectiveness of their products.
26.1.2 HYDRATION Hydration is a complex phenomenon influenced by intrinsic (i.e., age, anatomical site) and extrinsic (i.e., ambient humidity, chemical exposure) factors. These factors can affect the mechanical properties of the skin, and research has been performed to correlate hydration levels with the skin’s friction coefficient.24 Hydration studies have investigated how increases and decreases in skin hydration correlated with the friction coefficient. In past studies, researchers generally induced increases in skin hydration through water exposure. However, decreases in skin
Balloon; static weights Spring load Computer servofeedback control Computer servofeedback control
hydration were not experimentally induced and dehydration studies were performed between subjects with normal skin and subjects that had clinically dry skin.2,12
26.1.3 LUBRICANTS/EMOLLIENTS/MOISTURIZERS Much of the reviewed research has been devoted to ascertaining how the application of certain ingredients influences the skin surface, of interest to the cosmetic/moisturizer and lubricant industries. The studies focused on the effects of talcum powder,2,3 oils,2,3,5,14 and skin creams/moisturizers.7,14,17,24 Hills et al.15 analyzed how changes in the friction coefficient, following emollient application, differed with temperature.
26.1.4 PROBES As mentioned earlier, the probe geometry and material influence the measured value of the friction coefficient because friction is a probe–skin interaction phenomenon. Few studies have examined probe effects: El-Shimi2 studied probe roughness and Comaish and Bottoms3 probe roughness and material.
26.1.5 NORMAL LOAD Friction measurements can offer quantitative insight into changes on the skin surface, and the UMT offers technical advances over existing friction measurements. The control of the probe speed and the real-time monitoring of the normal load allow for real-time calculation of the friction
218
Handbook of Non-Invasive Methods and the Skin, Second Edition
Friction coefficient vs. normal load (50 g sensor)
TABLE 26.2 Reported Values of the Dynamic Friction μ) for Untreated Normal Skin Coefficient (μ In Vivo
1.2
0.8 μ
Author 0.4
Naylor1 El-Shimi2
0.0 0
5
10
15
20 25 30 Load (grams )
35
40
45
50
FIGURE 26.2 Friction coefficient vs. normal load. The friction coefficient increased as the normal load was decreased, suggesting that the skin does not follow Amonton’s law. The probe was moved at 5 mm min–1 (n = 4). (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
coefficient. As seen in Figure 26.2, the control of the load is important because the friction coefficient does not adhere to Amonton’s law. Wolfram18 theoretically deduced that the friction coefficient would relate to the normal load as follows: μ ∝ N–1/3 where N is the applied normal load to the skin. Sivamani et al.17 found that the friction coefficient related to the normal load as follows: μ ∝ N–0.32 and Koudine et al.4 found the dependence on the applied normal load to be μ ∝ N–0.28
26.2 SKIN FRICTION COEFFICIENT VALUES Friction is an important characteristic of skin because it allows us to execute many of our daily activities. In addition, friction studies offer insight into how skin and the skin surface change across age, gender, race, anatomical site, and chemical applications. This can provide better information about expected skin variations in the population and why certain topical applications are effective. Comparative studies are particularly useful in following how the skin mechanically changes under different conditions. Previous studies have reported various values for the skin’s friction coefficient. Dynamic friction coefficient measurements (Table 26.2) fall in the range 0.12 to 0.7; however, most fall in a narrower range of 0.2 to 0.5 (Figure 26.3). Besides natural variations in skin, the wide range
Comaish and Bottoms3
Koudine et al.4 Highley et al.5 Prall7 Cua et al.8
Johnson et al.9 Asserin et al.10 Elsner et al.11
m 0.5–0.6 0.2–0.4 (stainless steel, rough) 0.3–0.6 (stainless steel, smooth) 0.2 (Teflon) 0.45 (nylon) 0.3 (Polyethylene) 0.4 (wool) 0.24 (dorsal forearm) 0.64 (volar forearm) 0.2–0.3 0.4 0.34 (forehead) 0.26 (volar forearm) 0.21 (palm) 0.12 (abdomen) 0.25 (upper back) 0.3–0.4 0.7 0.48 (forearm) 0.66 (vulva)
in results may be due to differences in probe movement, geometry, controlled monitoring of the normal force, and material chosen to make the friction measurement. In designing the friction measurement apparatus, the two types of probe movement utilized were rotational and linear (Table 26.1). As a result, the linear probe constantly moves over untested skin and the rotational probe spins over tested skin. This can lead to discrepancies in reported values for the skin friction coefficient. Another important source of variation may be the ability to control the normal force while the probe is testing the skin surface. The skin friction instruments are designed to measure the frictional resistance of the skin, and it is assumed that the normal force is constant. During a test the normal force may not stay constant as a result of an uneven skin surface, inaccurate spring, or a nonuniform distribution of static weights placed above the probe head. Therefore, the assumption of a constant normal force may be incorrect and can lead to variation in the calculated friction coefficient. A third source for variation is the choice of the probe material. Because friction is a surface phenomenon between two materials, the choice of the probe influences the numerical value obtained for the friction coefficient.
26.2.1 HYDRATION Hydration studies revealed that drier skin had lowered friction while hydrated skin had an increased amount of
Tribological Studies on Skin: Measurement of the Coefficient of Friction
219
Ranges in the dynamic coefficient of friction measurements Naylor (1) El-shimi (2) Comaish bottoms (3) Koudine et al (4) Highley er al (5) Prall (7) Cua et al (8) Johnson et al (9) Asserin et al (10) Elsner et al (11) Sivamani et al (17) 0
0 .1
0 .2
0 .3 0 .4 0 .5 μ (dynamic coefficient of friction)
0 .6
0 .7
0 .8
FIGURE 26.3 Outline of the ranges in the dynamic coefficient of friction. These ranges reflect measurement of untreated normal skin friction in vivo. (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 227–234, 2003.)
friction (Table 26.3). However, the skin response is more complex, because very wet skin also has a lowered friction coefficient, much like the characteristics of dry skin.16 Most studies focus on an intermediate zone of hydration where the skin has been moistened without an appreciable slippery layer of water on the skin. Results in Table 26.3 show that the increases in friction were varied, and this possibly results from the various probes used. Although the addition of water increases the friction coefficient, this effect lasts for a period of minutes before the skin returns to its normal state.2,5,14,17 The water has an effect of softening the skin, and this in turn allows for a greater contact area between the probe and the skin. Also, water results
in adhesive forces between the water and the probe. Thus, there is more frictional resistance between the skin and the probe, resulting in a higher friction coefficient.18 Since the water evaporates in minutes, the skin returns to its normal state in the same time frame. For dry skin, the skin becomes less supple and the probe does not achieve as much contact area; this allows the probe to glide more easily over the skin surface. This results in a lowered friction coefficient, as seen in the isopropyl study17 and in prior studies involving subjects with clinically dry skin.2,12 The agreement between the experimentally induced dry skin and clinical dry skin is expected.18
TABLE 26.3 μ) with Increasing Comparative Studies of the Changes in Dynamic Friction Coefficient (μ Hydration (Hydration) and Decreasing Hydration (Dryness) Author Naylor1 El-Shimi2 Comaish and Bottoms3 Highley et al.5 Prall7 Johnson et al.9 Lodén et al.12 Nacht et al.14 Sivamani et al.17 a
Probe Material Polyethylene Stainless steel (rough), stainless steel (smooth) Wool, Teflon Nylon Glass Glass Stainless steel Teflon Stainless steel
% Increase Due to Hydration μmoist – μnormal)/μ μnormal} × 100 {(μ
% Decrease Due to Dryness μnormal – μdry)/μ μnormal} × 100 {(μ
80 100–200 (stainless steel rough)
— 28 (stainless steel, rough), 41 (stainless steel, smooth) — — —
40a (wool), 400a (Teflon) 500 200 100–233 — 45 55 (in vitro)
33 (hand), 41 (back), 14 (arm) — 10 (in vivo)
Comaish and Bottoms studied the change in the static friction coefficient in their hydration study.
220
Handbook of Non-Invasive Methods and the Skin, Second Edition
% Change in friction ceofficient
Pertrolatum Heavy mineral oil
Glycerin Baseline
50 40 30 20 10 0 −10 −20 −30
∗Time
−1
0
1
2 3 Time after application (hrs)
4
5
6
= −1 is immediately prior to application; Time = 0 is immediately after application
FIGURE 26.4 Effect of lubricant cosmetic ingredient on skin friction coefficient. Amount applied of each material: approximately 2 mg/cm2 (mean of five subjects, but p value was not published). Time = –1 is immediately prior to application; time = 0 is immediately after application. (Adapted from Nacht, S. et al., J. Soc. Cosmet. Chem., 32, 55–65, 1981.)
The studies on lubricants, emollients, and moisturizers are important for cosmetics and products developed to make the skin look and feel healthier. The literature reports that the important qualitative characteristics in skin topical agents are skin smoothness, greasiness, and moisturization.17,19 Previous research has tried to describe these subjective, qualitative descriptions in a quantitative fashion by correlating them against the friction coefficient. Prall7 tried to find a quantitative correlation for skin smoothness but was unable to make a direct correlation to the friction coefficient until he added skin topography and hardness into the analysis. Nacht et al.14 found a linear correlation between perceived greasiness and the friction coefficient (Figure 26.4). 26.2.2.1 Talcum Powder El-Shimi2 and Comaish and Bottoms3 showed that the friction coefficient decreased after the application of powder. El-Shimi2 found the friction coefficient to decrease by 50% after application; Comaish and Bottoms,3 in analyzing the static friction coefficient, observed an insignificant change for a wool probe and a 30% decrease in friction with a polyethylene probe. However, they also found that wetting the talc powder caused an increase in the measured friction. 26.2.2.2 Lubricant Oils A lowering in the friction coefficient is the initial effect after the application of oils and oil-based lubricants.2,5,14 Nacht et al.14 and Highley et al.5 also showed that after the initial decrease in friction, the oils eventually raised the
% Change in friction coefficient
26.2.2 LUBRICANTS/EMOLLIENTS/MOISTURIZERS
120 100 80 60 40 20 0 −20 −40
0 (Not greasy)
A B
C
D
1
2
E
F 3
4
5 (Very greasy)
Mean score of greasiness
FIGURE 26.5 Correlation between changes in the friction coefficient and the sensory perception of greasiness. A to F represent different creams that were applied to the skin. The reported percent change in the friction coefficient is immediately after application, and the greasiness scores were subjective evaluations. (Adapted from Nacht, S. et al., J. Soc. Cosmet. Chem., 32, 55–65, 1981.)
skin’s friction coefficient. The results of the lubricant cosmetic studies by Nacht et al.14 are shown in Figure 26.5. 26.2.2.3 Emollients and Moisturizers Prall7 and Nacht et al.14 found that the friction coefficient rises with the addition of emollients and creams in a similar fashion to water. However, the effects of the creams lasted for hours, while the water effects lasted for about 5 to 20 minutes.7,14,17 Sivamani et al.17 quantified the friction, greasiness, and stickiness of the skin following application of creams and treatments (Figure 26.6 to Figure 26.8). Hills et al.15 also studied emollients, but they examined how different emollients compared against one another and how changes in temperature changed the friction coefficient. At a higher temperature (45˚C), most emollients lowered the friction coefficient to a greater degree than at a lower temperature (18˚C).
Tribological Studies on Skin: Measurement of the Coefficient of Friction
2.5
Coefficient of friction
Mean 1.5
Amplitude
1
Amplitude/Mean
0.25
2
0
10
20
30 Time (sec)
40
50
60
FIGURE 26.6 Calculation of the amplitude/mean measurement. The mean refers to the mean value of the measured friction coefficient as indicated on the graph. The amplitude refers to the deviation seen during the friction coefficient measurement as indicated on the graph. Then the amplitude is divided by the mean to calculate the amplitude/mean. It has been suggested that this value represents the smoothness of the skin surface.4,17 (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
0.1 0.05
•
Amplitude/Mean
0.1
0.05
• Distal left forearm
Distal right Proximal right forearm forearm
All
FIGURE 26.7 Amplitude/mean on the untreated volar forearm. No significant differences were found for different anatomical sites between the left and right volar forearms or between distal and proximal sites on the same volar forearm. (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
When lubricant/moisturizers are applied to the skin, the skin friction is affected in three general ways:14,18 A large, immediate increase in the friction coefficient, similar to water application, that follows with a slow decrease in the friction coefficient. These agents can be interpreted to act by immediate hydration of the skin through some aqueous means to give the immediate increase in friction. In Figure 26.4, creams A, B, and C represent this type of lubricant/moisturizer.
∗
Petrolatum
Glycerin Occlusion Intervention
Untreated
FIGURE 26.8 Amplitude/mean measurements for interventions. The application of glycerin and the PVDC occlusion increased the amplitude/mean of the volar forearm. Also, the addition of glycerin raised the amplitude/mean significantly more than the PVDC occlusion. Petrolatum significantly decreased the amplitude/mean, and this is quantitative evidence of petrolatum’s greasiness (p < 0.001). (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
0.15
0
∗
0.15
0.5
0
∗
0.2
0
•
221
An initial decrease in the friction coefficient that is followed by an overall increase in the friction coefficient over time. These agents are fairly greasy products (Figure 26.5), and this greasiness causes the immediate decrease in the friction coefficient. The eventual rise in the friction coefficient is probably due to the occlusive effects of these agents.21 In other words, these products and ingredients act to prevent water loss from the skin, thereby increasing the hydration of the skin. Representations of a few ingredients that elicit this response are in Figure 26.5 and are represented as cream F in Figure 26.4. A small, immediate increase in the friction coefficient that then increases slowly with time. These agents are interpreted to act as a combination of effects seen in the previous two cases. These lubricants/moisturizers have ingredients and agents that serve to both hydrate the skin through some aqueous method and prevent water loss through some occlusive mechanism. Because of the presence of these occlusive agents, which tend to be more slippery, the immediate rise in the friction coefficient is lower than in products that fall into the first category listed above. In Figure 26.4, this is seen in creams D and E.
26.2.3 PROBES El-Shimi2 and Comaish and Bottoms3 compared probes (Table 26.2 and Table 26.3) and found that smoother probes gave higher friction coefficient measurements. ElShimi2 noted that higher friction coefficient measurements were made with a smoother stainless-steel probe than with
222
Handbook of Non-Invasive Methods and the Skin, Second Edition
a roughened stainless-steel probe. Comaish and Bottoms3 found a similar result with two types of nylon probes: a sheet probe and a knitted probe. The sheet probe (the smoother of the two) gave a higher friction coefficient measurement. El-Shimi2 postulates that the smoother probe forms more contact points with the skin and has a greater skin contact area than the rougher probe, resulting in more resistance from the skin and a larger measurement for the friction coefficient.
26.2.4 ANATOMIC REGION, AGE, GENDER, RACE
AND
Few studies address the effects of anatomic region, age, gender, or race as they pertain to the friction coefficient. To date, no significant differences have been found with regard to gender8,22,24 or race.23,24 The friction coefficient varies with anatomical site: Cua et al.8,22 found that friction coefficients varied from 0.12 on the abdomen to 0.34 on the forehead. Elsner et al.11 measured the vulvar friction coefficient at 0.66, whereas the forearm friction coefficient was 0.48. Sivamani et al.24 found that the proximal volar forearm had a higher friction coefficient than the distal volar forearm (Figure 26.9). Manuskiatti et al.23 studied skin roughness and found significant differences in skin roughness at various anatomical sites. Differences in environmental influences (e.g., sun exposure) and hydration may account for this. Elsner et al.11 showed that the more-hydrated vulvar skin had a 35% higher friction coefficient than the forearm, and this is in agreement with hydration studies that contend that skin has an increased friction coefficient under increased hydration. Age-related studies have been contradictory, where some authors found no difference8,22,24 (Figure 26.10) and others found differences.10,11 Cua et al.22 showed no differences in friction with respect to age except for friction measurements on the ankle. Elsner et al.11 also performed age-related tests and found no differences in the vulvar friction coefficient, but observed a higher forearm friction coefficient in younger subjects. They postulate that the skin on parts of the body that become exposed to sunlight can undergo photoaging, and thus forearm skin shows evidence of age-related differences while the light-protected vulvar skin does not.11 Asserin et al.10 concluded that younger subjects had a higher forearm friction coefficient than older subjects. There are few gender-related and racial friction studies. Cua et al.8,22 and Sivamani et al.24 found no significant friction differences between the genders. Manuskiatti et al.23 found no significant racial (black and white skin) differences in skin roughness and scaliness. Sivamani et al.24 found no differences in volar forearm friction among different ethnicities before and after chemical treatments (Figure 26.11 and Figure 26.12).
1 0.8 0.6 0.4 0.2 0
300 250 200 150 100 50 0
Coefficient of friction (untreated- all volunteers) ∗ ∗
Distal left forearm
Distal right arm
Proximal right arm
Electrical impedance (untreated- all volunteers) ∗ ∗
Distal left forearm
Distal right arm
Proximal right arm
FIGURE 26.9 Friction coefficient and electrical impedance. There were no significant differences between the distal left volar forearm and the distal right volar forearm. The proximal right volar forearm had a significantly higher friction coefficient and a significantly lower electrical impedance than the distal right volar forearm, and the proximal right arm friction and electrical impedance measurements were different from those of the distal right arm (p < 0.001). (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
COF between age groups (untreated skin) Young Old 1 0.8 0.6 0.4 0.2 0
∗ ∗
Distal left forearm
Distal right arm
Proximal right arm
El between age groups (untreated skin) 300 250 200 150 100 50 0
Distal left forearm
Distal right arm
Proximal right arm
FIGURE 26.10 Age-related comparisons of friction and electrical impedance. No significant differences were apparent between old and young skin on the volar forearm. Within each category, the proximal right arm friction and electrical impedance measurements were different from those of the distal right arm (p < 0.001). (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
26.3 CONCLUSION Although there have been limited studies dealing with the measurement of the skin friction coefficient, past studies
% Increase from untreated skin
Tribological Studies on Skin: Measurement of the Coefficient of Friction
Asians Caucasians
African Americans Hispanics/Latinos
400 300 200 100 0
% Decrease from untreated skin
of the test apparatus is an extremely important factor, because test design parameters can also have an influence on friction measurements. A better appreciation of the importance of the friction coefficient will become clearer as measurement methods improve and allow for greater accuracy.
REFERENCES Occlusion
Petrolatum Intervention
Glycerin
FIGURE 26.11 Coefficient of friction across ethnicity. Data represent the increase in friction when compared to untreated skin of the volar forearm. No significant differences were found between the different ethnic groups. Petrolatum and glycerin increased the friction coefficient significantly more than PVDC occlusion (p < 0.01). The increase in the friction coefficient due to petrolatum was not significantly different from the effect of glycerin. (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
Asians Caucasians
African Americans Hispanics/Latinos
0
−25
−50
−75
223
Occlusion
Petrolatum Intervention
Glycerin
FIGURE 26.12 Change in electrical impedance across ethnicity. Data represent the decrease in electrical impedance when compared to untreated skin of the volar forearm. No significant differences were found between the different ethnic groups. Glycerin lowered the electrical impedance significantly more than PVDC occlusion or petrolatum (p < 0.01). The decrease in the electrical impedance due to PVDC occlusion was not significantly different from the effect of petrolatum. (Adapted from Sivamani, R.K. et al., Skin Res. Technol., 9, 299–305, 2003.)
and our study17 show that differences in skin, because of various factors — such as age and hydration — can be correlated with the friction coefficient. Friction coefficient studies can serve as a quantitative method to investigate how skin differs on various parts of the body and how it differs between different people. It is also a useful method for tracking the changes resulting from environmental treatments, such as sunlight, and when various chemicals are applied to the skin, such as soaps lubricants and skin creams. The reviewed studies also indicate that the design
1. Naylor, P.F.D., The skin surface and friction, Br J Dermatol, 67: 239–248, 1955. 2. El-Shimi, A.F., In vivo skin friction measurements, J Soc Cosmet Chem, 28: 37–51, 1977. 3. Comaish, S., Bottoms, E., The skin and friction: deviations from Amonton’s laws, and the effects of hydration and lubrication, Br J Dermatol, 84: 37–43, 1971. 4. Koudine, A.A., Barquins, M., Anthoine, Ph., Auberst, L., Leveque, J.-L., Frictional properties of skin: proposal of a new approach, Int J Cosmet Sci, 22: 11–20, 2000. 5. Highley, D.R., Coomey, M., DenBeste, M., Wolfram, L.J., Frictional properties of skin, J Invest Dermatol, 69: 303–305, 1977. 6. Comaish, J.S., Harborow, P.R.H., Hofman, D.A., A hand-held friction meter, Br J Dermatol, 89: 33–35, 1973. 7. Prall, J.K., Instrumental evaluation of the effects of cosmetic products on skin surfaces with particular reference to smoothness, J Soc Cosmet Chem, 24: 693–707, 1973. 8. Cua, A., Wilheim, K.P., Maibach, H.I., Frictional properties of human skin: relation to age, sex and anatomical region, stratum corneum hydration and transepidermal water loss, Br J Dermatol, 123: 473–479, 1990. 9. Johnson, S.A., Gorman, D.M., Adams, M.J., Briscoe, B.J., The friction and lubrication of human stratum corneum, in Thin Films in Tribology, Dowson, D. et al., Eds., Elsevier Science Publishers, Amsterdam, 1993, pp. 663–672. 10. Asserin, J., Zahouani, H., Humbert, Ph., Couturaud, V., Mougin, D., Measurement of the friction coefficient of the human skin in vivo. Quantification of the cutaneous smoothness, Colloids Surf B Biointerfaces, 19, 1–12, 2000. 11. Elsner, P., Wilhelm, D., Maibach, H.I., Frictional properties of human forearm and vulvar skin: influence of age and correlation with transepidermal water loss and capacitance, Dermatologica, 181: 88–91, 1990. 12. Lodén, M., Olsson, H., Axéll, T., Linde, Y.W., Friction, capacitance and transepidermal water loss (TEWL) in dry atopic and normal skin, Br J Dermatol, 126: 137–141, 1992. 13. Sulzberger, M.B., Cortese, Jr., T.A., Fishman, L., Wiley, H., Studies on blisters produced by friction, J Invest Dermatol, 47: 456–465, 1966. 14. Nacht, S., Close, J., Yeung, D., Gans, E.H., Skin friction coefficient: changes induced by skin hydration and emollient application and correlation with perceived skin feel, J Soc Cosmet Chem, 32: 55–65, 1981.
224
Handbook of Non-Invasive Methods and the Skin, Second Edition
15. Hills, R.J., Unsworth, A., Ive, F.A., A comparative study of the frictional properties of emollient bath additives using porcine skin, Br J Dermatol, 130: 37–41, 1994. 16. Dawson, D., in Bioengineering of the Skin: Skin Surface Imaging and Analysis, Wilhelm, K.-P., Elsner, P., Berardesca, E., Maibach, H., Eds., CRC Press, Boca Raton, FL, 1997, pp. 159–179. 17. Sivamani, R.K., Goodman, J., Gitis, N.V., Maibach, H.I., Friction coefficient of skin in real-time, Skin Res Technol, 9: 235–239, 2003. 18. Wolfram, L.J., Friction of skin, J Soc Cosmet Chem, 34: 465–476, 1983. 19. Denda, M., in Dry Skin and Moisturizers: Chemistry and Function, Lodén, M., Maibach, H., Eds., CRC Press, Boca Raton, FL, 2000, pp. 147–153. 20. Wolfram, L.J., in Cutaneous Investigation in Health and Disease: Noninvasive Methods and Instrumentation, Leveque, J.-L., Ed., Marcel Dekker, New York, 1989, chap. 3.
21. Zhai, H., Maibach, H.I., Effects of skin occlusion of percutaneous absorption: an overview, Skin Pharmacol Appl Skin Physiol, 14: 1–10, 2001. 22. Cua, A.B., Wilhelm, K.-P., Maibach, H.I., Skin Surface lipid and skin friction: relation to age, sex, and anatomical region, Skin Pharm, 8: 246–251, 1995. 23. Manuskiatti, W., Schwindt, D.A., Maibach, H.I., Influence of age, anatomic site and race on skin roughness and scaliness, Dermatology, 196: 401–407, 1998. 24. Sivamani, R.K., Wu, G.C., Gitis, N.V., Maibach, H.I., Tribological testing of skin products: gender, age, and ethnicity on the volar forearm, Skin Res Technol, 9: 299–305, 2003. 25. Sivamani, R.K., Goodman, J., Gitis, N.V., Maibach, H.I., Coefficient of friction tribological studies in man: an overview, Skin Res Technol, 9: 227–234, 2003.
Friction Evaluation by 27 Skin Unidirectional Stress Using a Friction Tester Mariko Egawa and Motoji Takahashi Bioengineering Research Labs, Shiseido Co., Ltd., Yokohama, Japan
CONTENTS 27.1 Introduction............................................................................................................................................................225 27.2 Methodological Principle ......................................................................................................................................225 27.3 Skin Friction ..........................................................................................................................................................226 27.3.1 In Vivo Measurement of Skin Friction ......................................................................................................226 27.3.2 Relationship between Skin Friction and Other Physiological Parameters...............................................227 27.3.3 Skin Surface Friction and Sensory Evaluation .........................................................................................228 27.3.3.1 Skin Surface Friction after the Application of Emulsion..........................................................228 27.3.3.2 Skin Surface Friction and Sensory Evaluation by Experts .......................................................229 27.3.3.3 Skin Surface Friction and Sensory Evaluation (by Consumers)...............................................230 27.3.3.4 Skin Surface Friction and Fitting of Emulsion in Sensory Evaluation ....................................231 27.4 Recommendation ...................................................................................................................................................231 References .......................................................................................................................................................................231
27.1 INTRODUCTION Although sensory evaluation is important for evaluating cutaneous diseases, and efficacy of cosmetic products and drugs for external use, the differences with the examiner, physiological and physical fluctuations, reliability, and reproducibility must be taken into consideration. Since there are no convenient methods for measurement of sensory evaluation on human skin, objective measurement using instruments is necessary. Sensory properties have been considered to be related to skin friction,1–10 and various devices to measure skin friction have been developed in recent years. These devices have been categorized mainly into two types: one using a probe rotating on the skin and the other using a probe sliding on the skin. The portable friction meter11–13 is one of the many rotatingtype devices.14–21 This handheld device consists of a spring-loaded Teflon wheel that rotates at a constant speed with a constant load. Skin friction has been reported to increase with skin hydration15,16 and with application of moisturizers using this instrument.17 There are other reports using this type of instrument.13,18 The relationship between friction and capacitance or transepidermal water
loss (TEWL) in atopic dry skin and normal skin has also been reported by the oscillating method.19 On the other hand, there are a few sliding-type devices.22–26 As the sensor probe moves like the actual movement of fingers, this type should be more reliable in efficacy evaluation of cosmetics than the rotating type. But there have been few studies on the relation between tactile sensation and skin physiological parameters, including frictional properties.
27.2 METHODOLOGICAL PRINCIPLE The KES-SE Friction Tester25 (Kato Tech Co. Ltd., Kyoto, Japan), a sliding-type device, is a commercial one developed for evaluation of surface frictional property. The apparatus is shown in Figure 27.1. An arm holder is available in addition to the commercial specification of this device for measurement on the human forearm. A sensor holder including a fingerprint-type sensor was placed on a friction detector. In this condition, a fixed load (average force of the finger when touching an object) was added to the sensor. The load and measurement distance are changeable for this purpose. When the sensor moved, the coefficient of sliding friction was recorded. This device is 225
226
Handbook of Non-Invasive Methods and the Skin, Second Edition
30 mm
(b)
(a)
FIGURE 27.1 Friction measuring unit. (a) The general view of the friction measuring unit is shown. An arm holder was attached to measure skin friction on the ventral forearm in vivo. The contact probe was moved on the skin surface on the ventral forearm in the direction from the elbow to wrist. (b) The surface of the contact probe touching the skin is covered with 20 piano wires arranged in parallel and has an area of 100 mm2. The probe was moved in the direction shown by the arrow.
designed to correspond to human skin tactile with an internal second filter and can measure the kinetics of sliding friction. The low frequency below 1 Hz and direct current were eliminated by an active second filter (dumping factor = 0.6, cut frequency band = 1 Hz), based on the fact that the change of long wavelength is not felt rough to human touch. The signal passing through the internal second filter was magnified 10 times, and output was converted to voltage. This voltage was accumulated in an integrator and the result expressed numerically. The average coefficient of friction (μav) and its mean deviation (MD) are usually used with this instrument to indicate the surface friction numerically. μav has been reported to correspond to smoothness sensed by human touch and MD to asperity sense, also by human touch.24
27.3 SKIN FRICTION 27.3.1 IN VIVO MEASUREMENT
OF
SKIN FRICTION
A KES-SE friction tester equipped with an arm holder was used for measurement of skin surface friction. The following measurement conditions were used in our study: the moving speed of the sensor, 1.1 mm/sec; the shift distance, 30 mm; and load, 0.244 N. The sensor gently slid on the skin surface of the ventral forearm in the direction from elbow to wrist shown by the arrow in Figure 27.1, and the data were obtained from each measurement of the mid-zone from 5 mm of each end. The surface of
the sensor touching the skin is covered with 20 piano wires as an imitation of a fingerprint and is 100 mm2 (10 × 10 mm) in area. The parameters usually used with this device are μav and MD. μav is the average coefficient of friction for a 20mm length after cutting off 5 mm each from the starting point and the end point, where the total measurement length was 30 mm. MD indicates the mean deviation for the coefficient of friction for a measurement length of 20 mm, identical with the area used for the calculation of μav: ⎛ ⎛ L ⎞⎞ μ av = ⎜ 1 ⎜ max ⎟ ⎟ ⎝ ⎝ L max ⎠ ⎠ ⎛ ⎛ L ⎞⎞ MD = ⎜ 1 ⎜ max ⎟ ⎟ ⎝ ⎝ L max ⎠ ⎠
∫
L max
μ dL
L min
∫
L max
μ − μ av dL
L min
μ = coefficient of friction The skin surface on the ventral forearm as the test site was measured at 1 hour after washing once with soap. Then, the obtained μav and MD were measured, with the sensor load varying from 2.44 × 10–1 N to 4.89 × 101 N. The moving speed of the sensor was fixed at 1.1 mm/sec during this series of measurements. The effect of load of sensor on μav was stable in the range of load used in this study. This supports previous studies using almost the same applied load (load = 0.003 to 6 N).16,21
Skin Friction Evaluation by Unidirectional Stress Using a Friction Tester
μav, MD (x10)
Water content 0.6
60
0.4
40
0.2
20
0
Before Just after
1.5 4 Time (hr)
5.5
27.3.2 RELATIONSHIP BETWEEN SKIN FRICTION OTHER PHYSIOLOGICAL PARAMETERS
80
MD (x10)
7
Water content (a.u.)
μav
0.8
0
FIGURE 27.2 Effect of water content in the stratum corneum on μav or MD. The changes in μav and MD were measured before and after applying 2 μl/cm2 distilled water on the ventral forearm. μav and MD transiently increased after application of water, and then decreased with the decrease in water content of the stratum corneum measured with the Corneometer CM825.
Figure 27.2 shows the chronological changes of μav, MD, and water content in the stratum corneum27 measured by Corneometer CM825 (Courage+Khazaka Electronic Gmbh, Köln, Germany) after the application of water (2 μL/cm2) on the ventral forearm. μav and MD increased immediately after the application of water and later fell to almost the same level as before the application. These findings suggested that μav and MD on the ventral forearm would be influenced by water content in the stratum corneum.
AND
In total, 53 male or female Japanese volunteers in good health (ages 20 to 51 years) participated in this test. Test sites were on the ventral forearm. In order to avoid environmental influence, volunteers stayed in an air-conditioned room for 30 minutes after washing their forearms with soap. Then, the water content in the stratum corneum,27 skin friction, and viscoelastic properties (Cutometer SEM575, Courage+Khazaka Electronic Gmbh, Germany) were measured. Mechanical parameters are shown in Figure 27.3.28 Following these measurements, skin surface replicas29,30 were obtained from the same test sites. Aging of volunteers influenced neither μav nor MD, as shown in Figure 27.4. The relationship between μav or MD and other physiological parameters was examined by simple linear and stepwise multiple regression analysis, as shown in Table 27.1. A significant correlation was observed between μav and water content in the stratum corneum by simple linear regression analysis. In addition, by stepwise regression analysis it was shown that the value of μav was markedly influenced by water content in the stratum corneum and skin surface patterns, such as Ra(h), which is the arithmetic mean skin roughness value in the direction horizontal to that of the movement of the friction probe 2 9 ; KSD, which indicates average of skin roughness30; and VC1, which indicates anisotropy of skin furrows.30 By stepwise regression analysis, KSD and water content in the stratum corneum were recognized as the most influential factors on MD. Other parameters such as R0, which is a deviation length of the skin when it was
R3
R0 = e(a) R1 = e(a + b)
R0 = Uf
R2 = (e(a) − e(a + b)/e(a) = Ua/Uf Uv
Elongation (mm)
227
R3 = e((r∗a) + ((r −1)∗b)) R4 = e((a + b)∗r)
Ur
R5 = (e(a) − e(a + 0.1))/e(0.1) = Ur/Ue
e(0.1) = Ue
R6 = (e(a) − e(0.1))/e(0.1)) = Ur/Ue
Ua
R7 = (e(a) − e(a + 0.1))/e(a) = Ur/Uf
e(a + 0.1)
R8 = ((e(a)∗a∗100)/f(a) − 1)∗100 Ue Uf
f(x) = sum of e(a)
R4 R1
0 0.1
a a + 0.1
a+b Time (s)
(r∗a) + (r − 1)∗b
(a + b)∗r
FIGURE 27.3 Viscoelasticity parameter. The viscoelasticity parameter was measured using a cutometer: SEM575. The deformation of the cutaneous area was measured after mechanical suction, followed by release of pressure, repeated five times. The parameters were from R0 to R8, as generally used.
228
Handbook of Non-Invasive Methods and the Skin, Second Edition
0.04 r = 0.177 p = 0.205
MD
μav
0.8
0.4
0
0
20
40
60
Age (years)
r = 0.087 p = 0.579
0.02
0
0
20
40
60
Age (years)
FIGURE 27.4 Change in μav and MD with age. No age-related changes were found in the μav or MD values on the ventral forearm.
TABLE 27.1 Stepwise Multiple Regression Analysis of μav or MD vs. Other Physiological Parameters
μav
MD
r
R2
RMSE
p value
Water Ra(h) KSD VC1 R0
0.655 0.230 –0.051 –0.056 0.193
0.4293 0.4474 0.4673 0.4987 0.5271
0.086
<0.0001
KSD Water R0 Rz(h)
–0.518 0.379 –0.007 0.064
0.2682 0.4386 0.4751 0.4972
0.0048
<0.0001
Note: μ av, MD, and other physiological parameters obtained by measurement on the ventral forearm of 53 healthy individuals were subjected to simple linear and stepwise multiple regression analysis. r = simple correlation coefficient; R2 = multiple correlation coefficient; RMSE = residual mean square error; water = water content in the stratum corneum27; R0 = viscoelastic parameters, indicating a deviation length of the skin when it vacuumed28; Ra(h) = arithmetic mean skin roughness value in the direction horizontal to that of the movement of the friction probe29 (ISO 4287:1997); Rz(h) = the 10point average skin roughness peak height in the direction horizontal to that of the movement of the friction probe29 (ISO 4287:1997); KSD = average of skin roughness30; VC1 = anisotropy of skin furrows.30 Probability to enter: 0.250, leave: 0.100
vacuumed with 4 × 104 Pa28, and Rz(h), which is the 10point average skin roughness peak height in the direction horizontal to that of the movement of the friction probe,29 also affect MD. From these results, it was revealed that μav was mainly related to water content of stratum corneum, while MD was mainly related to skin roughness.
27.3.3 SKIN SURFACE FRICTION EVALUATION
AND
SENSORY
27.3.3.1 Skin Surface Friction after the Application of Emulsion Nine kinds of emulsions, which had a different texture when applied to the skin, were used: A to H and S. Contents of moisturizer and oil in each emulsion are shown in Figure 27.5. Emulsion S was used as a standard emulsion whose oil content was in the middle of the nine emulsions. Skin friction parameters and water content in the stratum corneum27 were evaluated by the same method as described above at immediately after and 7 hours after application of 2 μL/cm2. Eighteen normal, healthy volunteers washed their forearms with soap. Then 45 min later, each emulsion was applied to the forearms. Friction parameters and water content of the stratum corneum were measured immediately after and 1.5, 4, 5.5, and 7 h later in an air-conditioned room. Figure 27.6 shows the change in μav (a), MD (b), and water content in the stratum corneum (c) with time after the application of the emulsion. Regarding μav, such samples as A, F, and S, which have a high moisturizer content, showed a high friction coefficient value. Such samples as H and G, which had a low moisturizer content, showed a low friction coefficient value. Sample C, which had a low moisturizer content and high oil content, did not show as low a friction value as low moisturizer content emulsions like H and G did. Regarding the change in water content of stratum corneum, in the same manner as μav, samples A, F, and S showed high values and sample H showed a low value. The differences between each emulsion were strongly indicated by μav rather than MD. The higher the water content in the stratum corneum, the higher the μav, even after the application of emulsion. In general, two factors, adhesion term and projection term, affect fricional properties. The water molecules are thought to affect the adhesion term of the friction force. When emulsions were applied to the skin, water molecules on the skin surface would increase the skin friction coefficient because of the increase of the adhesion term. In
Skin Friction Evaluation by Unidirectional Stress Using a Friction Tester
229
B
F
S
H
10
C
G
0 20
0
F
60
D E
A
D
G
A Water content (a.u.)
Moisturizer (%)
20
40
S
B
40
20
S E
Oil (%)
FIGURE 27.5 Contents of moisturizer and oil in emulsions used in sensory evaluation.
0
Before
E
H
C A F 0
B G
C H
1.5 4 5.5 Time after the application (hr)
D
7
(c)
FIGURE 27.6 (Continued.)
0.5
27.3.3.2 Skin Surface Friction and Sensory Evaluation by Experts
A S
0.4 F 0.3 μav
B
C D
E
0.2
G S E
0.1 0
A F
B G
H D
C H
0 1.5 4 5.5 Time after the application (hr)
Before
7
(a)
FIGURE 27.6 (a) Changes in μav after the application of emulsions. (b) Changes in MD after the application of emulsions. (c) Changes in water content of the stratum corneum after the application of emulsions. 0.025 A
0.02 B
0.015
C
MD
G
F
0.01 0.005 0
S E Before
A F
B G
C H
0 1.5 4 5.5 Time after the application (hr)
D
S
D E H
7
(b)
FIGURE 27.6 (Continued.)
addition, an increase in skin viscosity that would increase the projection term would affect MD. These phenomena can explain the significant differences between each emulsion in vivo measurement.
Six experts used the same nine kinds of emulsions as in the first test (A to H and S). Emulsion S was used as a standard emulsion, and they filled out the questionnaire on the feeling when each emulsion was applied to the skin. The sample was applied to half of the face and the standard to the other half; then immediately and 4 h after the application, they filled out the questionnaire. The experts graded these items on a scale of 7, in comparison with the standard emulsion: +3 (very high), +2 (high), +1 (slightly high), 0 (medium), –1 (slightly low), –2 (low), and –3 (very low). Zero means the same grade as the standard emulsion. Items concerning sensory evaluation were sticky, oily, hydrate, refresh, soft, smooth, slippery, and firm for immediately after, and oily, hydrate, soft, and firm for 4 hours after application. Skin friction parameters and water content in the stratum corneum were evaluated as described above (Section 27.3.3.1). Table 27.2a shows the correlation between sensory evaluation and friction parameters immediately after application, and Table 27.2b shows the correlation 4 h after application. The values in the table are simple correlation coefficients between sensory evaluation and friction parameters or water content of stratum corneum. In the sensory evaluation at the time point of immediately after application, there was a positive correlation between μav and soft or hydrate. In addition, there was a positive correlation between MD and soft. Regarding the result at 4 h after application, there was a positive correlation between μav and oily or hydrate, which suggests that the experts feel more oily or hydrate the higher the coefficient of friction. Also, there was a negative correlation between μav and firm, which suggests that experts feel more firm skin the lower the coefficient of friction. Furthermore, oily and hydrate, which are conflicting tactile sensations, are both positively correlated to μav
230
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 27.2 Results of Simple Linear Regression Analysis of Sensory Evaluation by Expert vs. Skin Friction or Water Content (a) Immediately and (b) 4 Hours after the Application (a)
Sticky
Oily
Hydrate
Refresh
μav MD Water content
0.266 0.257 0.002
0.336 0.209 –0.253
0.477 0.418 –0.138
(b)
Oily
Hydrate
Soft
Firm
μav MD Water content
0.584* 0.264 0.507*
0.670* 0.582* 0.523*
0.032 –0.055 –0.004
–0.567* –0.359 –0.417
–0.33 –0.240 0.244
Soft 0.489* 0.535* 0.031
Smooth
Slippery
0.281 0.281 –0.05
–0.102 0.050 0.163
Firm –0.373 –0.293 0.158
Simple correlation coefficient, *p<0.05, **p<0.01
(Table 27.2b). Ηοwever, when using both μav and MD, they can be distinguished. Oily statistically correlated to only μav, while hydrate correlated to both μav and MD. So, we can distinguish oily and hydrate from skin friction properties. These indicate that it was possible to objectively evaluate sensory aluation using friction parameters. 27.3.3.3 Skin Surface Friction and Sensory Evaluation (by Consumers) A total of 149 Japanese female volunteers used the same emulsions at night that the experts evaluated and filled in the questionnaire immediately after application and before washing their face the next morning. Items concerning sensory evaluation were the same as those of the experts. The consumers made an absolute evaluation of these items on a scale of 4: 4 (high), 3 (medium), 2 (low), and 1 (very low). We examined the relationship between various sen-
sory evaluations and friction parameters of water content in the stratum corneum, namely, sensory evaluation vs. skin friction parameters measured immediately after application, and sensory evaluation the next morning vs. skin friction parameters measured at 7 h after application. Table 27.3 shows the correlation coefficients (a) immediately after application and (b) the next morning. The results were almost similar to those obtained in the experts’ evaluation. MD had a high correlation coefficient with soft immediately after application, while μav showed a high positive correlation coefficient with both oily and hydrate, and MD showed one only with hydrate in the next-morning evaluation. Like in the experts’ evaluation, the sensory evaluation of hydrate correlated with μav and MD, but oily did with only μav. This indicates that these contradictory terms in sensory evaluation can be distinguished using skin friction parameters.
TABLE 27.3 Results of Simple Linear Regression Analysis of Sensory Evaluation by Consumer vs. Skin Friction or Water Content (a) Immediately and (b) the Morning after the Application (a)
Sticky
Oily
Hydrate
Refresh
Soft
Smooth
Slippery
Firm
μav MD Water content
0.434 0.119 –0.243
0.431 0.106 –0.300
0.412 0.454 –0.182
–0.476 –0.241 0.151
0.391 0.550 0.243
0.377 0.446 0.158
–0.441 –0.132 0.246
–0.091 –0.013 0.259
(b)
Oily
Hydrate
Soft
Firm
μav MD Water content
0.654** 0.153 0.557*
0.664** 0.486* 0.633**
0.404 0.455 0.369
0.260 0.417 0.322
Simple correlation coefficient, *p<0.05, **p<0.01
Skin Friction Evaluation by Unidirectional Stress Using a Friction Tester
231
0.4 b
μav
0.3
d e
0.2
a
a b c d e
c 0.1 0
Before 0
30 60 90 120 150 180 210 240 270 300 Time after the application (sec) n = 10, mean ±SD
FIGURE 27.7 Changes in μav after the application of emulsions. (a, b) Fitting well. (c) Neither well nor poorly fitting. (d, e) Fitting poorly.
27.3.3.4 Skin Surface Friction and Fitting of Emulsion in Sensory Evaluation Two emulsions recognized as fitting well in sensory evaluation (a, b), one as neither well nor poorly fitting (c), and two fitting poorly (d, e) were selected (Figure 27.7). The feeling of fitting in sensory evaluation is considered to be related to the increased rate of resistance properties after the application of emulsion. An emulsion whose resistance property increases rapidly is fitting well. Ten normal, healthy volunteers were washing their forearms with soap, then 45 minutes after, skin friction measurement was done as before measurement. Each emulsion (2 μL/cm2) was applied to the forearms. Then the friction parameter was measured every 30 seconds up to 5 minutes after application. The moving speed of the sensor was 10.0 mm/sec in this experiment. Figure 27.7 shows the change in skin friction parameters with time after the application of the emulsion with different fitting properties. There was a rapid decrease in the friction coefficient immediately after the application, which was considered to be an effect of the oil in the emulsion. After that, the friction coefficient increased with time. The emulsions fitting well (a, b) showed a large rate of increase in the coefficient of friction in a short time before reaching a certain value, while the emulsions fitting poorly (c, d) showed a small change in the coefficient of friction (Figure 27.7). It was suggested to be possible to predict the experts’ evaluation of fitting by following the change in the coefficient of friction with time after application of the emulsion.
27.4 RECOMMENDATION Among several skin friction instruments, we considered the sliding probe to be better than the rotating probe for the sensory evaluation of skin, because it demonstrates the actual movement of the finger; that is, people feel their skin condition by sliding their fingers on the skin surface.
We have found that the coefficient of friction after the application of the emulsion was found to be correlated with sensory evaluation using a sliding probe. In the previous reports, the relationship between friction coefficient and sensory evaluation of creams and lotions was made in vitro without using skin.2,31,32 However, in actual skin, the variations in TEWL, water content in the stratum corneum, sebum content, and surface morphology are factors that contribute to the skin friction in a combined manner.33 Therefore, the measurements we made on actual skin34,35 are considered to be valuable. Because people feel that tactile sensation is complicated, it is not possible to express it perfectly with only the friction parameter. However, we could get the friction parameters between skin and the sensor made of piano wires with a friction tester, and the obtained results suggested that it was possible to objectively evaluate tactile sensation after cosmetic usage. For further study, it would be necessary to improve the sensor, load, and parameters to better express human touch sensation.
REFERENCES 1. Kawabata S., A concept of ìnumeriî (smoothness) in fabric hand, Fragrance J., 2: 17–21, 1995 (in Japanese). 2. Suganuma K. and Niwa M., Objective evaluation methods for the hand touch feeling of cosmetics, Fragrance J., 2: 42–54, 1995 (in Japanese). 3. Iida I., Koyanagi T., Isobe Y., and Shimizu J., Studies on typification of cosmetics: application of sensory evaluation to classify milky lotion, J. Soc. Cosmet. Chem. Jpn., 20: 225–231, 1987 (in Japanese). 4. Sulzberger M.B., Cortese T.A., Fishman L., and Wiley H.S., Studies on blisters produced by friction, J. Invest. Dermatol., 47: 456–465, 1996. 5. Zimmerer R.E., Lawson K.D., and Calvert C.J., The effects of wearing diapers on skin, Pediatr. Dermatol., 3: 95–101, 1986. 6. El-Shimi A.F., In vivo skin friction measurements, J. Soc. Cosmet. Chem., 28, 37–51, 1977.
232
Handbook of Non-Invasive Methods and the Skin, Second Edition
7. Armstrong T.J., Mechanical considerations of skin in work, Am. J. Ind. Med., 8: 463–472, 1985. 8. Aker W.A., Measurements of friction injuries in man, Am. J. Ind. Med., 8: 473–481, 1985. 9. Wilkinson D.S., Dermatitis from repeated trauma to the skin, Am. J. Ind. Med., 8: 307–317, 1985. 10. Buchholz B., Frederick L.J., and Armstrong T.J., An investigation of human plantar skin friction and the effects of materials, pinch force and moisture, Ergonomics, 31: 317–325, 1988. 11. Comaish J.S., Harborow P.R.H., and Horman D.A., A hand-held friction meter, Br. J. Dermatol., 89: 33–35, 1973. 12. Prall J.K., Instrumental evaluation of the effects of cosmetic products on skin surfaces with particular reference to smoothness, J. Soc. Cosmet. Chem., 24, 693–707, 1973. 13. Henricsson V., Svensson A., Plisson H., and Axell T., Evaluation of a new device for measuring oral mucosal surface friction, J. Dent. Res., 98: 529–536, 1990. 14. Wolfram L.J., Friction of skin, J. Soc. Cosmet. Chem. Jpn., 34: 465–476, 1983 (in Japanese). 15. Nacht S., Close J.A., Yeung D., et al., Skin friction coefficient: changes induced by skin hydration and emollient application and correlation with perceived skin feel, J. Soc. Cosmet. Chem., 32: 55–56, 1981. 16. Highley D.R., Coomey M., Denbeste M., and Wolfram I.J., Frictional properties of skin, J. Invest. Dermatol., 69: 303–305, 1977. 17. Comanish S. and Bottoms E., The skin and friction: deviations from Amontonís laws, and the effects of hydration and lubrication, Br. J. Dermatol., 84: 37–43, 1971. 18. Olsson H. and Axell T., Objective and subjective efficacy of saliva substitutes containing mucin and carboxymethyl-cellulose, Scand. J. Dent. Res., 9: 316–319, 1991. 19. Cua A.B., Wilhelm K.P., and Maibach H.I., Frictional properties of human skin: relation to age, sex and anatomical region, stratum corneum hydration and transepidermal water loss, Br. J. Dermatol., 123: 473–479, 1990. 20. Cua A.B., Wilhelm K.P., and Maibach H.I., Skin surface lipid and skin friction: relation to age, sex and anatomical region, Skin Pharmacol., 8: 246–251, 1995. 21. Loden M., Olsson H., Axell T., and Linde Y.W., Friction, capacitance and transepidermal water loss (TEWL) in dry atopic and normal skin, Br. J. Dermatol., 126: 137–141, 1992.
22. Nakajima K. and Narasaka H., Evaluation of skin surface associated with morphology and coefficient of friction, Int. J. Cosmet. Sci., 15: 135–151, 1993. 23. Kawabata S., Fuai hyouka no hyoujyunnka to kaiseki, 2nd ed., Nihon Senni Kikai Gakkai, Osaka, Japan, 1980 (in Japanese). 24. Tannba M., Science Research Found, Japan (No. 61300012), Report 162, 1998, (in Japanese). 25. Nomura T., Suzuki K., Suzuki R., and Kajiwara Y., A newly-developed tactile sensor and its application to the evaluation of cosmetics, J. Soc. Cosmet. Chem. Jpn., 31: 79–81, 1997 (in Japanese). 26. Sivamani R.K., Goodman J., Gitis N.V., and Maibach H.I., Friction coefficient of skin in real-time, Skin Res. Technol., 9: 235–239, 2003. 27. Berardesca E., EEMCO guidance for the assessment of stratum corneum hydration: electrical mehods, Skin Res. Technol., 3: 126–132, 1997. 28. Leveque J.L., Cutaneous Investigation in Health and Diesease: Noninvasive Methods and Instrumentation, Marcel Dekker, New York, 1989. 29. Gassmüller J., Kecskes A., and Jahn P., Stylus Method for Skin Surface Contour Measurement: Handbook of Noninvasive Methods and the Skin, CRC Press, Boca Raton, FL, 1995, pp. 83–87. 30. Takahashi M., Image analysis of skin surface contour, Acta. Derm. Venereol. (Stockh.), Suppl. 185: 9–14, 1994. 31. Suganuma K. and Niwa M., Relation between Physical property of message creams and skin surface smoothness, J. Soc. Cosmet. Chem. Jpn., 24: 212–219, 1990 (in Japanese). 32. Masaki K. and Tonani N., The evaluation of oily components by using the frictional feel analyzer, J. Soc. Cosmet. Chem. Jpn., 32: 59–64, 1998 (in Japanese). 33. Sakigara S. and Ishibashi T., Analysis of Frictional properties related to surface roughness of crepe fabric, Senni Gakkaishi, 50: 349–356, 1994 (in Japanese). 34. Egawa M., Oguri M., Hirao T., Takahashi M., and Miyakawa M., The evaluation of skin friction using a frictional feel analyzer, Skin Res. Technol., 8: 41–51, 2002. 35. Egawa M., Hirao T., and Takahashi M., Skin surface friction and sensory evaluation, J. Soc. Cosmet. Chem. Jpn., 37: 187–191, 2003 (in Japanese).
28 Haptic Finger M. Tanaka Department of Bioengineering and Robotics, Graduate School of Engineering, Tohoku University, Sendai, Japan
CONTENTS 28.1 Introduction............................................................................................................................................................233 28.2 Principal of the Haptic Finger...............................................................................................................................233 28.3 Signal Processing...................................................................................................................................................234 28.4 In Vivo Experiment ................................................................................................................................................234 28.5 Conclusion .............................................................................................................................................................235 References .......................................................................................................................................................................236
28.1 INTRODUCTION Touch is the most frequently used action to gather information outside the body. In our daily life, various kinds of things are touched by fingers, and their physical as well as morphological features are extracted and evaluated unconsciously. Thus, the sense of touch is indispensable to our life, as is the sense of sight. In the tactile function, the digital pulp of the finger is pressed against the object, and then the stroke/rubbing action is started over the surface of the object to feel the texture. It seems that most of the force sensors developed so far have been functional to the measurement of just force magnitude,1 while a few sensors are available for the rating of the texture on the object material.2–4 Skin conditions are typically evaluated subjectively through the senses, particularly through sight and touch. Assessment of the pharmaceutical actions of topical drugs on patients with dermatoses is a matter of clinical importance for dermatologists. In cosmetics, smoothness, softness, and firmness are very desirable skin attributes, and the products designed for improving them should be objectively evaluated The development of noninvasive technology in dermatology and cosmetology has advanced considerably over the past decade.5 Concerning the skin surface properties, some attempts were made either by measuring the skin friction coefficients6,7 or by characterizing the skin surface pattern by surfometry,8 image analysis,9 or fringe projection.10 All these methods present advantages, but on the one hand, they are hardly usable in clinics, and on the other hand, their results cannot be compared to the clinician hand sensations. Polyvinylidene fluoride (PVDF) piezopolymer film is flexible and very sensitive to variations in stress or strain.
Electrical current generated by mechanical stress in piezoelectric materials decays as a result of the movement of electric charge. The voltage signal takes the form of a very brief potential wave at the onset of the applied force, and a similar brief wave at termination. This signal increases with the application of increased force, but drops to zero when the applied force remains constant. No response is observed during the stationary plateau of the applied stimulus, and a negative voltage peak is seen as pressure is removed, subsequently decaying again to zero.11 This type of response resembles the output signal from Pacinian corpuscles in human skin.12 This chapter explains a device called a haptic finger,13 used for monitoring the condition of skin surface. PVDF polymer film was used as the receptor for the sensor that directly and noninvasively records the morphological features of the epidermis.
28.2 PRINCIPAL OF THE HAPTIC FINGER The geometry of the haptic finger is presented in Figure 28.1. This sensor is placed over the forefinger and can be Stainless steel
Strain gauge
Sponge rubber
Gauze
PVDF film
Cellophane film
FIGURE 28.1 Geometry of haptic finger. 233
234
Handbook of Non-Invasive Methods and the Skin, Second Edition
slid over the target skin surface. Thus, it is possible to use regardless of the measurement place. A thin stainless steel plate comprises the base of the sensor, with sponge rubber, PVDF foil, acetate film, and gauze layered above. A strain gauge is mounted inside the sensor to monitor force applied to the skin from the sensor. The output signals generated by the strain gauge and the PVDF foil are recorded on a digital oscilloscope through, respectively, a strain gauge amplifier and a band-pass filter (70 Hz to 25 kHz). The sensor is moved over the skin surface while maintaining relative constant speed and pressure. The output signal from the PVDF sensory film is measured for a period of 100.0 msec and sent to a digital storage oscilloscope at sampling times of 0.02 msec, then further stored for the signal processing by personal computer.
28.3 SIGNAL PROCESSING By using the haptic finger, the roughness and softness were focused to extract, although there is much tactile information. For the evaluation of roughness, the amplitude of sensor output was examined. Since the sensor’s raw output is the time response of voltage on PVDF film, which is caused from the deformation by stroking over the human skin surface, and the sensor output has a relation with surface conditions, the sensor output is considered to become larger when the surface is rougher. From these, the calculation of variance of sensor output was done as the parameter of roughness as follows: 1 V= N
N
∑ ( x( j ) − x )
2
(28.1)
j =1
Here, x(j) is the jth quantitized digital signal, x– is the average of x(j), and N is the total number of digital signals. The variance of raw data includes not only roughness, but also other information, such as hardness. In order to extract the parameter for evaluation of only roughness, regardless of the hardness, the extraction of sensor output of some frequency was examined via a wavelet multiplelevel decomposition analysis. The variance of signal of level 5 (790 Hz to 1.57 kHz), which was after the decomposition, was calculated. This frequency level was determined from the experimental conditions. It was confirmed that the variance of level 5 has the highest co-relations with the roughness among all levels. From the result of in vitro experiments of artificial skin made of sponge rubber covered with different silicon rubber sheets, every co-relation efficient of level 5 was more than 0.9. For the second parameter about hardness, the frequency component of the sensor output was investigated, and then fast Fourier transform (FFT) analysis was introduced. The parameter displaying the dispersion of power
spectrum density (PSD) in the frequency domain was calculated according to the following procedures. The PSD was initially calculated from the FFT analysis, and then the values of Fc_L and Fc_M satisfying the following equations were determined: Fc _ M
∑
L
PSD ( n ) =
n =0
Fc _ L
∑ n =0
∑
PSD ( n )
(28.2)
PSD ( n )
(28.3)
n = Fc _ M +1
Fc _ M
PSD ( n ) =
∑
n = Fc _ L +1
Here, L denotes the total data point number in the frequency domain (4096 points) and PSD(n) denotes the nth PSD output. Finally, as a parameter showing the dispersion of PSD in the frequency domain using the two obtained values Fc_L and Fc_M, FcR was calculated according to the following equation: FcR = ( Fc _ M − Fc _ L ) / Fc _ M
(28.4)
It is considered that a small FcR indicates small a power spectrum dispersion in the frequency domain. From the results of in vitro experiments using artificial skin made of the above-mentioned sponge rubber covered with different silicon rubber sheets, we confirmed that the values of FcR were found to be proportional to the hardness of the measured artificial skin, and it has a quite perfect linear relationship.
28.4 IN VIVO EXPERIMENT By using the same system, we tried to measure the human skin condition, and the comparison between the touch feeling of the clinician and sensor output was carried out. The experiment was conducted to measure skin surfaces from 30 people, 14 of whom displayed either xerosis (n = 5), atopy (n = 7), or psoriasis (n = 2). Typical output signals from the sensor (PVDF output) and power spectrum densities (PSDs) are shown in Figure 28.2, corresponding to a normal, healthy subject (smooth and soft) and an atopic patient (rough and hard), respectively. Results of the sensor output (FcR and WL5-Variance) were compared to subjective and clinical evaluations provided by a clinician who scored both roughness and hardness of the skin according to a 5-grade scale for both roughness and hardness (–2, –1, 0, 1, 2). For different reasons, the clinician only assessed 20 of the 30 volunteers.
Haptic Finger
235
50 PVDF output (mV)
PVDF output (mV)
50
0
−50
0
0
−50
0.1
0
0.1
Time (s)
Time (s) 2 PSD (× 103)
PSD (× 103)
6 4 2 0
1
0 0
1000
2000
0
1000
2000
Frequency (Hz)
Frequency (Hz)
FIGURE 28.2 Examples of sensor output and PSD vs. frequency. Left: Atopic dermatitis. Right: Normal, healthy skin. 30 20 Water transpiration quantity (g/m2h)
From the result of evaluations of the clinician in terms of hardness and softness for each patient, the skin condition was categorized as either group I (rough and hard; roughness > 1, hardness > 1), group II (smooth and soft; roughness > 1, hardness > 1), or group III (nongroup I and II; –1 < roughness < 1, –1 < hardness < 1). Transepidermal water loss (TEWL) and hydration of the stratum corneum of the patients were measured by using a Dermalab (Cortex Technology) and a capacitance meter (Skicon 200), respectively. Capacitance (water content) and TEWL for the three subjectively evaluated groups are shown in Figure 28.3. The distribution of the two parameters: WL5-Variance vs. FcR is shown in Figure 28.4. The repartition of the patients into three groups by the noninvasive methods (Figure 28.3) and by the haptic finger method (Figure 28.4) has to be compared with the repartition given by the clinician, who is the reference. This comparison is of course limited to the 20 patients classified in the three different approaches. Before looking at this comparison, we have to be aware that the two noninvasive methods considered here allow characterizing skin physiological parameters, while the clinician and the haptic finger method try to characterize clinical or cosmetic attributes of the skin. As the result of comparison, in Figure 28.4, the patients having rough and hard skin for the clinician (group I) are present in the upper-right cohort defined by the haptic finger. Smooth and soft skins by the clinician (group II) are placed in the lower-left part. At last, the rest is placed in the middle part between groups I and II. Except for the patient of No. 1, all data of haptic finger
Group I Group II Group III
Healthy Xerderma Atopic Psoriasis
I 20 18
III
19 10 21
10
1 17
0
23
4
II
8 16 3 2
11 7 6 25 24 15 9 28 29 22 14 27 12 30
0
26
5
50
13 100
Water content (μS)
FIGURE 28.3 Water content vs. water transpiration quantity.
are placed in the corresponding group. The reinvestigate to all measurement data was done, then it emerged that the force of measurement of only the patient of No. 1 did not achieve the regular one. Then it is considered possible to disregard the data of the patient of No. 1 as a regular one. From these results, the output from this haptic finger has the potential to categorize skin conditions.
28.5 CONCLUSION The haptic finger was designed using PVDF piezopolymer film as a sensory receptor. The effectiveness of this sensor for monitoring human skin conditions was investigated.
236
Handbook of Non-Invasive Methods and the Skin, Second Edition
Wavelet level 5_VAR
20
Healthy
Group I
Xerderma
Group II
Atopic
Group III
10
Psoriasis 21
4
19 23 7
2 1
10
5 29
6
8
24 3
11
9 27
13 28 26
25
17
18 22 16
30 12
0 0.2
20
15
0.4
14 0.6
FcR
FIGURE 28.4 Wavelet level 5 variance vs. FcR.
Signals obtained by sliding the sensor over skin surfaces were processed by wavelet analysis, and the dispersion of the power spectrum density in the frequency domain was obtained. Measurement of skin surfaces was undertaken, and comparisons made between sensor output and subjective and clinical evaluation. It was confirmed that the haptic finger can associate with roughness and hardness at the same time via signal processing. Furthermore, it is shown that the sensor has the potential to substitute as a tactile sensation of a clinician, through in vivo experiment.
REFERENCES 1. J. Jurczyk and K.A. Loparo, Mathematical transform and correlation techniques for object recognition using tactile data, IEEE Trans. Robotics Automation, 5, 359–362, 1989.
2. M. Tanaka, S. Chonan, Z.-W. Jiang, and T. Hikita, Measurement and valuation of touch sensation (AE sensor readings compared with tactile perception of forefinger), Studies Appl. Electromagn. Mech., 13, 289–292, 1998. 3. M. Tanaka, Measurement and valuation of touch sensation: texture measurement on underclothes, Studies Appl. Electromagn. Mech., 12, 53–58, 2002. 4. M. Tanaka and Y. Numazawa, Rating and valuation of human haptic sensation, Int. J. Appl. Electromagn. Mech., 19, 573–579, 2004. 5. H. Tagami, Perspective and prospective views of bioengineering of the skin, Fragrance J., 10, 11–15, 1993. 6. M. Egawa, M. Oguri, T. Hirao, M. Takahashi, and M. Miyakawa, The evaluation of skin friction using a frictional feel analyzer, Skin Res. Technol., 8, 41–51, 2002. 7. A.A. Koudine, M. Barquins, Ph. Anthoine, L. Aubert, and J.L. Lévêque, Frictional properties of skin: proposal of a new approach, Int. J. Cosmet. Sci., 22, 11–20, 2000. 8. S. Makki, J. Mignot, and H. Zahouani, Statistical analysis of the skin surface and three dimensional representation of human skin, J. Soc. Cosmet. Chem., 35, 311–317, 1985. 9. P. Corcuff and J.L. Lévêque, Skin surface replica image. Analysis of furrows and wrinkles, in Handbook of Noninvasive Methods and the Skin, J. Serup and G.B.E. Jemec, Eds., CRC Press, Boca Raton, FL, 1995, pp. 89–96. 10. M. Rohr and K. Schrader, Topometry of human skin (FOITS), SOFW J., 2, 52–59, 1998. 11. G. Harsanyi, Sensing effects and sensitive polymers, in Polymer Films in Sensor Applications, Technomic Publishing Company, Lancaster, PA, 1995, p. 97. 12. G.M. Shepherd, The somatic senses, in Neurobiology, 3rd ed., Oxford University Press, New York, 1994, p. 272. 13. M. Tanaka, J. Leveque, H. Tagami, K. Kikuchi, and S. Chonan, The haptic finger: a new device for monitoring skin condition, Skin Res. Techonol., 9, 131–136, 2003.
Epidermis Structure
Biopsy for Cytologic 29 Cyanoacrylate Evaluation of the Epidermis Jorge E. Arrese, Pascale Quatresooz, Claudine Piérard-Franchimont, and Gérald E. Piérard Department of Dermatopathology, University Hospital Sart Tilman, Liège, Belgium
CONTENTS 29.1 Introduction............................................................................................................................................................239 29.2 Critical Factors for Clinical Practicability of CSSS .............................................................................................239 29.3 Overall Microscopic Aspect of Normal Skin on CSSS........................................................................................239 29.4 Cytological Aspects of Normal Skin on CSSS.....................................................................................................240 29.5 Diagnostic CSSS in Inflammatory Conditions .....................................................................................................240 29.6 Diagnostic CSSS in Cutaneous Neoplasms ..........................................................................................................241 29.7 CSSS Assessment of Disease Severity and Therapeutic Activity ........................................................................241 29.8 Conclusions............................................................................................................................................................241 References .......................................................................................................................................................................241
29.1 INTRODUCTION Cyanoacrylate skin surface stripping (CSSS) is a timehonored method.1 After its clever discovery, it was soon applied for diagnostic purposes.2 Sampling on a polyethylene slide was a decisive improvement in the development of this method.3 Indeed, this kind of plastic sheet is preferable instead of a glass slide for two main reasons: (1) it is possible to get a perfect modeling of curved areas of the body, and (2) the adhesion of the stratum corneum is such that it is not lost during the staining procedure. The method consists of depositing a drop of cyanoacrylate liquid adhesive onto a supple transparent sheet of polyester of terephthalate polyethylene, 175 μm thick, cut to the size of a conventional coverslip (1.5 × 6 cm). The material is pressed firmly on the lesion. After 15 to 30 s, a sheet of stratum corneum can be easily removed. Because the adhesion mechanism is based on a chemical reaction, the depth of stratum corneum removed is determined by the depth of penetration of the adhesive before it hardens. The cleavage level is exclusively located inside the stratum corneum. Oozing and eroded lesions cannot be studied by CSSS.
29.2 CRITICAL FACTORS FOR CLINICAL PRACTICABILITY OF CSSS CSSS can be taken from any part of the body, with two main provisos. On the one hand, sampling from a hairy
area is painful because of pulling out hairs, and the CSSS quality may be inadequate owing to the poor contact with the stratum corneum. It is therefore advisable to shave these areas before a CSSS is harvested. On the other hand, intercorneocyte cohesion on the palms and soles is usually stronger than the cyanoacrylate bond, thus impairing the collection of a uniform sheet of corneocytes. However, a CSSS sampling on these sites is possible in certain physiopathological conditions in which the texture of the stratum corneum is compromised.
29.3 OVERALL MICROSCOPIC ASPECT OF NORMAL SKIN ON CSSS CSSS of normal skin reveals a regular network of highpeaked crests corresponding to the skin surface microdepressions composed of first- and second-order lines.4 Their pattern is typical for specific parts of the body. The primary lines of the skin surface correspond to grooves in the latticework papillary relief at the dermoepidermal junction.5–7 In young individuals, regular intersections of primary and secondary lines delimit regularly shaped polyhedral plateaus. With aging, this network progressively loses its configuration, aligning itself preferentially along the skin tension lines and ending by disappearing in the shallow wrinkles. It is therefore possible to assess the texture of the superficial dermis indirectly on a CSSS. As a result, the dermal aging, corticosteroid-induced 239
240
Handbook of Non-Invasive Methods and the Skin, Second Edition
atrophy, sclerosis, striae distensae, scars, and many other changes in the connective tissue may be observed noninvasively by CSSS. This morphological evaluation of the skin microrelief may be quantified by computerized image analysis using any profilometric method.8,9 When velus hairs are present on the examined site, they are captured with the CSSS. In addition, CSSS also collects follicular casts corresponding to the horny material present at the opening of the pilosebaceous follicles at the skin surface. It is therefore possible to assess the density of the follicles per unit of surface area and to observe the presence of follicular hyperkeratosis (kerosis) and that of comedones, trichostasis spinulosa, and intrafollicular bacteria and mites.8–19
29.4 CYTOLOGICAL ASPECTS OF NORMAL SKIN ON CSSS Cytological characteristics of stratum corneum cells are hardly visible on CSSS unless histological dyes are used.8,9,20 A number of stains are suitable. The most useful and simplest one is a mixture of toluidine blue and basic fuschin in 30% ethanol. Normal skin shows a regular cohesive pattern of adjacent anucleated corneocytes. Their boundaries are clearly stained by a thin polyhedral rim. Parakeratotic cells are normally rare, and they are not clustered on normal skin. They are recognized by the presence of a nucleus central to the polyhedral cell. Saprophitic microorganisms are normally present at the skin surface. Thus, they are encased in the cyanoacrylate bond during sampling, and they are not accessible to the staining procedure. As a result, the surface microflora is not seen on CSSS. By contrast, microorganisms present in the follicular casts can be collected distinctly from the skin surface microflora by scraping out these horny spiky structures appending to the CSSS. Viability of the intrafollicular bacteria can be assessed using flow cytometry.21,22 Melanin can be found in corneocytes of normal skin in phototype V and VI individuals. The dusty load can be specifically revealed using argentaffin-staining procedures. The relative darkness of these CSSSs can be assessed using corneomelametry.23,24 This method consists of measuring the reduction of light transmission through the CSSS using a photometric device adapted to a microscope. From a cytological point of view, it is important to distinguish melanin-laden anucleated corneocytes from neoplastic dendritic melanocytes having migrated inside the stratum corneum above a malignant melanoma.
Microorganisms in the stratum corneum -------------------------------------------No
Yes
Search for serum Diagnosis -------------------------------------------Yes
Spongiotic dermatitis Search for
No
Nonspongiotic dermatitis Search for
- Pattern of distribution of serum
- Parakeratosis
- Parakeratosis
- Inflammatory cells
- Inflammatory cells
- Xerosis
- Xerosis
- Neoplastic cells
FIGURE 29.1 Diagnostic steps and criteria for identifying skin disorders on CSSS.
29.5 DIAGNOSTIC CSSS IN INFLAMMATORY CONDITIONS It is obvious that the diagnostic indications for CSSS involve only disorders in which a change in the stratum corneum occurs. The most common conditions that can be diagnosed by CSSS are summarized in Figure 29.1.2,3,8,9,25–30 Definite diagnoses can be established in superficial infectious and parasitic skin diseases.2,3,8,9,25,29,30 Morphological examination, sometimes combined with fungal cultures, can be carried out to identify these types of disease. By essence, infectious agents that are made visible are not those adhering to the skin surface (see above), but rather those invading the stratum corneum. Fungi, including yeasts and dermatophytes, show their typical morphology, forming clusters or a network of globular or filamentous structures. In the group of parasitic disorders, scabies may pose a problem at the time of sampling.2 In fact, this diagnosis can be established only if the mite, its eggs, or its dejecta are present in the sample. Duplicate CSSS samplings should therefore be taken from a characteristic scabies burrow. The first one removes the roof of the burrow. The second one collects the parasite. Any samples taken outside this lesion, for example, from nonspecific prurigo, will be unhelpful because the diagnosis will simply suggest the presence of a spongiotic dermatitis.8,9,30 Demodex mites are easily recognized8,9,19 and are conveniently highlighted in the follicular casts by the Fite stain. Noninfectious erythemato-squamous disorders include spongiotic and parakeratotic dermatoses and
Cyanoacrylate Biopsy for Cytologic Evaluation of the Epidermis
xeroses.8,9 Spongiotic dermatoses represent superficial inflammatory reactions responsible for spongiosis, microvesiculation, and deposits of serosity inside the stratum corneum. Contact dermatitis, atopic dermatitis, and pityriasis rosea are examples that belong to this group. Parakeratotic dermatoses encompass id reactions, chronic eczema, and stable psoriasis. Seborrhoeic dermatitis also comes within this category in cases when Malassezia yeasts are rare. In active psoriasis, clusters of neutrophils are found in the center of parakeratotic foci.
29.6 DIAGNOSTIC CSSS IN CUTANEOUS NEOPLASMS Some epithelial neoplasms display characteristic aspects of CSSS. Seborrhoeic keratoses show spotty lenticular foci of soft hyperkeratosis. Widening of shallow furrows with hyperkeratosis can be seen. Samples of actinic keratosis often exhibit irregular thickness with interfollicular parakeratosis and xerosis. Actinic porokeratosis is revealed by the rim of cornoid lamellation and loss of the normal microrelief inside the lesion. Verrucous surfaces overlying melanocytic nevi and dermatofibromas are less pathognomonic, but sharp circumscriptions with normal surrounding skin and uniformity of changes in the texture of the horny layer usually are seen in a benign neoplasm. In pigmented neoplasms, melanin can be found inside corneocytes or in atypical melanocytes. Melanin located only in corneocytes is a feature of benign neoplasms, such as lentigos and solar lentigines. The presence of atypical melanocytes in the stratum corneum is strongly suggestive of malignant melanoma, but also, in rare instances, of a benign melanoacanthoma.8,9,28,31 Thus, CSSS proves to be highly sensitive and specific in terms of distinguishing between malignant melanoma and benign melanocytic tumors such as common melanocytic nevi, dysplastic nevi, or pigmented seborrhoeic keratoses. For research purposes, kariometry of neoplastic melanocytes can be performed on CSSS.3 Basal cell carcinomas and squamous cell carcinomas do not exhibit specific or suggestive features.
29.7 CSSS ASSESSMENT OF DISEASE SEVERITY AND THERAPEUTIC ACTIVITY Disease severity and improvement can be assessed noninvasively on CSSS showing specific features in the stratum corneum. An example is given by xeroses that correspond to various forms of predominantly orthokeratotic hyperkeratosis. This may involve what is commonly referred to as dry skin, but this appearance is also found to a more severe degree in ichthyoses.8,9,30,32–34 Several
241
types and grades of orthokeratotic hyperkeratosis can be detected on CSSS. Type 0 is the absence of hyperkeratosis, except for some discrete focal accumulation of corneocytes in the primary lines of the skin. Type 1a corresponds to a continuous linear hyperkeratosis of the primary lines. Type 1b is characterized by hyperkeratosis predominant at the site of adnexal openings either at hair follicles or at acrosyringia. Type 2 corresponds to focal hyperkeratosis of the skin surface plateaus representing less than 30% of the surface of the sampling. Type 3 resembles type 2, but with an altered area over 30% of the skin CSSS. Type 4 is defined by a homogeneous and diffuse hyperkeratosis with persistence of the primary lines. Type 5a resembles type 4, but with loss of recognizable primary lines. Type 5b corresponds to the most heterogeneous and diffuse hyperkeratosis, with loss or marked remodeling of the primary line network. Some quantifications of disease severity and therapeutic activity can be performed on CSSS using computerized image analysis. Quantifications of the fungal load in dermatomycosis can be performed similarly to what has been described in the corneofungimetry bioassay.35,36 Corneomelametry was described above.23,24 Comedometry allows the quantification of the number and size of follicular casts. This method finds application in the comedogenesis and comedolysis-related disorders and treatments.12,13,18,21,22,37
29.8 CONCLUSIONS Beside conventional biopsies and cytology of exudates, imprints, and scrapings, CSSS provides useful information in the field of dermatopathology. This simple and noninvasive method allows the clinician to avoid a conventional biopsy within limits of well-defined indications. Less than 3 min is necessary between sampling and examination. There are evident features and subtle characteristics discernible in the structure of the stratum corneum that enable a diagnosis to be made in a variety of skin diseases. It is important to stress that no single criterion should usually be relied upon for a definitive diagnosis, but rather a constellation of clues should be sought. Quantifications are made possible on CSSS using computerassisted image analysis.
REFERENCES 1. Marks, R. and Dawber, R.P.R. Skin surface biopsy: an improved technique for the examination of the horny layer. Br. J. Dermatol., 84, 117, 1971. 2. Agache, P., Mairey, J., and Boyer, J.P. Le stripping du stratum corneum au cyanoacrylate. Intérêt en physiologie et en pathologie cutanées. J. Med. Lyon, 53, 1017, 1972.
242
Handbook of Non-Invasive Methods and the Skin, Second Edition
3. Lachapelle, J.M., Gouverneur, J.C., Boulet, M., and Tennstedt, D. A modified technique (using polyester tape) of skin surface biopsy. Br. Dermatol., 97, 49, 1977. 4. Hashimoto, K. New methods for surface ultrastructure: comparative studies of scanning electron microscopy, transmission electron microscopy and replica method. Int. J. Dermatol., 13, 357, 1974. 5. Piérard, G.E., Hermanns, J.F., and Lapière, Ch.M. Stéréologie de l’interface dermo-épidermique. Observation de la plasticité de la membrane basale au microscope électronique à balayage. Dermatologica 149, 266, 1974. 6. Piérard, G.E., Franchimont, C., and Lapière, Ch.M. Le vieillissement, son expression au niveau de la microanatomie de la peau. Int. J. Cosmet. Sci., 2, 209, 1980. 7. Rochefort, A., Makki, S., and Agache, P. Anatomical location of human skin furrows. Clin. Exp. Dermatol., 11, 445, 1986. 8. Piérard-Franchimont, C. and Piérard, G.E. Assessment of aging and actinic damages by cyanoacrylate skin surface stripping. Am. J. Dermatopathol., 9, 500, 1987. 9. Piérard-Franchimont, C., and Piérard, G.E. Skin surface stripping in diagnosing and monitoring inflammatory, xerotic and neoplastic diseases. Pediatr. Dermatol., 2, 180, 1985. 10. Holmes, R.L., Williams, M., and Cunliffe, W.J. Pilosebaceous duct obstruction and acne. Br. J. Dermatol., 87, 327, 1972. 11. Marks, R. and Dawber, R.P.R. In situ microbiology of the stratum corneum. Arch. Dermatol., 105, 216, 1972. 12. Mills, O.H. and Kligman, A.M. A human model for assaying comedolytic substances. Br. J. Dermatol., 107, 543, 1982. 13. Mills, O.H. and Kligman, A.M. A human model for assessing comedogenic substances. Arch. Dermatol., 118, 903, 1982. 14. Mills, O.H. and Kligman, A.M. The follicular biopsy. Dermatologica, 167, 57, 1983. 15. Groh, D.G., Mills, O.H., and Kligman, A.M. Quantitative assessment of cyanoacrylate follicular biopsies by image analysis. J. Soc. Cosmet. Chem., 43, 101, 1992. 16. Pagnoli, A., Kligman, A.M., El Gammal, S., and Stoudemayer, T. Determination of density of follicles on various regions of the face by cyanoacrylate biopsy: correlation with sebum output. Br. J. Dermatol., 131, 862, 1994. 17. Piérard, G.E., Piérard-Franchimont, C., and Goffin, V. Digital image analysis of microcomedones. Dermatology, 190, 99, 1995. 18. Letawe, C., Boone, M., and Piérard, G.E. Digital image analysis of the effect of topically applied linoleic acid on acne microcomedones. Clin. Exp. Dermatol., 23, 56, 1998. 19. Forton, F., Seys, B., Marchal, J.L., and Song, A.M. Demodex folliculorum and topical treatment: acaricidal action evaluated by standardized skin surface biopsy. Br. J. Dermatol., 138, 461, 1998. 20. Marks, R. Histochemical applications of skin surface biopsy. Br. J. Dermatol., 86, 20, 1972.
21. Piérard-Franchimont, C., Gaspard, V., Lacante, P., Rhoa, M., Slachmuylders, P., and Piérard, G.E. A quantitative biometrological assessment of acne and hormonal evaluation in young women using a triphasic low-dose oral contraceptive containing gestodene. Eur. J. Contracep. Repr. Health Care, 5, 275, 2000. 22. Piérard-Franchimont, C., Goffin, V., Arrese, J.E., Martalo, C., Braham, C., Slachmuylders, P., and Piérard, G.E. Lymecycline and minocycline in inflammatory acne. A randomized, double-blind study on clinical and in vivo antibacterial efficacy. Skin Pharmacol. Appl. Skin Physiol., 15, 112, 2002. 23. Hermanns, J.F., Petit, L., Piérard-Franchimont, C., Paquet, P., and Piérard, G.E. Assessment of topical hypopigmenting agents on solar lentigines of Asian women. Dermatology, 204, 281, 2002. 24. Petit, L. and Piérard, G.E. Analytic quantification of skin lightening by a 2% hydroquinone-cyclodextrin formulation. J. Am. Acad. Dermatol. Venereol., in press. 25. Whiting, D.A. and Bisset, E.A. The investigation of superficial fungal infections by skin surface biopsy. Br. J. Dermatol., 91, 57, 1974. 26. Knudsen, E.A. The areal extent of dermatophyte infection. Br. J. Dermatol., 92, 413, 1975. 27. Piérard-Franchimont, C. and Piérard, G.E. Apport de la morphométrie et de la biopsie de surface au dépistage du mélanome malin. Rev. Med. Liège, 44, 610, 1989. 28. Piérard, G.E., Piérard-Franchimont, C., Arrese Estrada, J., Delvoye, P., Henry, C., Damseaux, M., Lê, T., Hermanns, J.F., Letot, B., Melotte, P., Van Cauwenberge, D., Ben Mosbah, T., and Gharbi, R. Cyanoacrylate skin surface strippings as an improved approach for distinguishing dysplastic nevi from malignant melanomas. J. Cutan. Pathol., 16, 180, 1989. 29. Piérard, G.E., Piérard-Franchimont, C., and Dowlati, A. La biopsie de surface en dermatologie clinique et expérimentale. Rev. Eur. Dermatol. MST, 4, 445, 1992. 30. Piérard-Franchimont, C. and Piérard, G.E. Biopsies de surface et maladies cutanées. Rev. Med. Liège, 50, 7, 1995. 31. Piérard, G.E., Ezzine Sebai, N., Fazaa, B., Nikkels-Tassoudji, N., and Piérard-Franchimont, C. Karyometry of malignant melanoma cells present in skin strippings. Skin. Res. Technol., 1, 177, 1995. 32. Piérard-Franchimont, C. and Piérard, G.E. Les xéroses: structure de la peau rêche. Int. J. Cosmet. Sci., 6, 47, 1984. 33. Piérard, G.E. EEMCO guidance for the assessment of dry skin (xerosis) and ichthyosis: evaluation by stratum corneum strippings. Skin. Res. Technol., 2, 3, 1996. 34. Piérard-Franchimont, C. and Piérard, G.E. Beyond a glimpse at seasonal dry skin. A review. Exog. Dermatol., 1, 3, 2002. 35. Piérard, G.E., Piérard-Franchimont, C., and Arrese Estrada, J. Comparative study of the activity and lingering effect of topical antifungals. Skin Pharmacol., 6, 208, 1993.
Cyanoacrylate Biopsy for Cytologic Evaluation of the Epidermis
36. Arrese, J.E., Fogouang, L., Piérard-Franchimont, C., and Piérard, G.E. Euclidean and fractal computer-assisted corneofungimetry. A comparison of 2% ketoconazole and 1% terbinafine topical formulations. Dermatology, 204, 222, 2002.
243
37. Uhoda, E., Piérard-Franchimont, C., and Piérard, G.E. Comedolysis by a lipohydroxyacid formulation in acne prone subjects. Eur. J. Dermatol., 13, 65, 2003.
Sonography of the 30 High-Resolution Epidermis In Vivo Stephan El Gammal and Claudia El Gammal Dermatological Clinic, Hospital Bethesda, Freudenberg, Germany
Peter J. Altmeyer Dermatological Clinic, Ruhr University Bochum, Bochum, Germany
Michael Vogt and Helmut Ermert Institute for High Frequency Techniques, Ruhr University Bochum, Bochum, Germany
CONTENTS 30.1 Introduction............................................................................................................................................................246 30.2 Methods..................................................................................................................................................................246 30.2.1 Image Processing .......................................................................................................................................246 30.2.2 Image Analysis ..........................................................................................................................................247 30.2.3 Three-Dimensional Sonography................................................................................................................247 30.3 Patients...................................................................................................................................................................247 30.3.1 Normal Palmar Skin ..................................................................................................................................247 30.3.2 Normal Glabrous Skin...............................................................................................................................247 30.3.3 Skin Diseases .............................................................................................................................................247 30.3.3.1 Psoriasis Vulgaris and Lichen Planus ........................................................................................247 30.3.3.2 Basal Cell Carcinoma.................................................................................................................248 30.3.4 Statistics .....................................................................................................................................................248 30.4 Results....................................................................................................................................................................248 30.4.1 Normal Palmar Skin ..................................................................................................................................248 30.4.1.1 The Horny Layer Is Represented as an Echo-Poor Band Below the Skin Entry Echo .........................................................................................................................248 30.4.1.2 The Stratum Corneum–Stratum Malpighii Interface Is an Echo-Rich Line; the Echo-Poor Band Beneath Represents the Viable Epidermis Together with the Papillary Dermis ..........................................................................................................249 30.4.1.3 Sweat Gland Ducts Are Visible as Echo-Rich Structures in the Horny Layer ........................250 30.4.2 Normal Glabrous Skin...............................................................................................................................250 30.4.2.1 The Horny Layer Is Sonographically Invisible; the Viable Epidermis Is Echo Poor ..............250 30.4.2.2 Hair Follicles Are Echo Poor in the Echo-Rich Reticular Dermis ...........................................250 30.4.3 Skin Diseases .............................................................................................................................................250 30.4.3.1 In Untreated Scaly Psoriatic Plaques, the Horny Layer Is Echo Rich; after Treatment with Petrolatum its Echo Density Decreases ...........................................................251 30.4.3.2 The Acanthotic Epidermis and the Dermis with the Inflammatory Infiltrate Are Represented as One Echo-Poor Band........................................................................................251 30.4.3.3 Tumor Parenchyma and Stroma Are Represented as One Echo-Poor Area .............................251 30.5 Discussion ..............................................................................................................................................................251 References .......................................................................................................................................................................253 245
246
Handbook of Non-Invasive Methods and the Skin, Second Edition
30.1 INTRODUCTION
30.2 METHODS An experimental ultrasound imaging unit was developed that can be operated with different transducers in a fre-
Resolution: δlateral ~ d D f0 δaxial ~ 1 Δf
D Object
fo Center-frequency Δf Bandwidth
Aperture
FIGURE 30.1 Physical parameters influencing the axial and lateral resolutions in ultrasound.
quency range of 20 to 250 MHz. Technical details have been published elswere (Paßmann et al. 1993; El Gammal et al. 1995, 1999). To study the epidermis, we used a 100MHz ceramic transducer. The characteristics of this transducer are summarized in Figure 30.2 and Figure 30.3. To obtain a high lateral resolution, we used a highly focused transducer (with a short length of the focus zone of only 400 μm). In order to obtain a sharp image not only from a stripe of 400 μm, but from a wider part of the skin, we developed a mechanical focusing procedure, which we call B/D-scan (brightness/depth scan). The principle of this method is to compose the sonogram of several 400μm-wide image stripes, which are recorded one after the other, each in the focus zone of the transducer (Paßmann et al. 1993, El Gammal et al. 1995). After the uppermost stripe is recorded by lateral movement of the transducer over the selected area, the transducer is moved vertically 400 μm toward the skin surface before the next image stripe is recorded. To eliminate movement artifacts between adjacent stripes, the overlapping parts (near field and far field) are used for adjustment by the computer program, which puts together the final image.
30.2.1 IMAGE PROCESSING Image processing involved two steps for every picture. First, the internal echoes of the 100-MHz transducer were
Relative amplitude
During the past 15 years, 25-MHz sonography of the skin has gained increasing importance as a noninvasive imaging method in dermatology. Clinical applications are the preoperative determination of the extension of skin tumors (Hoffmann et al. 1992b; Fornage et al. 1993; Gropper et al. 1993; Harland et al. 1993; El Gammal et al. 1993; Semple et al. 1995; Gupta et al. 1996a), the monitoring of inflammatory lesions (Di Nardo et al. 1992; Stiller et al. 1994; Vaillant et al. 1994; Hoffmann et al. 1995; Gupta et al. 1996b) and sclerotic processes (Cole et al. 1981; Serup 1984; Myers et al. 1986; Akesson et al. 1986; Hoffmann et al. 1992a, Lévy et al. 1993; Ihn et al. 1995), and the objective judgment of skin tests, such as patch test reactions (Serup and Staberg 1987; Seidenari and Di Nardo 1992; Seidenari 1995) and tuberculin test reaction (Beck et al. 1986), to name but a few. Today, 25-MHz ultrasound equipment has become widely available at reasonable cost. Sonograms of normal skin show at their upper border a thin, very echo-rich line, the so-called skin entry echo. Below, a broad, echo-rich band with scattered reflexes is seen, which corresponds to the dermis. The subcutaneous fatty tissue is echo lucent with obliquely oriented echorich connective tissue septae. Varying pathological processes (virtually all skin tumors, inflammatory infiltrates, edema, scar tissue, elastosis) as well as skin appendages and large blood vessels are represented as echo-poor areas within the echo-rich dermis (Altmeyer et al. 1992; El Gammal et al. 1993, Fornage et al. 1993). The epidermis cannot be visualized, and certainly structures within the epidermis cannot be differentiated (El Gammal et al. 1993; Fornage et al. 1993), due to the lack of resolution using commercially available transducers with center frequencies of 20 to 25 MHz (resolution of about 80 μm axially and 200 μm laterally). Figure 30.1 shows that the axial resolution is mainly determined by the bandwidth. The lateral resolution is proportional to the center frequency and indirectly proportional to the focal length. To investigate the epidermis, both the axial and lateral resolutions must be improved. By raising the center frequency and bandwidth of the ultrasound transducer, resolution increases, but the signal penetration depth into the skin is reduced (El Gammal et al. 1992a, 1992b, 1993). We modified the 100-MHz transducer technology in such a way that skin structures up to 2 mm depth can still be visualized. Using 100 MHz, we investigated the sonographic characteristics of normal epidermis and stratum corneum in different body regions.
d
1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0
0
50
100
150
200
250
MHz
FIGURE 30.2 Frequency spectrum of the 100-MHz transducer in biological tissue.
High-Resolution Sonography of the Epidermis In Vivo
Center-frequency:
lateral D axial
Bandwidth: Aperture: Focus length:
fo = 80 MHz Δf = 120 MHz D = 3.2 mm d = 4.3 mm
d1 F1 F2 F3
δlateral = 30 μm δaxial = 11 μm Penetration depth: 2–2. 5 mm (using z-scan) Resolution:
(a) Manufacturer Ultran Laboratories Inc., Boalsburg, PA Construction Ceramic transducer with a quartz coupling layer Piezoelectric Center frequency (influenced by the fm = 86 MHz properties water coupling medium)
Geometry
Resolution
Center frequency (after compensation for the water coupling medium) Bandwidth Radius of curvature (focus length) Diameter of the acoustic surface Angle of aperture Relative focus length Length of the focus zone Axial (theoretical) Axial (in practice*) Axial (in biological tissue) Lateral (in the focus zone)
fm = 95 MHz Δf = 107 MHz d = 4.3 mm D = 3.2 mm Θmax= 20˚ F = 1.34 df = 400 μm 6.6 μm 8.7 μm 11 μm 30 μm
* measured on a glass plate. (b)
FIGURE 30.3 (a) A strongly focused transducer with an excellent lateral resolution of 30 μm was used. This transducer has a small length of the focus zone (d1 = 400 μm). To overcome this disadvantage, we used B/D-scan technology, which reconstructs the final B-scan sonogram out of several B-scan image strips taken at different distances from the skin surface. (b) Characteristics of the 100-MHz transducer. The axial resolution was calculated using the width of the ±6-dB level of the received echo signal and assuming a mean sound speed of 1580 m/sec.
eliminated. The oscillation curves of all neighboring Ascans of the image were averaged, and the mean oscillation curve was then substracted from every single A-scan. Secondly, the A-scans were demodulated. The envelope curve is determined by two complex fast Fourier transformations for every A-scan. This procedure provides optimal results, but does not allow prompt viewing of the recorded data. For this purpose we additionally implemented a fast method of demodulation consisting of a digital rectification of the high-frequency A-scan combined with a nonrecursive digital filter of the order of 10. This linear phase filter has a passband cutoff frequency of 150 MHz.
247
of interest within the sonographic image (e.g., entry echo or echo-poor band) were manually delimited by polygons, using a position cursor. For every polygon the mean diameter in the y-axis was calculated as the average length of all A-scans in the polygon. The mean gray level of the polygon area was also determined; its value ranged from 0 (black) to 255 (white).
30.2.3 THREE-DIMENSIONAL SONOGRAPHY A total of 128 serial section gray-level sonograms were recorded and stored in a data block of 128 (lateral xdirection) × 128 (lateral y-direction) × 1024 (in-depth zdirection) points. A two-dimensional horizontal projection (128 × 128 points) was calculated from this data block by adding up the mean brightness intensity along the depth (z) axis.
30.3 PATIENTS All patients and volunteers gave informed consent for all examinations.
30.3.1 NORMAL PALMAR SKIN Ten right-handed volunteers with healthy skin (5 men and 5 women; age, 29 to 76; mean, 56.4) were investigated. Sonograms were taken from the palmar side of the distal segment of the right and left index fingers. Then on the left index finger a Finn® chamber filled with 0.1 ml of petrolatum was applied on the palmar side of the end phalanx, fixed with tape, and left in place for 180 minutes. After removal and wiping off of the petrolatum, sonography was performed again. In one person the horny layer of the palmar side of the right fourth finger was removed by successive stripping with adhesive tape. Sonograms were taken before and after 50, 100, 150, 200, 250, 300, and 350 tape strips. In three volonteers three-dimensional sonograms were recorded from the tip of the left index finger.
30.3.2 NORMAL GLABROUS SKIN Sonograms were taken from volunteers with healthy skin (age, 20 to 32; mean, 24.1) on the abdomen about 3 cm lateral from the umbilicus (n = 8), the upper back over the scapula (n = 11), the dorsal forearm (n = 9), and the calf (n = 14). In several persons we took sonograms on the volar wrist at the transition from palmar to glabrous skin.
30.3.3 SKIN DISEASES
30.2.2 IMAGE ANALYSIS
30.3.3.1 Psoriasis Vulgaris and Lichen Planus
For image analysis we used the program AnalySIS® (Soft Imaging Software GmbH, Münster, Germany). Structures
Thirty-five untreated, infiltrated, and slightly scaly psoriatic lesions on the extremities of 18 patients with chronic
248
Handbook of Non-Invasive Methods and the Skin, Second Edition
plaque-type psoriasis vulgaris and 10 lichen planus papules of 6 patients were investigated. Sonograms were taken in the center and the margin of the lesions and in the surrounding normal skin. In some patients with psoriasis vulgaris we took ultrasound pictures also after occlusive application of petrolatum for 60 minutes and after removal of the scales by tape stripping.
W
E EPB1
EPB2
30.3.3.2 Basal Cell Carcinoma We investigated superficial basal cell carcinoma in nine patients. In 26 patients (psoriasis plaques, n = 11; lichen planus papules, n = 6; basal cell carcinoma, n = 9) a biopsy was taken after sonography. In order to obtain exactly correlating sonographic and histologic images, a 10-mmlong line was drawn on the skin in the plane of the Bscan. After local anesthesia of the area, the skin was cut along this line down to the subcutis. Then a spindle-shaped excision was performed with this cut in the center. The two halves of the tissue spindle were separated and their central cutting planes placed on cardbord. This prevents warping of the tissue during fomalin fixation. In the histological sections, the thickness of the epidermis (from the stratum granulosum to the lowest points of the rete pegs), the thickness of the inflammatory infiltrate (from the uppermost parts of the dermal papillae downward), and the thickness of both the epidermis and the infiltrated dermis were measured, using the program AnalySIS®.
D
(a)
W E
EPB1 EPB2
D
(b)
30.3.4 STATISTICS Using the Wilcoxon test for paired observations, we compared the thickness of the entry echo, the EPB1, and the EPB2 (for definitions, see results) in the right vs. the left index finger, and in untreated skin of the left index finger vs. after the application of petrolatum under occlusion for 180 minutes. The U-test (Mann–Whitney–Wilcoxon) for unpaired observations was used to compare the thickness of the entry echo in normal skin vs. the thickness of the upper echo-rich band in psoriasis. p values of ≤0.05 were considered significant. By means of the linear regression analysis we checked whether there is a significant correlation between the thickness measurements in sonographic images and the corresponding histologic sections.
FIGURE 30.4 Palmar side of left index finger of a 30-year-old woman (distal phalanx). W, water (coupling medium); E, entry echo; EPB1, echo-poor band 1; EPB2, echo-poor band 2; D, dermis. (a) Before treatment. (b) Same location after occlusion with petrolatum for 180 minutes. EPB1 is markedly thickened.
E
EPB2
30.4 RESULTS 30.4.1 NORMAL PALMAR SKIN 30.4.1.1 The Horny Layer Is Represented as an Echo-Poor Band Below the Skin Entry Echo In 100-MHz sonograms of palmar skin, an echo-rich entry echo is seen at the upper border (Figure 30.4 and Figure 30.6). Its thickness on the index finger is shown in
FIGURE 30.5 Palmar side of the right distal fourth finger of a 33-year-old woman (distal phalanx). After more than 200 tape strips, the entry echo and the echo-rich line underneath form a homogeneous echo-rich band (right side of the arrow). *, remnants of EPB1 laterally to region of tape stripping.
High-Resolution Sonography of the Epidermis In Vivo
TABLE 30.1 Thickness of the Entry Echo, EPB1, and EPB2 on the Palmar Side of the Index Finger in 10 Right-Handed Persons
Right index finger (n = 10) Left index finger (n = 10) Left index finger after occlusion with petrolatum for 180 minute (n = 10) Abdomen (n = 8) Upper back (n = 11) Forearm (n = 9) Lower leg (n = 14) a
Entry Echoa
EPB1a
EPB2a
83 ± 8 76 ± 5 88 ± 14
158 ± 48 97 ± 15 194 ± 44
222 ± 25 203 ± 16 230 ± 31
75 91 82 75
± ± ± ±
11 30 12 14
97 110 126 121
± ± ± ±
17 23 16 20
Mean value (μm) ± SD.
W
E
EPB1 EPB2
y D x
proj(x, y) 1.0 0.75 0.5 0.25 0
249
TABLE 30.2 Thickness of the Entry Echo, EPB1, and EPB2 on the Palmar Side of the Distal Fourth Finger of a 33-Year-Old Woman before and after Tape Stripping of the Horny Layer Number of Tape Strips
Entry Echo (µm)
EPB1 (µm)
EPB2 (µm)
0 50 100 150 200 250 300 350
74 70 64 65 62 76 72 68
88 76 81 54 45 24 20 Not visible
158 150 167 167 148 148 158 141
1. In right-handed persons the right index finger has a significantly thicker EPB1 than the left index finger (p = 0.005; see Table 30.1). 2. Application of petrolatum on the distal segment of the index finger for 180 minutes under occlusion results in a 100% increase in the thickness of EPB1 (significant at p < 0.005; see Table 30.1 and Figure 30.4). The thickness of the entry echo does not change significantly (Table 30.1). 3. When the horny layer of the fingertip is gradually removed by tape stripping, the thickness of EPB1 decreases accordingly; finally, when the glistening layer becomes apparent, EPB1 is no longer detectable sonographically (Figure 30.5 and Table 30.2). The thickness of the entry echo does not change (Table 30.2).
FIGURE 30.6 Palmar side of the right index finger of a 39-yearold man at high magnification. W, water (coupling medium); E, entry echo; D, dermis; arrows, sweat gland duct orifices. Distance between two gradation marks = 100 μm. Insert: Horizontal brightness projection of a three-dimensional sonogram from the left index finger of a 29-year-old man (original size, 10 × 5 mm). On the crests of the dermatoglyphics, white points in a row are seen, which represent sweat gland ducts.
30.4.1.2 The Stratum Corneum–Stratum Malpighii Interface Is an Echo-Rich Line; the Echo-Poor Band Beneath Represents the Viable Epidermis Together with the Papillary Dermis
Table 30.1. Where the dermatoglyphics are crossly cut (which is mostly the case), the entry echo is wavy; in parts with longitudinally cut dermatoglyphics, it is a straight line. Below the entry echo, there is an echo-poor band, which we will call EPB1 (echo-poor band 1). It is followed by an echorich line, which runs parallel to the entry echo but is less intense (Figure 30.4 and Figure 30.6). The following observations and experiments give evidence that EPB1 represents the stratum corneum:
Below EPB1, a second echo-poor band is seen, which we will call EPB2 (echo-poor band 2). The thickness of EPB2 on the index finger is shown in Table 30.1. Neither removal nor swelling of the horny layer changes its thickness significantly, as seen in Figure 30.4a and b, Figure 30.5, Table 30.1, and Table 30.2. EPB2 is separated from EPB1 by an echo-rich line. Obviously, this line represents the interface between the water-poor stratum corneum and the moist, living part of the epidermis. The lower border of EPB2 is defined by the scattered reflexes of the dermis (Figure 30.4 to Figure 30.6). This border is too straight
250
Handbook of Non-Invasive Methods and the Skin, Second Edition
to correspond to the undulating dermoepidermal junction (Figure 30.4 and Figure 30.6); it rather represents the interface between the papillary and reticular dermis. 30.4.1.3 Sweat Gland Ducts Are Visible as EchoRich Structures in the Horny Layer In EPB1, twisted, echo-rich, about 100-μm-wide structures are seen, which cross EPB1 vertically. The distance between two of them is 800 to 950 μm, or a multiple. Each of them ends in a small dip on top of a dermatoglyphic crest (Figure 30.6). These structures represent eccrine sweat gland ducts. In EPB2 they are rarely visible; in the echo-rich dermis, they cannot be detected either.
30.4.2 NORMAL GLABROUS SKIN
30.4.2.2 Hair Follicles Are Echo Poor in the EchoRich Reticular Dermis Below EPB2 the reticular dermis is visible as an echorich zone with densely scattered, confluent echo reflexes. It is sharply demarcated from the very echo-poor subcutaneous fat. Within the dermis, hair follicle complexes are visible as homogeneous echo-poor structures (Figure 30.8). When they are cut longitudinally (Figure 30.8a), anagen from telogen follicles can be differentiated.
30.4.3 SKIN DISEASES Compared to normal skin, lesions of psoriasis vulgaris show distinct alterations of the entry echo, EPB1, and EPB2.
30.4.2.1 The Horny Layer Is Sonographically Invisible; the Viable Epidermis Is Echo Poor Figure 30.7 shows the transition from palmar to glabrous skin on the wrist: the upper and lower borders of EPB1 (entry echo and echo-rich line below) merge into one echo-rich line in glabrous skin, so that EPB1 disappears. The thin stratum corneum of glabrous skin obviously cannot be differentiated in 100-MHz sonograms. EPB2 remains as the only echo-poor band between the entry echo and the dermal reflexes. While the thickness of the entry echo is about the same as in palmar skin (Table 30.1), EPB2 of glabrous skin is markedly thinner than on the palms. On the lower extremities it is thicker than on the trunk (Table 30.1).
G
P
D
(a)
D
EPB1 EPB2 D
(b)
FIGURE 30.7 At the transition from palmar to glabrous skin the EBB1 disappears. Wrist of a 33-year-old woman. G, glabrous skin; P, palmar skin; D, dermis. Distance between two graduation marks = 100 μm.
FIGURE 30.8 Hair follicles are represented as echo-poor structures in the echo-rich dermis. (a) Thigh; *longitudinal section of hair follicles. Insert: Correlating histology. (b) Thigh; *cross sections of hair follicles. D, dermis. Distance between two graduation marks = 100 μm.
High-Resolution Sonography of the Epidermis In Vivo
30.4.3.1 In Untreated Scaly Psoriatic Plaques, the Horny Layer Is Echo Rich; after Treatment with Petrolatum its Echo Density Decreases In untreated scaly plaques, directly below the entry echo several parallel, echo-rich lines are seen (Figure 30.9a). They melt with the entry echo into an echo-rich band, which is significantly thicker than the entry echo in normal skin (111 ± 16 μm, n = 35, p < 0.001) and has a much more irregular surface (Figure 30.9a). The following observations show that the described band with varying echo density represents the hyper-
Normal skin
Psoriasis
EPB2
D
(a)
Normal skin
Psoriasis
EPB2
D
(b)
FIGURE 30.9 Border of a psoriatic plaque on the thigh of a 65year-old woman. (a) Before treatment. In the psoriatic plaque, the entry echo and the echo-rich lines underneath form a homogeneous echo-rich band (*), corresponding to the horny layer. Below, EPB2 cannot be separated from the dermis (D) because of a strong dorsal signal attenuation. (b) After occlusion with petrolatum for 60 minutes, the above-mentioned band (*) is increased in thickness and decreased in echo density. It is now demarcated by the entry echo and an echo-rich line at the bottom. EPB2, echo-poor band 2; D, dermis. Distance between two graduation marks = 100 μm.
251
keratotic horny layer: after application of petrolatum under occlusion for 60 minutes on a hyperkeratotic psoriatic plaque, the thickness of the band increases, and its echo density markedly decreases. In its whole length it can now be distinguished as an echo-poor band with the entry echo as the upper border and a thin echo-rich line as the lower border (Figure 30.9b). Repeated tape stripping of the scaly surface results in a gradual decrease in the thickness of the echo-rich band. When the scales are removed entirely, only a single echo-rich line remains. 30.4.3.2 The Acanthotic Epidermis and the Dermis with the Inflammatory Infiltrate Are Represented as One Echo-Poor Band At the border of a psoriatic lesion, EPB2 of normal skin widens into a broad echo-poor band (440 ± 69 μm, n = 35). The thickness of this band correlates very well with the thickness of the acanthotic epidermis plus the dermis with the inflammatory infiltrate in the corresponding histology (r = 0.94). No significant correlation was observed between the thickness of the echo-poor band and the epidermis, respectively, and the infiltrated dermis alone. The lower, quite straight border of EPB2 is defined by the scattered reflexes of the dermis. In lichen planus papules, EPB2 of normal glabrous skin focally wides into a spindle-shaped echo-poor area. The maximum thickness of this band correlates well with the maximum thickness of the acanthotic epidermis plus the dermis with inflammatory infiltrate in the corresponding histology (r = 0.86). Thick lichen planus papules (Figure 30.10) often exhibit an echo-poor line (EPB1) beneath the skin entry echo. 30.4.3.3 Tumor Parenchyma and Stroma Are Represented as One Echo-Poor Area To study whether the improved resolution at 100 MHz has an impact on visualization of skin tumor details, we evaluated thin basal cell carcinoma using image analysis. Correlation of sonograms and histology reveals that the tumor parenchyma and stroma, seen histologically as separate structures (Figure 30.11 insert), are summed up to a uniform spindle-shaped echo-poor area in the upper dermis in sonograms (Figure 30.11).
30.5 DISCUSSION Our 100-MHz ultrasound equipment allows a far more detailed visualization of the upper skin layers than 20MHz sonography. Especially with regard to the in vivo assessment of the horny layer, 100-MHz sonography is a valuable tool. Whereas in normal glabrous skin the stratum corneum is too thin (Kligman 1964, about 12 to 15 μm; Idson 1978, mean thickness of 15 μm for dry stratum
252
Handbook of Non-Invasive Methods and the Skin, Second Edition
Lichen planus EPB2
D
FIGURE 30.10 Lichen planus papule on the thigh of a 65-yearold woman. In the middle of papule an echo-poor line (*) is seen between the skin entry echo and the echo-rich line beneath.Corresponding histology (insert) exhibits that this region has a significant hyperkeratosis. Furthermore, the histology exhibits that EPB2 corresponds to the stratum Malpighii and the inflammatory infiltrate in the upper dermis. EPB2, echo-poor band 2; D, dermis. Distance between two graduation marks = 100 μm.
FIGURE 30.11 Basal cell carcinoma on the back of a 78-yearold woman. Corresponding histology (insert) reveals that tumor parenchyma and stroma (together, 912 μm thick) are summed up to a spindle-shaped echo-poor area (maximal thickness, 770 μm) in the upper dermis of the sonogram. Distance between two graduation marks = 100 μm.
corneum and 48 μm after hydration) to be separated from the entry echo in palmar skin and hyperkeratotic states, it is represented as a distinct band and its thickness can be easily determined. Our results suggest that the echo density of the horny layer depends on its water content: psoriatic scales, appearing silvery because of the included air, are, for example, much more echo rich than the moist stratum corneum of the palms. The significant impedance gap between the stratum corneum and the Malpighian layer — visible as an echo-rich line — can be explained by the different hydration states of these layers. As demonstrated by various methods, e.g., magnetic resonance
imaging (Ablett et al. 1996; Querleux et al. 1994) or measurement of the transepidermal water loss (Kalia et al. 1996), there is a sudden increase in water content at the border between the horny layer and the viable epidermis. In 20-MHz sonography, some authors have regarded the skin entry echo as the sonographic correlate of the epidermis or equated its thickening in hyperkeratotic states with acanthosis (Di Nardo et al. 1992; Schwartz and Murray 1991; Stiller et al. 1994). However, this has been defeated by others, who found no correlation between the sonometric thickness of the entry echo and the histometric thickness of the stratum corneum or the epidermis (Hoffmann et al. 1994; Murakami and Miki, 1989). Today most authors agree that the skin entry echo is an artifact caused by the change in impedance between the coupling water and the horny layer (Gniadecka et al. 1994; Hoffmann et al. 1994; Seidenari 1995). This hypothesis is confirmed by our results: the thickness and echo density of the entry echo remain constant, no matter whether the horny layer is stripped, occluded with topical agents, or entirely removed. In most 20-MHz studies, the echo signal was strongly amplified to reach a high signal depth penetration. This, however, leads to a significant blurring of the entry echo. This effect as well as the low lateral resolution of only 200 μm entails that the entry echo is represented as a 100- to 250-μm-thick relatively homogeneous band (Hoffmann et al. 1994). During image aquisition of the 100-MHz sonograms we applied the B/D-scan technology, which allows for selection of a specific amplification for each of the four to eight horizontal stripes that compose the sonographic image. An overamplification of the entry echo is thus avoided. Its thickness is about 80 μm, and due to the high lateral resolution of 27 μm, it reflects even fine irregularities of the skin surface, like the dermatoglyphics or the rough surface of psoriatic lesions. In 20-MHz sonograms of normal glabrous skin, the dermal reflexes are directly adjacent to the entry echo; the viable epidermis, which is about 80 μm thick, cannot be visualized. At 100 MHz, the resolution is sufficient to show in normal skin a thin echo-poor band above the dermal reflexes. Its thickness and its straight lower border suggest that it represents the viable epidermis together with the papillary dermis. At the transition from normal skin to a psoriatic plaque this band widens into a 400- to 500-μm-thick echo-lucent band. An echo-lucent band of comparable thickness can be observed also in 20-MHz sonograms of psoriatic lesions (Di Nardo et al. 1992; Fornage et al. 1993; Hoffmann et al. 1995; Seidenari 1995; Stiller et al. 1994; Vaillant et al. 1994). Conflicting theories have been proposed regarding its nature. While some authors equate it with the sum of acanthosis and the upper dermis with the inflammatory infiltrate (Fornage et al. 1993; Murakami and Miki 1989; Hoffmann et al. 1995), others interprete it as a correlate of the papillary dermis (Di Nardo et al. 1992; Stiller et al. 1994; Vaillant et al.
High-Resolution Sonography of the Epidermis In Vivo
1994). Our results favor the first hypothesis: comparison with the corresponding histology revealed an excellent correlation between the thickness of this band and the histometric thickness of the Malpighian layer plus the inflammatory infiltrate. Moreover, in 100-MHz sonograms this echo-poor band always has a fairly straight upper and lower border. If it represented the viable epidermis only, we would expect an undulating lower border; if it was the correlate of the papillary dermis, the upper border would be wavy, especially in psoriatic lesions where there are prominent rete pegs. The lateral resolution of the 100MHz transducer is high enough to depict structures of this dimension, as the cross sections of the dermatoglyphics in palmar skin demonstrate. We can conclude that both the viable epidermis and the infiltrated dermis are echo poor and cannot be differentiated from each other. These reflections illustrate that it is not only a question of resolution whether a structure is visualized sonographically. As we could show, the resolution of 100-MHz sonography allows detection of structures as small as a sweat gland duct in the horny layer. On the other hand, the viable epidermis cannot be distinguished from the papillary dermis, and as we learned from the study of skin tumors, the stroma of basal cell carcinoma cannot be distinguished from the tumor cell nests (Figure 30.11). How can this be explained? According to Fields and Dunn (1973), echoes are only reflected from the border between two tissues when they have a different acoustic impedance at the applied frequency. Obviously, there is no difference in impedance between the viable epidermis, compact tumor masses, dense lymphocytic infiltrate, and fine fibrillary connective tissue at 100 MHz, but only between these structures and the reticular dermis. These acoustic tissue properties are the reason why Gupta et al. (1996a) and Semple et al. (1995), using 40- to 60-MHz transducers, could not improve imaging of skin turmors compared with 20 MHz. An entirely exact correlation of histometry and sonometry cannot be expected, as various artifacts influence the measurements in both methods. Histological processing leads to tissue shrinkage; the stratum corneum shows the characteristic basket-weave structure, which does not correspond to the in vivo anatomy. Sonographic examination requires water as a coupling medium, which itself may lead to swelling of the horny layer. Sonometry is also influenced by the sound speed, which is taken as the basis for distance calculations. In dermatological sonography, distance calculations from the echo signal time lapse are usually based on the sound speed of the dermis (1580 m/sec) (Alexander and Miller 1979; Beck et al. 1986). In the nail plate, however, Finlay et al. (1987) found a sound speed of 2140 m/sec, comparing 20-MHz sonography and thickness measurements by a micrometer screw. Jemec and Serup (1989) divided the nail into two
253
compartments with different speeds, an upper dry one (3103 m/sec) and a lower humid inner one (2125 m/sec). Obviously, a similar situation must be postulated for the stratum corneum, which consists of keratin and has a low content of water, thereby having similar properties as the upper nail compartment. We can conclude that high-resolution sonography enables in vivo examination of fine structural details of the epidermis. In order to obtain images with significant information, understanding of the basic physics of ultrasound principles is indispensible. Only optimal preamplification, implementation of a z-scan to increase in-depth penetration at 100 MHz, and knowledge about the influence of resolution on the echo character enable correct interpretation of the sonograms.
REFERENCES 1. Ablett S, Burdett NG, Carpenter TA, Hall LD, Salter DC. Schort echo time MRI enables visualisation of the natural state of human stratum corneum water in vivo. Magn Reson Imaging 13, 357, 1996. 2. Akesson A, Forsberg L, Hederström E, Wollheim E. Ultrasound examination of skin thickness in patients with progressive systemic sclerosis (scleroderma). Acta Radiol Diagn 27, 91, 1986. 3. Alexander HD, Miller L. Determining skin thickness with pulsed ultrasound. J Invest Dermatol 72, 17, 1979. 4. Altmeyer P, Hoffmann K, Stücker M, Goertz S, el-Gammal S. General phenomena of ultrasound in dermatology. In Ultrasound in Dermatology, Altmeyer P, elGammal S, Hoffmann K, Eds. Springer-Verlag, Berlin, 1992, p. 55. 5. Beck JS, Speace VA, Lowe JG, Gibbs JH. Measurement of skin swelling in the tuberculin test by ultrasonography. J Immunol Methods 86, 125, 1986. 6. Cole CW, Handler SJ, Burnett K. The ultrasonic evaluation of skin thickness in sklerederma. J Clin Ultrasound 9, 501, 1981. 7. Di Nardo A, Seidenari S, Giannetti A. B-scanning evaluation with image analysis of psoriatic skin. Exp Dermatol 1, 121, 1992. 8. El Gammal S, Auer T, Hoffmann K, Altmeyer P, Paßmann C, Ermert H. Grundlagen, Anwendungsgebiete und Grenzen des hochfrequenten (20–50 MHz) Ultraschalls in der Dermatologie. Zbl Haut 162, 817, 1993. 9. El Gammal S, Auer T, Hoffmann K, Matthes U, Altmeyer P. Möglichkeiten und Grenzen der hochauflösenden (20 und 50 MHz) Sonographie in der Dermatologie. Akt Dermatol 18, 197, 1992a. 10. El Gammal S, Auer T, Hoffmann K, Matthes U, Hammentgen R, Altmeyer P, Ermert H. High-frequency ultrasound: a non-invasive method for use in Dermatology. In Noninvasive Methods in Dermatology, Frosch P, Kligman AM, Eds. Springer Verlag, Heidelberg, 1993, p. 104.
254
Handbook of Non-Invasive Methods and the Skin, Second Edition
11. El Gammal S, Auer T, Hoffmann K, Paßmann C, Ermert H. High resolution ultrasound of the human epidermis. In In Vivo Examination of the Skin: A Handbook of Noninvasive Methods, Serup J, Jemec GBE, Eds. CRC Press, Boca Raton, FL, 1995, p. 125. 12. El Gammal S, El Gammal C, Kaspar K, Pieck C, Altmeyer P, Vogt M, Ermert H. Sonography of the skin at 100 MHz enables in-vivo-visualization of stratum corneum and viable epidermis in palmar skin and psoriatic plaques. J Invest Dermatol 113, 821, 1999. 13. El Gammal S, Hoffmann K, Auer T, Korten M, Altmeyer P, Höss A, Ermert H, A 50 MHz high-resolution imaging system for dermatology. In Ultrasound in Dermatology, Altmeyer P, el-Gammal S, Hoffmann K, Eds. Springer Verlag, Heidelberg, 1992b, p. 297. 14. Fields S, Dunn F. Correlation of echographic visualizability of tissue with biological composition and physiological state. J Acoust Soc Am 54, 809, 1973. 15. Finlay AY, Moseley H, Duggan TC. Ultrasound transmission time: an in vivo guide to nail thickness. Br J Dermatol 117, 765, 1987. 16. Fornage BD, McGavran MH, Duvic M, Waldron CA. Imaging of the skin with 20-MHz US. Radiology 189, 69, 1993. 17. Gniadecka M, Gniadecki R, Serup J, Søndergaard J. Ultrasound structure and digital image analysis of the subepidermal low echogenic band in aged human skin: diurnal changes and interindividual variability. J Invest Dermatol 102, 362, 1994. 18. Gropper CA, Stiller MJ, Shupack JL, Driller J, Rorke M, Lizzi F. Diagnostic high-resolution ultrasound in dermatology. Int J Dermatol 32, 243, 1993. 19. Gupta AK, Turnbull DH, Foster FS, et al. High-frequency 40-MHz ultrasound. A possible noninvasive method for the assessment of boundary of basal cell carcinoma. Dermatol Surg 22, 131, 1996a. 20. Gupta AK, Turnbull DH, Karasiewicz KA, Shum DT, Watteel GN, Fister FS, Sauder DN. The use of highfrequency ultrasound as a method of assessing the severity of a plaque of psoriasis. Arch Dermatol 132, 658, 1996b. 21. Harland CC, Bamber JC, Gusterson BA, Mortimer PS. High frequency, high resolution B-scan ultrasound in the assessment of skin tumours. Br J Dermatol 128, 525, 1993. 22. Hoffmann K, Dirschka T, Schwarze H, el-Gammal S, Matthes U, Hoffmann A, Altmeyer P. 20 MHz sonography, colorimetry and image analysis in the evaluation of psoriasis vulgaris. J Dermatol Sci 9, 103, 1995. 23. Hoffmann K, el-Gammal S, Gerbaulet U, Schatz H, Altmeyer P. Examination of circumscribed scleroderma using 20 MHz B-scan ultrasound. In Ultrasound in Dermatology, Altmeyer P, el-Gammal S, Hoffmann K, Eds. Springer Verlag, Heidelberg, 1992a, p. 231. 24. Hoffmann K, el-Gammal S, Winkler K, Jung J, Pistorius K, Altmeyer P. Skin tumours in high-frequency ultrasound. In Ultrasound in Dermatology, Altmeyer P, elGammal S, Hoffmann K, Eds. Springer Verlag, Heidelberg, 1992b, p. 181.
25. Hoffmann K, Stücker M, Dirschka T, Görtz S, el-Gammal S, Dirting K, Hoffmann A, Altmeyer P. Twenty MHz B-scan sonography for visualization and skin thickness measurement of human skin. J Eur Acad Dermatol 3, 302, 1994. 26. Holbrook KA, Odland GF. Regional differences in the thickness (cell layers) of the human stratum corneum: an ultrastructural analysis. J Invest Dermatol 62, 415, 1974. 27. Idson B. Hydration and percutaneous absorption. Curr Probl Dermatol 7, 132, 1978. 28. Ihn H, Shimozuma M, Fujimoto M, Sato S, Kikuchi K, Igarashi A, Soma Y, Tamaki K, Takehara K. Ultrasound measurement of skin thickness in systemic sclerosis. Br J Rheumatol 24, 535, 1995. 29. Jemec GBE, Serup J. Ultrasound structure of the human nail plate. Arch Dermatol 125, 643, 1989. 30. Kalia YN, Pirot F, Guy RH. Homogeneous transport in a heteregeneous membrane: water diffusion across human stratum corneum in vivo. Biophys J 71, 2692, 1996. 31. Kligman AM. The biology of the stratum corneum. In The Epidermis, Montagna W, Lobitz WC, Eds. Academic Press, New York, 1964, p. 387. 32. Lévy J, Gassmüller J, Audring H, Brenke A, AlbrechtNebe H. Darstellung der subkutanen Atrophie bei der zirkumskripten Sklerodermie im 20 MHz-B-scan Ultraschall. Hautarzt 44, 446, 1993. 33. Murakami S, Miki Y. Human skin histology using highresolution echography. J Clin Ultrasound 17, 77, 1989. 34. Myers SL, Cohen JS, Sheets PW, Bies JR. B-mode ultrasound evaluation of skin thickness in progressive systemic sclerosis. J Rheumatol 13, 577, 1986. 35. Paßmann C, Ermert H, Auer T, Kaspar K, el-Gammal S, Altmeyer P. In vivo ultrasound biomicroscopy. In IEEE Ultrasonics Symposium Proceedings, 1993, p. 1015. 36. Querleux B, Lévêque JL, de Rigal J. In vivo crosssectional ultrasonic imaging of the skin. Dermatologica 177, 332, 1988. 37. Querleux B, Richard S, Bittoun J, Jolivet O, Idy-Peretti I, Bazin R, Lévêque JL. In vivo hydration profile in skin layers by high-resolution magnetic resonance imaging. Skin Pharmacol 7, 210, 1994. 38. Schwartz SR, Murray RA. Assessment of epithelial thickness by ultrasonic imaging. Decubitus 4, 29, 1991. 39. Seidenari S. High-frequency sonography combined with image analysis: a non-invasive objective method for skin evaluation and description. Clin Dermatol 13, 349, 1995. 40. Seidenari S, Di Nardo A. B scanning evaluation of irritant reactions with binary transformation and image analysis. Acta Derm Venereol Suppl (Stockh) 175, 9, 1992. 41. Semple JL, Gupta AK, From L, Harasiewicz KA, Sauder DN, Foster FS, Turnbull DH. Does high-frequency (40–60 MHz) ultrasound imaging play a role in the clinical management of cutaneous melanoma? Ann Plast Surg 34, 599, 1995.
High-Resolution Sonography of the Epidermis In Vivo
42. Serup J. Decreased skin thickness of pigmented spots appearing in localized scleroderma (morphoea): measurement of skin thickness by 15 MHz pulsed ultrasound. Arch Dermatol Res 276, 135, 1984. 43. Serup J, Staberg B. Ultrasound for assessment of allergic and irritant patch test reactions. Contact Derm 17, 80, 1987.
255
44. Stiller MJ, Gropper CA, Shupack JL, Lizzi F, Driller J, Rorke M. Diagnostic ultrasound in dermatology: current uses and future potential. Cutis 53, 44, 1994. 45. Vaillant L, Berson M, Machet L, Callens A, Pourcelot L, Lorette G. Ultrasound imaging of psoriatic skin: a non-invasive technique to evaluate treatment of psoriasis. Int J Dermatol 33, 786, 1994.
Coherence Tomography in 31 Optical Dermatology Andreas Tycho OCT Innovation ApS., Roskilde, Denmark
Peter Andersen and Lars Thrane Optics and Plasma Research Department, Risø National Laboratory, Roskilde, Denmark
Gregor B.E. Jemec Department of Dermatology, Roskilde Hospital, University of Copenhagen, Copenhagen, Denmark
CONTENTS 31.1 Introduction............................................................................................................................................................257 31.1.1 Time-Domain OCT....................................................................................................................................258 31.1.2 Spatial Resolution......................................................................................................................................258 31.1.3 Penetration Depth ......................................................................................................................................258 31.2 Technical Considerations.......................................................................................................................................258 31.3 Clinical Considerations..........................................................................................................................................260 31.3.1 Probe ..........................................................................................................................................................260 31.3.2 Accurate Positioning of Images ................................................................................................................261 31.3.3 Imaging Speed ...........................................................................................................................................261 31.3.4 Patient Interface .........................................................................................................................................261 31.4 OCT in Dermatology.............................................................................................................................................262 31.4.1 General Imaging ........................................................................................................................................263 31.4.2 Normal Skin Anatomy...............................................................................................................................264 31.4.3 Inflammatory Skin Diseases ......................................................................................................................264 31.4.4 Doppler Imaging........................................................................................................................................264 31.4.5 Skin Tumors...............................................................................................................................................264 31.5 Conclusion .............................................................................................................................................................264 References .......................................................................................................................................................................264
31.1 INTRODUCTION Optical coherence tomography (OCT) has developed rapidly since its potential for applications in clinical medicine was first demonstrated in 1991.1 OCT performs high-resolution, cross-sectional tomographic imaging of the internal microstructure in tissue or other biologic systems by measuring backscattered or backreflected light. In this context, high resolution is on a subcellular scale and the penetration in soft tissue is typically in the millimeter range, depending on the center wavelength of the light source.
The origin of OCT lies in the early work on whitelight interferometry that led to the development of optical coherence-domain reflectometry (OCDR), a one-dimensional optical ranging technique.2 OCDR uses short coherence length light and interferometric detection techniques to obtain high-sensitivity, high-resolution range information. OCDR was developed for finding faults in fiber-optic cables and network components.2 However, its ability to perform ranging measurements in the retina3,4 and other eye structures3–5 was soon recognized. Huang et al.1 then extended the technique of OCDR to tomographic imaging in biological systems by developing the OCT system. The 257
258
Handbook of Non-Invasive Methods and the Skin, Second Edition
OCT system performs multiple longitudinal scans at a series of lateral locations to provide a two- or three-dimensional map of reflection sites in the sample. Recently, a thorough review of OCT technology and systems was published.6 The book6 also includes a comprehensive overview of the various clinical applications of OCT at the time of publication. For a thorough review of the OCT technology and various system realizations, including Fourier domain OCT, refer to Fercher et al.
31.1.1 TIME-DOMAIN OCT The basic concept of OCT imaging can be understood from considering time-domain OCT. The key issue is that with the system the optical path length differences of the internal structures of the sample are measured with respect to a reference path length by using a Michelson interferometer.8 In its concept, it is analogous to time-delay echo known from ultrasound measurements, which are widely known and used clinically. In short, OCT is an imaging technology measuring the backscattering properties of tissues.1 In OCT, the amplitude of light reflected off a sample is interfered with a reference wave, utilizing a technique known as lowcoherence interferometry. An interference signal can only arise when the optical path length difference between the two arms of a Michelson interferometer is smaller than the coherence length of the source. In a layered tissue such an interference signal will be obtained from each boundary crossing, e.g., the epidermis–dermis junction. Therefore, the longitudinal resolution in OCT is proportional to the temporal coherence length of the light source in the system. By using broadband sources it is possible to obtain longitudinal resolution as high as 2 μm.9 As a fiber-based technology, OCT is also readily integrated into many catheter-based endoscopic applications. A longitudinal scan of the sample, called an A-scan, is then performed by scanning the reference mirror position and simultaneously recording the interferometric signal. The interferometric signal is demodulated using bandpass filtering and envelope detection, then digitized and stored on a computer for postprocessing. To acquire data for a two-dimensional image, called a B-scan, a series of longitudinal scans (A-scans) are performed with the optical beam position translated laterally between scans. The data set is then displayed as either a false color or grayscale image. Taking advantage of the advances in technology, video-rate OCT imagery is feasible and has been demonstrated.13
31.1.2 SPATIAL RESOLUTION The longitudinal and lateral resolutions are, in contrast to microscopy, independent of one another. The longitudinal resolution of an OCT image is determined by the
coherence length of the light source. For a light source having a Gaussian spectrum, the coherence length lc is given by
lc =
2 ln 2 λ 2 π Δλ
where λ and Δλ are the center wavelength and the fullwidth-at-half-maximum spectral bandwidth, respectively.11 The first OCT system had a longitudinal resolution of 17 μm in air.1 In 2001, high-resolution OCT, i.e., longitudinal resolution below 2 μm, in ophthalmology was demonstrated in vivo,12 and in 2002 submicron OCT was demonstrated.13 The lateral resolution of an OCT image is determined by the spot size of the sample beam at the depth being probed in the tissue. In a random medium-like tissue, it is necessary to take the scattering of the light into account when determining the spot size.14
31.1.3 PENETRATION DEPTH The spatial resolution of OCT covers the gap between ultra-high-frequency ultrasound (on the order of 50 μm)15 and confocal microscopy,16 as illustrated in Figure 31.1. The penetration depth is mainly determined by the choice of center wavelength of the light source and the optical properties of the tissue being investigated. Furthermore, the following system parameters are of utmost importance: the optical power of the source and the sensitivity of the detector. The optical power impinging on the tissue must not exceed the limits set by the laser safety standards.17 The available source power, detector sensitivity, and other losses in the optical system are characterized by the dynamic range for the system: the better the system, the higher the dynamic range. To obtain the highest penetration depth in tissue, the dynamic range should be maximized. Therefore, in this context, the main limiting properties are the optical scattering coefficient of the tissue and the absorption coefficient. For examples of tissue optical properties, refer to Welch and van Gemert.18 Taking human skin as an example, the penetration depth at center wavelength 1300 nm is approximately 2 mm.
31.2 TECHNICAL CONSIDERATIONS The purpose of this section is to describe the main requirements and drawbacks associated with an OCT system when one wants to obtain a higher resolution, a higher penetration depth, or a higher imaging speed. Differences between time-domain OCT (TD),1 Fourier domain OCT with a swept source (FD1),19 and Fourier domain OCT with a broadband source and a spectrometer (FD2)19 are pointed out.
Optical Coherence Tomography in Dermatology
259
Standard clinical
1 mm Ultrasound Resolution (log)
100 μm
10 μm
Ultra-high frequency
Confocal microscopy
Optical coherence tomography
1 μm
Penetration depth (log) 1 mm
1 cm
10 cm
FIGURE 31.1 Resolution and penetration depth of ultrasound, confocal microscopy, and optical coherence tomography.
TABLE 31.1 OCT System Specifications with Corresponding Requirements, Drawbacks, and Comments System Specifications Higher resolution: Axial
Lateral
Requirements
TD: larger source bandwidth FD1: larger frequency scanning range FD2: larger source bandwidth Larger numerical aperture of focusing optics in probe
Higher penetration depth
Larger dynamic range Larger wavelength
Higher imaging speed
TD: faster reference scanner FD1: faster frequency scanning FD2: higher detector array readout speed
Table 31.1 summarizes the requirements and drawbacks. Regarding the resolution, we have to distinguish between the axial and lateral resolutions. For a TD OCT system, a higher axial resolution is obtained by using a larger source bandwidth. In FD1 and FD2, a larger frequency scanning range and a larger source bandwidth are needed, respectively. No drawbacks are associated with a higher axial resolution. For all three types of OCT systems, a higher lateral resolution is obtained with a larger numerical aperture of the focusing optics in the probe, giving a smaller spot size. For systems with a fixed focus position, this results in a smaller penetration depth due to a smaller depth of focus. Thus, a compromise between a desired lateral resolution and the available depth of focus has to be found. A way out of that trade-off is to use dynamic focus tracking,20 where the focus position moves together with the coherence gate, thus increasing the pen-
Drawbacks
Comments
Fixed focus: smaller penetration depth
Dynamic focus tracking may increase the penetration depth compared to a fixed focus for TD systems Larger wavelength ⇒ lower scattering ⇒ lower signal attenuation (scattering dominates absorption in skin tissue)
Lower dynamic range
etration depth compared to a fixed focus. This is only possible in TD systems. In a highly scattering medium like human skin, the lateral resolution is influenced by the scattering, decreasing it as a function of depth. This dependence of the lateral resolution on the scattering properties of the skin may be calculated using the analysis presented in Thrane et al.14 Another important system specification is the penetration depth, which is the maximum depth in the tissue at which OCT imaging is possible. A higher penetration depth in skin may be obtained by having a larger dynamic range or a larger wavelength. The dynamic range of a system determines the sensitivity and depends upon its design and data acquisition speed. For a typical dynamic range of 100 dB, a sensitivity to reflected signals of –100 dB can be achieved, corresponding to the detection of signals that are 10–10 of the incident optical power. Thus,
260
Handbook of Non-Invasive Methods and the Skin, Second Edition
a system with a larger dynamic range may detect smaller signals, thereby increasing the penetration depth. Another system parameter that determines the penetration depth is the center wavelength of the light source. In a highly scattering tissue like skin, the OCT signal attenuation is normally dominated by the light scattering. The light scattering decreases nearly monotonically with increasing wavelength. Thus, for OCT imaging in skin, a center wavelength of 1300 nm is normally chosen due to a lower light scattering, compared to 800 nm, and a water absorption that is still much smaller than the scattering. The primary requirements for a higher imaging speed are rather different for the three types of OCT systems. A TD OCT system needs a faster reference scanner to increase the imaging speed, whereas FD1 and FD2 OCT systems need a faster frequency scanning and a higher detector array readout speed, respectively. In all three cases, this will result in a lower dynamic range due to the higher noise level associated with the higher bandwidths needed. However, it was shown recently that Fourier domain OCT systems provide a superior dynamic range compared with time-domain OCT systems.21–23 Thus, by using high-speed tunable lasers24 or high-speed line scan cameras,25 the A-scan rate has recently been pushed into the 15- to 30-kHz range without sacrificing sensitivity. Further development along this line will have a huge impact on image acquisition rates, penetration depth, and the ability to realize real-time three-dimensional imaging.
31.3 CLINICAL CONSIDERATIONS In the previous section we described how the system may be optimized to obtain images with a desired specification. In this section we will discuss the remaining design considerations that we have found relevant in our clinical studies.
31.3.1 PROBE With the use of optical fibers, i.e., thin cables for light transport, it is possible to have a hand piece as the probe of an OCT system. Such hand pieces are well known from other clinical equipment based on optics, such as surgical lasers. Furthermore, a fiber-based OCT system offers higher stability, possible compactness, and easier implementation than a system based on traditional free-space optics. However, for OCT there are three main limitations of optical fiber that must be considered in the design of the system. First, optical fibers have a limited bandwidth, which must be considered together with the choice of light source. Typically, standard single-mode fibers for communication are suitable for light sources around 1300 nm, and these have a bandwidth of about 1290 to 1600 nm. For other wavelengths, special fiber must be used at an extra expense. Second, optical fiber exhibits dispersion,
i.e., slightly different propagation speeds for different wavelengths, resulting in a degraded resolution of the images. This is usually not a serious effect for light sources with a bandwidth up to about 70 nm, but for sources with a higher bandwidth fiber dispersion must be considered — either by using short fiber lengths or by compensating, e.g., by using a rapid scanning optical delay line (RSOD) as a reference scanner.26 Finally, twisting of an optical fiber introduces a polarization change resulting in a reduced signal. With our handheld probe, our experience is that the signal strength may vary as much as 50% between measurements due to twisting of the fiber alone. However, the signal is stable in the individual image, so if the absolute values of the signal are not important for your application, polarization effects due to the fiber may be ignored. Polarization effects may be compensated by using quadrature detection, which requires the use of extra optics and an extra photodetector in the system.27 This has the advantage of also eliminating polarization effects introduced by so-called birefringence in the skin, which may give rise to imaging artifacts.28 Furthermore, a quadrature detection system may be modified to measure the birefringence, yielding a so-called polarization-sensitive OCT (PS-OCT) system.29,30 As discussed in Section 31.4.5, birefringence measurements may have clinical value for some applications. With a fiber-based probe it is advantageous to minimize the probe size for easier application to the correct site and reduced strain on the operator. However, there is a trade-off between size and the number of necessary functionalities, such as mechanical control of the focusing lens for dynamic focusing (see Section 31.2) or the integration of a camera. The probe currently in use in our research system is shown in Figure 31.2. Besides the optics for focusing and scanning the beam over the tissue, the probe also includes a video camera for registration of the scan site. The probe is connected to the OCT system by an optical fiber and control cables for the camera and
FIGURE 31.2 The handheld probe of the transportable research system currently in use in our clinical studies.
Optical Coherence Tomography in Dermatology
261
S
P
100 μm
FIGURE 31.4 The dark band of the stratum corneum is disrupted by two spiraled eccrine ducts (S). Within the epidermal–dermal junction the papillary structures (P) are visible. (Image courtesy of ISIS-Optronics GmbH.)
FIGURE 31.3 The probe head of the commercial SkinDex300 system. The entire optical part of the OCT system is incorporated into the probe. (Photo courtesy of ISIS-Optronics GmbH.)
scanner. These cables are enclosed in a flexible steel tube to minimize stretching and twisting. A different approach is to integrate the OCT system and the probe and mounting the system on a mechanical arm. An example of such a system is the commercial SkinDex300® shown in Figure 31.3. The advantages of such an approach are that the drawbacks of using optical fibers may be avoided and that the larger probe head allows for the incorporation of more advanced technology. The SkinDex300 includes a sophisticated system for dynamic focusing, which produces images with close to the highest resolution demonstrated in dermatology (see Figure 31.4). A drawback of this approach may be that a larger probe is less practical for certain applications, such as imaging around the facial features.
31.3.2 ACCURATE POSITIONING
OF IMAGES
One potential advantage of OCT over the excision biopsy is the potential to follow the development of microstructures in the skin over time. This, however, requires accurate registration of the scanning site. One approach is to implement a camera or video camera to capture an image surface of the skin being scanned. Even so, it is often necessary to pinpoint the location of an OCT image accurately by marking the skin with, e.g., a surgical marker or some other longer-lasting pigment. In the choice of a suitable camera, a wealth of compact imaging technology is available from the development of digital cameras and video cameras. It is beyond the scope of this text to give a review of this area, but it is obvious
that the size of the camera chip must be considered in relation to the probe size and the desired imaging quality and speed.
31.3.3 IMAGING SPEED The impact of the scanning speed on the quality of the OCT images was discussed in Section 31.2. However, imaging speed also has clinical implications. In our current system, it takes about 5 seconds to acquire an OCT image with 400 A-scans. This allows for a relatively high penetration depth of about 1.3 mm in most skin tissue, with a dynamic range of 110 dB. At this speed there is no difficulty in avoiding motion artifacts due to the operator or the patient. However, we have found that for some applications exact selection of the scanning position is crucial to finding distinguishing pathologies. It may therefore be advantageous to implement real-time imaging. With such a system the operator may inspect the OCT images while changing the probe position or angle to obtain optimum. Once there, several images may be collected to allow for frame averaging and, in this way, to some degree, compensate for the reduced dynamic range in a fast scanning system.
31.3.4 PATIENT INTERFACE The patient interface is the part of the OCT probe in contact with the patient and through which the scanning is performed. One may choose to either allow the scanning light beam directly on the skin, e.g., through a hole in the probe, or have a glass plate as the interface to the patient. The latter yields control over where the scanning begins and has the advantage that it is possible to use contact media such as ultrasound gel. As discussed above, an OCT image probes differences in refractive indices in the skin similarly to the differences
262
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 31.2 Different Configurations of the Patient Interface Patient interface Air
Glassplate
Advantages The surface of the skin is unrestricted and therefore the true shape is scanned.
Because the surface is pressed against the glass plate it is easier to control where the reference scan should start.
Glassplate + water
The skin surface is allowed to maintain its shape with the water filling any gaps. With the reduction in entrance echo the contrast is increased. GlassEasier to control on the skin than water. plate + ultrasound Very little is absorbed by the skin, so that gel measurements are not affected by the contact medium. GlassGlycerol is absorbed plate + into the skin and glycerol increases penetration depth by decreasing scattering (31).
Glassplate + glycerol after 20 minutes
Disadvantages
OCT scan of human palm
A high entrance reflection reduces the contrast in the remaining scan. This is most noticeable in the transition between epidermis and dermis (ED). The compression of the skin alters the morphological shape of the skin and the high entrance echo is still present. It is difficult to control the water on the skin.
None.
The tissue is affected by the contact medium and therefore absolute measurements of the tissue scattering should be done with care.
Same as above. Glycerol has been reported to increase contrast 20–30 minutes after (31).
ED
0.00 mm 0.20 mm 0.40 mm 0.60 mm 0.80 mm 1.00 mm
0.20 mm
ED
0.40 mm 0.60 mm 0.80 mm 1.00 mm 1.20 mm 1.40 mm 0.20 mm 0.40 mm 0.60 mm 0.80 mm 1.00 mm 1.20 mm 1.40 mm
ED
0.40 mm 0.60 mm 0.80 mm 1.00 mm 1.20 mm 1.40 mm
ED
ED
ED
0.40 mm 0.60 mm 0.80 mm 1.00 mm 1.20 mm 1.40 mm 1.60 mm 0.40 mm 0.60 mm 0.80 mm 1.00 mm 1.20 mm 1.40 mm 1.60 mm
to the acoustic impedance probed by an ultrasound system. Due to the abrupt change in refractive index from air to tissue, there is often a large entrance reflection from the surface of the skin. This abrupt change may be smoothed by the use of a contact medium, thus improving the image contrast in the upper layers of the skin. In Table 31.2 images of a human palm are shown for different configurations of the patient interface together with comments on the pros and cons of each configuration.
31.4
OCT IN DERMATOLOGY
A limited number of studies have been performed using this imaging modality in dermatology. Preliminary data,
however, suggest that OCT imaging may play an important role in future dermatological research by providing high-resolution imaging of the skin in vivo.32 The published studies have relied on purpose-built experimental technology, thereby simultaneously adding flexibility and suffering from lack of standardization. Initial studies have been applied in a manner similar to that of early studies of skin using high-frequency ultrasound, i.e., normal anatomy and tumors. Currently the B-mode images produced by OCT are very similar to those produced by high-resolution ultrasound, and may therefore find immediate resonance with users of ultrasound. The methods, however, differ greatly in their potential for achieving a higher resolution, as discussed. Large
Optical Coherence Tomography in Dermatology
263
0.20 mm
F PD
0.40 mm 0.60 mm 0.80 mm
RD
1.00 mm 1.20 mm 1.40 mm
FIGURE 31.5 OCT scan of normal skin (volar forearm). The scan was recorded using glycerol as a contact medium to reduce the entrance reflection. The epidermis (E) has a low reflectivity relative to the papillary epidermis (PD). The signal fades due to attenuation in the reticular dermis.
0.00 mm E
S
0.20 mm 0.40 mm PD
0.60 mm 0.80 mm RD 1.00 mm 1.20 mm
FIGURE 31.6 The OCT scan of thick skin (palm) is similar to that of thin skin. The epidermis (E) is considerably thicker and penetrated by two spiraling sweat glands (S).
0.80 mm S
1.00 mm
C N
1.20 mm 1.40 mm
R
1.60 mm 1.80 mm
FIGURE 31.7 OCT scan of a human nail (N), limited by the cuticle (C). The root of the nail (R) can be distinguished underneath the overlying skin (S).
prospective studies are currently lacking, and the present understanding of the potentials and limitations of the method is therefore based on early pilot projects.
31.4.1 GENERAL IMAGING The B-mode imaging exemplified by Figure 31.5 through Figure 31.7 shows that the reflected signal has a high peak at the surface of the skin, and is subsequently quickly atten-
uated. The pictures will therefore generally consist of a light band depicting the surface of the skin, underneath which structures are displayed in various tones of gray. Reflective structures such as eccrine gland ducts will show in thick skin, and follicles and hairs may also be visualized.33 Because of scattering and absorption, the signal is weakened in the lower parts of the skin, where the image correspondingly becomes darker and the contrast lower.
264
Handbook of Non-Invasive Methods and the Skin, Second Edition
31.4.2 NORMAL SKIN ANATOMY In contrast to conventional high-frequency ultrasound, epidermal details can be visualized in OCT as seen in Figure 31.5 and Figure 31.6. This is obviously particularly easy in thick stratum corneum, such as that of palms and soles, see Figure 31.6. The penetration of the light to deeper structures is enhanced by application of water, glycerol, or ultrasound gel to the skin. See Table 31.2. In addition to these intraepidermal or superficial dermal structures, OCT can also be used to provide imaging of human nails, where the higher resolution of the OCT confirms the existence of a bilamellar structure, as seen in earlier ultrasound studies (see Figure 31.7). The normal anatomy of the skin is well established, and the level of detail allows the assessment of, e.g., moisturizing effects of the skin, making this methodology suitable not only in clinical medicine, but also for substantiating claims in cosmeceutics or actual cosmetic products. Similar to what has been described in highfrequency ultrasound, age-related changes can be visualized in OCT, except that a white fibrous layer appears in OCT in contrast to the subepidermal low-echogenic band seen in ultrasound. In addition, various short-term experimental interventions have been visualized by OCT: freezing, thermal damage, changes in blood flow, and wound healing.34–37 Specific changes appear to occur, and may further the practical use of this technique.
31.4.3 INFLAMMATORY SKIN DISEASES The measurement of epidermal thickness makes assessment of a number of inflammatory skin diseases possible. Acanthosis and spongiosis can therefore both be quantified directly. At the same time, associated edema improves the optical properties of the tissue and inflammation, therefore possibly allowing a deeper look by OCT. Systematic data are, however, lacking currently, but will become available with greater availability of OCT systems.
31.4.4 DOPPLER IMAGING Similar to ultrasound, an OCT system may also incorporate imaging of blood flow by using Doppler-shift detection.38,39 Only few studies of the techniques utility in OCT have been published,40 but with this capability it can determine local blood flow in the skin, thereby significantly enhancing the potential for determining local viability over earlier methods that provide summation values for volumes of tissue.
31.4.5 SKIN TUMORS Whereas the skin biopsy will most likely remain the gold standard for a number of inflammatory diseases, because
of their strong pathogenic and histological similarities, greater hope is attached to the in vivo diagnosis of skin tumors and OCT. With current development levels, skin tumors appear as homogenous translucent structures surrounded by the normal skin. For the identification of basal cell carcinoma it has been suggested that imaging may be enhanced by polarization-sensitive OCT.41 The differences in A-scans of these tumors may allow the distinction between nonmelanoma skin cancer and precursor lesions, and with increasing resolution of newer systems, it is hoped that this can also be accomplished for melanocytic lesions.
31.5 CONCLUSION Optical coherence tomography (OCT) is an optical equivalent to ultrasound where a near-infrared light beam irradiates the tissue instead of a sonic wave. Due to the extremely short wavelength of light resolution, down to 1 μm can be obtained, but at the cost of a lower penetration depth of about 1 to 1.5 mm. The technique has received great attention in ophthalmology and cardiology in recent years due to its ability to non-invasively produce images with resolution close to histopathology. In this chapter we have reviewed the specifications on an OCT system that influence the image quality and the suitability for use in dermatological studies. The clinical testing of OCT in dermatology is still in its infancy, but the results are promising. With the continuing development of the technology, we expect that OCT will introduce the possibility of giving quantitative measures for a wide range of dermatological pathologies as well as offer diagnostics non-invasively, similar to histopathology.
REFERENCES 1. D. Huang, E.A. Swanson, C.P. Lin, J.S. Schuman, W.G. Stinson, W. Chang, M.R. Hee, T. Flotte, K. Gregory, C.A. Puliafito, and J.G. Fujimoto, Optical coherence tomography, Science 254, 1178–1181, 1991. 2. R.C. Youngquist, S. Carr, and D.E.N. Davies, Optical coherence-domain reflectometry: a new optical evaluation technique, Opt. Lett. 12, 158–160, 1987. 3. A.F. Fercher, K. Mengedoht, and W. Werner, Eye-length measurement by interferometry with partially coherentlight, Opt. Lett. 13, 186–188, 1988. 4. C.K. Hitzenberger, Optical measurement of the axial eye length by laser Doppler interferometry, Inv. Ophthalmol. Vis. Sci. 32, 616–624, 1991. 5. D. Huang, J.P. Wang, C.P. Lin, C.A. Puliafito, and J.G. Fujimoto, Micron-resolution ranging of cornea anteriorchamber by optical reflectometry, Lasers Surg. Med. 11, 419–425, 1991.
Optical Coherence Tomography in Dermatology
6. B. Bouma and G. Tearney, Eds., Handbook of Optical Coherence Tomography, Marcel Dekker, New York, 2002. 7. A.F. Fercher, W. Drexler, C.K. Hitzenberger, and T. Lasser, Optical coherence tomography: principles and applications, Rep. Prog. Phys. 66, 239–303, 2003. 8. B.E.A. Saleh and M.C. Teich, Fundamentals of Photonics, J. Wiley & Sons, New York, 1991. 9. W. Drexler, U. Morgner, F.X. Kärtner, C. Pitris, S.A. Boppart, X.D. Li, E.P. Ippen, and J.G. Fujimoto, In vivo ultrahigh-resolution optical coherence tomography, Opt. Lett. 24, 1221–1223, 1999. 10. A.M. Rollins, M.D. Kulkarni, S. Yazdanfar, R. Ungarunyawee, and J. Izatt, In vivo video rate optical coherence tomography, Opt. Express 3, 219–229, 1998. 11. J.M. Schmitt, Optical coherence tomography (OCT): a review, IEEE J. Select. Topics Quantum Electron. 5, 1205–1215, 1999. 12. W. Drexler et al., Ultrahigh-resolution ophthalmic optical coherence tomography, Nat. Med. 7, 502–507, 2001. 13. B. Povazay, K. Bizheva, A. Unterhuber, B. Hermann, H. Sattmann, A.F. Fercher, W. Drexler, A. Apolonski, W.J. Wadsworth, J.C. Knight, P. St. J. Russell, M. Vetterlein, and E. Scherzer, Submicrometer axial resolution optical coherence tomography, Opt. Lett. 27, 1800–1802, 2002. 14. L. Thrane, H.T. Yura, and P.E. Andersen, Analysis of optical coherence tomography systems based on the extended Huygens-Fresnel principle, J. Opt. Soc. Am. A 17, 484–490, 2000. 15. John G. Webster, Ed., Medical Instrumentation: Application and Design, J. Wiley & Sons, New York, 1998. 16. J. Powley, Handbook of Biological Confocal Microscopy, Plenum Press, New York, 1995. 17. D. Sliney and M. Wolbarsht, Safety with Lasers and Other Optical Sources, Plenum Press, New York, 1980. 18. A.J. Welch and M.J.C. van Gemert, Eds., Appendix to Chapter 8, in Optical-Thermal Response of Laser-Irradiated Tissue, Plenum Press, New York, 1995. 19. A.F. Fercher, C.K. Hitzenberger, G. Kamp, and S.Y. ElZaiat, Measurement of intraocular distances by backscattering spectral interferometry, Opt. Commun. 117, 43–48, 1995. 20. J.M. Schmitt, S.L. Lee, and K.M. Yung, An optical coherence microscope with enhanced resolving power in thick tissue, Opt. Commun. 142, 203–207, 1997. 21. R. Leitgeb, C.K. Hitzenberger, and A.F. Fercher, Performance of Fourier domain vs. time domain optical coherence tomography, Opt. Express 11, 889–894, 2003. 22. M.A. Choma, M.V. Sarunic, C. Yang, and J.A. Izatt, Sensitivity advantage of swept source and Fourier domain optical coherence tomography, Opt. Express 11, 2183–2189, 2003. 23. J.F. de Boer, B. Cense, B.H. Park, M.C. Pierce, G.J. Tearney, and B.E. Bouma, Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography, Opt. Lett. 28, 2067–2069, 2003. 24. S.H. Yun, G.J. Tearney, J.F. de Boer, N. Iftimia, and B.E. Bouma, High-speed optical frequency-domain imaging, Opt. Express 11, 2953–2963, 2003.
265
25. N.A. Nassif, B. Cense, B.H. Park, S.H. Yun, T.C. Chen, B.E. Bouma, G.J. Tearney, and J.F. de Boer, In vivo human retinal imaging by ultrahigh-speed spectral domain optical coherence tomography, Opt. Lett. 29, 480–482, 2004. 26. K.F. Wong, D.Yankelevich, K.C. Chu, J.P. Heritage, and A. Dienes, 400-hz mechanical optical delay line, Opt. Lett. 18, 558–560, 1993. 27. J. Schmitt, D. Kolstad, and C. Petersen, Intravascular optical coherence tomography, Opt. Phys. News 15, 20–25, 2004. 28. J.F. de Boer, T.E. Milner, M.G. Ducros, S.M. Srinivas, and J.S. Nelson, Polarization-sensitive OCT, in Handbook of Optical Coherence Tomography, B.E. Bouma and G.J. Tearney, Eds., Marcel Dekker Publ., New York, 2001, pp. 237–274. 29. M.R. Hee, D. Huang, E.A. Swanson, and J.G. Fujimoto, Polarization-sensitive low-coherence reflectometer for birefringence characterization and ranging, J. Opt. Soc. Am. B 9, 903–908, 1992. 30. C.E. Saxer, J.F. de Boer, B.H. Park, Y. Zhao, Z. Chen, and J.S. Nelson, High-speed fiber-based polarizationsensitive optical coherence tomography of in vivo human skin, Opt. Lett. 25, 1355–1357, 2000. 31. Y.H. He and R.K. Wang, Dynamic optical clearing effect of tissue impregnated with hyperosmotic agents and studied with optical coherence tomography, J. Biomed. Opt. 9, 200–206, 2004. 32. J. Welzel, Optical coherence tomography in dermatology: a review, Skin Res. Technol. 7, 1–9, 2001. 33. J. Welzel, C. Reinhardt, E. Lankenau, C. Winter, and H.H. Wolff, Changes in function and morphology of normal human skin: evaluation using optical coherence tomography, Br. J. Dermatol. 150, 220–225, 2004. 34. B. Choi, T.E. Milner, J. Kim, J.N.Goodman, G. Vargas, G. Aguilar, and J.S. Nelson, Use of optical coherence tomography to monitor biological tissue freezing during cryosurgery, J. Biomed. Opt. 9, 282–286, 2004. 35. M.C. Pierce, R.L. Sheridan, B. Hyle Park, B. Cense, and J.F. De Boer, Collagen denaturation can be quantified in burned human skin using polarization-sensitive optical coherence tomography, Burns 30, 511–517, 2004. 36. L.L. Otis, D. Piao, C.W. Gibson, and Q. Zhu, Quantifying labial blood flow using optical Doppler tomography, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod. 98, 189–194, 2004. 37. A.T. Yeh, B. Kao, W.G. Jung, Z. Chen, J.S. Nelson, and B.J. Tromberg, Imaging wound healing using optical coherence tomography and multiphoton microscopy in an in vitro skin-equivalent tissue model, J. Biomed. Opt. 9, 248–253, 2004. 38. J.A. Izatt, M.D. Kulkarni, S. Yazdanfar, J.K. Barton, and A.J. Welch, In vivo bidirectional color Doppler flow imaging of picoliter blood volumes using optical coherence tomography, Opt. Lett. 22, 1439–1441, 1997.
266
Handbook of Non-Invasive Methods and the Skin, Second Edition
39. Z. Chen, T.E. Milner, S. Srinivas, X. Wang, A. Malekafzali, M.J.C. van Gemert, and J.S. Nelson, Noninvasive imaging of in vivo blood flow velocity using optical Doppler tomography, Opt. Lett. 22, 1119–1121, 1997. 40. Y. Zhao, Z. Chen, C. Saxer, S. Xiang, J.F. de Boer, and J.S. Nelson, Phase-resolved optical coherence tomography and optical Doppler tomography for imaging blood flow in human skin with fast scanning speed and high velocity sensitivity, Opt. Lett 25, 114–116, 2000.
41. J. Strasswimmer, M.C. Pierce, B.H. Park, V. Neel, and J.F. deBoer, Polarization-sensitive optical coherence tomography of invasive basel cell carcinoma, J. Biomed. Opt., 9, 292–298, 2004.
Vivo Reflectance Mode Confocal 32 In Microscopy in Clinical and Surgical Dermatology Salvador González Dermatology Service, Memorial Sloan-Kettering Cancer Center, New York, New York, and Department of Dermatology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, and Clinica La Luz, Madrid, Spain
Yolanda Gilaberte-Calzada San Jorge Hospital, Huesca, Spain
Pedro Jaén-Olasold Dermatology Service, Ramon y Cajal Hospital and Clinica La Luz, Madrid, Spain
Milind Rajadhyaksha Dermatology Service, Memorial Sloan-Kettering Cancer Center, New York, New York
Abel Torres Division of Dermatology, Loma Linda University Hospital, Loma Linda, California
Allan Halpern Dermatology Service, Memorial Sloan-Kettering Cancer Center, New York, New York
CONTENTS 32.1 32.2 32.3 32.4
Introduction............................................................................................................................................................268 Laser Scanning Confocal Microscopy: Basic Principles of Reflectance .............................................................268 Reflectance Confocal Microscopy Findings of Normal Skin...............................................................................269 Reflectance Confocal Microscopy Findings of Inflammatory Skin Lesions and Cutaneous Infections .............270 32.4.1 Acute Contact Dermatitis ..........................................................................................................................270 32.4.2 Psoriasis .....................................................................................................................................................271 32.4.3 Cutaneous Infections .................................................................................................................................271 32.5 Reflectance Confocal Microscopy Findings of Neoplastic Skin Lesions ............................................................271 32.5.1 Nonpigmented Lesions ..............................................................................................................................271 32.5.1.1 Actinic Keratosis and Squamous Cell Carcinoma ....................................................................271 32.5.1.2 Basal Cell Carcinoma.................................................................................................................272 32.5.2 Pigmented Lesions.....................................................................................................................................272 32.5.2.1 Melanocytic Nevi .......................................................................................................................272 32.5.2.2 Malignant Melanoma .................................................................................................................272 32.6 Use of Reflectance Confocal Scanning Laser Microscopy as Adjunct to Standard Therapy..............................274 32.6.1 Use of Confocal Microscopy for Margin Assessment..............................................................................274 32.6.2 Use of Confocal Microscopy for Evaluating Response to Treatment......................................................274 267
268
Handbook of Non-Invasive Methods and the Skin, Second Edition
32.7 Difficulties and Potential Solutions.......................................................................................................................274 32.8 Summary ................................................................................................................................................................274 Acknowledgments ...........................................................................................................................................................275 References .......................................................................................................................................................................275
32.1 INTRODUCTION Conventional histopathology for screening or diagnosing diseases is an invasive method that involves excising a small piece of tissue (biopsy), fixing, cutting into thin slices (sectioning), and staining with dyes. The stained sections are then examined with a light microscope. While high resolution of detail can be accomplished by this method, biopsies are painful, scarring, expensive, and time-consuming. New imaging techniques potentially obviate these problems and offer physicians a non-invasive high-resolution analysis of lesions and operate in real time. These include optical coherence tomography,1 highfrequency ultrasound,2 magnetic resonance imaging,3 and reflectance confocal microscopy (RCM).4–6 Of these, RCM offers the highest resolution imaging comparable to routine histology.
32.2 LASER SCANNING CONFOCAL MICROSCOPY: BASIC PRINCIPLES OF REFLECTANCE The first report of the use of confocal scanning laser microscopy for reflectance imaging of human skin in vivo appeared in 1995.6 The advantages of in vivo RCM over conventional histology are several-fold. The imaging is painless and non-invasive, causing no tissue damage, and the tissue is not altered in any way by processing or staining, thus minimizing artifact. A very important aspect is the rapidity with which the data can be collected and also the ability to evaluate dynamic changes such as disease evolution in real time.7–11 The use of a point source of light illuminating a small spot within tissue, followed by detection of backscattered light through an optically conjugate aperture (pinhole) is the basic principle of confocal microscopy. The backscattered light must pass through a pinhole, such that only the in-focus light reaches the detector, while out-of-focus light is rejected. Hence, only the single plane within the specimen that is in focus is detected (Figure 32.1). The numerical aperture of the objective lens, the wavelength, and the detection aperture (pinhole) size determine the resolution of the images produced by RCM. Lasers of different wavelengths may be used as the light source for reflectance confocal microscopy. Longer, near-infrared wavelengths penetrate deeper into the skin but provide lower resolution, compared to shorter, visible wavelengths. Backscattering of light occurs due to local variations of the refractive index within the tissue and, for a specific organelle or
Beamsplitter Skin site
Laser source
Small aperture
Objective lens
Detector
FIGURE 32.1 Scheme of a reflectance mode confocal microscope illustrating noninvasive imaging of a thin (focused) plane of skin. Backscattered light is detected from the skin rather than transmitted light. The small aperture (pinhole) in front of the detector collects only the light in focus, while rejecting light that is out of focus.
structure, depends on the refractive index relative to that of the immediate surrounding environment, as well as the size relative to the wavelength. Melanosomes produce strong backscattering with visible (400 to 700 nm) and near-infrared (700 to 1064 nm) wavelengths used in our laboratories because of a high refractive index compared to the surrounding epidermis. Hence, cells containing melanin, such as basal keratinocytes and melanocytes, image brightly. The commercially available RCM that we use has a wavelength of 830 nm and a 30× objective lens of NA 0.9, which provides a lateral resolution of approximately 1 μm and an axial resolution (section thickness) of 3 μm. With this system it is possible to image normal skin to a depth of 200 to 350 μm, sufficient for imaging the epidermis and upper dermis (papillary and upper reticular dermis). The laser power used in the commercial device is less than 30 mW and causes no tissue damage or eye injury. Water immersion lenses are used since the refractive index of water (1.33) is close to that of the epidermis (1.34), and this minimizes spherical aberrations caused by the overlying epidermal cell layers when imaging deeper in the dermis the light passes through the tissue–air interface. Water-based gels may be used as immersion media, particularly if imaging a scaly or hyperkeratotic lesion since the gel settles between disrupted corneocytes, reducing irregularities in refraction.7 Gel is also useful if imaging a skin site that is not particularly flat, since the gel does not run off of the skin in the same way as water can. A skin contact device is used to reduce motion artifact and contain the water or gel interface when imaging.7 This
In Vivo Reflectance Mode Confocal Microscopy in Clinical and Surgical Dermatology
device consists of a metal ring that is fixed to the patient’s skin with adhesive, and is coupled to the microscope housing with a magnet during imaging. It has a concave shape to hold the immersion medium. By moving the objective lens in the z (vertical) direction with respect to the skin, it is possible to image at different horizontal planes in depth within the tissue since the focal plane is progressively moved deeper. Images can be captured to produce static or time-sequenced (digitized video) images of horizontal skin sections, as well as recorded on analog videotape (20 to 30 Hz) to produce movies to demonstrate dynamic events such as blood flow.7,9
32.3 REFLECTANCE CONFOCAL MICROSCOPY FINDINGS OF NORMAL SKIN In order to identify pathological skin, one must be familiar with the appearance of images from normal skin. Major differences from conventional histology are that the images are oriented horizontal or parallel to skin surface (en face) and they are gray-scale images. The field of view with RCM depends on the microscope, but usually is 250 to 500 μm across. The depth level being imaged can be ascertained by the morphological appearance of tissue or by using a micrometer attached to the z stage of the objective lens. When imaging the skin in real time starting from the surface and progressing deeper, the most superficial images obtained are from the stratum corneum. Superficial images are very bright due to the refractive difference at the interface between the immersion medium (for example, water at 1.33) and the stratum corneum (1.54), which results in a large amount of backscattered light. Low laser illumination power helps to minimize this. The morphological appearance of this stratum is anucleated polygonal corneocytes measuring 10 to 30 μm in size, and grouped in “islands” separated by skin folds, which appear very dark (Figure 32.2a). The next layer (section) seen is the stratum granulosum, consisting of two to four layers of cells, each cell measuring 25 to 35 μm in size. Here the nuclei can be appreciated as dark central ovals within the cell, surrounded by bright grainy cytoplasm (Figure 32.2b). The stratum spinosum is located at 20 to 100 μm below the top surface. This consists of a tight honeycomb pattern of smaller cells, each cell measuring 15 to 25 μm in size, with well-demarcated cell borders (Figure 32.2c). The deepest layer of the epidermis, the basal layer, is seen as bright clusters of cells, each cell measuring about 7 to 10 μm. When imaging a little deeper, the suprapapillary epidermal plate at the dermoepidermal junction is apparent as round or oval rings of bright basal cells surrounding dark dermal papillae, which often show a central area of blood flow consistent with papillary dermal vascular loops
269
A B
C D E F
A2
A1
B1
B2
C1
C2 p
p
p H
p
D1
D2
E
F
FIGURE 32.2 Correlation between vertical and transverse sections (H&E and in vivo confocal) of normal skin. In confocal images, the corneocytes appear as bright polygonal shapes (A2, arrows) and are 10 to 30 μm in size. Granular cells (B2, arrows) are regularly seen at depths of 10 to 15 μm. The dark oval areas correspond to nuclei within the bright cytoplasm. Spinous keratinocytes (C2, arrows) are seen at 20 to 100 μm below the stratum corneum. Note that basal keratinocytes (D2, arrows), located around of a dermal papillae (p, D), are brighter than keratinocytes of spinosum (ss, C2) and granulosum (sg, B2) layers. Blood vessels (arrows, E) and collagen bundles (arrows, F) are also seen (A1, A2, B1, C1, D1, 40X/0.65 NA dry objective lens; B2, C2, D2, E, F, 30X/0.9 NA water immersion objective lens; scale bar, 25 μm). (From Rajadhyaksha, M. et al., J. Invest. Dermatol., 113, 293–303, 1999. Reprinted by permission of Blackwell Science, Inc.)
(Figure 32.2d). Therefore, the superficial (papillary) dermis can be seen to consist of a network of reticulated fibers
270
Handbook of Non-Invasive Methods and the Skin, Second Edition
and small blood vessels. Other features that can be observed in normal skin include eccrine ducts, which appear as bright, centrally hollow structures that spiral through the epidermis and dermis, and hair shafts with pilosebaceous units. The latter appear as whorled, centrally hollow structures with elliptical elongated cells at the circumference and a central refractile long hair shaft. The appearance of normal skin varies depending on the site and skin color being imaged.12 For example, healthy skin from sun-exposed or darkly pigmented areas appears generally brighter because of what appears to be more pigment at the basilar layer. Chronically sunexposed skin also demonstrates a thicker and more fissured or wrinkled stratum corneum, more randomly arranged and irregularly shaped dermal papillae, and clumping of the dermal reticulated pattern, consistent with collagen and elastic fibers. Variations in numerical density of keratinocytes are also apparent, being greater in sunprotected sites. The palms and soles are characterized for having an extremely thick stratum corneum and a greater number of eccrine ducts.
32.4 REFLECTANCE CONFOCAL MICROSCOPY FINDINGS OF INFLAMMATORY SKIN LESIONS AND CUTANEOUS INFECTIONS
32.4.1 ACUTE CONTACT DERMATITIS Typical features of contact dermatitis such as spongiosis, microvesicle formation, inflammatory infiltrate, and patchy epidermal necrosis have been visualized in realtime using RCM.13,14 Importantly, several features that may enable differentiation of acute irritant vs. allergic contact dermatitis have been observed (Figure 32.3). In this sense, sodium lauryl sulfate-induced irritant contact dermatitis showed greater stratum corneum disruption, with gaps in the surface microtopography, more parakeratosis, clear demarcation and separation of individual corneocytes, and superficial inflammatory infiltrate. These superficial changes developed 24 hours after patch testing, and were sometimes seen even where no clinical response was appreciated. These findings suggest that RCM could be a good method to provide a valid assessment of the stratum corneum barrier.15 RCM has been shown to be of value in the study and differentiation of irritant and allergic contact dermatitis. Furthermore, RCM has been used to evaluate ethnic differences in acute irritant contact dermatitis induced by sodium lauryl sulfate14 and Ivory soap,16 demonstrating a more resistant skin in black individuals, and to assess the kinetic changes in acute allergic contact dermatitis.17 ACD RCM
Table 32.1 illustrates main skin processes evaluated by RCM. The great advantage of this technique is that it allows real-time appreciation of dynamic processes, like inflammatory skin disorders, and very importantly, the evolution of responses to therapy. With time, RCM will be potentially useful for non-invasive diagnosis.
TABLE 32.1 Skin Processes Studied by RCM Inflammatory dermatoses Acute contact dermatitis Psoriasis Cutaneous infections Dermatophytoses Warts Folliculitis Herpes simplex Cutaneous neoplasms Nonpigmented Actinic keratosis Basal cell carcinoma Pigmented Melanocitic nevi Malignant melanoma
ICD RCM
H&E
ICD
a ACD
b ACD
c
FIGURE 32.3 Features common to ACD and ICD observed with RCM and correlated by routine histology. (a) Spongiosis: Increased intercellular brightness apparent on RCM. (b) Inflammatory cell infiltrate: Bright structures 12 to 15 μm in size interspersed between keratinocytes. Arrows denote inflammatory cells. (c) Intraepidermal vesicle formation: Dark spaces in the epidermis containing inflammatory cells and necrotic keratinocytes (arrows denote vesicles). Scale bars = 50 μm. (From Swindells, K. et al., J. Am. Acad. Dermatol., 50, 220–28, 2004. Reprinted by permission of Mosby, Inc.)
In Vivo Reflectance Mode Confocal Microscopy in Clinical and Surgical Dermatology
271
32.4.2 PSORIASIS
32.5.1 NONPIGMENTED LESIONS
The RCM features of stable psoriasis vulgaris have been characterized and correlated with routine histology.18 RCM demonstrates the typical histological features of psoriasis, including parakeratosis, Munro microabcesses, acanthosis, capillary dilatation, and papillomatosis. It was also possible to define histologically the boundary or edge of a lesion.19 This information may be very useful in the future to allow not only for use in modulating therapy, but also for understanding pathogenesis of this often intractable disease.
32.5.1.1 Actinic Keratosis and Squamous Cell Carcinoma
32.4.3 CUTANEOUS INFECTIONS Dermatophyte infections, including onychomycosis and tinea pedis, although common, can be difficult to diagnose and, if culture is necessary, may be delayed because of the length of time required by the hiphae to growth. RCM enables rapid real-time identification of branched hyphae and inflammatory infiltrate in vivo or in vitro on nail clippings or scrapings.20,21 Folliculitis imaged with RCM can be conclusively diagnosed by direct demonstration of intraepidermal pustules, inflammatory infiltrate, spongiosis, and capillary dilatation.22 Warts have also been imaged with RCM. The hyperkeratotic stratum corneum and the presence of multiple highly refractile round structures measuring 20 to 40 μm in size within the lesion allow a rapid and conclusive diagnosis of the common wart. These typical round structures may be keratohyaline granules or perhaps viral particles within infected keratinocytes, based on their size (unpublished observation). Cutaneous herpes infections may be atypical and severe, especially in an immunocompromised host. RCM has been probed to be a useful tool in their diagnoses. The presence of pleomorphic ballooned keratinocytes and multinucleated giant cells in a loose aggregate of keratinocytes, inflammatory cells and debris is the most important finding.23
32.5 REFLECTANCE CONFOCAL MICROSCOPY FINDINGS OF NEOPLASTIC SKIN LESIONS RCM characterization of neoplastic lesions is an important area for research, with the potential to aid in the noninvasive diagnosis and management of a variety of skin cancers. With the advent of newer, less invasive or topical therapies, it is desirable to use a non-invasive diagnostic tool that can allow high-resolution accurate identification of tumor subtypes and tumor margins, and response to treatment.
We have demonstrated the key histopathological features of actinic keratoses using RCM24 (Figure 32.4). These include architectural disarray, epidermal nuclear enlargement with pleomorphism, and parakeratosis, all resulting in orderly chaos. Depth of penetration currently imposes a major limitation on RCM in the diagnosis of actinic keratosis, especially when dealing with hyperkeratotic lesions that prevent accurate assessment of the dermoepidermal junction and detection of superficially invasive squamous cell carcinoma (SCC). When not hampered by a parakeratotic, hyperkeratotic stratum corneum, observation of full thickness dyplastic features on RCM is suggestive of SCC. Other changes suggesting SCC, such as vascular patterns and appearance of keratin pearls, are being evaluated. Current studies are under way to evaluate
*
A1
A1
B1
B2
C1
C2
D1
D2
A B C D
FIGURE 32.4 Vertical H&E section of actinic keratosis. Lines represent depths in the epidermis corresponding to horizontal sections. Left column corresponds to en face conventional histopathology, while right column corresponds to confocal images. (A) Stratum corneum showing irregular hyperkeratosis. (B) Stratum granulosum demonstrates uniform, evenly spaced, broad keratinocytes. (C) Stratum spinosum shows enlarged, pleomorphic nuclei with haphazard orientation. (D) Stratum basale shows enlarged, pleomorphic nuclei with haphazard orientation. Dermal papillae appear as well-demarcated dark holes in epidermis (arrow), containing blood vessels. 30×, 0.9 NA water immersion objective lens; scale bar = 25 μm. (From Agasshi, D. et al., J. Am. Acad. Dermatol., 43, 42–48, 2000. Reprinted by permission of Mosby, Inc.)
272
Handbook of Non-Invasive Methods and the Skin, Second Edition
the response of actinic keratoses to both photodynamic therapy25 and the topical imiquimod. 32.5.1.2 Basal Cell Carcinoma RCM characteristics of basal cell carcinoma (BCC), the most common skin tumor in humans, have been well defined.26 The presence of islands of monomorphic tumor cells that are elongated in shape with characteristic elongated nuclei oriented along the same axis, producing a polarized appearance (Figure 32.5), seems to be characteristic. This uniform polarized cell pattern persists through the thickness of the epidermis, losing the normal honeycomb pattern, and the dermal papillae architecture. Abundant blood vessels demonstrating prominent tortuosity as well as prominent, predominantly mononuclear inflammatory infiltrate admixed or in close apposition with basal cell carcinoma cells. Trafficking of leukocytes can be visualized in real time.9 We have recently com-
A
x
y
pleted a large retrospective multicenter study that has shown significant accuracy of RCM features for diagnosing BCC in vivo.27 In this regard, RCM was able to realtime detect the presence of residual or clinically equivocal BCC.28 Moreover, the effectiveness of topical imiquimod in the treatment of this tumor has been confirmed by RCM.29,30
32.5.2 PIGMENTED LESIONS 32.5.2.1 Melanocytic Nevi The large amount of melanin present in melanocytic lesions makes them ideal for RCM imaging and diagnosis. RCM features of common melanocytic nevi include the presence of small monomorphous round or oval brightly refractile cells within which nuclei, if visualized, are centrally positioned.31,32 In junctional nevi these cells are seen within the epidermis at the dermoepidermal junction, typically surrounding dermal papillae (Figure 32.6A). On the other hand, cells consistent with nevomelanocytes may be seen at the dermoepidermal junction, but also in the superficial dermis in compound nevi (Figure 32.6B). These are often grouped in rounded clusters (nests) containing several cells, often near blood vessels (Figure 32.6B). In both these nevus types, the melanocytes arranged in small nests or as single cells may also be seen higher in the epidermis, but the architecture of the stratum corneum, granulosum, spinosum, and basal cell layer remains otherwise unchanged and as previously described for normal epidermis. In contrast, dysplastic lesions show focal loss of the cell–cell keratinocyte borders at the dermoepidermal junction. Fine bright granules within the epidermis that probably represent melanin bodies are characteristically seen (Figure 32.6C). Also, these dysplastic lesions show a greater variety in nevomelanocyte size and shape, though cells still tend to be rounded or oval rather than dendritic (Figure 32.6D).
B B
32.5.2.2 Malignant Melanoma
A
C
bv
C
FIGURE 32.5 Confocal images of BCC, superficial type. (A) The en face section of the overlying epidermis to the tumor showing actinic epidermal damage in the form of pleomorphism. (B) The en face section of the tumor showing uniform population of elongated nuclei (arrow) (nuclear polarization). (C) The en face section at dermal level showing inflammatory cells (arrows) and dilated blood vessels (bv) (30×, 0.9 NA water immersion objective lens; width, 250 μm).
RCM demonstrates the presence of pleomorphic bright cells within the epidermis and dermis.26 These cells may be stellate in shape and possess coarse branching dendritic processes and eccentrically placed large nuclei (Figure 32.7).31,33,34 Additionally, the regular honeycombing architectural pattern of the stratum spinosum is disrupted with indistinct cell borders and bright grainy particles, probably melanin, distributed within the epidermis. Based on our preliminary studies, the successful distinction between melanoma and nevi appears to be quite easy to accomplish. Moreover, even intraepidermal melanoma can be recognized by these techniques, using the same criteria as established for conventional histology. Confocal images of these melanomas showed an increased number of intraepidermal enlarged (atypical) melanocytes in solitary
In Vivo Reflectance Mode Confocal Microscopy in Clinical and Surgical Dermatology
273
*
* * A1
A1
A2
A2
* M BV
c
N B1
*
A4
A3
B2
*
*
* * C1
C2
B1
B2
* *
* D1
B3
B4
D2
FIGURE 32.6 Confocal images of benign and dysplastic nevi. (A1) The histologic section of a melanocytic nevus showing melanocytes and keratinocytes in the basal layer. (A2) The en face section of a benign nevus showing a uniform pattern of bright monomorphic cells (arrows) and pigmented basal keratinocytes. (B1) The histologic section of a compound nevus demonstrating blood vessels (BV), nevomelanocytes (N), and melanophages (M). (B2) The en face section of the same compound nevus showing nests of bright monomorphous cells (arrows). The dermal location of these nests can be determined by identification of the blood vessels (labeled c). (C1) The histologic section of a lentiginous compound dysplastic nevus. Note the nest of novemelanocytes at the dermoepidermal junction showing uniform cytoplasm and nuclei but variation in size. (C2) Nevomelanocytes that can be identified by their bright, highly refractive nature (arrows). In comparison to common melanocytic nevi, the displastic nevi have more variation in their size and less brigthness. Asterisks indicate focal loss of keratinocyte borders. Melanin dusts can also be noted (arrowheads). (D1) The histologic section of a lentiginous compound dysplastic nevus showing the dermal component. Note the variation in melanin content. (D2) The nest of nevomelanocytes in the papillary dermis (arrows). A1, B1, C1, and D1, H&E-stained sections; scale bar, 50 μm. A2, B2, C2, and D2, in vivo confocal images, 100×, 1.2 NA water immersion objective lens; scale bar, 25 μm. (From Langley, R.G.B. et al., J. Am. Acad. Dermatol., 45, 365–376, 2001. Reprinted by permission of Mosby, Inc.)
FIGURE 32.7 Confocal images of melanoma. (A1 to A3) Confocal images of a superficial spreading malignant melanoma. Loss of keratinocyte border (asterisks in A1 to A3), bright granular highly refractile particles (arrowheads in A1), and atypical stellate cells (arrows A1 and A2) are seen. (A4) Histologic section of intraepidermal component of superficially spreading malignant melanoma showing confluence of tumor cells at the dermoepidermal junction as well as individual cells of varying levels of the epidermis, so-called pagetoid spread (asterisks). (B1 to B3) Confocal images of a melanoma in situ. Focal disappearance of keratinocyte cell borders (asterisks in B1 to B3), presence of individual polymorphic cells (arrows in B1 and B2), and fine granular particules (arrowheads in B1 and B2) can be identified. Note the eccentric melanin staining (arrowheads in B3). (B4) Histologic section of malignant melanoma in situ, superficial spreading type. Note the confluent nest of markedly atypical tumor cells with varying quantities of melanin in their cytoplasm. A1 to A3 and B1 to B3, in vivo confocal images, 100×, 1.2 NA water immersion objective lens; scale bar, 25 μm. A4 and B4, H&E-stained sections; scale bar, 50 μm. (From Langley, R.G.B. et al., J. Am. Acad. Dermatol., 45, 365–376, 2001. Reprinted by permission of Mosby, Inc.)
274
Handbook of Non-Invasive Methods and the Skin, Second Edition
units at all layers of the epidermis, including the upper spinous and granular cell layers.35 Interestingly, some features of melanomas can also be identified in amelanotic melanomas using RCM. In two cases reported in the literature and several others subsequent studies, RCM allowed recognition of an abnormal intraepidermal melanocytic proliferation that was distinctly different from normal skin,36 presumably because of the presence of some melanin in premelanosomes.33 These RCM features correlated well with conventional histology. It remains to be seen whether malignant cells with small amounts of pigment can be accurately identified. Further studies are being performed to clarify this.
32.6 USE OF REFLECTANCE CONFOCAL SCANNING LASER MICROSCOPY AS ADJUNCT TO STANDARD THERAPY 32.6.1 USE OF CONFOCAL MICROSCOPY MARGIN ASSESSMENT
FOR
The ability of RCM to perform non-invasive evaluation of skin lesions means that it has the potential to define lesion margins before surgical or nonsurgical therapy. This could be particularly helpful in margin assessment of tumors with radial growth phases, including lentigo maligna melanomas or some basal cell carcinomas, or tumors that are difficult to see clinically, such as amelanotic melanomas33–36 or sclerosing infiltrative basal cell carcinomas. Limiting factors to date have been the depth of penetration, which prevents accurate imaging below the superficial dermis, and the presence of inflammatory cells’ refractivity, among others. Lack of contrast can also pose a problem, which might be solved by the future development of exogenous contrast agents, like aluminum chloride.37 Despite present limitations, RCM can help to identify atypical areas that need biopsy to provide histological confirmation. Furthermore, RCM may aid in the rapid establishment of tumor margins by examining excised specimens during procedures such as Mohs’ micrographic surgery.37–39
32.6.2 USE OF CONFOCAL MICROSCOPY FOR EVALUATING RESPONSE TO TREATMENT To assess histological changes that occur after new treatments, a biopsy would normally be required. The ability to use RCM to image in real time without discomfort to the patient makes it a powerful tool to assist in the responses to treatments.10 There are ongoing RCM studies evaluating the histopathological responses of actinic keratoses treated with 5-aminolevulanic acid photodynamic therapy, as well as those of basal cell carcinomas treated with imiquimod. To date, RCM has proven to be accurate
in establishing the presence of BCC and its responsiveness to the treatment regimen with imiquimod.30 RCM has demonstrated to be a sensitive tool for characterizing and quantifying histological changes of the epidermis and papillary dermis due to aging.40 Therefore, it is a promising method to evaluate the cutaneous response to new antiaging therapies.
32.7 DIFFICULTIES AND POTENTIAL SOLUTIONS Although we have discussed several of the benefits of using RCM, there are some limitations. The current technology is complex and expensive, and thus not widely available to researchers and clinicians. However, several groups in both academia and industry are developing simpler and less expensive scanning methods that should result in smaller, handheld, including endoscope-like, confocal reflectance microscopes. Problems related to imaging include a depth of penetration limited to the superficial dermis, or even less when a lesion is particularly hyperkeratotic, mainly due to scattering and spherical aberration. This may be improved by experimenting with different immersion media of suitable refractive indices, illumination wavelengths, and aberration correction methods. The current microscopic-to-skin contact device is unwieldy and difficult to use on awkward or nonflat anatomical sites. Better devices must be designed to enable efficient contact with various anatomical sites. Work is ongoing to facilitate the creation of vertically oriented sections that would be similar to those seen in histology, which would vastly increase the potential of RCM. The utility of oblique sections (that maintain adequate resolution and pixilation), as a reasonable representation of vertical sections, is investigated. A combination of an en face and oblique section will better exploit the three-dimensional imaging capability of RCM. A significant scientific challenge is image understanding — the ability to read, interpret, and analyze images to extract useful clinical and histologic information. Several groups worldwide are performing detailed studies to characterize confocal images and correlate to histology. This should lead to sensitivity and specificity studies toward potential clinical and surgical utility for RCM.
32.8 SUMMARY RCM offers tremendous potential applications for the advancement of medical research and clinical care (Table 32.2). In research it can be used for the study of normal or pathophysiologic processes in real time non-invasively and by the same technique sequentially over time. Immunologic events previously only studied ex vivo or by static images can be traced from their inception to completion.
In Vivo Reflectance Mode Confocal Microscopy in Clinical and Surgical Dermatology
TABLE 32.2 Main Utilities of Reflectance Mode Confocal Microscopy as a Complementary Technique to Conventional Histopathology in Basic and Clinical Dermatology It opens a new way to view and diagnose skin lesions, particularly skin tumors. It may be used as an adjunct to dermatology surgery (margin demarcation, biopsy guide, etc.). It is a useful technique for evaluation of dynamic cutaneous processes (leukocyte trafficking, monitoring of noninvasive treatment of skin cancer, etc.).
The ability of RCM to non-invasively image in real time without any discomfort to the patient makes it a powerful and user-friendly tool to assist in the management of skin diseases. Nevertheless, like early x-ray and ultrasound imaging, RCM is in its infancy. Time and continued persistent research will lead to similar success and utility for RCM.
ACKNOWLEDGMENTS Partially supported by grants from the NIH (NIOSH grant RO1 OH04029) to S.G., from the Whitaker Foundation to M.R., and from the Department of Energy to both M.R. and S.G.
REFERENCES 1. Aguirre AD, Hsiung P, Ko TH, Hartl I, and Fujimoto JG. High-resolution optical coherence microscopy for high-speed in vivo cellular imaging. Opt. Lett. 28, 2064–2066, 2003. 2. Mansotti L. Basic principles and advanced technical aspects of ultrasound imaging. In Physics and Engineering of Medical Imaging, Guzzardi R, Ed. Boston: Martinus Nijhoff Publishers, 1987, pp. 263–317. 3. Markisz JA and Aquilia MG. Technical Magnetic Resonance Imaging. Stanford, CA: Appleton & Lange, 1996. 4. New KC, Petroll WM, Boyde A, Martin L, Corcuff P, Leveque JL, Lemp MA, Cavanagh HD, and Jester JV. In vivo imaging of human teeth and skin using real-time confocal microscopy. Scanning 13, 369–372, 1991. 5. Corcuff P and Leveque JL. In vivo vision of the human skin with the tandem scanning microscope. Dermatology 186, 50–54, 1993. 6. Rajadhyaksha M, Grossman M, Esterowitz D, Webb RH, and Anderson RR. In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J. Invest. Dermatol. 104, 946–952, 1995.
275
7. Rajadhyaksha M, González S, Zavislan J, Anderson RR, and Webb RH. In vivo confocal scanning laser microscopy of human skin. II. Advances in instrumentation and comparison to histology. J. Invest. Dermatol. 113, 293–303, 1999. 8. González S, White WM, Rajadhyaksha M, Anderson RR, and González E. Confocal imaging of sebaceous gland hyperplasia in vivo to assess efficacy and mechanism of pulsed dye laser treatment. Lasers Surg. Med. 25, 8–12, 1999. 9. González S, Sackstein R, Anderson RR, and Rajadhyaksha M. Real-time evidence of in vivo leukocyte trafficking in human skin by reflectance confocal microscopy. J. Invest. Dermatol. 117, 384–386, 2001. 10. Agasshi D, Anderson RR, and González S. Timesequence histologic imaging of laser-treated cherry angiomas using in vivo confocal microscopy. J. Am. Acad. Dermatol. 43, 37–41, 2000. 11. Agasshi D, González E, Anderson RR, Rajadhyaksha RR, and González S. Elucidating the pulsed dye laser treatment of sebaceous hyperplasia in vivo using realtime confocal scanning laser microscopy. J. Am. Acad. Dermatol. 43, 49–53, 2000. 12. Huzaira M, Rius F, Rajadhyaksha M, Anderson RR, and González S. Topographic variations in normal skin histology, as viewed by in vivo reflectance confocal microscopy. J. Invest. Dermatol. 116, 846–852, 2001. 13. González S, González E, White WM, Rajadhyaksha M, and Anderson RR. Allergic contact dermatitis: correlation of in vivo confocal imaging to routine histology. J. Am. Acad. Dermatol. 40, 708–713, 1999. 14. Hicks S, Swindells K, Middelkamp-Hup M, Sifakis MA, González E, and González S. Confocal histopathology of irritant contact dermatitis in vivo and the impact of skin color. J. Am. Acad. Dermatol. 48, 727–734, 2003. 15. Swindells K, Burnett N, Rius-Díaz F, González E, Mihm MC, and González S. Reflectance confocal microscopy may differentiate acute allergic and irritant contact dermatitis in vivo. J. Am. Acad. Dermatol. 50, 220–228, 2004. 16. Burnett N, Astner S, González E, Rius-Díaz F, Doukas A, and González S. Ethnic variability in the skin response induced by Ivory soap. A non-invasive evaluation. Submitted. 17. Astner S, González E, Cheung A, Rius-Díaz F, Doukas A, Farinelli W, and González S. Non-invasive evaluation of the kinetics of allergic and irritant contact dermatitis. J. Invest. Dermatol. 124(2), 351–359, 2005. 18. González S, Rajadhaksha M, Rubinstein G, and Anderson RR. Charaterization of psoriasis in vivo by reflectance confocal microscopy. J. Med. (J. Med. Clin. Exp. Mol.) 30, 337–356, 1999. 19. González S, Rajadhyaksha M, and Anderson RR. Confocal imaging of proliferative cutaneous lesion margen in vivo. J. Invest. Dermatol. 111, 538–539, 1998. 20. Markus R, Huzaira M, Anderson RR, and González S. A better KOH prep? In vivo diagnosis of tinea with confocal microscopy. Arch. Dermatol. 137, 1076–1078, 2001.
276
Handbook of Non-Invasive Methods and the Skin, Second Edition
21. Hongcharu W, Dwyer P, González S, and Anderson RR. Confirmation of onychomychosis by confocal microscopy. J. Am. Acad. Dermatol. 42, 214–216, 2000. 22. González S, Rajadhyaksha M, González-Serva A, White WM, and Anderson RR. Confocal reflectance imaging of folliculitis in vivo. Correlation of confocal imaging to routine histology. J. Cutan. Pathol. 26, 201–205, 1999. 23. Goldgeier M, Fox CA, and Muhlbauer JE. Immediate non-invasive diagnosis of herpes virus by confocal scanning laser microscopy. J. Am. Acad. Dermatol. 46, 783–785, 2002. 24. Agasshi D, Anderson RR, and González S. Confocal laser microscopic imaging of actinic keratoses in vivo: a preliminary report. J. Am. Acad. Dermatol. 43, 42–48, 2000. 25. Trehan M, Swindells K, Taylor CR, Racette AL, and González S. Confocal Microscopy Imaging of Actinic Keratoses Post-Photodynamic Therapy with 5-ALA (abstract). Paper presented at the 20th World Congress of Dermatology, Paris, France, July 1–5, 2002. 26. González S and Tannous Z. Real-time in vivo confocal reflectance microscopy of basal cell carcinoma. J. Am. Acad. Dermatol. 47, 869–874, 2002. 27. Nori S, Rius Diaz F, Cuevas J, Jaen P, Torres A, Goldgeier MH, and González S. Sensitivity and specificity of reflectance confocal microscopy for in vivo diagnosis of basal cell carcinoma: a multicenter study. J. Am. Acad. Dermatol., 51(6), 923–930, 2004. 28. Marra D, Torres A, Schanbacher CF, and González S. Detection of residual basal cell carcinoma by in vivo confocal microscopy. Dermatol. Surg. 31(5), 538–541, 2005. 29. Goldgeier M, Fox CA, Zavislan JM, Harris D, and González S. Noninvasive imaging, treatment, and microscopic confirmation of clearance of basal cell carcinoma. Dermatol. Surg. 29, 205–210, 2003. 30. Torres A, Niemeyer A, Berkes B, Marra D, Schanbacher CF, González S, Owens M, and Morgan B. Imiquimod 5% cream and reflectance-mode confocal microscopy as adjunct modalities to Mohs micrographic surgery for the treatment of basal cell carcinoma. Derm. Surg. 30(12 pt. 1), 1462–1469, 2004. 31. Langley RGB, Rajadhyaksha M, Dwyer PJ, Sober AJ, Flotte TJ, and Anderson RR. Confocal scanning laser microscopy of benign and malignant melanocytic skin lesions in vivo. J. Am. Acad. Dermatol. 45, 365–376, 2001.
32. Busam KJ, Charles C, Lee G, and Halpern AC. Morphologic features of melanocytes, pigmented keratinocytes, and melanophages by in vivo confocal scanning laser microscopy. Mod. Pathol. 14, 862–868, 2001. 33. Busam KJ, Hester K, Charles C, Sachs DL, Antonescu C, González S, and Halpern A. Detection of clinically amelanotic malignant melanoma and assessment of its margins by in vivo confocal scanning laser microscopy. Arch. Dermatol. 137, 923–929, 2001. 34. Tannous Z, Mihnn M, Flotte T, and González S. In vivo examination of lentigo maligna, in situ malignant melanoma, lentigo maligna type by near-infrared reflectance confocal microscopy: comparison of in vivo confocal images with histologic sections. J. Am. Acad. Dermatol. 46, 260–263, 2002. 35. Busam KJ, Charles C, Lohmann CM, Marghoob A, Goldgeier M, and Halpern AC. Detection of intraepidermal malignant melanoma in vivo by confocal scanning laser microscopy. Melanoma Res. 12, 349–355, 2002. 36. Curiel-Lewandrowski C, Williams CM, Swindells KJ, Tahan SR, Astner S, Frankenthaler RA, and González S. Use of in vivo confocal microscopy in malignant melanoma: an aid in diagnosis, and assessment of surgical and non-surgical therapeutic approaches. Arch. Dermatol., 140(9), 1127–1132, 2004. 37. Tannous Z, Torres A, and González S. In-vivo real-time, in-vivo confocal reflectance microscopy, a surgical tool guide for Mohs micrographic surgery facilitated by aluminum chloride, an excellent contrast enhancer. Dermatol. Surg. 29, 839–846, 2003. 38. Rajadhyaksha M, Menaker G, Flotte T, Dwyer P, and González S. Confocal examination of non-melanoma cancers in skin excisions to potentially guide Mohs micrographic surgery without histopathology. J. Invest. Dematol. 117, 1137–1143, 2001. 39. Chung VQ, Dwyer PJ, Nehal KS, Rajadhyaksha M, Menaker GM, Charles C, and Jiang SB. Use of ex vivo confocal scanning laser microscopy during Mohs surgery for non-melanoma skin cancers. Dermatol. Surg. 30(12 pt. 1), 1470–1478, 2004. 40. Sauermann K, Clemann S, Jaspers S, Gambichler T, Altmeyer P, Hoffmann K, and Ennen J. Age related changes of human skin investigated with histometric measurements by confocal laser scanning microscopy in vivo. Skin Res.Technol. 8, 52–56, 2002.
Vivo Reflectance Mode Confocal 33 In Laser Microscopy of Melanocytic Skin Lesions Giovanni Pellacani and Stefania Seidenari Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 33.1 33.2 33.3 33.4
Introduction............................................................................................................................................................277 CSLM Studies on Healthy and Diseased Skin .....................................................................................................277 CSLM Image Acquisition Method for Dermoscopic and Histopathologic Correlation ......................................278 Characteristic CSLM Features of Melanocytic Lesions .......................................................................................279 33.4.1 Superficial Epidermal Layers ....................................................................................................................279 33.4.2 Basal Cell Layer and Dermal-Epidermal Junction ...................................................................................280 33.4.3 Cells Inside Dermal Papillae and Melanocytic Nest Features .................................................................280 33.4.4 Upper Dermis Features..............................................................................................................................282 33.5 Conclusions............................................................................................................................................................282 References .......................................................................................................................................................................282
33.1 INTRODUCTION Confocal scanning laser microscopy (CSLM) represents a novel approach for the in vivo study of the skin, enabling the visualization of the epidermis and superficial dermis at a nearly histologic resolution (Vivascope 1000, Lucid, Inc., Rochester, NY; Optiscan F900, Optiscan Pty. Ltd., Notting Hill, VIC, Australia).1,2 A low-power laser beam illuminates a point inside the object. Reflectance mode CSLM works by detecting a single backscattered photon from the illuminated in-focus section through a pinhole-size filter and rejecting light reflected from out-of-focus portions of the object. A laser beam is then scanned on the horizontal plane producing two-dimensional pictures representing parallel sections of the skin. CSLM enables the non-invasive imaging of skin structures with cellular-level resolution (0.5 to 1.0 μm in the lateral dimension and 4 to 5 μm in the axial ones) to a depth limited to 200 to 300 μm, in relation to the wavelength of the employed laser light, corresponding to the level of papillary dermis in normal skin. In reflectance mode CSLM, contrast is provided by differences in refraction index of organelles and other microstructures that are bright, contrasting with the background. Melanin and melanosomes are a strong source of contrast, rendering melanocytic cells, particularly evident
by means of this technique.3 In fluorescence mode CSLM the laser beam at an appropriate wavelength excites the intradermally injected fluorophore and the emitted fluorescence signal is detected simultaneously.2 Many fluorescent agents employed for the visualization of selectively stained structures in the skin ex vivo and in vitro by means of fluorescence mode CSLM are not suitable candidates for in vivo study.4 Some fluorescent agents, tested on animal models, demonstrated a range of potential clinical applications, although to date these have not been tested on humans,2,5–8 with the exception of fluorescin sodium 0.2% solution.9 Owing to its safety and non-invasiveness, the majority of the reports in literature regard the employment of reflectance mode CSLM.
33.2 CSLM STUDIES ON HEALTHY AND DISEASED SKIN CSLM was applied for imaging physiological and pathological conditions of the skin. The study of healthy skin by means of CSLM and the correlation of the observed features enabled the characterization of the different epidermal layers and papillary dermis structures. The stratum corneum is constituted by large, polygonal, and nonnucleated structures, with a diameter ranging from 10 to 30 μm, 277
278
Handbook of Non-Invasive Methods and the Skin, Second Edition
and a grossly grainy aspect, aggregated in a cobblestone pattern, that corresponds to corneocytes. Going deeper into the epidermis, polygonal cells, with a diameter ranging from 10 to 20 μm, are observable at spinous and granular cell layers, forming a cohesive honeycomb pattern, with bright granular cytoplasm centered by dark oval to round structures corresponding to nuclei. Basal layer keratinocytes form rings of cohesive polygonal cells around dermal papillae. The bright granular material in the cytoplasm of basal keratinocytes is often eccentrically located and oriented away from the papillae, forming a bright “hat” on the nucleus. Dermal papillae are homogeneous nonrefractive structures sometimes containing canalicular structures corresponding to capillaries. A network of coarse bundles of collagen can be observed below the epidermis. Other appendageal structures (hair follicles, eccrine glands) are also visible,1,10 Moreover, site- and age-dependent variations of the confocal aspect of normal skin have been recently reported,11,12 evidencing increased cell brightness and epidermal flattening in sun-exposed sites, and increased epidermal thickness associated with smaller granular and basal cells in aged skin. Since in vivo CSLM provides non-invasive high-resolution imaging of cytoarchitectural details in human skin, it has been explored for use in the diagnosis of a variety of cutaneous diseases.12–18 Recently, the application of CSLM in dermatological oncology has enabled the accurate identification of pigmented basal cell carcinoma19,20 and the possibility of monitoring the efficacy of the treatment in its eradication, either surgically excised or treated with topical therapy, such as immune response modifiers.21,22 Since pigment melanin and melanosomes are among the strongest sources of contrast in CSLM, this technique seems to be particularly indicated for the study of melanocytic lesions. The clear and simple identification of melanocytes, usually round to oval, but sometimes fusiform or dendritic in shape, within the epidermis and superficial dermis may allow the immediate recognition of a pigmented lesion as melanocytic in origin.10 Although preliminary and based on few cases, the analysis of cytological and architectural features of melanocytic lesions by means of CSLM may enable the identification of distinct patterns of benign and malignant melanocytic lesions, as first described by Langley and coworkers.23 Common melanocytic nevi were characterized by small, bright, regular cells located between basal cells at the dermoepidermal junction or clustered into cohesive nests within the papillary dermis. Unlike common nevi, atypical nevi had a more varied nevus cell size, although the cells still appeared monomorphic relative to melanoma. In melanomas, a disarray of the epidermal pattern was observable, leading to the disappearance of the normal honeycomb aspect. Moreover, the presence of cells, polymorphic in size and shape, occasionally with branching dendritic-like
structures, mainly located in the basal layer, but also spreading upward in a pagetoid fashion, and the predominance of single cells over nests were the most striking findings.23,24 Irregular highly refractive cells, sometimes with dendritic-like structures, corresponding to atypical melanocytes, were also observed in one case of melanoma in situ, lentigo maligna type.25 Moreover, identification of characteristic melanoma CSLM features in two cases of amelanotic melanomas makes this technique promising for the in vivo characterization of difficult-to-diagnose lesions.26
33.3 CSLM IMAGE ACQUISITION METHOD FOR DERMOSCOPIC AND HISTOPATHOLOGIC CORRELATION Reflectance confocal imaging by means of the commercially available Vivascope 1000 device (Lucid Inc., Rochester, NY), which employs a diode laser at 830 nm, with power inferior to 35 mW, and a 30× objective lens (numerical aperture of 0.9), is performed attaching an adaptor ring on the skin with a double-sided adhesive tape. After covering the ring hole with a thin layer of water, the arm of the instrument is placed onto the adaptor ring in order to obtain the immobilization of the examined skin area. This procedure enables the examination of an approximately 8 mm2 skin area by moving the lens along horizontal (X- and Y-axes) and vertical planes (Z-axis) with the micrometer screws. A snag in this procedure is the fact that the objective lens moves inside a closed camera; therefore, the corresponding clinical aspect of the portion of tissue or lesion imaged on the monitor is not observable. In order to allow an exact correlation of the observed CSLM features with the corresponding clinical dermoscopic aspects of the pigmented skin lesion examined, it is advisable to previously acquire a 50-fold digital dermoscopic image, positioning the probe of a videomicroscope (VideoCap 100, DS Medica, Milan, Italy) onto the CSLM adaptor ring. Then, by means of CSLM, a sequence of 30 block images, each corresponding to a montage picture of 16 contiguous horizontal images with a 1.9 × 1.4 mm in vivo field of view, is acquired for each lesion at the dermal-epidermal junction level and mounted by a dedicated software to obtain a “reconstructed image” with the same 7.60 × 6.65 mm field of view of the dermoscopic one. Subsequently, the dermoscopic image can be resized and rotated for pointby-point correspondence with the reconstructed confocal image. By means of this procedure, a high-resolution confocal image can be positioned onto the XY plane of the dermoscopic image for exact pattern correlation. Moreover, sequences of confocal sections, starting from the stratum corneum and going into the papillary dermis, are recorded at areas of interest of the lesions employing the
In Vivo Reflectance Mode Confocal Laser Microscopy of Melanocytic Skin Lesions
a
b
279
c
FIGURE 33.1 Superficial layers: (a) honeycombed appearance in a melanocytic nevus; (b) cobblestone appearance in a melanocytic nevus; (c) disarrangement of the normal architecture in a melanoma.
highest instrument resolution (640 × 480 pixels for an effective 475 × 350 μm field of view). A silk suture at one pole of the specimen can be positioned in all excised lesions to make their orientation easier for histopathologic examination.27 Afterwards, a series of equidistant histopathologic sections, starting from the silk suture, enable the correlation of dermoscopic and confocal aspects with histopathology. With this method the aspects of epidermal superficial layers, architecture of the dermal-epidermal junction, morphology and distribution of single cells, presence and features of clusters of melanocytic cells, and structures inside papillary dermis can be described in detail, correlating confocal features with dermoscopic and histopathologic aspects.
33.4 CHARACTERISTIC CSLM FEATURES OF MELANOCYTIC LESIONS
a
b
FIGURE 33.2 Cells infiltrating superficial layers of the epidermis in melanomas: (a) round to oval cells with eccentric dark nucleus and bright cytoplasm; (b) irregularly shaped cells with dendritic-like branches.
TABLE 33.1 Evaluation of the Presence of Large Nucleated Cells within the Spinosum and Granulosum Layers in Melanocytic Lesions
33.4.1 SUPERFICIAL EPIDERMAL LAYERS We have recently evaluated the aspect of superficial layers and the presence and morphology of large, bright nucleated cells within the epidermis on 93 melanocytic lesions belonging to 31 melanomas (15 of which had a Breslow’s thickness inferior to 0.5 mm), 49 melanocytic nevi, and 13 Spitz nevi. As previously reported, usually in melanocytic nevi the stratum spinosum and granulosum were altered with respect to control skin, constituted by large (10- to 30-μm) polygonal cells with dark nuclei and bright cytoplasm and cell borders giving rise to a honeycombed appearance.23 In some benign melanocytic lesions, superficial layers consisted of small polygonal cells with refractive cytoplasm separated by a less refractive border, giving rise to a cobblestone appearance (Figure 33.1). Thin melanomas did not present substantial differences in superficial structures compared with common nevi and Spitz nevi, whereas invasive melanomas presented a disarray of the normal architecture of the superficial layers in 7 of 16 cases, characterized by unevenly distributed bright granular particles and cells, irregular in shape and size (Figure 33.1c). The presence of large cells with bright cytoplasm and dark eccentric nuclei in superficial layers was suggestive of pagetoid infiltration.28 Whereas the observation of
Number of lesions with cells infiltrating superficial layers of the epidermis Cell density: Few Many Cell aspect: Oval Dendritic Both aspects
Melanomas (31)
Melanocytic Nevi (49)
Spitz Nevi (13)
27 (87%)
6 (12%)
4 (31%)
8 19
3 3
3 1
10 1 16
0 0 4
1 5 1
a great amount of large round to oval cells, or of irregularly shaped cells with dendritic-like branches, in pagetoid fashion, was strictly correlated to melanoma diagnosis (Figure 33.2), large melanocytic cells round to oval or dendritic in shape within spinous and granular layers can be occasionally seen in benign lesions (Table 33.1).27,28 Moreover, Spitz nevi may present ovoid homogeneously bright structures in basal and suprabasal layers, correlated with large coarse pigment conglomerates inside the epidermis at histopathology.27
280
Handbook of Non-Invasive Methods and the Skin, Second Edition
33.4.2 BASAL CELL LAYER AND DERMAL-EPIDERMAL JUNCTION By means of CSLM, basal cells appear at a depth of approximately 50 to 100 μm below the stratum corneum, usually brighter than other epidermal cells. Going indepth, dermal papillae appear as dark round to oval areas circumscribed by refractive cells, corresponding to melanocytes and melanin-rich keratinocytes. The aspect and the morphology of dermal papillae and basal keratinocytes seem to fit with the dermoscopic pigment network, enabling the identification of its cytological substrate.29 In common nevi characterized by a typical dermoscopic network, dermal papillae appear as dark round to oval, smallto medium-size areas. In the majority of cases dermal papillae are homogeneously featured. In 20 examined cases, dermal papillae were circumscribed by a rim of refractive cells (edged papillae) (Figure 33.3a), corresponding to small melanocytes and melanin-rich keratinocytes, lacking cytological atypia and appearing as bright rings separated by a thin, structureless, slightly refractive space, sharply contrasting with the dark background (Table 33.2). Refractive rings corresponded to the external darker lines of the pigment network grids, as observed by means of high magnification dermoscopy.30 Moreover, atypical melanocytic nevi, frequently presenting an irregular pigment network at dermoscopy, usually consisting of broad grids and wide meshes, are predominantly characterized by small- to medium-size irregularly shaped dermal papillae. In 11 of 15 atypical nevi edged papillae were observable, whereas the remaining four lesions were characterized by small- to medium-size dermal papillae without a demarcated rim of bright cells (nonedged papillae), but separated by a series of large reflecting cells (Table 33.2).29 In some cases round or oval medium to large cells, with bright cytoplasm and peripheral nuclei, corresponding to a mild cytological atypia, were also observable. On the other hand, melanomas are usually characterized by an irregularly broadened pigment network with wide nonhomogeneous meshes at dermoscopy. Malignant
a
TABLE 33.2 Confocal Microscopy Features of Dermal Papillae Evaluated at the Dermal-Epidermal Junction in Melanocytic Lesions
Papillary contour Edged papillae Nonedged papillae Both
Melanomas (15)
Atypical Nevi (15)
Typical Nevi (20)
1 (7%) 11 (73%) 3 (20%)
11 (73%) 4 (27%) 0 (0%)
20 (100%) 0 (0%) 0 (0%)
lesions presented nonhomogeneous small- to medium-size dermal papillae without a demarcated rim of bright cells (nonedged papillae) (Figure 33.3b) in all cases except one (Table 33.2).29 Nonedged papillae, separated by a series of large bright cells, corresponded to a broadened darkly pigmented network at dermoscopy and to a disarrangement of the dermal-epidermal architecture, characterized by enlarged rete ridges, small dermal papillae, and epidermal flattening at histopathology. Moreover, in approximately two thirds of cases, the confocal reticular architecture, represented by lines of cells and dermal papillae, was occasionally interrupted by stretches of numerous irregularly shaped, highly reflective cells, corresponding to pigment blotches at dermoscopy and to epidermal flattening with basal proliferation of large malignant cells at histopathology. At CSLM examination, melanomas were characterized by larger and more irregular cells than acquired nevi. A marked cellular atypia was a specific marker of malignancy, present in approximately 50% of melanomas, whereas in the remaining malignant lesions a mild atypia was usually present (Figure 33.4).
33.4.3 CELLS INSIDE DERMAL PAPILLAE MELANOCYTIC NEST FEATURES
AND
Inside dermal papillae, separated or clustered reflecting cells are sometimes observable. Single large roundish
b a
FIGURE 33.3 Confocal aspect of (a) dermal papillae edged by a rim of refractive polygonal cells (edged papillae) in a melanocytic nevus, and (b) nonhomogeneous dermal papillae without a demarcated rim of bright cells (nonedged papillae) in an in situ melanoma.
b
FIGURE 33.4 Confocal aspect of (a) marked cytological atypia disarranging dermal-epidermal architecture in a melanoma and (b) mild cellular atypia (white arrowheads) within typical cellular architecture in a melanoma.
In Vivo Reflectance Mode Confocal Laser Microscopy of Melanocytic Skin Lesions
a
b
FIGURE 33.5 Confocal aspect of (a) nucleated cells, corresponding to malignant melanocytes (white arrowheads), infiltrating dermal papilla in an invasive melanoma, and (b) plump bright cells with ill-defined borders, corresponding to melanophages (white arrowheads) in a melanocytic nevus.
refractive cells, with bright cytoplasm and dark eccentric nucleus, corresponded to melanocytic cells infiltrating dermal papillae, whereas irregularly shaped plump bright cells with ill-defined cytoplasmic borders corresponded to melanophages.10,29 Large irregular cells with refractive cytoplasm and eccentric dark nucleus infiltrating dermal papilla are frequently observable in more than 50% of invasive melanomas, whereas they are seldom present in atypical nevi (Figure 33.5). Immediately below the basal cell layer, the presence of aggregates of refractive cells forming oval to roundish structures was reported in melanocytic nevi10,23 and in lentigo maligna.25 Since the description of cytological and architectural features of melanocytic nests is relevant for the histopathologic diagnosis, CSLM examination of cell clusters represents a fundamental step for the in vivo characterization of melanocytic lesions. On reconstructed CSLM images it is possible to evaluate the shape and distribution of cell clusters, whereas high-resolution images are employed for the characterization of single nests. According to their aspect, cellular clusters were divided into three different types (Figure 33.6).31 Compact aggregates of large polygonal cells, with hyporeflecting nuclei and fine granular cytoplasm, forming polyhedral structures, were defined as dense clusters, whereas roundish nonreflecting structures with a well-demarcated border, containing isolated round to oval cells with dark nucleus and reflecting cytoplasm, sometimes having a
a
b
281
multilobate aspect, were defined as sparse cell clusters. Finally, cellular clusters consisting of confluent aggregates of low reflecting polygonal or elongated structures separated by a low reflecting rim, cerebriform in appearance, where cellular nuclei and contours cannot usually be distinguished, were called cerebriform clusters. In melanomas, the distribution of cell clusters was not specific, appearing not only as isolated or focally aggregated nests, but also diffusely distributed on the examined area. Cell clusters in melanomas were frequently constituted by large (maximum diameter greater than 500 μm) cerebriform clusters or by small- to medium-size (maximum diameter inferior to 500 μm) sparse cell clusters, sometimes intercalated with thin fibrillar structures multilobate in appearance (Table 33.3).31 On the other hand, melanocytic nevi were predominantly characterized by numerous small- and medium-size dense clusters frequently distributed on the whole lesion, sometimes forming large central conglomerates (Table 33.3).31 Spitz nevi often presented a peripheral rim of medium-size peripheral clusters, with a dense or sparse cell appearance, sometimes observable also on the whole lesion area (Table 33.3).27,31 By means of CSLM, cellular clusters correlated with dermoscopic pigment globules in the majority of cases, with rare exceptions in heavily pigmented lesions, in which the presence of large dark blotches overshadowed the underlying structures. At histopathologic examination, dense clusters corresponded to well-circumscribed nests comprising large epithelioid cohesive cells with fine dusty melanin, located in the lower epidermis or in the upper dermis. Sparse cell clusters correlated with well-circumscribed nests consisting of unevenly pigmented cells, characterized by a marked cellular discohesion, both in melanomas and in Spitz nevi at histology. Thin refractive fibers inside the sparse cell clusters were observed only in melanomas, resulting in a multilobate appearance by both CSLM and histology, probably due to the confluence of malignant cell aggregates. Corresponding to CSLM cerebriform clusters, the most specific finding for melanoma diagnosis, large confluent nests formed by compact malignant cells containing fine dusty melanin were observable in the upper dermis.31
c
FIGURE 33.6 Confocal aspect of (a) dense clusters in a melanocytic nevus; (b) sparse cell clusters in a melanoma; (c) cerebriform clusters in a melanoma.
282
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 33.3 Frequency of Different Types of Cellular Clusters in Melanocytic Lesions as Observed by Confocal Microscopy Melanomas (10)
Melanocytic Nevi (10)
Spitz Nevi (6)
1 3 1 5
10 0 0 0
5 0 1 0
Dense cluster Sparse cell cluster Dense + sparse cell cluster Cerebriform cluster
33.4.4 UPPER DERMIS FEATURES By means of CSLM, only the upper dermis at a maximum depth of approximately 300 μm can be investigated, owing to the limited penetration of the 830-nm laser beam. Inside the papillary dermis of normal skin, we can distinguish a fine texture of thin refractive fibers representing the distribution of the collagen bundles, and canalicular structures corresponding to blood capillaries, in which small refractive cells flow inside. In melanocytic lesions, besides melanocytic nests or malignant cells infiltrating dermal papilla, the presence of plump, bright cells with ill-defined borders, corresponding to melanophages,10 was also reported and correlated with inflammatory infiltration.10,27 Moreover, in both benign and malignant melanocytic lesions, dilated capillaries with high blood flow within the papillary dermis, sometimes associated with melanocytic nests, are frequently observable.23,27 Whereas no peculiarity of collagen texture was reported in the pigmented component of melanocytic lesions, a fibrillar pattern with thin, irregularly distributed refractive bundles, sometimes intercalated by refractive nucleated cells, in association with the loss of the typical dermal-epidermal profile was noticed inside regression areas in melanomas (unpublished data) (Figure 33.7).
33.5 CONCLUSIONS In spite of the limited number of cases studied so far, CSLM seems to represent an efficient approach for the study of skin physiology and skin diseases. CSLM is an in vivo histologic examination of the skin without surgery, tissue transport, and processing delays, enabling the visualization of skin structures at a cellular-level resolution. Its application includes the study of all skin conditions involving epidermis or superficial dermis. Characterization of CSLM features of melanocytic lesions may improve diagnostic accuracy in dermatological oncology, together with knowledge on their biology. Further studies and systematic statistical analyses are needed to exactly
FIGURE 33.7 Confocal aspect of collagen texture at upper dermis level as observed in a regression area of an invasive melanoma, characterized by thin refractive bundles intercalated by nucleated cells with bright cytoplasm.
determine the improvement in diagnostic accuracy obtained with CSLM. However, the possibility of a strict correlation between confocal images and dermoscopy, as enabled by the CSLM image reconstruction method proposed by us, represents the missing link between dermoscopy and histopathologic examination, allowing the visualization of the cytological and architectural features of difficult-to-interpret clinical dermoscopic patterns.
REFERENCES 1. Rajadhyaksha, M., Gonzalez, S., Zavislan, J.M., Andersson, R.R., and Webb, R.H., In vivo confocal scanning laser microscopy of human skin. II. Advances in instrumentation and comparison with histology, J. Invest. Dermatol., 113, 293, 1999. 2. Delaney, P.M., Harris, M.R., and King, R.G., Novel microscopy using fiber optic confocal imaging and its suitability for subsurface blood vessel imaging in vivo, Clin. Exp. Pharmacol. Physiol., 20, 197, 1993. 3. Rajadhyaksha, M., Grossman, M., Esterowitz, D., Webb, R.H., and Andersson, R.R., In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast, J. Invest. Dermatol., 104, 946, 1995. 4. Veiro, J.A. and Cummins, P.G., Imaging of skin epidermis from various origins using confocal laser scanning microscopy, Dermatology, 189, 16, 1994. 5. Bussau, L.J., Vo, L.T., Delaney, P.M., Papworth, G.D., Barkla, D.H., and King, R.G., Fibre optic confocal imaging (FOCI) of keratinocytes, blood vessels and nerves in hairless mouse skin in vivo, J. Anat., 192, 187, 1998. 6. Papworth, G.D., Delaney, P.M., Bussau, L.J., Vo, L.T., and King, R.G., In vivo fibre optic confocal imaging of microvasculature and nerves in the rat vas deferens and colon, J. Anat., 192, 489, 1998.
In Vivo Reflectance Mode Confocal Laser Microscopy of Melanocytic Skin Lesions
7. Vo, L.T., Papworth, G.D., Delaney, P.M., Barkla, D.H., and King, R.G., In vivo mapping of the vascular changes in skin burns of anaesthetised mice by fibre optic confocal imaging (FOCI), J. Dermatol. Sci., 23, 46, 2000. 8. Anikijenko, P., Vo, L.T., Murr, E.R., Carrasco, J., McLaren, W.J., Chen, Q., Thomas, S.G., Delaney, P.M., and King, R.G., In vivo detection of small subsurface melanomas in athymic mice using noninvasive fiber optic confocal imaging, J. Invest. Dermatol., 117, 1442, 2001. 9. Swindle, L.D., Thomas, S.G., Freeman, M., and Delaney, P.M., View of normal human skin in vivo as observed using fluorescent fiber-optic confocal microscopic imaging, J. Invest. Dermatol., 121, 706, 2003. 10. Busam, K., Charles, C., Lee, G., and Halpern, A.C., Morphological features of melanocytes, pigmented keratinocytes, and melanophages by in vivo confocal scanning laser microscopy, Mod. Pathol., 14, 862, 2001. 11. Huzaira, M., Rius, F., Rajadhyaksha, M., Andersson, R.R., and Gonzales, S., Topographic variations in normal skin, as viewed by in vivo reflectance confocal microscopy, J. Invest. Dermatol., 116, 864, 2001. 12. Gonzalez, S., Gonzalez, E., White, M., Rajadhyaksha, M., and Andersson, R.R., Allergic contact dermatitis: correlation of in vivo confocal imaging to routine histology, J. Am. Acad. Dermatol., 40, 708, 1999. 13. Gonzalez, S., Rajadhyaksha, M., Gonzalez-Serva, A., White, M., and Andersson, R.R., Confocal reflectance imaging of folliculitis in vivo: correlation with routine histology, J. Cutan. Pathol., 26, 201, 1999. 14. Gonzalez, S., Rubinstein, G., Mordovsteva, V., Rajadhyaksha, M., and Andersson, R.R., In vivo abnormal keratinization in Darier-white’s disease as viewed by realtime confocal imaging, J. Cutan. Pathol., 26, 504, 1999. 15. Gonzalez, S., Rajadhyaksha, M., Rubinstein, G., and Andersson, R.R., Characterization of psoriasis in vivo by reflectance confocal microscopy, J. Med., 30, 337, 1999. 16. Aghassi, D., Andersson, R.R., and Gonzalez, S., Confocal laser microscopic imaging of actinic keratoses in vivo: a preliminary report, J. Am. Acad. Dermatol., 43, 42, 2000. 17. Tachihara, R., Choi, C., Langley, R.G.B., Andersson, R.R., and Gonzalez, S., In vivo confocal imaging of pigmented eccrine poroma, Dermatology, 204, 185, 2002. 18. Hicks, S.P., Swindells, K.J., Middelkamp-Hup, M.A., Sifakis, M.A., Gonzalez, E., and Gonzalez, S., Confocal histopathology of irritant contact dermatitis in vivo and the impact of skin color (black vs. white), J. Am. Acad. Dermatol., 48, 727, 2003. 19. Gonzalez, S. and Tannaus, Z., Real time, in vivo confocal reflectance microscopy of basal cell carcinoma, J. Am. Acad. Dermatol., 47, 869, 2002. 20. Charles, C.A., Marghoob, A.A., Busam, K.J., ClarkLoeser, L., and Halpern, A.C., Melanoma or pigmented basal cell carcinoma: a clinical-pathologic correlation with dermoscopy, in vivo confocal scanning laser microscopy, and routine histology, Skin. Res. Technol., 8, 282, 2002.
283
21. Goldgeier, M., Fox, C.A., Zavislan, J.M., Harris, D., and Gonzalez, S., Noninvasive imaging, treatment, and microscopic confirmation of clearance of basal cell carcinoma, Dermatol. Surg., 29, 205, 2003. 22. Tannous, Z., Torres, A., and Gonzalez, S., In vivo realtime confocal reflectance microscopy: a noninvasive guide for Mohs micrographic surgery facilitated by aluminum chloride, an excellent contrast enhancer, Dermatol. Surg., 29, 839, 2003. 23. Langley, R.G.B., Rajadhyaksha, M., Dwyer, P.J., Sober, A.J., Flotte, T.J., and Andersson, R.R., Confocal scanning laser microscopy of benign and malignant melanocytic skin lesions in vivo, J. Am. Acad. Dermatol., 45, 365, 2001. 24. Busam, K.J., Charles, C., Lohmann, C.M., Marghoob, A., Goldgeier, M., and Halpern, A.C., Detection of intraepidermal malignant melanoma in vivo by confocal scanning laser microscopy, Melanoma Res., 12, 349, 2002. 25. Tannous, Z.S., Mihm, M.C., Flotte, T.J., and Gonzalez, S., In vivo examination of lentigo maligna and malignant melanoma in situ, lentigo maligna type by near-infrared reflectance confocal microscopy: comparison of in vivo confocal images with histologic sections, J. Am. Acad. Dermatol., 46, 260, 2002. 26. Busam, K.J., Hester, K., Charles, C., Sachs, D.L., Antonescu, C.R., Gonzalez, S., and Halpern, A.C., Detection of clinically amelanotic malignant melanoma and assessment of its margins by in vivo confocal scanning laser microscopy, Arch. Dermatol., 137, 923, 2001. 27. Pellacani, G., Cesinaro, A.M., Grana, C., and Seidenari, S., In-vivo confocal scanning laser microscopy of pigmented Spitz nevi. Comparison of in-vivo confocal images with dermoscopy and routine histopathology, J. Am. Acad. Dermatol., 51, 371, 2004. 28. Pellacani, G., Cesinaro, A.M., and Seidenari, S., Reflectance-mode confocal microscopy for the in vivo characterization of superficial epidermal layers in benign and malignant melanocytic lesions, pagetoid melanocytosis in melanomas and nevi, J. Invest. Dermatol., 125, 532, 2005. 29. Pellacani, G., Cesinaro, A.M., Longo, C., Grana, C., and Seidenari, S., Microscopic in vivo description of cellular architecture of dermoscopic pigment network in nevi and melanomas, Arch. Dermatol., 141, 147, 2005. 30. Puppin, D., Salomon, D., and Saurat, J.H., Amplified surface microscopy. Preliminary evaluation of a 400fold magnification in the surface microscopy of cutaneous melanocytic lesions, J. Am. Acad. Dermatol., 28, 923, 1993. 31. Pellacani, G., Cesinaro, A.M., and Seidenari, S., In-vivo confocal reflectance microscopy for the characterization of melanocytic nests and correlation with dermoscopy and histology, Br. J. Dermatol., 152, 384, 2005.
Vivo Confocal Microscopy of the 34 In Skin Surface Using Fluorescent Markers Steven G. Thomas Optiscan Pty. Ltd., Notting Hill, Victoria, Australia
CONTENTS 34.1 34.2 34.3 34.4
Introduction............................................................................................................................................................285 Applications of In Vivo Fluorescence Confocal Microscopy ...............................................................................286 Fluorophore Selection............................................................................................................................................286 Methods/Routes of Administration .......................................................................................................................287 34.4.1 Topical Application....................................................................................................................................287 34.4.2 Intradermal Administration........................................................................................................................287 34.4.3 Intravenous Administration........................................................................................................................290 34.5 In Vivo Immunofluorescence Labeling..................................................................................................................291 34.6 Functional Imaging................................................................................................................................................291 34.6.1 Skin Barrier Function ................................................................................................................................291 34.6.2 Transdermal Drug Delivery.......................................................................................................................291 34.6.3 GFP Expression In Vivo ............................................................................................................................292 34.6.4 Thermal Burn-Induced Autofluorescence .................................................................................................292 34.7 Future Directions ...................................................................................................................................................292 References .......................................................................................................................................................................293
34.1 INTRODUCTION Confocal microscopy allows the visualization of subsurface microscopic structures in thick translucent tissue specimens by optically rejecting out-of-focus information. Only light returning from the exact plane of focus in the tissue is permitted to pass through to the detector, thus optically isolating the focal plane from the surrounding structures.1 This enables the focal plane to be positioned below the surface of the specimen to obtain subsurface images. The ability to optically section intact tissue without the need to physically section it allows the physical structure to be preserved, and enables tissues and dynamic tissue events to be studied in vivo in real time.2–4 The optical sections obtained from a confocal microscope are parallel to the tissue surface (horizontal or en face sections). This is in contrast to conventional histological sections that are traditionally produced by taking vertical sections from paraffin-embedded and chemically stained skin samples. As confocal microscopy is noninvasive, the
same tissue site may be imaged repeatedly to monitor tissue changes over a period, such as tissue growth, wound healing, lesion progression, or effect of treatments.5,6 The mechanisms and features of different confocal instruments specifically designed for in vivo confocal microscopy of skin are detailed by Rajadhyaksha et al.7 and Suihko et al.8 Fluorescence confocal microscopy relies on the differential distribution of fluorescent molecules within the tissue to produce contrast.9 Fluorescence is the result of a three-stage process that occurs in certain molecules called fluorophores or fluorescent dyes. In confocal microscopy, a laser light source is used to illuminate the tissue.10 When a photon from the laser is absorbed by a fluorophore molecule, that molecule is excited to an electronic singlet state. The excited state exists for a few nanoseconds, during which time some of the energy is dissipated, resulting in a relaxed singlet excited state. A photon is emitted when the fluorophore returns to its ground state. Due to the loss of energy during the excited-state lifetime, the energy of the emitted photon is lower than the excitation photon, 285
286
Handbook of Non-Invasive Methods and the Skin, Second Edition
and therefore a longer wavelength. The difference in energy (or wavelength) between the excitation photon and the emitted photon is referred to as the Stokes shift. This wavelength shift allows the emitted fluorescence to be separated from the excitation photons, allowing detection against a low background. The fluorescence process is cyclical, and the same fluorophore can be repeatedly excited and detected, unless the fluorophore is irreversibly destroyed while in the excited state. This destruction of the fluorophore results in a loss of fluorescence known as photobleaching.
34.2 APPLICATIONS OF IN VIVO FLUORESCENCE CONFOCAL MICROSCOPY Through the selection of appropriate exogenous fluorophores, in vivo fluorescence confocal microscopy offers a unique view of the skin in vivo. The application of fluorophores to the tissue can provide a variety of results, depending on the fluorophore chosen and its method of application. This has allowed the development of new and highly specific techniques to probe tissues using fluorescent markers to study the structure, function, and dynamics of the skin in vivo.11 In vivo fluorescence confocal microscopy studies using animal models have demonstrated a range of potential in vivo clinical applications.12 These include characterization of epidermal layers and subcellular structure of skin;13 mucous crypt architecture and subcellular detail of colonic mucosa;14,15 blood flow in superficial vasculature of the skin,4,9,13 brain,16–20 and colon;9,14,21,22 microvasculature changes in skin pathologies,11,23 angiogenesis,24 and thermal injury;25–27 arterial thrombus formation in mice;28 innervation of skin13 and vas deferens,4,29 epithelial desquamation30 and wound healing31 of cornea in rabbits; structure and function of renal tubules in rat kidney;32 mouse peritoneum;33,34 use of fluorescently labeled antibodies to detect human melanoma xenografts on nude mice11,23,25,36 and cholinergic neurons in brain;37 penetration of substances into skin;13,38 bone remodeling;39–41 and detection of green fluorescent protein (GFP) expression in the brain42 and epidermis43 of transgenic animals. Recently, some of these techniques have been extended into clinical studies where human tissues have been imaged in vivo. These include studies of the epidermis and stratum corneum of normal skin44–46 and melanocytic lesions,47,48 penetration of substances into skin,49,50 microvascularization in human choroidal melanomas,51 retinal angiography,52 detection of cervical neoplasia,53 and endoscopic examination of the gastrointestinal mucosal architecture in both normal and diseased tissue.54–56 These studies have employed a range of fluorescent contrast agents, each selected according to the
instrument used to perform the study and the biological application.
34.3 FLUOROPHORE SELECTION Fluorescence contrast can provide a highly sensitive and specific discrimination of cellular microstructure through the appropriate choice of fluorophore, from both a biological and optical/instrument point of view. The biological compatibility and specificity of a fluorophore can significantly affect the choice of a fluorophore for in vivo imaging, as can the information desired from the tissue. The primary considerations for selecting a suitable fluorophore for use in living cells or tissue are similar to the fundamental concepts used in traditional in vitro fluorescence microscopy. These include matching the fluorophore excitation/emission spectra with laser lines and spectral bandwidths of the detectors available on the confocal microscope; specificity of the fluorophore for tissue structures or processes of interest; high fluorescence quantum yield; and minimal photobleaching and delivery technique required to label the site of interest.57,58 When considering in vivo use of fluorescent contrast agents for confocal microscopy, the fluorophore must also be nontoxic, and not produce unwanted pharmacological effects. Many fluorophores used in ex vivo fluorescence microscopy are precluded due to the lack of toxicity data, or history of prior use in vivo, particularly if the study is to be performed in humans. However, there are still a number of fluorescent agents and their derivatives that have properties that make them suitable as potential in vivo contrast agents for use in human studies. Laser scanning microscopes require fluorescent probes that are excitable at a single fixed wavelength (rather than broad-spectrum excitation). The majority of fluorescence confocal microscopes are equipped with a 488-nm (blue) laser line from an argon ion laser, or more recently, the solid-state equivalent. This laser line is ideally suited to the excitation of blue excited, green-emitting fluorophores. Some of the agents used to date for in vivo fluorescence confocal imaging that can be excited at 488 nm include fluorescein and its analogs, acridine orange,13,15 acriflavine/proflavine,56,60 eosin,15 tetracyclines,22 4-di-2-ASP,13,59 green fluorescent protein (GFP),42 cresyl violet, and, to some extent, rhodamine B. Other fluorophores that have been used in vivo for confocal studies, such as methylene blue, indocyanine green, and Texas Red11 conjugates, have used excitation wavelengths provided by other lasers: 635-nm solid-state diode laser, 780-nm diode laser, and 568-nm krypton/argon laser, respectively. Physiological and pharmacokinetic processes that occur in living tissue, such as metabolism, fluorophore clearance, cellular and tissue permeability characteristics, or disease state of individual cells, affect the distribution
In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers
and time course of contrast agents, and therefore the cellular patterns observed during the examination. These factors need to be taken into account in the interpretation of images. Local physiologic factors, such as altered tissue physiology (e.g., changes in permeability or clearance of a fluorophore resulting in redistribution), and factors that affect the fluorescence properties of the fluorophore, such as pH or chemical environment, will all influence the pattern of fluorescence observed.61–63 While the dynamic interaction between the exogenous contrast agent and the living tissue can present complexities in the interpretation of images,57,61 it also highlights the ability to perform functional imaging in vivo, where such interactions may be studied. Photochemical interactions have a potential to produce toxic effects, causing negative alterations in the in vivo cellular environment. These factors need to be considered in fluorescence confocal microscopy.64,65
34.4 METHODS/ROUTES OF ADMINISTRATION The route of administration or method of application of a fluorescent contrast agent can significantly affect the tissue labeling that is produced, and hence the cellular patterns observed in the confocal images. The route of administration that is used may be dictated by physicochemical properties of the fluorophore (solubility, stability, pharmacology, etc.), the region of tissue or tissue compartment to be observed (e.g., skin surface, microvasculature, etc.), the extent of labeling (local vs. widespread or systemic), pharmacokinetics, or, in the case of dynamic studies, the biological application (e.g., monitoring transdermal delivery).
34.4.1 TOPICAL APPLICATION Topical application of a fluorescent contrast agent may be used to image the surface of the skin and morphological structure of the stratum corneum.66 One fluorophore that has been commonly used is fluorescein sodium (FNa).44,46–50 Fluorescein sodium is a low-molecularweight water-soluble fluorescent contrast agent that is routinely used in clinical applications in ophthalmological and vasculature diagnostic studies.67,68 When administered in concentrations of 5% w/v or greater via the oral,68 topical,46,50,69 intrathecal,70 and intravenous71,72 routes, fluorescein sodium is generally regarded as effective and safe in clinical practice. Fluorescein has the advantages of high absorption and quantum yield, and a peak maximum absorption that closely matches the 488-nm spectral line of an argon ion laser.73 Topical administration of fluorescein sodium (0.1% w/v solution) to the skin (mouse or human) results in heterogenous distribution of the fluorescent contrast agent molecules between the individual corneocytes, and in the wrinkles and hairs,44,46,49 with further penetration beyond
287
the stratum corneum into the epidermis limited by the barrier properties of skin.50 Using this method, sharp, high-contrast images of the stratum corneum and associated structures, such as the openings of hair follicles and sweat pores and the scale structure of hair shafts and skin surface texture, may be observed (see Figure 34.1A). Fluorescence confocal microscopy of topically applied fluorescent substances is also well suited to the in vivo analysis of penetration pathways and kinetics in the superficial skin layers,13,38,49 and observing the effects of cosmetic treatments such as exfoliation to remove surface squames and moisturizers that alter the cornocyte morphology and volume. Similar results are obtained with the use of other topically applied water-soluble fluorophores, such as the food additive D&C Orange 5 (salt of dibromofluorescein, tribromofluorescein, and tetrabromofluorescein; Figure 34.2A), acridine orange, rhodamine B, eosin, curcumin49 (Figure 34.2B), and others (Figure 34.2C). This technique is useful for assessing penetration of substances into the skin, and to directly observe the effectiveness of the skin’s barrier function.
34.4.2 INTRADERMAL ADMINISTRATION Superficial intradermal injection of fluorescein sodium (~10 μl of 0.1% w/v solution in water for injection or saline) into human skin in vivo overcomes the barrier function of the stratum corneum and introduces the fluorescent contrast agent directly into the viable epidermis.44 While this technique bypasses the barrier function of the skin, it does present several other challenges. It is necessary to place the tip of the needle as superficially as possible within the skin, to deposit the fluorescein solution into the epidermis rather than the dermis per se. In the studies to date using this technique, a 0.5-ml tuberculin syringe fitted with a 12.5-mm 29-gauge needle was used.45,48,50 In addition, as it is the aim to image the microscopic structure of the epidermis, the volume of injection must be very small to reduce structural damage as a result of introducing a large volume of fluid into the tissue. The injection is made by applying gentle pressure to the syringe plunger and allowing the fluorescein to diffuse into the tissue over 15 to 20 seconds. Rather than injecting a predetermined bolus dose, the quantity of fluorescein solution injected into the tissue is judged by observing the size of the color change to the skin as the contrast agent diffuses into the tissue. Under most circumstances, an area 3- to 5-mm in diameter infiltrated with fluorescein sodium is sufficient for confocal imaging. When injected in this manner, the fluorescein sodium diffuses into the tissue and does not raise a bleb at the injection site. While this technique is practical in human skin, intradermal injection of FNa into mouse skin is extremely difficult, as the skin is too thin. In human skin, fluorescein sodium distributes mainly in the aqueous extracellular spaces between the
288
Handbook of Non-Invasive Methods and the Skin, Second Edition
A Skin surface
B
Epidermis
PD
SS
Dermis C PD = Papillary dermis BC = Basal cells SS = Stratum spinosum C = Capillary
SS
BC
D SS
PD
C PD
BC E
FIGURE 34.1 Images from normal human skin in vivo following topical (A) and intradermal (B to E) administration of fluorescein sodium (0.1%). There is a continuum of change in cell size, shape, and morphology with increased imaging depth into the epidermis. Due to the undulating structure of the human epidermis and the planar en face images produced by optical sectioning, cells from multiple different epidermal layers/different depths in the epidermis may be observed in a single confocal image. (A) The contrast media distributes in the extracellular spaces, and the corneocytes in the stratum corneum (most superficial skin layer) are large, flattened, polygonal, and non-nucleated. Hair, skin creases, and debris on the surface label clearly. (B) Keratinocytes in the stratum granulosum are smaller than corneocytes, flattened and diamond shaped. Cytoplasmic organelles (keratohyaline granules and membrane-coating granules) in this layer give a granular appearance. (C) Smaller polyhedral-shaped keratinocytes are arranged in a regular honeycomb array in the stratum spinosum. The dermal papillae may be visible in deep stratum spinosum. (D) Cells in the stratum basale are small, dark, densely packed cuboidal-shaped cells. Cytoplasm of individual cells is often darker at this level (especially in darker skin types). Melanocytes are scattered throughout this layer (not specifically identified). (E) Dermal papillae project into the epidermis and appear as bright circular or elliptical islands surrounded by a single layer of darker basal cells. Capillary loops are visible within papillae; blood cells appear dark and contrast with bright plasma. Movement of blood cells through dermal capillaries was observed in real time.
In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers
A
B
C
FIGURE 34.2 The food dyes D&C Orange 5 Alum Lake (A), curcumin (B), and riboflavin (C) (50 μl of 1% w/v) held under occlusion for 20 minutes prior to imaging (FOV = 250 μm). (Images courtesy of E. Murr, Optiscan, unpublished results.)
289
keratinocytes in the deeper layers of the living epidermis46–48 (Figure 34.1B to E). In the fibrous tissue of the dermal papillae, FNa distributes uniformly (Figure 34.1E). The basal cells surrounding the papillae appear darker than the surrounding keratinocytes in the stratum spinosum.46 Blood cells may be observed flowing through dermal capillary loops as the FNa is cleared by diffusion into the capillaries46 (Figure 34.1E). If the fluorescein is injected carefully using this technique, and no solution escapes onto the surface of the skin, only the viable epidermis will be observed during confocal imaging, as the fluorescein is unable to cross the epidermal barrier, and hence remains trapped below it. This is the opposite situation to that observed for the topical application of a highly water-soluble fluorescent contrast agent. Due to the precision required in needle placement, the intradermal injection technique can be subject to operator variability. The distribution of the fluorescein sodium in the epidermis is dynamic. Initially it distributes in the extracellular space, thus contrasting the cell boundaries with little to no intracellular details detectable. Over the course of a few minutes, it is possible to observe diffusion of the fluorescein into the cytoplasm of keratinocytes in the living epidermis, and superficial dermis at the level of the dermal papillae. Approximately 10 to 15 minutes from the time of administration, sodium fluorescein provides a differential contrast of the epidermal keratinocytes whereby the extracellular space remains bright, with the cytoplasm mid-gray and the nuclei dark due to a differential access and distribution into these cellular compartments. With additional time, the relative concentration of the fluorescein in these compartments is altered as the fluorescein is redistributed, and the fluorescein is removed from the extracellular space into the cytoplasm of the keratinocytes, and thus the extracellular compartment appears darker than the contrasting cytoplasm.46 Fluorescein is cleared from the tissue by diffusion into the draining capillaries, and ultimately excreted in the urine. As the fluorescein drains into the dermal capillaries, blood flow in the dermal capillary loops may be observed, as the blood cells that do not take up the fluorescein contrast strongly against the fluorescein in the plasma. Other mechanical methods of delivering fluorescein sodium into human skin have been assessed in an attempt to avoid the use of an injection. These have included the use of needleless injectors (Panjet® and Injex® high-pressure devices), microneedle arrays, and sonophoresis.48 While these techniques have promise for reduced invasiveness, they have produced greater variability in our experience than the intradermal injection method. Initial pilot studies using intradermally administered fluorescein sodium (~10 μl of 0.1% w/v) in minimally erythemal sunburn of type 3 human skin in vivo have shown a population of brightly labeled inflammatory cells in the stratum granulosum and stratum spinosum that are not
290
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 34.3 Fluorescein sodium, ~10 μl of 0.1% w/v, was injected intradermally into sunburnt (minimal erythema) type 3 skin 24 hours post sun exposure. A population of brightly labeled inflammatory cells can be observed in the stratum granulosum and stratum spinosum in vivo. These cells are not observed in normal skin. FOV = 150 μm.
observed in normal skin (Figure 34.3). These cells are consistent with the inflammatory cell infiltrate observed in inflammatory lesions of the colonic mucosa during confocal endomicroscopy using i.v. fluorescein sodium.74 By combining the topical and intradermal techniques with fluorescein sodium, it is possible to image both the stratum corneum and underlying viable epidermis to the level of the mid-papillary dermis43–46,48 (Figure 34.1). It is not common to image as deep as the reticular dermis, although occasionally the fibrous structure of the reticular dermis is observed in thin human skin in vivo.
34.4.3 INTRAVENOUS ADMINISTRATION Fluorescein sodium is routinely used intravenously in ophthalmology for fluorescence angiography.71,72 Due to the small size of the fluorescein molecule and its pharmacokinetics, it leaks rapidly from the peripheral vasculature into the tissues, and distributes throughout the peripheral tissues, resulting in a yellow discoloration of the skin (and other peripheral tissues) within minutes. To date, intravenous administration of fluorescein sodium has not been used clinically for skin imaging studies with confocal microscopy, but it has been used for in vivo fluorescence confocal imaging of the colonic and gastric mucosa in humans,54–56,60 and for confocal retinal angiography.52 When used for confocal endomicroscopy in the gastrointestinal tract, the fluorescein is administered via the intravenous cannula used for routine sedation using a standard opthalmological dose (10 ml of 10% sodium fluorescein w/v). Following i.v. administration of the fluorescein, it is possible to image the gastrointestinal mucosa and
microcirculation within a minute or so, and the tissue fluorescence remains viable for at least 1 h. One advantage of using intravenously administered fluorescein sodium is that all peripheral tissues are rendered fluorescent, and thus it is possible to image many tissue sites using a single administration of a fluorescent contrast agent. This is in contrast to locally administered contrast agents, such as intradermally injected fluorescein sodium, where the localized fluorescence limits the area that can be examined using a single application of contrast agent. Like intradermal administration, intravascular administration results in delivery of the fluorescent contrast agent to the deeper skin structures. The fluorophore then diffuses toward the tissue surface. By exploiting the chemical or physical properties of various fluorescent contrast agents, it is possible to selectively label specific biological compartments. Intravenous administration of a fluorescently labeled, highmolecular-weight dextran (e.g., FITC-dextran, 70 to 150 kDa) in animals results in fluorescent labeling of the plasma compartment. The large dextran molecules remain trapped within the vasculature and allow the microcirculation to be imaged in vivo in real time.4,11,13,22,23,25,26,29 The blood cells exclude the labeled dextran, and therefore appear dark and contrast against the fluorescent plasma. While the FITC-dextran is confined to the vascular space in normal murine vasculature, changes in vascular permeability resulting from changes in the local cytokine environment, as occurs in tumor growth, or as a result of disease (e.g., diabetes) or injury (e.g., thermal or inflammation), result in the FITC-dextran leaking from the vasculature into the surrounding tissue.4,11,23,25,26,29 The frame rates of fluorescence confocal microscopes are typically not high enough to measure blood flow velocity; however, it is possible to determine the direction of flow, measure vessel diameter, and monitor vasoreactivity to various substances or stimuli. This technique has been used to study the cutaneous vascular changes in response to thermal injury,25,26 and angiogenesis in human melanomas xenografted onto nude mice11,23 using confocal imaging. In the mouse, the dermal microvasculature runs parallel to the skin surface, and therefore it is possible to examine and map the microvasculature architecture.26 In contrast, the capillaries in the human papillary dermis are organized orthogonal to the skin surface. Consequently, the imaging depth limitations of confocal microscopy in the visible part of the spectrum only allow in vivo imaging of the capillary loops in the dermal papillae.46 The deeper dermal vasculature that runs parallel to the skin surface is beyond the imaging depth capabilities of confocal microscopes. To date, fluorescently labeled dextrans have not been used intravenously in humans for microvascular imaging; however,
In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers
291
they have been used subcutaneously for fluorescence microlymphography studies.78–81
34.6 FUNCTIONAL IMAGING
34.5 IN VIVO IMMUNOFLUORESCENCE LABELING
Fluorescence confocal microscopy may be used to monitor the penetration and distribution of fluorescent or fluorescently labeled substances into the skin, and alterations in this behavior as a result of either disease or intervention. A simple experiment that highlights the functional barrier of the skin to water-soluble substances involves injecting fluorescein sodium intradermally into human volar forearm skin using the method described previously (~10 μl of 0.1% w/v), taking care to avoid any fluorescein spilling onto the skin surface. Another water-soluble fluorescent contrast agent, rhodamine B, is then applied topically (1% w/v) to the same site. These contrast agents can then be imaged using a two-channel confocal instrument that simultaneously excites the fluorescein sodium at 488 nm and the rhodamine at 568 nm, and detects the green and red fluorescence emissions, respectively, on separate detectors. The fluorescein sodium (green) remains localized in the living epidermis (Figure 34.4A), while the topically applied rhodamine (red) cannot penetrate beyond the stratum corneum (Figure 34.4B), and thus the skin barrier keeps the two fluorescent agents separated on either side of the barrier (Figure 34.4C). The images in Figure 34.4 were obtained by progressively collecting images stepwise into the tissue (three-dimensional data set) and then processing a cross-sectional image (XZ image) from the resultant data set. This results in a vertical image that shows the relative distribution of the red and green fluorophores.
An area of great potential for in vivo fluorescence confocal microscopy lies in the ability to target specific subcellular molecules (including proteins) using in vivo immunofluorescence labeling. Preliminary studies using targeted fluorescent probes in animal models of melanoma have demonstrated that pathologic protein overexpression can be microscopically observed in vivo.11,23,35,36 In these experiments, human melanoma cells were xenografted onto nude mice. A fluorescently labeled monoclonal antibody directed against a cell surface antigen that is overexpressed in the tumor cells was injected intravenously into anesthetized mice. The antibody was allowed to circulate for a few minutes, to localize to the melanoma cells before imaging the cells in vivo from the surface of the skin. This technique exploits the specificity of the antibody to bind to the target antigen, and thus specifically localize the fluorescent reporter to the melanoma cells. This technique has the potential to be used clinically as a highly specific and sensitive tumor cell detection method, or to characterize the protein expression repertoire of tumor cells as a means of staging the progression of disease state or tumor differentiation in vivo. If a multichannel confocal instrument that is capable of detecting two spectrally separated fluorophores is used, it is possible to combine immunodetection of implanted melanoma cells with microvascular imaging using a fluorescently labeled dextran,11,23,36 as described above. In these studies the xenografted melanoma cells were labeled with an FITC-labeled antibody against αVβIII integrin administered intravenously and a Texas Red-labeled dextran administered intravenously. These two fluorescent agents were imaged simultaneously to examine the relationship of the tumor-associated angiogenesis (red fluorescence) with the implanted melanoma cells (green fluorescence). Not only can the extent and morphology of the tumor vascularization be observed, but permeability changes in the tumor-associated vasculature are evidenced by leakage of the fluorescent dextran from the capillaries into the surrounding tissue. This appears as pools of contrast agent that are homogenous and acellular. These techniques can be used to examine the effects of antiangiogenesis treatments, or manipulation of the cytokine environment of tumors, and their effects on tumor progression.11 Similar dual-channel imaging techniques have been used to image the colon in vivo using topically applied achromycin and intravenous FITC-dextran.22
34.6.1 SKIN BARRIER FUNCTION
34.6.2 TRANSDERMAL DRUG DELIVERY The barrier function of the skin is particularly important when considering transdermal delivery of agents across the skin, for either therapeutics, diagnostics, or cosmetics. Manipulating the barrier function of the skin, or modifying the chemical properties of transdermal preparations, is often required to enable transdermal delivery. As in vivo fluorescence confocal microscopy allows direct visualization of fluorescent or fluorescently labeled substances in the skin, it can be used to monitor the dynamic distribution and clearance of substances applied to the skin.38,49,50 Studies by White et al.38 have used in vivo fluorescence confocal microscopy to examine the subcellular distribution of a fluorescently labeled antisense oligonucleotide applied topically to human skin xenografted onto mice. When the oligonucleotide was applied topically with no pretreatment, it did not penetrate the stratum corneum. In contrast, the use of tape stripping to remove the superficial layers of the grafted skin (thus altering the barrier function) prior to applying the oligonucleotide allowed penetration down to the basal keratinocytes, where it localized in the nuclei of these cells.38
292
Handbook of Non-Invasive Methods and the Skin, Second Edition
34.6.4 THERMAL BURN-INDUCED AUTOFLUORESCENCE
A
B
C
FIGURE 34.4 Fluorescence confocal microscopy can be used to monitor the penetration and distribution of fluorescent, or fluorescently labeled, substances into the skin. To highlight the barrier function of human skin to water-soluble substances, fluorescein sodium was injected intradermally (~10 μl of 0.1% w/v) and rhodamine B (0.1% w/v) was applied topically to the same site. Both fluorophores were simultaneously imaged using a twochannel fluorescence confocal microscope (detection of fluorescein, 505 to 550 nm; rhodamine B, >585 nm; excitation at 488 and 568 nm, respectively). The fluorescein sodium remains localized in the living epidermis (A), while the topically applied rhodamine cannot penetrate beyond the stratum corneum (B). (C) Both images overlayed. These images are oriented vertically relative to the skin surface and were obtained by progressively collecting images stepwise into the tissue (three-dimensional data set) and then processing a cross-sectional image (XZ image) from the resultant data set. FOV = 300 × 72 μm.
34.6.3 GFP EXPRESSION IN VIVO Currently, green fluorescent protein (GFP) is widely used as a reporter of protein expression in transgenic animals. GFP has similar fluorescent properties to fluorescein and is ideally suited to in vivo fluorescence confocal imaging with 488-nm laser illumination. Often, generating such transgenic animals is difficult and time-consuming. In vivo fluorescence confocal microscopy can be used to image the microscopic distribution of GFP expression in vivo without the need to sacrifice precious transgenic animals. While nonmagnifying observation of GFP expression in vivo has been widely reported, the lack of instruments specifically designed for in vivo microscopic examination of fluorescence has limited the ability to examine the microscopic distribution of GFP expression in vivo.
If skin is subjected to thermal injury, the collagen in the skin is denatured and becomes autofluorescent.27 In this case, the application of an exogenous fluorescence contrast agent is not necessary. This technique has been used to monitor thermal damage, assess burn depth and severity in mouse models, and test the effects for first aid intervention.27
34.7 FUTURE DIRECTIONS One area of research in skin disease detection and treatment is photodynamic therapy (PDT). PDT is a technique whereby a photosensitizer compound is administered either locally or systemically. The photosensitizer selectively accumulates in tumor cells. When the tissue is irradiated with the appropriate wavelength of light, the photosensitizer molecules are photoactivated to destroy the cell. Many of these photosensitizers are fluorescent, including protoporphyrin IX (PpIX). When excited at an appropriate wavelength, the photosensitizer fluoresces, allowing those cells containing the photosensitizer to be observed.75 It is therefore feasible to image these photosensitizer molecules in living tissue using confocal techniques to determine which cells contain the photosensitizer, the subcellular distribution, and the relative concentration of the photosensitizer. This technique may be useful in studying PDT photosensitizer and light activation dosimetry to achieve the desired PDT effect. To date, only very limited pilot studies have been done in which the microscopic distribution of PpIX has been examined using in vivo microscopy.76,77 The results have been variable and range from no visible fluorescence to bright abundant fluorescence. At this stage, the source of this variability is not clear. FITC-dextran has been used in human fluorescence microlymphography studies to examine the cutaneous lymphatic drainage, in the study of both the lymphatics78,81 and lymphoedma.79 To date, this use of FITC-dextran has not been examined using in vivo confocal microscopy. FITC-dextran has been used intravenously in animal models, as mentioned previously, to examine the microvasculature of a range of tissues in vivo. Together, these two pieces of information suggest that the study of lymphatic structure and function using confocal microscopy in vivo is possible, and that there is a precedent for use of FITCdextran in humans, which suggests that microvascular imaging with this agent may be feasible clinically. Conventional bench-top confocal microscopes usually feature more than one laser illumination wavelength and multiple detectors. This allows more than one fluorescent agent to be imaged simultaneously. There are two commercially available confocal systems specifically designed
In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers
for in vivo dermatological confocal imaging: a reflection confocal system by Lucid, Inc.,7 and a fluorescence confocal system by Optiscan Pty. Ltd.8 Both of these instruments are single-channel systems. The production of a multichannel instrument for in vivo use that is capable of simultaneously detecting both a reflection signal and one or more fluorescence signals would greatly enhance the quantity and quality of information that can be obtained from the tissue.82 For example, reflection confocal images of the skin derive contrast from endogenous skin substances such as keratin and melanin,83 and therefore clearly identify the skin surface and melanocytes, with other cell types difficult to distinguish. The fluorescence images generated using intradermal fluorescein, however, show all cells within the epidermis,44 with little specificity for individual cell types. By simultaneously acquiring and superimposing both a reflection and fluorescence image of the skin, it is possible to combine the benefits of both imaging modalities.82 Currently, there is a range of diagnostic immunohistochemical techniques used in the assessment of skin pathologies84 that have the potential to be adapted to in vivo application with fluorescent reporters, rather than the enzymatic or color reactions currently used in vitro. While such techniques are routine in vitro, to date the use of these immunoreagents has not been tested in vivo. One of the challenges of in vivo immunolabeling is the size of the antibody molecules, which prevents them from crossing biological membranes easily. Therefore, the in vivo immunofluorescence studies performed to date using confocal detection have used antibodies directed against cell surface antigens that are presented on the extracellular surface of the target cells.11,23,35,36 Recent advances in antibody fragment technology have allowed the production of very small, highly specific mono- and polyvalent immunoprobes, such as diabodies and triabodies.85–91 These molecules have an advantage in that they essentially contain only the antigen binding sites, and do not possess the remainder of the antibody molecule. The small size of these molecules offers the potential, if tagged to a small fluorescent dye, to allow intracellular immunolabeling in vivo. If combined with a multichannel confocal instrument, it is theoretically possible to use several different immunolabels, each tagged with a different colored fluorescent marker, to probe a tissue to determine if certain cells are expressing particular combinations of proteins. Such a technique would be powerful for studying the progression of tumors or clinically staging diseases in vivo, and manipulating treatments accordingly.
REFERENCES 1. Minsky, M. Microscopy Apparatus. U.S. Patent 3013467, 1961 (filed November 7, 1957).
293
2. Amos, W.B., White, J.G., and Fordham, M. Use of confocal imaging in the study of biological structures, Appl. Opt., 26, 3239–3243, 1987. 3. Aspres, N., Egerton, I.B., Lim, A.C., and Shumack, S.P. Imaging the skin. Aust. J. Derm., 44, 19–27, 2003. 4. Papworth, G.D., Delaney, P.M., Bussau, L.J., Vo, L.T., and King, R.G. In vivo fibre optic confocal imaging of microvasculature and nerves in the rat vas deferens and colon. J. Anat., 192, 489–495, 1998. 5. Gonzalez, S., Gonzalez, E., White, W.M., Rajadhyaksha, M., and Anderson, R.R. Allergic contact dermatitis: correlation of in vivo confocal imaging to routine histology. J. Am. Acad. Dermatol., 40, 7086–7136, 1999. 6. Aghassi, D., Anderson, R.R., and Gonzalez, S. Timesequence histologic imaging of laser-treated cherry angiomas with in vivo confocal microscopy. J. Am. Acad. Dermatol., 43, 37–41, 2000. 7. Rajadhyaksha, M., González, S., Zavislan, J.M., Anderson, R.R., and Webb, R.H. In vivo confocal laser microscopy of human skin II: Advances in instrumentation and comparison with histology. J. Invest. Dermatol., 113, 293–303, 1999. 8. Suihko, C., Swindle, L.D., Thomas, S.G., and Serup, J. Fluorescence fibre-optic confocal microscopy of skin in vivo: microscope and fluorophores. Skin Res. Tech., 11, 254–267, 2005. 9. Delaney, P.M., Harris, M.R., and King, R.G. Novel microscopy using fibre optic confocal imaging and its suitability for subsurface blood vessel imaging in vivo. Clin. Exp. Pharmacol. Physiol., 20, 197–198, 1993. 10. Johnson, I. Introduction to fluorescence technique. In Handbook of Fluorescent Probes and Research Chemicals, 6th ed., Haugland, R.P., Ed. Molecular Probes, Inc., Eugene, OR, 1996. 11. Anikijenko, P., Vo, L.T., Murr, E., Carrasco, J., McLaren, W.J., Chen, Q., Thomas, S.G., Delaney, P.M., and King, R.G. In vivo detection of small subsurface melanomas in athymic mice using non-invasive fibre optic confocal imaging (FOCI). J. Invest. Derm., 117, 1442–1448, 2001. 12. King, R.W. and Delaney, P.M. Confocal microscopy in pharmacological research. TiPS, 15, 275–279, 1994. 13. Bussau, L.J., Vo, L.T., Delaney, P.M., Papworth, G.D., Barkla, D.H., and King, R.G. Fibre optic confocal imaging (FOCI) of keratinocytes, blood vessels and nerves in hairless mouse skin in vivo. J. Anat., 192, 187–194, 1998. 14. Delaney, P.M., King, R.G., Lambert, J.R., and Harris, M.R. Fibre optic confocal imaging (FOCI) for subsurface microscopy of the colon in vivo. J. Anat., 184, 157–160, 1994. 15. McLaren W., Anikijenko P., Barkla D., Delaney P.M., and King R.G. In vivo detection of experimental ulcerative colitis in rats using fiberoptic confocal imaging (FOCI). Dig. Dis. Sci., 46, 2263–2276, 2001. 16. Villringer, A., Haberl, R.L., Dirnagl, U., Anneser, F., Verst, M., and Einhaupl, K.M. Confocal laser microscopy to study microcirculation on the rat brain surface in vivo. Brain Res., 504, 159–160, 1989.
294
Handbook of Non-Invasive Methods and the Skin, Second Edition
17. Villringer, A., Dirnagl, U., Them, A., Schurer, L., Krombach, F., and Einhaupl, K.M. Imaging of leukocytes within the rat brain cortex in vivo. Microvasc. Res., 42, 305–315, 1991. 18. Villringer, A., Them, A., Lindauer, U., Einhaupl, K., and Dirnagl, U. Capillary perfusion of the rat brain cortex: an in vivo microscopy study. Circ. Res., 75, 55–62, 1994. 19. Dirnagl, U., Villringer, A., Gebhardt, R., Haberl, R.L., Schmiedek, P., and Einhaupl, K.M. Three-dimensional reconstruction of the rat brain cortical microcirculation in vivo. J. Cereb. Blood Flow Metab., 11, 353–360, 1991. 20. Dirnagl, U., Villringer, A., and Einhaupl, K.M. In-vivo confocal scanning laser microscopy of the cerebral microcirculation. J. Microsc., 165, 147–157, 1992. 21. Delaney, P.M., Papworth, G.D., and King, R.G. Fibre optic confocal imaging (FOCI) for in vivo subsurface microscopy of the colon. In Methods in Disease: Investigating the Gastrointestinal Tract, Preedy, V.R. and Watson, R.R., Eds. Greenwich Medical Media, London, 1998. 22. McLaren, W.J., Anikijenko, P., Thomas, S.G., Delaney, P.M., and King, R.G. In vivo detection of morphological and microvascular changes of the colon in association with colitis using fibreoptic confocal imaging (FOCI). Digest. Dis. Sci., 47, 2424–2433, 2002. 23. Anikijenko, P., Vo, L.T., Thomas, S.G., Delaney, P.M., and King, R.G. Changes in dermal vascular morphology in vivo associated with melanoma: imaged using fibre optic confocal microscopy (FOCI). Proc. Aust. Soc. Clin. Exp. Pharm. Tox., 6, 127, 1999 (Poster 2-13). 24. Merchant, F.A., Aggarwal, S.J., Diller, K.R., and Bovik, A.C. In-vivo analysis of angiogenesis and revascularization of transplanted pancreatic islets using confocal microscopy. J. Microsc., 176, 262–275, 1994. 25. Vo, L.T., Papworth, G.D., Delaney, P.M., Barkla, D.H., and King, R.G. A study of vascular response to thermal injury on hairless mice by fibre optic confocal imaging, laser Doppler flowmetry and conventional histology. Burns, 24, 319–324, 1998. 26. Vo, L.T., Papworth, G.D., Delaney, P.M., Barkla, D.H., and King, R.G. In vivo mapping of the vascular changes in skin burns of anaesthetised mice by fibre optic confocal imaging (FOCI). J. Derm. Sci., 23, 46–52, 2000. 27. Vo, L.T., Anikijenko, W.J., Delaney, P.M., Barkla, D.H., and King, R.G. Autofluorescence of skin burns detected by fibre optic confocal imaging (FOCI): evidence that cool water treatment limits progressive thermal damage in anaesthetised hairless mice. J. Trauma, 51, 98–104, 2001. 28. Shahrokh, F., Gross, P., Merrill-Skoloff, G., Furie, B.C., and Furie, B. Real-time in vivo imaging of platelets, tissue factor and fibrin during arterial thrombus formation in the mouse. Nat. Med., 8, 1175–1181, 2002. 29. Papworth, G.D., Delaney, P.M., Bussau, L.J., and King, R.G. In vivo subsurface microvascular imaging using fibre optic confocal microscopy (FOCI). Proc. Aust. Physiol. Pharm. Soc., 26, 200P, 1995.
30. Ichijima, H., Petroll, W.M., Jester, J.V., and Cavanagh, H.D. Confocal microscopic studies of living rabbit cornea treated with benzalkonium chloride. Cornea 11, 221–225, 1992. 31. Moller-Pedersen, T., Li, H.F., Petroll, W.M., Cavanagh, H.D., and Jester, J.V. Confocal microscopic characterization of wound repair after photorefractive keratectomy. Invest. Ophthalmol. Vis. Sci., 39, 487–501, 1998. 32. Boyde, A., Capasso, G., and Unwin, R.J. Conventional and confocal epi-reflection and fluorescence microscopy of the rat kidney in vivo. Exp. Nephrol., 6, 398–408, 1998. 33. Sabharwal, Y.S., Rouse, A.R., Donaldson, L., Hopkins, M.F., and Gmitro, A.F. Slit-scanning confocal microendoscope for high-resolution in vivo imaging. Appl. Opt., 38, 7133–7144, 1999. 34. Gmitro, A.F., Rouse, A.R., and Sabharwal, Y.S. In situ optical biopsy with a confocal microendoscope. In Proceedings of the 22nd Annual EMBS International Conference, Chicago, 2000, pp. 1040–1042. 35. Carrasco, J.C., Vo, L.T., Anikijenko, P., Delaney, P.M., and King, R.G. Fibre optic confocal imaging (FOCI) for early in vivo detection of melanoma in athymic mice. Proc. Aust. Soc. Clin. Exp. Pharm. Tox., 6, 127, 1999 (Poster 2-14). 36. Vo, L.T., Carrasco, J.C., Delaney, P.M., Chen, Q., and King, R.G. In vivo detection of small sub-surface human melanomas in mice. G.I.T. Imaging Microsc., 3, 20–22, 2001. 37. Kacza, J., Grosche, J., Seeger, J., Brauer, K., Bruckner, G., and Hartig, W. Laser scanning and electron microscopic evidence for rapid and specific in vivo labelling of cholinergic neurons in the rat basal forebrain with fluorochromated antibodies. Brain Res., 867, 232–238, 2000. 38. White, P.J., Fogarty, R.D., Liepe, I.J., Delaney, P.M., Werther, G.A., and Wraight, C.J. Live confocal microscopy of oligonucleotide uptake by keratinocytes in human skin grafts on nude mice. J. Invest. Derm., 112, 887–892, 1999. 39. Boyde, A., Jones, S.J., Taylor, M.L., Wolfe, L.A., and Watson, T.F. Fluorescence in the tandem scanning microscope. J. Microsc., 157, 39–49, 1990. 40. Boyde, A., Wolfe, L.A., Maly, M., and Jones, S.J. Vital confocal microscopy in bone. Scanning, 17, 72–85, 1995. 41. Jones, S.J. and Taylor, M.L. Confocal fluorescence microscopy: some applications in bone cell biology. J. Microsc., 158, 249–259, 1990. 42. Ilyin, S.E., Flynn, M.C., and Plata-Salaman, C.R. Fiberoptic monitoring coupled with confocal microscopy for imaging gene expression in vitro and in vivo. J. Neurosci. Methods, 108, 91–96, 2001. 43. Kishimoto, J., Ehama, R., Ge, Y., Kobayashi, T., Nishiyama, T., Detmar, M., and Burgeson, R.E. In vivo detection of human vascular endothelial growth factor promoter activity in transgenic mouse skin. Am. J. Pathol., 157, 103–110, 2000.
In Vivo Confocal Microscopy of the Skin Surface Using Fluorescent Markers
44. Thorn Leeson, D., Lynn Meyers, C., and Subramanyan, K. In vivo confocal fluorescence imaging of skin surface cellular morphology. OSA Biomedical Topical Meetings Technical Digest PD8-1, 2002. 45. Swindle, L., Delaney, P., Thomas, S., and Freeman, M. View of human skin in vivo as observed using fluorescence confocal microscopic imaging. Proceedings of the 20th World Congress of Dermatology, Paris, July 2002, P1198, p. 182. 46. Swindle, L.D., Thomas, S.G., Freeman, M., and Delaney, P.M. View of normal human skin in vivo as observed using fluorescent fiber-optic confocal microscopic imaging. J. Invest. Derm., 121, 706–712, 2003. 47. Swindle, L., Feit, N., Hazan, C., Busam, K., and Halpern, A. Features of melanocytic lesions as demonstrated on reflectance and fluorescence confocal imaging. Proceedings of the 61st Annual Meeting of the American Academy of Dermatology, 2003, poster session 491, p. 5. 48. Swindle, L., Freeman, M., Jones, B., and Thomas, S. Fluorescence confocal microscopy of normal human skin and skin lesions in vivo. Skin. Res. Technol., 9, 167, 2003. 49. Lademann, J., Richter, H., Otberg, N., Lawrenz, F., Blume-Peytavi, U., and Sterry, W. Application of a dermatological laser scanning microscope for investigation in skin physiology. Laser Phys., 13, 1–5, 2003. 50. Thomas, S.G., Murr, E.R., Anikijenko, P., Reese, A., and Delaney, P.M. The use of fluorescence in vivo confocal imaging in the evaluation of transdermal delivery methods. Skin. Res. Technol., 9, 162, 2003. 51. Mueller, A.J., Freeman, W.R., Folberg, R., Bartsch, D.U., Scheider, A., Schaller, U., and Kampik, A. Evaluation of microvascularization pattern visibility in human choroidal melanomas: comparison of confocal fluorescein with indocyanine green angiography. Graefes. Arch. Clin. Exp. Ophthalmol., 237, 448–456, 1999. 52. Freeman, W.R., Bartsch, D.U., Mueller, A.J., Banker, A.S., and Weinreb, R.N. Simultaneous indocyanine green and fluorescein angiography using a confocal scanning laser opthalmoscope. Arch. Opthalmol., 116, 455, 1998. 53. McLaren, W., Tan, J., and Quinn, M. Detection of cervical neoplasia using non-invasive fibre-optic confocal microscopy. 5th International Multidisciplinary Congress Eurogin 2003, Paris, April 2003, pp. 213–217. 54. Kiesslich, R., Burg, J., Vieth, M., Gnaendiger, J., Enders, M., Delaney, P., Polglase, A., McLaren, W., Janell, D., Thomas, S., Nafe, B., Galle, P.R., and Neurath, M.F. Confocal laser endoscopy for diagnosing intraepithelial neoplasias and colorectal cancer in vivo. Gastroenterology, 127(3), 706–713, 2004. 55. Kiesslich, R., Gossner, L., Dahlmann, A., Veith, M., Stolte, M., Hofmann, A., Jung, M., Nafe, B., Galle, P.R., Ell, C., and Neurath, M.F. In vivo histology of Barrett’s esophagus and associated neoplasias by confocal laser endomicroscopy. Gastrointestinal Endoscopy, 61(5), AB101, 2005.
295
56. Kiesslich, R., Goetz, M., Burg, J., Stolte, M., Siegel, E., Maeurer, M., Thomas, S., Strand, D., Galle, P.R., and Neurath, M.F. Diagnosing Helicobacter pylori in vivo by confocal laser endoscopy. Gastroenterology, 128(7), 2119–2123, 2005. 57. Cullander C. Fluorescent probes for confocal microscopy. Methods Mol. Biol., 122, 59–73, 1999. 58. Mullins, J.M. Overview of fluorophores. Methods Mol. Biol., 34, 107–116, 1994. 59. Papworth, G.D., Delaney, P.M., Bussau, L.J., and King, R.G. In vivo subsurface fibre optic confocal microscopy: selected applications of the fluorescent dye 4-(4-diethylaminostyryl)-n-methyl pyridinium iodide. Proc. Aust. Physiol. Pharm. Soc., 27, 19, 1996. 60. Polglase, A.L., McLaren, W.J., Skinner, S.A., Kiesslich, R., Neurath, M.F., and Delaney, P.M. A fluorescence confocal endomicroscope for in vivo microscopy of the upper- and lower-GI tract. Gastrointestinal Endoscopy, 62(5), 686–695, 2005. 61. Rajadhyaksha, M. and Gonzalez, S. Real-time in vivo confocal fluorescence microscopy. In Handbook of Biological Fluorescence, Mycek, M.-A. and Pogue, B.W., Eds. Marcel Dekker, New York, 2003. 62. Sjoback, R., Nygren, J., and Kubista, M. Absorbtion and fluorescence properties of fluorescein. Spectrochim. Acta, A51, 7,1995. 63. Song, L., Hennink, E.J., Young, I.T., and Tanke, H.J. Photobleaching kinetics of fluorescein in quantitative fluorescence microscopy. Biophys. J., 68, 2588–2600, 1995. 64. Saetzler, R.K., Jallo, J., Lehr, H.A., Philips, C.M., Vasthare, U., Arfors, K.E., and Tuma, R.F. Intravital fluorescence microscopy: impact of light-induced phototoxicity on adhesion of fluorescently labeled leukocytes. J. Histochem. Cytochem., 45, 505–513, 1997. 65. Zhang, J.L.,Yokoyama, S., and Ohhashi, T. Inhibitory effects of fluorescein isothiocyanate photoactivation on lymphatic pump activity. Microvasc. Res., 54, 99–107, 1997. 66. Papworth, G.D., Delaney, P.M., and King, R.G. Use of topically applied dyes for fibre optic confocal imaging (FOCI) in vitro and in vivo. Clin. Exp. Pharm. Physiol., Suppl. 1, 55, 1993. 67. Kohno, T., Miki, T., and Hayashi, K. Choroidopathy after blunt trauma to the eye: a fluorescein and indocyanine green angiographic study. Am. J. Ophthalmol., 126, 248–260, 1998. 68. Hara, T. and Inami, M. Efficacy and safety of fluorescein angiography with orally administered sodium fluorescein. Am. J. Ophthalmol., 126, 560–564, 1998. 69. Elman, M.J., Fine, S.L., Sorenson, J., Yannuzzi, L., Hoopes, J., Weidenthal, D.T., and Singerman, L.J. Skin necrosis following fluorescein extravasation. A survey of the Macula Society. Retina, 7, 89–93, 1987. 70. Scavone, J. and Fraser, D. Intrathecal sodium fluorescein. Drug Intel. Clin. Pharm., 17, 186–187, 1983. 71. Yannuzzi, L.A., Rohrer, K.T., Tindel, L.J., Sobel, R.S., Costanza, M.A., Shields, W., and Zang, E. Fluorescein angiography complication survey. Opthalmology, 93, 611–617, 1986.
296
Handbook of Non-Invasive Methods and the Skin, Second Edition
72. Cavallerano, A.A. Ophthalmic fluorescein angiography. Optom. Clin., 5, 1–23, 1996. 73. Haugland, R.P. Handbook of Fluorescent Probes and Research Chemicals, 6th ed. Molecular Probes, Eugene, OR, 1996. 74. Thomas, S.G. and Kiesslich, R. Personal communication. 75. Olivo, M., Lau, W., Manivasager, V., Hoon, T.P., and Christopher, C. Fluorescence confocal microscopy and image analysis of bladder cancer using 5-aminolevulinic acid. Int. J. Oncol., 22, 523–528, 2003. 76. Swindle, L, Thomas, S., Delaney, P., and Freeman, M. Personal communication. 77. Manifold, R. and Anderson, C. Personal communication. 78. Bollinger, A., Jager, K., Sgier, F., and Seglias, J. Fluorescence microlymphography. Circulation, 64, 1195–1200, 1981. 79. Mellor, R.H., Stanton, A.W.B., Azarbod, P., Sherman, M.D., Levick, J.R., and Mortimer, R.S. Enhanced cutaneous lymphatic network in the forearms of women with postmastectomy oedema. J. Vasc. Res., 37, 501–512, 2000. 80. Leu, A.J., Husmann, M.J.W., Held, T., Frisullo, R., Hoffmann, U., and Franzeck, U. Measurement of the lymphatic clearance of the human skin using a fluorescent tracer. J. Vasc. Res., 38, 423–431, 2001. 81. Stanton, A.W.B., Kadoo, P., Mortimer, P.S., and Levick, J.R. Quantification of the initial lymphatic network in normal human forearm skin using fluorescence microlymphography and stereological methods. Microvasc. Res., 54, 156–163, 1997. 82. Delaney, P., Pattie, R., and Thomas, S. Directions for in vivo fibre optic confocal microscopy. Skin Res. Technol., 9, 186–187, 2003.
83. Rajadhyaksha, M., Grossman, M., Esterowitz, D., Webb, R., and Anderson, R.R. In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J. Invest. Derm., 104, 946–952, 1995. 84. Kanitakis, J. Immunohistochemistry of normal human skin. In Diagnostic Immunohistochemistry of the Skin: An Illustrated Text, Kanitakis, J., Vassileva, S., and Woodley, D., Eds. Chapman & Hall, London, 1998. 85. Power, B.E. and Hudson, P.J. Synthesis of high avidity antibody fragments (scFv multimers) for cancer imaging. J. Immunol. Methods, 242, 193–204, 2000. 86. Nuttall, S.D., Irving, R.A., and Hudson, P.J. Immunoglobulin VH domains and beyond: design and selection of single-domain binding and targeting reagents. Curr. Pharm. Biotechnol., 1, 253–263, 2000. 87. Hudson, P.J., and Souriau, C. Recombinant antibodies for cancer diagnosis and therapy. Expert Opin. Biol. Ther., 1, 845–855, 2001. 88. Kortt, A.A., Dolezal, O., Power, B.E., and Hudson, P.J. Dimeric and trimeric antibodies: high avidity scFvs for cancer targeting. Biomol. Eng., 18, 95–108, 2001. 89. Todorovska, A., Roovers, R.C., Dolezal, O., Kortt, A.A., Hoogenboom, H.R., and Hudson, P.J. Design and application of diabodies, triabodies and tetrabodies for cancer targeting. J. Immunol. Methods, 248, 47–66, 2001. 90. Power, B.E., Kortt, A.A., and Hudson, P.J. Generation of recombinant multimeric antibody fragments for tumor diagnosis and therapy. Methods Mol. Biol., 207, 335–350, 2003. 91. Hudson, P.J., and Souriau, C. Engineered antibodies. Nat. Med., 9, 129–134, 2003.
Vivo Confocal Microscopy 35 In Application in Product Research and Development Toyonobu Yamashita and Motoji Takahashi Bioengineering Research Labs, Shiseido Co., Ltd., Yokohama, Japan
CONTENTS 35.1 35.2 35.3 35.4
Introduction............................................................................................................................................................297 Background of Reflectance Mode CM .................................................................................................................297 Object.....................................................................................................................................................................297 Methodological Principles .....................................................................................................................................298 35.4.1 Epidermal Thickness Measurement ..........................................................................................................298 35.4.2 Visualization of Melanin Distribution.......................................................................................................299 35.5 The Application of Confocal Microscopy Images................................................................................................299 35.5.1 Evaluation of Changes of Epidermal Thickness.......................................................................................299 35.5.2 Evaluation of Sunscreen Agent against Ultraviolet Radiation .................................................................300 35.6 Suitability of RCM for Skin Research..................................................................................................................302 References .......................................................................................................................................................................303
35.1 INTRODUCTION It is a trend today for biologists to visualize the human tissue structure and to investigate morphological and functional changes in vivo. Recently, also in the dermatological field, many imaging tools have been developed for diagnostic purposes, such as ultrasound imaging,1,2 multiphoton imaging,3,4 and optical coherence topography (OCT).5,6 Especially, microscopy combined with confocal technique is an outstanding diagnostic tool that allows observing the specific skin layer as an optical horizontal section.7–10 The confocal microscopy (CM) has many advantages: realtime observation, noninvasiveness, and microimaging at the cellular level in vivo. Therefore, not only dermatologists but also skin researchers have been attracted by CM as a tool for the evaluation of skin internal conditions. Furthermore, the confocal technique will potentially be found to be a more powerful technique when combined with other detection systems, such as multiphoton,3,4 second harmonic generation, 4,11,12 and Raman spectroscopy.13,14 In this chapter, we introduce our trial in the examination of the skin changes under various circumstances using reflectance mode CM (RCM). Moreover, we discuss the possibilities of the RCM application to basic
research in healthy human skin and cosmetic products development.
35.2 BACKGROUND OF REFLECTANCE MODE CM The RCM for dermatology was created by Rajadhyaksha et al.7,8 and improved by Lucid, Inc.15,16 So far, RCM has become a commercially available and extensively used equipment for skin research. Recently, fiber-type CM has been developed as a convenient tool.17 González et al.18–20 evaluated various skin disorders using RCM and provided RCM a wider clinical significance as a useful diagnostic tool in the 1990s. It was reported that RCM images of human pigmented lesions are correlated with pathological features observed on the dermoscopic and histologic examination.21
35.3 OBJECT Various diagnostic tools equipped with imaging systems have been developed. Especially, microscopic visualization in vivo is crucial for understanding cutaneous changes and differences at the cellular level. From this point of 297
298
Handbook of Non-Invasive Methods and the Skin, Second Edition
view, RCM is one of the most suitable tools, enabling the showing of clear images of individual cells at real time. However, it should be taken into account that the information provided by RCM is based on reflectance light, which depends on the feature of biomaterial. In RCM, there is no detection of specific targets with fluorescent probes. Therefore, when developing applications for RCM, its advantages as well as limitations should be considered. RCM has two advantages. First, the confocal effect allows getting structural information of the skin from a cellular to a skin layer level. Shift of field view of the target skin layer is easily controlled by the objective lens, and a spatial understanding of the skin structure is possible. Second, the reflectance light information is obtainable from one particular skin material. Rajadhyaksha et al.7 reported that melanin is a strong reflective agent. Actually, in the RCM images, dark skin provides bright images in the epidermal layer, contrasting with the skin color. Our major objective is to widen the applicability of RCM imaging in the evaluation of the skin reactions against environmental changes and the efficacy of cosmetics or pharmaceuticals.
35.4 METHODOLOGICAL PRINCIPLES 35.4.1 EPIDERMAL THICKNESS MEASUREMENT One of the advantages of RCM is acquisition of images at a specific depth. The lateral and axial resolutions for one confocal point are 0.5 to 1.0 μm and 3 to 5 μm, respectively.8 To obtain spatial images, the skin must be scanned in each axis by the laser. Originally, the XY-axes are scanned by a Galvanometric mirror and a polygon mirror. However, concerning the Z-axis, there is no laser scanning to this direction, and the only way to move in the Z-direction is using the objective lens controlled by manual or electrical pulse. Therefore, there have been only a few reports showing optical vertical sections obtained by RCM. However, analysis of vertical images having much significant information regarding specific skin layers should be done. Thus, we built a Z-actuator controlled software for RCM, which is combined with a stacking system for XY-optical sections. One pulse signal in the Zaxis corresponds to 3.6-μm movement of depth. Using the automatically constructed entire image stack, the vertical section is viewed using the other exclusive XZ two-dimensional viewing software. In this method, a resolution higher than 3.6 μm is not possible. Nevertheless, this resolution is still quite high, compared to the resolution of vertical images obtained by other diagnostic tools, for instance, ultrasound imaging and OCT. Our optical vertical image allows two-dimensional measurement of the epidermal thickness and the degree of elevation of the rate ridge of dermis (Figure 35.1). In addition, stacking and
A
B
50 μm
FIGURE 35.1 Two-dimensional optical images of ventral forearm obtained by RCM. (A) Horizontal image. (B) Constructed vertical image.
imaging procedures for measurement of the skin thickness have some advantages: 1. Prevention of errors owing to trembling of hands when touching the Z-actuator in manual thickness measurement 2. Precise analysis of skin thickness based on multiple points measurement 3. Visual assessment of the vertical section Furthermore, using configured RCM XY-optical sections, the three-dimensional structure of the epidermis can also be constructed (Figure 35.2). However, the threedimensional display of the reflectance image does not lead to a suitable three-dimensional image for understanding the spatial configuration. Preferably, edge detection, which utilizes the difference of brightness in images due to the refractive index difference between the epidermis and dermis, should be performed. An image analysis method of three-dimensional edge detection and threedimensional reconstruction based on the Snakes method22 allows visualization of the surface of the stratum corneum and the dermoepidermal junction. Three-dimensional edge analysis of the reconstructed epidermal structure23 made it possible to analyze the thickness of the epidermis and the undulation of the dermoepidermal junction in detail.
In Vivo Confocal Microscopy Application in Product Research and Development
299
50 μm
FIGURE 35.2 A reconstructed three-dimensional image of the epidermal layer based on the Snakes method. Upper layer: Skin surface. Lower layer: Epidermal-dermal junction.
35.4.2 VISUALIZATION
OF
MELANIN DISTRIBUTION
The contrast in RCM images is mainly due to the differences of refractive indexes between individual biomaterials. Especially, melanin has a high refractive index (n = 1.70)24 in the skin and provides strong reflectance compared with the background (epidermis, n = 1.34).25 Therefore, the pigmented area can be visualized as bright areas in the RCM image. The differences in melanin contents can be recognized when comparing different racial groups, outer vs. inner forearm, or moles vs. surrounding skin tissue. The melanocytes of animals like rodents microstructually hold abundant melanosomes, which make the observation of melanocytes easier. On the other hand, human melanocytes transfer the majority of melanosomes to surrounding keratinocytes.26 The melanosomes are localized in the epidermal basal layer as “melanin hats”7 or “supranuclear melanin caps”27 (Figure 35.3), which protect the nuclei of basal cells. Therefore, it is difficult to clearly detect the shape of melanocytes in human skin except in pathological conditions when a high concentration of melanosomes is present in melanocytes. Attention must be paid to the fact that the high reflectance is not limited to melanin alone. In RCM images, keratin is also a strong reflectance agent, and the stratum corneum (n = 1.55),28 which is rich in keratin, causes high reflectance as well. Moreover, in the dermis, which has a comparatively low refractive index (n = 1.40),25 erythrocytes and immuno-related cells are observed as highly reflectant cells. Therefore, it should be kept in mind which layer of the skin is being examined. Regarding melanin examination, one should focus on the epidermal basal layer, where melanin is concentrated.
35.5 THE APPLICATION OF CONFOCAL MICROSCOPY IMAGES 35.5.1 EVALUATION THICKNESS
OF
CHANGES
OF
EPIDERMAL
Generally, strong surfactants stimulate the proliferation of epidermal basal cells and thicken the epidermis. For evaluation of rough skin, we measured the epidermal thickness as described above using experimentally induced dry skin. The subjects consisted of three men aged 32 to 40 years. We used the skin of the ventral aspect of the forearm for this study. Treatments were tape stripping and sodium dodecyl sulfate (SDS) application. To induce rough skin, subjects repeated tape strippings 30 times on one forearm and applied 3% SDS solution for 8 hours under occlusion on the other forearm by means of an 8-mm Finn chamber®. The treated skin sites were analyzed by RCM at the baseline and at the following times after inducing rough skin: 3, 10, and 25 days. Additionally, skin surface images were photographed by a videomicroscope, HM-2200 MD (Hirox, Tokyo, Japan), at 100× magnification. The measurement of epidermal thickness was performed in several sites (Figure 35.4). It is obvious that the microtopograph obtained by the videomicroscope shows a disorganized skin surface pattern after both treatments, but the disorder is almost recovered 25 days after treatment (Figure 35.5 and Figure 35.6). On the other hand, the measurement of thickness by RCM showed that the epidermal layer was thicker at the treated area 3 and 10 days after tape stripping or SDS application (Figure 35.7). Although the thicker epidermis had a tendency to become milder 25 days after treatment, it could
300
Handbook of Non-Invasive Methods and the Skin, Second Edition
Epidermis A B Dermis
Vertical section
Epidermis
Dermis
20 μm Horizontal section A
Horizontal section B
FIGURE 35.3 The supranuclear caps in the epidermal basal layer. RCM images (A and B) of human forearm skin are shown. Supranuclear caps (arrows) localized in the basal layer show strong backscattering light, but melanocytes are not clearly recognized.
Stratum corneum A Epidermis Dermis
B C
FIGURE 35.4 Epidermal thickness measurement at the dorsal forearm. The epidermal thickness was measured at points A (shallow) and B (deep) in their adjoining points. The depth of the rate ridge (C) was also measured.
not be recovered completely. These results show that treatment with tape stripping or SDS causes persistent effects in the internal skin, although appreciable recovery was observed on the surface.
35.5.2 EVALUATION OF SUNSCREEN AGENT AGAINST ULTRAVIOLET RADIATION Excessive exposure to ultraviolet rays (UVR) causes acute effects such as sunburn and pigmentation, as well as epidermal hyperproliferation. The sun protectors against UVR can prevent erythema and skin tanning. Many sunscreen products, which contain chemical UV absorber or physical UV scattering agent, have been developed as UV protectors. However, the evaluation of sunscreen in
human subjects is limited to visual judgments, such as the sun protection factor (SPF) evaluation. We investigated the role of sunscreens in preventing the internal skin from damage using RCM. First, phototest was performed for the determination of the minimum erythema dose (MED) using a solar simulator (circles of 1 cm diameter). After phototest, the sites on both ventral forearms were imaged by RCM. These served as a baseline for following RCM imaging. One of the forearms received a topical application of sunscreen lotion that contained physical sunscreen agents for reflection and scattering of UVR. The other one was treated in the absence of the lotion (control). After the application, the subject was exposed to 2 MED on both ventral forearms.
In Vivo Confocal Microscopy Application in Product Research and Development
Before treatment
3 days after treatment
10 days after treatment
25 days after treatment
301
FIGURE 35.5 Images of the changes in skin surface caused by tape-stripping treatment obtained by videomicroscope (100×).
Before treatment
3 days after treatment
10 days after treatment
25 days after treatment
FIGURE 35.6 Images of the changes in skin surface treated with 3% SDS aqueous solution obtained by videomicroscope (100×).
302
Handbook of Non-Invasive Methods and the Skin, Second Edition
Tape stripping A
μm
80 70
140
60
120
50
100
40
80
30
60
C
B
160
90 80 70 60
0
5
10 15 Days
20
25
50 40 30 0
5
10 15 Days
20
0
25
5
10 15 Days
20
25
20
25
3% SDS aqueous solution A
μm
80
B
160
70
140
60
120
50
100
40
80
C
90 80 70 60
30
50 40 30
60 0
5
10 15 Days
20
25
0
5
10 15 Days
20
25
0
5
10 15 Days
FIGURE 35.7 Changes of epidermal thickness induced by treatment of tape stripping or application of 3% SDS aqueous solution.
The RCM imaging at different times (3 and 10 days) were obtained, as well as clinical photographs by videomicroscope at 100× magnification. Three days after UVR exposure, the sunscreen nontreated site showed drastic changes on RCM images (Figure 35.8). The outline of keratinocytes became obscure and darker. At this point, the skin surface microtopography showed a persistent erythema occurring from 1 day after UV exposure. These changes on RCM images suggested that the affected configuration of internal cells might be due to epidermal hyperproliferation. The above effects were gradually relieved, and the RCM images showed an almost normal feature at 10 days after UVR exposure. The only exception was the increased reflectance of the basal layer due to melanin newly produced by melanocytes (Figure 35.9). On the other hand, no effects of UVR were observed on the sunscreen-treated sites. Therefore, it was demonstrated by RCM that the sunscreen agents used herein prevented UVR damage at the cellular level.
35.6 SUITABILITY OF RCM FOR SKIN RESEARCH Deeper penetration of incident light of the scanning laser and minimization of unnecessary reflectance light
reaching the detector are required to obtain clear RCM images. Large differences of refractive indexes between the stratum corneum and immersion medium for objective lenses on the skin (water, oil, or none) lead to strong reflectance light from their interface. The stratum corneum as “the skin entrance” is known as being highly refractive (n = 1.55)28 for scanning laser. Consequently, the skin surface image, but not the internal skin image, can be obtained using a normal objective lens, which requires no immersion medium. This high reflection is due to the interface between the stratum corneum and air. Conventionally, the RCM developed by Lucid, Inc., is equipped with a water immersion lens, which is less prone to high reflectance. However, in our experience, the selection of an oil immersion lens is better for the acquisition of good images than water ones, maybe because the refractive indexes between oil and the stratum corneum are similar. Furthermore, in the oil-immersed RCM examination, the scanning laser reaches deeper skin sites and brighter images are obtained (Figure 35.10). Therefore, attention to the refractive index is crucial for obtaining clear RCM images. In order to acquire a series of continuous vertical or horizontal RCM images for two- or three-dimensional reconstruction of the internal skin, it is necessary to find a stable skin position and condition for RCM examination.
In Vivo Confocal Microscopy Application in Product Research and Development
A-1
B-1
A-2
B-2
303
50 μm
FIGURE 35.8 Prevention effect of sunscreen on the spinous layers from UVB damage. Ventral forearm skin was exposed to 2 MED UVB with or without sunscreen lotion and imaged 3 days later. A-1 (skin surface image, 100×) and A-2 (CM image) show the sunscreen lotion applied site. B-1 (skin surface image, 100×) and B-2 (CM image) show the sunscreen lotion nonapplied site.
The investigator should avoid a position influenced by pulsatory motions or areas on the bone, which easily transmit vibration. In this regard, the conventional RCM should be equipped with the piezo-type translator, which allows rapid and accurate vertical positioning.29
REFERENCES 1. Serup, J., Keiding, J., Fullerton, A., Gniadecka, M., and Gniadecki, R. High-frequency ultrasound examination of skin: introduction and guide. In Handbook of NonInvasive Methods and the Skin, Serup, J.J. and Jemec, G.B.E., Eds., CRC Press, Boca Raton, FL, 1995. 2. el-Gammal, S., Hoffmann, K., Auer, T., Korten, M., Altmeyer, P., Hoss, A., and Ermert, H. A 50MHz high-resolution ultrasound imaging system for dermatology. In Ultrasound in Dermatology, Altmeyer, P., el-Gammal, S., and Hoffmann, K., Eds., Springer-Verlag, Berlin, 1992. 3. Ragan, T.M., Huang, H., and So, P.T. In vivo and ex vivo tissue applications of two-photon microscopy. Methods Enzymol., 361, 481–505, 2003.
4. Konig, K. and Riemann, I. High-resolution multiphoton tomography of human skin with subcellular spatial resolution and picosecond time resolution. J. Biomed. Opt., 8, 432–439, 2003. 5. Pagnoni, A., Knuttel, A., Welker, P., Rist, M., Stoudemayer, T., Kolbe, L., Sadiq, I., and Kligman, A. Optical coherence tomography in dermatology. Skin Res. Technol., 5, 83–87, 1999. 6. Aguirre, A.D., Hsiung, P., Ko, T.H., Hartl, I., and Fujimoto, J.G. High-resolution optical coherence microscopy for high-speed, in vivo cellular imaging. Opt. Lett., 28, 2064–2066, 2003. 7. Rajadhyaksha, M., Grossman, M., Esterowitz, D., Webb, R.H., and Anderson R.R. In vivo confocal scanning laser microscopy of human skin: melanin provides strong contrast. J. Invest. Dermatol., 104, 946–952, 1995. 8. Rajadhyaksha, M., González, S., Zavislan, J.M., Anderson, R.R., and Webb, R.H. In vivo confocal scanning laser microscopy of human skin: advances in instrumentation and comparison with histology. J. Invest. Dermatol., 113, 293–303, 1999.
304
Handbook of Non-Invasive Methods and the Skin, Second Edition
A-1
A-2
B-1
B-2
50 μm
FIGURE 35.9 Prevention effect of sunscreen on the basal layers from UVB damage. Ventral forearm skin was exposed to 2 MED UVB with or without sunscreen lotion and imaged 1 week later. A-1 (skin surface image, 100×) and A-2 (CM image) show the sunscreen lotion applied site. B-1 (skin surface image, 100×) and B-2 (CM image) show the sunscreen lotion nonapplied site. 9. Corcuff, P., Betrand, C., and Leveque, J.L. Morphometry of human epidermis in vivo by real-time confocal microscopy. Arch. Dermatol. Res., 285, 475–481, 1993. 10. Corcuff, P. and Leveque, J.L. In vivo vision of the human skin with tandem scanning microscope. Dermatology, 186, 50–54, 1993. 11. Zoumi, A., Yeh, A., and Tromberg, B.J. Imaging cells and extracellular matrix in vivo by using second-harmonic generation and two-photon excited fluorescence. Proc. Natl. Acad. Sci. U.S.A., 99, 11014–11019, 2002. 12. Cox, G., Kable, E., Jones, A., Fraser, I., Manconi, F., and Gorrell, M.D. 3-dimensional imaging of collagen using second harmonic generation. J. Struct. Biol., 141, 53–62, 2003. 13. Caspers, P.J., Lucassen, G.W., Carter, E.A., Bruining H.A., and Puppels, G.J. In vivo confocal Raman microspectroscopy of the skin: noninvasive determination of molecular concentration profiles. J. Invest. Dermatol., 116, 434–442, 2001. 14. Caspers, P.J., Lucassen, G.W., and Puppels, G.J. Combined in vivo confocal Raman spectroscopy and confocal microscopy of human skin. Biophys. J., 85, 572–580, 2003.
15. Rajadhyaksha, M. and Zavislan, J.M. Confocal laser microscope images tissue in vivo. Laser Focus World, 33, 119–127, 1997. 16. Rajadhyaksha, M. and Zavislan, J.M. Confocal reflectance microscopy of unstained tissue in vivo. Retinoids, 14, 26–30, 1998. 17. Swindle, L.D., Thomas, S.G., Freeman, M., and Delaney, M. View of normal human skin in vivo as observed using fluorescent fiber-optic confocal microscopic imaging. J. Invest. Dermatol., 121, 706–712, 2003. 18. González, S., Rajadhyaksha, M., and Anderson, R.R. Non-invasive (real-time) imaging of histologic margins of a proliferative skin lesion in vivo. J. Invest. Dermatol., 111, 538–539, 1998. 19. González, S., González, E., White, W.M., Rajadhyaksha, M., and Anderson, R.R. Allergic contact dermatitis: correlation of in vivo confocal imaging to routine histology. J. Am. Acad. Dermatol., 40, 708–713, 1999. 20. González, S., White, W.M., Rajadhyaksha, M., Anderson, R.R., and Gonzalez, E. Confocal imaging of sebaceous gland hyperplasia in vivo to assess efficacy of pulsed dye laser treatment. Lasers Surg. Med., 25, 8–12, 1999.
In Vivo Confocal Microscopy Application in Product Research and Development
A-1
B-1
A-2
B-2
305
100 μm
FIGURE 35.10 Comparison of the CM images obtained by water and oil immersion lenses. The water immersion lens (Olympus UMPlanFL20 × water, images A-1 and A-2) and the oil immersion lens (Olympus UPlanApo20 × oil, images B-1 and B-2) are shown. A-1 and B-1 show the skin surface, and A-2 and B-2 show the basal layer. 21. Charles, C.A., Marghoob, A.A., Busam, K.J., ClarkLoeser, L., and Halpern, A.C. Melanoma or pigmented basal cell carcinoma: a clinical-pathologic correlation with dermoscopy, in vivo confocal scanning laser microscopy, and routine histology. Skin Res. Technol., 8, 282–287, 2002. 22. Kass, M., Witkin, A., and Terzopoulos, D. Snakes: active contour models. Int. J. Comput. Vision, 1, 321–331, 1988. 23. Wang, J., Saito, H., Ozawa, S., Kuwahara, T., Yamashita, T., and Takahashi, M. Extraction of dermo-epidural surface from 3D volumetric images of human skin. Int. J. Image Graphics, 3, 589–608, 2003. 24. Vitkin, I.A., Woolsey, J., Wilson, B.C., and Anderson, R.R. Optical and thermal characterization of natural melanin. Photochem. Photobiol. 59, 455–462, 1994. 25. Tearney, G.J., Brezinski, M.E., Southern, J.F., Bouma B.E., Hee, M.R., and Fujimoto, J.G. Determination of the refractive index of highly scattering human tissue by optical coherence tomography. Opt. Lett., 20, 2258–2260, 1995.
26. Yamashita, T., Kuwahara, T., González, S., and Takahashi, M. Non-invasive visualization of melanin and melanocytes by reflectance-mode confocal microscopy. J. Invest., Dermatol., 124, 235–240, 2005. 27. Kobayashi, N., Nakagawa, A., Muramatsu, T., Yamashita, Y., Shirai, T., Hashimoto, M.W., Ishigaki, Y., Ohnishi, T., and Mori, T. Supranuclear melanin caps reduce ultraviolet induced DNA photoproducts in human epidermis. J. Invest. Dermatol., 110, 806–810, 1998. 28. Scheuplein, R.J. A survey of some fundamental aspects of the absorption and reflection of light by tissue. J. Soc. Cosmet. Chem., 15, 111–122, 1964. 29. Caspers, P.J., Lucassen, G.W., Bruining H.A., and Puppels, G.J. Automated depth-scanning confocal Raman microspectroscopy for rapid in vivo determination of water concentration profiles in human skin. J. Raman Spectrosc., 31, 813–818, 2000.
Magnetic Resonance (NMR) 36 Nuclear Examination of the Epidermis In Vivo Bernard Querleux Laboratoires de Recherche de L’Oréal, Aulnay-sous-bois, France
CONTENTS 36.1 Introduction............................................................................................................................................................307 36.2 Object.....................................................................................................................................................................307 36.3 Methodological Principle ......................................................................................................................................308 36.3.1 In Vivo High-Resolution MR Imaging ......................................................................................................308 36.3.1.1 Equipment...................................................................................................................................308 36.3.1.2 Application for Imaging of the Epidermis and Dermis ............................................................308 36.3.1.3 Application for Imaging of the Pathological Epidermis ...........................................................308 36.3.2 In Vivo Measurement of the NMR Parameters in Skin Layers................................................................310 36.3.2.1 Method........................................................................................................................................310 36.3.2.2 Application for Characterization of the Epidermis and Dermis ...............................................310 36.3.2.3 Application for Studies of Water Behavior in the Epidermis and Dermis ...............................311 36.3.2.4 Application for Reconstruction of the Hydration Profile of the Stratum Corneum and Epidermis17 ..........................................................................................................312 36.4 Sources of Error.....................................................................................................................................................312 36.5 Correlation with Other Methods ...........................................................................................................................313 36.6 Recommendations..................................................................................................................................................313 Acknowledgments ...........................................................................................................................................................313 References .......................................................................................................................................................................313
36.1 INTRODUCTION The acquisition of sectional images of the epidermis in vivo is a difficult task, which at the present time may be achieved by three recent techniques: ultrasound, magnetic resonance (MR) imaging, and confocal microscopy. The oldest one, about 10 years old, is high-frequency ultrasound imaging. With an axial resolution in the order of 30 to 60 mm, it is well adapted for the visualization of the whole skin, but also differentiates epidermis from dermis in some cases.1,2 The most recent technique is in vivo confocal microscopy, about 2 years old, which is characterized by an excellent spatial resolution of about 1 mm.3,4 This technique is very efficient for visualizing the stratum corneum, but optical improvements are still in progress in order to increase the signal-to-noise ratio on images of the inner layers of the epidermis. The first studies of skin by high-resolution MR imaging date from 1987–1988.5,6 As with ultrasound, high-resolution MR imaging, obtained
by modifying a standard whole-body MR scanner, is more adapted to the visualization of the whole skin, but with an in-depth resolution in the order of 35 to 70 mm, epidermis can be clearly delineated and thus analyzed.
36.2 OBJECT Over the last decade the nuclear magnetic resonance (NMR) technique has become a powerful method in medical diagnosis. This technique is of great interest because one can noninvasively obtain not only a spatial localization of the different tissues, but also quantitative information on tissues by measuring their proton relaxation times T1 and T2, and proton density N(H). These are intrinsic parameters of each tissue, providing, for instance, useful information about correlation with water content or interactions of water protons with macromolecules and, more generally, in the understanding of the molecular organization. 307
308
Handbook of Non-Invasive Methods and the Skin, Second Edition
The aim of this chapter is to present some applications of in vivo high-resolution MR imaging to epidermis examination.
36.3 METHODOLOGICAL PRINCIPLE 36.3.1 IN VIVO HIGH-RESOLUTION MR IMAGING 36.3.1.1 Equipment The main requirement for imaging the different skin layers is high spatial resolution, at least in the direction perpendicular to the skin surface, as the typical thickness of epidermis is less than 100 mm. In a conventional wholebody MR imaging system and in clinical use, the pixel size is limited to about 300 mm, which corresponds to a field of view of 8 cm. In the depth direction, it is insufficient to observe the epidermis. To increase the resolution in this direction, we designed a high-strength surface gradient coil of small dimensions, allowing a decrease of the field of view in this direction to 18 mm, corresponding to a pixel size of 70 microns.7 A small-surface radio frequency coil was designed in order to improve the signalto-noise ratio. This specific imaging module (Figure 36.1) is connected in place of the standard corresponding coils of the system, allowing us to obtain high-resolution images of the skin on most parts of the body.
FIGURE 36.2 Whole-body MR imager with the specific imaging module for high-resolution imaging of the skin in vivo. For each acquisition, the subject lies on the examination cradle and the skin area of interest is centered on the surface of the module.
the skin layers: epidermis and its annex, dermis, hypodermis and its fibrous septa, and even a thickened stratum corneum on the palm, as well as on the heel (Figure 36.3a to c). Such a high spatial resolution allows us to measure various thicknesses of the epidermis and to characterize shape, size, and density of the pilosebaceous units, according to the location.
36.3.1.2 Application for Imaging of the Epidermis and Dermis
36.3.1.3 Application for Imaging of the Pathological Epidermis
During the acquisition time, the skin area to be investigated lies on the imaging module (Figure 36.2) and motion is avoided by stabilizing the body with straps, and by surrounding the skin area with a double-sided adhesive tape on the module. High-resolution MR images are obtained with an 18 × 50 mm2 field of view, corresponding to a pixel size of 70 × 310 mm2 and a slice thickness of 3 mm in two-dimensional acquisition and 0.7 mm in threedimensional acquisition.7 We are thus able to differentiate
If large cutaneous lesions can be studied with conventional MR scanners,8–10 pathologic epidermis imaging must be performed in vivo with high-resolution imaging systems, and in vitro on biopsy samples with very high field microscopy NMR imaging systems.11
Gradient power supply
X RF coil
a Reference plate
To receiver Surface gradient coil
Y
FIGURE 36.1 Schematic diagram of the specific imaging module comprising a high-strength surface gradient coil and a small surface radio frequency coil, 3 cm in diameter.
1 mm
Z
2 mm
FIGURE 36.3 (a) MR imaging of skin on the calf. Layers of skin are clearly delineated: upper bright layer represents epidermis. The dermis appears as a thicker layer of hypointensity. Pilosebaceous units are seen as inclusions of epidermis inside dermis (arrowheads). The hypodermis and its fibrous septa are clearly visible.
Nuclear Magnetic Resonance (NMR) Examination of the Epidermis In Vivo
309
2 mm
Our group has some preliminary results on pathological epidermis combined with inflammatory processes such as eczema or psoriasis. Figure 36.4a represents a psoriatic lesion before any treatment, located on the calf, and Figure 36.4b, the healthy contralateral area. A thickened outer bright layer, about 1 mm in thickness, well differentiates the healthy area from the pathological one. However, studies are still in progress in order to establish whether this bright layer only represents a thickened epidermis, or whether it also represents the inflammatory process of the upper dermis.
5 mm b
2 mm
FIGURE 36.3 (b) MR image of skin on the back. The dermis is much thicker than on the calf and pilosebaceous units are more numerous and thinner. Hypodermis invaginations are also visible.
5 mm a
c
2 mm
2 mm
FIGURE 36.4 (a) MR image of a psoriatic plaque on the calf.
5 mm 5 mm b
FIGURE 36.3 (c) MR image of skin on the heel. The upper layer is the stratum corneum, about 1 mm thick. The brighter layer between stratum corneum and dermis is epidermis, which is also particularly thick. The dermis appears more homogeneous and without visible epidermis annexe.
FIGURE 36.4 (b) MR image of the healthy contralateral area. A thickened outer bright layer, about 1.1 mm thick, differentiates the pathologic epidermis from the healthy contralateral one, about 0.15 mm thick.
310
Handbook of Non-Invasive Methods and the Skin, Second Edition
S = k N(H) exp(–TE/T2) (1 – exp(–TR/T1))
a
FIGURE 36.5 (a) MR image of a squamous cell carcinoma on the heel.
(36.1)
where k is a function of many parameters constant during the protocol, N(H) is proportional to the mobile proton density or free water content as determined by the MR imaging technique, T1 and T2 are MR tissue relaxation times, and TE and TR, parameters of the imaging sequence, are echo time and repetition time, respectively. T1 measurements are carried out by acquiring a set of two-dimensional spin echo (SE) images only varying TR (TE = 16 msec, values of TR ranging from 100 to 4000 msec). T2 measurements were carried out in a first step by acquiring a set of three-dimensional gradient echo images in order to achieve short echo time. At the present time, owing to many potential sources of artifacts related to the gradient echo sequence, T2 measurements are currently carried out by acquiring a set of two-dimensional SE images (TR = 500 msec, values of TE ranging from 16 to 70 msec). T1 and T2 are calculated by fitting the signal intensities from a region of interest (ROI) to Equation 36.1. 36.3.2.2 Application for Characterization of the Epidermis and Dermis14
The case of a squamous cell carcinoma on the heel gives an idea of the feasibility to delineate the exact border of epidermis tumors noninvasively. Morphological information about the shape, depth, and location of the tumor between high-resolution MR imaging (Figure 36.5a) and histology (Figure 36.5b) is in good agreement. Consequently, with such a high resolution, MR imaging may be useful to discriminate tumors with regular or irregular boundaries, and diffused tumors.
36.3.2 IN VIVO MEASUREMENT OF THE NMR PARAMETERS IN SKIN LAYERS 36.3.2.1 Method If high-resolution imaging is required to visualize in detail both the epidermis and dermis, the NMR imaging technique is also of great interest to obtain complementary information on the physicochemical properties of tissues by measuring their proton relaxation times T1 and T2,12 and proton density N(H). The received signal intensity can be approximately expressed as:13
T1 measurements of calf skin layers 300 Signal intensity (a.u.)
FIGURE 36.5 (b) Histology of the tumor. Dimensions, shape, and location are quite similar with the two methods.
We measured T1 and T2 in the different skin layers on two different locations: calf and heel, which is characterized by a thickened stratum corneum of about 1 mm in thickness. Typical data on skin layers in the calf are presented in Figure 36.6 and on the heel in Figure 36.7. Mean values on nine healthy volunteers (four women, five men; mean age ± standard deviation, 35 ± 6 years) are summarized in Table 36.1. Results have shown that epidermis and dermis are characterized by shorter T2 relaxation times than other biological soft tissues, and that dermis and epidermis can be differentiated by their mean
200
100
0
Epidermis Dermis Hypodermis 1000
2000 3000 Repetition time (ms)
4000
(a)
FIGURE 36.6 (a) Plots of signal intensities of the skin layers on the calf and the fitted curves obtained for T1 calculation.
Nuclear Magnetic Resonance (NMR) Examination of the Epidermis In Vivo
T2 measurements of calf skin layers
T1 measurements of heel skin layers 300
Epidermis Dermis Hypodermis
200
100
20
40 Echo time (ms)
Signal intensity(a.u.)
Signal intensity(a.u.)
300
0
311
60
Epidermis Dermis Stratum corneum
200
100
0
3000
1000 2000 Repetition time (ms)
(b)
(a)
FIGURE 36.6 (b) Plots of signal intensities and the fitted curves obtained for T2 calculation. Data presented are issued from one of the five ROIs for the dermis and hypodermis and from averaged 40 pixels in the epidermis. Error bars, standard deviation between pixels of the selected ROI.
FIGURE 36.7 (a) Plots of signal intensities of skin layers of the heel and the fitted curves obtained for T1 calculation. T2 measurements of heel skin layers
TABLE 36.1 Mean Values of the Relaxation Times T1 and T2 of the Different Skin Layers in Heel and Calf Heel T1 (msec)
Calf T2 (msec)
T1 (msec)
T2 (msec)
Stratum corneum 313 ± 47a 10.2 ± 2 – – Epidermis 720 ± 53 36.6 ± 5 887 ± 92 22.3 ± 7 Dermis 728 ± 104 17.9 ± 5 870 ± 143 13 ± 2.4 Hypodermis — — 393 ± 34 35.4 ± 3.6
Signal intensity (a.u.)
500 Epidermis Dermis Stratum corneum 300
100 0
10
20 Echo time (ms)
30
(b)
a
Standard deviation calculated from nine subjects, five ROIs per subject.
T2 values. These short T2 values may be essentially assigned to their fibrous protein content, particularly of the dermis. In contrast, the measured T1 values differentiate neither epidermis and dermis from other tissues, nor epidermis from dermis. Concerning the location dependence, heel epidermis T2 is lengthened and is closer to usual T2 values. This longer value, compared to the calf epidermis value, would be related to physiological differences induced by friction.15 Finally, stratum corneum on heel presents short values not only of T2, but also of T1, characteristic of a low water content. 36.3.2.3 Application for Studies of Water Behavior in the Epidermis and Dermis16 In another study, and in addition to T1 and T2 measurements, we evaluated the relative proton density N(H). This parameter, which is proportional to the mobile
FIGURE 36.7 (b) Plots of signal intensities and the fitted curves obtained for T2 calculation. Data presented are issued from one of the five ROIs in each layer. Error bars, standard deviation between pixels of the selected ROI.
proton density or free water content as determined by the MR imaging technique, allowed us to quantify the mobile water fractions in the epidermis and dermis. So a maximum signal intensity for each skin layer was computed according to Equation 36.1 using signal intensity for infinite TR value corrected by T2 decay for each layer. The quantity thus obtained was still dependent on the instrument’s receiver gain, k, which was optimized for each subject. So the quantity N(H) had to be normalized in order to obtain comparable N(H) values between subjects. For this purpose, an external reference constant was introduced for the entire experiment. This reference, mounted inside the radio frequency coil, was simultaneously scanned with each subject, and every subject’s relative proton density was normalized to that of the external reference.
312
Handbook of Non-Invasive Methods and the Skin, Second Edition
0.80
1.0 0.8
0.60
p < 0.05 N (H)(a.u.)
N (H)(a.u.)
Untreated Treated
0.70
0.6 0.4
0.50 0.40 0.30 0.20
0.2
0.10 0.00
0.0
Epidermis
Outer dermis
Inner dermis
Total dermis
FIGURE 36.8 Relative proton density N(H) mean values of the skin layers on the thigh, normalized with the external reference and thus expressed in arbitrary units. The outer dermis corresponds to a subepidermal layer of 200 μm in thickness. The contribution of cutaneous appendages has been excluded from dermis measurements by semiautomatic image analysis. Relative proton density, i.e., mobile water fraction, is significantly higher in epidermis than in dermis (p < .01) and is significantly higher in the outer dermis than in the inner dermis (p < .01). N(H) is significantly higher in aged outer dermis than in young outer dermis (p < .05). Hatched columns, young subjects; solid columns, old subjects. Error bars, standard deviation.
After medical examination, two groups of healthy volunteers were involved in the study: 10 young women (25 to 40 years; mean ± standard deviation, 32 ± 4) and 10 elderly women (70 to 81 years; mean ± standard deviation, 74 ± 4). Results have confirmed in vivo skin layer differentiation through relaxation times performed previously. Moreover, relative proton density quantification (Figure 36.8) has shown that epidermal mobile water is at least twice as abundant as dermal mobile water. So, as it is well established that epidermis and dermis have a total water content of the same order of magnitude, differences in N(H) values reflect differences in bound water content, much more important in dermis than in epidermis. About differences between the two groups, if the main result concerns the upper part of the dermis (about 200 mm in thickness), which contains more mobile water protons in chronologically aged skin than in young adult skin, no clear difference could have been established in the epidermis. 36.3.2.4 Application for Reconstruction of the Hydration Profile of the Stratum Corneum and Epidermis17 We applied the method of measuring mobile proton density by high-resolution MR imaging to analyze modifications of hydration of the stratum corneum in vivo under treatment. The images were acquired on the heel, where stratum corneum is particularly thick compared to our
0.000
1.000
2.000 Depth (mm)
3.000
FIGURE 36.9 In vivo hydration profile of stratum corneum and epidermis by high-resolution MR imaging: effect of a bath. An increase in hydration is recorded in the outer layer of the thick stratum corneum on the heel.
in-depth pixel size of 70 mm. The images were obtained by using a chemical shift imaging sequence,18 which allows one to obtain two images: one related to the mobile water fraction within the tissues, the second one related to the lipid fraction. Signal intensity, extracted from the images related to the mobile water fraction, was measured for every depth increment equal to the pixel size. Then hydration profiles vs. depth were plotted after T1 and T2 fitting and normalization. Hydration profiles of heel stratum corneum and epidermis, before and after immersion of the heel for 15 min in water at 30˚C, are presented in Figure 36.9. An increase in hydration on the treated heel is clearly recorded in the outer layer of the stratum corneum. The main interest of this method lies in the fact that the physical signal is perfectly located, as spatial encoding is the basis of in vivo imaging. This method differs from other noninvasive methods that acquire an averaged signal from a nondelimited volume of interest. This preliminary study tends to delineate two structures within the stratum corneum: an outer layer where hydration can be modified by external mechanisms and an inner layer where hydration is not altered. In the future this methodology will be available for in vivo measurement, in thick stratum corneum (palms and soles), of mobile water content vs. depth. Such a measure could be very useful for the follow-up of topical moisture product effects and for the fundamental understanding of stratum corneum physiology in health or disease.
36.4 SOURCES OF ERROR With in vivo imaging, whatever the technique used, ultrasound, x-rays, or magnetic resonance imaging, one major source of errors is spatial distortions induced by the encoding method. In MR imaging, these distortions are related
Nuclear Magnetic Resonance (NMR) Examination of the Epidermis In Vivo
to the use of gradients, nonlinear over the volume of interest. This problem is more important when using a high-intensity asymmetric surface gradient coil, because with such a design linearity cannot be as good as with a conventional symmetric gradient coil. Consequently, many experiments have to be carried out in order to specify the volume of interest in which the linearity of such a gradient is acceptable. Another source of error is motion artifacts, which are potentially critical in high-resolution imaging. This problem was well overcome by stabilizing the skin to investigate relative to the gradient coil by straps and double-sided adhesive tape, in such a way that no shift on images acquired at different times has been recorded. Other sources of errors, specific to the MR technique,19 have to be analyzed in order to assess their possible influence on high-resolution studies. These artifacts may be related to very short T2, susceptibility, and chemical shift. Very short T2 values induce a loss in spatial resolution, whereas susceptibility and chemical shift introduce spatial distortions. In fact, only the chemical shift artifact is visible on our MR images. It corresponds to a shift of 1.5 pixels, only localized at the boundary between water-rich and lipid-rich tissues, such as the dermohypodermis junction, and thus does not alter the visualization of a thin epidermis. Finally, there is a source of error concerning epidermis thickness measurement. If the in-depth resolution corresponds to the pixel size, i.e., 70 mm, we have to keep in mind that the slice thickness is in the range of 0.7 to 3 mm. With such an isotropic voxel, measurement of epidermis thicknesses may be corrupted and overestimated due to the partial-volume effect, which could happen if the direction of the slice thickness is not quite parallel to the skin surface. In conclusion, even if partial-volume effect seems to be the most important source of error, no clear degradation of the image quality arises from the potential sources of artifacts listed above.
36.5 CORRELATION WITH OTHER METHODS At the present time, no study has been published about a comparison of in vivo NMR examination of the epidermis and other imaging methods: ultrasound or confocal microscopy in vivo and histology in vitro. In comparison to ultrasound, where the dermoepidermis junction is more or less echogenic, in MR imaging the contrast is very important, so the interest in comparing it to other methods is less evident and should only concern epidermis thickness measurements, in order to evaluate the overestimation probably induced by the partial-volume effect.
313
Unfortunately, we have to keep in mind that the comparison with the gold standard, histology, suffers from some difficulties: differences in slice thickness, effects of shrinkage, influence of dehydration, and deparaffination. Thus, we cannot expect an exact correlation in epidermis thickness measurement or visualization of its annex.
36.6 RECOMMENDATIONS If skin imaging may be performed in some cases with conventional whole-body imaging systems, in order to study large cutaneous tumors,20 epidermis examination requires a modification of standard systems. We have proposed the use of not only a specific small-surface radio frequency coil, but also a high-strength small-surface gradient coil, which seems to be a very efficient method to obtain high-resolution images while maintaining a short echo time. We are thus able to obtain a pixel size in the range of 35 to 70 mm in the direction perpendicular to the skin surface, which is necessary for examining the epidermis in vivo. With such a high spatial resolution, morphological characteristics may give complementary information for the diagnosis of tumors. More particularly, differentiation between benign and malignant melanomas will be one of the most important challenges in assessing the utility of this noninvasive technique. Nevertheless, we think that tissue characterization by quantitative measurements of the NMR parameters is the main interest of this technique. It already allows one to differentiate the skin layers, to quantify modifications of the physicochemical properties in normal and pathologic epidermis, and even to measure hydration processes in the thick stratum corneum. Consequently, if microscopic examination of a biopsy specimen is a simple way to evaluate the epidermis, tissue characterization by highresolution MR imaging is an in vivo noninvasive method that makes it possible to follow, on the same patient, physiologic processes or evolution of different pathologic conditions under treatment.
ACKNOWLEDGMENTS All the works presented are the results of collaborations between our laboratory with I’Institut d’Electronique Fondamentale-CNRS-Orsay, France, and Centre Inter-Etablissement de Résonance Magnétique, CIERM, Hôpital de Bicêtre, Le Kremlin-Bicêtre, France.
REFERENCES 1. Querleux, B., Lévêque, J.L., and de Rigal, J., In vivo cross-sectional ultrasonic imaging of human skin, Dermatologica, 177, 332, 1988.
314
Handbook of Non-Invasive Methods and the Skin, Second Edition
2. el-Gammal, S., Hoffmann, K., Auer, T., Korten, M., Altmeyer, P., Höss, A., and Hermert, H., A 50-MHz high-resolution ultrasound imaging system for dermatology, in Ultrasound in Dermatology, Altmeyer, P., elGammal, S., and Hoffman, K., Eds., Springer-Verlag, Berlin, 1992, p. 97. 3. New, K.C., Petroll, W.M., Boyde, A., Martin, L., Corcuff, P., Lévêque, J.L., Lemp, M.A., Cavanagh, H.D., and Jester, J.V., In vivo imaging of human teeth and skin using real-time confocal microscopy, Scanning, 13, 369, 1991. 4. Corcuff, P. and Lévêque, J.L., In vivo vision of the human skin with the tandem scanning microscope, Dermatology, 186, 50, 1993. 5. Hyde, J.S., Jesmanowicz, A., and Kneeland, J.B., Surface coil for MR imaging of skin, Magn. Reson. Med., 5, 456, 1987. 6. Querleux, B., Yassine, M.M., Darrasse, L., Saint-Jalmes, H., Sauzade, M., and Lévêque, J.L., Magnetic resonance imaging of the skin. A comparison with the ultrasonic technique, Bioeng. Skin, 4, 1, 1988. 7. Bittoun, J., Saint-Jalmes, H., Querleux, B., Darrasse, L., Jolivet, O., Idy-Peretti, I., Wartski, M., Richard, S., and Léêque, J.L., In vivo high-resolution imaging of the skin in a whole-body MR system at 1.5T, Radiology, 176, 457, 1990. 8. Zemtsov, A., Lorig, R., Bergfeld, W.F., Bailin, P.L., and Ng, T.C., Magnetic resonance imaging of cutaneous melanocytic lesions, J. Dermatol. Surg. Oncol., 15, 854, 1989. 9. Schwaighofer, B.W., Fruehwald, F.X.J., Pohl-Markl, H., Neuhold, A., Wicke, L., and Landrum, W.L., MRI evaluation of pigmented skin tumors. Preliminary study, Invest. Radiol., 24, 289, 1989. 10. Takahashi, M. and Kohda, H., Diagnostic utility of magnetic resonance imaging in malignant melanoma, J. Am. Acad. Dermatol., 27, 51, 1992.
11. Aygen, S., el-Gammal, S., Bauermann, T., Hartwig, R., and Altmeyer, P., Tissue characterization and 3-D visualization of human skin tumors by high resolution proton NMR-microscopy at 9.4 Tesla, in Eleventh Annual Meeting of the Society of Magnetic Resonance in Medicine, Berlin, 1992, p. 4605. 12. Bottomley, P.A., Hardy, C.J., Argersinger, R.E., and Allen-Moore, G., A review of 1H nuclear magnetic resonance relaxation in pathology: are T1 and T2 diagnostic? Med. Phys., 14, 1, 1987. 13. Breger, R.K., Wehrli, F.W., Charles, H.C., MacFall, J.R., and Haughton, V.M., Reproducibility of relaxation and spin-density parameters in phantoms and the human brain measured by MR imaging at 1.5 T, Magn. Reson. Med., 3, 649, 1986. 14. Richard, S., Querleux, B., Bittoun, J., Idy-Peretti, I., Jolivet, O., Cermakova, E., and Lévêque, J.L., In vivo proton relaxation times analysis of the skin layers by magnetic resonance imaging, J. Invest. Dermatol., 97, 120, 1991. 15. MacKenzie, I.C., The effects of frictional simulation on mouse ear epidermis. I. Cell proliferation, J. Invest. Dermatol., 87, 187, 1972. 16. Richard, S., Querleux, B., Bittoun, J., Jolivet, O., IdyPeretti, I., de Lacharrière, O., and Lévêque, J.L., Characterization of the skin in vivo by high resolution magnetic resonance imaging: water behavior and age-related effects, J. Invest. Dermatol., 100, 1993. 17. Querleux, B., Richard, S., Bittoun, J., Jolivet, O., IdyPeretti, I., Bazin, R., and Lévêque, J.L., In vivo hydration profile in skin layers by high resolution magnetic resonance imaging, Skin Pharmacol., in press. 18. Dixon, W.T., Simple proton spectroscopic imaging, Radiology, 153, 189, 1984. 19. Bellon, E.M., Haacke, E.M., Coleman, P.E., Sacco, D.C., Steiger, D.A., and Gangarosa, R.E., MR artifacts: a review, Am. J. Roentgenol., 147, 1271, 1986. 20. Zemtsov, A. and Dixon, L., Magnetic resonance in dermatology, Arch. Dermatol., 129, 215, 1993.
Intracutaneous 37 Spectrophotometric Imaging (SIAscopy): Method and Clinical Applications E. Claridge School of Computer Science, University of Birmingham, Birmingham, United Kingdom
S. Cotton Astron Clinica, The Mount, Cambridge, United Kingdom
M. Moncrieff and P.N. Hall Department of Plastic Surgery, Addenbrooke’s Hospital, Cambridge, United Kingdom
CONTENTS 37.1 Introduction............................................................................................................................................................315 37.2 Background ............................................................................................................................................................315 37.3 The Science of SIAscopy ......................................................................................................................................317 37.3.1 General Outline of the Method .................................................................................................................317 37.3.2 Optical Model of the Normal Skin ...........................................................................................................317 37.3.3 Forward Predictive Model of Skin Coloration..........................................................................................318 37.3.4 Inversion and Computation of SIAgraphs.................................................................................................319 37.3.5 Malignant Melanoma.................................................................................................................................319 37.3.6 Validation ...................................................................................................................................................320 37.4 Image Acquisition Using a SIAscope ...................................................................................................................320 37.5 Applications ...........................................................................................................................................................320 37.5.1 Clinical Applications .................................................................................................................................320 37.5.2 Nonclinical Applications ...........................................................................................................................321 37.5.3 Future Potential..........................................................................................................................................321 37.6 Discussion ..............................................................................................................................................................321 37.7 Conclusions............................................................................................................................................................324 References .......................................................................................................................................................................324
37.1 INTRODUCTION SIAscopy — Spectrophotometric Intracutaneous Analysis — is a new optical skin imaging method in which computer reconstructed images reveal a number of aspects of skin histology, namely the concentration of epidermal melanin, the concentration of dermal blood, the thickness of the papillary dermis and, in pathological cases, the presence of dermal melanin. Due to its ability to provide a unique insight into the skin histology in vivo, SIAscopy is becoming a preferred tool for the diagnosis of pigmented skin lesions and early melanoma detection.
Other applications, both clinical and not, are in the experimental stage of development.
37.2 BACKGROUND The study of any clinical dermatology textbook will reveal that color is an important diagnostic indicator (e.g., [1]). This is because the colors seen on the skin surface reflect many aspects of its internal structure and composition. For example, the reddening of the skin, erythema, indicates increased amounts of dermal blood. 315
316
Handbook of Non-Invasive Methods and the Skin, Second Edition
(a)
(d)
(b)
(e)
(c)
FIGURE 37.1 (a) A color image of a melanoma together with parametric maps showing. (b) Total melanin (darker = more). (c) Dermal melanin, whose presence suggest abnormality (brighter = more). (d) Papillary dermis, showing that collagen was displaced by melanocytes and the melanin they produce; also a peripheral increase indicating fibrosis in the region where cancer is developing actively (brighter = more). (e) Dermal blood, showing an absence in the centre, suggesting necrosis; also an increase on the periphery, suggesting angiogenesis in the area of active cancer growth (darker = more). These features are typical for melanoma and can be easily seen in the maps.
Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications
If the skin is considered as a physical system comprising components with particular optical properties, then its color can be analyzed by using the principles of physics. Considered in this way, skin color is a function of the spectral characteristics of light remitted from the skin. This, in turn, is the result of the interaction of light with the structures in the skin and depends on the optical properties of the skin components and the spectral composition of the light shone on the skin surface (the incident light). Given information about the optical properties of the skin, the color of the remitted light can be computed using physics equations describing a light transport theory [2–3]. In this way it is possible to predict the skin colors corresponding to a given particular skin structure. This approach has been used in a number of applications [4–6]. However, in SIAscopy the process is reversed, allowing predictions of the skin structure to be made from skin colors captured by a digital camera. From calibrated images captured in the red, green, blue and near-infrared parts of the spectrum, the computer reconstructs four images showing respectively the concentration of epidermal melanin, the concentration of blood within the papillary dermis, the thickness of collagen in the papillary dermis, and the presence of dermal melanin. These quantities are represented as parametric maps (SIAgraphs) which can be either viewed directly or subjected to further image analysis. Information contained in these images is complementary to that obtained through normal visual examination and thus it is likely to help in differential diagnosis. The method is totally non-invasive as it uses only visible and near-infrared light. As an example, Figure 37.1 shows an advanced melanoma and a set of SIAgraphs showing the concentration of dermal and epidermal melanin, blood and collagen thickness across the imaged
Histological component quantities
Light interaction
317
lesion and its surrounding skin. A detailed description and interpretation of the maps are given in Section 37.5.1
37.3 THE SCIENCE OF SIASCOPY 37.3.1 GENERAL OUTLINE
OF THE
METHOD
The key concept underpinning SIAscopy is a model of skin coloration. It is a set of correspondences between the parameters characterizing the skin and its colors. The model is constructed by computing the spectra remitted from the skin from parameters specifying its structure and optical properties, and then computing equivalent RGB colors. This step needs to be carried out only once. The model is then used to perform the inversion process, i.e., to infer the combination of histological parameters that lead to a particular color of tissue. Figure 37.2 shows schematically the process of image formation. As a final step, semi-quantitative information about the histological parameters is represented in a form of parametric maps (SIAgraphs). These are gray-scale images showing the magnitude of the parameters at each pixel of the original skin image.
37.3.2 OPTICAL MODEL
OF THE
NORMAL SKIN
The normal skin can be optically modeled as five layers, each with different optical properties. The stratum corneum is the most superficial layer of the epidermis, consisting of keratin-impregnated cells. Its thickness varies considerably between individuals as well as between different sites on the body. Optically it scatters the light forward, but it does not change its spectral characteristics.
Reflectance
Image acquisition
Image
Inversion (computational method)
FIGURE 37.2 The process of image formation. Histological components present in the skin give rise to a reflectance spectrum as a result of the light interaction with the tissue. The spectrum is specific to the particular skin composition. The remitted light is captured by a camera and stored as a digital image. The digital image is normally composed of three (red, green, and blue) or more spectral bands. SIAgraphs are derived by a computational method that takes the digital image as input and from it derives the quantities of the histological components present in the tissue.
Molar extinction coefficient (1000 L/Moles/mm)
318
Handbook of Non-Invasive Methods and the Skin, Second Edition
•
40 Haemoglobin Oxyhaemoglobin Melanin
30
• • •
20
• •
10
0 400
37.3.3 FORWARD PREDICTIVE MODEL COLORATION 600 800 Wavelength (nm)
1000
FIGURE 37.3 The absorption spectra for melanin, hemoglobin and oxyhemoglobin. In the model a mixture of 50% hemoglobin and 50% oxyhemoglobin has been used.
The remaining part of the epidermis is mostly cellular, containing dividing keratinocytes and basel cells, in addition to immune cells and melanocytes, all of which are transparent to the visible light. The melanocytes produce melanin, a pigment which strongly absorbs light in the blue part of the visible spectrum and in the ultraviolet (see Figure 37.3). All light not absorbed by melanin can be assumed to pass into the dermis, a distinct, deeper layer of the skin separated from the epidermis by a basement membrane known as the dermal-epidermis junction. The dermis consists of two structurally different layers, papillary and reticular, which also differ optically. The papillary dermis consists of loosly packed collagen fibers with a small diameter. This arrangement makes this layer highly backscattering so that a large proportion of incoming light is directed back toward the skin surface. The reticular dermis is composed of significantly larger collagen fibers arranged into tightly knit and highly organized bundles. The diameter of these bundles is comparable to or greater than the wavelengths of visible light. Structures of this size cause scattering which is highly forward-directed. Thus, any light which reaches this layer is passed on deeper into the subcutis, where is is absorbed and does not contribute to the spectrum remitted from the skin. The papillary dermis contains an extensive network of fine boood vessels, and the hemoglobins present in blood act as selective absorbers of light in both blue and green parts of the spectrum (see Figure 37.3). For the above description of the skin structure, the following parameters are required to model the optical properties of the normal skin: • •
the absorption coefficients for hemoglobin, μha (λ) the hemoglobin concentration, ch the scatter coefficient for collagen in the papillary dermis, μpd s the thickness of the papillary dermis collagen layer, dpd the scatter coefficient for collagen in the reticular dermis, μrds the thickness of the reticular dermis collagen layer, drd
a set of wavelength (λ) dependent absorption coefficients for melanin, μma (λ) the melanin concentration, cm
OF
SKIN
Given the above parameters, a remitted light spectrum can be computed using physics-defined light transport equations. The optical characteristics of the skin tissue are such that a number of different methods can be used. This work implements a two-flux method [7] based on KubelkaMunk theory [2]. It computes the remitted (R) and transmitted (T) light separately for each layer i: RI and Ti.
(1 − β ) ( e 2
Ri =
Ti =
(1 + β )
2
(1 + β )
2
e
Kd
e
Kd
Kd
− e− Kd
)
2
− (1 − β ) e− Kd 4β
where K = k ( k + 2 s ) , β =
2
− (1 − β ) e− Kd k , d is a layer thickk + 2s
ness, k ∝ μa and s ∝ μs. For an n layered system, values for R12…n and T12…n are computed recursively [8]: R12...n = R12...n−1 + T12,,,n−1 =
T122 ...n−1 Rn 1 − R12...n−1 Rn
and
T12...n−1Tn 1 − R12...n−1 Rn
The method requires that the incident light is diffuse. This condition is fulfilled because, as explained in Section 37.3.1, light is diffused as it passes through stratum corneum. A complete model of coloration for the normal skin is computed by considering all the possible values of parameters defining the skin. In the set of parameters given in Section 37.3.2, the absorption and scatter coefficients (μma (λ), μha (λ), μpds , μrds ) are treated as constants because they characterise the entire tissue type rather than a specific instance of the tissue. Their specific values are based on published data (e.g., [9]). The thickness of the reticular dermis, drd, can be assigned an arbitrary value because due to its strong forward scattering properties even a thin layer
Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications
will prevent any remission of light. The parameters characterizing a specific instance of the skin are the melanin and blood concentration cm and ch, and the thickness of the papillary dermis, dpd. Thus the model of skin coloration computes a color vector for each combination of the above three parameters, where the parameters are drawn from the entire range of histologically valid concentrations and thicknesses. In this way a cross-reference between histology and color is formed. The range of wavelengths used for computing the remitted spectra, from λ = 400 nm to 1300 nm, covers the whole visible spectrum and a small range of near infrared radiation. The wavelengths used for the computation are taken at equal intervals of 30 nm, giving 30 discrete points for each spectrum. The incident light is white, i.e., it has equal contributions from each discrete wavelength. The [r g b nir] vectors are derived from the computed spectra using a set of filter response functions equivalent to the physical filters used by the SIAscope.
37.3.4 INVERSION AND COMPUTATION SIAGRAPHS
OF
Figure 37.4 illustrates the relationship between the two reference systems: color, [r g b]; and histological param-
5 R Pap. Derm.
4 3.5 3
Blood
2.5 2 G
1.5 0
Melanin
B
2 4 6
3.5
3
2.5
TABLE 37.1 The Range of Histological Parameter Values Corresponding to the Normal Skin Parameter
Symbol
Range from
Range to
cm
0.00
8.69
ch
0.00
4.34
dpd
0.05
0.40
Epidermal melanin concentration (10–1 mMol/l) Haemoglobin concentration (in dermal blood) (mMol/l) Thickness of the papillary dermis (10–4 m)
eters: melanin concentration, blood concentration and thickness of the papillary dermis, [cm ch dpd]. These relationships constitute the model of skin coloration. Since there is a one-to-one mapping between the colors and the parameters [10], the parameter values can be retrieved from the model, given the color vector obtained from each point in a color image of the skin. The magnitude of each parameter is then displayed at each pixel location in a separate image, giving three SIAgraphs: melanin, hemoglobin and papillary dermis, as shown in Figure 37.1 and Figure 37.6(b), (d), and (e).
37.3.5 MALIGNANT MELANOMA
Model of skin colouration
4.5
319
2
1.5
1
0.5
0
FIGURE 37.4 A graphical representation of the model of skin coloration that encodes the relationships between the quantities of histological components and the RGB colors. The main axes labeled R, G and B represent the color system. The histological parameter system is represented by smaller axes labeled Melanin, Blood and Pap. Derm. Each dot represents a specific instance of the skin tissue. It is characterized by the specific quantities of melanin, blood and thickness of the papillary dermis, as indicated by its position in the histological parameter system; and by the specific color, as indicated by its position in the color system. During the model construction the colors are computed from the parameters using light transport equations. During the image interpretation, the parameters corresponding to the image colors are “looked up” from the model.
The model of coloration described above represents all instances of the normal skin. Differences related to racial type, age, or gender are all in terms of the quantities of pigments and thicknesses of the layers and hence are all represented in the model. However, certain pathologies change the structure of the skin in such a way that it is no longer represented by the normal model. An example is malignant melanoma where malignant melanocytes have penetrated into the dermal layer and created melanin deposits there. Optically, this not only increases the overall melanin concentration at the site of the deposit, but also decreases the total scatter as melanin replaces the highly back-scattering collagen in the papillary dermis. The colors of the skin with melanin deposits in the dermis no longer fit the model of normal coloration. In the schematic representation shown in Figure 37.4 such colors appear outside the “surface” which corresponds to a particular thickness of the papillary dermis. This non-conformance to the model identifies colors as “abnormal” and provides a highly sensitive diagnostic sign. The presence of dermal melanin is represented in the fourth SIAgraph (Figure 37.1 and Figure 37.6(c)). There are also a number of histological changes associated with melanoma, such as angiogenesis and fibrosis, which induce color changes within the bounds of the normal coloration. These secondary signs have been found diagnostically important, as will be described in Section 37.5.1.
320
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 37.2 Percentage Errors Associated with the Recovery of the Three Main Parameters, Excluding Zero Concentrations
Melanin Blood Papillary dermis
mean
st dev
median
14.0 5.0 3.3
14.0 8.2 3.4
13.9 3.4 3.2
37.3.6 VALIDATION One of the advantages of physics based methods is that it is possible to estimate the accuracy of their results if errors on the input data can be estimated. The two main sources of error associated with the computation of the parametric maps are the uncertainties in the absorption and scatter coefficients (σspec) [9] and the camera digitisation error (σcam), derived from the signal-to-noise ratio (SNR) characteristics of a particular camera. As the two sources are independent, the respective errors can be combined to give the overall error estimate [10]. 2 σ pk = σ spec + σ 2cam
Errors associated with each of the skin parameters can then be calculated using standard error propagation [11]. Table 37.2 shows the percentage errors derived in this way. Practical validation has established that, firstly, the skin colors predicted by Kubelka Munk equations were consistent with skin colors measured from the skins of volunteers of various racial origins [7]. The validity of parametric maps was established through indirect comparisons and measurements [12–13].
37.4 IMAGE ACQUISITION USING A SIASCOPE The SIAscope is a commercial, clinically approved device [14] operating according to the principles of the imaging method described above. Images of the skin can be taken over an area of 24 × 24 mm or 12 × 12 mm, corresponding to spatial resolutions of 40 μm or 20 μm per pixel respectively. Images are acquired using a lightweight handset connected to a laptop via a IEEE-1394 connection. A number of spectrally filtered images are obtained in the range between 400 nm to 1000 nm, calibrated and processed to compute the parametric maps. These are then displayed in the form of SIAgraphs on the screen together with the color image of a lesion. SIAscope II is shown in Figure 37.5.
FIGURE 37.5 The SIAscope II — a commercial device for skin imaging and computation of SIAgraphs. Images are acquired using a handset (at the bottom right of the case) connected to a laptop computer. The image interpretation is implemented on the computer and SIAgraphs are displayed on its screen.
37.5 APPLICATIONS 37.5.1 CLINICAL APPLICATIONS The main current application area for the SIAscope is an aid for diagnosis of pigmented skin lesions, and in particular for early detection of cutaneous melanoma. It is particularly suited for this role because the SIAgraphs make explicit a number of primary and secondary signs associated with this cancer. Table 37.3 lists the features together with their sensitivity and specificity as melanoma predictors. The lesion shown in Figure 37.1 shows all the characteristic signs of invasive cutaneous melanoma. The presence of melanin in the dermis is a very sensitive sign, indicating invasion into the papillary dermis. Also coinciding with dermal melanin is a “collagen hole,” an area with no papillary collagen present, indicating destruction or replacement of the papillary dermis by melanoma. The collagen map also shows fibrosis (the increase in collagen thickness and irregularity) on the lesion periphery, often associated with early, invasive melanoma [1]. The blood distribution map shows the increase in blood levels on the lesion periphery, (erythematous blush), which is indicative of angiogenesis, often associated with invasive skin tumors. In addition, there is a total lack of blood in the lesion center in the area that coincides with the dermal melanin, further indicating destruction or replacement of the papillary dermis. While other benign skin lesions may demonstrate either erythematous blush or blood displacement, it is the
Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications
321
A lesion is suspicious if it has TABLE 37.3 Diagnostic Features Shown in SIAgraphs and Their Individual Sensitivity and Specificity as Melanoma Predictors Diagnostic Feature Presence of dermal melanin Areas within the lesion with no blood present (“blood displacement”) Increase in blood level on the lesion periphery (“erythematous blush”) Areas within the lesion with no collagen present (“collagen holes”) Asymmetry
Sensitivity (%)
Specific (%)
Dermal melanin Dermal blood
96.2
56.8
75.0
70.3
Dermal blood
75.0
65.5
Collagen thickness
78.8
74.0
Total melanin
76.9
62.2
SIAgraph
presence of the two signs together in the same lesion that is highly specific for invasive melanoma. The lesion in Figure 37.6 is small and fairly pale, however, the SIAgraphs strongly suggest that it is a malignant case. Many significant features are present, including dermal melanin, blood displacement with erythematous blush together with evidence of fibrosis. Both lesions were histologically diagnosed as melanomas. A number of larger studies have demonstrated that SIAscopy is an excellent diagnostic aid. Diagnosis using three SIAscopy features (dermal melanin, blood displacement and peripheral blush) together with the assessment of the lesion size yielded sensitivity of 80% and specificity of 83% on a set of 348 lesions [15]. Diagnosis using a simple scoring method that considers two SIAscopy features (dermal melanin and blood displacement with peripheral blush) together with the assessment of the lesion size and the patient’s age yielded on the same lesion set sensitivity of 90.4% and specificity of 74.0% [12]. Another study, using a scoring method which incorporates patient’s age, lesion diameter, the presence of dermal melanin, collagen hole and blood displacement with peripheral blush, achieved specificity of 87% (78%) for sensitivity level of 80% (100%) respectively, on a set of 214 lesions [16]. The most recent study [17] have shown that the SIAscope outperforms dermatoscopy in the diagnosis of melanoma producing a sensitivity of 91% and specificity of 91% compared with a sensitivity of 91% and specificity of 54% for a panel of experts using dermatoscopy. This study introduced the assessment of the dermo-epidermal junction to the analysis of a pigmented lesion and produced a simple clinical checklist:
•
An irregular dermal epidermal junction
•
Dermal melanin with any of
or
• Irregular dermal melanin • Presence of linear vessels or dots • Regression structures
The use of the SIAscope has also been investigated in the assessment of psoriasis and eczema [18–19] where it has been shown to be useful in both differential diagnosis and the assessment of treatment regimes.
37.5.2 NONCLINICAL APPLICATIONS Although the imaging method was designed with clinical applications in mind, the generic nature of the underpinning science means that it is potentially applicable in many different areas. For example, in a cosmetics application, a modified method was used to image facial skin in order to detect spots and blemishes [20]. Images in Figure 37.7 show the original image together with SIAgraphs of blood and melanin. The blood image shows the increased levels in the lips and also clearly shows a number of “spots.” The melanin image highlights a number of small moles (common nevi). In a sport sciences applications, a small study was carried out in which the blood supply levels at various body locations were measured as a function of time during strenuous exercise. In this application SIAscopy provided the scientists with a noninvasive indirect method of studying a thermoregulatory mechanisms in the human body. Figure 37.8 shows four blood SIAgraphs of two subjects.
37.5.3 FUTURE POTENTIAL Due to its physical and physiological underpinnings, SIAscopy is a truly generic imaging method. The range of its applications is growing steadily. In the medical field the use of the method to determine excision margins for basal cell carcinoma [21] is currently under investigations. Preliminary studies have also begun on the monitoring of wound healing (Figure 37.9), the development of ulcers and the assessment of port wine stains prior to laser treatment. The ability to understand the composition and structure of the skin has also been of interest in the assessment of cosmetic treatments for the skin.
37.6 DISCUSSION The method in its current form is principally two-dimensional, and SIAgraphs show the surface distribution and the magnitudes of individual tissue components. This
322
Handbook of Non-Invasive Methods and the Skin, Second Edition
(a)
(d)
(b)
(e)
(c)
FIGURE 37.6 (a) A color image of an early stage melanoma; parametric maps showing. (b) Total melanin (darker = more). (c) Dermal melanin (black = none, brighter = more); small speckles of dermal melanin can be seen in the center of the lesion. (d) Thickness of the papillary dermis (brighter = more); note the increased amount in all areas of increased epidermal melanin and one very distinct nest in the vicinity of the dermal melanin deposits. (e) Blood (darker = more); note an erythematous reaction in the area corresponding to dermal melanin and a small decrease in the area corresponding to the dermal melanin deposits.
Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications
(a)
(b)
aspect makes it different from tomographic methods (e.g., OCT [22] and confocal microscopy [23]) which are capable of showing the three-dimensional structure of the skin. The depth resolution of SIAscopy is limited by both absorption and scatter within the skin tissue. Its resolution within-the-surface varies depending on scatter, so that small details are not always resolved. However, the maps showing the gross distribution of histological parameters are still providing useful diagnostic insight into the tissue properties.
323
(c)
FIGURE 37.7 (a) A color image of the face. (b) A blood SIAgraph showing hemoglobin concentration across the image. (c) A total melanin SIAgraph showing melanin concentration across the image. The region of interest 1 shows the location of a skin “spot”, note that it is clearly visible in the blood image, but not in the melanin image. The region of interest 2 shows the location of a small “mole” (common nevus); note that it is clearly visible in the melanin image but not in the blood image.
The detection of the presence of dermal melanin in vivo is one of the unique features of SIAscopy, which in principle can detect melanin penetration at depths smaller than 30 μm from the dermal–epidermal junction. However, the accuracy of measurement decreases with increase in depth and increase in melanin concentration. Although this is a limitation in a general sense, it does not adversely affect the method’s capability to detect the very thin, early melanomas that are still a challenge to other methods. The encouraging diagnostic results achieved with the aid of SIAscopy can be attributed to the inclusion of the clinically appropriate variables in the scoring methods. For instance, increased diameter of a pigmented lesion is a risk factor for early melanoma actively enlarging; the presence of dermal melanin and blood displacement with erythematous blush are strong risk factors for early, invading melanoma and, like most cancers, the risk of developing melanoma increases with age. Further advantage of these features is that they are rapidly identifiable in a repeatable and reproducible manner. Studies comparing intra- and interobserver error of SIAscopy features demonstrated “excellent” or “almost perfect” scores in their identification [12]. This, together with the speed with which these features can be learned, is in direct contrast with the identification of features employed by other methods in the diagnosis of cutaneous melanoma [26, 27].
324
Handbook of Non-Invasive Methods and the Skin, Second Edition
At rest
Exercise time
8 mins
FIGURE 37.8 Four blood SIAgraphs from two subjects showing the effect of exercise on superficial blood supply.
and thus increase understanding as to why various skin diseases manifest themselves through particular visual signs. As SIAgraphs show views directly related to histology, they can be easily understood by medical personnel and their interpretation does not require much training. A substantial body of experimental work showed that the method can help improve the diagnosis of melanoma. A range of further applications, in skin imaging and beyond [24–25], is currently being investigated.
REFERENCES
FIGURE 37.9 A blood SIAgraph taken around the edge of a healing punch biopsy wound (top left). Small punctate blood vessels can be seen in the surrounding skin. They have formed due to the release of angiogenic stimulants in the wound vicinity. As the wound heals these dots can be seen “stretching” into linear vessels around the wound edge.
37.7 CONCLUSIONS SIAscopy is a novel skin imaging method showing significant promise as a tool for in vivo inspection of the normal and abnormal skin. Unlike traditional image analysis methods, which operate primarily on the image itself, this method adds and exploits a physics based model of skin coloration. SIAgraphs reconstructed from color images using the model of coloration, characterize the internal structure and composition of the skin rather than its external appearance and thus are far more informative than images obtained by simpler imaging methods. In clinical applications, the model can explain the skin appearance
1. Menzie, S.W, Crotty, K.A., Ingvar C. et al. An Atlas of Surface Microscopy of Pigmented Skin Lesions: McGraw-Hill, 1996. 2. Egan, W.G., Hilgeman, T.W. Optical Properties of Inhomogeneous Materials: Academic Press, 1979. 3. Prahl, S.A., Keijzer, M., Jacques, S.L., and Welch, A.J. A Monte Carlo model of light propagation in tissue. SPIE Institute Series IS, 5:102–111, 1989. 4. Angelopoulou, E., Molana, R., and Danilidis, K. Multispectral skin colour modelling, in IEEE Conference On Computer Vision and Pattern Recognition, 635–642, 2001. 5. Tsumurra, N., Haneishi, H., and Miyake, Y. Independent component analysis of skin color image, J Am Opt Soc 6, 2169–2187, 1999. 6. Shimada, M., Masuda, Y., Yamada, M.Y. et al., Explanation of human skin color by multiple linear regression analysis based on the Modified Lambert-Beer law, Opt. Rev. 7, 348–352, 2000. 7. Cotton, S.D., Claridge, E. Developing a predictive model of human skin colouring. Proceedings of the SPIE Medical Imaging 1996, (Vanmetter, R.L., Beutel, J. Eds.), vol. 2708, 815–825, 1996. (Available from http://www.cs.bham.ac.uk/~exc/Research/Papers/spie9 6.pdf.)
Spectrophotometric Intracutaneous Imaging (SIAscopy): Method and Clinical Applications
8. Spiegel, M.R. Theory and Practice of Advanced Calculus: McGraw-Hill, 1962. 9. Anderson, R., Parrish, B.S., Parrish, J. The optics of the human skin. J Invest Dermatol 77(1), 13–19, 1981. 10. Claridge, E., Preece, S.J. An inverse method for the recovery of tissue parameters from colour images. Information Processing in Medical Imaging (IPMI), Taylor C and Noble JA (Eds.) LNCS 2732, 306–317. Springer, 2003. (Available from http://www.cs.bham.ac.uk/~exc/ Research/Papers/ipmi03.pdf.) 11. Kendall, M.G. and Stuart, A. The Advanced Theory of Statistics. Volume 1: Distribution Theory, 3rd Ed. London: CharlesGriffin & Co. Ltd, 1969. 12. Moncrieff, M. The Clinical Application of Spectrophotometric Intracutaneous Analysis for the Diagnosis of Cutaneous Malignant Melanoma. MD Thesis, The University of East Anglia, 2001. 13. Claridge, E., Cotton, S., Hall, P., Moncrieff, M. From colour to tissue histology: Physics based interpretation of images of pigmented skin lesions. Med Image Anal 7(4), 489–502, 2003. 14. www.siascope.com 15. Moncrieff, M., Cotton, S., Claridge, E., Hall, P. Spectrophotometric intracutaneous analysis — a new technique for imaging pigmented skin lesions. Br J Dermatol 146(3), 448–457, 2002. 16. Powell, J., Moncrieff, M., Hall, P. Assessment of pigmented skin lesions using SIAscopy and Dermoscopy. European Academy of Dermatology & Venerealogy Annual Meeting, Prague, 2002. 17. Schultz, H.J., Klaas, S. SIAscopy for in vivo diagnosis of atypical pigmented skin lesions superior to dermoscopy. Submitted to the J Invest Dermatol. 18. Novakovic, L. and Hawk, J. Spectrophotometric intracutaneous analysis: a novel technique in the differential diagnosis of psoriasis and eczema. Br J Dermatol 147(suppl. 62), 104, 2002.
325
19. Novakovic, L. and Hawk, J. Assessing the treatment efficacy of Puva treatment for psoriasis using spectrophotometric intracutaneous analysis. J Eur Acad Dermatol Venereol , 2003. 20. Preece, S.J., Cotton, S., Claridge, E. Imaging the pigments of human skin with a technique which is invariant to changes in surface geometry and intensity of illuminating light. Medical Image Understanding and Analysis 2003, Barber, D., Ed., 145148, 2003. 21. Bjerring, P., Obitz, E.R., Cotton, S. In vivo spectrophotometric evaluation of the skin tumours using a new skin chromophore imaging system SIAscope. Melanoma Res 11, S180, 2001. 22. Happe, M., Hoffmann, K., v. Düring, M., et al. Optical coherence tomography (OCT): a new noninvasive diagnostic imaging technique in dermatology. Melanoma Res 7, 84–85, 1997. 23. Rajadhyaksha, M., Grossman, M., Esterowitz, D., Webb, R. In-vivo confocal scanning laser microscopy of human skin — melanin provides strong contrast. J Invest Dermatol 104, 946–952, 1995. 24. Orihuela, F., Claridge, E. A new optical imaging method for the ocular fundus. Med Phys 30, 1540, 2003. 25. Hidovi, D. and Claridge, E. Predictive model of colon tissue colouration based on the physics of image formation. Submitted to IEEE International Symposium on Biomedical Imaging 2004. 26. Binder, N., Schwartz, N., Winkler, A. et al., Epiluminescence microscopy. A useful tool for the diagnosis of pigmented lesions for formally trained dermatologists. Archives Dermatol 131, 286–291, 1995. 27. Morton, C. and Mackie, R., Clinical accuracy of the diagnosis of cutaneous malignant melanoma. Br J Dermatol 138, 283–287, 1998.
Epidermis Hydration
Hydration: Measurement of 38 Epidermal High-Frequency Electrical Conductance Hachiro Tagami Department of Dermatology, Tohoku University School of Medicine, Sendai, Japan
CONTENTS 38.1 Introduction ..........................................................................................................................................................329 38.2 Objective...............................................................................................................................................................329 38.3 Skin Impedance ....................................................................................................................................................330 38.4 Methodological Principle of the High-Frequency Method .................................................................................330 38.5 Assessment of Accuracy ......................................................................................................................................331 38.6 Sources of Error ...................................................................................................................................................331 38.7 Measurements in Normal Skin under Various Conditions..................................................................................332 38.8 Measurements in Lesional Skin...........................................................................................................................332 38.9 Water Sorption–Desorption Test of the Skin Surface In Vivo.............................................................................333 38.10 Assessment of the Efficacy of Skin Moisturizers ...............................................................................................334 38.11 Correlation with Other Methods..........................................................................................................................334 38.12 Recommendations ................................................................................................................................................334 References .......................................................................................................................................................................334
38.1 INTRODUCTION The primary function of the epidermis is to produce the stratum corneum (SC) that protects our body from desiccation and invasion of various kinds of external attacks. It consists of about 15 to 20 tightly stacked layers of corneocytes, flattened dead bodies of epidermal cells. Although the SC is a thin membrane only about 20 μm thick, it is an efficient barrier to water and other substances. Hence, even in a dry environment, only a small amount of water is lost from the body. Surprisingly, just beneath it, hydrated living epidermal tissue functions to sustain our existence. The SC plays another important role for human skin. It always keeps our skin surface soft and smooth. It allows free body movement without the skin surface becoming cracked or fissured. About 50 years ago, Blank1 made interesting observations in vitro. He found that the isolated fragments of plantar SC became hard and brittle when dehydrated. Attempts to soften them with petrolatum or olive oil, which are clinically used for the treatment of rough, scaly skin, completely failed. Only after absorption
of water did they become soft and flexible, and we can regard water as the ultimate moisturizer that improves subjective perception of the mechanical properties of human skin. By contrast, there is always a water supply from the underlying hydrated living tissue in vivo, even in an atmosphere with extremely low relative humidity (RH), from which it is difficult for the SC to take up water. Despite the lack of water, occlusive agents such as petrolatum exert a softening and smoothing clinical effect on the skin surface by preventing the water loss. Thus, maintaining an appropriate water content in the SC is an important clinical and cosmetic concern.
38.2 OBJECTIVE It is easy to measure the water content of the SC in vitro by simple gravimetry.1 In contrast, it is difficult to measure the absolute amount of water contained in the SC in vivo because of the presence of a concentration gradient of water within the SC.2,3 Until 1980, we lacked adequate methodologies to assess the state of skin surface hydration in vivo. Recent urgent demands for a practical technique 329
330
Handbook of Non-Invasive Methods and the Skin, Second Edition
to objectively evaluate the efficacy of different moisturizers have prompted development of various modalities of techniques that measure the skin properties that are influenced by the water content of the SC, e.g., electrical, mechanical, thermal, and spectroscopic properties. These techniques measure water content at poorly defined locations within skin, and with few exceptions, they have provided only qualitative information on changes in water content. Most of all, we should know that there exists a concentration gradient of water within the SC in vivo, highest in the lowermost layer and lowest in the uppermost portion. The SC is the rate-limiting barrier between the watersaturated viable tissue and the dry outer environment, and diffusion of water takes place as a purely passive process through the SC.2,3 The superficial portion of the SC remains supple and flexible, as long as its water-holding capacity is intact. Small, water-soluble metabolites, proteinaceous structural components, and sebum constitute the main components that bind water in the SC.4 Ceramides, the main intercellular lipid component of SC, which is a key factor in the SC barrier function,5 also play a crucial role in the water-holding capacity of the SC by preventing easy water passage through it.6 Moreover, sebum that covers the skin surface prevents water evaporation,7 and its component, glycerol, plays also an important role with its high hygroscopic properties.8 By contrast, SC deficient in the water-holding capacity, such as that found in pathologic skin conditions, becomes brittle enough to break on flexing or stretching, which results in fissures and scaling, even under the normal ambient conditions of temperature and humidity.9 The efficacy of various topical agents or cosmetics depends mainly on their effectiveness in increasing the skin surface hydration state, which helps to recover softness and smoothness in dry skin. Thus, we urgently need techniques to measure the hydration state of the exposed portion of the SC, the skin surface hydration state, quickly and quantitatively.
38.3 SKIN IMPEDANCE Among the diversity of techniques to evaluate the hydration state of the skin surface, the most widely used ones are those involving the measurement of skin impedance. Impedance (Z), the total electrical opposition to the flow of an alternating current, depends on two components, resistance (R) and capacitance (C), and their relationship may be formulated as follows: Z = [R2 + (1/2 π fC)2]1/2 where f stands for a frequency of an applied alternating current. In the past, for reasons of technical simplicity, many researchers studied the impedance of human skin.
Tregear10 speculated that when the skin surface is not deliberately hydrated, the reciprocal of specific impedance should be a measure of the hydration of its surface position. However, most measurements of the skin impedance used to employ damp contact using electrode paste between the electrodes and the skin because of the high impedance of human skin, which is chiefly due to the properties of the SC.11 Certainly, such an approach not only is cumbersome, but also has a great influence on the later functional properties of the SC. We found that by employing dry electrodes, we can evaluate the hydration state of the skin surface quickly and quantitatively in a noninvasive way using high-frequency electric current.12 Leveque and de Rigal13 also reported good sensitivity to the water content of the skin surface of a similar instrument developed by them.
38.4 METHODOLOGICAL PRINCIPLE OF THE HIGH-FREQUENCY METHOD The instrument used by us operates at 3.5 MHz. It depends on the high-frequency apparatus developed by Masuda et al.14 that enables us to measure conductance (= 1/R) and capacitance separately. It consists of a main recording body and a long flexible cable, at the end of which a probe is attached. The skin probe consists of two concentric electrodes of 1 and 6 mm external diameter, respectively, separated by a dielectric. As soon as the probe is placed on the skin, both conductance and capacitance show a rapid initial increase for a few seconds, followed by a gradual increase if the contact is maintained. The level of the initial increase represents the hydration state of the skin surface at the time of application of the probe, and the later slow increase is due to accumulation of water beneath the probe resulting from transepidermal water loss. Thus, we use the initial value for the evaluation of the hydration of the skin surface. In general, skin conductance and capacitance show a very similar behavior (r = 0.95; p < 0.01). The only exception is the palmoplantar skin surface, where the values of capacitance are disproportionately low compared with conductance.12 Thus, we can assess the hydration state of the skin surface more accurately with the measurement of conductance alone, in terms of micro-mho or microsiemens. Our later study also demonstrated that high-frequency measurement correlated much better with the hydration dynamics taking place in normal skin surface than capacitance recording with a Corneometer.15 However, the latter seems to be superior to the high-frequency conductance measurement for the measurements of hydration changes that occur in extremely dry skin surface, such as scaly psoriatic lesions. The original skin surface hygrometer (Skicon-200, IBS Ltd., Hamamatsu, Japan) was a commercially
Epidermal Hydration: Measurement of High-Frequency Electrical Conductance
FIGURE 38.1 Skin surface hygrometer and a probe incorporated with a graduated spring mechanism.
Conductance (micromho) 0
1000
2000
Bulla Erosion Crust Scale Adjacent normal skin
FIGURE 38.2 High-frequency conductance values measured at various skin lesions (dotted column) in comparison with those obtained at adjacent normal skin. Data are expressed as a mean and standard deviation (bar).
available model of this apparatus. It automatically records measured values 3 seconds after application of a probe 6 mm in diameter (Figure 38.1). The reason for the relatively small size of the probe is that investigators can minimize the occlusive effect of the electrode placed on the skin. We can replace the conventional probe with a more sensitive one that contains a larger central electrode 2 mm in diameter; the sensitivity is three times greater than that of the conventional probe. Currently, only one other similar instrument (Skicon 200EX), which has a built-in computer, is commercially available from the same company.
38.5 ASSESSMENT OF ACCURACY As mentioned above, there is a concentration gradient of water within the SC. The isolated piece of uniformly hydrated SC, which has been used in all experiments on the hydration state of the SC so far,1 cannot be used as an in vivo model of SC. We have devised a simple simulation model of an in vivo SC, in which a concentration gradient of water exists between the surface and the lowermost portion.16 It consists of an isolated sheet of SC that tightly
331
occludes the underlying water-saturated filter paper placed as in a diffusion chamber. The filter paper is mounted on a glass slide, and all free edges of the SC sheet are sealed to the glass with a removable frame of adhesive vinyl tape. The surface of the SC is exposed to the ambient atmosphere, and passage of water is allowed only through this portion of the SC. The underlying water-saturated filter paper, like fluid-saturated cutaneous tissues in vivo, is the water source of overlying SC and is also the conducting medium that allows the formation of an adequate electric field. Conductance values recorded with only one sheet of underlying filter paper were quite low. With an increase in the number of sheets of paper, the conductance value increased until five sheets of paper were in place. At this point the readings reached a plateau. The total thickness of five sheets of filter paper saturated with water was approximately 5 mm. Thus, to obtain an optimal reading, the high-frequency current of 3.5 MHz should extend at least 5 mm into the wet and electrically conductive substances. This condition is always attained in vivo. We performed gravimetric determination of water content in the SC, together with the high-frequency conductometry. As a result, we confirmed that the recorded conductance values correlated well with the actual water content of the SC (r = 0.94). Moreover, by using a model consisting of five overlapped SC sheets instead of one, we corroborated that there is a close correlation between the high-frequency conductance values and the water content in the uppermost SC sheet (r = 0.98).
38.6 SOURCES OF ERROR In this instrument the high frequency of 3.5 MHz flows between the concentric electrodes via skin tissue. Thus, close contact between the probe and the skin surface is an important factor to obtain reproducible results. We have also found that even a small additional manual pressure on the probe greatly increases the observed values. To obtain reproducible results, the probes of commercially available skin surface hygrometers are now incorporated with a graduated spring mechanism to ensure the same pressure to be applied to the skin each time (Figure 38.1). Even with it, however, sufficient fitting with irregularly contoured skin surface is difficult, particularly when the skin surface is dry and firm, and it should be kept in mind that the recorded values tend to indicate a lower hydration state than the actual one when the measurements are performed on lesional skin covered by scales. To obtain better contact even with a rough, scaly skin surface or hairy skin such as the scalp, replacement of the probe with an interdigitated electrode (MT-8C probe, Measurement Technologies Co., Cincinnati, OH) has been developed.17 When it is used in hydration measurements
332
Handbook of Non-Invasive Methods and the Skin, Second Edition
of dry, scaly skin surface such as atopic xerosis and senile xerosis, it seems to be comparable or more sensitive than the capacitance recording with a Corneometer (CM420, Khzaka, Cologne, Germany). Because the SC is exposed to the atmosphere, its surface hydration state is greatly influenced by the ambient relative humidity. Even simply breathing upon the skin induces an instantaneous increase in conductance. Thus, covered areas such as the trunk show much higher values than the exposed areas in the dry winter season due to the effect of thick, airtight clothes, when the measurement is performed just after removal of the clothes. At least 15 minutes should be allowed for the skin surface to adapt to the atmosphere. Measured conductance values sometimes vary greatly, even between sites only slightly apart from each other, particularly when the probe is applied to the sites rich in sweat glands such as the palmoplantar surface, axilla, and, in some individuals, even the forehead. Generally, the highest values are found on the face and the lowest values on the distal portion of the limbs on the body surface. We should avoid the skin areas such as the palmoplantar skin for comparative study, because these areas are always under the influence of mental sweating. Our comparison made between summer and winter in the same subjects at our outpatient clinic with 21 to 26˚C room temperature showed that high environmental relative humidity of the summer, and possibly invisible sweating, induced higher conductance values in the summertime.18 Therefore, the measurements should ideally be conducted in the identical environmental conditions, namely, by using a climate-controlled chamber. Such a study performed in identical subjects consisting of different age groups using a climate-controlled chamber maintained at 21˚C and 50% relative humidity definitely demonstrated that the high-frequency conductance is significantly lower in winter than in summer either on the exposed area (the cheek) or on the covered area (the flexor surface of the forearm). These data indicate that the skin surface conditions are greatly influenced by weathering of the skin in the dry and cold winter.19 An exogenous supply of water on a skin surface results in a remarkable elevation in conductance value. However, this increase is not influenced whether we apply distilled water or a highly concentrated buffer solution; namely, it is not affected by the presence of other electrolytes in the applied water due to the fact that the skin surface is already rich in various electroconductive substances. (To totally remove such electroconductive substances from a fragment of SC, it was necessary to soak it in distilled water for at least 14 days.)
38.7 MEASUREMENTS IN NORMAL SKIN UNDER VARIOUS CONDITIONS Removal of the SC from normal skin surface by stripping with adhesive cellophane tape demonstrates clearly that the principal hydration detected by this method is that in the outermost portion of the SC. The conductance value progressively increased as deeper layers of the SC were serially exposed by cellophane tape stripping, eventually reaching a certain high level, which presumably represents the water content of the fully hydrated viable epidermis.12 Again, such a progressive increase after serial tape stripping is much smaller with the corneometry with capacitance evaluation than with the conductometry.15 It took about 14 days for the elevated conductance to return to the levels of the adjacent normal skin; it was the time point when the stripped skin finally resumed an almost no-scaly appearance.9 Removal of skin surface lipids with ethyl ether or acetone induces a marked decrease in conductance corresponding to the length of the lipid extraction. Extraction for 3 minutes was required to reach a nadir.18 Recovery to the untreated levels was noted after 30 minutes on the skin rich in sebaceous glands, such as the facial skin or the scalp of adults, in contrast to other areas, where it took a much longer time.20 Thus, skin surface lipids play a crucial role in maintaining the skin surface hydration state.8,20 Children show low skin conductance levels among various age brackets, probably reflecting the low sebum secretion rate.21 Much lower levels are found in neonates.22,23 This neonatal xerosis disappears within 2 weeks of birth.23 In addition to the fact that they cannot sweat properly, even in a warm environment, their SC, which has been under continuous exposure to the amniotic fluid, shows defective water-holding capacity. Xerotic skin in elderly people shows decreased conductance in relation to clinical severity.24 There was a significant correlation between the reduced amino acid content and skin conductance level.25 However, photoaging-associated alterations in the SC function seem to occur only mildly by chronic sun exposure, which causes skin photoaging on the exposed skin areas.26 There is no marked sexual difference in conductance values when compared in the same age groups.
38.8 MEASUREMENTS IN LESIONAL SKIN Accumulation of tissue fluid beneath normal SC does not affect the readings made on the skin surface as long as the covering SC is intact. There is hardly any difference between bullous lesions covered by intact SC and adjacent normal skin (Figure 38.3). In contrast, the changes in the SC greatly affect the results. Scaly lesions noted in various dermatoses always show lower conductance values than those recorded in the adjacent normal skin27 (Figure 38.4).
Epidermal Hydration: Measurement of High-Frequency Electrical Conductance
(mS) 1500
n = 10 Control Vesicle 1000
Water
500
0
0
30
60
90
120 Time (sec)
FIGURE 38.3 Water sorption–desorption test performed on bullous lesions covered by intact stratum corneum. No significant differences are noted before and after application of a water droplet for 10 seconds on the skin surface and at measurements conducted thereafter every 30 seconds, up to 120 seconds. Data are expressed as a mean (circle) and standard deviation (bar).
(mS)
1000
n = 10 Control Thin scale
Water
500
0
0
30
60
90 120 Time (sec)
FIGURE 38.4 Water sorption–desorption test performed on the skin covered by thin scale. Significantly lower conductance values are observed before and after application of a water droplet for 10 seconds and at measurements conducted every 30 seconds thereafter, up to 120 seconds. Data are expressed as a mean (circle) and standard deviation (bar). *, p < 0.05; **, p < 0.01.
Even in the lesions of pityriasis alba, rough and hypopigmented macules on the face and trunk of children, where it is difficult to visualize clearly the presence of very fine scales on the skin surface, we can find low conductance levels.28 This is also the case with atopic xerosis, the dry, clinically noninflamed skin of patients with atopic
333
dermatitis.27,29 Such skin shows deficient water barrier function measured as transepidermal water loss (TEWL). By performing simultaneous measurements of TEWL and skin conductance in patients who had scaly lesions of various grades of severity, such as eczematous dermatitis or psoriasis, we obtained data indicating that there is an inverse relationship between these parameters, i.e., between the water barrier function of the SC and skin surface hydration state.9 Similar to the removal of the SC by tape stripping to expose viable epidermis, even a small scratch wound in the test area results in a marked increase in conductance. Measurements in erosive lesions always yield a prominently high reading. Even in such lesions conductance becomes almost zero after the formation of dry crust.9 In contrast, the skin surface hydration state is much higher on fresh scars, hypertrophic scars, and keloids, where immature corneocytes due to an enhanced SC turnover rate play a much more important role in the barrier dysfunction than do changes in intercellular lipids.30,31
38.9 WATER SORPTION–DESORPTION TEST OF THE SKIN SURFACE IN VIVO The principal hydration detected by this method is in the outermost portion of the SC, which is always under the influence of the relative humidity of the atmosphere. The superficial portion of the SC in normal skin takes up (= sorb) water quickly due to its hygroscopicity, but it releases water rather slowly opposing a dehydration process (= desorption) due to its inherent water-holding capacity. In contrast, a dry, scaly skin surface can absorb only a small amount of water because of the deficient hygroscopicity of its SC and releases it quickly to the environment due to deficient water-binding capacity. Hence, we have devised a rapid functional assay of the SC, the in vivo water sorption–desorption skin test, which furnishes information about the hygroscopicity and waterholding capacity of the surface SC in a short time.32 The test procedure is simple, consisting of serial electromeasurements before and after artificial hydration of the skin surface. Because the whole procedure takes only 2 minutes, we can easily repeat the test several times in nearby areas to confirm reproducibility as well as to obtain mean values. The hygroscopic property of the superficial portion of the SC is evaluated by the increase in conductance to attain a maximal value immediately after blotting a water droplet placed at the site for 10 seconds. The ability of the superficial SC to retain the absorbed water, i.e., water-holding capacity, is measured from analysis of the subsequent desorption measurements of conductance at 30-second intervals for 2 minutes (Figure 38.3 and Figure 38.4). The desorption curve is approximated to an exponential curve, from which the desorption rate constant
334
Handbook of Non-Invasive Methods and the Skin, Second Edition
for water can be calculated.7 More accurately, the entire area beneath the desorption curve is measured. Computer analysis is applicable for these procedures.
38.10 ASSESSMENT OF THE EFFICACY OF SKIN MOISTURIZERS Immediately after application of moisturizing agents, the skin surface shows an increase in conductance value depending on the water content of the agents.33,34 This is followed by a rapid decrease due to evaporation of excess water from the skin surface. Thereafter, the conductance values are maintained at certain increased levels according to the efficacy of the agents for several hours if undisturbed. In contrast, no initial increase is observed after application of emollients such as petrolatum, which does not contain water. However, there is a gradual increase in conductance until reaching a plateau after 2 hours due to accumulation of water beneath it. To obtain reproducible results, we apply 20 μl of the agent in a 4 × 4 cm skin area. Repeated daily applications of moisturizers for several days induce an increase in conductance or capacitance that is demonstrable even several days after cessation of treatment, depending on the efficacy of the moisturizers.35,36 Moreover, such repeated applications of a moisturizer not only increase the hydration state of the skin surface, but also improve barrier function of the facial skin impaired by the dry and cold winter air.19,37
38.11 CORRELATION WITH OTHER METHODS The Corneometer is considered able to depict changes of hydration much deeper in the skin than the high-frequency method.15 Thus, we have compared measurements with these instruments on various sites of involved and uninvolved skin of patients with psoriasis. Conductance revealed a wider range of distribution on uninvolved skin, whereas capacitance tended to show a wider range of distribution than high-frequency conductance when measured on dry skin such as the involved psoriatic skin. Although there was a positive correlation between the obtained values (r = 0.60; p < 0.001), the X-Y intercept was clearly different from 0. As mentioned above, in a simulation model of in vivo SC, the high-frequency conductance showed a close correlation with the hydration state of the surface SC (r = 0.99) compared with capacitance (r = 0.79). Both devices were insensitive to changes of hydration taking place in deeper viable skin tissue, e.g., the accumulated tissue fluids in suction blisters. We have had a chance to compare the results of highfrequency conductometry with parameters of the skin surface obtained with other instrumental techniques.38 When the skin surface is observed under 200× magnification
with a special video camera, we can observe the presence of small fragments of detaching corneocytes even on the normal skin surface. The number of such desquamating corneocytes is higher on dry skin surface. Image analysis for quantification of these detaching clumps of corneocytes showed an inverse correlation between them and high-frequency conduction levels. Firmness and elasticity of the skin surface are clinically evaluated by simply touching it with fingers. They can be measured with a new type of tactile sensor designed to detect a shift in resonant frequency generated in the piexoelectric element.39 We have also found that the data obtained with this instrument showed a close correlation with the high-frequency conductance.
38.12 RECOMMENDATIONS As mentioned above, care should be taken to apply the probe properly, perpendicularly to the skin surface. It is also recommended to perform measurements at least three times in sites close together to get a mean value, rather than to record after only one, because even unnoticed poor contact between the probe and the skin surface causes a large decrease in obtained value. Furthermore, such small series of reading can be completed in a short time. The skin surface quickly absorbs water. Even invisible sweating or high relative humidity in the atmosphere can cause a great increase in conductance. Therefore, the skin of the forehead and palmoplantar surface is not suitable, because mental sweating occurs so easily on these sites, making it impossible to obtain reproducible results. A mid-portion of the flexor surfaces of the forearms is most suitable for comparative studies. A marked increase in conductance also occurs when humidity rises above 60%. Rather constant readings are obtained at a room temperature between 19 and 22˚C, but a steady increase in conductance is also observed above 22˚C.18 The ambient conditions with temperature around 20˚C and relative humidity between 40 and 50% are recommended for ordinary measurements. For the time course study ranging more than 1 day conducted on lesional skin or skin treated with topical agents, a climatecontrolled chamber with a room temperature around 20˚C and 40 or 50% relative humidity should be used. If such a chamber is unavailable, measurements should also be performed on uninvolved or nontreated control sites to observe daily fluctuations of conductance induced by environmental changes.
REFERENCES 1. Blank IH. Factors which influence the water content of the stratum corneum. J Invest Dermatol 18: 433, 1952.
Epidermal Hydration: Measurement of High-Frequency Electrical Conductance
2. Blank IH, Moleney J, Emslie A, Simon I, Apt C. The diffusion of water across the stratum corneum as a function of its water content. J Invest Dermatol 82: 188, 1984. 3. Warner RR, Myers MC, Taylor DA. Electron probe analysis of human skin: determination of the water concentration profile. J Invest Dermatol 90: 218, 1988. 4. Middleton JD. The mechanism of water binding in stratum corneum. Br J Dermatol 80: 437, 1968. 5. Elias PM. Lipids and the epidermal permeability barrier. Arch Dermatol Res 270: 95, 1981. 6. Imokawa G, Akasaki S, Hattori M, Yoshizuki N. Selective recovery of deranged water-holding properties by stratum corneum lipids. J Invest Dermatol 87: 785, 1986. 7. O’goshi K, Iguchi M, Tagami H. Functional analysis of the stratum corneum of scalp skin: studies in patients with alopecia areata and androgenetic alopecia. Arch Dermatol Res 292: 605, 2000. 8. Fluhr JW, Mao-Qiang M, Brown BE, Wertz PW, Crumrine D, Sundberg JP, Feingold KR, Elias PM. Glycerol regulates stratum corneum hydration in sebaceous gland deficient (asebia) mice. J Invest Dermatol 120: 728, 2003. 9. Tagami H, Yoshikuni K. Interrelationship between water barrier and reservoir functions of pathologic stratum corneum. Arch Dermatol 181: 642, 1985. 10. Tregear RT. The interpretation of skin impedance measurements. Nature 205: 600, 1965. 11. Clar EP, Her CP, Sturelle CG. Skin impedance and moisturization. J Cosmet Chem 26: 337, 1973. 12. Tagami H, Ohi M, Iwatsuki K, Kanamaru Y, Yamada M, Ichijo B. Evaluation of the skin surface hydration in vivo by electrical measurement. J Invest Dermatol 75: 500, 1980. 13. Leveque JL, de Rigal J. Impedance methods for studying skin moisturisation. J Soc Cosmet Chem 34: 419, 1983. 14. Masuda Y, Nishikawa M, Ichijo B. New methods of measuring capacitance and resistance of very high loss materials at high frequencies. IEEE Trans Instrum Meas IM-29: 28, 1980. 15. Hashimoto-Kumasaka K, Takahashi K, Tagami H. Electrical measurement of water content of the stratum corneum in vivo and in vitro under various conditions: comparison between skin surface hygrometer and Corneometer in evaluation of the skin surface hydration state. Acta Derm Venereol 73: 335, 1993 16. Obata M, Tagami H. Electrical determination of water content and concentration profile in a simulation model of in vivo stratum corneum. J Invest Dermatol 92: 854, 1989. 17. Sasai S, Zhen YX, Tagami H. High-frequency conductance measurement of the skin surface hydration state of dry skin using new probe studded with needle-form electrodes (MT-C). Skin Res Technol 2: 173, 1995. 18. Tagami H. Impedance measurement for evaluation of the hydration state of the skin surface. In Cutaneous Investigation in Health and Disease. Noninvasive Methods and Instrumentation, Leveque J-L, Ed. Marcel Dekker, New York, 1989, p. 79.
335
19. Kikuchi K, Kobayashi H, le Fur I, Tschachler E, Tagami H. The winter season affects more severely the facial skin than the forearm skin: comparative biophysical studies conducted in the same Japanese females in later summer and winter. Exog Dermatol 1: 32, 2002. 20. O’goshi K, Iguchi M, Tagami H. Functional analysis of the stratum corneum of scalp skin: studies in patients with alopecia areata and androgenetic alopecia. Arch Dermatol Res 292: 605, 2000. 21. Tagami H. Aging and the hydration state of the skin. In Cutaneous Aging, Kligman AM, Takase Y, Eds. University Tokyo Press, Tokyo, 1988, p. 99. 22. Saijo S, Tagami H. Dry skin of newborn infants: functional analysis of the stratum corneum. Pediatr Dermatol 8: 2, 1991. 23. Tagami H, Kikuchi K, O’goshi K, Kobayashi H. Electrical properties of new born skin. In Neonatal Skin. Structure and Function, 2nd ed, Hoath SB, Maibach HI, Eds. Marcel Dekker, New York, 2003, p. 179. 24. Hara M, Kikuchi K, Watanabe M, Dende H, Koyama J, Horii I, Tagami H. Senile xerosis: functional, morphological, and biochemical studies. J Geriatr Dermatol 1: 111, 1993. 25. Horii I, Nakayama Y, Obata M, Tagami H. Stratum corneum hydration and amino acid content in xerotic skin. Br J Dermatol 121: 587, 1989. 26. Kikuchi-Numagami K, Suetake T, Yanai M, Tahahashi M, Tanaka M, Tagami H. Functional and morphological studies of photodamaged skin on the hands of middleaged Japanese golfers. Eur J Dermatol 10: 277, 2000. 27. Kobayashi H, Tagami, H. Functional analysis of the stratum corneum of patients with atopic dermatitis: comparison with psoriasis vulgaris. Exog Dermatol 2: 33, 2003. 28. Urano-Suehisa S, Tagami H. Functional and morphological analysis of the horny layer of pityriasis alba. Acta Derm Venereol (Stockh) 65: 164, 1985. 29. Watanabe M, Tagami H, Horii I, Takahashi M, Kligman AM. Functional analyses of the superficial stratum corneum in atopic xerosis. Arch Dermatol 127: 1689, 1991. 30. Suetake T, Sasai S, Zhen Y-X, Ohi T, Tagami H. Functional analysis of the stratum corneum in scars. Sequential studies after injury and comparison among keloids, hypertrophic scars, and atrophic scars. Arch Dermatol 132: 1453, 1996. 31. Kunii T, Hirao T, Kikuchi K, Tagami H. Stratum corneum lipid profile and maturation pattern of corneocytes in the outermost layer of fresh scars: the presence of immature corneocytes plays a much more important role in the barrier dysfunction than do changes in intercellular lipids. Br J Dermatol 149: 749, 2003. 32. Tagami H, Kanamaru Y, Inoue K, Suehisa S, Inoue F, Iwatsuki K, Yoshikuni K, Yamada M. Water sorptiondesorption test of the skin in vivo for functional assessment of the stratum corneum. J Invest Dermatol 78: 77, 1982. 33. Blichmann CW, Serup J. Assessment of skin moisture. Acta Derm Venereol (Stockh) 68: 284, 1988.
336
Handbook of Non-Invasive Methods and the Skin, Second Edition
34. Serup J, Winther A, Blichmann CW. Effects of repeated application of a moisturizer. Acta Derm Venereol (Stockh) 69: 457, 1989. 35. Tagami H. Quantitative measurements of water concentration of the stratum corneum in vivo by high-frequency current. Acta Derm Venereol Suppl (Stockh) 185: 29, 1994. 36. Blichmann CW, Serup J, Winther A. Effects of single application of a moisturizer: evaporation of emulsion water, skin surface temperature, electrical conductance, electrical capacitance, and skin surface (emulsion) lipids. Acta Derm Venereol 69: 327, 1989. 37. Tabata N, O’Goshi K, Zhen YX, Kligman AM, Tagami H. Biophysical assessment of persistent effects of moisturizers after their daily applications: evaluation of corneotherapy. Dermatology 200: 308, 2000.
38. Tanita Y, Tagami H, Osanai O, Kawai M. Skin surface microscopy with a magnifying television unit (Direct Skin Analyzer®). Acta Derm (Kyoto) 84: 647, 1989. 39. Sasai S, Zhen X, Suetake T, Taninta Y, Omata S, Tagami H. Palpation of the skin with a robot finger: an attempt to measure skin stiffness with a probe loaded with a newly developed tactile vibration sensor and displacement sensor. Skin Res Technol 5: 237, 1999. 40. Kikuchi K, Tagami H. Comparison of the effects of daily applications between topical corticosteroid and tacrolimus ointments on normal skin: evaluation with noninvasive methods. Dermatology 205: 378, 2002.
of Epidermal 39 Measurement Capacitance André O. Barel and Peter Clarys Laboratory of General and Biological Chemistry, Faculty of Physical Education and Physiotherapy, Vrije Universiteit Brussel, Brussels, Belgium
CONTENTS 39.1 Introduction............................................................................................................................................................337 39.2 Object of This Study .............................................................................................................................................338 39.3 Measuring System and Description of the Corneometer......................................................................................338 39.3.1 Principle .....................................................................................................................................................338 39.3.2 Measuring Electrode..................................................................................................................................338 39.3.3 Description of the Instrument....................................................................................................................339 39.3.4 Measurements ............................................................................................................................................339 39.3.4.1 Accuracy and Sensitivity............................................................................................................339 39.3.4.2 Reproducibility ...........................................................................................................................339 39.3.4.3 Depth of the Detection of Hydration of the Skin .....................................................................340 39.3.4.4 Influence of External Environmental Factors ............................................................................340 39.3.4.5 Seasonal Influences on the Hydration .......................................................................................340 39.3.4.6 Anatomical Skin Areas for Testing............................................................................................340 39.3.4.7 Sex and Age ...............................................................................................................................341 39.3.4.8 Influence of Stripping.................................................................................................................341 39.4 Measuring System and Description of the MoistureMeter SC-2 .........................................................................341 39.5 Dermato-Cosmetic Applications............................................................................................................................341 39.5.1 Skin Diseases and Lesions ........................................................................................................................341 39.5.1.1 Dry Scaly Lesions ......................................................................................................................341 39.5.1.2 Irritation of the Skin...................................................................................................................341 39.5.2 Evaluation of the Efficiency of Cosmetic Moisturizing Products ............................................................342 39.5.2.1 Short-Term Effects .....................................................................................................................342 39.5.2.2 Long-Term Effects......................................................................................................................343 39.6 Conclusion .............................................................................................................................................................343 Acknowledgments ...........................................................................................................................................................343 References .......................................................................................................................................................................343
39.1 INTRODUCTION The presence of an adequate amount of water in the stratum corneum is important for the following properties of the skin: general appearance of a soft, smooth, well-moisturized skin in contrast to a rough and dry skin, of a flexible skin in contrast to a brittle and scaly skin, and of an intact barrier function allowing a slow release of transepidermal water loss (TEWL) in a dry environment. There is no universally accepted theory for explaining the situation of dry skin. Some authors consider that dry
skin is related to disorders of corneocyte adhesion and desquamation (rough and scaly surface), to modifications in the composition of certain epidermal lipids, or to disorders of the water-retaining properties of the horny layer (presence of natural moisturizing factors).1–6 There are very few data confirming that a situation of dry skin is linked solely to a diminution of the water content of the horny layer. However, the positive physiological effect of application of water and moisturizers to the skin surface to relieve the condition of dry skin has been repeatedly confirmed.7,8 As a consequence, given the fact that the 337
338
Handbook of Non-Invasive Methods and the Skin, Second Edition
presence of an adequate amount of water is an essential prerequisite for the maintenance of the normal structure and function of the stratum corneum, the in vivo determination of the degree of hydration of the horny layer is an important factor in the characterization of normal and pathological situations of this layer, of an actinic aged skin, of irritated skin conditions, and finally, in the assessment of the efficiency of various moisturizing topical products. It has been known for a long time that the electrical properties of the skin are related to the water content of the horny layer.9–13 Considering the simple electrical model of the skin as a resistor in parallel with a capacitor, these two components contribute to the total impedance or electrical opposition to an alternating current applied on the surface of the skin. By using an adequate geometry of the measuring electrodes, frequency of the current and design of the oscillating electronic circuit, the commercially available electrical instruments measure either (1) the conductance contribution (reciprocal of resistance) of the impedance, (2) the contribution of the capacitance impedance, or (3) the total impedance (combination of both parameters).13–18
39.2 OBJECT OF THIS STUDY Various experimental electrical instruments were developed in order to measure the water content of the horny layer (review articles have been published16,19). The measurement of the impedance of the skin has been studied extensively and is the simplest technique to assess the hydration state of the skin surface. A certain number of instruments are commercially available and mostly used in dermato-cosmetic research. There are three instruments based on the conductance/impedance method: the Skicon (versions 200 and 200-EX, ISBS Co., Hamamatsu, Japan20), the Dermalab (Cortex Technology, Hadsund, Denmark21), and the multifrequency IMP Spectrometer (SciBase AB, Huddinge, Sweden22,23). The Nova Dermal Phase Meter Nova DPM 9003 (Nova Technology Corporation, Portsmouth, NH24) operates on impedance-based capacitance technology. Finally, two instruments based on the measurement of the capacitance: the Corneometers CM 820 and 825 from Courage-Khazaka (Köln, Germany)25 and the MoistureMeter from Delphin Technologies Ltd. (Kuopio, Finland) are commercially available.26 This chapter is mainly concerned with the description of the measurements of the epidermal capacitance using the capacitance method (Corneometer). Since the development of the MoistureMeter is very recent, a limited amount of scientific information is available.26 Different aspects of the use of the capacitance apparatus, such as accuracy, reproducibility, range of measurements, and influence of external environmental factors on
the measurements, will be described. It will be shown that the capacitance apparatus is a reliable and very easy to use instrument that has been widely used for hydration measurements in healthy and pathological skin conditions and for quantitative assessment of the efficiency of various moisturizing products. As has been described earlier,27 reliable and reproducible hydration measurements are obtained only if the experiments with the capacitance method are carried out under well-controlled standardized experimental conditions. A few comparative studies about the measuring capabilities of the conductance method and the capacitance method have been published.28–30
39.3 MEASURING SYSTEM AND DESCRIPTION OF THE CORNEOMETER 39.3.1 PRINCIPLE The total impedance of the skin Z, when the skin is submitted to an applied alternating current of frequency F, depends on the contribution of the resistance R and the capacitance C according to the following relation:13,28,31 Z = (R2 + 1/2πFC2)1/2 In agreement with Mosely et al.,11 by using an adequate design of the measuring electrodes and the oscillating electronic circuit, the apparatus measures primarily the capacitance contribution of the skin in contact with the measuring electrodes.28,31 In vitro measurements on cellulose filters showed clearly a correlation between the capacitance values and the dielectric constant of the liquid saturating the filter paper.29,31
39.3.2 MEASURING ELECTRODE The measuring probe consists of an interdigital grid of gold-covered electrodes. The active part of the electrode covers a surface of 7 × 7 mm (Figure 39.1). The electrodes are 50 μm wide with an interdigital spacing of 75 μm. The active part of the electrode is covered by a low dielectric vitrified material of 20 μm thickness. The total probe (surface area, 0.95 cm2) is applied with a constant pressure with a spring system. The pressure of the CM 820 and CM 825 analog detection system is 1.6 N/m2, whereas the new digital system works at a lower pressure (less than 1 N/m2). Since there is no direct galvanic contact between the gold-plated electrodes and the skin surface, no galvanic current occurs through the skin. Only an electric field of variable frequency is established in the upper parts of the skin (horny layer and epidermis).
Measurement of Epidermal Capacitance
Lateral view of the electrode
Front view of the electrode
Skin surface
FIGURE 39.1 Schematic representation of the measuring probe of the capacitance apparatus (Corneometer CM 820 PC). Front and lateral views of the interdigital electrodes. The electrodes are covered with a vitrified material and are consequently not in direct contact with the skin surface.
39.3.3 DESCRIPTION
OF THE INSTRUMENT
The form and depth of the electric field present in the skin depends on the geometry of the electrodes and the dielectric material covering the electrodes (constant capacitance, Co) and of the capacitance of the biomaterial in contact with the electrode (variable capacitance, C). As a consequence, the whole system (electrode, horny layer, and upper parts of the epidermis) works as a variable capacitor. The total capacitance is only influenced by changes in the dielectric constant of the biomaterial in contact with the electrodes. A dry horny layer is a dielectricum medium. When the horny layer is hydrated, a significant change in the dielectric properties of this medium is observed. As a consequence, the total capacity of this system is changed in function of the degree of hydration of the skin (mostly in the horny layer). A resonating system in the instruments measures the shift in frequency of the oscillating system, which results from the changes in the total capacity of the skin surface. In the case of the old version of the Corneometer, the CM 820, the frequency shifts from 40 KHz for a hydrated medium to 75 KHz for a dry medium. In the case of the new versions of the Corneometer (CM 825, with the analog and digital probe), the frequency shifts from 0.95 MHz for a hydrated medium to 1.15 MHz for a dry medium. The capacity of the skin surface is automatically measured after 1.0 to 1.5 s of application of the probe on the skin. The variable total capacitance of the skin surface is converted in arbitrary units (a.u.) of skin hydration. The theoretical range of the apparatus varies from 0 to 120 a.u. The new apparatus (CM 825) is connected to a central multiprobe unit MPA 5, which operates on Windows software. Hydration results are displayed either as 10 independent consecutive values (mean value + SD) or kinetic curves in function of time (at 3 and 7 s and 7 min).
39.3.4 MEASUREMENTS 39.3.4.1 Accuracy and Sensitivity The instrument indicates changes in capacitance of the skin that are expressed in arbitrary hydration units (a.u.)
339
varying from 0 to 120; arbitrary units are not directly related to real electrical units correlated in a simple way to the water content of the horny layer. Calibration of the instrument is carried out in the factory using three standard materials with known capacitance. Direct calibration of the capacitance instrument with various sophisticated in vitro model materials of known thickness and water content (simulating the real epidermis) is very complex and has not yet been published. However, in vitro calibration of the instrument with simple model systems such as cellulose filters of known thickness (from 15 μm to 1 mm) impregnated with aqueous solutions and with solutions of known dielectric constant is possible.25,29,31 With a cellulose filter disc impregnated with a 0.15 M NaCl aqueous solution one obtains a maximal hydration value of ±120 a.u. When covering the surface of the saturated filter with a layer of a low dielectric plastic film of 15 μm thickness one measures a hydration value of about 30 a.u. The capacitance method shows in vivo a broad range of sensitivity from 20 to 110 a.u. Unfortunately, when comparing the data of the old version (CM 820) and the new version (CM 825), systematically significant higher values (±60%) are observed with the older version.16 In a multicenter study the analog and digital versions of the Corneometer CM 825 were compared and a very high correlation (r = 0.92) was noticed. The analog and digital probes give identical hydration values.30 Comparison of the hydration measurements obtained by the capacitance method with data obtained by the impedance method (Skicon-100), which can be directly calibrated with a reference of known conductance, showed a very high degree of correlation (r = 0.94) between both instruments over a broad range of hydration values.28 Practically, it is considered that for the cornemeter CM 825 values of hydration below 40 a.u. correspond to very dry skin; between 40 and 55, dry skin; and higher than 55, normal skin.30 In agreement with many authors, the capacitance method is very sensitive for measurements at low hydration and less sensitive in the range of very high hydration values (above 110 a.u.).26,28 39.3.4.2 Reproducibility The reproducibility of the capacitance method was tested on different skin sites (forearm, forehead, crow’s-feet, abdomen, thigh, hand, leg, sole, etc.) either on the same individual or on a large group of individuals of the same age group. In agreement with other studies,16,27 the reproducibility of the hydration measurements on the same individual varies from 4 to 5% for most skin sites under a broad range of skin conditions (very dry to moist skin). When measurements of reproducibility were carried out on a larger heterogeneous group of individuals of both
Handbook of Non-Invasive Methods and the Skin, Second Edition
sexes in the same age group, a coefficient of variability of 10% was observed for most of the skin sites. In accordance with other studies,27 the results of the capacitance measurements are very reproducible. 39.3.4.3 Depth of the Detection of Hydration of the Skin In order to investigate the depth of the detection of the capacitance method, experiments were performed in vitro on cellulose impregnated with an aqueous solution covered with increasing sheets of plastic foil of 15 μm thickness each and in vivo on the forearm skin through increasing sheets of 25 μm thickness each.29,31 About 75% of the reduction of the initial values for the CM 820 and CM 825 was obtained with only one layer (15 μm). In vivo measurements showed a sharp decrease in hydration a.u. after successive layers of tape were placed on the skin. Starting from a typical value of 90 a.u. of capacitance at the skin surface, at a depth of 125 μm (five layers of tape) the capacitance response of the instrument is lowered to 10 a.u.31 These results and others29 indicate that the capacitance method detects variations in the dielectric constant of the skin at a depth corresponding to the superficial part of the epidermis, with a main contribution due to the horny layer. 39.3.4.4 Influence of External Environmental Factors The effect of external factors such as temperature and relative humidity of the ambient air on the hydration values of the horny layer has been described before.9,13 Figure 39.2 shows the results of the influence of external humidity on the hydration of the forearm skin. The hydration of the skin (volar part of the forearm) was examined under constant temperature (20 ± 2˚C) in function of an increase of relative humidity (from 37 to 87%). A linear relation (r = 0.98) was observed between the hydration, as measured by the capacitance method and the external relative humidity. This result is more or less in agreement with previous skin conductance measurements.8,9,11,13,27 As a consequence, hydration measurements with the capacitance method must be performed in an experimental room where the relative humidity is kept more or less constant (50 ± 5% RH). When hydration measurements are carried out in variable conditions of relative humidity, the obtained data must be corrected with a correction factor taken from the slope of Figure 39.2. Similarly, a steady increase of capacitance was observed by Rogiers et al.27 in function of temperature above 22˚C, which corresponds to a higher hydration of the horny layer and the invisible beginning of sweating. The temperature of the experimental room must be kept constant and preferably below 22˚C (ideally 20˚C).
90 r = 0.98
Hydration (a.u.)
340
80
70
60 30
40
50
60
70
80
90
Relative humidity (%)
FIGURE 39.2 Effect of external relative humidity (%) on the capacitance hydration values of the forearm skin. Temperature = 22 ± 2˚C (n = 15; age group, 18 to 30 years).
39.3.4.5 Seasonal Influences on the Hydration As a result of the higher external temperature and higher values of relative humidity occurring during summer, much higher hydration values are measured during this season on all skin sites.2,7,13 In the winter, due to very low values of relative humidity, typical symptoms of very dry skin appear on all the exposed areas. With older individuals, capacitance hydration values as low as 30 to 40 are measured on the lower legs (winter xerosis). Due to the interference of external high values of relative humidity and possibility of small amounts of sweating on the skin surface observed in summer, it is very difficult to carry out reproducible and reliable hydration experiments at this time of year. As a consequence, quantitative efficiency experiments with moisturizing products are never carried out in the summer in our laboratory. In order to minimize the interferences of all these external factors, it is important that all the volunteers rest a minimum of 30 min in the experimental room prior to the experimental procedure starting. 39.3.4.6 Anatomical Skin Areas for Testing The choice of the skin area selected for hydration measurements is important, since there are large variations in the hydration of the horny layer in function of body region.9,11–27 High hydration values are obtained at the forehead and the palm of the hand; lower values are observed at the abdomen, thigh, and lower leg. The hydration status of symmetrical sites of the body is generally identical, a situation that allows contralateral left–right comparative studies. In agreement with other researchers,27 significant differences in hydration were observed when comparing skin sites located respectively at the distal part and proximal part on the volar side of the forearm.
Measurement of Epidermal Capacitance
It is important to point out that comparative contralateral hydration measurements must be performed exactly on the same identical anatomical skin sites. 39.3.4.7 Sex and Age In agreement with other studies carried out with the conductance method3,13 and the capacitance method,27 both sexes when compared within the same age group show identical hydration values at all anatomical skin areas studied. Different authors have shown that starting from adults (18 to 25 years) there is a slow but steady decrease in hydration of the stratum corneum at all skin sites in function of age.3 One observes in elderly volunteers of both sexes a typical situation of xerosis on the lower legs (arbitrary units of skin capacitance around 30 to 40). 39.3.4.8 Influence of Stripping Removal of the horny layer by stripping with an adhesive tape was examined by the skin capacitance method. The hydration values of the skin located at the forearm are progressively increased as deeper layers of the horny layer are serially removed by successive tape strippings.13,14 Finally, after 20 to 25 strippings, a high value of 110 to 120 a.u. is obtained, which corresponds to the fully hydrated horny layer located near the viable part of the epidermis.
39.4 MEASURING SYSTEM AND DESCRIPTION OF THE MOISTUREMETER SC-2 As already mentionned in this chapter, the development of the MoistureMeter is very recent; only one source of scientific information, published in 2004, is available.26 The MoistureMeter is a novel capacitive device operating at 1.25 MHz with two concentric electrodes. The instrument shows arbitrary capacitance units (normalized hydration values) that correspond to units of a dielectric constant divided by the thickness of the horny layer. The system works with a variable contact pressure going from 0.69 to 2.06 N. The device is factory calibrated using a 100 pF capacitor. In vitro calibration is possible using a cellulose filter impregnated with a saline solution covered with a 10-μm polyethylene foil giving a measured value of 21 normalized hydration units. An in vivo comparison between the MoistureMeter and the old version of the Corneometer, CM 820, was carried out, indicating a significant good correlation (r = 0.75) between both instruments. According to this study, the relative range of the MoistureMeter is greater than that of the Corneometer (from 18 to 72 units and from 67 to 104 units, respectively), suggesting that this instrument is a more sensitive
341
device. A comparative study between the MoistureMeter and the new digital version of the Corneometer CM 825 should be carried out in the future.
39.5 DERMATO-COSMETIC APPLICATIONS Besides the fundamental interest of studying in vivo the hydration properties of the superficial layers of the skin surface located at various anatomical sites in healthy and diseased skin situations, the determination of skin hydration is also important for assessing the therapeutic efficacy of topical skin products in various skin lesions and for the objective evaluation of the efficiency of various moisturizing preparations in cosmetics.
39.5.1 SKIN DISEASES
AND
LESIONS
39.5.1.1 Dry Scaly Lesions Hydration measurements carried out on the skin with dry scaly lesions (psoriasis, eczematous dermatitis) always reveal lower values of skin hydration (conductance measurements).11,13 Similar results were obtained by Werner1 and Berardesca et al.6 using the capacitance method on atopic dermatitis and psoriatic skin situations. 39.5.1.2 Irritation of the Skin When the skin is exposed to various chemical irritants such as surfactants, one observes a complex phenomenon of skin irritation. In addition to typical irritation symptoms (swelling and redness), the barrier function of the stratum corneum is partially destroyed and the water content of this layer is significantly lower.5 An objective assessment of the irritant character of some household detergent solution can be carried out by following TEWL, skin color, and the hydration of the horny layer of the forearm and the dorsal part of the hands in function of consecutive exposures to these products in the hand/forearm immersion test.32 A significant decrease in hydration of the horny layer (forearm and hand) was observed after two and four consecutive exposures during 30 minutes at 40˚C to two different household cleaning products (Figure 39.3). When hydration measurements are carried out respectively before and in function of time after two consecutive exposures to the same cleaning products, the hydration of the skin returns progressively to normal values, depending of the mildness of the surfactant (Figure 39.4). When capacitance measurements are carried out on the skin in the immersion test under standardized experimental conditions, it is possible to discriminate between a mild surfactant (product 1) and a more irritant detergent (product 2) (see Figure 39.3 and Figure 39.4 for comparison).32,33
342
Handbook of Non-Invasive Methods and the Skin, Second Edition
100
1-hand 1-arm 2-hand 2-arm
50
1
2
3
4
5
Number of exposures
FIGURE 39.3 Effect of exposure of the skin (dorsal part of the hands and volar part of the forearm) to detergents (30 min, each exposure at 40˚C). Hydration measurements (arbitrary capacitance units) were performed initially and 24 hours after the second and fourth exposure to detergent (n = 12; age group, 18 to 25 years). Solution 1 = 1% solution of a mild dishwashing liquid containing alkylethoxy sulfate and amphoteric surfactants (open symbols); solution 2 = 1% solution of a more irritant dishwashing liquid containing linear alkylbenzene sulfonate and sodium lauryl sulfate surfactants (filled symbols).
Hydration (a.u.)
60
50
1-hand 1-arm 2-hand 2-arm
40
30 0
10
20
90 Blanco O/W 80
70 −30
40 0
Hydration (a.u.)
Hydration (a.u.)
60
30
Time after exposure (hours)
FIGURE 39.4 Recovery effect of the skin (dorsal part of the hands and volar part of the forearm) after two consecutive exposures to detergents (30 min, each exposure at 40˚C). Hydration measurements (arbitrary capacitance units) were performed initially and in function of time after two exposures. Same symbols as in Figure 39.3.
39.5.2 EVALUATION OF THE EFFICIENCY OF COSMETIC MOISTURIZING PRODUCTS Many review articles concerning hydration studies, measurements, claim support of moisturizers, and possible pitfalls have been published.16,17,19,34 39.5.2.1 Short-Term Effects The short-term efficiency of moisturizing products can be readily analyzed from the increase of hydration after application of the product on the skin. Most experiments are carried out either on the volar side of the forearm or on the frontal side of the lower legs. Both anatomical sites
0
30
60 90 120 150 180 Time (min.)
FIGURE 39.5 Changes in hydration (arbitrary capacitance units) of the forearm skin after a single application of an O/W emulsion containing 2% urea as humectant. Treated area (O/W, filled symbols) in comparison with nontreated area (blanco, open symbols) (n = 15; age group, 18 to 30 years).
allow contralateral at-random comparison (untreated control skin areas vs. treated skin areas). The various hydrating products (oil-in-water (O/W) and water-in-oil (W/O) emulsions, hydro- and lipogels, occlusive creams, etc.) are gently applied by rubbing at a concentration of 1 to 2 mg product per cm2 skin area on the test area (4 × 4 cm). Recordings of the hydration of the control and the test area are taken during a certain time (10 to 30 minutes) before application of the product on the test area in order to ascertain the hydration state of the skin (baseline value). Following application, recordings are performed on the treated and untreated skin areas every 10 to 15 min during a period lasting from 60 to 180 min depending on the efficacy of the hydration product. Figure 39.5 shows the typical short-term effect of an O/W emulsion containing 2% urea as moisturizing factor on the hydration of the forearm skin. As has been previously described,13,27–29 immediately after application of the O/W emulsion the hydration (capacitance arbitrary units) shows a significant increase that corresponds to the application of water present in the O/W emulsion on the skin surface. As a consequence of the evaporation of the excess of water present on the skin surface, a decrease in hydration of the horny layer is observed, and after a certain time, a constant increase in hydration is observed. The level and duration of the increased hydration of the horny layer are a measure of the efficacy of the moisturizing preparation. Figure 39.6 shows the effect of a single application of a water-in-oil emulsion containing 4% urea as humectant on the hydration of the forearm skin. One observes again a significant increase in hydration of the skin immediately after application of the product, followed by a decrease. A significant increase in hydration (plateau value) is maintained during a long time, reflecting the temporary occlusion effect on the skin surface of the W/O emulsion.
Measurement of Epidermal Capacitance
343
widely used in dermatological, pharmacological, and skin care research.
100 Hydration (a.u.)
95 90 Blanco W/O
85 80 75 70 −30
0
30
60 90 120 150 180 Time (min.)
FIGURE 39.6 Changes in hydration (arbitrary capacitance units) of the forearm skin after a single application of a W/O emulsion containing 4% urea as humectant. Treated area (W/O, filled symbols) in comparison with nontreated area (blanco, open symbols) (n = 15; age group, 18 to 30 years).
39.5.2.2 Long-Term Effects Long-term hydration studies can also be quantitatively assessed by skin capacitance measurements. In a longterm study the moisturizing products are generally tested in a group (n = 12 to 30) of middle-age to older women who present symptoms of dry to very dry skin (skin capacitance values of around 50 to 60 on the forearms and lower legs). Testing on volunteers can be carried out in the laboratory or in an at-home study.7 The skin areas (volar part of the forearms or frontal part of the lower legs) are twice or three times daily treated with topical moisturizing products for 2 to 3 weeks. Contralateral skin sites are either untreated or treated with a placebo. Skin capacitance measurements are carried out before and succesively after 1, 2, and 3 weeks’ treatment with the products. Good correlations were obtained in our laboratory between clinical visual dryness scores, participant perception of skin dryness, and skin capacitance values during the progress of alleviating the dryness of the skin.7,16,17,19
39.6 CONCLUSION The capacitance apparatus is a simple-to-use, convenient, low-cost instrument that measures in vivo noninvasively the hydration of the skin at the superficial layers of the epidermis. The capacitance method, when used under well-controlled standard experimental conditions, is accurate and reproducible enough for skin hydration measurements in normal and pathological skin situations. Since the capacitance hydration values show a linear response in function of external relative humidity, the capacitance data can be corrected in function of the external relative humidity. The apparatus is suitable for quantitative evaluation of shortterm and long-term hydration effects on the skin after treatment with various topical dermato-cosmetic moisturizing preparations. As a consequence, this instrument is
ACKNOWLEDGMENTS The authors thank Mr. W. Courage and Mr. G. Khazaka for their valuable technical support in this study. The authors also acknowledge the support of Dr. B. Gabard (Basel, Switzerland) and thank Prof. Alain Barel (VUB, Brussels, Belgium) for valuable theoretical advice in this work.
REFERENCES 1. Werner, Y., The water content of the stratum corneum in patients with atopic dermatitis. Measurement with the Corneometer CM 420, Acta Derm. Venereol. (Stockh.), 66, 281, 1986. 2. Lévêque, J.L., Grove, G., de Rigal, J., Corcuff, P., Kligman, A.M., and Saint Leger, D., Biophysical characterization of dry facial skin, J. Soc. Cosmet. Chem., 82, 171, 1987. 3. Lévêque, J.L., Méthodes expérimentales d’étude du vieillissement cutané chez l’homme, In Vivo Acta Dermatol. Venereol., 144, 1279, 1987. 4. Saint Léger, D., François, A.M., Lévêque, J.L., Stoudemayer, T.J., Grove, G.L., and Kligman, A.M., Age-associated changes in stratum corneum lipids and their relation to dryness, Dermatologica, 177, 159, 1988. 5. Imokawa, G., Akasaki, S., Minematsu, Y., and Kawai, M., Importance of intercellular lipids in water-retention properties of the stratum corneum: induction and recovery study of surfactant dry skin, Arch. Dermatol. Res., 281, 45, 1989. 6. Berardesca, E., Fidelli, D., Borroni, G., Rabbrosi, G., and Maibach, H., In vivo hydration and water retention capacity of stratum corneum in clinically uninvolved skin in atopic and psoriatic patients, Acta Derm. Venereol. (Stockh.), 70, 400, 1990. 7. Prall, J.K., Theiler, R.F., Bowser, P.A., and Walsh, M., The effectiveness of cosmetic products in alleviating a range of skin dryness conditions as determined by clinical and instrumental techniques, Int. J. Cosmet. Sci., 8, 159, 1986. 8. Batt, M.D., Davis, W.B., Fairhurst, E., Gerraid, W.A., and Ridgde, B.D., Changes in the physical properties of the stratum corneum following treatment with glycerol, J. Soc. Cosmet. Chem., 39, 367, 1988. 9. Clar, E.J., Her, C.P., and Sturell, C.G., Skin impedance and moisturization, J. Soc. Cosmet. Chem., 26, 337, 1975. 10. Lévêque, J.L. and de Rigal, J., Impedance methods for studying skin moisturization, J. Soc. Cosmet. Chem., 34, 419, 1983. 11. Mosely, H., English, J.S., Coghill, G.M., and Mackie, R.M., Assessment and use of a new skin hygrometer, Bioeng. Skin, 1, 177, 1985.
344
Handbook of Non-Invasive Methods and the Skin, Second Edition
12. Blichman, C.W. and Serup, J., Assessment of skin moisture. Measurement of electrical conductance, capacitance and transepidermal water loss, Acta Derm. Venereol. (Stockh.), 68, 284, 1988. 13. Tagami, H., Impedance measurements for evaluation of the hydration state of the skin surface, in Cutaneous Investigation in Health and Disease, Noninvasive Methods and Instrumentation, Lévêque, J.L., Ed., Marcel Dekker, New York, 1989, chap. 5. 14. Barel, A.O., Clarys, P., Wessels, B., and de Romsée, A., Non-invasive electrical measurement for evaluating the water content of the horny layer: comparison between the capacitance and the conductance measurements, in Prediction of Percutaneous Penetration: Methods, Measurements, Modelling, Scott, R.C., Guy, R.H., Hadgraft, J., and Boddé, H.E., Eds., IBC Technical Services Ltd., London, 1991, p. 238. 15. Barel, A.O. and Clarys, P., In vitro calibration of the capacitance method (Corneometer CM 825) and conductance method (Skicon-200) for the evaluation of the hydration state of the skin, Skin Res. Technol., 3, 107, 1997. 16. Wilhelm, K.P., Possible pitfalls in hydration measurements, in Skin Bioengineering: Techniques and Applications in Dermatology and Cosmetology, Elsner, P., Barel, A.O., Berardesca, E., Gabard, B., and Serup, J., Eds., Karger, Basel, 1998, p. 223. 17. Barel, A.O., Clarys, P., and Gabard, B., In vivo evaluation of the hydration state of the skin: measurements and methods for claim support, in Cosmetics: Controlled Efficacy Studies and Regulations, Elsner, P., Merk, H.F., and Maibach, H.I., Eds., Springer, Berlin, 1999, p. 57. 18. Bernengo, J.C. and de Rigal, J., Techniques physiques de mesure de l’hydratation du Stratum Corneum, in Physiologie de la peau et explorations fonctionelles cutanée, Agache, P., Ed., Editions Médicales Internationales, Cachau, France, 2000, p. 117. 19. Barel, A.O., Product testing: moisturizers, in Bioengineering of the Skin: Skin Biomechanics, Elsner, P., Berardesca, E., Wilhelm, K.P., and Maibach, H.I., Eds., CRC Press, Boca Raton, 2002, p. 241. 20. Skicon, ISBS Co., Hamamatsu, Japan, Technical information of the version 200 and 200-EX, 2004. 21. Dermalab, Cortex Technology, Hadsund, Denmark, Technical information of the Dermalab, 2004. 22. Nicander, I., Nyrèn, M., Emtestan, L., and Ollman, S., Baseline electrical impedance measurements at various skin sites, Skin Res. Technol., 3, 252, 1997. 23. Multifrequency IMP Spectrometer, SciBase AB, Huddinge, Sweden, 2004. 24. The Nova Dermal Phase Meter Nova DPM 9003, Nova Technology Corporation, Portsmouth, NH. Technical information of the Nova DPM 9003, 2004.
25. Khazaka, G. and Courage, W., Courage-Khazaka Electronic, Köln, Germany, Technical information concerning the analog and digital measuring probes of the Corneometer CM 825, 2004. 26. Alanen, E., Nuutinen, J., Nicklén, K., Lahtinen, T., and Mönkkönen, J., Measurement of hydration in the stratum corneum with the MoistMeter and comparison with the Corneometer, Skin Res. Technol., 10, 32, 2004. 27. Rogiers, V., Derde, M.P., Verleye, G., and Roseeuw, D., Standardized conditions needed for skin surface hydration measurements, Cosmet. Toiletries, 105, 73, 1990. 28. Clarys, P., Barel, A.O., and Gabard, B., Non-invasive electrical measurements for the evaluation of the hydration state of the skin: comparison between three conventional instruments: the Corneometer, the Skicon and the Nova DPM, Skin Res. Technol., 5, 14, 1999. 29. Fluhr, J.W., Gloor, M., Lazzerini, S.L., Kleesz, P., Grieshaber, R., and Berardesca, E., Comparative study of five instruments measuring stratum corneum hydration (Corneometer CM 820 and CM 825, Skicon-200, Nova DPM 9003 and Dermalab). Part I. In vitro. Part II. In vivo, Skin Res. Technol., 5, 161, 171, 1999. 30. Heinrich, U., Koop, U., Leneveu-Duchemin, M.C., Osterrieder, K., Bielfeldt, S., Charkaut, C., Degwert, J., Häntschel, D., Jaspers, S., Nissen, H.P., Rohr, M., Schneider, G., and Tronnier, H., Members of the DGK Task Force “Skin Hydration,” multi-center comparison of skin hydration in terms of physical, physiological and product dependent parameters by the capacitive method (Corneometer CM 825), J. Cosmet. Sci., 25, 31, 2003. 31. Barel, A.O. and Clarys, P., In vitro calibration of the capacitance method (Corneometer CM 825) and conductance method (Skicon-200) for the evaluation of the hydration state of the skin, Skin Res. Technol., 3, 107, 1997. 32. Barel, A.O., Clarys, P., Wessels, B., and van de Straat, R., Quantitative Biophysical Measurements of the Mildness Properties of Cleaning and Detergent Products in the Hand Immersion Test, paper presented at the International Symposium on Irritant Contact Dermatits, Groningen, The Netherlands, October 3–5, 1991. 33. Clarys, P., van de Straat, R., Boon, A., and Barel, A.O., The Use of the Hand/Forearm Immersion Test for Evaluating Skin Irritation by Various Detergent Solutions, paper presented at the First Congress of the European Society of Contact Dermatitis, Brussels, October 8–10, 1992. 34. Treffel, P. and Gabard, B., Skin hydration, in DermatoPharmacology of Topical Preparations, Gabard, B., Elsner, P., Surber, C., and Treffel, P., Eds., Springer, Berlin, 2000, p. 317.
as a Noninvasive 40 Bioimpedance Method for Measuring Changes in Skin I. Nicander Department of Dermatology, Huddinge University Hospital, Huddinge, Sweden
P. Åberg and S. Ollmar Karolinska Institutet, Medical Engineering, Novum Research Park, Huddinge, Sweden
CONTENTS 40.1 Introduction............................................................................................................................................................345 40.2 What Causes Changes in Bioimpedance?.............................................................................................................346 40.3 Bioimpedance and Hydration ................................................................................................................................346 40.4 Skin Barrier and Skin Diseases.............................................................................................................................347 40.5 Impedance Imaging ...............................................................................................................................................348 References .......................................................................................................................................................................349
40.1 INTRODUCTION In order to measure electrical bioimpedance (electrical impedance in a biomedical application), electric energy (electric current or electromagnetic wave) has to be passed through the tissue under test. Alternating current (AC) is always used, since direct current (DC) would imply mass transportation; i.e., charged entities would continuously travel to opposite poles depending on the sign of the charge, thereby altering the composition of the sample. At DC only one parameter, the resistance (or its inverse, the conductance), would be measured; i.e., only one number is needed. At AC it takes two parameters to fully describe the impedance, because there are both a size relationship and a time lag between the voltage and current. These relations are frequency dependent. Thus, for AC two numbers are required: the magnitude and the phase (lag) or, mathematically equivalent, the real part and the imaginary part of the impedance at each frequency. Sometimes the inverse of impedance is used, called admittance. At low or medium–high frequencies (up to several megahertz), impedance (or admittance) parameters are typically used, while at very high frequencies, where the conduction mechanism is better viewed as electromagnetic wave propagation, often dielectric parameters are used. For a
comprehensive overview, refer to the textbook by Grimnes and Martinsen.1 In living systems, a number of frequency-dependent conduction mechanisms are at play. The major mechanisms were sorted out in 1957 by Schwan,2 who named them the α-, β-, and γ-dispersions. In relatively simple and homogeneous tissues, such as blood, these dispersions are reasonably separated along the frequency axes, and in a limited frequency range (encompassing only one relatively pure dispersion), impedance properties of many tissues can be fitted to the so-called Cole equation, which yields a part of a semicircle if displayed as a Nyquist plot (imaginary part vs. real part of the impedance). For skin, however, with its multilayer structure, the dispersions are partly overlapping, and there are also subdivisions that make adequate modeling increasingly complicated. In this context it is pertinent to emphasize that the stratum corneum, which is the dominating factor of impedance of intact skin, does not obey the Cole equation, as pointed out by Martinsen et al.,3,4 and is demonstrated in Figure 40.1. Careless use of the Cole model will filter away useful information by trying to squeeze the biomedical reality into too narrow a space. Electrodes or some sort of antenna system have to be employed to contact the tissue under test. It is a well345
346
Handbook of Non-Invasive Methods and the Skin, Second Edition
80 tape strips
20 tape strips 45 40
Imaginary part (kOhm)
200 180 160 140 120 100 80 60 40 20
Imaginary part (kOhm)
Imaginary part (kOhm)
0 tape strips
35 30 25 20 15 10 5
50
100
150
200
Real part (kOhm)
5 10 15 20 25 30 35 40 45 Real part (kOhm)
4.5 4 3.5 3 2.5 2 1.5 1 0.5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 Real part (kOhm)
FIGURE 40.1 Nyquist plots (imaginary part vs. real part) of bioimpedance of skin of a healthy volunteer in the frequency range 1.6 kHz to 1 MHz. The stratum corneum (SC) was incrementally removed using tape stripping. It is seen that both the real part and imaginary part, and thereby the magnitude of bioimpedance, go down with the number of strips, and that the graphs gradually take on the shape of a section of a Cole semicircle typical of living tissue, while the otherwise dominating non-Cole behavior of the SC disappears. At 80 strips, sections of two different dispersions are discernible. The skin was inundated with saline solution before each measurement, so that the graphs really show the effect of stripping and not differences in hydration between the outermost dry layers and the inner naturally wet layers of the SC.
known fact that when measuring, e.g., ECG or EEG, the skin–electrode interface and the skin itself constitute a major obstacle to the desired signals, which is handled by preparation of the skin with abrasion or somewhat aggressive electrolytic contact gel. While aiming at extraction of information about the skin using bioimpedance, such enforced alterations cannot be allowed. One might think that contact artifacts could be eliminated using so-called four-pole measurement, where current injection electrodes are separated from voltage detection electrodes, but this is typically not practical due to the dimensions of the small area of a skin site under investigation, the thickness of the skin layers, and a desire not to include tissue below the skin in the measured volume. Therefore, two-pole measurement (with variations) is normally used. Thus, properties of electrode material, size, and shape have to be carefully considered when interpreting the resulting data and when comparing results from different measurement systems, which may be straightforward for in vitro measurement, but it is far from straightforward in noninvasive in vivo applications.
40.2 WHAT CAUSES CHANGES IN BIOIMPEDANCE? All charge carriers, mainly ions, will contribute to the conduction of electricity through the tissue according to their concentration and mobility. This is straightforward in aqueous solutions at low frequencies, but any structural barriers between compartments of electrolytes will introduce polarization phenomena, and at higher frequencies polar entities will interact with the field. Thus, changes in concentrations of electrolytes, as well as changes in the mobility of electric field interacting components, will
modulate the baseline impedance. Structural changes, shape, size, and packing density on the cellular level, or in pore or duct size, as well as the integrity of cell membranes or other barriers, will also influence the bioimpedance.
40.3 BIOIMPEDANCE AND HYDRATION Several instruments employ a special case of impedance, the conductivity or the dielectric capacity, to estimate skin hydration. No doubt, the hydration of the stratum corneum is a dominating factor for its conductive properties, as demonstrated already in 1945 by Rosendal,5 and expanded upon by Yamamoto and Yamamoto in 1976.6 Fluhr et al.7,8 and Fischer et al.9 have compared a variety of devices intended to estimate skin hydration. Intuitively, the meaning of skin hydration might seem obvious, but it is actually ill-defined. Is it free water, more or less bound water, in only the upper layers of the stratum corneum, or in the whole stratum corneum (including the functional skin barrier), will living epidermis or deeper skin layers influence the measurement, etc.? Since each of the devices uses a different frequency (in the β-dispersion range) and has a different electrode size and shape, the devices cannot be directly compared. Although there is an obvious correlation to some sort of hydration, they are influenced by other factors in different proportions, and a major difference is related to depth penetration, which is dependent on both distance between electrodes and frequency.1,10 Another factor to consider in the context of skin hydration is the time of measurement, since moisture will gradually build up under occlusion of the electrode system due to the transepidermal water loss. Taking data at an exact point of time after application of the probe will keep this
Bioimpedance as a Noninvasive Method for Measuring Changes in Skin
variable under control, and may be handled automatically by a device. Martinsen and Grimnes4 have suggested using only one aspect of impedance, the susceptance, at only one very low frequency (in the α-range) to estimate skin hydration, based on a physically reasonable model of skin parameters, which should exclude deeper layers of the skin, while others have suggested using dielectric properties of water at very high frequencies (in the γ-range), which necessarily will penetrate deeper. The estimated hydration is thus strongly dependent on both method and device design, but may yield consistent results with correlation to some sort of hydration about which no real consensus exists. In addition, correlation between blood glucose and skin hydration has been reported by Elden,11 using a hydration meter (arbitrary units). Since blood glucose can vary rapidly, operators should also consider this phenomenon while designing protocols for the study of skin hydration.
40.4 SKIN BARRIER AND SKIN DISEASES Due to the many factors behind the function of the skin barrier, hydration being just one, and in the absence of a clear definition of skin hydration, it may be wise to study hydration and skin barrier function together. Multifactorial studies call for multifrequency methods. Due to the dramatic difference in conductivity between the stratum corneum and the living strata of skin, information pickup depth will be highly frequency dependent.10 However, if the skin barrier has been degraded physically, chemically, or by disease, information resident in deeper strata will become more accessible, while at the same time information about the degree of degradation of the skin barrier will come out. If skin hydration is not the primary interest, wet electrodes will reduce the impedance of the stratum corneum, but the medium has to be chosen carefully. Many electrode gels contain polyethylene glycol (PEG) or similar long polymers, and these will interfere with impedance measurements above around 10 kHz.12 Plain saline solution of physiological concentration will work well up to many megahertz. Inundation will also effectively eliminate some environmental factors and artifacts due to sweat. Most biologically relevant information will be found in the β-range of frequencies. Multifrequency measurements yield impedance spectra, and interpretation of data calls for some sort of data reduction. In several studies our group has applied simple indices to condense information in electrical impedance spectra. The four indices introduced are based on magnitude and phase at only two frequencies, and reflect major aspects of changes in impedance space.13
(
Magnitude index MIX = abs Z 20kHz
)
(
abs Z 500kHz
)
347
Phase index PIX = arg ( Z 20kHz ) – arg ( Z 500kHz )
(
Real part index RIX = Re Z 20kHz
(
)
(
abs Z 500kHz
Imaginary part index IMIX = Im Z 20kHz
)
(
)
abs Z 500kHz
)
where abs(Zi) is the magnitude of the impedance at the frequency i, arg(Zi) the argument (phase angle) in degrees, and Re(Zi) and Im(Zi) the real and imaginary parts, respectively. These definitions entail a normalization at the cost of one degree of freedom, and have been found to extract most of the information in spectra from skin with an intact or almost intact barrier — where a Nyquist plot is almost a straight line in the applied range of frequencies, and therefore Cole models are out of place. The usefulness of both bioimpedance as such, as well as the simplified data analysis with indices, became apparent when comparing modest skin reactions to different irritants.14 A correspondence between histological classification and patterns of the indices, i.e., bioimpedance information, was demonstrated, and the obvious question was raised whether this could be further elaborated aiming at diagnostic applications. In addition to classical bioimpedance, the device used by our group is also equipped with a mechanism to achieve controlled depth penetration (Figure 40.2).15,16 This new way to collect multidepth bioimpedance data
FIGURE 40.2 Clinical setting. The impedance spectrometer under computer control and probes for skin evaluation. The device used by our group (SciBase II, SciBase AB, Stockholm, Sweden), in addition to normal impedance spectra, incorporates depth penetration control, which adds more information to the data sets, thereby facilitating enhanced power of discrimination between various skin conditions. We have proposed the term electronic biopsies, although each diagnostic application has to be validated, and until this has been done, standard histology or established tests are the basis for diagnosis.
348
Handbook of Non-Invasive Methods and the Skin, Second Edition
yields data sets with more information about tissue structure. It was also established that bioimpedance could detect alterations below the visual threshold,17 confirming earlier observations with less sophisticated equipment.18 One interesting question is whether irritant reactions can be separated from allergic reactions. At an early stage, allergic reactions have not yet involved the skin barrier, and this may be the reason bioimpedance, at least during the early phase of reactions, seems to provide such discriminative information.19–22 Baseline levels vary with body location, age, and sex, and some major areas have been mapped.23 There are also seasonal variations.24 Proper use of reference sites (ipsilateral or contralateral) or simple monitoring of a site repeatedly over time, avoiding experiments at seasons with significant environmental alterations in exposure to sun and humidity, helps to minimize artifactual biological variations. Using the four indices, or the raw data of magnitude and phase at the same frequencies, a number of studies on relevant skin parameters, e.g., lipid content, efficacy of cosmetic products (including skin hydration preparations), and different reactivity of healthy volunteers and patients to irritants, have demonstrated significant differences.25–31 As mentioned above,11 blood glucose seems to modulate skin impedance. Would normal-looking diabetic skin differ from normal skin of nondiabetics? There is a subtle difference, but the information residing in the four indices is not sufficient to show that, but it was possible using more information from the spectra. There are several ways to do that, e.g., artificial neural networks or multivariate statistics, while Cole models should be avoided because they filter away relevant information. The diabetic skin question was resolved using multivariate statistics,32 and the difference did not appear until age and sex factors had been sorted out. Using parallel factor analysis, it was possible to demonstrate a significant difference between the inner side and outer side of the volar forearm.33 This finding should alert any study protocol designer about the importance of randomizing test sites, even if elicited responses are much above the level detected in this study. Would it make sense to extend applications of bioimpedance to applications where biopsy is normally required, i.e., could electronic biopsy sometimes replace the invasive histological method? There is already some evidence that this is feasible, although extensive clinical trials will have to validate the potential and map limitations.34,35 Due to the complexity of differential diagnostics, it is predicted that large amounts of data will be needed, i.e., many frequencies, and probably various aspects of the tissue, such as depth differentiation. Recent studies using depth-selective impedance spectroscopy on skin cancer, classical statistics as well as data analysis using neural networks, look promising.36–38 Playing with impedance parameters and
Benign nevi (n = 520) Basal cell carcinoma (n = 82) Squamous cell carcinoma (n = 6) Actinic keratosis (n = 10)
FIGURE 40.3 Using modern number crunching, it is possible to deduce a number of parameters from the multidepth–multifrequency spectra. In this example, three parameters separating various nonmelanoma skin cancers and actinic keratosis from benign lesions are displayed. The raw data set behind each point contains 310 parameters (magnitude and phase at 31 frequencies, at 5 depths). Playing with sensitivity and specificity, it is possible to set sensitivity close to 100% and achieve close to 90% specificity for the cancer types in this study.
more robust multivariate techniques, our group has already achieved a specificity of 87% at 100% sensitivity with nonmelanoma lesions (Figure 40.3). Is there an exact definition of noninvasive? The question is relevant in this context, because the skin barrier residing in the stratum corneum reduces the amount and quality of information about deeper layers of the skin, where many diseases, including skin cancer, take place. Obviously, all disorders related to the skin barrier are best studied with a totally noninvasive electrode system. However, if slight penetration through the stratum corneum, but not into the dermis, with tiny spikes is acceptable, direct contact with, e.g., skin tumors could be achieved and information would not be diluted by the otherwise dominating stratum corneum. Such devices can be manufactured using modern microelectromechanical systems (MEMS) technology, 39 and clinical research is in progress.40
40.5 IMPEDANCE IMAGING It is possible to construct images from impedance data recorded from a multitude of electrodes, methods known
Bioimpedance as a Noninvasive Method for Measuring Changes in Skin
as electrical impedance tomography (EIT), which are used mainly for functional imaging of large internal organs.41 Due to practical limitations, only one or very few frequencies are employed in EIT. A special case, not EIT, but parallel point impedance mapping, has been attempted with blunt surface electrodes as well as with somewhat invasive electrodes.42,43 A novel MEMS device, originally intended to read fingerprints with an array of 360 × 256 sensing elements, is based on electrical capacitance that is a special case of electrical impedance.44 It will be interesting to see which applications will benefit most from depth resolution, lateral resolution or multifrequency resolution for diagnostic power.
REFERENCES 1. Grimnes S, Martinsen ØG. Bioimpedance and Bioelectricity Basics. Academic Press, London, 2000. 2. Schwan HP. Electrical properties of tissue and cell suspensions. In Advances in Biological and Medical Physics, Vol. 5. Academic Press, New York, 1957, pp. 147–224. 3. Martinsen ØG, Grimnes, Sveen O. Dielectric properties of the epidermal stratum corneum. Med Biol Eng Comput 35: 172–176, 1997. 4. Martinsen ØG, Grimnes S. Facts and myths about electrical measurement of stratum corneum hydration state. Dermatology 202: 87–89, 2001. 5. Rosendal T. Concluding studies on the conducting properties of human skin to alternating current. Acta Physiol Scand 9: 39–45, 1945. 6. Yamamoto T, Yamamoto Y. Electrical properties of the epidermal stratum corneum. Med Biol Eng 14: 151–158, 1976. 7. Fluhr JW, Gloor M, Lazzerini S, Kleesz P, Grieshaber R, Berardesca E. Comparative study of five instruments measuring stratum corneum hydration (Corneometer CM 820 and CM 825, Skicon 200, Nova DPM 9003, DermaLab). Part I. In vitro. Skin Res Technol 5: 161–170, 1999. 8. Fluhr JW, Gloor M, Lazzerini S, Kleesz P, Grieshaber R, Berardesca E. Comparative study of five instruments measuring stratum corneum hydration (Corneometer CM 820 and CM 825, Skicon 200, Nova DPM 9003, DermaLab). Part II. In vivo. Skin Res Technol 5: 171–178, 1999. 9. Fischer TW, Wigger-Alberti W, Elsner P. Assessment of “dry skin”: current bioengineering methods and test designs. Skin Pharmacol Appl Skin Physiol 14: 183–195, 2001. 10. Martinsen ØG, Grimnes S, Haug E. Measuring depth depends on frequency in electrical skin impedance measurements. Skin Res Technol 5: 179–181, 1999. 11. Elden HR. U.S. Patent 5,890,489, 1999. 12. Clar EJ, Her CP, Sturelle CG. Skin impedance and moisturization. J Soc Cosmet Chem 26: 337–353, 1975.
349
13. Ollmar S, Nicander I. Information in multi-frequency measurement on intact skin. Innov Technol Biol Med 16: 745–751, 1995. 14. Nicander I, Ollmar S, Eek A, Lundh Rozell B, Emtestam L. Correlation of impedance response patterns to histological findings in irritant skin reactions induced by various surfactants. Br J Dermatol 134: 221–228, 1996. 15. Ollmar S, Eek A, Sundström F, Emtestam L. Electrical impedance for estimation of irritation in oral mucosa and skin. Med Prog Technol 21: 29–37, 1995. 16. Nicander I, Ollmar S, Lundh Rozell B, Eek A, Emtestam L. Electrical impedance measured to five skin depths in mild irritant dermatitis induced by sodium lauryl sulphate. Br J Dermatol 132: 718–724, 1995. 17. Nicander I, Ollmar S. Mild and below threshold skin responses to sodium lauryl sulphate assessed by depth controlled electrical impedance. Skin Res Technol 3: 259–263, 1997. 18. Ollmar S, Emtestam L. Electrical impedance applied to non-invasive detection of irritation in skin. Contact Derm 27: 37–42, 1992. 19. Nyrén M, Ollmar S, Nicander I, Emtestam L. An electrical impedance technique for assessment of weals. Allergy 51: 923–926, 1996. 20. Nicander I, Ollmar S, Lundh Rozell B, Emtestam L. Allergic contact reactions in the skin assessed by electrical impedance: a pilot study. Skin Res Technol 3: 121–125, 1997. 21. Nyrén M, Hagströmer L, Emtestam L. Instrumental measurement of the Mantoux test: differential effects of tuberculin and sodium lauryl sulphate on impedance response patterns in human skin. Dermatology 201: 212–217, 2000. 22. Nyrén M, Kuzmina N, Emtestam L. Electrical impedance as a potential tool to distinguish between allergic and irritant contact dermatitis. J Am Acad Dermatol 48: 394–400, 2003. 23. Nicander I, Nyrén M, Emtestam L, Ollmar S. Baseline electrical impedance measured at various skin sites: related to age and sex. Skin Res Technol 3: 252–258, 1997. 24. Nicander I, Ollmar S. Electrical impedance measurements at different skin sites related to seasonal variations. Skin Res Technol 6: 81–86, 2000. 25. Nicander I, Norlén L, Brockstedt U, Lundh Rozell B, Forslind B, Ollmar S. Electrical impedance and other physiological parameters as related to lipid content of human stratum corneum. Skin Res Technol 4: 213–221, 1998. 26. Norlén L, Nicander I, Lundh Rozell B, Ollmar S, Forslind B. Differences in human stratum corneum lipid content related to physical parameters of skin barrier function in vivo. J Invest Dermatol 112: 72–77, 1999. 27. Kuzmina N, Hagströmer L, Emtestam E. Urea and sodium chloride in moisturizers for skin of the elderly: a comparative, double-blind, randomised study. Skin Pharmacol Appl Skin Physiol 15: 166–174, 2002.
350
Handbook of Non-Invasive Methods and the Skin, Second Edition
28. Nicander I, Åberg P, Ollmar S. The use of different concentrations of betaine as a reducing irritation agent in soaps monitored visually and non-invasively. Skin Res Technol 9: 43–49, 2003. 29. Nicander I, Rantanen I, Lundh Rozell B, Söderling E, Ollmar S. The ability of betaine to reduce the irritating effects of detergents assessed visually, histologically and by bioengineering methods. Skin Res Technol 9: 50–58, 2003. 30. Kuzmina N, Hagströmer L, Nyrén M, Emtestam L. Basal electrical impedance in relation to sodium lauryl sulphate-induced skin reactions: a comparison of patients with eczema and healthy controls. Skin Res Technol 9: 357–362, 2003. 31. Kuzmina N, Duval C, Johnsson S, Boman A, Lindberg M, Emtestam L. Assessment of irritant skin reactions using electrical impedance: a comparison between 2 laboratories. Contact Derm 49: 26–31, 2003. 32. Lindholm-Sethson B, Han S, Ollmar S, Nicander I, Jonsson G, Lithner F, Bertheim U, Geladi P. Multivariate analysis of skin impedance data in long-term type 1 diabetic patients. Chemometrics Intell Lab Syst 44: 381–394, 1998. 33. Åberg P, Geladi P, Nicander I, Ollmar S. Variation of skin properties within human forearms demonstrated by non-invasive detection and multi-way analysis. Skin Res Technol 8: 194–201, 2002. 34. Emtestam L, Nicander I, Stenström M, Ollmar S. Electrical impedance of nodular basal cell carcinoma: a pilot study. Dermatology 197: 313–316, 1998. 35. Ollmar S. Making electronic biopsies into a viable future for non-invasive diagnostics with electrical impedance. Med Biol Eng Comput 37 (Suppl. 2): 116–117, 1999.
36. Beetner DG, Kapoor S, Manjunath S, Zhou X, Stoecker WV. Differentiation among basal cell carcinoma, benign lesions, and normal skin using electric impedance. IEEE Trans Biomed Eng 50: 1020–1025, 2003. 37. Åberg P, Nicander I, Holmgren U, Geladi P, Ollmar S. Assessment of skin lesions and skin cancer using simple electrical impedance indices. Skin Res Technol 9: 257–261, 2003. 38. Dua R, Beetner DG, Stoecker WV, Wunsch DC. Detection of basal cell carcinoma using electrical impedance and neural networks. IEEE Trans Biomed Eng, 51: 66–71, 2004. 39. Griss P, Enoksson P, Tolvanen-Laakso HK, Meriläinen P, Ollmar S, Stemme G. Micromachined electrodes for biopotential measurements. IEEE J Microelectromech Syst 10: 10–16, 2001. 40. Åberg P, Nicander I, Ollmar S. Minimally invasive electrical impedance spectroscopy of skin exemplified by skin cancer assessment. In Proceedings of the IEEE EMBS, Cancun, September 17–21, 2003, pp. 3211–3214. 41. Metherall P, Barber DC, Smallwood RH, Brown BH. Three-dimensional electrical impedance tomography. Nature 380: 509–512, 1996. 42. Babich YF, Bakay EA. The skin electrophysiological imaging (SED): some multiparameter observations. Med Biol Eng Comput 35 (Suppl. 1): 349, 1997. 43. Glickman YA, Filo O, David M, Yayon A, Topaz M, Zamir B, Ginzburg A, Rozenman D, Kenan G. Electrical impedance scanning: a new approach to skin cancer diagnosis. Skin Res Technol 9: 262–268, 2003. 44. Lévêque JL, Querleux B. SkinChip©, a new tool for investigating the skin surface in vivo. Skin Res Technol 9: 343–347, 2003.
of Commercial Electrical 41 Comparison Measurement Instruments for Assessing the Hydration State of the Stratum Corneum Bernard Gabard Egerkingen, Switzerland
Peter Clarys and André O. Barel Faculty of Physical Education and Physiotherapy, Vrije Universiteit Brussel, Brussels, Belgium
CONTENTS 41.1 Introduction............................................................................................................................................................351 41.2 Principle of Measuring the Electrical Properties of the Skin...............................................................................351 41.3 Description and Comparison of the Commercial Electrical Instruments.............................................................352 41.3.1 Corneometer...............................................................................................................................................352 41.3.2 DermaLab Moisture Unit ..........................................................................................................................353 41.3.3 MoistureMeter SC-4 and D-3....................................................................................................................353 41.3.4 Nova Dermal Phase Meter DPM 9003 and Petite....................................................................................353 41.3.5 Skicon 200 and 200EX Conductance .......................................................................................................354 41.3.6 Multifrequency Impedance Spectrometer .................................................................................................354 41.3.7 SkinChip: Skin Surface Capacitance Measurements ................................................................................354 41.4 Comparison of the Electrical Measurements ........................................................................................................355 41.4.1 Correlations................................................................................................................................................355 41.4.2 Accuracy ....................................................................................................................................................355 41.4.3 Sensitivity Range .......................................................................................................................................355 41.4.4 Coefficient of Variation..............................................................................................................................356 41.4.5 Depth of Measurement ..............................................................................................................................356 41.5 Conclusions............................................................................................................................................................356 References .......................................................................................................................................................................356
41.1 INTRODUCTION The presence of an adequate amount of water is an essential prerequisite for the maintenance of the normal structure and function of the stratum corneum. The in vivo determination of the degree of hydration of the horny layer is an important factor in the characterization of normal and pathological situations, of an actinic aged skin, of irritated skin conditions, and finally in the assessment of the efficiency of various moisturizing topical products.1–7
41.2 PRINCIPLE OF MEASURING THE ELECTRICAL PROPERTIES OF THE SKIN It has been known for a long time that the electrical properties of the skin are related to the water content of the horny layer.3–11 Therefore the measurement of the impedance of the skin, the total electrical resistance of the skin to an alternating current of frequency F, has been studied extensively and is the most widely used technique to assess the hydration state of the skin surface. The skin, 351
352
Handbook of Non-Invasive Methods and the Skin, Second Edition
like many biological materials, presents both resistive (R) and capacitive (C) properties, which results in a time shift between the sinusoidal curve of voltage and the sinusoidal curve for intensity of current. This time shift is called the phase angle or phase shift. As a consequence, the total impedance (Z) depends on two components, a resistance (R) and a capacitance (C) in agreement with a simple theoretical model where the skin is modeled as an electrical alternating circuit with a resistance in parallel with a capacitor.4–7 The outcome of the electrical measurements is influenced by numerous factors that may be roughly separated in biological and physical parameters. Dry stratum corneum is a typical medium of weak electrical conduction (medium of low dielectric constant or low permittivity). When this medium is hydrated, great changes in its electrical properties occur. Among biological factors, many proteins and other chemicals are involved in these electrical properties, which are related to hydration.3,9,12,13 Matrix proteins present in the corneocytes are dipolar molecules and are more or less hydrated. Various ions are also contained in the intracellular polar space of the corneocytes and in the intercellular hydrophylic space of the multilamellar lipid structures located between the corneocytes. Hydrogen-bonded dipolar structures of the water molecules are also involved. The electrical impedance of the horny layer does not depend only on the water content, but can also be further influenced by compounds other than water (ions, glycerine, emollients, etc.). Thus, as pointed out by many authors, the complex electrical impedance properties of the horny layer are dependent on the water content of the horny layer, but are also dependent on a variety of other factors.3,9,12,13 Among physical factors, the frequency of the electrical current used (kHz or MHz domain, single or multiple frequency), geometry of applied electrode on the skin surface (μm or mm distance gap between the two electrodes), design of the oscillating electronic circuit, direct galvanic contact or indirect contact through a dielectric medium between electrode, and skin surface may play important roles. By using an adequate geometry of the measuring electrodes, frequency of the current, and design of the oscillating electronic circuit, the commercially available electrical instruments measure either (1) the conductance contribution (reciprocal of resistance) of the impedance, (2) the capacitance contribution of the impedance, or (3) the total impedance (combination of both parameters).4–6 There are actually many different electrical commercially available instruments for assessing skin hydration based on the different modes of measurement: Corneometer, Nova DPM, DermaLab, Skicon, MoistureMeter, and IMP SciBase. It is the purpose of this chapter to describe in a comparative way the hardware of these different instruments and their in vivo measuring performances.
41.3 DESCRIPTION AND COMPARISON OF THE COMMERCIAL ELECTRICAL INSTRUMENTS A limited number of comparative studies about the measuring capabilities of the impedance and the capacitance method have been published.14–19 Furthermore, the conductance, capacitance, and bioimpedance methods are extensively described in Chapters 38, 39, and 40 of this book, respectively.20–22 For the description of each instrument we will consider the following items when possible and when informations is available: principle of measurement, single or multifrequency, description of the electrodes, units and range, calibration, accuracy, repeatability, and reproducibility.
41.3.1 CORNEOMETER There are different versions of the Corneometer in use: old analog versions of the CM 820 and CM 825 and, more recently, a digital version of the CM 825. The instruments are manufactured by CourageKhazaka Electronic GmbH, Köln, Germany. The principle of this instrument is based on the capacitance method. Technical descriptions of the instrument and its use have been published by many research groups4,6,14–20,23–25 and in Chapter 39 of this book.21 The measuring probe (0.7 × 0.7 cm; surface, 0.49 cm2) consists of an interdigital grid of gold-covered electrodes (75 mm wide) with interdigital spacing of 75 mm. The interdigital electrode is covered by a low dielectric vitrified material of 20 μm thickness. As a consequence, there is no direct galvanic contact between the electrode and the skin surface. A constant application pressure is applied on the skin surface through a spring system. The pressure of the CM 820 and CM 825 analog detection systems is 1.6 N/m2, whereas the new digital system works at a lower pressure (less than 1 N/m2). The form and depth of the electric field present in the skin depend on the geometry of the electrodes and the dielectric material covering the electrodes (constant capacitance, Co) and of the capacitance of the biomaterial in contact with the electrode (variable capacitance, C). As a consequence, the whole system (electrode, horny layer, and upper parts of the epidermis) works as a variable capacitor. The total capacitance is only influenced by changes in the dielectric constant of the biomaterial in contact with the electrodes. A dry horny layer is a dielectric medium. When the horny layer is hydrated, a significant change in the dielectric properties of this medium is observed. As a consequence, the total capacity of this system is changed in function of the degree of hydration of the skin (mostly in the horny layer).
Comparison of Commercial Electrical Measurement Instruments
A resonating system in the instruments measures the shift in frequency of the oscillating system that results from the changes in the total capacity of the skin surface. In the case of the CM 820, the frequency shifts from 40 KHz for a hydrated medium to 75 KHz for a dry medium. In the case of the new versions of the Corneometer (CM 825 with the analog and digital probe), the frequency shifts from 0.95 for a hydrated medium to 1.15 MHz for a dry medium. The capacity of the skin surface is automatically measured after 1.0 to 1.5 seconds of application of the probe on the skin. The variable total capacitance of the skin surface is converted in arbitrary units (a.u.) of skin hydration varying from 0 to 120; arbitrary units are not directly related to real electrical units or related in a simple way to the water content of the horny layer. Calibration of the instrument is carried out in the factory using three standard materials with known capacitance. In vitro calibration is possible using a cellulose filter impregnated with a 0.15 M NaCl solution and subsequently covered with a 15-μm plastic foil, giving values of 120 and 10 to 20 a.u., respectively. The new apparatus (digital version of the Corneometer CM 825) is connected to a central multiprobe unit MPA 5, which operates on Windows software. Hydration results are displayed as either 10 independent consecutive values (mean value + SD) or kinetic curves in function of time (at 3 and 7 seconds and 7 minutes). The capacitance method shows in vivo a broad range of sensitivity from 20 to 110 a.u. Unfortunately, when comparing the data of the old version (CM 820) and the new version (CM 825), systematically significant higher values (±60%) are observed with the older version.4 In a multicenter study, the analog and digital versions of the Corneometer CM 825 were compared and a very high correlation (r = 0.92) was noticed. The analog and digital probes give identical hydration values.24 Practically, for the Cornemeter CM 825, values of hydration below 30 a.u. correspond to very dry skin; between 30 and 40, dry skin; and higher than 40, normal skin.24
41.3.2 DERMALAB MOISTURE UNIT The DermaLab is manufactured by Cortex Technology (Hadsund, Denmark). Technical descriptions of the instrument and its use have been published by a few researchers.18,26 The instrument is based on impedance measurements operating at a single frequency (100 kHZ) and delivers impedance units from 0 to 10,000 μmho (μS). The calibration is carried out in the factory and the instrument can be checked or calibrated using a phantom. The instrument is a combination instrument that allows measurements of elasticity, transepidermal water loss (TEWL), and hydration. It consists of a central unit
353
with a printer connected to a moisture probe. The central unit can be connected to a PC using a standard software of an acquisition program. There are two moisture probes available: a classical flat-faced moisture probe and a moisture pin probe. The flat-faced electrode (12.5 mm diameter, 1.23 cm2 surface) features three concentric gold-covered ring electrodes separated by a distance of about 1 mm. There is a direct galvanic contact between the electrodes and the skin. A spring ensures a constant application pressure (90 g/cm2) on the skin and triggers the measurement of impedance. The pin probe electrode (same external dimension) features a central circular electrode surrounded by eight small pins. This electrode is recommended by the manufacturer for use on a rough and scaly skin surface and for minimizing moisture accumulation under the probe surface during prolonged measurements.
41.3.3 MOISTUREMETER SC-4
AND
D-3
These instruments are manufactured by Delfin Technologies Ltd. (Kuopio, Finland). The development of the MoistureMeter is very recent; only one published source of scientific information is available,20 and technical information can be obtained from the manufacturer.27 The MoistureMeter SC-4 is a novel capacitive device operating at 1.25 MHz with two concentric electrodes. The inner electrode has a diameter of 5 mm, and the inner and outer diameters of the concentric electrode are 9 and 22 mm, respectively. The system works with a variable contact pressure from 0.69 to 2.06 N. A recommended contact pressure corresponds to 140 to 210 g or 1.37 to 2.06 N. The instrument shows arbitrary capacitance units (normalized hydration values) that correspond to units of a dielectric constant divided by the thickness of the horny layer. The device is factory calibrated using a 100-pF capacitor. In vitro calibration is possible using a cellulose filter impregnated with a saline solution covered with a 10-μm polyethylene foil, giving a measured value of 21 normalized hydration units. Another model with probes with increasing distances between the inner and outer electrode has been developed by Delphin (MoistureMeter D-3). With increasing dimensions of the probe (sensor contact diameter going from 10 to 55 mm), it is possible to increase the effective measuring depth from 0.5 to 5 mm.
41.3.4 NOVA DERMAL PHASE METER DPM 9003 AND PETITE These instruments are manufactured by Nova Technology Corporation (Portsmouth, NH). Technical descriptions of the instruments29 and their uses have been published by two research groups.15,18,28
354
Handbook of Non-Invasive Methods and the Skin, Second Edition
The instrument measures impedance-based capacitive reactance of the skin at preselected frequencies up to 1 MHz from the observed signal phase delays. By integrating measurements at the preselected frequencies, capacitive reactance is calculated from the signal and phase delay using a proprietary integrated circuit in the instrument. The final readout is given in arbitrary DPM units, ranging from 90 to 999 DPM units, which are directly related to the capacitance. The standard DPM 9103 sensor probe features two concentric brass ring electrodes separated by an isolator (with respective inner and outer diameters of 4.34 and 8.76 mm). The distance between the inner and outer electrode is 1 mm. There is a direct galvanic contact between the electrodes and the skin. Other probes are available: the smaller DPM 9105 sensor (with respective inner and outer diameters of 5.08 and 2.54 mm) and the very small DPM 9107 sensor (with respective inner and outer diameters of 3.81 and 1.52 mm). An individual use/disposable tip of the probe (DPM 9111) and an articulated probe tip are available for clinical and medical studies. Constant pressure of application is assured by means of a sensor switch spring system. An automatic calibration takes place, ensuring standardization of the instrument before taking any readings. The measuring principle and the automatic calibration are considered proprietary and not accessible for publication. Recently, an easy portable Nova Petite instrument has been developed, working on a Palm PDA system employing a small probe. The instrument can be interconnected with a PC or to the Internet.
41.3.5 SKICON 200
AND
200EX CONDUCTANCE
The instruments based on the experimental device developed by Tagami3 and coworkers are manufactured by ISBS Co. (Hamamatsu, Japan). Technical descriptions of the instruments and their use have been published by different authors3,13,14,16,17,19,30,31 and by the manufacturer.32 The Skicon 200 measures the conductance of a single high-frequency current of 3.5 MHz. The measuring probe (surface, 0.28 cm2) consists of two concentric gold-covered electrodes (with external diameters of 2 and 5 mm, respectively). The distance between the inner and outer electrode is 1 mm. There is a direct galvanic contact between the electrodes and the skin. The apparatus gives readings of the conductance of the skin in microsiemens (μS) units, ranging from 0 to 1999 μS. An internal device with a standard conductance of 300 μS is used to calibrate the instrument. The conductance of the horny layer, in particular, may be considered a representative parameter of the skin surface. The Skicon 200EX is a more recent version of the Skicon and also measures the conductance of a single high-frequency current of 3.5 MHz. A comparison
between the two versions of the Skicon and the Corneometer CM 825 has been published.19 The new measuring probe consists of concentric interdigital electrodes of 75 μm and a gap of 200 μm between the electrodes. The outer diameter of the electrode is 6 mm (surface of the probe, 0.65 cm2). The manufacturer claims that due to the small distance between the electrodes (200 μm), a better contact is established between the electrodes and the skin surface, particularly in the case of a scaly and rough surface of the skin. Pressure application of the electrode on the skin surface is constant using a spring (190 g/cm2). There is a direct galvanic contact between the electrodes and the skin. The apparatus gives readings of the conductance of the skin in microsiemens units, ranging from 0 to 2000 μS.
41.3.6 MULTIFREQUENCY IMPEDANCE SPECTROMETER The impedance spectrometer, based on the experimental device developed by Ollmar and coworkers,33 is manufactured by SciBase AB (Huddinge, Sweden). Technical descriptions of the instrument and its use have been published by Ollmar and coworkers33–35 and others6 and also by the manufacturer.36 The instrument works on an impedance-based mode, which records the magnitude (in ohms) and phase (in degrees) at 31 logarithmically distributed frequencies between 1 kHz and 1 MHz and from five different depth settings (from 0.1 to 2.0 mm). The measurement probe consists of two concetric electrodes. The instrument delivers through a SciBase software four impedance indices that emphasize different aspects of the impedance properties of the living tissues: magnitude index (MIX, dimensionless), phase index (PIX, in degrees), real part index (RIX, dimensionless), and imaginary part index (IMIX, dimensionless). It is possible to select the range of depth measurements: from 0.1 to 2.0 mm. The four indices are related to the impedance properties of the epidermis and the superficial dermis, and much less to the hydration state of the horny layer. Unfortunately, it appears that this device is not frequently used in Europe or the U.S.
41.3.7 SKINCHIP: SKIN SURFACE CAPACITANCE MEASUREMENTS This experimental device has been developed by the l’Oréal research group37 and is based on the fingerprintsensing technology. The sensor is composed of an array of 360 × 256 active capacitance microsensors. For each microsensor, the effective feedback capacitance is modulated by the contact between the skin surface and the sensor surface. One obtains capacitance images of the skin surface with a resolution of 50 μm. At the present time, this device is experimental and not commercially available. With few doubts, it will have a promising future in dermato-cosmetic domains.
Comparison of Commercial Electrical Measurement Instruments
41.4 COMPARISON OF THE ELECTRICAL MEASUREMENTS It must be remembered that the commercially available electrical instruments measure either (1) the conductance contribution (reciprocal of resistance) of the impedance, (2) the capacitance contribution of the impedance, or (3) the total impedance (combination of both parameters) of the upper layers of the skin (mostly due to the horny layer). The instruments deliver either real electrical units, microsiemens, ohms, or arbitrary units. In both cases, these complex electrical impedance properties of the horny layer are dependent on the water content of the horny layer, but also dependent on a variety of other factors. As a consequence the electrical results are in most cases not related in a simple way only to water content (no linear relation between results and percent water content).
41.4.1 CORRELATIONS As already mentioned in the introduction, very few comparative studies using different commercial instruments have been carried out. Only in such studies the real performances of the instruments could be compared. Comparison of the hydration measurements obtained by the capacitance method (Corneometer CM 820) with data obtained by the impedance method (Skicon 100), which can be directly calibrated with a reference of known conductance, showed a very high degree of correlation (r = 0.94) between both instruments over a broad range of hydration values.14 Gabard and Treffel15 investigated the performances of the Nova DPM 9003 and the Corneometer CM 820. The respective performances of the Corneometer CM 820 and 825, Skicon 200, and Nova were investigated by Barel and Clarys16 and Clarys et al.17 Excellent correlations were noticed between the three instruments (r varies from 0.89 to 0.97). Fluhr et al.18 compared the performances of the Corneometer CM 820 and 825, the Skicon 200, the Nova DPM 9003, and the DermaLab. Excellent correlations were noticed between the four different instruments (r varies from 0.87 to 0.94). Alanen et al.20 compared the performances of the new MoistureMeter and the Corneometer CM 820. A good correleation of 0.75 was noticed between both instruments. In a multicenter study, the analog and digital versions of the Corneometer CM 825 were compared and a very high correlation (r = 0.92) was noticed. The analog and digital probes give identical hydration values.24
41.4.2 ACCURACY The instruments deliver either real electrical units, microsiemens, ohms, or arbitrary units. In the case of real electrical units, a calibration with phantoms of known electrical units is possible (Skicon 200 and DermaLab)
355
or can be calculated (MoistureMeter). In the case of arbitrary units (Corneometer, Nova), a calibration with a standard of known electrical properties is impossible, and this criterium is not applicable. Furthermore, the principle of internal calibration in the manufacturing is considered by many companies as proprietary and not accessible for publication. However, some in vitro calibration can be carried out using simple model systems such as cellulose filters of known thickness (from 150 μm to 1 mm) impregnated with aqueous solutions and with solutions of known dielectric constant. A cellulose filter disc impregnated with 0.15 M NaCl aqueous solution gives a maximal hydration value of ±120 Corneometer a.u.17,18 When covering the surface of the saturated film with a layer of a low dielectric plastic film of 15 μm thickness,25 one measures a hydration value of about 20 a.u. The capacitance method (CM 820 and CM 825) shows in vivo a range of sensitivity from 20 to 110 a.u. However, it must be noted that when comparing the data of the old version (CM 820) and the new version (CM 825), systematically significant higher values (±60%) are observed with the older version.4 This means that comparison of absolute hydration values between the CM 820 and CM 825 versions is not possible. Furthermore, in vitro measurements on cellulose filters showed clearly high correlations, respectively, between the capacitance values and the dielectric constant of the liquid saturating the filter paper and between the capacitance values and the amount of water impregnating the filter in a sorption test.17 Similar results showing high correlations for both criteria (dielectric constant and amount of water) were noticed for the Corneometer, DermaLab, and Nova DPM.18 In conclusion, the determination of absolute values of hydration is generally not very useful; these devices are more useful in measuring relative variations (after a dermato-cosmetic treatment, after irritation, or when comparing healthy with diseased skin surfaces).
41.4.3 SENSITIVITY RANGE Taking in account the results of the studies of Barel et al.16,17,19 on the Corneometer CM 820 and CM 825 and the Skicon 200 and 200EX, the results of Fluhr et al.18 on the Corneometer CM 820 and 825, the Skicon 200, the Nova DPM 9003, and the DermaLab, and the recent study of Alanen et al.20 on the new MoistureMeter, the sensitivity range of the different instruments can be compared. Our group16,17,19 and others20 considered in vivo four different states of the skin surface (at different anatomical sites): very dry, dry, hydrated, and very hydrated. The four states of the horny layer were evaluated by the four instruments, and the following ranges of measuring capabilities were noticed:
356
Handbook of Non-Invasive Methods and the Skin, Second Edition
Corneometer CM 820: 67 to 104 a.u.; range, 1.55 Corneometer CM 825: 40 to 100 a.u.; range, 2.50 Nova: 113 to 350 a.u.; range, 3.09 Skicon 200: 17 to 500 μS; range, 29.41 Skicon 200EX: 10 to 586 μS; range, 58.60 MoistureMeter: 18 to 72 a.u.; range, 4.00 Similar conclusions concerning the sensitivity range of the CM 820 and CM 825, Skicon, Nova, and DermaLab can be drawn from the moisture accumulation tests and from the sorption–desorption tests (Fluhr et al.18). As an example, the moisture accumulation test gave the following results: Corneometer CM 820: 76 to 115 a.u.; range, 1.51 Corneometer CM 825: 41 to 69 a.u.; range, 1.68 Nova: 156 to 323 a.u.; range, 2.09 DermaLab: 123 to 427 μS; range, 3.47 Skicon: 32 to 295 μS; range, 9.22 The following rank order for range of variation in the hydration state can be proposed: CM 820 < CM 825 < Nova < DermaLab < Skicon 200 < Skicon 200EX In agreement with many authors,4–7,11,14–19,21,23 the capacitance method is very sensitive for measurements at low hydration and less sensitive in the range of very high hydration values (above 110 a.u.).28,30 The Skicon 200 lacks sensitivity at very low hydration values and is not very suited for quantifying a situation of very dry skin or skin with keratosis. The performances of the Skicon 200EX with small concentric interdigital electrodes are better for evaluating dry skin.The DPM also lacks some sensitivity at very low hydration levels. With their much larger range of hydration values, the DermaLab and particularly the two models of Skicon are well suited for quantifying states of hydrated and very hydrated horny layers.
41.4.4 COEFFICIENT
OF
VARIATION
The mean values of coefficients of variation (in percent) for different anatomical sites of the five different instruments were evaluated by Fluhr et al.18: CM 820, 15%; CM 825, 25%; DermaLab, 34%; Nova, 55%; and Skicon, 91%. These mean values of coefficients of variation were more or less confirmed by our group16,17 and others15: CM 825, 14%; Nova, 20%; Skicon 200, 47%; and Skicon 200EX, 35%. The reproducibility of the Corneometer, Nova, and DermaLab, respectively, is reasonably good. As pointed out by many authors,16–18 the coefficient of variation is relatively high for the two versions of the Skicon. This is
a major drawback of these instruments; only large variations in the hydration state of the skin can be demonstrated in a significant way with these instruments.
41.4.5 DEPTH
OF
MEASUREMENT
In vitro hydration measurements were carried out on cellulose filter paper saturated with a 0.15 M NaCl aqueous solution at the surface of the filter and through increasing layers of a low dielectric plastic film of 15 to 20 μm thickness.16,18 The electrical field of measurement for the different instruments can be evaluated from the thickness of the foil necessary to reach the 0 or minimum hydration level. The instruments cover the following depths: less than 15 or around 20 μm for the two versions of Skicon, around or less than 15 μm for the DermaLab, 95% reduction after 30 to 45 μm for the CM 820, after 20 to 30 μm for the CM 825, and after 40 to 60 μm for the Nova. The new probe of the Corneometer CM 825 has a lower depth of detection (20 to 30 μm) than the previous model (CM 820). The Skicon 200 and 200EX measure hydration very superficially (less than 20 μm).
41.5 CONCLUSIONS Taking in account the comparative performances, such as high degree of correlation between the instruments, the possibility of in vitro calibration on model systems, the range of sensitivity, the depth of electrical field, and the reasonably good variability, this review demonstrates the validation of the Corneometer, DermaLab, Nova, and Skicon for assessing the hydration state of the horny layer. The two other devices, MoistureMeter and multifrequency impedance meter, are also valuable but are less used in routine. Routine instrumental hydration measurements appear to be rather simple to carry out. But in order to substantiate claims with these commercial devices, it is necessary to emphasize the importance of standardization and calibration of the measurements, which must be conducted under well-controlled conditions. The final choice of the instrument will be made in function of the following criteria: price, availability, scientific information about the measuring principle, and technical service offered by the manufacturer in case of technical problems.
REFERENCES 1. Lévêque, J.L., Méthodes expérimentales d’étude du vieillissement cutané chez l’homme, In Vivo Acta Dermatol. Venereol., 144, 1279, 1987. 2. Lévêque, J.L., Grove, G., de Rigal, J., Corcuff, P., Kligman, A.M., and Saint Leger, D., Biophysical characterization of dry facial skin, J. Soc. Cosmet. Chem., 82, 171, 1987.
Comparison of Commercial Electrical Measurement Instruments
3. Tagami, H., Impedance measurements for evaluation of the hydration state of the skin surface, in Cutaneous Investigation in Health and Disease, Noninvasive Methods and Instrumentation, Lévêque, J.L., Ed., Marcel Dekker, New York, 1989, chap. 5. 4. Wilhelm K.P., Possible pitfalls in hydration measurements, in Skin Bioengineering: Techniques and Applications in Dermatology and Cosmetology, Elsner, P., Barel, A.O., Berardesca, E., Gabard, B., and Serup, J., Eds., Karger, Basel, 1998, p. 223. 5. Barel, A.O., Clarys, P., and Gabard, B., In vivo evaluation of the hydration state of the skin: measurements and methods for claim support, in Cosmetics: Controlled Efficacy Studies and Regulations, Elsner, P., Merk, H.F., and Maibach, H.I., Eds., Springer, Berlin, 1999, p. 57. 6. Bernengo, J.C. and de Rigal, J., Techniques physiques de mesure de l’hydratation du Stratum Corneum, in Physiologie de la peau et explorations fonctionelles cutanée, Agache, P., Ed., Editions Médicales Internationales, Cachau, France, 2000, p. 117. 7. Barel, A.O., Product testing: moisturizers, in Bioengineering of the Skin: Skin Biomechanics, Elsner, P., Berardesca, E., Wilhelm, K.P., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 2002, p. 241. 8. Clar, E.J., Her, C.P., and Sturell, C.G., Skin impedance and moisturization, J. Soc. Cosmet. Chem., 26, 337, 1975. 9. Lévêque, J.L. and de Rigal, J., Impedance methods for studying skin moisturization, J. Soc. Cosmet. Chem., 34, 419, 1983. 10. Mosely, H., English, J.S., Coghill, G.M., and Mackie, R.M., Assessment and use of a new skin hygrometer, Bioeng. Skin, 1, 177, 1985. 11. Blichman, C.W. and Serup, J., Assessment of skin moisture. Measurement of electrical conductance, capacitance and transepidermal water loss, Acta Derm. Venereol. (Stockh.), 68, 284, 1988. 12. Loden, M. and Lindeberg, M., Product testing of moisturizers, in Bioengineering of the Skin: Water and Stratum Corneum, Elsner, P., Berardesca, E., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1994, p. 275. 13. Tagami, H., Hardware and measuring principle: skin conductance, in Bioengineering and the Skin: Water and the Stratum Corneum, Elsner, P., Berardesca, E., and Maibach, H., Eds., CRC Press, Boca Raton, FL, 1994, p. 197. 14. Barel, A.O., Clarys, P., Wessels, B., and de Romsée, A., Non-invasive electrical measurement for evaluating the water content of the horny layer: comparison between the capacitance and the conductance measurements, in Prediction of Percutaneous Penetration: Methods, Measurements, Modelling, Scott, R.C., Guy, R.H., Hadgraft, J., and Boddé, H.E., Eds., IBC Technical Services Ltd., London, 1991, p. 238. 15. Gabard, B. and Treffel, P., Hardware and measuring principle: the Nova DPM 9003, Bioengineering and the Skin: Water and the Stratum Corneum, Elsner, P., Berardesca, E., and Maibach, H., Eds., CRC Press, Boca Raton, FL, 1994, p. 177.
357
16. Barel, A.O. and Clarys, P., In vitro calibration of the capacitance method (Corneometer CM 825) and conductance method (Skicon 200) for the evaluation of the hydration state of the skin, Skin Res. Technol., 3, 107, 1997. 17. Clarys, P., Barel, A.O., and Gabard, B., Non-invasive electrical measurements for the evaluation of the hydration state of the skin: comparison between three conventional instruments: the Corneometer, the Skicon and the nova DPM, Skin Res. Technol., 5, 14, 1999. 18. Fluhr, J.W., Gloor, M., Lazzerini, S.L., Kleesz, P., Grieshaber, R., and Berardesca, E., Comparative study of five instruments measuring stratum corneum hydration (Corneometer CM 820 and CM 825, Skicon-200, Nova DPM 9003 and Dermalab). Part I. In vitro. Part II. In vivo, Skin Res. Technol., 5, 161, 171, 1999. 19. Barel, A.O., Lambrecht, R., and Clarys, P., In Vivo Evaluation of the Hydration State of Dry Skin Using Electrical Methods. Comparison of the Skicon 200 and 200EX and the Corneometer CM825, paper presented at Skin in Health and Disease: The Noninvasive Approach, Joint ISBS, ISCS and SFIC Congress, Paris, June 27–28, 2002. 20. Alanen, E., Nuutinen, J., Nicklén, K., Lahtinen, T., and Mönkkonem, J., Measurement of hydration in the stratum corneum with the MoistureMeter and comparison with the Corneometer, Skin Res. Technol., 10, 32, 2004. 21. Barel, A.O. and Clarys, P., Measurement of capacitance, in Handbook of Non-invasive Methods and the Skin, 2nd ed., Serup, J., Jemec, G.B.E., and Grove, G., Eds., CRC Press, Boca Raton, FL, 2005, Chapter 2.5.5. 22. Nicander, I. and Ollmar, S., Bioimpedance as a noninvasive method for measuring changes in skin, in Handbook of Non-Invasive Methods and the Skin, 2nd ed., Serup, J., Jemec, G.B.E., and Grove, G., Eds., CRC Press, Boca Raton, FL, 2005, Chapter 40. 23. Courage, W., Hardware and measuring principle: Corneometer, in Bioengineering and the Skin: Water and Stratum Corneum, Elsner, P., Berardesca, E., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1994, p. 171. 24. Heinrich, U., Koop, U., Leneveu-Duchemin, M.C., Osterrieder, K., Bielfeldt, S., Charkaut, C., Degwert, J., Häntschel, D., Jaspers, S., Nissen, H.P., Rohr, M., Sc hneider, G., and Tronnier, H., Members of the DGK Task Force “Skin Hydration,” multi-center comparison of skin hydration in terms of physical, physiological and product dependent parameters by the capacitive method (Corneometer CM 825), J. Cosmet. Sci., 25, 31, 2003. 25. Courage-Khazaka Electronic GnbH, Köln, Germany, Technical information concerning the analog instruments CM820 and CM825 and the digital version of the CM 825, 2004. 26. Dermalab, Cortex Technology, Hadsund, Denmark, Technical information of the Dermalab, 2004. 27. Tagami, H., Measurement of electrical conductance and impedance, in Handbook of Non-invasive Methods and the Skin, 2nd ed., Serup, J., Jemec, G.B.E., and Grove, G., Eds., CRC Press, Boca Raton, FL, 2005, Chapter 38.
358
Handbook of Non-Invasive Methods and the Skin, Second Edition
28. Murray, B.C. and Wickett, R.R., Sensitivity of Cutometer data to stratum corneum hydration level. A preliminary study. Skin Res. Technol., 2, 167, 1996. 29. The Nova Dermal Phase Meter Nova DPM 9003, Nova Technology Corporation, Portsmouth, NH, Technical information of the Nova DPM 9003, 2004. 30. Tagami, H., Quantitative measurements of water concentration of the stratum corneum in vivo by high frequency current, Acta Derm. Venereol. (Stockh.), 185, 29, 1994. 31. Moss, J., The effect of 3 moisturizers on skin surface hydration. Electrical conductance (Skicon-200), capacitance (Corneometer CM420) and transepidermal water loss (TEWL), Skin Res. Technol., 2, 32, 1996. 32. Skicon, ISBS Co., Hamamatsu, Japan, Technical information of the version 200 and 200-EX, 2004.
33. Nicander, S., Ollmar, S., Lundh Rozell, B., Eek, A., and Emtestan, L., Electrical impedance measured to five skin depths in mild irritant dermatitis induced by sodium lauryl sulphate, Br. J. Dermatol., 132, 718, 1995. 34. Ollmar, S. and Nicander, I., Information on multi-frequency measurement on intact skin, Innov. Bio. Med., 16, 745, 1995. 35. Nicander, I., Nyrèn, M., Emtestan, L., and Ollmar, S., Baseline electrical impedance measurements at various skin sites, Skin Res. Technol., 3, 252, 1997. 36. Multifrequency IMP Spectrometer, SciBase AB, Huddinge, Sweden, Technical information about the instrument, 2004. 37. Lévêque, J.L. and Querleux, B., SkinChip, a new tool for investigating the skin surface in vivo, Skin Res. Technol., 9, 343, 2003.
Desquamation
to Determine Desquamation 42 Methods Rate C. Edwards Cardiff Biometrics Ltd., Cardiff, United Kingdom
CONTENTS 42.1 Introduction............................................................................................................................................................361 42.2 Measurement Techniques ......................................................................................................................................362 42.2.1 Collection Techniques................................................................................................................................362 42.2.1.1 The Chamber Technique ............................................................................................................362 42.2.1.2 Forced Desquamation.................................................................................................................362 42.3 Staining Techniques...............................................................................................................................................363 42.3.1 Visual and Photometric Techniques ..........................................................................................................363 42.3.2 Image Analysis Technique.........................................................................................................................365 42.3.3 Microscopic Techniques ............................................................................................................................365 42.4 Tape and Adhesive Methods..................................................................................................................................366 42.5 Cohesion ................................................................................................................................................................367 42.6 Conclusion .............................................................................................................................................................368 References .......................................................................................................................................................................368
42.1 INTRODUCTION The integrity of the stratum corneum is vital to the constancy of the mammalian internal environment. Its barrier functions are quite remarkable when it is recognized that they are undisturbed by the continuous formation of the structure in its deeper parts, as well as the loss of cells in its superficial zone. The loss of stratum corneum at the skin surface (desquamation) is an integral and essential part of epidermal physiology that takes place in a controlled manner by the loss of single-horn cells (corneocytes). It is clear that there is a relationship between epidermal cell production and desquamative loss, but the details of the link between these two facets of the population dynamics of the epidermis are obscure. Factors influencing cell production, including cytokine and mediator release, as well as the action of endocrine secretions, do so in a different time frame than factors influencing desquamation, such as the external mechanical stimulation of clothing, toilet, or social contacts. Not only do the two sets of factors differ as far as time frame is concerned, but it appears likely that production and desquamation each have a separate buffering capacity. It seems highly likely that the epidermal cell population size stays constant over a period of several
days, but short-term inequalities may well occur in one or another of the compartments. Regardless of the link between epidermal cell population and desquamation, measurement of that at which the latter occurs is an important descriptor of this part of the skin. In particular, determination of the rate of desquamation is helpful to the understanding of the vagaries of normal skin physiology and disease processes and in interpreting the actions of drugs on the skin. The mechanical integrity of the stratum corneum is an essential part of its protective function and is largely determined by the forces that bind individual corneocytes to their neighbors. This intracorneal cohesive force must, however, reduce to zero in order that corneocytes at the surface can be shed. Intracorneal cohesion is therefore an integral part of the mechanism of desquamation. The study of how intracorneal cohesion varies with depth in the stratum corneum, and how it varies with different physiological, ambient, and disease states, is of interest to skin biology, dermatology, and cosmetic science. One further comment before detailed descriptions of the methods are given is that it is important to distinguish desquamation from scaling. Scaling is due to the individual corneocytes failing to separate one from the other at the skin surface, so that clumps of horn cells come off, 361
362
Handbook of Non-Invasive Methods and the Skin, Second Edition
and is not directly related to the rate of desquamation. Measurement of the rate of desquamation can be direct, by counting the number of squames released at the surface, or indirect, in which the rate of loss of a stain in the stratum corneum is measured and taken to indicate epidermal turnover, or desquamation.
42.2 MEASUREMENT TECHNIQUES 42.2.1 COLLECTION TECHNIQUES 42.2.1.1 The Chamber Technique The passive or chamber technique of measurement of passive desquamation relies on the collection of all squames shed from the skin over a predetermined period. Roberts and Marks1 described the use of chambers consisting of Perspex cylinders of 1 cm internal diameter, or a 3.14 cm2 area containing a bung with a glass fiber insert to allow free flow of water vapor, thus avoiding occluding the sample skin site. Nonirritant adhesive material was used to attach the chamber to the skin, and the whole assembly was covered with a foam pad. The enclosed volume was 2 ml. After 48 hours the shed and loosely adherent squames were harvested. A total of 2 ml of 0.06 M Triton X-100 phosphate-buffered solution was carefully introduced into the chamber. Triton X-100 is a nonionic surfactant that acts to prevent clumping of the squames and helps to keep them in suspension. The solution was then withdrawn and aliquots placed into a hemocytometer to count the corneocytes. Results were quoted as numbers of corneocytes collected per square centimeter per hour. 42.2.1.2 Forced Desquamation Forced desquamation occurs when the measurement process applies some force to remove surface squames. Generally, this is done by gently scrubbing or rubbing the surface, or by the application of some sort of adhesive, on either tape or glass slides.
Early methods for quantitating desquamation were based on techniques first developed to collect cutaneous microflora.2 These cup scrub techniques consisted of holding an open-ended cup onto the skin, placing a wash fluid in the cup, and agitating or scrubbing the skin surface to release loose material. McGinley et al.3 used a glass cylinder of 3.8 cm2 held tightly against the skin. One milliliter of Triton X-100 phosphate-buffered solution was introduced and the skin surface rubbed vigorously with a smooth Teflon rod for 1 minute. These authors stated that the scrubbing should be with firm pressure, and that experience is necessary before repeatability is obtained. The wash fluid was removed with a pipette, and the wash–scrub procedure repeated with clean wash fluid. The two wash samples were then pooled. The squames were then stained with a solution made from equal volumes of Hucker’s crystal violet and basic fuschin.4 This was done by dropping 0.05 ml of stain into 1 ml of wash sample, and the stained suspended squames were counted in a hemocytometer. Nicholls and Marks5 improved this method by building a motorized scrubber. A motor-driven smooth Perspex blade was held against the skin with a known and adjustable pressure in a modified cup apparatus. They used the same Triton X-100 phosphate-buffered solution described earlier, instilling this through a side arm, but did not stain the solution before counting suspended squames in a hemocytometer. In a later report, Roberts and Marks1 described a handheld version of the motorized scrub apparatus, which they called the desquamator. Scrub application force was monitored as the torque required to rotate the blade against the skin and was kept constant by an electronic feedback circuit. Both the torque and the time of scrubbing were adjustable, and typical settings were 2 g-cm torque and a scrub time of 10 seconds. This latter study demonstrated a site variation for corneocyte shedding and a good relationship between the passive chamber technique and the forced desquamation technique (Table 42.1).
TABLE 42.1 Comparison of Regional Variations in Rates of Corneocyte Loss Using Passive and Scrub Techniques with Dansyl Chloride Fluorescence Extinction Times Site
Scrub Technique (corneocytes/cm2/10-second scrub)
Chamber Technique (corneocytes/cm2/hour)
Forearm Upper arm Abdomen Thigh Back
104,243 (SD 15,570) 63,660 (SD 5893) 55,623 (SD 8215) 67,638 (SD 4736) 81,962 (SD 12,929)
1309 (SD 328) 795 (SD 141) 646 (SD 85) 878 (SD 140) 1094 (SD 290)
Note: DC = dansyl chloride.
DC Extinction Time (days) 18.5 25.8 25.9 26.5 23.6
(SD (SD (SD (SD (SD
4.6) 2.7) 5.4) 5.4) 5.2)
Methods to Determine Desquamation Rate
Corcuff et al.6 used an air current generated by a turbine device to collect naturally shed corneocytes. In a later development, Corcuff et al.7 described a similar device that incorporated a woolen pad as the friction element on a motor-driven stainless-steel disc. The wool pad was chosen to resemble human hair in its biochemical nature and morphological organization, and was used to simulate the action of shampooing. The ratio of numbers of corneocytes collected by shampooing to the number of spontaneously shed corneocytes (24:1 in normal subjects, 32:1 in subjects suffering from dandruff) was similar to the ratio of wool scrubbed to spontaneous shed (22:1) or hair rubbed to spontaneous shed (24:1) The wool turbine device was also used to study the kinetics of desquamation after ultraviolet radiationinduced injury.
42.3 STAINING TECHNIQUES 42.3.1 VISUAL
AND
PHOTOMETRIC TECHNIQUES
The cup or chamber scrub techniques, although very useful direct measures of numbers of corneocytes released, have the disadvantage of interfering in a physical or mechanical way with the normal environment and external forces experienced by the skin on different body sites. Staining methods aim to apply a substantive stain to the stratum corneum, the intensity of which can in some way be assessed or quantified. If the stain penetrates and is substantive, the entire thickness of the stratum corneum is labeled, and when the stained stratum corneum has been completely shed, the stain will have disappeared and the time taken is the stratum corneum turnover time. If the intensity of staining is measured several times over a period of a few weeks, then a graph of stain remaining vs. time can be constructed. This allows for a line-of-bestfit procedure to be applied to the data, and a best estimate of complete disappearance of the stain will be given by the intercept on the time axis. Applications of the method differ in the stains used and in the methods of measuring the amount or distribution of remaining stain. Sutton9 stained the horny layer with silver nitrate and observed the time required for disappearance of the black color. This method was modified by Roberts and Marks,1 who used a photographic developer to reduce the nitrate to metallic silver and then measured the intensity of staining with standardized photographic photometry. They found that the stain thus obtained was not wholly substantive, when sites washed in the normal way were compared to unwashed sites. Baker and Kligman10 noted that silver nitrate can cause irritant responses and can also stimulate mitotic activity. They introduced the use of a fluorescent dye, tetrachlorsalicylanilide (TCSA), in a vehicle of ethylene
363
glycol monomethyl ether (EGME). This was an antiseptic that bound strongly to the stratum corneum. Their technique consisted of first establishing the time of penetration of the whole stratum corneum of each volunteer by applying the dye for 1, 1.5, and 2 hours on different sites. These sites were then stripped down to the glistening layer (the stratum granulosum) to check penetration. The time of application to the first site to show dye penetration to and including the glistening layer was used in subsequent tests as the application time. After staining, the sites to be investigated were examined daily using a Woods lamp (UVA light of around 365 nm) to excite fluorescence. These authors checked that the dye was sharply localized to the stratum corneum by examination of transverse frozen sections of stained skin under the fluorescence microscope. The main problems occurred when it was found that about 50% of volunteers had an irritant reaction to the dye, which excluded them from any study. Contact sensitization occurred in about 15% of volunteers. However, despite these difficulties, stratum corneum turnover times could be estimated from the time taken for complete extinction of fluorescence. Reported times ranged from 6.3 (SD 1.4) days for forehead sites to 20.8 (SD 2.3) days for the back-of-hand sites. Jansen et al.11 introduced the use of another fluorescent dye, dansyl chloride (5-dimethyl-amino-1-naphthalene-sulfonyl chloride). They used a 5% by weight of dansyl chloride in white petrolatum suspension, which was applied on a cotton patch under occlusion for 24 hours. They recommended lifting the patch after 6 to 8 hours, rubbing the site, and refastening the patch. The endpoint for measurement of turnover time was time elapsed until complete disappearance of fluorescence under the Woods lamp. Results confirmed earlier studies12,13 that the forehead showed the shortest turnover time (about 8.5 days), with other body sites showing average turnover times of about 14 days. The authors also checked the binding of the dye to the stratum corneum, concluding that it must bind to the insoluble fibrous proteins that form the central keratin structure. Finlay et al.14 reported a fluorescence photographic photometric technique to measure objectively the intensity of fluorescence of dansyl chloride-stained skin. A fluorescent standard consisting of tiles of uniformly increasing fluorescence arranged in a circular fashion around a central aperture was placed over the area of unknown fluorescence. The site was photographed using a film holder with an integral calibrated step neutral density wedge.15 The optical density of the standard tiles could be measured and plotted against their relative fluorescence on a log scale. A best-fit slope was calculated and the unknown site fluorescence determined by reading off the log percentage fluorescence axis value at the optical density of
364
Handbook of Non-Invasive Methods and the Skin, Second Edition
the unknown area measured with reference to the internal step wedge standard. This method was objective and accurate, and the results were in good agreement with other reports. However, the fluorescent standards were not absolute or widely available (they were a selection of paper of different fluorescent qualities), and the method required considerable specialist photographic experience and equipment. It was not therefore easily accessible for more routine investigations. Marks et al.16 introduced a fluorescence comparator device for the assessment of skin fluorescence due to dansyl chloride staining. The device consisted of a viewing tube with an eyepiece at one end and an open aperture at the other. Half of the lumen of the tube was taken up by a linear varying optical density film strip (a gray wedge). The gray wedge was placed above a fluorescent standard positioned at the open end, so as to be on the skin surface when the tube was placed on the skin. The standard was made from a skin surface biopsy (SSB)17 from a site previously treated with a 5% dansyl chloride in white petrolatum suspension. The gray wedge could be moved by a rack and pinion across the tube and its position accurately measured by reference to an accurate scale. In use, the instrument is placed over the dansyl chloridetreated area of skin with the standard fluorescent SSB adjacent to it, while both are illuminated by a handheld Woods lamp placed between 4 and 8 in. away from the site being examined. Through the simple eyepiece lens the test area and fluorescent standard can be viewed simultaneously as equal halves of a circle divided by a septum in the tube. The gray wedge is then adjusted so as to attenuate the brightness of the standard until it appears to match the test site. The reading of the wedge position is then recorded as the arbitrary unit of fluorescence of the site. These authors checked the method against fluorescence photographic photometry.14 They also measured inter- and intraobserver errors. Coefficients of variation for the intraobserver error and the means of the coefficients of variation for the interobserver error differences were between 4 and 6%. The extinction time of dansyl chloride fluorescence was determined by plotting the measured arbitrary fluorescence units for the dansyl chloridetreated site and an adjacent site (for a measure of background skin fluorescence) against time. Standard estimates of best fit allowed a mean slope to be fitted to the data and extrapolated to the time at which the slope of the treated site intercepted the background site slope. This was taken to be the extinction time. Takahashi et al.,18 using this fluorescence comparator device, demonstrated a circadian rhythm to the rate of decrease of dansyl chloride fluorescence and the effect of
TABLE 42.2 Stratum Corneum Renewal Time in Days Estimated Using the Fluorescence Comparator Technique for Protected and Unprotected Flexor Forearm Sites Mean SC Turnover Time Unprotected site Gauze-protected site Ratio Unprotected site Chamber-protected site Ratio
14.5 23.7 1.75 13.2 29.1 2.21
(SD (SD (SD (SD (SD (SD
3.8) 6.7) 0.19) 1.7) 12.2) 0.77)
Range 8.9–18.8 16.9–31.5 1.50–1.96 11.0–15.0 17.8–45.8 1.59–3.32
TABLE 42.3 Circadian Rhythm in Stratum Corneum Turnover Times as Estimated from Decrease in Dansyl Chloride Fluorescence Measured in Arbitrary Units Using a Comparator Device
Morning (9–10 A.M.) Period: 9–6 P.M. Evening (5–6 P.M.) Period: 9 A.M.–6 P.M.
Comparator Units/Day
Comparator Units/Hour
2.51 (SD 1.37)
0.17
3.00 (SD 1.55)
0.33
protection by a gauze pad and a Finn chamber. A twofold difference was found in the rate of fluorescence loss between evening and morning readings (Table 42.2 and Table 42.3). Takahashi et al.19 later developed an electronic fluorimeter containing sources of ultraviolet light, peaking at 338 nm, a photomultiplier detector, and suitable filters to measure fluorescence of the skin at 446 nm (dansyl chloride fluorescence peak) without effect from the UV light. Readings were presented on a digital display. This device was totally objective and easy to use. Paye et al.20 used the rapid reduction of dansyl chloride fluorescence after application of soaps and detergents as an indicator of the potential of these substances to cause skin irritation. Effendy and colleagues21 used dansyl chloride-determined turnover time to assess the irritant properties of calcipitriol, as compared to sodium lauryl sulfate and its own vehicle base. They found that calcipitriol reduced stratum corneum turnover less than SLS, but more than its vehicle base.
Methods to Determine Desquamation Rate
365
42.3.2 IMAGE ANALYSIS TECHNIQUE
Of further interest was the study of the pattern of fluorescence decay within the 12-mm areas of skin studied. There were clear differences between small areas of skin within the 12-mm stained study sites, with some areas losing fluorescence more rapidly than others. The authors described six patterns of variation. For example, different patterns of fluorescence behavior are caused by staining of lipids or keratotic plugs.
Ratio of fluorescence intensity (%)
Hayashi and colleagues22 have taken the fluorescence staining assessment of turnover rate further, by using image analysis to study the details of fluorescence over small areas of skin. They used standardized 24-hour occlusive dansyl chloride staining on 12-mm-diameter areas of skin on eight healthy Japanese men between 25 and 40 years of age. The sites studied were forehead, cheek, forearm, opsithenar, back, and lower leg. They imaged the UV-induced fluorescence using a sensitive charge coupled device (CCD) video camera with appropriate filtering so that the incident UV spectrum matched the absorption peak of dansyl chloride, and the camera was filtered at 450 nm to separate the fluorescence signal from the illumination wavelengths. By averaging the fluorescence over the 12-mm areas, an overall estimate of stratum corneum turnover time could be measured at each site. They found that fluorescent intensity declined in a linear fashion at the back and lower leg sites, but a faster decline followed by a slower fading was observed on the forehead, cheek, forearm, and opsithenar. Turnover rates were said to be consistent with other reports, although there were some discrepancies between their text and their graphs of turnover times. Taking the data from their graphs, turnover times (Figure 42.1) can be estimated as follows: forehead, 9 days; cheek, 14 days; and other sites, 16 to 18 days.
42.3.3 MICROSCOPIC TECHNIQUES Staining techniques tend to be easier to perform than direct estimates but are probably less sensitive to small and transient changes. Microscopic evaluation of the dansyl chloride staining of individual corneal layers has been reported.23 This involved staining the skin with the standard 5% dansyl chloride preparation under occlusion for 24 to 48 hours, then taking biopsies at various intervals from 0 to 10 days after staining. Examination of transverse sections under the fluorescence microscope allowed identification of stained and new, unstained corneal layers. In this way, the number of newly formed corneal layers was assessed and a rate of formation per 24 hours calculated. It was claimed that an increase in turnover rates greater than about 15 to 32% above normal was detectable using this method. An average rate of formation from a group of 10 normal
100
100
100
80
80
80
60
60
60
40
40
40
20
20
20
0
0
5
10 Day
15
20
0
0
5
(a)
10 Day
15
20
0
(b) 100
80
80
80
60
60
60
40
40
40
20
20
20
5
10 Day (d)
15
20
0
0
5
10 Day (e)
10
15
20
15
20
(c)
100
0
5
Day
100
0
0
15
20
0
0
5
10 Day (f )
FIGURE 42.1 Fluorescence extinction in different body sites: (a) forehead, (b) cheek, (c) opsithenar, (d) forearm, (e) back, (f) lower leg. (From Hayashi, S. et al., Skin Res. Technol., 4, 109–120, 1998. With permission from Blackwell Publishing.)
366
Handbook of Non-Invasive Methods and the Skin, Second Edition
volunteers was found to be 1.15 (SE 0.09) corneal layers per 24 hours. Assuming the average stratum corneum to have 15 to 20 layers would translate to a turnover time of between 13 and 17 days, which is in accord with renewal times reported for the noninvasive dansyl chloride method. Pierard24 reported a method for the assessment of dansyl chloride fluorescence of SSB samples from stained skin using a fluorescence microscope. In this study fluorescence on each test site used was clinically graded using a 0 to 10 analog scale. Then an SSB was taken from that site. Photographs of the fluorescence microscope images were taken. The distribution of stain was reported to be at hair follicle openings, in primary and secondary lines, and on the plateaux. The extent of fluorescent and nonfluorescent areas was quantified using image analysis, and the ratio of these areas was used as a rate of fluorescence extinction. Pierard reports dense and diffuse fluorescence immediately after staining, with labeling persisting in the adnexal openings, the primary and secondary lines, and on large patchy areas of the plateaux, but with areas of focal extinction. Occlusion with some soaps and cosmetics revealed that some were capable of removing the dansyl chloride stain well before the stratum corneum could have been shed. The author concludes that dansyl chloride may not be suitable for testing the effects on stratum corneum turnover of all soaps, and that the reported increased desquamation measured by Ridge and colleagues25 using this test could be due to removal of the dye by the products themselves. Pierard26 reports on the use of fading of pigmentation induced by dihydroxyacetone (DHA) as a measure of stratum corneum renewal time. This product is widely used as a self-tanning agent. It reacts with stratum corneum proteins to form brown substances known as melanoids.27 For use as a turnover marker, 10% DHA solution is applied for 2 hours under occlusion. This is then repeated 6 hours later. After this application the site is left uncovered. The evaluation of the intensity of browning of the skin is assessed using an industrial colorimeter based on the tristimulus method.28 The CIE L* (lightness), a*, and b* (chromaticity) color space parameters are used to measure differences between test sites and adjacent unstained skin.29 The most sensitive and linear measures of fading of skin color are reported to be the change in b* chromaticity coordinate and the change in the overall color parameter (calculated from [ΔL*2 + Δa*2 + Δb*2]–1/2). These parameters are reported to be significant up to 18 days after application. In their attempt to measure the pigmentation effect throughout the stratum corneum, the authors tape-stripped the pigmented and adjacent skin using DSquame® tape and measured the color differences of subsequent pairs of D-Squame corneocyte collections. The progressive decrease in color found probably reflected a decreasing gradient of concentration of melanoids, as well
as a decreasing amount of collected corneocytes from lower and more cohesive stratum corneum layers. The DHA-induced pigmentation fading was compared to visual assessment of extinction of fluorescence from dansyl chloride-treated skin. Although the authors say that a good correlation exists between the two methods, they do not quantify this.
42.4 TAPE AND ADHESIVE METHODS Adhesive-coated tapes, discs, or glass slides can be used to harvest surface squames. The bonding strength of the adhesive to the stratum corneum and to its substrate material will determine the intracorneal cohesive force at which the sample is taken. Controlling these adhesive properties is essential if reproducible results are to be obtained. For example, differences introduced by surface contamination (by dust, grease, etc.), temperature, moisture, or manufacturing variables can have major effects on the binding of the adhesive to the stratum corneum. Bashir and colleagues30 studied the tape-stripping method in some detail, comparing Micropore tape, Transpore tape, and D-Squame adhesive discs. They quantified the mass of stratum corneum removed as a function of number of tape strips and related this to the thickness of stratum corneum removed. They found that the amount of stratum corneum removed per tape strip declined in an exponential manner as more tape strips were taken. This illustrates the change in intracorneal cohesion with depth and shows that this change is not linear, but decreases sharply in the outermost layers of the stratum corneum. They also measured transepidermal water loss (TEWL) and showed that this rises more sharply after about 4 to 5 microns of stratum have been removed. Currently the most common method for assessing desquamation involves the use of D-Squame adhesive discs.31 The methods and uses of D-Squames are presented in greater detail elsewhere in this book, and so only a brief overview will be presented here. D-Squames are small discs of transparent polyester support film coated with a clear pressure-sensitive adhesive. They are available in various sizes and are carefully manufactured to ensure constancy of the adhesive properties. The method consists of careful preparation of the sample site, then application of the D-Squame disc for a set time (5, 10, and 30 seconds have been used) under controlled pressure (10 to 25 kPa or 100 to 250 g/cm2). On removal, some squames remain stuck to the disc. The disc is then placed on a standard background and the amount of adherent squames assessed. The manufacturer of the D-Squame discs supplies a card with photographic standards of varying density of adherent squames for a visual grading of the result. This is the easiest and quickest method, but the results are only on an ordinal scale, so that no attempt is made to measure the actual amount of
Methods to Determine Desquamation Rate
desquamation, or the size distribution of the adherent squames. For more quantitative analysis of the amount of squames, a gravimetric method can be used, with careful weighing of the discs before and immediately after removal from the skin.32 This method, while quite crude, does at least attempt a quantitation of the amount of material on the discs. It can be affected by extraneous materials, such as hairs, and if the reweighing is not carefully controlled, then hydration effects on the adhered squames may introduce some errors. Quantitative assessments can be made using image analysis techniques.33,34 The overall amount of reflection can be measured and taken to represent the amount of adherent scales. More sophisticated methods involve measuring the distribution and sizes of the individual reflecting squames, and relating the reflectivity to an estimate of the thickness of squame. The adhesive disc method has been used to study the effects of diabetes on desquamation. 35 Yoon and colleagues35 used both visual grading and image analysis of D-Squames in diabetic xerosis. They found that fine scaling, representing the background desquamation process, was significantly increased in their diabetic subjects.
42.5 COHESION The binding forces between corneocytes at the skin surface are a function of the rate of desquamation. It must be remembered that it is the release of this intracorneal cohesion (ICC) that allows desquamation to occur. This binding force can be measured using a device known as a cohesograph. This employs a piston of 50 mm2 area that is stuck to the skin surface with cyanoacrylate adhesive,
367
and a means of pulling the piston upward, thus removing a known area of stratum corneum. The force required to distract this segment of stratum corneum from the surface (about two cells thick) is then measured. The first version of this device36,37 used a manual piston retraction, but a motor-driven version followed (Figure 42.2). Changes in cohesion force were measured at increasing depths in the stratum corneum by measuring ICC after tape stripping. The mean force of cohesion of the forearm of a group of 10 normal subjects rose steeply from 100 gmF at the surface to about 200 gmF after 18 strips (about half the stratum corneum depth), when the rate of change slowed considerably.38 There exists a (weak) relationship between the measured rate of forced desquamation and the ICC.39 Also, an increase in ICC from skin from ichthyotic and other scaling disorders, compared to normal, can be measured with cohesography.40 Long and Marks41 used cohesography to investigate senile pruritis. They found an increase in cohesion in aged subjects with senile pruritis compared to a sex- and agematched control population. They concluded that elderly patients with pruritis may have an acquired abnormality of keratinization. Chapman and colleagues42 investigated the roles of extracellular lipids, corneosomes (modified stratum corneum desmosomes), and corneocyte lipid envelopes in corneocyte cohesion on adult pig ears. Cohesion was measured directly by cohesometry. They concluded that corneosomes were the major contributor to the cohesive forces measured. Serizawa et al.43 also used cohesography to try to elucidate the origins of stratum corneum cohesion. They
FIGURE 42.2 Cohesograph: Left panel shows device in use; right panel illustrates 50 mm2 piston and guard ring.
368
Handbook of Non-Invasive Methods and the Skin, Second Edition
compared cohesive forces between palm and upper arm, and then measured the amounts of cholesterol sulfate. They found that differences in cholesterol sulfate levels were unable to account for the observed sevenfold increase in cohesion in the palm site compared to the upper arm.
42.6 CONCLUSION The rate of loss of corneocytes from the skin surface is an important parameter of the population dynamics of the epidermis. Methods to measure desquamation have been devised and include direct counting techniques and the rate of clearance of substantive stratum corneum stains. The latter are the easier to perform, but may be less sensitive than the more direct techniques. Recent advances in quantitative image analysis have allowed semiautomated objective measurement of some aspects of desquamation as assessed via adhesive discs. This technique has become the most widely used in the estimation of parameters related to desquamation and dry skin disorders. No new methods for direct in vivo measurement of the forces involved in stratum corneum cohesion have been developed in the last 25 years, and this is an area of bioengineering that is ripe for review.
REFERENCES 1. Roberts, D., Marks, R., The determination of regional and age variations in the rate of desquamation: a comparison of four techniques, J Invest Dermatol, 74, 13–16, 1980. 2. Williamson, P., Kligman, A.M., A new method for the quantitative investigation of cutaneous microflora, J Invest Dermatol, 45, 498–503, 1965. 3. McGinley, K.J., Marples, R.R., Plewig, G., A method for visualising and quantitating the desquamating portion of the human stratum corneum, J Invest Dermatol, 53(2), 107–111, 1969. 4. Kolmer, J.A., Spaulding, E.H., Robinson, H.W., Approved Laboratory Technique, 5th ed., Appleton-Century-Crofts, New York, 1951. 5. Nicholls, S., Marks, R., Novel techniques for the estimation of intracorneal cohesion in vivo, Br J Dermatol, 96, 595–602, 1977. 6. Corcuff, P., Delasalle, G., Schaefer, H., Quantitative aspects of corneocytes, J Soc Cosmet Chem, 33, 1–17, 1982. 7. Corcuff, P., Chatenay, F., Saint-Leger, D., Hair skin relationships: a new approach to desquamation, Bioeng Skin, 1, 133–139, 1985. 8. Corcuff, P., Chatenay, F., Leveque, J.-L., Desquamation of the stratum corneum: kinetics following U.V. induced injury, Acta Derm Venereol (Stockh) (Suppl), 134, 35–39, 1987. 9. Sutton, R.L., Early epidermal neoplasia, Arch Derm Syph, 37, 738, 1938.
10. Baker, H., Kligman, A.M., Technique for estimating turnover time of human stratum corneum, Arch Dermatol, 95, 408–411, 1967. 11. Jansen, L.H., Hojyo-Tomoko, M.T., Kligman, A.M., Improved fluorescence staining technique for estimating turnover time of the human stratum corneum, Br J Dermatol, 90, 9–12, 1974. 12. Weinstein, G.D., Van Scott, E.J., Autoradiographic analysis of turnover time of normal and psoriatic epidermis, J Invest Dermatol, 45, 257–262, 1965. 13. Rothberg, S., Crounse, R.G., Lee, J.L., Glycine-14 incorporation into the proteins of normal stratum corneum and the abnormal stratum corneum of psoriasis, J Invest Dermatol, 37, 497, 1961. 14. Finlay, A.Y., Marshall, R.J., Marks, R., A fluorescence photographic photometric technique to assess stratum corneum turnover rate and barrier function in vivo, Br J Dermatol, 107, 35–42, 1982. 15. Marshall, R.J., Infrared and ultraviolet photography in a study of the selective absorption of radiation by pigmented lesions of skin, Med Biol Illus, 26, 71, 1976. 16. Marks, R., Black, D., Hamami, I., Caunt, A., Marshall, R.J., A simplified method for measurement of desquamation using dansyl chloride fluorescence, Br J Dermatol, 111, 265–270, 1984. 17. Marks, R., Dawber, R.P.R., Skin surface biopsy: an improved technique for the examination of the horny layer, Br J Dermatol, 84, 117–123, 1971. 18. Takahashi, M., Black, D., Hughes, B., Marks, R., Exploration of a quantitative dansyl chloride technique for measurement of the rate of desquamation, Clin Exp Dermatol, 12, 246–249, 1987. 19. Takahashi, M., Machida, Y., Marks, R., A new apparatus to measure rate of desquamation using dansyl chloride fluorescence, Arch Dermatol Res, 279, 281–282, 1987. 20. Paye, M., Simion, A., Pierard, G.E., Dansyl chloride labelling of stratum corneum: its rapid extraction from skin can predict skin irritation due to surfactants and cleansing products, Contact Derm, 30, 91–96, 1994 21. Effendy, I., Kwangsuksith, C., Chiappe, M., Maibach, H.I., Effects of calcipotriol on stratum corneum barrier function, hydration and cell renewal in humans, Br J Dermatol, 135(4), 545–549, 1996. 22. Hayashi, S., Matsue, K., Takiwaki, H., Image analysis of the distribution of turnover rate in the stratum corneum, Skin Res Technol, 4, 109–120, 1998. 23. Johannesson, A., Hammar, H., Measurement of the horny layer turnover after staining with dansyl chloride: description of a new method, Acta Dermatovener (Stockh), 58, 76–79, 1978. 24. Pierard, G.E., Microscopic evaluation of the dansyl chloride test, Dermatology, 185, 37–40, 1992. 25. Ridge, B.D., Batt, M.D., Palmer, H.E., Jarett, A., The dansyl chloride technique for stratum corneal renewal as an indicator of changes in epidermal mitotic activity following topical treatment, Br J Dermatol, 118, 167–174, 1988. 26. Pierard, G.E., Pierard-Franchimont, C., Dihydroxyacetone test as a substitute for the dansyl chloride test, Dermatology, 186, 133–137, 1993.
Methods to Determine Desquamation Rate
27. Maibach, H.I., Kligman, A.M., Dihydroxyacetone. A suntan-simulating agent, Arch Dermatol, 82, 505–507, 1960. 28. el-Gammal, S., Hoffmann, K., Steiert, P., Gassmuller, J., Dirschka, T, Altmeyer, P., Objective assessment of intra- and inter-individual skin colour variability: an analysis of human skin reaction to sun and UVB, in The Environmental Threat to the Skin, Marks, R., Plewig, G., Eds., Martin Dunitz, London, 1992, pp. 99–115. 29. Robertson, A.R., The CIE color difference formulas, Col Res Appl, 2, 7–11, 1977. 30. Bashir, S.J., Chew, A., Anigbogu, A., Dreher, F., Maibach, H.I., Physical and physiological effects of stratum corneum tape stripping, Skin Res Technol, 7, 40–48, 2001. 31. Miller, D.L., D-Squame® adhesive discs, in Bioengineering of the Skin: Skin Surface Imaging and Analysis, Wilhelm, K.P., Elsner, P., Berardesca, E., Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1997, pp. 39–46. 32. Leveque, J.-L., Corcuff, P., deRigal, J., et al., In vivo studies of the evolution of physical properties of the human skin with age, Int J Dermatol, 23, 322–329, 1984. 33. El-Gamal, C., Pagnoni, A., Kligman A.M., El-Gammal, S., A model to assess the efficacy of moisturisers: the quantitation of soap-induced xerosis by image analysis of adhesive discs (D-Squame®s), Clin Exp Dermatol, 21, 338–343, 1996. 34. Wilhelm, K.-P., Kaspar, K., Schumann, F., Articus K., Development and validation of a semiautomatic image analysis system for measuring skin desquamation with D-Squame®s, Skin Res Technol, 8, 98–105, 2002. 35. Yoon, H.S., Baik, S.H., Oh, C.H., Quantitative measurement of desquamation and skin elasticity in diabetic patients, Skin Res Technol, 8, 250–254, 2002.
369
36. Nicholls, S., Marks, R., Novel techniques for the estimation of intracorneal cohesion in vivo, Br J Dermatol, 96, 595–602, 1977. 37. Marks, R., Nicholls, S., Fitzgeorge, D., Measurement of intracorneal cohesion in man using in vivo techniques, J Invest Dermatol, 69, 299–302, 1977. 38. Marks, R., Quantification of desquamation and stratum corneum cohesivity, in Cutaneous Investigation in Health and Disease, Leveque, J.-L., Ed., Marcel Dekker, New York, 1989, pp. 33–47. 39. King, C.S., Nicholls, S., Marks, R., Relationship of intracorneal cohesion to rates of desquamation in the scaling disorders, in Bioengineering and the Skin, Marks, R., Payne, P.A., Eds., MTP Press, Lancaster, 1981, pp. 237–244. 40. Nicholls, S., King, C.S., Marks, R., Is there a relationship between corneocyte size and stratum corneum function in vivo?, in Bioengineering and the Skin, Marks, R., Payne, P.A., Eds., MTP Press, Lancaster, 1981, pp. 227–235. 41. Long, C.C., Marks, R., Stratum corneum changes in patients with senile pruritis, J Am Acad Dermatol, 27, 560–564, 1992. 42. Chapman, S.J., Walsh, A., Jackson, S.M., Friedmann, P.S., Lipids, proteins and corneocyte adhesion, Arch Dermatol Res, 283, 167–173, 1991. 43. Serizawa, S., Osawa, K., Togashi, K., Yamamoto, A., Ito, M., Hamanaka, S., Otsuka, F., Relationship between cholesterol sulfate and intercellular cohesion of the stratum corneum: demonstration using a push-pull meter and an improved high-performance thin-layer chromatographic separation system of all major stratum corneum lipids, J Invest Dermatol, 99(2), 232–236, 1992.
of Adhesive Techniques to 43 Application Harvest Stratum Corneum Material David L. Miller Bionet Incorporated, Dallas, Texas
CONTENTS 43.1 43.2 43.3 43.4 43.5 43.6
Introduction............................................................................................................................................................371 Object.....................................................................................................................................................................371 Methodological Principle ......................................................................................................................................371 Sources of Error.....................................................................................................................................................372 Correlation with Other Methods ...........................................................................................................................372 Recommendations..................................................................................................................................................372 43.6.1 Sticky Slides ..............................................................................................................................................372 46.6.2 Adhesive Tape............................................................................................................................................372 43.6.3 D-SQUAME Discs ....................................................................................................................................372 43.7 D-SQUAME Sampling Disc Applications ............................................................................................................373 References .......................................................................................................................................................................373
43.1 INTRODUCTION
43.3 METHODOLOGICAL PRINCIPLE
The stratum corneum is both a barrier to the physical and chemical insults of the external environment and a record reflecting disease of the epidermis. In many cases examination of the skin surface, particularly the nature of the stratum corneum, is more precisely accomplished by first separating a portion of the stratum corneum from the underlying tissue, then applying certain test modalities. Examples include techniques where the light absorbing/scattering properties of the surface are to be measured1–3 (to estimate the scaliness), exfoliative cytology to elucidate certain features of the stratum corneum as a component of the diagnostic process,4–7 and analytical methods for determining endogenous or exogenous chemical components of the barrier.8,9 This chapter deals with the means of effecting convenient and reproducible samples of the desquamating stratum corneum.
Pressure-sensitive adhesives are high-molecular-weight organic substances that flow under applied pressure so as to form an intimate mechanical bond with a substrate, in this case the stratum corneum surface. As long as the adhesion of this interface, the internal cohesion of the adhesive material, and the adhesion to its support surface are all higher than the internal cohesiveness of the stratum corneum, some portion of the stratum corneum will be separated when the support surface is peeled away from the skin. The use of pressure-sensitive adhesive tape (i.e., support surface = a flexible plastic film or fabric) to accomplish a sampling of the superficial stratum corneum for dermatolgical study probably originated with the publications by Wolf10 beginning in 1939. Depending on the particular kind of adhesive and support employed, rather complex steps had to be taken to prepare samples for microscopy.7 Many adhesives and their support materials are strongly stained by biological dyes. Goldschmidt and Kligman11 coated an adhesive directly onto ordinary glass slides, making a very rigid support. This eliminated the necessity of transferring the sample to a light-transmitting surface for observation, but required selection of an adhesive that did not interfere with the ultimate procedure. The very stiff nature of the support restricts sampling to flat areas. In any case, the preparation and storage of
43.2 OBJECT The object of these harvesting methods is to easily and reproducibly obtain a sample of the stratum corneum. To some extent, the ultimate fate of the sample places different requirements on the sampling technique.
371
372
Handbook of Non-Invasive Methods and the Skin, Second Edition
consistent coatings free of contamination from the air is an art in itself. Also, the nondrying adhesive film gradually loses tackiness due to oxidation. For the past 20 years, pressure-sensitive adhesive discs with a moderately flexible plastic support specifically designed for dermatological use have become commercially available under the trade name D-SQUAME®* discs.12 The discs are made from a very clear grade of polyester support film and an aggressive, super clear adhesive. The combination of film and adhesive results in very high contrast between the optical properties of the adhering stratum corneum and the sampling medium. The clear carrier sheet on which the discs are supplied forms an effective barrier to dust and oxidation.
43.4 SOURCES OF ERROR Contamination of the adhesive surface by dust or dirt prior to use can be a problem with sticky slides and adhesive tape. The design of the D-SQUAME skin sampling system tends to reduce this problem. Insufficient or inconsistent application pressure will result in samples that are difficult to compare. This error can be reduced by using a calibrated application pressure2,3,13 and can be controlled by adopting a consistent application technique.1,7 Preparation of the skin surface is a critical step. While a clean, dry surface will provide maximal adhesion, samples taken for certain applications will be rendered useless by a cleaning step.
43.5 CORRELATION WITH OTHER METHODS There are only two other common methods of sampling the skin surface: scraping with a blade or other tool6 and the use of quick-setting cements such as the cyano-acrylate type.14 Scraping the skin is not very useful in preparing a sample with any quantitative value, as it is difficult to both control the extent of the area scraped and prevent loss of scraped material by air currents. Cyanoacrylate cements have been used for preparing so-called skin surface biopsies.14 Because the adhesion mechanism is based on a chemical reaction, the depth of skin removed will be determined by the depth of penetration of the adhesive before it hardens. Generally, more stratum corneum is removed by this method than by the pressure-sensitive adhesives. This type of sampling technique is described in detail elsewhere in this handbook. * Registered trademark of CuDerm Corporation, Dallas, TX.
43.6 RECOMMENDATIONS The basic sampling technique is essentially the same regardless of the material used. The adhesive surface is pressed against the selected skin site for a few seconds, during which time a sufficient pressure is applied to ensure contact and flow of the adhesive substance. Then the adhesive support surface is peeled away from the skin, and the sample set aside for further processing. The three approaches are discussed below.
43.6.1 STICKY SLIDES Preparation of the slides in the laboratory requires obtaining a source of adhesive solution, then coating the solution on to glass slides and allowing the organic solvent to evaporate.11 Producing an even, consistent coating will require skill and practice. Care must be taken in storing and handling the prepared slides to prevent contamination of the surface. The useful life of the slides after preparation is limited due to gradual air oxidation of the adhesive surface. Adhesive and precoated slides are available commercially.15
43.6.2 ADHESIVE TAPE There are a number of commercially produced adhesive tapes that have been used for skin sampling,6,7,13,16 but they are not necessarily well characterized with respect to component properties. In quantitative applications requiring a fixed sampling area,8,9 tape should be precut under clean conditions.
43.6.3 D-SQUAME DISCS This skin sampling system eliminates many of the difficulties related to sticky slides and adhesive tape. It is a specially formulated adhesive system readily available worldwide directly from the manufacturer.12 At the time of this writing, the discs were available in two sizes: the standard 22-mm-diameter disc fits on a ordinary glass slide, and the 13-mm-diameter disc allows sampling from small areas such as the area covered by an experimental treatment patch like the large Finn chamber or hilltop chamber, without including untreated skin. Both sizes are produced on a clear carrier sheet, making storage and use of the discs easy. Figure 43.1 illustrates the appearance of skin samples made with the discs. The samples were obtained from the lower leg, then immediately placed on black background cards for viewing. The view is through the clear adhesive disc, looking at the top surface of the skin. Note that differing scale thickness is represented in the photographs by differing degrees of scattered light intensity.
Application of Adhesive Techniques to Harvest Stratum Corneum Material
(a)
(b)
373
(c)
FIGURE 43.1 Photographic comparison of D-SQUAME skin surface samples representing three levels of scaliness: (a) fine scales, (b) medium scales, and (c) coarse scales; 22-mm diameter discs against a black background viewed by scattered light.
43.7 D-SQUAME SAMPLING DISC APPLICATIONS The simplicity of use and uniformity of the D-SQUAME sampling disc has enabled many interesting skin research applications. The discs have been used in investigations involving variations in activity of epidermal antioxidant enzymes,17 exploring the route of delivery to the skin surface of endogenous substances,18 and assessing trace exogenous compounds remaining on the skin.19 Others have employed the discs in studying regulation of biochemical processes in the stratum corneum20,21 and the distribution of lipids in the stratum corneum.22 Validation studies of analytical techniques applied to the measurement of scaliness using the D-SQUAME discs have been completed,23,24 and aspects of the structure and biological dynamics of the stratum corneum (XLRS) accessable via the D-SQUAME stripping sample (SACD) have been reviewed.25 Detailed investigations of the application of the D-SQUAME sampling disc in the efficacy screening of hydration products,26 quantification of soap-induced xerosis,27 and hydration state of the stratum corneum28 have been reported. The treatment effect of shampoos on dandruff has been studied using this method.29 The D-SQUAME method has proved to be a versatile sampling technique and deserves serious consideration by scientists interested in studying the scaling stratum corneum, its diseases, conditions, composition, and biochemistry.
REFERENCES 1. Prall, J.K. The scaliness of human skin. Arch. Biochem. Cosmetol. 9, 87, 1966.
2. Piérard, G.E., Piérard-Franchimont, C., Saint Léger, D., and Kligman, A.M. Squamometry: the assessment of xerosis by cyanoacrylate surface biopsies and colorimetry of D-SQUAME adhesive discs. J. Soc. Cosmet. Chem. 47, 297, 1992. 3. Serup, J., Winther, A., and Blichman, C. A simple method for the study of scale pattern and effect of a moisturizer. Qualitative and quantitative evaluation by D-SQUAME tape in comparison with parameters of epidermal hydration. Clin. Exp. Dermatol. 14, 277 1989. 4. Piérard-Franchimont, C. and Piérard, G.E. Skin surface stripping in diagnosing and monitoring inflammatory, xerotic, and neoplastic diseases. Pediatr. Dermatol. 2, 180, 1985. 5. Piérard-Franchimont, C. and Piérard, G.E. Assessment of aging and actinic changes by cyanoacrylate skin surface strippings. Am. J. Dermatopathol. 9, 500, 1987. 6. Barr, R.J. Cutaneous cytology. J. Am. Acad. Dermatol. 10, 163, 1984. 7. Jenkins, H.L. and Tresise, J.A. An Adhesive-tape stripping technique for epidermal histology. J. Soc. Cosmet. Chem. 1, 1969. 8. Pershing, L.K. New approaches to assess topical corticosteroid bioequivalence: pharmacokinetic evaluation. Int. J. Dermatol. 38 (Suppl. 1), 14, 1992. 9. Rougier, A., Lotte, C., and Dupuis, D. An original predictive method for in vivo percutaneous absorption studies. J. Soc. Cosmet. Chem. 38, 397, 1987. 10. Wolf, J. Das innere Struktur der Zellen des Stratum desquamans der menschlichen Epidermis. Z. Mikr. Anat. Forsch. 46, 170, 1939. 11. Goldschmidt, H. and Kligman, A.M. Exfoliative cytology of human horny layer. Arch. Dermatol. 96, 572, 1967. 12. CuDerm Corporation, Box 797686, Dallas, TX 753797686. Website: www.cuderm.com. 13. Klaschka, F. and Norenberg, M. Individual transparency patterns of adhesive-tape strip series of the stratum corneum. Int. J. Dermatol. 16, 836, 1977.
374
Handbook of Non-Invasive Methods and the Skin, Second Edition
14. Marks, R. and Dawber, R.P.R. Skin surface biopsy, an improved technique for the examination of the horny layer. Br. J. Dermatol. 84, 117, 1971. 15. Durotak adhesive and Durotak adhesive slides from Delasco, 608 13th Ave., Council Bluffs, IA 51501-6401. Website: www.delasco.com. 16. Prall, J.K., Theiler, R.F., Bowser, P.A., and Walsh, M. The effectiveness of cosmetic products in alleviating a range of dryness conditions as determined by clinical and instrumental techniques. Int. J. Cosmet. Sci. 8, 159, 1986. 17. Hellemans, L., Corstjens, H., Neven, A., Declercq, L., and Maes, D. Antioxidant enzyme activity in human stratum corneum shows seasonal variation with an agedependent recovery. J. Invest. Dermatol., 120, 434, 2003. 18. Thiele, J., Weber, S., and Packer, L. Sebaceous gland secretion is a major physiologic route of vitamin E. Delivery to skin. J. Invest. Dermatol. 113, 1006, 1999. 19. Kawakami, S., Callicott, R.H., and Zhang, N. Trace analysis of benzalkonium chloride on skin by flow injection ionspray mass spectrometry-mass spectrometry. Analyst 123, 489, 1998. 20. Fluhr, J.W., Kao, J., Jain, M., Ahn, S.K., Feingold, K.R., and Elias, P.M. Generation of Free fatty acids from phospholipids regulates stratum corneum acidification and integrity. J. Invest. Dermatol. 117, 44, 2001. 21. Hachem, J.-P., Crumrine, D., Fluhr, J.W., Brown, B.E., Feingol, K.R., and Elias, P.M. pH directly regulates epidermal permeability barrier homeostasis, and stratum corneum integrity/cohesion. J. Invest. Dermatol. 121, 345, 2003.
22. Weerheim, A. and Ponec, M. Determination of stratum corneum lipid profile by tape stripping in combination with high-performance thin-layer chromatography. Arch. Dermatol. Res. 293, 191, 2001. 23. Lagard, J.M., Black, D., Gall, Y., and Del Pozo, A. Image analysis of scaly skin using D-SQUAME® samplers: technical and physiological validation. Int. J. Cosmet. Sci. 22, 53, 2000. 24. Wilhelm, K.-P., Kaspar, K., Schumann, F., and Articus, K. Development and validation of a semi-automatic image analysis system for measuring skin desquamation with D-SQUAMES®. Skin Res. Technol. 8, 98, 2002. 25. Piérard-Franchimont, C., Henry, F., and Piérard, G.E. The SACD method and the XLRS squamometry tests revisited. Int. J. Cosmet. Sci. 22, 437, 2000. 26. De Paepe, K., Janssens, K., Hachem, J.-P., Roseeuw, D., and Rogiers, V. Squamometry as a screening method for the evaluation of hydrating products. Skin Res. Technol. 7, 184, 2001. 27. El Gammal, C., Pagnoni, A., Kligman, A.M., el Gammal, S. A model to assess the efficacy of moisturizers: the quantification of soap-induced xerosis by image analysis of adhesive-coated discs (D-Squames). Clin. Exp. Dermatol. 21, 338, 1996. 28. Gasser, P., Peno-Mazzarino, L., Lati, E., and Djian, B. Original semiologic standardized evaluation of stratum corneum hydration by Diagnoskin® stripping sample. Int. J. Cosmet. Sci. 26, 1, 2004. 29. Piérard-Franchimont, C., Uhoda, E., Loussouarn, G., Saint-Léger, D., and Piérard, G.E. Effect of residence time on the efficacy of antidandruff shampoos. Int. J. Cosmet. Sci. 25, 267, 2003.
Skin and Scaling Evaluated by 44 Dry D-Squames and Image Analysis Harald Schatz and Peter J. Altmeyer Dermatologische Klinik der Ruhr, Universität Bochum, Bochum, Germany
Albert M. Kligman Department of Dermatology, University of Pennsylvania, Philadelphia, Pennsylvania
CONTENTS 44.1 Introduction............................................................................................................................................................375 44.1.1 Object.........................................................................................................................................................375 44.1.2 Quantification of Dry Skin ........................................................................................................................375 44.2 Methodological Principle ......................................................................................................................................376 44.2.1 D-Squame Image Analysis ........................................................................................................................376 44.2.1.1 Sampling Desquamating Horny Cells on Adhesive Discs ........................................................376 44.2.1.2 Obtaining the Video Image ........................................................................................................376 44.2.1.3 Illumination ................................................................................................................................377 44.2.1.4 Image Analysis Procedures ........................................................................................................377 44.2.1.5 Assessment of Leg Dryness by the D-Squame Method ...........................................................377 44.2.2 Correlation with Other Methods ...............................................................................................................377 44.2.3 Objectiveness and Reproducibility............................................................................................................378 44.3 Recommendations..................................................................................................................................................378 References .......................................................................................................................................................................378
44.1 INTRODUCTION 44.1.1 OBJECT It is a well-known fact that the sampling and visual examination of tape strippings of the stratum corneum can reveal differences in the extent of dryness not noticeable by inspection of the skin surface. Following this simple method, a new image analysis technique was developed to objectively analyze the desquamating portion of the horny layer. Skin surface sampling discs (D-Squames) were employed to sample loose cells and scales from the superficial stratum corneum. Placing the discs against the black background of a storage card provides the maximum of contrast while evaluating the desquamation patterns. The discs are then illuminated in a light box and viewed by a charge coupled device (CCD) video camera attached to a stereomicroscope. The video image of the sample is captured by a frame grabber in a personal computer and then processed with the aid of an image analysis program. Within seconds, the computer generates a few numbers,
which represent the quantitative and qualitative properties of the sample. One of the values, introduced as desquamation index, is a calculation of the disc area occupied with corneocytes and the individual thickness of the scales. The desquamation index provides exact, fast, and reproducible characterization of the D-Squames and is especially valuable for the assessment of dry skin.
44.1.2 QUANTIFICATION
OF
DRY SKIN
The epidermis, the outer layer of the skin, is responsible for a wide range of properties that help to maintain the integrity of the body. It absorbs and excretes liquids, acts as a barrier against microorganisms, and provides cosmetic functions. Epidermal lipids and the sebum help to condition the stratum corneum and regulate the turnover of the horny layer. The shedding or accumulation of stratum corneum in visible flakes is called desquamation. Under normal circumstances, the epidermis is completely replaced every 25 to 30 days, depending on the skin site. 375
376
Handbook of Non-Invasive Methods and the Skin, Second Edition
At the end of the process of keratinization, corneocytes get piled up to the stratum corneum. Abnormal shedding, produced by underlying conditions like parakeratosis (psoriasis) or loss of lipids and hydration (xerosis), leads to squamous eruptions. Quantifying the rate of desquamation is one way to determine pathological changes in the keratinization of the epidermis. It has been used in the assessment of skin care products, to quantify xerosis, and to measure the degree of hydration of the stratum corneum.1–4 Considering the high incidence of xerosis in the population and the growing interest in skin care, it was necessary to characterize and quantify dryness for clinical and commercial studies. Even though most efforts have been made in the field of cosmetic dermatology, we realize more and more that the evaluation of the stratum corneum by means of noninvasive techniques becomes an important issue for clinical dermatology. In combination with other skin physiology techniques, like transepidermal water loss measurements, lipid assessments, pH, and conductance measurements, the quantification of desquamation is important to characterize the condition of skin in health and disease. There are three possible approaches to characterize the process of desquamation: the assessment of the stratum corneum turnover, the measurement of intracorneal cohesion, and the quantification of scaling. Measuring the stratum corneum turnover can be achieved by application of dansyl chloride fluorescence dye on the skin.5 The substance penetrates the entire thickness of the stratum corneum and can be detected easily under a Woods lamp. Consecutive observations under ultraviolet radiation up to the time of extinction of the fluorescence allow determination of the duration of one full turnover measured in days. A number of methods and modifications have been developed to measure the degree of fluorescence.6 The measurement of stratum corneum cohesivity is much more complicated and requires a special apparatus to record the force that is required to rupture the stratum corneum after fixing a cyanoacrylate stub to the skin.7 However, the method is very complicated and susceptible to environmental factors. The most direct way to evaluate dryness is to look at the scale pattern itself. Since visual grading of scaliness is liable to environmental changes and subjectiveness, collecting corneocytes is a suitable approach toward an objective assessment of dry skin. Efforts were made to collect corneocytes by a washing technique. In a glass chamber placed on the skin, filled with water or a detergent (0.1% Triton X-100), loose scales and horny cells were sampled from the skin surface. The yield could be enhanced by using a spatula to scrub the skin manually.8 The material was then ready for quantitative measurements (weighting, cell counter) or biochemical assessments (lipid extractions).
The tape-stripping technique was first described over 50 years ago.9 Readily available adhesive tape is used to strip the surface of the skin. After placement on a black background, an experienced observer can grade the extent of scaliness on the tape. However, this elegant technique has never come into fashion. Later, adhesive-coated glass slides (sticky slides) were utilized to obtain a sample of the loose desquamating portion of the outer horny layer.10,11 When appropriately stained, scales, which are really clumps of corneocytes, develop an intense color. Much depends on the nature of the adhesive and the thickness of the coating on the slide. The sticky slides themselves were highly variable, depending on the thickness of the adhesive, its source, and the uniformity of the adhesive coating. A further step toward objectiveness was made in 1989 by Serup et al.,12 who described a method to measure the amount of scales on adhesive tapes by the attenuation of transmitted light, using a projector and a light meter. To utilize the full capacity of desquamation sampling, we developed a rigidly controlled video and image analysis technique to evaluate the amount and structure of desquamating material on adhesive tapes.
44.2 METHODOLOGICAL PRINCIPLE 44.2.1 D-SQUAME IMAGE ANALYSIS 44.2.1.1 Sampling Desquamating Horny Cells on Adhesive Discs The sampling device is a 22-mm crystal-clear adhesivecoated disc (D-Squame, CuDerm Corporation, Dallas, TX) with a homogeneous adhesive layer. The medicalgrade adhesive safely removes superficial corneocytes and provides optimum visibility of adhering skin cells. After peeling of the protective seal, the D-Squame is pressed to the skin surface. The degree of pressure can be precisely controlled, using the device described by Serup et al.12 However, with experience, one can obtain a reliable sample manually with firm finger pressure. The disc is then peeled from the skin with tweezers and placed on a black storage card included in the D-Squame kit. 44.2.1.2 Obtaining the Video Image An image analysis system consists of four components: the image source (live through video camera), a video screen that displays the image, a video digitizing board called a frame grabber, and a computer with monitor to run the software. In our setup, the video images of the samples were taken by a high-resolution black-and-white CCD video camera (Dage-MTI CCD72, Michigan City, IN) connected to a stereomicroscope (ZEISS, West Germany). A separate video control panel with manual gain
Dry Skin and Scaling Evaluated by D-Squames and Image Analysis
and black level controls guarantees constant video processing under identical conditions. The image is then captured by an image analysis program. We use the Java (Jandel Scientific, CA) in combination with the frame grabber board (Truevision Targa-M8 Frame Grabber), both installed in a Unisys personal computer. However, any high-quality image analysis system available on the market can be used to grab the image. Through a sampling process, the frame grabber translates the image into 512 × 480 picture elements (pixels). Each pixel will be given a numeric value according to its intensity on a gray-level scale from 0 to 255. 44.2.1.3 Illumination A crucial factor in video image analysis is a perfectly discriminating, homogeneous, and constant illumination. The best results are achieved using a reflecting white light box illuminated from two sides by means of two fiberoptic light carriers. By scattering the light through white translucent glass, an enhanced contrast of the corneocyte clusters on the discs can be accomplished. Due to their loose structure, the scales disperse the light and appear as shiny objects against the black background. The degree of brightness is proportional to the thickness of the scales. Before beginning the measurements, the illumination has to be calibrated using a reference gray card or by employing an automatic light meter, which controls the power supply of the lamps by a feedback mechanism. 44.2.1.4 Image Analysis Procedures The program applies a mask to the image, to define a measurement area of 200 mm2. No contrast enhancement is necessary. The next step is to apply a lookup table to the image. This substitutes a new set of numeric values for the default gray scale so that ranges of gray levels are represented by single values. Each pixel is assigned to one of five arbitrary thickness levels of the corneocyte clusters. Transformations can be used to calculate the number of pixels in each thickness group as a percent value, as well as the total area occupied with cells. We also determine the percentage area occupied by corneocytes. These two functions are integrated to yield the desquamation index according to the following formula: 5
2A + D.I. =
∑ T * ( n − 1) n
n =1
6
(44.1)
D.I. is the desquamation index, A the percent area covered by corneocytes, Tn the percentage of corneocytes in relation to thickness, and n the thickness level (1 to 5). The macrofacility of the program is used to record the
377
(a)
(b)
(c)
FIGURE 44.1 (a–c) Ordinary video camera illustrations of samples obtained from (a) nondry, (b) moderately dry, and (c) severely dry skin from the leg.
sequences of all measurement functions and transformations and to perform them automatically with one keystroke. 44.2.1.5 Assessment of Leg Dryness by the DSquame Method Leg samples from nondry, moderately dry, and severely dry skin were captured by the video camera (Figure 44.1a to c) and then processed by the image analysis (Figure 44.2a to c). Figure 44.1 and Figure 44.2 illustrate the findings for these three levels of xerosis. The five thickness levels are reflected by five different colors. The computer outputs on these samples are presented in Figure 44.2. The differences are very striking. In nondry skin the area occupied by scales is only 22%, in contrast to 97% for severely dry skin. Likewise, the thickness levels are very different, resulting in desquamation indices that for nondry, moderately dry, and severely dry skin are 7.6, 28.3, and 70.0, respectively.
44.2.2 CORRELATION
WITH
OTHER METHODS
The assessment of dry skin (xerosis), whether for purposes of classification, diagnosis, or therapeutic evaluations, has been notoriously handicapped by the traditional grading systems based on visual and tactile scoring. Such systems are highly subjective and suffer from unacceptable variability by inconsistencies from grader to grader and also from poor reproducibility. Environmental changes jeopardize such subjective grading systems since hydration swells the outer horny layer and leads to a camouflage of scaling and dryness. D-Squames are certainly a technical advance in these respects, standardizing the collection of scales from the outermost portion of the horny layer, actually the presumptive desquamating layer, where the loosened horny cells are ready to be shed. To demonstrate the usefulness of the D-Squame disc, we developed procedures based on computerized image analysis that enable quantification of the degree of scaling,
378
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 44.2 (a–c) Same samples as Figure 44.1, however, after image analysis with five scale thickness levels shown in five different colors (here presented in black and white). Note that the contrast and grading have been improved in comparison with Figure 44.1.
i.e., the level of dryness. We measured both the percentage of area covered by corneocytes and the distribution of scales according to five levels of thickness. From these one can calculate the desquamation index, which expresses the degree of xerosis in one integrated value.
44.2.3 OBJECTIVENESS
AND
REPRODUCIBILITY
With the computer analysis of adhesive tapes, we can add an objective method without the possible diversions in human judgment. According to their subjective nature, visual grading can only produce scores in the form of ordinal values, with the restriction that the differences between the values are not consistent, and therefore not meaningful. The data cannot be analyzed by parametric statistical tests, and the use of mean scores is not permitted. The image analysis provides ratio measurements with consistent differences between the values. The data can be analyzed with parametric statistical tests like t-tests or analysis of variance (ANOVA) and can easily be summarized in mean values. The desquamation index quantitatively and qualitatively characterizes the amount and pattern of desquamation with a high reproducibility.
44.3 RECOMMENDATIONS We have developed a procedure that enhances the appearance of scaling by delipidization of the skin surface. Equal parts of ether:acetone are applied in a small glass cup, 2 cm in diameter, for 1 min. Even when scaling is not
clinically apparent, the site becomes bright white, in proportion to the level of occult scaling. This enhancement is particularly notable on the face where sebum tends to mask scaling. We have demonstrated that the desquamation index increases when D-Squames are applied to the delipidized surface. Stereophotographs of delipidized sites are very useful for clinical grading. Finally, one can add another dimension to the DSquame analysis by removing the scales from the tape, and after sonication to obtain a unicellular dispersion, determining the number of corneocytes per mm2. We have found a good correlation between the desquamation index and the corneocyte count.
REFERENCES 1. Boisits, E.K., Nole, G.E., and Cheney, M.C., The refined regression method, J. Cut. Aging Cosmet. Dermatol., 1, 155, 1989. 2. Kligman, A.M., Regression method for assessing the efficacy of moisturizers, Cosmet. Toilet., 93, 27, 1978. 3. Kligman, A.M., Lavker, R.M., Grove, G.L., and Stoudemeyer, T., Some aspects of dry skin and its treatment, in Safety and Efficacy of Topical Drugs and Cosmetics, Kligman, A.M., Leyden, J., Marks, R., Black, D., Hamami, I., Count, A., and Marshall, R.J., Br. J. Dermatol., 111, 265, 1984. 4. Marks, R., Quantification of desquamation and stratum corneum cohesivity, in Cutaneous Investigation in Health and Disease, Leveque, J.-L., Ed., Marcel Dekker, New York, 1989, p. 33.
Dry Skin and Scaling Evaluated by D-Squames and Image Analysis
5. Jansen, L.H., Hoiyo-Tomoko, M.T., and Kligman, A.M., Improved fluorescent staining technique for estimating turnover of the human stratum corneum, Br. J. Dermatol., 90, 9, 1974. 6. Finlay, A.Y., Marshall, R.J., and Marks, R., Br. J. Dermatol., 107, 35, 1982. 7. Nicholls, S. and Marks, R., Br. J. Dermatol., 96, 595, 1977. 8. McGinley, K.J., Marples, R.R., and Plewig, G., J. Invest. Dermatol., 53, 107, 1969. 9. Wolf, J., Die innere Struktur der Zellen des Stratum Desquamans der menschlichen Epidermis, Z. Mikrosk. Anat. Forsch., 46, 170, 1936.
379
10. Goldschmidt, H. and Kligman, A.M., Exfoliative cytology of human horny layer, Arch. Dermatol., 96, 572, 1967. 11. Grove, G.L., Exfoliative cytological procedures as a nonintrusive method for dermatogerontological studies, J. Invest. Dermatol., 73, 67, 1979. 12. Serup, J., Winther, A., and Blichmann, C., A simple method for the study of scale pattern and effect of a moisturizer: qualitative and quantitative evaluation by D-Squame tape compared with parameters of epidermal hydration, Clin. Exp. Dermatol., 14, 277, 1989.
Barrier Functions and Gradients
of Transepidermal 45 Measurement Water Loss by Semiopen Systems R.A. Tupker Department of Dermatology, St. Antonius Hospital, Nieuwegein, The Netherlands
J. Pinnagoda Ministry of Public Health, Singapore
CONTENTS 45.1 Introduction............................................................................................................................................................384 45.2 Object.....................................................................................................................................................................384 45.2.1 Evaluation of Barrier Function in Clinical Conditions.............................................................................384 45.2.2 Sweat Gland Activity in Clinical Conditions............................................................................................384 45.2.3 Exposure to Irritants, Predictive Irritancy Testing ....................................................................................384 45.2.4 Composition of the Epidermal Barrier......................................................................................................385 45.2.5 Evaluation of Effects of Moisturizers and Protective Creams .................................................................385 45.3 Methodological Principle ......................................................................................................................................385 45.3.1 The Theory.................................................................................................................................................385 45.3.2 The Method................................................................................................................................................386 45.4 Sources of Error.....................................................................................................................................................387 45.4.1 Validity of Method and Associated Variables ...........................................................................................387 45.4.2 Instrument-Related Variables.....................................................................................................................387 45.4.2.1 Start-Up and Use .......................................................................................................................387 45.4.2.2 Zeroing.......................................................................................................................................387 45.4.2.3 Measuring ..................................................................................................................................387 45.4.2.4 Zero Drift...................................................................................................................................388 45.4.2.5 Humidity Changes.....................................................................................................................388 45.4.2.6 Temperature Changes ................................................................................................................388 45.4.2.7 The Surface Plane .....................................................................................................................388 45.4.2.8 Contact Pressure ........................................................................................................................388 45.4.2.9 Use of the Probe Protection Covers..........................................................................................389 45.4.2.10 Intra- and Interinstrumental Variability.....................................................................................389 45.4.2.11 Calibration .................................................................................................................................389 45.4.2.12 Accuracy ....................................................................................................................................389 45.4.2.13 Performance of the Different Types of Instruments.................................................................389 45.4.3 Environment-Related Variables .................................................................................................................390 45.4.3.1 Air Convections.........................................................................................................................390 45.4.3.2 Ambient Air Temperature .........................................................................................................390 45.4.3.3 Ambient Air Humidity ..............................................................................................................390 45.4.3.4 Seasonal Variation .....................................................................................................................390 45.4.3.5 Direct Light ...............................................................................................................................390 45.4.4 Individual-Related Variables......................................................................................................................390 45.4.4.1 Sweating ....................................................................................................................................391 45.4.4.2 Skin Surface Temperature .........................................................................................................391 45.5 Conclusion .............................................................................................................................................................391 References ......................................................................................................................................................................391 383
384
Handbook of Non-Invasive Methods and the Skin, Second Edition
45.1 INTRODUCTION Measurement of transepidermal water loss (TEWL) is used in many research centers for studying the water barrier function of the skin. Different methods for TEWL measurement from local skin sites have been described.1 Unventilated-chamber (closed-chamber) methods are not capable of continuous measurement, as they tend to occlude the skin. Ventilated-chamber methods, using dry or moistened carrier gas, are capable of measuring TEWL continuously. However, both methods interfere with the microclimate overlying the surface of the skin, thereby influencing the water loss to varying extents. Thus, these methods have certain inherent drawbacks. The open-chamber gradient estimation method provides continuous measurement in ambient air, with little alteration of the microclimate overlying the skin surface.1 This method is therefore preferable and has gained wide use in the evaluation of skin barrier function. As a consequence of the different measuring principles, results obtained with different methods cannot be directly compared with any accuracy. In this chapter, attention is focused on the open-chamber gradient estimation method of the commercially available Evaporimeter EP1/EP2 (ServoMed), Tewameter (Courage and Khazaka), and DermaLab (Cortex Technology). The total amount of water vapor passing the skin can be divided into water vapor passing the stratum corneum by passive diffusion and water vapor loss as a result of sweating.2 Originally, the term transepidermal water loss was applied to indicate the amount of water vapor passing the stratum corneum by passive diffusion.2 However, nowadays TEWL refers to the total amount of water loss through the skin. Therefore, it must be kept in mind that TEWL is a reflection of stratum corneum barrier function for water only when there is no sweat gland activity.
45.2 OBJECT Nowadays there is an overwhelming amount of literature on various subjects in which TEWL measurement is used. In fact, TEWL measurement is one of the most frequently used non-invasive methods in skin irritancy studies, showing an exponential growth in the number of publications. Searching for “transepidermal water loss” in PubMed yielded 740 hits at the end of 2003, whereas the years 1993 and 1983 yielded about 240 and 60 hits, respectively. TEWL measures barrier function, and therefore this method is preeminently useful in those conditions that are characterized by stratum corneum impairment. Stratum corneum is impaired in several clinical conditions and as a result of exposure to irritants in the laboratory and occupational setting. Since TEWL measures total water vapor loss, another application of TEWL measurement may be to assess the degree of sweating in various conditions.
In view of the abundance of literature on this subject, this chapter can offer only a very limited selection of studies, to be considered as a short introduction.
45.2.1 EVALUATION OF BARRIER FUNCTION CLINICAL CONDITIONS
IN
A higher TEWL has been noticed on the involved skin in various types of dermatitis than with uninvolved sites. Shahidullah et al.3 have observed an increased TEWL on the involved and uninvolved skin in dermatitis, being related to the severity of the disease.3 Uninvolved skin of patients with (a history of) atopic dermatitis was demonstrated to have a higher TEWL than the same skin region of subjects without dermatitis.4,5 In psoriasis, TEWL levels at the plaques were higher than in uninvolved skin, dependent on the type of psoriasis.6 The erythrodermic type had higher values than the active plaque type, which in turn had higher values than the inactive plaque type. In patients with ichthyosis, the magnitude of increase in TEWL paralleled the severity of the scaling.7 TEWL measurement has been applied for monitoring the epidermal repair process in (burn) wounds. Soon after the burning, TEWL was extremely elevated, dependent on the degree of burn.8 Hammarlund et al. have found a linear relationship between TEWL and ambient percent relative humidity (RH) in neonates.9 By determining TEWL from different parts of the body and calculating the areas of the corresponding surfaces, the total cutaneous water loss could be obtained.
45.2.2 SWEAT GLAND ACTIVITY CONDITIONS
IN
CLINICAL
Lower TEWL on the hands and forearms in patients with generalized scleroderma were observed.10 In contrast, a higher TEWL was noticed in patients with Parkinson’s disease than in healthy subjects, as a result of sweating.11 In a patient with cholinergic urticaria, we have demonstrated an increased TEWL on gustatory provocations of increasing strength, followed by the emergence of urticarial lesions.12
45.2.3 EXPOSURE TO IRRITANTS, PREDICTIVE IRRITANCY TESTING The sensitivity of TEWL as a screening technique for early signs of irritancy in the laboratory setting was found to be superior compared with visual scoring,13,14 as well as with laser Doppler flowmetry, colorimetry, and skin thickness by ultrasound A-scan.14 A wide variety of irritants, such as detergents, solvents, etc., exert their damaging influence on the skin by impairing the barrier function of the stratum corneum. In a multiple 3-week repeated
Measurement of Transepidermal Water Loss by Semiopen Systems
exposure model we have demonstrated increasing TEWL values over time due to the cumulative irritating action of detergents and a solvent on the epidermal barrier.13 This increase in TEWL was different for the various irritants. Thus, it was possible to rank the irritants according to their irritant potential.13 However, for irritants that act largely on targets beneath the stratum corneum (for example, dimethyl sulfoxide and phenol), TEWL had a poor correlation with visual scoring.15 Predictive irritancy models were not only aimed to select the least irritant substance, but also to select a population that is at risk of developing chronic irritant contact dermatitis. In the laboratory setting, several factors determining a high skin susceptibility have been investigated. A close correlation has been demonstrated between preexposure TEWL level and TEWL after a single 24-h sodium lauryl sulfate (SLS) application.16,17 A similar correlation was found between preexposure TEWL and TEWL after 4 days of repeated SLS exposure.16,18 However, in a long-term repeated exposure model, lesser correlations between preand postexposure TEWL have been noted, which is probably due to the development of adaptation.19 It has been proven that existing dermatitis, irrespective of which type, in another location of the body may enhance reactivity to various irritants.3 Only those patients with active hand eczema had an increased susceptibility to SLS on the upper arm, as opposed to chronic and healed eczema patients and normal controls.17 We have confirmed the validity of this phenomenon for atopic dermatitis, where more severe dermatitis was accompanied with higher levels of reactivity.5 In contrast to the findings in experimental skin irritancy studies, prospective occupational field studies could not demonstrate the importance of preexposure barrier function as a risk factor for hand dermatitis,20,21 probably by the fact that multiple repeated exposures to irritant stimuli in certain occupations may result in adaptation.
45.2.4 COMPOSITION
OF THE
EPIDERMAL BARRIER
Lower TEWL values have been found in uninvolved atopic dermatitis skin, which was related to a decreased level of ceramides in the stratum corneum.22 UVB irradiation or feeding with an essential fatty acid-deficient diet caused a higher TEWL level in hairless rats, together with a reduction in covalently bound epidermal ceramides and incomplete intercellular multilamellar structures.23
45.2.5 EVALUATION OF EFFECTS OF MOISTURIZERS AND PROTECTIVE CREAMS Many studies have addressed the role of moisturizers in irritant-exposed skin. It was demonstrated that moisturizers could prevent irritant reactions induced by detergents, and also accelerate recovery of detergent-treated skin.24,25
385
One barrier cream was effective in the prevention of skin responses by SLS, tested in a repetitive irritation test, in which TEWL measurement appeared to be the most discriminating evaluation technique.26
45.3 METHODOLOGICAL PRINCIPLE 45.3.1 THE THEORY In the absence of forced convections, the human skin surface is surrounded by a water vapor boundary layer.27 This layer, which forms a physical barrier against the environment, constitutes the transition zone for transportation of moisture and heat from the body to the ambient air. Considering the skin surface as a water-permeable surface, the process of water exchange through this zone is described by Fick’s diffusion law: dm/dt = –D·A·dp/dx where m = water transport (g), t = time (h), D = diffusion constant, A = surface (m2), p = vapor pressure of the atmosphere (mmHg), and x = distance from skin surface to point of measurement (m). The diffusion flow dm/dt indicates the mass per square centimeter being transported in a period. This diffusion flow can be expressed in terms of vapor pressure gradient.28 This gradient is approximately constant in the absence of forced convection and under steady-state conditions. The vapor pressure gradient is computed from the difference between the vapor pressures measured at two different fixed heights situated perpendicularly above the skin surface and within the zone of diffusion.1 The actual vapor pressure at each fixed height of measurement above the skin is calculated from the formula p = RH·psat, where p is the water vapor pressure (Pa), RH is the relative humidity (%), and psat is the saturation vapor pressure (Pa). The RH is measured with a capacitive sensor based on an organic polymer with dielectric sensitivity to changes in RH. The saturated vapor pressure, which is a function of the temperature alone, is calculated from the measured temperature value obtained with a fast thermistor.1 These sensors are mounted on the measuring probe head, as mentioned previously (Figure 45.1). The calculated difference in the vapor pressure at the two fixed heights of measurement is the estimated vapor pressure gradient of the boundary layer of diffusion. From this gradient, the evaporative TEWL value, in g/m2h, is calculated by the signal processing units in the probe handle or main unit, and digitally displayed. This estimation is valid only within this boundary layer,1,29 its depth depending on the site, air speed, and convections, forced or free.29 In the absence of convection currents or drafts, a mean depth of about 10 mm may be assumed for this boundary layer.1,29
386
Handbook of Non-Invasive Methods and the Skin, Second Edition
Tewl
Skin
FIGURE 45.1 Diagram of the probe of the ServoMed Evaporimeter with an open measuring chamber and two sensors mounted 3 and 9 mm over the skin surface.
45.3.2 THE METHOD Nowadays there are three different types of instruments capable of measuring TEWL using the open-chamber technique: the Evaporimeter EP1 and EP2 (ServoMed, Stockholm, Sweden), Tewameter TM210 and TM300 (Courage and Khazaka, Cologne, Germany), and DermaLab (Cortex Technology, Hadsund, Denmark). The construction of the latter two instruments is based on that of the ServoMed Evaporimeter. All instruments can be connected with a personal computer (PC) for data management. The measuring head of the probe is placed on the selected skin surface, and a small area (1 cm2) of skin is limited for measuring the TEWL. The electronic controls and digital display unit offer various measuring possibilities and ranges. Further details of the technical descriptions and accuracies of the instruments are given elsewhere.1,30–32 The Evaporimeter EP1 and EP2 (ServoMed) consist of a detachable measuring probe connected by a cable to a portable main signal processing unit (Figure 45.2). The sensor arrangement within the probe head is represented diagrammatically in Figure 45.1. The Teflon capsule of the probe head has a cylindrical measuring chamber, open at both ends, with a diameter of 12 mm and height of 15 mm. Within this chamber, relative humidity sensors (hygrosensors) are paired with temperature sensors (ther-
FIGURE 45.2 The ServoMed EP1 Evaporimeter, measurement box and pen recorder.
mistors), at two fixed heights of 3 and 9 mm above the skin surface, i.e., within the boundary layer of diffusion. The probe head of the Tewameter has a diameter of 10 mm and a height of 20 mm, with the sensors at a distance of 4.5 and 10.5 mm from the skin surface. The Tewameter displays curves of the actual measuring TEWL value and values for average and standard deviation TEWL, or temperature and relative humidity of each sensor, simultaneously. Furthermore, the skin surface water loss can be shown for sorption–desortpion experiments. The newer Tewameter TM300 version has all electronic circuits in its probe, which can be used in combination with a traditional display main unit, or as stand-alone, connected with a small display device, or with a PC. Figure 45.3 shows the curves for the actual TEWL value, its average value, and its standard deviation, as shown on the PC. The probe can be fixed with a supplied holder. The DermaLab has a display that shows the actual TEWL value in figures, and the mean and standard deviation values over the last eight measurement cycles. There are four measurement cycles per second. It is possible to adjust that the instrument stops measuring after a certain period or after a certain standard deviation has been
FIGURE 45.3 Curves of actual TEWL values, its average and its standard deviation, as shown on the PC, connected with the Tewameter TM210.
Measurement of Transepidermal Water Loss by Semiopen Systems
387
FIGURE 45.4 Curve of the actual TEWL and figures of average and standard deviation values, as shown on the PC, connected with the DermaLab.
reached. The DermaLab has a built-in thermoprinter. Figure 45.4 depicts the graphic display on the PC of the actual TEWL, and figures of average and SD values.
45.4.2 INSTRUMENT-RELATED VARIABLES
45.4 SOURCES OF ERROR
45.4.2.1 Start-Up and Use
45.4.1 VALIDITY OF METHOD VARIABLES
AND
ASSOCIATED
The measurement of TEWL with the Evaporimeter is valid only within the boundary layer of diffusion surrounding the human body.1 The depth of this boundary layer is therefore crucial and depends on the environmental conditions (see Section 45.3). Thus, it is apparent that any environment- or instrument-related variables that influence the depth of this boundary layer would affect the gradient, and therefore the measured TEWL value. Furthermore, due to the extreme sensitivity of the instrument, any variations in the microclimate, whether due to the instrument, environment, or individual, are immediately detected and instantly displayed as a fluctuation, indicating an error in the measured TEWL level. A detailed account of these influencing variables associated with the method of measuring TEWL using the Evaporimeter is given in the Guidelines of the European Contact Dermatitis Society.33 In this chapter, guidelines as to what should be considered good laboratory practice are given for the measurement of TEWL. To perform accurate and precise TEWL measurements, sources of variation due to the instrument, environment, and individual need to be known and taken into account.33
The commonly occurring and important instrumentrelated variables are discussed in this section.
The instruments should be turned on at least 15 min before measurements are performed, and if the instrument is being used intermittently during the day, it should not be switched off between measurements.30–32 45.4.2.2 Zeroing After the warm-up period, the instrument should be zeroed only if the values exceed +0.5 g/m2h or are lower than –0.5 g/m2h during a zero measurement.31,33 Regularly calibrated and well-maintained instruments will not require this zeroing daily if the offset knob is not used for zeroing in between measurements (see Section 45.4.2.4 on zero drift). 45.4.2.3 Measuring Stabilization of the TEWL value is usually reached by 30 to 45 sec after the start of measuring.34,35 This prescribed stabilization period is for baseline TEWL measurements only. If, however, measurements are made on excessively diseased or damaged skin sites, where high water evaporation rates are expected, or at high ambient relative humidities, a longer stabilization period may be necessary. This prescribed period is not a hard-and-fast rule, as it may also vary from instrument to instrument.35
388
Handbook of Non-Invasive Methods and the Skin, Second Edition
The skin area measured is approximately 1 cm2, and the smallest amount measured corresponds to 0.00000006 g/cm2sec (2.16 g/m2h).36 Such precision also means that disturbances in the microclimate are immediately detected as a fluctuation in TEWL. Therefore, further fluctuations after stabilization may occur, and as a rule, it is easier to read the TEWL value from a curve that shows the time course, because the average value is more readily comprehended from an analog than from a digital reading. For this purpose, a pen recorder can be used, or the graphic display on the Tewameter’s main unit, or the graphics on the PC, connected with the measuring instrument. The ServoMed Evaporimeter built-in damping filters may be used to smooth these fluctuations in TEWL.30 Two filters giving time constants of 10 and 20 sec are available, and thus the instrument can be operated with variable time constants, 0, 10, 20, and 30 sec (10 + 20 sec). When using the filters, press filter button 10 after the 45-sec stabilization period, wait about 5 sec, and then press button 20.36 The TEWL value registered (recorded or displayed) during the 30-sec period after stabilization is to be considered the measured value, only if it shows a straight horizontal line.35 The Tewameter and DermaLab instruments possess software that enables the assessment of average values over a range from 2 to 60 sec, to be adjusted prior to the measuments.31,32 Averages of 20 or 30 sec are advised. From each measurement series the standard deviation is also calculated. This standard deviation can be adjusted beforehand to a desired value, thus helping the investigator to judge when a stable reading is achieved. However, with the filters on or using averages, the investigator obtains no information about fluctuations and artifacts (e.g., slipping off the exposed site). Therefore, in that case, direct recording of the TEWL values on a pen recorder or otherwise, as mentioned above, is recommended.33 45.4.2.4 Zero Drift Displacement of the water evaporation (WE) zero level, in between measurements, is attributed to the abrupt humidity changes as well as to the temperature changes of the probe, as a result of the measurement itself (see below). 45.4.2.5 Humidity Changes After a TEWL measurement has been made, condensation vapor will remain within the funnel of the probe for some time, causing a moisture gradient to persist.30,31 Therefore, the instrument will for some time continue to indicate a nonzero WE value, which will of course disappear when the water vapor in the probe has evaporated. To accelerate the evaporation of this moisture, the probe may be carefully waved, vertically up and down. A well-maintained
instrument will return to zero within about 2 to 4 min postmeasurement.35 As this phenomenon is entirely a matter of a delayed response, and not a real zero displacement, avoid using the offset button in between measurements for zeroing the instrument. Allow it to reach (almost) the zero level on its own, before the next measurement is made. 45.4.2.6 Temperature Changes The temperature-dependent variability of the sensors of the probe, as well as the amplifiers in the probe handle, is of importance when measuring TEWL.30,31 During a measurement, the temperature of the probe increases as a result of heating from the measured skin surface, and also due to the operator’s hand. The warmth from the hand influences the amplifiers in the handle of the probe, and thereby also the TEWL measurement.30,31,37 A displacement in the WE zero level of the order of ±1 to 2 g/m2h may result due to a 5-min-long measurement, in which the probe is handheld. Therefore, avoid holding the probe directly by hand, particularly in repeated measurements. The probe is best handled with an insulating glove, the holed rubber stopper of the ServoMed Evaporimeter’s calibration set, a lightweight (laboratory) burette clamp with rubber-covered ends,33 or the Tewameter’s supplied holder.31 Full control of the position of the probe and its contact with the skin surface should still be possible when using the above-mentioned accessories. 45.4.2.7 The Surface Plane A standing person, warmer or cooler than the surrounding ambient air, will act as a chimney and cause an increased convection of air close to the surface. To avoid this chimney effect, place the measuring surface in a horizontal plane and apply the probe parallel to this surface.28 Moreover, with a horizontal plane, it is easier to control the probe position and pressure against the skin, thereby avoiding probe movement during the measurement. 45.4.2.8 Contact Pressure Variations in the contact pressure between the probe and the surface of the skin may cause alterations in the TEWL, due to changes in the distance between the skin and the sensors, and due to changes in the water permeability of the skin.1 If the pressure is too light, gaps arise between the probe head and the skin surface. If the pressure is too heavy, there is a risk that the lower sensor comes into contact with the skin, which will have a negative effect on its function. Therefore, a constant light pressure should be applied when holding the probe against the skin.
Measurement of Transepidermal Water Loss by Semiopen Systems
45.4.2.9 Use of the Probe Protection Covers Although it is recommended that these protection covers, supplied with the ServoMed Evaporimeter, be applied whenever possible during measurements,30 it is important to recognize the variables that they introduce.37 The use of the protection cover with the screen and the grid (no. 2107, ServoMed) elevates the probe, and therefore the sensors, above the water vapor boundary layer surrounding the skin (see Section 45.3), due to the added height (6 to 7 mm) of the stainless-steel screen. This will influence absolute TEWL measurement,1,29,37 and has no importance only if the Evaporimeter is used for relative measurements.33 However, measurements made with and without the screen (relative or absolute) cannot be directly compared. With the screen, the TEWL values will be somewhat lower than without, the difference becoming greater as the TEWL rate increases.37,38 Use of these protection covers should thus be stated clearly in reports and publications. For measuring absolute TEWL levels, the uncovered probe or the protection cover without the screen and grid (no. 2108, ServoMed; disposable plastic rings, Cortex Technology) is thus recommended. If it is foreseeable that contaminations (ointments, oils, sweat, etc.) from the measuring skin surface may arise, or if a sterile probe surface is required, the protection cover without the stainless-steel screen and grid (no. 2108, ServoMed), which can be sterilized, should be used. The elevation of the probe above the measuring surface due to this attachment is negligible.37 However, in hairy body regions, hair follicles may come into contact with the sensors. Dust and evaporated substances, including solvents, may also damage the sensors. In such conditions, the protection covers with the screen and grid (no. 2107, ServoMed) should be used,33 or these hairy regions should be shaved some days prior to the first measurement. 45.4.2.10 Intra- and Interinstrumental Variability A high reproducibility of results was found for individual ServoMed Evaporimeters, i.e., a low intrainstrumental variability.35 However, large differences were found between the Evaporimeters, i.e., a high interinstrumental variability.35 Interinstrumental variability is dependent on the age of the Evaporimeter. The newer instruments respond faster, i.e., stabilization time is shorter. The older versions appear to measure much lower TEWL values.35 This may be attributed to the sensors of the probe, which are reported to undergo a slight aging.30 To keep this in check, the calibration of the instrument should be checked from time to time (see Section 45.4.2.11 on calibration). To enable successful and reliable interlaboratory comparison of results, overcoming the effect of the interinstrumental variability, an additional calibration procedure using a standard constant water evaporation device can be
389
adopted.33 However, it is recognized that there is a certain degree of inaccuracy inherent to this method,33 due to the extreme sensitivity of the sensors. 45.4.2.11 Calibration As a general rule, calibration of the instruments, according to the manufacturer’s specifications, should be performed at regular intervals.30–32 For details of calibration and maintenance, refer to the operation handbooks.30–32 45.4.2.12 Accuracy It was reported that the ServoMed Evaporimeter may underestimate the water evaporation rate.29 At evaporation rates of about 20 g/m2h, the underestimation is only about 10%. When the evaporation rates exceed 80 g/m2h, however, the underestimation may be about 50%, even in still air. It was concluded that the presence of the probe may restrict the flux of water vapor, particularly at the high evaporation rates.29 This was also reported by Scott et al.,39 who stated that this underestimation is at evaporation rates above 75 g/m2h. The same holds true for the other instruments. Most of the changes in the barrier function that are caused by detergent damage or diseased states result in water evaporation rates within the range of 20 to 60 g/m2h. Therrfore, the consequences of a potential underestimation in water evaporation rates may be acceptable. However, more severe damage to the barrier, such as burns or wounds, gives rise to much higher evaporation rates (above 100 g/m2h), where the underestimation of TEWL is relevant and should be considered in the interpretation of the results. 45.4.2.13 Performance of the Different Types of Instruments Important measures of performance are mean, standard deviation, and coefficient of variation values of a series of measurements, as tested in in vitro and in vivo conditions. In a study on four ServoMed Evaporimeters using in vitro and in vivo measurements, it was found that one instrument had a much lower mean value than the others,35 being beyond the permissible range (15%) of accuracy specified by the manufacturer.30 In a study on TEWL measurements at eight sites of both foreams of 30 individuals, the coefficient of variation was 15.4% for intraindividual comparisons and 35.9% for interindividual comparisons, using the ServoMed system.40 Leaving both distal sites (near the wrists) out of the calculations yielded coefficient of varation values of 11.0 and 36.4% for intraand interindividual comparisons, respectively.40 Comparing the performance of one ServoMed Evaporimeter with one Tewameter on repetitive measurements on various body regions, it was noted that the mean values obtained from the Tewameter were more than two times as high as
390
Handbook of Non-Invasive Methods and the Skin, Second Edition
those from the Evaporimeter.41 The coefficient of variation values for intraindividual comparisons were in the range of 3 to 8% for both instruments.41 Values of both instruments showed a very close correlation (R = 0.97).41 In another study, five ServoMed Evaporimeters and four DermaLab instruments were evaluated by repetitive measurements.42 The ServoMed and DermaLab instruments showed the same mean values for relative humidity, corresponding to the values from a thermohygrometer. However, the standard deviation values of the ServoMed Evaporimeters were high (3.08) compared with those of DermaLab (0.63).42 Repetitive determinations of evaporative water loss on a standard device showed a close correlation (R = 0.97) between the ServoMed and DermaLab. The absolute mean values of the five ServoMed and four DermaLab devices, however, were not similar, as the ServoMed had higher values than the DermaLab, which was especially true for the lower measuring range.42 The coefficient of variation values were again higher in the ServoMed instruments.42 In determining reproducibility and accuracy of these instruments in in vivo conditions, the ServoMed probes were equipped with the protection covers model 2107, which elevate the probe head about 6 mm, whereas the DermaLab probes were equipped with plastic discs that elevate the probe only 0.25 mm. The systematic error introduced by this difference in elevation hampers a good comparability, and therefore the results are erroneous.42 In conclusion, large differences in mean values were noted among the instruments from the different manufacturers, and even among instruments from the same company. In contrast to these absolute comparisons, measurement values from ServoMed and Tewameter, and ServoMed and DermaLab showed close correlations. Thus, comparisons between instruments from different manufacturers can be considered reliable in relative measurement outcomes (e.g., when different detergents are tested), but not when absolute data are given (e.g., TEWL values of untreated skin of various body regions).
This variable is mainly determined by variations in ambient air temperature and relative humidity. Provided that measurements are made at room temperatures between 20 and 22˚C, a seasonal variation in TEWL is mainly attributable to the seasonal variation in ambient relative humidity and skin hydration state, the latter itself influenced by the former.33 Publications should provide information about the season of the year in which the TEWL measurements were made.
45.4.3 ENVIRONMENT-RELATED VARIABLES
45.4.3.5 Direct Light
The effects of the most important environment-related variables are discussed in this section.
Direct light warms up the surface of the object, which in turn warms up the air close to the object, and an air convection is created.30,31 TEWL measurements should not be made under direct light sources or close to windows with direct sunlight.33
45.4.3.1 Air Convections This is the main source of disturbance resulting in rapid fluctuations of the measurements.33 It is commonly produced by disturbances in the room, such as people moving about, opening and closing doors, breathing across the measurement zone, air conditioners, etc. As these disturbances are difficult to avoid, some form of an enclosure (a measuring box) to serve as a draft shield is
recommended.33 A box or an incubator with an open top and sides preferably of Perspex (for visibility), with holes for the placement of the forearms (subject and investigator), will protect the measurement zone from rapid air movements (Figure 45.2). 45.4.3.2 Ambient Air Temperature The most important effect of the temperature of the ambient air is that it influences the skin temperature both directly (by convection) and indirectly (by central thermoregulatory effects).2,43–45 With increasing ambient air temperature, the skin surface temperature increases, and the TEWL is almost double at the high ambient air temperature of 30˚C, in comparison with that at the lower ambient air temperature of 22˚C.28 A room temperature of 20 to 22˚C is recommended (see below). 45.4.3.3 Ambient Air Humidity Ambient relative humidity is a complex and important variable that influences TEWL measurements.33 Ambient room relative humidity should be registered during TEWL measurements, reported in publications, and taken into consideration whenever TEWL results are compared. If climate room facilities are available, the relative humidity should be regulated to about 40%.33 45.4.3.4 Seasonal Variation
45.4.4 INDIVIDUAL-RELATED VARIABLES The effects of the most important individual-related variables are discussed in this section. The effects of other common individual-related variables are given elsewhere.33
Measurement of Transepidermal Water Loss by Semiopen Systems
45.4.4.1 Sweating Physical, thermal, and emotional sweating are important variables to control, in order to make accurate TEWL measurements. If the ambient air temperature is below 20˚C, and the skin temperature is below 30˚C, thermal sweat gland activity is unlikely, provided that the skin is not exposed to forced convection and no excessive body heat is produced (result of physical exercise).3,46,47 Therefore, a premeasurement 15- to 30-min rest in a measurement room with an ambient air temperature regulated to about 20˚C, possibly by an air conditioner, is best suited for accurate TEWL measurements.47 Emotional sweating may be controlled by performing a couple of dummy TEWL measurements to put the test subject at ease.48 45.4.4.2 Skin Surface Temperature Skin surface temperature is one of the essential factors dictating the rate of TEWL in normal skin,28 and preconditioning of the test person is required (see above). At ambient room temperatures around 20 to 22˚C, however, the normal range of skin surface temperature is between 28 and 32˚C, which may be considered too little a variation to influence TEWL significantly.28,33
45.5 CONCLUSION TEWL measurement is a frequently used method to determine the barrier function of the stratum corneum. With the development of the open-chamber technique by Nilsson1 there has been much progress in making this method easier to handle than with the closed-chamber techniques. However, TEWL measurement is a very delicate technique for which it is of utmost importance to be aware of the many pitfalls in the methodology of measurement and interpretation of the results, in order to achieve reliable outcomes. Therefore, researchers should adhere to the guidelines published on this subject.33 When authors decide not to follow a particular aspect of the guidelines, they should make this absolutely clear by stating that they have not adhered to the guidelines in this respect, and also give the reason for this. Nowadays, there are three different manufacturers who make open-chamber instruments. The mean values, accuracy, and reproducibility of the various types, however, show different results. This necessitates good independent studies on these performance parameters on all types of instruments, incorporating the use of a standard evaporation device for gravimetrically controlled observations, as a gold standard. Comparisons between instruments from different manufacturers can be considered reliable in relative measurement outcomes only, but not for absolute data.
391
REFERENCES 1. Nilsson, G.E., Measurement of water exchange through skin, Med. Biol. Eng. Comput., 15, 209, 1977. 2. Rothman, S., Insensible water loss, in Physiology and Biochemistry of the Skin, University of Chicago Press, Chicago, 1954, p. 233. 3. Shahidullah, M., Raffle, E.J., Rimmer, A.R., and FrainBell, W., Transepidermal water loss in patients with dermatitis, Br. J. Dermatol., 81, 722, 1969. 4. Van der Valk, P.G.M., Nater, J.P., and Bleumink, E., Vulnerability of the skin to surfactants in different groups of eczema patients and controls as measured by water vapour loss, Clin. Exp. Dermatol., 10, 98, 1985. 5. Tupker, R.A., Coenraads, P.J., Fidler, V., de Jong, M.C.J.M., van der Meer, J.B., and De Monchy, J.G.R., Irritant susceptibility and wheal and flare reactions to bioactive agents in atopic dermatitis: I. Influence of disease severity, Br. J. Dermatol., 133, 358, 1995. 6. Ghadially, R., Reed, J.T., and Elias, P.M., Stratum corneum structure and function correlates with phenotype in psoriasis, J. Invest. Dermatol., 107, 558, 1996. 7. Frost, P., Weinstein, G.D., Bothwell, J.W., and Wildnauer, R., Ichthyosiform dermatosis. III. Studies on transepidermal water loss, Arch. Dermatol., 98, 230, 1968. 8. Lamke, L.-O., Nilsson, G.E., and Reithner, H.L., The evaporative water loss from burns and the water-vapour permeability of grafts and artificial membranes used in the treatment of burns, Burns, 3, 159, 1977. 9. Hammarlund, K., Nilsson, G.E., Oberg, P.A., and Sedin, G., Transepidermal water loss in newborn infants. I. Relation to ambient humidity and site of measurement, and estimation of total transepidermal water loss, Acta Pediatr. Scand., 66, 553, 1977. 10. Serup, J. and Rasmussen, I., Dry hands in scleroderma. Including studies of sweat gland function in healthy individuals, Acta Dermatol. Venereol. (Stockh.), 65, 419, 1985. 11. Turkka, J.T. and Myllyla, V.V., Sweating dysfunction of Parkinson’s disease, Eur. Neurol., 26, 1, 1987. 12. Tupker, R.A. and Doeglas, H.M.G., Water vapour loss threshold and induction of cholinergic urticaria, Dermatologica, 181, 23, 1990. 13. Tupker, R.A., Pinnagoda, J., Coenraads, P.J., and Nater, J.P., The influence of repeated exposure to surfactants on the human skin as determined by transepidermal water loss and visual scoring, Contact Derm., 20, 108, 1989. 14. Agner, T. and Serup, J., Sodium lauryl sulphate for irritant patch testing: a dose-response study using bioengineering methods for determination of skin irritation, J. Invest. Dermatol., 95, 543, 1990. 15. Van der Valk, P.G.M., Kruis-de Vries, M.H., Nater, J.P., Bleumink, E., and De Jong, M.C.J.M., Eczematous (irritant and allergic) reactions of the skin and barrier function as determined by water vapour loss, Clin. Exp. Dermatol., 10, 185, 1985.
392
Handbook of Non-Invasive Methods and the Skin, Second Edition
16. Pinnagoda, J., Tupker, R.A., Coenraads, P.J., and Nater, J.P., Prediction of susceptibility to an irritant response by transepidermal water loss, Contact Derm., 20, 341, 1989. 17. Agner, T., Skin susceptibility in uninvolved skin of hand eczema patients and healthy controls, Br. J. Dermatol., 125, 140, 1991. 18. Tupker, R.A., Coenraads, P.J., Pinnagoda, J., and Nater, J.P., Baseline transepidermal water loss (TEWL) as a prediction of susceptibility to sodium lauryl sulfate, Contact Derm., 20, 265, 1989. 19. Tupker, R.A., Pinnagoda, J., Coenraads, P.J., and Nater, P.J., Susceptibility to irritants: role of barrier function, skin dryness and history of atopic dermatitis, Br. J. Dermatol., 123, 199, 1990. 20. Smit, H.A., van Rijssen, A., Vandenbroucke, J.P., and Coenraads, P.J., Individual susceptibility and the incidence of hand dermatitis in a cohort of hairdressers and nurses, Scan. J. Work Envir. Health, 20, 113, 1994. 21. Berndt, U., Hinnen, U., Iliev, D., and Elsner, P., Is occupational irritant contact dermatitis predictable by cutaneous bioengineering methods? Results of the Swiss Metalworkers’ Eczema Study (PROMETES), Dermatology, 198, 351, 1999. 22. Imokawa, G., Abe, A., Jin, K., Higaki, Y., Kawashima, M., and Hiddano, A., Decreased level of ceramides in SC of atopic dermatitis: an etiologic factor in atopic dry skin, J. Invest. Dermatol., 96, 523, 1991. 23. Meguro, S., Arai, Y., Masukawa, Y., Uie, K., and Tokimitsu, I., Relationship between covalently bound ceramides and transpediermal water loss (TEWL), Arch. Dermatol. Res., 292, 463, 2000. 24. Ramsing, D.W. and Agner, T., Preventive and therapeutic effects of a moisturizer. An experimental study of human skin, Acta Derm. Venereol., 77, 335, 1997. 25. Lodén, M., Barrier recovery and influence of irritant stimuli in skin treated with a moisturizing cream, Contact Derm., 36, 256, 1997. 26. Frosch, P.J., Kurte, A., and Pilz, B., Efficacy of skin barrier creams. III. The repetitive irritation test (RIT) in humans, Contact Derm., 29, 113, 1993. 27. Gates, D.M., in Humidity and Moisture, Vol. 2, Wexler, A. and Amdur, E.J., Eds., Reinhold, New York, 1965, p. 33. 28. Nilsson, G.E., On the Measurement of Evaporative Water Loss. Methods and Clinical Applications, Thesis, Linkoping University Medical Dissertations, No. 48, Linkoping, Sweden, 1977. 29. Wheldon, A.E. and Monteith, J.L., Performance of a skin Evaporimeter, Med. Biol. Eng. Comput., 18, 201, 1980. 30. ServoMed Evaporimeters, Operation Handbook, ServoMed, Vallingby, Stockholm, Sweden, 1981. 31. Information and operating instructions for the Tewameter TM210 and the software for Windows NT, Courage and Khazaka, Cologne, Germany, 2001.
32. DermaLab user’s manual, Cortex Technology, Hadsund, Denmark, 2001. 33. Pinnagoda, J., Tupker, R.A., Agner, T., and Serup, J.,Guidelines for transepidermal water loss (TEWL) measurement, Contact Derm., 22, 164, 1990. 34. Blichman, C.W. and Serup, J., Reproducibility and variability of transepidermal water loss measurements, Acta Dermatol. Venereol. (Stockh.), 67, 206, 1987. 35. Pinnagoda, J., Tupker, R.A., Coenraads, P.J., and Nater, J.P., Comparability and reproducibility of the results of water loss measurements: a study of 4 evaporimeters, Contact Derm., 20, 241, 1989. 36. A guide to water evaporation rate measurement, ServoMed, Vallingby, Stockholm, Sweden. 37. Nilsson, G.E., personal communication, 1987. 38. Agner, T. and Serup, J., Transepidermal water loss and air convection, Contact Derm., 22, 120, 1990. 39. Scott, R.C., Oliver, G.J.A., Dugard, P.H., and Singh, H.J., A comparison of techniques for the measurement of transepidermal water loss, Arch. Dermatol. Res., 274, 57, 1982. 40. Pinnagoda, J., Tupker, R.A., Smit, J.A., Coenraads, P.J., and Nater, J.P., The intra- and inter-individual variability and reliability of transepidermal water loss measurements, Contact Derm., 21, 255, 1989. 41. Barel, A.O. and Clarys, P., Study of the stratum corneum barrier function by transepidermal water loss measurements: comparison between two commercial instruments: Evaporimeter and Tewameter, Skin Pharmacol., 8, 186, 1995. 42. Grove, G.L., Grove, M.J., Zerweck, C., and Pierce, E., Comparative metrology of the Evaporimeter and the DermaLab TEWL probe, Skin Res. Technol., 5, 1, 1999. 43. Grice, K.A., Transepidermal water loss, in The Physiology and Pathophysiology of the Skin, Vol. 6, Jarret, A., Ed., Academic Press, London, 1980, p. 2121. 44. Lamke, L.-O. and Wedin, B., Water evaporation from normal skin under different environmental conditions, Acta Derm. Venereol., 51, 111, 1971. 45. Rothman, S., The role of the skin in thermoregulation: factors influencing skin surface temperature, in Physiology and Biochemistry of the Skin, University of Chicago Press, Chicago, 1954, p. 258. 46. Baker, H. and Kligman, A.M., Measurement of transepidermal water loss by electrical hygrometry. Instrumentation and responses to physical and chemical insults, Arch. Dermatol., 96, 441, 1967. 47. Pinnagoda, J., Tupker, R.A., Coenraads, P.J., and Nater, J.P., Transepidermal water loss: with and without sweat gland inactivation, Contact Derm., 21, 16, 1989. 48. Pinnagoda, J., A Pilot Field Study for the Assessment of the Practicability of the Recommendations for the Transepidermal Water Loss Measurements and the Skin Patch Testing Technique, Transepidermal Water Loss, Thesis, University of London, 1990, p. 172.
of Transepidermal 46 Measurement Water Loss by Closed-Chamber Systems J. Nuutinen Delfin Technologies Ltd., Kuopio, Finland
CONTENTS 46.1 Introduction............................................................................................................................................................393 46.2 Measurement Principles of the Present Closed-Chamber Systems......................................................................393 46.3 Calibration..............................................................................................................................................................394 46.4 Update to the Guidelines of TEWL Measurement ...............................................................................................395 References .......................................................................................................................................................................396
46.1 INTRODUCTION Conventionally, instruments measuring transepidermal water loss (TEWL) or skin surface water loss (SSWL) are based on the measurement of water vapor pressure gradient on the upper surface of the skin by open chambers.1 The steep vapor pressure gradient then indicates a high water evaporation rate. The gradient is measured inside a cylindrical chamber with two humidity sensors at fixed distances from the skin surface. Due to an open-chamber principle, the air inside the chamber is influenced by external airflows and body-induced air currents.2,3 It has been found that the disturbance of the measuring probe on the microclimate near the skin4 at high evaporation rates of the palms, soles, or diseased skin sites will result in underestimation of the TEWL values.2,5 Due to a stabilization of the instrumental readings, long measurement times have been described.6 Evidently, the water vapor pressure gradient does not stay similar for different probe contact angles, since the TEWL values are dependent on whether the measurement site, and thus the probe, is facing upward or downward.3 Recently, a new generation of the closed-chamber systems for the measurement of TEWL has been introduced.3,7,8 A major concern with the closed chambers relates to the blocking of normal evaporation through the skin. This problem has been solved either by the use of a water vapor absorber7 or by minimizing the blocking effect by using a short measurement time of less than 10
sec.8 A 20-sec application time may be too long at high evaporation rates.3 From the practical point of view, the closed-chamber systems offer obvious advantages. In this chapter a brief introduction to the closed-chamber principle, and especially to the unventilated methods, is presented. Finally, an update to the recommendations originally established for the open-chamber systems is presented to include the measurement site and the range of the TEWL values.
46.2 MEASUREMENT PRINCIPLES OF THE PRESENT CLOSED-CHAMBER SYSTEMS Recently, three new closed-chamber systems have been introduced.3,7,8 The devices consist of a closed air volume for measuring either the increase of relative humidity in the chamber or the evaporated water by a condenser element. Based on technical descriptions and publications, the major differences are related to the calibration procedure and the calculation of the TEWL value. Tagami et al.3 introduced the portable TEWL device H4300 (Nikkiso-Ysi Co. Ltd., Tokyo, Japan). The device consists of cylindrical space with a volume of 0.904 cm3, a main unit, a 150-cm cable, and a measuring probe with a humidity sensor and thermistor. The measuring time is reported to be 18 sec. The exact method to calculate the evaporation rate has not been published. The AquaFlux system (Biox Systems Ltd., London) utilizes an electronically cooled condenser inside the 393
Handbook of Non-Invasive Methods and the Skin, Second Edition
chamber.7 The probe consists of a cylindrical chamber with a diameter of 8 mm and a length of 12 mm. The condenser, at the top of the chamber, is a metal plate with a temperature below the freezing point of water in order to absorb water vapor from the air to the cool condenser. The aim of the cooler is to create a controlled microclimate and maintain the humidity at a known value inside the chamber. The calculation of the evaporation rate with the AquaFlux applies the one-dimensional form of Fick’s first law:
60 Relative humidity (%)
394
50 40 30 20
dρ dz
4
6
8
10
12
14
8 10 Time (sec)
12
14
(a)
60 TEWL = 220 g/m2h
50 40 30 20
0
2
4
6
(b)
FIGURE 46.1 Increase of the relative humidity (RH) inside the closed chamber of the VapoMeter (a) at low and (b) at high evaporation rates. The vertical lines show the period when the determination of the evaporation value is performed. Note the beginning of the saturation effect at high evaporation rate after the 10-sec contact time. This demonstrates that the measurement time should be very short when measuring high water loss (Table 46.2). (From Nuutinen, J. et al., Skin Res. Technol., 9, 85, 2003. With permission.)
Δm = (constant) k
(46.4)
(46.2)
where RHa is an initial relative humidity of the chamber and ambient air at t = 0. k is the slope of the rising relative humidity, and thus is directly proportional to water loss through the skin. During a short-term experiment, Equation 46.2 approximates an increase of relative humidity inside the chamber. A too long measuring time would generate a saturation effect. Relative humidity is related to the mass of the water inside the chamber. Hence, Equation 46.2 can be rewritten for the mass of water m(t) at time t: m(t) = ma + Δm t A = ma + (constant) k t A (46.3) since
2
(46.1)
where D is the mass diffusion coefficient for water vapor and z is an axis parallel with the axis of the chamber. According to Equation 46.1 the water vapor flux density J (the evaporation) can be determined from the gradient of water vapor density (dρ/dz). Applying a one-dimensional solution for Equation 46.1, the evaporation rate can be determined. The calculation requires the temperature of the cooler, relative humidity and the temperature at a fixed place inside the chamber (for example, midway between the measured surface and the cooler). The VapoMeter (Delfin Technologies Ltd., Kuopio, Finland) measures the increase of relative humidity inside a closed chamber with an air volume of 2.0 cm3 and an open contact area of 1.0 cm2. The humidity and temperature sensors are mounted on a portable handheld (weight, 150 g), battery-operated, and microprocessor-controlled electronic unit with no measuring cables. The device is provided with a readout for the TEWL value and an optional data interface to an external PC. During a 7- to 9-sec application time of a closed chamber, the increase of relative humidity RH(t) against time t (Figure 46.1) can be approximated as a simple linear model: RH(t) = RHa + k t
0
Time (sec)
Relative humidity (%)
J = −D
TEWL = 26 g/m2h
ma is the initial mass of water, Δm the increase of mass per unit time and surface area, and A the surface area of evaporation. The quantity Δm is directly proportional to the slope k according to Equation 46.4 and can be expressed in units of g/m2h. The proportionality constant related to chamber dimensions and the mass of saturated water vapor can be determined from the slope k by evaporating a known amount of water and measuring the respective decrease in the mass of water.
46.3 CALIBRATION The calibration of the TEWL device H43003 has not been described. The AquaFlux system is calibrated using a 2 μl drop of water measured by a micropipette. The water is
Measurement of Transepidermal Water Loss by Closed-Chamber Systems
allowed to evaporate and diffuse to the condenser. During the calibration procedure the flux density is registered against time. An area under the curve represents the response of the device for the known amount of water. The calibration procedure of the VapoMeter consists of a measurement of a change of a known mass of water and the respective evaporation rate. The calibrator is covered with either a semipermeable membrane or a porous plate in order to change the evaporation rate of the calibrator.8 Due to the sensitive humidity sensors, the probe should be situated close to the evaporative calibrator surface just before the calibration measurement. The sensitive element of the probe should not be unnecessarily near the hands; i.e., the unintended humidity should not reach the sensors and the quantity RHa (Equation 46.2) should stay constant. Naturally, the same precautions should be applied in the use of the devices. Furthermore, additional instructions given by the manufacturers for temperature effects should be followed in order to avoid warming effects due to a continuous handling of the probe or a close proximity to the external source of heat. An obvious advantage for all manufacturers and users of the TEWL instruments would be a calibration procedure that would be independent from the applied techniques or instruments. An attempt to achieve a generally accepted method to calibrate the evaporation devices is at present ongoing (http://www.skin-forum.org.uk/abstracts/ helen-packham.html). The aim of the project is to generate an accurate and reproducible calibration method for the evaporation measurements. At present, a suitable calibration device introduced in the 1990 guidelines is available.2 An integrated heating system for this standard evaporation device is recommended to assess the TEWL readings at high evaporation rates (until 200 g/m2h). Since the evaporation rates of this device are dependent on ambient air humidity,2 the calibration circumstances should also be standardized.
46.4 UPDATE TO THE GUIDELINES OF TEWL MEASUREMENT A number of variables affecting the TEWL measurements have been reported. Many of them are related to the conventional open-chamber techniques,2,9 since until recently practical closed-chamber techniques have not been commercially available. Therefore, variables associated with warm body-induced airflow and air convection in the investigation room, instability effects related to zero drift, the temperature of the probe, and the horizontal position of the probe are absent or negligible with the closedchamber techniques. Due to a small probe size of some devices, a freely selectable position, and an adequate contact of the probe against the measured surface, the closed-
395
chamber techniques have enlarged the range of potential applications to many anatomically difficult sites, like nails, lips, eyelids, scalp, and axilla.10 The evaporation rate measurements are now possible at almost any probe angle, not just on the upper side of the investigated surface.3 Evidently, the further advantage of some closed-chamber systems relates to their ability to measure high TEWL values up to 200 g/m2h without a saturation problem.8 Underestimation of the evaporation rate reported to 10% at an evaporation rate of 20 g/m2h and 50% at 80 g/m2h may lead to serious misinterpretations of the results. 2,4 Recently, it has been demonstrated that one of the openchamber systems starts to saturate when the evaporation rate exceeds 80 g/m2h.8 In 1990, the report on the guidelines for transepidermal water loss (TEWL) measurements was published by the Standardization Group of the European Society of Contact Dermatitis.2 Recently, the European Group for Efficacy Measurements on Cosmetics and Other Topical Products (EEMCO) updated these guidelines.9 Both reports take into account the potential influence of several external and environmental-related factors, like room temperature and relative humidity, probe temperature, air circulation effect, seasonal, diurnal, or geographical effect, light effect, probe handling and occlusion, measuring angle, and skin cleansing effect on the registered value. The guidelines still form an adequate basis for assessing individual-related variables like age-, sex- and race-related factors, as well as factors associated with the anatomical site, sweating, vascular effects, and skin surface temperature, since these are mainly independent from the applied instrumentation. The most obvious need for the updated guidelines concerns the measurement site and the range of TEWL values. According to Table 46.1, a closed-chamber technique is recommended for small anatomical sites, nonhorizontal probe position, or curved measurement sites, while for flat surfaces and larger measurement sites all present instruments are applicable. Modifications violating the established measurement principle should not be used to avoid erroneous interpretation of the obtained results. Table 46.2 updates the recommendations for different ranges of the evaporation rates. The open-chamber and closed-chamber evaporimeters are applicable instruments at evaporation rates between 5 and 20 g/m2h, while due to the reported increasing underestimation of evaporation rates above 20 g/m2h,4 the closed-chamber systems are recommended. The linear range for the TEWL devices should reach up to 200 g/m2h. The lowest reproducible TEWL values are not well reported, but the estimated lower limit of the present instruments lies at 2 to 5 g/m2h. Although the TEWL devices have an important role in the investigation of human epidermal barrier function in dermatological, pharmacological, toxicological, and
396
Handbook of Non-Invasive Methods and the Skin, Second Edition
REFERENCES TABLE 46.1 Update to Guidelines for the TEWL Measurement: Measurement Site Measurement Sites Flat sites Curved sites Small anatomical area (diameter < 8 mm) Horizontal probe position Any other position
Recommended Technique Open chamber, closed chamber Closed chamber Closed chamber Open chamber, closed chamber Closed chamber
TABLE 46.2 Update to Guidelines for the TEWL Measurement: Measurement Range (g/m2h) Measurement Range 2–5 5–20 20–80 80–200
Recommended Technique No objective data available Open chamber, closed chamber Closed chamber if provided with short measurement time Closed chamber if provided with very short measurement time
cosmetic research,2,9 the same devices are also used to investigate water loss of many other objects, like animals, plants, or fruits. Also, the industry developing products for personal care and hygiene is utilizing the water loss measurements in a wide range of applications.11 Therefore, recommendations on the correct use, calibration, and reliable range of the TEWL values should be available in order to avoid erroneous results and misinterpretations.
1. Nilsson, G.E., Measurement of water exchange through skin, Med. Biol. Eng. Comput., 15, 209, 1977. 2. Pinnagoda, J., Tupker, R.A., Agner, T., and Serup, J., Guidelines for transepidermal water loss (TEWL) measurement. A report from the Standardization Group of the European Society of Contact Dermatitis, Contact Derm., 22, 164, 1990. 3. Tagami, H., Kobayashi, H., and Kikuchi, K., A portable device using a closed chamber system for measuring transepidermal water loss: comparison with the conventional method, Skin Res. Technol., 8, 7, 2002. 4. Wheldon, A.E. and Monteith, J.L., Performance of a skin Evaporimeter, Med. Biol. Eng. Comput., 18, 201, 1980. 5. Scott, R.C., Oliver, G.J.A., Dugard, P.H., and Singh H.J., A comparison of techniques for the measurement of transepidermal water loss, Arch. Dermatol. Res., 274, 57, 1982. 6. Blichman, C. and Serup, J., Reproducibility and variability of transepidermal water loss, Acta DermatoVenereol., 67, 206, 1987. 7. Imhof, R.E, O’Driscoll, D., Xiao, P., and Berg, E.P., New sensor for water vapour flux, in Sensors and Their Applications X, White, M.N. and Augousti, A.T., Eds., IoP Publishing, Bristol, 1999, p. 173. 8. Nuutinen, J., Alanen, E., Autio, P., Lahtinen, M.-R., Harvima, I., and Lahtinen T., A closed unventilated chamber for the measurement of transepidermal water loss, Skin Res. Technol., 9, 85, 2003. 9. Rogiers, V. and the EEMCO Group, EEMCO guidance for the assessment of transepidermal water loss in cosmetic sciences, Skin Pharmacol. Appl. Skin Physiol., 14, 117, 2001. 10. Nuutinen, J., Harvima, I., Lahtinen, M.-R., and Lahtinen T., Water loss through lip, nail, eyelid skin, scalp skin and axillar skin measured with a closed-chamber evaporation principle, Br. J. Dermatol., 148, 839, 2003. 11. Zhai, H., Ebel, J.P., Chatterjee, R., Stone, K.J., Gartstein, V., Juhlin, K.D., Pelosi, A., and Maibach, H.I., Hydration vs. skin permeability to nicotinates in man, Skin Res. Technol., 8, 13, 2002.
of Transcutaneous 47 Measurement Oxygen Tension Hirotsugu Takiwaki Department of Dermatology, The University of Tokushima School of Medicine, Tokushima, Japan
CONTENTS 47.1 47.2 47.3 47.4 47.5
Introduction............................................................................................................................................................397 Object and Methodological Principle....................................................................................................................397 Sources of Error.....................................................................................................................................................399 Correlation with Other Methods ...........................................................................................................................399 Clinical and Experimental Applications................................................................................................................400 47.5.1 Fundamental Study ....................................................................................................................................400 47.5.2 Clinical Application ...................................................................................................................................400 47.5.3 Analysis of Dynamics of Change in TcPO2 .............................................................................................402 47.6 Recommendation ...................................................................................................................................................403 References .......................................................................................................................................................................403
47.1 INTRODUCTION So-called cutaneous respiration, which means the absorption of oxygen and the elimination of carbon dioxide through the skin surface, was discovered over a century ago and was vigorously investigated in the early 1930s by Shaw and Messer.1 After 40 years, Evans and Naylor2 and Huch et al.3 measured the partial pressure of oxygen diffusing from the capillaries in the skin with a newly developed device consisting of a polarographic electrode set at the skin surface, and demonstrated that the oxygen tension, namely, transcutaneous PO2 (TcPO2), is close to zero. Since the direct oxygen supply from the atmosphere to the skin is excluded by the probe placed at the skin surface in this method, this finding indicates that almost all oxygen supplied by blood is consumed in the skin. However, vasodilatation induced by topically applied chemicals2,3 or by heat4 is found to increase the TcPO2 value and eventually to correlate with the PO2 in arterial blood (PaO2).4 By taking advantage of this, a compact polarographic electrode incorporated with a heating element was developed to monitor PaO2 noninvasively. Systems of this type were quickly commercialized and have been widely used in the management of premature or sick infants. Although their role in neonatal care units seems to have been more or less taken over by pulse oximeters that are easier to handle and less expensive,
TcPO2 monitors are still used for quantitative evaluation of peripheral circulation in various clinical and experimental studies. From a dermatological viewpoint, they offer useful information about microcirculation and respiration of the skin as well, since TcPO2 is determined not only by PaO2, but also by various cutaneous factors affecting oxygen flux to the skin surface. In this chapter, emphasis is not placed on the TcPO2 measurement simply as a noninvasive monitoring of PaO2, but rather as a method for the assessment of the respirocirculatory state of the skin.
47.2 OBJECT AND METHODOLOGICAL PRINCIPLE Transcutaneous PO2 is the partial pressure of oxygen measured with an electrode placed on the skin surface, i.e., a reflection of PO2 in the viable epidermis. When the skin temperature rises up to about 43˚C, TcPO2 values become close to the PaO2, especially in neonates. Therefore, TcPO2 measurement is most advantageously utilized for noninvasive monitoring to detect sudden changes in PaO2. If the systemic PaO2 is normal, abnormal TcPO2 values indicate that they are significantly influenced by peripheral or local factors, such as cutaneous blood flow, oxygen diffusibility, and cellular respiration, of the skin on which the sensor is placed. 397
398
Handbook of Non-Invasive Methods and the Skin, Second Edition
The TcPO2 measurement is based on the polarographic technique using the Clark-type electrode, which consists of a platinum cathode, a silver (Ag/AgCl) anode, and electrolyte solution (KCl, etc.) covered with a hydrophobic membrane (Teflon, polypropylene, etc.) permeable to only gases. If a suitable voltage (polarizing voltage) is applied between the cathode and the anode, the following reactions take place at the electrodes: O2 + 2H2O + 4e– = 4OH– 4Ag + 4Cl– = 4AgCl + 4e–
FIGURE 47.1 Instruments for transcutaneous PO2 measurement. Upper: two types of sensors, electrolyte solution, and gaspermeable membrane covering electrodes (arrows). In the upper sensor, a glass electrode for TcPCO2 measurement is incorporated. Lower: Control panel and recorder.
150 Occlusion
100
tcPo2
mmHg
The resulting current is proportional to the concentration of oxygen. In order to transform the output voltage signal into the unit of pressure (mmHg or Torr or kPa, 1 mmHg = 1 Torr = 133 Pa), the sensor has to be calibrated with gases, usually air and pure nitrogen, the oxygen concentration of which is known. In order to supply heat to the skin and to regulate temperature of the electrode, a heating element controlled by thermistors is set in the electrode. The sensor temperature can be optionally selected in the range of 37 to 45˚C in most commercially available instruments. These elements are incorporated into a small probe head that can be applied to most body regions with double-sided adhesive ring tape (Figure 47.1). Recent models contain electrodes for both oxygen and carbon dioxide in one sensor to obtain the transcutaneous PCO2 value simultaneously. These instruments have been provided by many companies (Linde Medical, Basel, Switzerland; Radiometer, Copenhagen, Denmark; Perimed, Järfälla, Sweden; Koken, Tokyo, Japan, etc.). By heating the skin, not only the cutaneous blood flow increases, but also the red blood cells dissociate more oxygen owing to the shift of the oxygen dissociation curve to the right by the temperature effect,5 both of which induce more flux of oxygen from capillaries to the skin surface. Transcutaneous PO2 therefore depends on skin temperature. About 15 to 20 min is needed before we can determine the TcPO2 value after placing the sensor on the skin, because some time is required to achieve maximal hyperemia. But once the steady state is achieved, TcPO2 quickly responds to the change of PO2 in the skin unless the sensor is placed on a hyperkeratotic region. For instance, a rapid fall of TcPO2 on the forearm is observed following a cuff occlusion of the proximal part of the arm with suprasystolic pressure, which indicates that an anoxic state takes place locally within a few minutes after the cuff inflation (Figure 47.2). The fall rate is regarded as the oxygen consumption rate of the skin.6,7 Figure 47.3 shows the expected profile of PO2 in the heated normal skin at 43˚C during the steady state.8 The PO2 in capillaries is higher than PaO2 owing to the temperature effect; then, being consumed by epidermal cells, it decreases gradually through the epidermis. The diffusion resistance of the
tcPco2
50
0
10
20
30 min.
40
50
FIGURE 47.2 Simultaneous recording of TcPCO2 and of TcPCO2 monitored on the volar side of the forearm of an adult. Sensor temperature: 44˚C. About 25 min was required to achieve the steady state of TcPO2 after attachment of the sensor (first arrow). Suprasystolic pressure was maintained during cuff occlusion of the arm for 10 min.
stratum corneum also contributes to the decrease in PO2, since the Clark-type electrode itself consumes oxygen and produces a diffusion field in front of it.8,9 It finally becomes close to the PaO2 value at the skin surface, and this value is the TcPO2 discussed here. If measured in adults,
Measurement of Transcutaneous Oxygen Tension
Electrode
Po2 1
100
4 Pco2
50
mmHg
2 3
399
0
Electrode membrane
Stratum corneum
Viable part
Epidermis
Papillary dermis
FIGURE 47.3 PO2 and PCO2 profiles of the skin during heatinduced hyperemia: (1) temperature effect, (2) oxygen consumption, (3) diffusion error by oxygen consumption of electrode, (4) CO2 production. (Modified from Lübbers, D.W., Birth Defects Orig. Article Ser., 15, 13, 1979.)
however, TcPO2 is usually lower than PaO2, which is explained by insufficient hyperemia, higher diffusion resistance of the stratum corneum, and thicker epidermis of adult skin.
47.3 SOURCES OF ERROR The most serious error might be a wrong interpretation of a TcPO2 value, especially when it is obtained in adults. For example, significantly lower TcPO2 values are found on the facial and glabrous skin of adults than on other regions.10,11 This is not due to the decrease in blood flow in these areas, but most likely is explained by increasing oxygen consumption of the facial skin where pilosebaceous units are dense and highly developed, and by the low diffusibility of oxygen through the thick stratum corneum of the glabrous skin.11 Many kinds of skin lesions, for example, inflammatory skin diseases, also show low TcPO2 irrespective of their pathological differences.12 Even if a therapeutic agent increases TcPO2 on the lesion of leg ulcer, no one can safely say that this drug improved cutaneous circulation because it might have improved only inflammation that existed there. Therefore, the rigid conception that TcPO2 is a parameter determined only by PaO2 and cutaneous perfusion might consequently be misleading. Since little skill is required for the TcPO2 measurement, technical errors seldom occur as long as an examiner handles the instrument properly according to the operation manual. Negligence of periodic exchange of electrolyte solution and a membrane of the sensor can be a source of errors. If one forgot calibration of the sensor prior to actual measurement, the data obtained are not reliable at all. Extraordinary TcPO2 values higher than PaO2 suggest an
influx of ambient air due to incomplete attachment of the sensor to the skin surface. The posture of subjects, especially when the sensor is placed on the extremities, also affects TcPO2,13 so that a standardized position is desirable during the measurement. If the sensor is sited over bony prominences like ribs, sudden intermittent falls might take place during the monitoring, probably due to the pressure on the skin electrode.14 When normal control subjects are needed, heavy smokers should be excluded because a significant difference in TcPO2 is found between smokers and nonsmokers.15 It is reported that the reproducibility of TcPO2 was not necessarily good when values were compared on the right and left lower extremities of normal adults,16 and when compared at adjacent sites on the dorsal foot of patients with arterial occlusive diseases, although the short-term (24 to 48 h) reproducibility of TcPO2 was relatively good.17 Although the temperature of the test site is kept constant by the thermostatic probe, it is preferable to avoid the examination in a too hot or too cold room. Physical or mental stress on a subject before or during the examination also should be avoided, since it may affect the cutaneous blood flow or the respiration rate.
47.4 CORRELATION WITH OTHER METHODS Close correlations are reported between TcPO2 at a sensor temperature of 44˚C and PaO2 at 37˚C in neonates, except in the high PaO2 range.18,19 A good correlation was also found in Japanese children of school age, with the exception of data for patients with atopic dermatitis.20 In adults, however, a rather poor correlation is obtained, with far lower TcPO2 values in the forearm relative to PaO2 values. According to a multicenter study reported by Palmisano and Severinghaus,19 which used a large number of patients with a wide age range, the TcPO2 is useful only to indicate changes of PaO2 in an individual if he or she is older than neonates. The correlation becomes even poorer when PO2 is greater than about 80 mmHg (Figure 47.4). The relationship between TcPO2 and PO2 of capillary blood from the ear lobe is relatively good in adults, but not so clearcut.21 The PaO2-TcPO2 gradient is unrelated to skinfold thickness, body surface area, and body mass indices determined by body weight and by height in adults.22 Transcutaneous PO2 is a parameter that is especially influenced by the cutaneous blood flow (CBF) at the test site. Since CBF depends on the heat supplied by the sensor, TcPO2 is positively correlated to both sensor temperature and CBF.23,24 At a constant sensor temperature of 43˚C, a positive correlation is reported between the change in TcPO2 and that in CBF during orthostatic changes of leg.13 However, the correlation becomes negative if the
400
Handbook of Non-Invasive Methods and the Skin, Second Edition
mmHg
TABLE 47.1 TcPO2 and TcPCO2 (mmHg, mean ± SD) at Eight Body Sites of 10 Adult Males
Id en tit y
160
120
tcPo2
Site 80
40
40
80
120 Pao2
160
200 mmHg
FIGURE 47.4 Relationship between TcPO2 and PaO2 of 723 samples from 251 patients with wide age range. Sensor temperature is 44˚C. Regression lines are shown for PaO2 below and above 80 mmHg. (Redrawn from Palmisano, B.W. and Siveringhaus, J.W., J. Clin. Monit., 6, 189, 1990. With permission.) 133
CBF is measured by the Xe washout technique under unheated conditions, indicating that heating abolishes autoregulation of CBF following orthostatic changes.25 Therefore, the relationship between TcPO2 and CBF differs depending on whether the CBF is measured in the heated condition or in the unheated condition.
47.5 CLINICAL AND EXPERIMENTAL APPLICATIONS Besides practical use as a noninvasive PaO2 monitoring device for sick neonates, measurement of TcPO2 has been utilized for various kinds of fundamental and clinical studies as follows.
47.5.1 FUNDAMENTAL STUDY In healthy adults, mean TcPO2 values are similar at various anatomical sites, with the exception of the face, palm, and probably the sole (Table 47.1).10,11 Women show higher values than men.11,26 The difference in the mean TcPO2 between the cheek and forearm is very small in children, indicating that the low TcPO2 value in the adult face is based on the anatomical characteristics of adult facial skin, i.e., probably related to the well-developed pilosebaceous units.11 Transcutaneous PO2 values usually increase with cutaneous warming from 37 to 44˚C, but they paradoxically decrease at the plantar skin.27 The oxygen consumption rate of the skin, which was found to be age dependent,36 was estimated to be about 0.18 to 0.37 ml O2/100 g/min at 43 to 45˚C by some authors7,28–30 from the reduction rate of TcPO2 during arterial occlusion. This consumption rate on skin with a thick
Forehead Cheek Forearm, volar Abdomen Back Crus, anterior Crus, posterior Palm
TcPO2 26.6 29.6 69.6 63.8 60.6 66.6 67.3 26.4
± ± ± ± ± ± ± ±
21.0 9.8 5.3 6.4 12.2 6.1 6.8 6.6
TcPCO2 67.2 69.2 62.6 63.7 63.9 61.1 62.8 60.5
± ± ± ± ± ± ± ±
2.4 2.8 3.3 2.4 3.1 4.1 2.9 3.1
Note: Probe temperature: 44˚C. At 37˚C, TcPO2 at forearm, 8 ± 4; TcPCO2, 43 ± 6 (Haisjackl et al.68). The approximate values of PO2 and PCO2 (mmHg) in the airway and blood are as follows:72
Tracheal air Alveolar gas Arterial blood Venous blood
PO2
PCO2
149.2 104 100 40
0.3 40 40 46
From Takiwaki, H. et al., Br. J. Dermatol., 125, 243, 1991. With permission.
stratum corneum, such as the palm, markedly increases after removal of the stratum corneum (Figure 47.5).11 A mathematical simulation using a simplified one-dimensional model can be applied for the explanation (Figure 47.6).31 Therefore, if the real oxygen consumption rate of hyperkeratotic skin is required, the thick stratum corneum should be removed. In addition, TcPO2 values increase after the removal of the stratum corneum by average values of 8 to 30 mmHg,11,29,32 depending on the test region. Therefore, we should always be conscious of the interference of the stratum corneum with the diffusion of oxygen when measuring TcPO2 and examining the dynamics of its changes. In addition, the thickening of the viable layer of the epidermis also reduces TcPO2 by the increase in oxygen consumption, and the change of PO2 across the layer (from the granular layer to the basal layer) is estimated at 0.26 mmHg/μm.33
47.5.2 CLINICAL APPLICATION Transcutaneous PO2 measurement is reasonably applied for quantitative estimation of the efficient cutaneous circulation and for the evaluation of the severity of leg ulcers, since TcPO2 reflects the actual efficiency of oxygen supply by tissue perfusion and since oxygen availability is considered a limiting factor in wound healing. In arterial occlusive diseases, including diabetic ischemic ulcers,
Measurement of Transcutaneous Oxygen Tension
50 tcPO2 = 37 mmHg
60 40 20 0
Off
0
Fall 40 mmHg/min 100
Cuff on 50 Off
0 (b)
FIGURE 47.5 Actual records of the TcPO2 fall during arterial occlusion before (A) and after (B) removal of the stratum corneum of palmar skin. The occlusion was started at the point of “Cuff on” and released at “off.” (From Takiwaki, H., Acta Derm. Venereol. (Stockh.), 185 (Suppl.), 21, 1994. With permission.)
several authors have reported that TcPO2 values of the lower leg or foot can be a good predictor of failure in wound healing after treatments or a method to determine amputation level.34–38 Its validity remains controversial,39 however, and interobserver and intraobserver reproducibility of TcPO2 measurement is not always good.40 There are discrepancies between the results of TcPO2 measurements when used for evaluating venous ulcers41–46 because the extent of the decrease in TcPO2 differs considerably from report to report. Although pericapillary fibrin deposits resulting from persistent venous hypertension may account for the reduction in TcPO2,43,44 this does not appear to be the only reason for the low TcPO2 value in venous ulcers. Differences in capillary density47 and the presence of lipodermatosclerosis also affect TcPO2.48 In addition, Dodd et al.49 reported that mean TcPO2 was higher in patients than in controls in the recumbent position when the sensor temperature was set at 37˚C. Another utilization of TcPO2 is done by plastic surgeons50–52 for intra- and postoperative management of flaps and for determination of the optimal time for cutting pedicle flaps. TcPO2 measurements have also been utilized
4
60 40 20 0
60
40
40
20
20 1
2 3 min.
1
4
0
2 3 min.
4
0.3 mm
80
60
0
tcPO2 = 78 mmHg
2 3 min.
0.02 mm
80
(a)
1 min.
1
0.1 mm
80 tcPo2(mmHg)
Fall 9 mmHg/min
Cuff on
tcPO2 (mmHg)
Thickness of the horny layer 0 mm
80
1 min.
tcPo2(mmHg)
tcPO2 (mmHg)
100
401
1
2 3 min.
4
FIGURE 47.6 Calculated TcPO2–time curves during arterial occlusion using a one-dimensional simplified model, where the electrode consumes no oxygen and the O2 consumption rate in the epidermis is constant. The diffusion equation ∂PO2/∂t = D(∂2PO2/∂x2) is used for computation, where the diffusion coefficient of the stratum corneum (D) at 44˚C is estimated to be 1.4 × 10–6 cm2/sec from the data of skin O2 conductance. (From Takiwaki, H., Acta Derm. Venereol. (Stockh.), 185 (Suppl.), 21, 1994. With permission.)
for the prevention of pressure ulcers by assessing of the effect of position and mattress quality on microcirculation in the sacral or trochanteric areas.53,54 In the field concerning dermatology, sclerotic skin lesions such as systemic scleroderma (SSc),55–58 morphea,57 and hypertrophic scar55,71 have been especially examined, since there are few parameters available for the objective evaluation of sclerotic change. The results show significantly reduced TcPO2 at sensor temperatures of 40.5 to 44˚C in these lesions, except for one report57 on SSc. Moreover, some authors55,56,59 demonstrated a negative relationship between the severity of lesions and TcPO2 values. At 37˚C, however, no difference in TcPO2 was found between SSc and the control.55 The cause of reduced TcPO2 is most likely explained by the reduced diffusibility of oxygen through the capillary wall and through the sclerotic dermis,56,59 or by the poor response of the blood vessels to the heat stimulus.55,58 Low TcPO2 under heating is never specific to sclerotic lesions, but is common to various skin lesions, such as atopic dermatitis, psoriasis, discoid lupus erythematosus, cutaneous lymphomas, herpes, pemphigoid, etc.12,31 It is probably because common pathological changes observed in various skin lesions, such as hyperkeratosis, acanthosis, cellular infiltrate or proliferation, fibrosis, and vascular changes, are all expected to reduce TcPO2 from the
402
Handbook of Non-Invasive Methods and the Skin, Second Edition
Treated with sodium amytal
Po2(mmHg)
No treatment 500
0
500
0.5
0
0.5
Skin thickness(mm)
FIGURE 47.7 Cutaneous PO2 against skin thickness during exposure to pure oxygen (•) and air () in an airtight chamber controlled at 37˚C. The experiment was performed using rat skin cut in various thicknesses. PO2 was measured at the dermal side of skin samples, which were directly attached to the sensor surface. To block the cellular respiration, samples had been soaked in isotonic sodium amytal solution for 20 min. (From Takiwaki, H., Acta Derm. Venereol. (Stockh.), 185 (Suppl.), 21, 1994. With permission.)
theoretical point of view. Therefore, we should not use the TcPO2 value as a parameter representing a specified pathological or functional change in the skin unless the cause of the TcPO2 change is clarified. It is a marker only indicating the extent of overall interference of the skin on oxygen flux from capillaries to the skin surface. Nevertheless, it might be a sensitive, though nonspecific, marker indicating the severity of skin lesions if measured on the same lesion continually, since it becomes close to the normal value as the lesion improves clinically.12 Skin test reactions, such as tuberculin and patch tests, and their time courses have been evaluated by monitoring TcPO2 or TcPCO2 at the test sites.60,61 Because the TcPO2 and visual grading values at positive test sites show a negative correlation, this method can be used for quantitative analysis of the skin test responses. Unlike other noninvasive bioengineering techniques, however, it may be a weak point of this method that 10 to 20 min is required before obtaining a steady-state value. It is most likely that the severe edema around capillaries and increased oxygen consumption of infiltrating inflammatory cells result in a relative decrease in local blood flow and low TcPO2.60,61 Although some authors have speculated that the decrease in TcPO2 results from the inflammation-induced increase in the distance between the capillary lumen and the skin surface, the results of several studies indicate that respiration of cells seems to be a far more important factor than the distance (Figure 47.7).31,62
47.5.3 ANALYSIS TCPO2
OF
DYNAMICS
OF
CHANGE
IN
Monitoring the dynamic response of TcPO2 to some loads is more advantageous for the investigation of the functional state of the microcirculation and the respiration
of the skin. Hyperkeratotic body sites should be avoided for this purpose. As already mentioned, the oxygen consumption rate of the skin can be estimated by the fall rate of TcPO2 during transient arterial occlusion using a pressure cuff. It should be kept in mind that this fall rate is smaller during breathing in air than in 100% oxygen, since it is influenced by the nonlinear relationship between PO2 and the oxygen content of blood. Although this fall rate (mmHg/min) is converted to the consumption rate (ml/100 g/min) by multiplying it by the oxygen solubility coefficient (ml O2/ml water/atm at given temperature) in the high TcPO2 range, where hemoglobin is fully saturated, it becomes necessary to consider the oxygen fraction bound to hemoglobin when TcPO2 is relatively low during air breathing. Thus, the oxygen consumption rate is usually estimated from data obtained during pure oxygen inhalation. It is determined not only by the respiration of epidermal cells, but also by that of inflammatory cells infiltrating the dermis. A few minutes of arterial occlusion usually leads TcPO2 to its minimum level close to zero, and the release of the cuff restores TcPO2 exponentially to the previous level. The recovering time of TcPO2 is regarded as a parameter reflecting the time required to refill tissue with blood flow, though influenced by the resistance of the skin to oxygen diffusion, and is found to be delayed in peripheral arterial occlusive disease.16,34,63,64 The maximum level of TcPO2 during oxygen inhalation becomes lower in ischemic limbs with arterial occlusive disease and diabetic patients.36,65 If the monitoring is carried out at a sensor temperature of 37˚C, the physiological vasoreactions, such as reflex vasoconstrictor response to posture change, vasodilation response to exercise, and reactive hyperemia after arterial occlusion, are reflected in the fast response of TcPO2 following the stimuli or loads mentioned.25,66–70 The sympathetic autoregulation of blood flow to thermal change is also detectable by examining the response time of TcPO2 to a sudden change in the sensor temperature from 45 to 37˚C.71 These physiological vascular responses are never obtained under conditions of heatinduced maximal hyperemia at 44˚C, where every kind of vasomotion is inhibited. Since oxygen consumption and diffusibility of the skin do not vary abruptly in a test site, TcPO2 changes in these examinations result from the change in blood flow or perfusion of the skin. Therefore, these examinations are valuable for the evaluation of weakened or absent vascular responses or reflexes that are seen in diabetic patients,69,70 in patients with severe venous incompetence,25,66 and in critically ill patients.68 By lowering the sensor temperature, however, the TcPO2 value is so markedly reduced that any changes of it need to be interpreted with caution.44
Measurement of Transcutaneous Oxygen Tension
47.6 RECOMMENDATION It is especially recommended to measure not only the resting TcPO2 value, but also its dynamic responses to some stimuli, and to measure other parameters such as TcPCO2 and CBF simultaneously for more precise investigation of the respirocirculatory function of the skin, because the TcPO2 value is a complex parameter influenced by many factors. Transcutaneous PCO2, which is measured on the heated skin surface just like TcPO2, is far less influenced by cutaneous factors than TcPO2,11 since CO2 is readily eliminated owing to far more solubility in water than oxygen, and since the stratum corneum does not affect the TcPCO2 value because the TcPCO2 sensor (pH-sensitive glass electrode) does not consume CO2. Moreover, the PCO2 can be kept stable by the buffering effect of the blood. However, arterial occlusion induces gradual but continuous increase in TcPCO2 (Figure 47.2). The extraordinarily high TcPCO2 value at TcPO2 close to zero therefore implies a severe disturbance of gas exchange in the skin. This finding is observed on blisters or severely edematous erythema, as seen in bullous pemphigoid and on prenecrotic skin of patients with arterial occlusive disease or necrotizing fasciitis.12 Since these lesions frequently fall into necrosis of the epidermis or dermis, simultaneous measurement of TcPCO2 provides more reliable information to predict necrotic change of the skin than TcPO2 measurement alone. The measurement of CBF is also preferable in order to get pure information on cutaneous microcirculation at the test site. Since some of the commercially available laser Doppler flowmeters have a thermostat probe holder by which the skin can be heated to a constant temperature, they are especially convenient for the assessment of blood flow under heated conditions corresponding to the circumstance of TcPO2 measurement. Laser Doppler flowmetry does not necessarily measure the perfusion involved in gas exchange since it appears to measure blood flow in arteriovenous anastomoses that does not participate in nutrition of the tissue, but it is not influenced by cellular respiration and oxygen diffusibility. In contrast, TcPO2 is more or less influenced by these factors, but skin perfusion affecting TcPO2 is concerned directly with gas exchange. Consequently, simultaneous measurements of both most likely give us more precise information on the microcirculation of the skin.
REFERENCES 1. Shaw, L.A. and Messer, A., Cutaneous respiration in man. III.The permeability of skin to carbon dioxide and oxygen as affected by altering their tension in the air, Am. J. Physiol., 98, 93, 1931.
403
2. Evans, N.T.S. and Naylor, P.F.D., The systemic oxygen supply to the surface of human skin, Respir. Physiol., 3, 21, 1967. 3. Huch, R., Lübbers, D.W., and Huch, A., Quantitative continuous measurement of partial oxygen pressure on the skin of adults and newborn babies, Pflügers Arch., 337, 185, 1972. 4. Huch, A., Huch, R., Arner, B., and Rooth, G., Continuous transcutaneous oxygen measured with a heat electrode, Scand. J. Clin. Lab. Invest., 31, 269, 1973. 5. Lübbers, D.W. and Grossmann, U., Gas exchange through the human epidermis as a basis of TcPO2 and TcPCO2 measurement, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 1. 6. Naylor, P.F.D. and Evans, N.T.S., The action of locally applied barbiturates on skin oxygen tension and rate of oxygen utilization, Br. J. Dermatol., 82, 600, 1970. 7. Severinghaus, J.W., Stafford, M., and Thunstrom, A.M., Estimation of skin metabolism and blood flow with tcPo2 and tcPco2 electrodes by cuff occlusion of the circulation, Acta Anaesth. Scand., 68 (Suppl.), 9, 1978. 8. Lübbers, D.W., Cutaneous and transcutaneous Po and Pco and their measuring conditions, Birth Defects Orig. Article Ser., 15, 13, 1979. 9. Baumgärtl, H., Grunewald, W., and Lübbers, D.W., Polarographic determination of the oxygen partial pressure field by Pt microelectrodes using the O2 field in front of a Pt electrode as a model, Pflügers Arch., 347, 49, 1974. 10. Orenstein, A., Mazkereth, R., and Tsur, H., Mapping of the human body skin with the transcutaneous oxygen pressure method, Ann. Plast. Surg., 20, 419, 1988. 11. Takiwaki, H., Nakanishi, H., Shono, Y., and Arase, S., The influence of cutaneous factors on the transcutaneous pO2 and pCO2 at various body sites, Br. J. Dermatol., 125, 243, 1991. 12. Takiwaki, H., Arase, S., Nakanishi, H., and Takeda, K., Transcutaneous Po2 and Pco2 measurements in various skin lesions, J. Dermatol., 18, 311, 1991. 13. Eickhoff, J.H. and Jacobsen, E., Correlation of transcutaneous oxygen tension to blood flow in heated skin, Scand. J. Clin. Lab. Invest., 40, 761, 1980. 14. Whitehead, M., Pollitzer, M., and Reynolds, E.O.R., Artefactal hypoxemia during estimation of Pao2 by skin electrode, Lancet, ii, 157, 1979 (letter). 15. Workman, W.T. and Sceffield, P.J., Continuous transcutaneous oxygen monitoring in smokers under normobaric and hyperbaric oxygen conditions, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 649. 16. Lusiani, L., Visona, A., Nicolin, P., Papesso, B., and Pagnan, A., Transcutaneous oxygen tension (TcPo2) measurement as a diagnostic tool in patients with peripheral vascular disease, Angiology, 39, 873, 1988. 17. Rooke, T. and Osmundson, P.J., Variability and reproducibility of transcutaneous oxygen tension measurements in the assessment of peripheral vascular disease, Angiology, 40, 695, 1989.
404
Handbook of Non-Invasive Methods and the Skin, Second Edition
18. Nilsson, E., Redham, I., Lagercrantz, H., and Olsson, P., The validity of transcutaneous Po2 monitoring (tcPo2) as compared to intraarterial Po monitoring in newborn infants, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 259. 19. Palmisano, B.W. and Severinghaus, J.W., Transcutaneous Pco2 and Po2: a multicenter study of accuracy, J. Clin. Monit., 6, 189, 1990. 20. Oda, Y., Sano, Y., Sato, M., Ishizuka, T., Adachi, K., and Suzuki, H., Correlation between arterial Po2 and transcutaneous Po2 in children of school age, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 281. 21. Wimberley, P.D., Pedersen, K.G., Thode, J., FoghAnderson, N., Sørensen, A.M., and Siggaard-Andersen, O., Transcutaneous and capillary pCO2 and pO2 measurements in healthy adults, Clin. Chem., 29, 1471, 1983. 22. Rafferty, T.D. and Morrero, O., Skin-fold thickness, body mass, and obesity indexes and the arterial to skinsurface Po2 gradient, Arch. Surg., 118, 1142, 1983. 23. Eberhard, P., Mindt, W., and Schäfer, R., Reliability of methods for skin blood flow measurement during transcutaneous Po2 monitoring, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 69. 24. Enkema, L., Holloway, G.A., Piraino, D.W., Harry, D., Zick, G.L., and Kenny, M.A., Laser Doppler velocimetry vs. heater power as indicators of skin perfusion during transcutaneous O2 monitoring, Clin. Chem., 27, 391, 1981. 25. Dodd, H.J., Gaylarde, P.M., and Sharkany, I., Skin oxygen tension in venous insufficiency of the lower leg, J. R. Soc. Med., 78, 373, 1985. 26. Monterio, L., Contreiras, P., and Leal, A., Transcutaneous flow related variables measured in vivo: the effects of gender, BMC Dermatol (Epub), 1, 4, 2001. 27. Smith, D.G., Boyko, E.J., Ahroni, J.H., Stensel, V.L., Davigon, D.R., and Pecoraro, R.E., Paradoxical transcutaneous oxygen response to cutaneous warming on the plantar foot surface: a caution for interpretation of plantar foot tcPO2 measurements, Foot Ankle Int, 16, 787, 1995. 28. Evans, N.T.S. and Naylor, P.F.D., The oxygen tension gradient across human epidermis, Respir. Physiol., 3, 38, 1967. 29. Jaszczak, P. and Sejrsen, P., Oxygen tension and consumption measured by a tcPo2 electrode on heated skin before and after epidermal stripping, Acta Anaesthesiol. Scand., 31, 362, 1987. 30. Horio, H., Tamura, H., Hasegawa, T., and Ohminato, S., Oxygen transport model for transcutaneous Po2 measurement, Med. Electr. Bioeng., 22, 31, 1984 (in Japanese). 31. Takiwaki, H., Transcutaneous Po2 and Pco2 measuring in dermatology, Acta Derm. Venereol. (Stockh.), 185 (Suppl.), 21, 1994.
32. Fallenstein, F., Nef, W., Huch, A., and Huch, R., Effect of skin stripping on the level and variability of transcutaneous Po2, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 161. 33. Falstie-Jensen, N., Spaun, E., Brøchner-Mortensen, J., and Falstie-Jensen, S., The influence of epidermal thickness on transcutaneous oxygen pressure measurements in normal persons, Scand. J. Clin. Lab. Invest., 48, 519, 1988. 34. Franzeck, U.K., Talke, P., Bernstein, E.F., Golbranson, F.L., and Fronek, A., Transcutaneous Po2 measurements in health and peripheral arterial occlusive disease, Surgery, 91, 156, 1982. 35. Christensen, K.S. and Klarke, M., Transcutaneous oxygen measurement in peripheral occlusive disease. An indicator of wound healing in leg amputation, J. Bone Joint Surg., 68-B, 423, 1986. 36. Bongard, O. and Krahenbuhl, B., Predicting amputation in severe ischemia: the value of transcutaneous Po2 measurement, J. Bone Joint Surg., 70-B, 465, 1988. 37. Bunt T.J. and Holloway, A., TcPO2 as an accurate predictor of therapy in limb salvage, Ann. Vasc. Surg., 10, 224, 1996. 38. Kalani, M., Brismar, K., Fagrell, B., Ostergren, J., and Jorneskog, G., Transcutaneous oxygen tension and toe blood pressure as predictors for outcome of diabetic foot ulcers, Diabetes Care, 22, 147, 1999. 39. Falstie-Jensen, N., Christensen, K.S., and BrøchnerMortensen, J., Selection of lower limb amputation level not aided by transcutaneous pO2 measurements, Acta Orthop. Scand., 60, 483, 1989. 40. Graaff, J.C., Ubbink, D.T., Legemate, D.A., Haan, R.J., and Jacobs, M.J.H.M., Interobserver and intraobserver reproducibility of peripheral blood and oxygen pressure measurements in the assessment of lower extremity arterial disease, J. Vasc. Surg., 33, 1033, 2001. 41. Nemeth, A.J., Eaglstein, W.H., and Falanga, V., Clinical parameters and transcutaneous oxygen measurements for the prognosis of venous ulcers, J. Am. Acad. Dermatol., 20, 186, 1989. 42. Mannarino, E., Pasqualini, L., Maragoni, G., Sanchini, R., Regni, O., and Innocente, S., Chronic venous incompetence and transcutaneous oxygen pressure: a controlled study, VASA J. Vasc. Dis., 17, 159, 1988. 43. Clyne, C.A.C., Ramsden, W.H., Chant, A.D.B., and Webster, J.H.H., Oxygen tension on the skin of the gaiter area of limbs with venous disease, Br. J. Surg., 72, 644, 1985. 44. Mani, R., Gorman, F.W., and White, J.E., Transcutaneous measurements of oxygen tension at edges of leg ulcers: preliminary communication, J. R. Soc. Med., 79, 650, 1986. 45. Cheatle, T.R., Mcmullin, G.M., Farrah, J., Smith, P.D.C., and Scurr, J.H., Three tests of microcirculatory function in the evaluation of treatment for chronic venous insufficiency, Phlebology, 5, 165, 1990. 46. Roberts, G., Hammand, L., Collins, C., Shearman, C., and Mani, R., Some effects of sustained compression on ulcerated tissues, Angiology, 53, 451, 2002.
Measurement of Transcutaneous Oxygen Tension
47. Franzeck, U.K., Bollinger, A., Huch, R., and Huch, A., Transcutaneous oxygen tension and capillary morphologic characteristics and density in patients with chronic venous incompetence, Circulation, 70, 806, 1984. 48. Roszinski, S. and Schmeller, W., Differences between intracutaneous and transcutaneous skin oxygen tension in chronic venous insufficiency, J. Cardiovasc. Surg., 36, 407, 1995. 49. Dodd, H.J., Gaylarde, P.M., and Sharkany, I., Skin oxygen tension in venous insufficiency of lower leg, J. R. Soc. Med., 78, 373, 1985. 50. Achauer, B.M. and Black, K.S., Transcutaneous oxygen and flaps, Plast. Reconstr. Surg., 74, 721, 1984. 51. Tuominen, H.P., Asko-Seljavaara, S., Svartling, N.E., and Härmä, A., Cutaneous blood flow in the TRAM flap, Br. J. Plast. Surg., 45, 261, 1992. 52. Nabawi, A., Gurlek, A., Patrick, C.W., Amin, A., Ritter, E., Elsharaky, M., and Evans, G.R., Measurement of blood flow and oxygen tension in adjacent tissues in pedicled and free flap head and neck reconstruction, Microsurgery, 19, 254, 1999. 53. Colin, D., Abraham, P., Preault, L., Bregeon, D., and Saumet, J.L., Comparison of 90 and 30 degrees laterally inclined positions in the prevention of pressure ulcers using transcutaneous oxygen and carbon dioxide pressures, Adv. Wound Care, 9, 35, 1996. 54. Colin, D., Loyant, R., Abraham, P., and Saumet, J.L., Changes in sacral transcutaneous oxygen tension in the evaluation of different mattresses in the prevention of pressure ulcers, Adv. Wound Care, 9, 25, 1996. 55. Takiwaki, H., Transcutaneous Po2 measurements on clinically sclerotic lesions in scleroderma and hypertrophic scar, Nishinihon J. Dermatol., 49, 492, 1987 (in Japanese). 56. Silverstein, J.L., Steen, V.D., Medsger, T.A., and Falanga, V., Cutaneous hypoxia in patients with systemic sclerosis, Arch. Dermatol., 124, 1379, 1988. 57. Kalis, B., De Rigal, J., Léonard, F., Lévêque, J.L., Riche, O., Le Corre, Y., and De Lacharriere, O., In vivo study of scleroderma by non-invasive techniques, Br. J. Dermatol., 122, 785, 1990. 58. Valentini, G., Leonardo, G., Moles, D.A., Apasia, M.R., Maselli, R., Tirri, G., and Del Guercio, R., Transcutaneous oxygen pressure in systemic sclerosis: evaluation at different sensor temperatures and relationship to skin perfusion, Arch. Dermatol. Res., 283, 285, 1991. 59. Berry, R.B., Tan, O.T., Cooke, E.D., Gaylarde, P.M., Bowcock, S.A., Lamberty, B.G.H., and Hackett, M.E.J., Transcutaneous oxygen tension as an index of maturity in hypertrophic scars treated by compression, Br. J. Plast. Surg., 38, 163, 1985.
405
60. Spence, V.A. and Swanson, J.B., Transcutaneous measurement of Po2 and Pco2 in the dermis at the site of the tuberculin reaction in healthy human subjects, J. Pathol., 155, 289, 1988. 61. Prens, E.P., van Joost, V., and Steketee, J., Quantification of patch test reactions by transcutaneous Po2 measurement, Contact Derm., 16, 142, 1987. 62. Carnochan, F.M.T., Abott, N.C., Beck, J.S., Spence, V.A., and James, P.B., The influence of histamine and PGE2-induced hyperaemia and oedema on respiratory metabolism in normal human forearm skin, Agents Actions, 29, 292, 1990. 63. Slagsvold, C.E., Stranden, E., Rosen, L., and Kroese, A.J., The role of blood perfusion and tissue oxygenation in the postischemic transcutaneous pO2 response, Angiology, 43, 155, 1992. 64. Diamantopoulos, E.J., Stavreas, N.P., Roussis, D.P., Charitos, D.N., Vasdekis, S.N., and Raptis, S.A., Simultaneous laser Doppler and transcutaneous oxygen tension measurements in claudicant patients, Int. Angiol., 14, 53, 1995. 65. Breuer, H.W.M., Breuer, J., and Berger, M., Transcutaneous oxygen pressure measurements in type I diabetic patients for early detection of functional diabetic microangiopathy, Eur. J. Clin. Invest., 18, 454, 1988. 66. Sarkany, I., Dodd, H.J., and Gaylarde, P.M., Surgical correction of venous incompetence restores normal skin blood flow and abolishes hypoxia during exercise, Arch. Dermatol., 125, 223, 1989. 67. Rooth, G., Ewald, U., and Caligara, F., Transcutaneous Po2 and Pco2 monitoring at 37˚C, Adv. Exp. Med. Biol., 220, 23, 1987. 68. Haisjackl, M., Hasibeder, W., Klaunzer, S., Alterberger, H., and Koller, W., Diminished reactive hyperemia in the skin of critically ill patients, Crit. Care Med., 18, 813, 1990. 69. Ewald, U., Tuvemo, T., and Rooth, G., Early reduction of vascular reactivity in diabetic children detected by transcutaneous oxygen electrode, Lancet, i, 1287, 1981. 70. Kobbah, M., Ewald, U., and Tuvemo, T., Vascular reactivity during the first year of diabetes in children, Acta Ped. Scand., Suppl. 320, 56, 1985. 71. Weindorf, N., Schultz-Ehrenburg, U., and Altmeyer, P., Plaqueförmige kutane Muzinose mit Teleangiektasien, Hautarzt, 39, 589, 1988. 72. Comroe, J.H., Physiology of Respiration, Year Book Medical Publishers, Chicago, 1974, p. 9.
48 Measurement of Transcutaneous P
CO2
Carsten N. Nickelsen Hvidovre Hospital, University of Copenhagen, Hvidovre, Denmark
CONTENTS 48.1 Introduction............................................................................................................................................................407 48.2 Object.....................................................................................................................................................................407 48.3 Methodological Principle ......................................................................................................................................407 48.3.1 The Electrochemical Electrode .................................................................................................................408 48.3.2 Electrode Calibration .................................................................................................................................409 48.3.3 Electrode Application to the Skin .............................................................................................................409 48.4 Sources of Error.....................................................................................................................................................409 48.5 Correlation with Other Methods ...........................................................................................................................409 48.6 Recommendations..................................................................................................................................................410 References .......................................................................................................................................................................410
48.1 INTRODUCTION In 1793 John Abernethy1 demonstrated for the first time carbon dioxide exchange through the intact skin by submerging his arm into mercury. He analyzed the gas bubbles accumulating above the mercury and found carbon dioxide. Sixty years later Gerlach2 demonstrated changes in the gas mixture of a horse bladder glued to his chest. However, transcutaneous monitoring of the carbon dioxide tension was not possible until the construction of an electrochemical CO2 electrode by Stow and Randall in 1954,3 improved by Severinghaus and Bradley in 1956,4 by Johns et al. in 1969,5 and modified by Huch et al. in 1973.6
48.2 OBJECT The acid–base balance is usually described by the pH, PCO2, and the base excess (BE). The pH is defined as the negative logarithm of the hydrogen activity and is the resultant of two factors, a respiratory component and a metabolic component. The respiratory component is expressed as the PCO2 in kPa (or in mmHg), being an index of dissolved CO2. The metabolic component consists of all other acids or bases, which affects the pH, and can be expressed as base excess in meq/l or mmol/l. According to these parameters, different clinical changes of human acid–base balance are defined. Whereas respiratory acidosis is characterized by accumulation of carbon dioxide in the organism, with elevated PCO2, decreased pH, and unchanged BE values, metabolic
acidosis is characterized by accumulation of noncarbonic acid in the organism, with decreased pH and BE and primarily unchanged PCO2 values. During respiratory alkalosis and metabolic alkalosis the changes are primarily in the opposite direction. These primary changes are followed by both chemical and physiological compensation in order to stabilize the pH. The compensation may be respiratory caused by hyperventilation or renal by means of increased excretion of bicarbonate. The chemical compensation is caused by the buffering system of the blood and the extracellular fluid. The major buffering system in the organism is the bicarbonate based in the changes in the equation CO2 + H2O ↔ H2CO3 ↔ H+ + HCO–3 The same chemical reaction occurs when blood with low pH flows into the skin, leading to a PCO2 increase in the skin; for instance, when metabolic acidosis is produced in another tissue during exercise, hypoperfusion, or in ketoacidosis, there may be an elevation of the carbon dioxide tension in the skin. Therefore, the PCO2, being mainly an indicator for respiratory disturbances, when measured in the skin can also reflect metabolic disturbances.7
48.3 METHODOLOGICAL PRINCIPLE The electrode for transcutaneous carbon dioxide monitoring measures the partial pressure of CO2 in a small lumen outside the skin. The lumen is usually filled with a contact 407
408
Handbook of Non-Invasive Methods and the Skin, Second Edition
liquid, an electrolyte solution, which very rapidly equilibrates with the carbon dioxide tension in the outer layer of the epidermis. The measured value is a reflection of the capillary blood PCO2 value, but it is modified by the capillary blood flow, the temperature, and the production of CO2 in the tissue between the capillaries and the electrode surface. During normal conditions the PCO2 of arterial blood is close to the PCO2 of venous blood, the difference only being 1 kPa (7.5 mmHg).8 At increasing capillary blood flow the capillary blood PCO2 is decreasing toward arterial values, but at decreasing capillary blood flow the local capillary blood PCO2 value may increase to much higher values, and even to values higher than central venous blood values. Therefore, a capillary blood flow above a certain limit is necessary in order to obtain transcutaneous values correlating to the arterial values. In order to obtain a satisfactory capillary blood flow, vasodilatation usually is induced using heated transcutaneous electrodes. Beside vasodilatation and a faster diffusion, heating of the electrode and the tissue below it will also induce a temperature-dependent elevation of the measured PCO2 value, just as PCO2 values increase with temperature in blood samples.9 In blood samples this temperature-dependent elevation is described by the anaerobic temperature coefficient of PCO2 in blood, a. When calculating the value at 37˚C [PCO2(37)] from the measured value at t˚C [PCO2(t)] the following formula is used:
PCO2 (37) = PCO2 (t )
1 10 α ( t –37)
The anaerobic temperature coefficient of CO2 in blood (a) is 0.021˚C–1.9 According to that, the correction factor is calculated for different temperatures in Table 48.1. Besides correction for the elevated electrode temperature, the transcutaneous value must be corrected for the CO2 production in the tissue between the capillaries and the electrode on the skin surface in order to correlate it to the capillary blood value of PCO2. This metabolic contribution to the transcutaneous value has been found to be approximately 0.5 kPa (3 to 4 mmHg).10,11 The metabolic contribution may be changing when different electrode temperatures are used as the rate of metabolism increases with the temperature, but the increasing diffusion speed of CO2 at increasing temperature will balance the change to some extent, making the metabolic contribution to the transcutaneous carbon dioxide value almost temperature independent. Theoretically, the transcutaneous value of P CO 2 (tcPCO2) can be calculated as the product of the capillary PCO2 (measured at 37˚C) and the temperature factor (Table 48.1) added to the metabolic contribution (0.5 kPa), but a
TABLE 48.1 Temperature Correction Factor for PCO2, Measured at Different Temperatures t°C 39
40
41
Correction from t° to 37° 0.9078 8.8650 0.8241 Correction from 37° to t° 1.1015 1.1561 1.2134
42
43
44
45
0.7852
0.7482
0.7129
0.6742
1.2735
1.3366
1.4028
1.4723
sufficient capillary blood flow is required for the use of this calculation. Five different principles of transducers have been used for monitoring of tcPCO2,11 but the use of electrochemical electrodes has been predominant.
48.3.1 THE ELECTROCHEMICAL ELECTRODE The electrochemical electrode used for tcPCO2 monitoring is a thermostated Stow–Severinghaus electrode. This electrode consists of a glass pH electrode, a silver–silver chloride reference electrode placed around the pH electrode, and a thermostated heating element. A CO2-permeable membrane covers the electrodes, occluding an electrolyte chamber between the membrane and the electrodes. An O-ring secures the membrane to the electrode housing (Figure 48.1). During monitoring the CO2 released from the skin diffuses through a contact fluid through the gas-permeable membrane and into the electrolyte solution in the electrolyte chamber. The carbon dioxide reacts with water to
9
1
2
8 3 4
5 6
7
FIGURE 48.1 Electrochemical transcutaneous carbon dioxide electrode (E 5230, Radiometer Copenhagen). (1) Temperature sensor, (2) O-ring securing the membrane, (3) CO2-permeable membrane, (4) electrolyte, (5) inner buffer solution, (6) inner reference Ag/AgCl electrode, (7) pH-sensitive glass membrane, (8) outer reference Ag/AgCl electrode, (9) heating element.
Measurement of Transcutaneous PCO2
409
form carbonic acid (H2CO3), which immediately dissociates to hydrogen ions (H+) and bicarbonate ions (HCO3–). Following the diffusion of CO2 into the electrolyte, the pH of the electrolyte will change according to the Henderson–Hasselbalch equation:
pH = pK + log
cHCO 3– a PCO2
where pK is the dissociation constant of carbonic acid, cHCO3– the concentration of HCO3–, a the solubility coefficient of dissolved CO2, and PCO2 the partial pressure of CO2. The pH change (as a result of the PCO2 change) in the electrolyte is measured as a change in the potential between the glass electrode and the reference electrode. The monitor therefore in principle is a pH meter, but the display is usually changed and the value displayed is the PCO2 value. In the monitor a microprocessor is usually integrated controlling the heating of the electrode, and often also performing the calibration of the electrode more or less automatically. In some cases the microprocessor modifies the measured value for the elevated electrode temperature or the metabolic contribution before displaying the result.
48.3.2 ELECTRODE CALIBRATION The calibration procedure is usually recommended as a two-step calibration with two different gas mixtures (usually 5%/95% and 10%/90% CO2/N2 gas mixtures). Using this procedure the electrode and the monitor are adjusted (with the first gas mixture) and the actual sensitivity of the electrode is calculated (with the second gas mixture). The sensitivity of an electrode is, however, unchanged during most of its lifetime, and it is possible to perform a one-step calibration, in which a preset standard electrode sensitivity is used during monitoring. The possible measurement inaccuracy following one-step calibration is below 0.2 kPa when measuring in the range from 0 to 10 kPa (below 2%), and this was found acceptable.13 Nevertheless, it is recommended to check the electrode sensitivity now and then by exposing the electrode to two different calibration gas mixtures, in order to avoid the use of old electrodes with low sensitivity. The one-point calibration is performed rapidly (approximately 5 to 10 minutes vs. 15 to 20 minutes by two-point calibration).
48.3.3 ELECTRODE APPLICATION
TO THE
SKIN
In the neonate and the adult the electrode application to the skin is performed using a fixation ring attached to the skin by adhesive tape. The fixation ring is attached to the skin, contact liquid is filled into the ring, and the electrode is locked into the fixation ring superseding some of the
contact liquid. In this way the electrode membrane is positioned in direct contact with the small liquid-filled lumen on the skin. The same method is not possible for fetal monitoring, primarily because the adhesive tape does not attach to the wet fetal skin, but also because it is not possible to fill the contact liquid into the fixation ring when it is attached to the presenting part of the fetus (usually the fetal scalp) in the vagina. A special fixation ring for suction fixation is used for fetal monitoring.13 The ring is made of soft plastic material and attached to the skin with a vacuum of 20 kPa. No contact liquid is used for fetal monitoring.
48.4 SOURCES OF ERROR Insufficient fixation of the electrode will usually cause false low values. When applying the electrode to the skin, an area where the skin is thin and where movements causing dislocation of the electrode are few should be chosen. In the adult and the neonate the electrode is usually applied on the chest; in the fetus the electrode can only be applied on the presenting part, usually the fetal scalp. Insufficient blood flow in the skin below the electrode may cause false elevated values, only slowly reacting to changes in the PCO2 values of blood in central blood values. The need for vasodilation is especially important in adult monitoring, but in healthy adults a short period of preheating to 45˚C permits sufficient vasodilatation for monitoring at 37˚C electrode temperature (in contrast to transcutaneous PO2 monitoring).14 However, in the critically ill patient a higher electrode temperature causing local vasodilatation is preferable. In the fetus, an electrode temperature of 44˚C should be used.15 The electrode must be placed in an area where mechanical pressure on the back of the electrode is avoided, as compression of the area under the electrode can cause false elevation of the PCO2 value.
48.5 CORRELATION WITH OTHER METHODS In adults close correlations are found between transcutaneous carbon dioxide tension and capillary blood PCO2.14 After correction for the elevated electrode temperature and the metabolic contribution from the skin, the values are almost identical. In the neonate a similar close correlation has been demonstrated even during respiratory insufficiency between transcutaneous PCO2 and arterial PCO2.16 During fetal monitoring transcutaneous PCO2 has been correlated both to capillary blood PCO2 values17 and to PCO2 values in umbilical artery blood,15 and close correlations were found in all cases.
410
Handbook of Non-Invasive Methods and the Skin, Second Edition
48.6 RECOMMENDATIONS Although most studies have concluded that transcutaneous carbon dioxide tension measurement is both useful and sensitive for continuous respiratory monitoring, there are some limitations for the method. Especially in adults it seems that there are individual differences in either skin permeability or central regulation of respiration,18 which may limit the use of the method for the assessment of respiratory depression in extubated, spontaneously breathing patients recovering from general anesthesia.19 The electrode temperature should be 42 to 43˚C for adult monitoring. Transcutaneous PCO2 monitoring is especially useful as a continuous measurement for respiratory depression in the neonate, where the correlation to the arterial PCO2 value is very close, and the need for repeated blood gas measurements on arterial or capillary blood is thus diminished. The electrode temperature can be set as low as 38˚C in the neonate, but as the electrode often is combined with a PO2 electrode, higher temperatures are used. In the fetus during labor, transcutaneous PCO2 monitoring gives valuable information on fetal distress, but the method cannot be used as the only method because of a low specificity. For fetal monitoring an electrode temperature of 44˚C is necessary for accurate measurements.15 In newborns with arterial–venous shunts or compromised peripheral blood flow, it is possible by use of two transcutaneous electrodes placed in different areas noninvasively to diagnose the hemodynamic change. Further, the method is increasingly used in a variety of clinical situations, such as assessment of skin flab viability during evaluation of wound healing and selection of amputation level in peripheral vascular diseases.
REFERENCES 1. Abernethy, J., An essay of the nature of the matter perspired and absorbed from the skin, in Surgical and Physiological Essays, Part 2, London, 1793, p. 107. 2. Gerlach, J.V., Über das Hautathmen, Arch. Anat. Physiol. Leipzig, 431, 1851. 3. Stow, R.W. and Randall, B.F., Electrical measurement of the pCO2 of blood, Am. J. Physiol., 179, 678, 1954. 4. Severinghaus, J.W. and Bradley, A.F., Electrodes for blood pO2 and pCO2 determination, J. Appl. Physiol., 13, 515, 1958. 5. Johns, R.J., Lindsay, W.J., and Shephard, R.H., A system for monitoring pulmonary ventilation, Biomed. Sci. Instrum., 5, 119, 1969. 6. Huch, A., Lübbers, A.W., and Huch, R., Patientenüberwachung durch transcutane pCO2 Messung bei gleichzeitiger Kontrolle der relative lokalen Perfusion, Anaesthesist, 22, 379, 1973.
7. Rooth, G., Ewald, U., and Caligara, F., Transcutaneous pO2 and pCO2 monitoring at 37˚C. Cutaneous pO2 and pCO2, in Continuous Transcutaneous Monitoring, Advances in Experimental Medicine and Biology, 220, Huch, A., Huch, R., and Rooth, G., Eds., Plenum Press, New York, 1986, p. 23. 8. Lübbers, D.W., Cutaneous and transcutaneous pO2 and pCO2 and their measuring conditions, Birth Defects, 15, 13, 1979. 9. Siggaard-Andersen, O., The Acid-Base Status of the Blood, 4th ed., Munksgaard, Copenhagen, 1974. 10. Hazinski, T.A. and Severinghaus, J.W., Transcutaneous analysis of arterial pCO2, Med. Instrum., 16, 150, 1982. 11. Severinghaus, J.W., Stafford, M., and Bradley, A.F., TcpCO2 electrode design, calibration and temperature gradient problems, Acta Anaesthesiol. Scand. Suppl., 68, 118, 1978. 12. Nickelsen, C., Thomsen, S.G., and Weber, T., Fetal carbon dioxide tension during human labour, Eur. J. Obstet. Gynecol. Reprod. Biol., 22, 205, 1986. 13. Nickelsen, C. and Weber, T., Suction fixation of the tcpCO2 electrode for fetal monitoring, J. Perionat. Med., 15, 383, 1987. 14. Wimberley, P.D., Pedersen, K.G., Olsson, J., and Siggaard-Andersen, O., Transcutaneous carbon dioxide and oxygen tension measured at different temperatures in healthy adults, Clin. Chem., 31, 1611, 1985. 15. Nickelsen, C., Monitoring of fetal carbon dioxide tension during human labour, Dan. Med. Bull., 36, 537, 1989. 16. Frederiksen, P.S., Wimberley, P.D., Melberg, S.G., WittHansen, J., and Friis-Hansen, B., Transcutaneous pCO2 at different temperatures in newborns with respiratory insufficiency: comparison with arterial pCO2, in Continuous Transcutaneous Blood Gas Monitoring, Huch, R. and Huch, A., Eds., Marcel Dekker, New York, 1983, p. 227. 17. Schmidt, S., Langner, K., Gesche, J., Dudenhausen, J.W., and Saling, E., Correlation between transcutaneous pCO2 and the corresponding values of fetal blood: a study at a measuring temperature of 39˚C, Eur. J. Obstet. Gynecol. Reprod. Biol., 17, 387, 1984. 18. Lehmann, K.A., Asoklis, S., Grond, S., and Huttarsch, H., Entwiklung eines kontinuierlichen Monitorings der Spontanatmung in der postoperativen Phase. I. Normalwertbereiche für kutane pO2- und pCO2-partialdrucke sowie pulsoxymetrisch bestimmte Sauerstoffsattigungen bei gesunden jungen Erwaachsenen, Anaesthesist, 41, 121, 1992. 19. Kavanagh, B.P., Sandler, A.N., Turner, K.E., Wick, V., and Lawson, S., Use of end-tital pCO2 and transcutaneous pCO2 as noninvasive measurement of arterial pCO2 in extubated patients recovering from general anesthesia, J. Clin. Monit., 8, 226, 1992.
Surface pH: Mechanism, 49 Skin Measurement, Importance Joachim Fluhr and Lora Bankova Department of Dermatology, Friedrich Schiller University, Jena, Germany
Shabtay Dikstein Unit of Cell Pharmacology, School of Pharmacy, The Hebrew University of Jerusalem, Jerusalem, Israel
CONTENTS 49.1 Introduction............................................................................................................................................................411 49.1.1 Phospholipid-to-Free Fatty Acid Pathway ................................................................................................412 49.1.2 Membrane Antiporters ...............................................................................................................................412 49.1.3 Histidine-to-Urocanic Acid Pathway.........................................................................................................412 49.2 Methodological Aspects ........................................................................................................................................413 49.2.1 Principle .....................................................................................................................................................413 49.2.2 Instrument ..................................................................................................................................................413 49.2.3 Method .......................................................................................................................................................413 49.2.4 Procedure ...................................................................................................................................................413 49.2.5 Sources of Error.........................................................................................................................................413 49.3 Influence of Exogenous and Endogenous Factors ................................................................................................413 49.3.1 Influence of External Environmental Factors ...........................................................................................413 49.3.2 Circadian Rhythm......................................................................................................................................414 49.3.3 Age Differences .........................................................................................................................................414 49.3.4 Gender and Racial Differences .................................................................................................................414 49.3.5 Anatomical Skin Areas for Testing ...........................................................................................................414 49.3.6 Influence of Tape Stripping .......................................................................................................................414 49.4 Importance of pH for the Barrier ..........................................................................................................................415 49.4.1 pH and Antimicrobial Function.................................................................................................................415 49.4.2 Changes in pH Could Influence Key SC Functions .................................................................................415 49.5 pH and Skin Disorders ..........................................................................................................................................416 49.5.1 Eczematous Diseases .................................................................................................................................416 49.5.2 Ichthyosis ...................................................................................................................................................416 49.5.3 Other Diseases ...........................................................................................................................................417 49.6 Summary ................................................................................................................................................................417 References .......................................................................................................................................................................417
49.1 INTRODUCTION Skin surface pH is a key parameter of the stratum corneum (SC) and an important regulator of the epidermal barrier homeostasis.1,2 The optimal stratum corneum pH is a prerequisite for the activation of the lipid hydrolases in the
cornified layer. It is responsible for the postsecretory processing of lamellar bodies, which is an important step in the formation of the cutaneous barrier.3 The skin surface has been known for more than one century.4 Furthermore, the upper stratum corneum exhibits pH values lower than the physiologic pH, characteristic
411
412
Handbook of Non-Invasive Methods and the Skin, Second Edition
Barrier homeostasis
Bilamellar structures
Enzyme activity
sPLA2 Phospholipids
FFA
Antimicrobial activity
pH
Serine proteases Na/H Proton pump CD Desintegration
Cohesion/ desquamation
FIGURE 49.1 Skin pH as a regulation parameter for barrier homeostasis, stratum corneum cohesion, and desquamation. (Adapted from Fluhr, J.W. and P.M. Elias, Exogenous Dermatol., 1: 163–175, 2002.)
of the body’s internal environment. Another characteristic feature of the cutaneous pH is its sharp rising gradient within the stratum corneum and further into the epidermis, reaching the body’s internal pH values at the level of the stratum granulosum. The acid mantle is now shown to be formed as a result of the interaction of a variety of endogenous and exogenous mechanisms (Figure 49.1). The different mechanisms described below lead to a new definition of the socalled acid mantle. Thus, we suggest the term acid stratum corneum buffer system. Exogenous factors affecting the stratum corneum acidification are sweat and sebum5 as well as hydrolytic products of filaggrin, microbial metabolites,6 and lactic acid from ecrine glands.7 Currently, three endogenous mechanisms have been identified that not only influence SC pH, but also regulate one or more key SC functions.
49.1.1 PHOSPHOLIPID-TO-FREE FATTY ACID PATHWAY A major pathway contributing to SC acidification is the secretory phospholipase A2 pathway. The product of this pathway is a pool of free fatty acids (FFA). Processing of phospholipids to free fatty acids by sPLA2 generates nonessential FFA, one of the three key families of SC barrier lipids. Applications of inhibitors of sPLA2 exert profound effects on barrier homeostasis and epidermal lipid processing.8,9 These inhibitors not only perturb permeability barrier homeostasis, but also elevate SC bulk pH, with additional functional alterations in SC integrity/cohesion.2,10
Pertinently, coapplications of an acidic buffer with the sPLA2 inhibitors corrected the integrity/cohesion abnormality, but not the critical requirement for FFA as structural components of the barrier. The phospholipid-to-free fatty acid pathway contributes directly to SC acidification, impacting both SC barrier function and integrity/cohesion (Figure 49.1).11
49.1.2 MEMBRANE ANTIPORTERS Non-energy-requiring, integral membrane transporters that extrude protons in exchange for sodium (NHE) are ubiquitous in mammalian tissues.12–16 Mammals possess a minimum of eight functional genes for NHE. The isoforms are expressed in a tissue/cell-specific manner and are localized differentially to discrete membrane compartments (cell surface or endomembrane organelles).14,15,17 The only one expressed in keratinocytes is the NHE1.16,18,19 It has been demonstrated that protons generated by NHE1 are an important source of SC acidity and play a crucial role in the acidification of the SC in the first postnatal days.20,21 It is upregulated under ambient conditions present in the outer epidermis, high calcium, high osmolarity, or acidity,12,13 i.e., could acidify extracellular membrane domains selectively near the SG-SC interface or in membrane domains in the lower SC.
49.1.3 HISTIDINE-TO-UROCANIC ACID PATHWAY The histidine-to-urocanic acid pathway is largely responsible for the SC hydration.22,23 Early in cornification, filaggrin, the predominant, histidine-enriched basic protein in keratohyalin granules, disperses around keratin filaments
Skin Surface pH: Mechanism, Measurement, Importance
413
within the stratum compactum.24 Fillagrin undergoes proteolysis to free amino acids (histidine, glutamine, and arginine). The succeeding enzymatic processes of nonoxidative histidine deamination have been identified to potentially contribute to the SC acidification.25–27 The key enzyme in this process is the histidine ammonia-lyase (histidase). It is found primarily in the liver and the skin.26,28–32 In the skin it is almost entirely located in the SC and functions optimally at a neutral pH.27,29–31 The main product of the histidine deamination is the transurocanic acid (tUCA). It has been shown to play an important role not only in skin hydration, but also in several non-pH-dependent skin-protective functions: the tUCA acts as an endogenous sunscreen and plays an important role in immunosuppression through its metabolite cis-UCA. Furthermore, cis-UCA is hypothesized to favor the development of skin cancer.26,27 Although a potent mechanism in SC acidification, there are certain indirect clues indicating that it does not suffice to explain the acidity of the skin surface.
whereby the liquid in the external electrode sheath should be higher than the level in the internal electrode. The electrode is stored in a covering also filled with KCl solution; in case of a prolonged dry storage, a minimum of 12 h in a KCl solution is required.38
49.2 METHODOLOGICAL ASPECTS 49.2.1 PRINCIPLE The skin surface pH is generally measured by a flat glass electrode at the skin surface with a hydrated skin–electrode interface. The symbol pH is used to indicate the concentration of hydrogen ion (H+) in a solution. More specifically, pH is defined as the negative logarithm of the hydrogen ion concentration in an aqueous solution. In pure water, at 25oC, the ionic product of water is defined as [H+] = [OH–] = 10–7 M, and pH = –log [H+] = 7. Proceeding from this equation, the pH scale is divided into 14 ranks, from 0 (acidic) to 14 (basic). Concerning skin bioengineering measurements, a specified definition of pH was introduced: pH as measured by flat glass electrode at the skin surface with a hydrated skin–electrode interface.
49.2.2 INSTRUMENT The most common method applied currently is the potentiometric assessment with a flat glass electrode. 33 Recently, more sophisticated high-resolution methods that focus on microdomains, such as fluorescence lifetime imaging microscopy (FLIM)20,34–36 and electron spin resonance imaging,37 have been introduced.
49.2.3 METHOD Prior to measurement, the flat electrode can be calibrated with standard buffer solutions with known and stable pH (calibration prior to initial application and subsequently every 5 to 15 days). The electrode has to be filled with electrolyte liquid (usually 3 mol/l solution of KCl),
49.2.4 PROCEDURE The measuring electrode is dipped in distilled water and then dried with a filter paper, leaving its surface wet, but avoiding an excessive amount of water. Subsequently, it is placed onto the selected skin area for a duration of about 10 sec with slight pressure.39
49.2.5 SOURCES
OF
ERROR
Measurement errors could be caused by surplus of water or dry electrode surface. Distilled water in the covering could influence the accuracy of measurements. Sweat can compromise the results, as well as contamination. Contact with hard objects should be avoided. Washing procedures have an alkalizing effect on the skin surface pH, independent of the pH of the washing solution. This effect can be observed after a single washing, and it usually normalizes within 8 h.40
49.3 INFLUENCE OF EXOGENOUS AND ENDOGENOUS FACTORS 49.3.1 INFLUENCE FACTORS
OF
EXTERNAL ENVIRONMENTAL
Washing procedures have an alkalizing effect on the skin surface pH, independent of the pH of the washing solution. This effect can be observed after a single washing, and usually normalizes within 8 h.40 Even short-term washing with tap water results in an increase in surface pH in infants.41 However, repetitive washing with sodium lauryl sulfate (SLS) showed a more pronounced alkalization effect than the single washing procedure.42–44 Furthermore, the more alkaline surface pH persists for more than 12 h in a repeated washing model.45 A crossover study with acidic and neutral cleanser showed an increase of propionibacteria in alkaline cleanser-treated skin, while acidic cleanser reduced the propionibacteria colonization.46 Thus, the use of a neutral or acidic cleanser is less harmful to the surface pH than alkaline soap. However, it remains unclear whether the increase in SC pH is due to a direct effect of the soaps on SC or if soaps add base equivalents to SC (e.g., OH–). An increase in surface pH can be observed both after 3 days of occlusion (e.g., from pH 4.9 to 7.1)47 and after 5 days (i.e., from 4.38 to 7.05).48 The pH increase was paralleled in both studies by an increase of bacterial colonization. A negative correlation has been shown between the sweat inhibition and skin surface pH
414
Handbook of Non-Invasive Methods and the Skin, Second Edition
pH (mean +/− SD)
7.0 6.5 6.0 5.5 5.0
Newborns
Childhood
Ye ar s
60
–7
3
M on th 4– M on 12 th 12 –2 Ye 4 ar s1 Ye –6 ar s8 –9
D ay 8 D ay 28
D ay 1
4.5
Adults
FIGURE 49.2 pH values at different ages.
values; i.e., a good sweat inhibition results in low pH values.49 An older paper, in contrast, reported the influence of sweating resulting in a more acidic skin surface pH.50 It has been shown that diaper dermatitis is correlated with high pH.51 Children wearing diapers showed higher pH values early after birth.52
49.3.2 CIRCADIAN RHYTHM Recent studies reported a circadian rhythm of skin surface pH in forearm, shin, axillae, and facial regions.53–55 Yosipovitch et al.55 reported SC pH maxima between 2:00 and 4:00 P.M., with minimal values around 8:00 P.M., with a variation over 24 hours for both men and women of 0.4 pH units. Le Fur et al.53 found the lowest values around 4:00 A.M. with a plateau during daytime. The axillae showed a signficant decrease during a single evening series of measurements (5.49 ± 0.23) compared to a single morning measurement (5.87 ± 0.23).54 Again, these data suggest the influence of hormones underlaying a circadian cycle, like glucocorticosteroids.
49.3.3 AGE DIFFERENCES The pH of the stratum corneum of full-term mammalian infants is known to be near neutral (6.5),52 and an acidic skin surface pH develops postnatally. In the first postnatal month a rapid decline in the pH is observed, reaching values around 5.4 early after birth.34,52,56–58 These changes are followed by a more gradual change in the next 6 to 12 months until the steady-state characteristic of the adults is reached. Aged persons show elevated SC pH values42,59,60 (Figure 49.2).
49.3.4 GENDER
AND
RACIAL DIFFERENCES
Only few investigators have reported a gender-related difference. A number of studies reported women to be more
acidic,61–63 while others could not prove any significant differences in the pH values.59,60 Controversial data exist regarding the influence of race on skin pH. The skin surface pH in African Americans displayed significantly lower values in the upper SC, while deeper sections of the SC assessed after sequential tape stripping did not reveal any difference in the pH gradient.64 Similar results were presented in an early study of Warrier et al.65 In contrast, an early publication from Draize showed that the average pH on multiple areas was higher in black (pH 5.21) males than in white males (pH 4.85).66
49.3.5 ANATOMICAL SKIN AREAS
FOR
TESTING
The standard test site for pH measurements is the volar forearm, with some differences depending on the position of the test site with lowest values detected close to the wrist.63 It has been shown that skin surface pH is relatively constant in different body areas,67,68 except for the intertriginous areas (axilla, inguinal, and submammary folds and finger webs) and sebaceous gland-rich areas (breast skin and back).61,62,69 Differences between dominant and nondominant forearms in women were shown in a study of Treffel et al.,70 with the dominant arm showing slightly higher pH values than the nondominant arm. Table 49.1 summarizes the skin surface pH on the skin of healthy adults.
49.3.6 INFLUENCE
OF
TAPE STRIPPING
Skin pH increases with tape stripping. The pH gradient is steepest in the area of the stratum corneum directly above the granular layer71 — an increase of 0.2 to 0.3 units was shown for each step.64,71,72 Such a sharp decline over a distance of only 10 to 15 μm requires a 100- to 1000-fold increase in protons (i.e., H+ ion concentration).
Skin Surface pH: Mechanism, Measurement, Importance
415
TABLE 49.1 pH Values at Different Anatomical Localizations in Humans Number of Subjects
Mean Value (Males)
Number of Subjects
22 14
5.06 ± 0.11 5.07 ± 0.62
18 14
5.26 ± 0.12 5.07 ± 0.62
61 67
Cheek
282
5.1 ± 0.8
292
5.2 ± 0.9
68
Axilla
27
6.13 ± 0.13
13
6.15 ± 0.16
62
Inguinal area
22
6.22 ± 0.13
18
6.22 ± 0.16
61
Breast
22
5.54 ± 0.1
18
5.83 ± 0.17
61
Back
27
5.04 ± 0.08
13
5.58 ± 0.15
62
15
5.99 ± 0.45
69
Volar forearm
Vulva Forehead
Inframammary
PH AND
Reference
27 282
4.95 ± 0.08 4.7 ± 0.7
13 292
5.22 ± 0.09 4.8 ± 0.8
59 68
22
5.54 ± 0.1
18
5.83 ± 0.17
61
49.4 IMPORTANCE OF PH FOR THE BARRIER 49.4.1
Mean Value (Females)
ANTIMICROBIAL FUNCTION
The density and composition of the skin flora are dependent on a variety of factors; pH is preeminent48,51,57,73,74 While propionibacteria grow well at pH 6.0 to 6.5, growth slows at pH 5.5.46,75 In contrast, Staphylococcus aureus grows best at pH 7.5, but continues to proliferate slowly at pH 5.0 to 6.0.46,75,76 Thus, the acidic pH of the SC restricts colonization by pathogenic flora and encourages persistence of normal microbial flora. Pertinently, both intertriginous and inflamed skin display an increased skin pH.42,51,61,77,78 The developmental delay in acidification in neonates often is complicated by increased hydration and urea/fecal contamination, which can further increase SC pH, with a subsequent increased risk of infection and diaper dermatitis.48,51,52,78 Thus, SC pH is closely linked to microbial colonization and possible pathogenesis.
49.4.2 CHANGES IN PH COULD INFLUENCE KEY SC FUNCTIONS Barrier repair after acute perturbations proceeds normally at an acidic pH, while recovery is delayed at a neutral pH, independent of the ionic environment of the bathing solution.1,10,79 Moreover, lamellar body (LB) secretion is unimpeded at neutral pH; instead, the postsecretory processing of newly secreted polar lipids into mature lamellar bilayers is impeded at a neutral pH. Furthermore, exposure to a neutral pH correlates with
inhibition of β-GlcCer’ase activity in the lower SC, as shown by in situ zymography.79 Additionally, neonatal SC exhibits a neutral pH with normal basal barrier function,52,56,58 but a delay in barrier recovery after acute perturbations, in conjunction with decreased lipid processing (preliminary observations from our laboratory). Together, these results support the hypothesis that changes in extracellular pH influence barrier homeostasis by modulating the expression or activity of lipid processing enzymes. Altered surface pH values after tape stripping regain normal values in humans within 3 to 4 days.80 Pretreatment with acidic -hydroxy acids (pH 3.4 to 4.4 compared to an alkaline vehicle) revealed significantly higher resistance against a barrier challenge with SLS.81 Ultrastructural findings support the positive effect of glycolic acid on the degradation of desmosomes.82 These studies show again the importance of an acidic SC pH for barrier homeostasis and SC integrity and cohesion. Permeability barrier homeostasis requires the secretion of both lipid precursors and colocalized hydrolytic enzymes from epidermal lamellar bodies (LBs), and the subsequent, postsecretory processing of polar lipids into their more nonpolar products.83 Whereas changes in extracellular ions regulate LB secretion,84 the epidermal hydrolytic enzymes generate lipid products required to form the mature membrane structures that mediate barriers: β-glucocerebrosidase, acidic sphingomyelinase, secretory phospholipase A2, and steroid sulfatase. Two of these key enzymes, secretory phospholipase A285,86 and steroid sulfatase, 87 exhibit neutral pH optima;
416
Handbook of Non-Invasive Methods and the Skin, Second Edition
49.5 PH AND SKIN DISORDERS
epidermal β-glucocerebrosidase and acidic sphingomyelinase display acidic pH optima.88,89 Blockade of these enzymes results in abnormal SC structure and function in human diseases, such as Gaucher disease,89–91 and recessive x-linked ichthyosis (RXLI)91 in murine transgenic knockout models of these disorders88,89 and in inhibitor-based models.89 Furthermore, it has been shown that the defects in the barrier could be corrected by acidification of the stratum corneum, indicating that the abnormal permeability barrier is due to failure to acidify the SC, and not to other developmental defects.92 Thus, it appears that the activity of the lipid processing enzymes, critical for barrier integrity, is dependent upon the SC localized pH for appropriate activities. The acidity of SC is shown to influence both directly and indirectly the barrier homeostasis and integrity/cohesion.1–3,93 Corneocyte adhesion is mainly dependent on the corneodesmosome riveting. Anyhow, detergent extraction studies have shown that not only corneodesmosomes, but also extracellular lipids (whose processing is pH dependent), regulate SC cohesion.94,95 A recent study has demonstrated that pH directly influences SC cohesion by means of serine-protease-mediated degradation of corneodesmosomes1 (Figure 49.1). Based upon in vitro inhibitor studies, two enzymes located in the stratum corneum are responsible for the epidermal desquamation: chemotryptic enzyme and tryptic enzyme.1,96–99 The chemotryptic enzyme activation is strictly controlled under basal conditions by the requirement for precursor activation by tryptic enzyme100 and by endogenous serine proteinase inhibitors, the colocalization of certain SC lipids, and the physiologic state of SC under basal conditions,96 i.e., low pH, low Ca++, and low hydration (above the stratum compactum) (Figure 49.1).
49.5.1 ECZEMATOUS DISEASES Eczematous skin disease are associated with an elevated skin surface pH. Eczematous skin of the volar forearm shows an elevated surface pH (compared to healthy skin) in a large group of atopic dermatitis children (ages 8 to 9).101 Seidenari and Giusti102 showed elevated pH values in both involved and uninvolved skin of children with atopic dermatitis, compared with the skin of nondiseased children. At the same time, involved skin was characterized by higher values than uninvolved skin.102 Beare et al.77 reported an alkaline pH in children affected by seborrheic dermatitis, compared to healthy children in both affected and unaffected sites. In cases with acute eczema with erosions, the skin surface pH is near neutral (pH 7.3 to 7.4)38 (Table 49.2).
49.5.2 ICHTHYOSIS Ichthyotic patients were reported to have an elevated surface pH in different areas in both involved and uninvolved skin, compared to healthy subjects.103 The only exception was the axillary regions displaying lower pH than normal patients, probably due to the reduced sweating of these patients. The higher pH in ichthyosis vulgaris was paralleled by an impaired alkali resistance test.103 More recently, Ohman and Vahlquist71 confirmed the elevated surface pH in ichthyosis vulgaris compared to X-linked recessive ichthyosis patients and healthy volunteers. However the pH gradient showed markedly lower pH values in X-linked recessive ichthyosis (RXLI) and slightly lower values in ichthyosis vulgaris. One possible explanation for
TABLE 49.2 pH Values of the Skin in Patients with Skin Disorders Number of Subjects Acute eczema Atopic dermatitis Involved skin Uninvolved skin Seborrheic dermatitis Ichthyosis Ichthyosis vulgaris X-linked recessive ichthyosis Fungal infections (Tinea manuum) Hemodialysis patients Diabetus mellitus
Mean Value 7.3–7.4
55 45
5.54 ± 0.63 5.23 ± 0.74
7 6 68 41 50
5.3 4.6 5.2 5.50 5.2
± ± ± ± ±
0.7 0.4 0.57 (hand) 0.10 (forearm) 0.1 (forearm)
Reference 38 102 102
71 71 74 62 61
Skin Surface pH: Mechanism, Measurement, Importance
417
theses findings in ichthyosis might be the increased activity of serum protease enzymes that could occur with elevated pH. In contrast, a low pH would decrease the protease activity, and consequently lead to SC retention, as occurs in RXLI (Table 49.2).
6. Korting, H.C. et al. Influence of the pH-value on the growth of Staphylococcus epidermidis, Staphylococcus aureus and Propionibacterium acnes in continuous culture. Zentralbl Hyg Umweltmed, 193: 78–90, 1992. 7. Patterson, M.J., S.D. Galloway, and M.A. Nimmo. Variations in regional sweat composition in normal human males. Exp Physiol, 85: 869–875, 2000. 8. Mao-Qiang, M. et al. Secretory phospholipase A2 activity is required for permeability barrier homeostasis. J Invest Dermatol, 106: 57–63, 1996. 9. Mao-Qiang, M. et al. Extracellular processing of phospholipids is required for permeability barrier homeostasis. J Lipid Res, 36: 1925–1935, 1995. 10. Fluhr, J.W. et al. Stratum corneum acidification in neonatal skin: secretory phospholipase A2 and the NHE1 antiporter acidify neonatal rat stratum corneum. J Invest Dermatol, 122: 320–329, 2004. 11. Fluhr, J.W. and P.M. Elias. Stratum corneum pH: formation and function of the “acid mantle.” Exogenous Dermatol, 1: 163–175, 2002. 12. Demaurex, N. and S. Grinstein. Na+/H+ antiport: modulation by ATP and role in cell volume regulation. J Exp Biol, 196: 389–404, 1994. 13. Mahnensmith, R.L. and P.S. Aronson. The plasma membrane sodium-hydrogen exchanger and its role in physiological and pathophysiological processes. Circ Res, 56: 773–788, 1985. 14. Orlowski, J. and S. Grinstein. Na+/H+ exchangers of mammalian cells. J Biol Chem, 272: 22373–22376, 1997. 15. Orlowski, J., R.A. Kandasamy, and G.E. Shull. Molecular cloning of putative members of the Na/H exchanger gene family. cDNA cloning, deduced amino acid sequence, and mRNA tissue expression of the rat Na/H exchanger NHE-1 and two structurally related proteins. J Biol Chem, 267: 9331–9339, 1992. 16. Sarangarajan, R. et al. Molecular and functional characterization of sodium-hydrogen exchanger in skin as well as cultured keratinocytes and melanocytes. Biochim Biophys Acta, 1511: 181–192, 2001. 17. Orlowski, J. and S. Grinstein. Diversity of the mammalian sodium/proton exchanger SLC9 gene family. Pflugers Arch, 2003. 18. van Hooijdonk, C. et al. Demonstration of an Na+/H+ exchanger in mouse keratinocytes measured by the novel pH-sensitive fluorochrome SNARF-calcein. Cell Proliferation, 30: 351–363, 1997. 19. Behne, M. et al. Functional role of the sodium/hydrogen antiporter, NHE1, in the epidermis: pharmacologic and NHE1 null-allele mouse studies. J Invest Dermatol, 114: 797, 2000. 20. Behne, M. et al. NHE1 regulates the stratum corneum permeability barrier homeostasis. J Biol Chem, 277: 47399–47406, 2002. 21. Fluhr, J.W. et al. Functional consequences of a neutral stratum corneum pH in neonates. J Invest Dermatol, 123: 140–151, 2004. 22. Rawlings, A.V. et al. Stratum corneum moisturization at the molecular level. J Invest Dermatol, 103: 731–741, 1994.
49.5.3 OTHER DISEASES Higher skin surface pH was observed with patients suffering fungal infections in the toe webs.74 Alkalinization of the stratum corneum, as observed in the urea-soaked skin of diaper dermatitis, is an important antecedent of bacterial infections.78,104 Additionally, other disease states that are associated with accelerated aging show altered skin pH values. For example, patients with non-insulin-dependent diabetes mellitus showed higher pH values in intertriginous regions than normal control subjects.61 Increased pH, in turn, might be one factor that increases susceptibility of diabetic skin to bacterial and yeast colonization or infections. Yosipovitch et al.62 also showed an elevated skin surface pH in patients undergoing hemodialysis, compared to healthy control subjects, even though hemodialyzed patients display an overall decrease in blood pH (Table 49.2).
49.6 SUMMARY Skin surface pH is one of the fundamental skin bioengineering parameters giving information about the buffering capacity of the skin. This regulatory mechanism has profound effects on stratum corneum integrity and cohesion, barrier homeostasis, as well as antimicrobial function of the stratum corneum. Skin pH measurements have proven to be essential in the overall assessment of the physiologic condition of the skin.
REFERENCES 1. Hachem, J.P. et al. pH directly regulates epidermal permeability barrier homeostasis, and stratum corneum integrity/cohesion. J Invest Dermatol, 121: 345–353, 2003. 2. Fluhr, J.W. et al. Generation of free fatty acids from phospholipids regulates stratum corneum acidification and integrity. J Invest Dermatol, 117: 44–51, 2001. 3. Mauro, T. et al. Barrier recovery is impeded at neutral pH, independent of ionic effects: implications for extracellular lipid processing. Arch Dermatol Res, 290: 215–222, 1998. 4. Heuss, E. Die Reaktion des Scheisses beim gesunden Menschen. Monatsh Prakt Dermatol, 14: 343–359, 1892. 5. Lieckfeldt, R. et al. Apparent pKa of the fatty acids within ordered mixtures of model human stratum corneum lipids. Pharm Res, 12: 1614–1617, 1995.
418
Handbook of Non-Invasive Methods and the Skin, Second Edition
23. Harding, C. et al. Review article: dry skin, moisturization, and corneodesmolysis. Int J Cosmet Sci, 22: 21–52, 2000. 24. Harding, C.R. and I.R. Scott. Histidine-rich proteins (filaggrins): structural and functional heterogeneity during epidermal differentiation. J Mol Biol, 170: 651–673, 1983. 25. Krien, P.M. and M. Kermici. Evidence for the existence of a self-regulated enzymatic process within the human stratum corneum: an unexpected role for urocanic acid. J Invest Dermatol, 115: 414–420, 2000. 26. Baden, H.P. and M.A. Pathak. The metabolism and function of urocanic acid in skin. J Invest Dermatol, 48: 11–17, 1967. 27. Schwarz, W. et al. Distribution of urocanic acid in human stratum corneum. Photodermatology, 3: 239–240, 1986. 28. Kuroda, Y. et al. A new sensitive method for assay of histidase in human skin and detection of heterozygotes for histidinemia. Clin Chim Acta, 96: 139–144, 1979. 29. Sano, H. et al. Isolation of a rat histidase cDNA sequence and expression in Escherichia coli: evidence of extrahepatic/epidermal distribution. Eur J Biochem, 250: 212–221, 1997. 30. Baden, H.P. and L. Gavioli. Histidase activity in rat liver and epidermis. J Invest Dermatol, 63: 479–481, 1974. 31. Bhargava, M.M. and M. Feigelson. Studies on the mechanisms of histidase development in rat skin and liver. I. Basis for tissue specific developmental changes in catalytic activity. Dev Biol, 48: 212–225, 1976. 32. Taylor, R.G., H.L. Levy, and R.R. McInnes. Histidase and histidinemia. Clinical and molecular considerations. Mol Biol Med, 8: 101–116, 1991. 33. Dikstein, S. and A. Zlotogorski. Measurement of skin pH. Acta Derm Venereol Suppl, 185: 18–20, 1994. 34. Behne, M. et al. Neonatal development of the stratum corneum pH gradient: localization and mechanisms leading to emergence of optimal barrier function. J Invest Dermatol, 120: 998–1006, 2003. 35. Hanson, K.M. et al. Two-photon fluorescence lifetime imaging of the skin stratum corneum pH gradient. Biophys J, 83: 1682–1690, 2002. 36. Masters, B.R., P.T.C. So, and E. Gratton. Multiphoton excitation microscopy of in vivo human skin: functional and morphological optical biopsy based on three-dimensional imaging, lifetime measurements and fluorescence spectroscopy. Ann NY Acad Sci, 838: 58–67, 1998. 37. Kroll, C. et al. Influence of drug treatment on the microacidity in rat and human skin: an in vitro electron spin resonance imaging study. Pharm Res, 18: 525–530, 2001. 38. Parra, J.L. and M. Paye. EEMCO guidance for the in vivo assessment of skin surface pH. Skin Pharmacol Appl Skin Physiol, 16: 188–202, 2003. 39. Courage + Khazakaelectronic GmbH, Bedienungsanleitung und Information zum Skin-pH-Meter. 40. Post, A., M. Gloor, and W. Gehring. Über den Einfluss der Hautwaschung auf den Haut-pH-Wert. Dermatol Monatsschr, 178: 216–222, 1992.
41. Gfatter, R., P. Hackl, and F. Braun. Effects of soap and detergents on skin surface pH, stratum corneum hydration and fat content in infants. Dermatology, 195: 258–262, 1997. 42. Thune, P. et al. The water barrier function of the skin in relation to the water content of stratum corneum, pH and skin lipids. The effect of alkaline soap and syndet on dry skin in elderly, non-atopic patients. Acta Derm Venereol, 68: 277–283, 1988. 43. Korting, H.C. et al. Influence of repeated washings with soap and synthetic detergents on pH and resident flora of the skin of forehead and forearm. Results of a crossover trial in health probationers. Acta Derm Venereol, 67: 41–47, 1987. 44. Grunewald, A.M. et al. Damage to the skin by repetitive washing. Contact Derm, 32: 225–232, 1995. 45. Grunewald, A.M. et al. Efficacy of barrier creams. Curr Probl Dermatol, 23: 187–197, 1995. 46. Korting, H.C. et al. Differences in the skin surface pH and bacterial microflora due to the long-term application of synthetic detergent preparations of pH 5.5 and pH 7.0. Results of a crossover trial in healthy volunteers. Acta Derm Venereol, 70: 429–431, 1990. 47. Hartmann, A.A. Effect of occlusion on resident flora, skin-moisture and skin-pH. Arch Dermatol Res, 275: 251–254, 1983. 48. Aly, R. et al. Effect of prolonged occlusion on the microbial flora, pH, carbon dioxide and transepidermal water loss on human skin. J Invest Dermatol, 71: 378–381, 1978. 49. Hölzle, E. Antiperspirants. Dermatopharmacology of Topical Preparations, P. Treffel, Ed. Springer, Heidelberg, 2000, pp. 401–416. 50. Bernstein, E.T. and F. Hermann. The acidity on the surface of the skin. NY State J Med, 42: 436–442, 1942. 51. Berg, R.W., M.C. Milligan, and F.C. Sarbaugh. Association of skin wetness and pH with diaper dermatitis. Pediatr Dermatol, 11: 18–20, 1994. 52. Visscher, M.O. et al. Changes in diapered and nondiapered infant skin over the first month of life. Pediatr Dermatol, 17: 45–51, 2000. 53. Le Fur, I. et al. Analysis of circadian and ultradian rhythms of skin surface properties of face and forearm of healthy women. J Invest Dermatol, 117: 718–724, 2001. 54. Burry, J., H.F. Coulson, and G. Roberts. Circadian rhythms in axillary skin surface pH. Int J Cosmet Sci, 23: 207–210, 2001. 55. Yosipovitch, G. et al. Time-dependent variations of the skin barrier function in humans: transepidermal water loss, stratum corneum hydration, skin surface pH, and skin temperature. J Invest Dermatol, 110: 20–23, 1998. 56. Behrendt, H. and M. Green. Skin pH pattern in the newborn infant. J Dis Child, 95: 35–41, 1958. 57. Giusti, F. et al. Skin barrier, hydration, and pH of the skin of infants under 2 years of age. Pediatr Dermatol, 18: 93–96, 2001. 58. Fox, C., D. Nelson, and J. Wareham. The timing of skin acidification in very low birth weight infants. J Perinatol, 18: 272–275, 1998.
Skin Surface pH: Mechanism, Measurement, Importance
419
59. Wilhelm, K.P., A.B. Cua, and H.I. Maibach. Skin aging. Effect on transepidermal water loss, stratum corneum hydration, skin surface pH, and casual sebum content. Arch Dermatol, 127: 1806–1809, 1991. 60. Zlotogorski, A. Distribution of skin surface pH on the forehead and cheek of adults. Arch Dermatol Res, 279: 398–401, 1987. 61. Yosipovitch, G. et al. Skin surface pH in intertriginous areas in NIDDM patients. Possible correlation to candidal intertrigo. Diabetes Care, 16: 560–563, 1993. 62. Yosipovitch, G. et al. Skin surface pH, moisture, and pruritus in haemodialysis patients. Nephrol Dial Transplant, 8: 1129–1132, 1993. 63. Ehlers, C. et al. Females have lower skin surface pH than men. A study on the surface of gender, forearm site variation, right/left difference and time of the day on the skin surface pH. Skin Res Technol, 7: 90–94, 2001. 64. Berardesca, E. et al. Differences in stratum corneum pH gradient when comparing white Caucasian and black African-American skin. Br J Dermatol, 139: 855–857, 1998. 65. Warrier, A.G. et al. A comparison of black and white skin using noninvasive methods. J Soc Cosmet Chem, 47: 229–240, 1996. 66. Draize, J.H. The determination of the pH of the skin of man and common laboratory animals. J Invest Dermatol, 5: 77–85, 1942. 67. Fluhr, J.W. et al. Impact of anatomical location on barrier recovery, surface pH and stratum corneum hydration after acute barrier disruption. Br J Dermatol, 146: 770–776, 2002. 68. Zlotogorski, A. and S. Dikstein. Measurement of skin surface pH, in Handbook of Non-Invasive Methods and the Skin, J. Serup and G.B.E. Jemec, Eds., CRC Press, Boca Raton, FL, 1995, pp. 223–225. 69. Elsner, P. and H.I. Maibach. The effect of prolonged drying on transepidermal water loss, capacitance and pH of human vulvar and forearm skin. Acta Derm Venereol, 70: 105–109, 1990. 70. Treffel, P. et al. Hydration, transepidermal water loss, pH and skin surface parameters: correlations and variations between dominant and non-dominant forearms. Br J Dermatol, 130: 325–328, 1994. 71. Ohman, H. and A. Vahlquist. The pH gradient over the stratum corneum differs in X-linked recessive and autosomal dominant ichthyosis: a clue to the molecular origin of the “acid skin mantle”? J Invest Dermatol, 111: 674–677, 1998. 72. Ohman, H. and A. Vahlquist. In vivo studies concerning a pH gradient in human stratum corneum and upper epidermis. Acta Derm Venereol, 74: 375–379, 1994. 73. Jensen, P.J. et al. Plasminogen activator inhibitor type 2: an intracellular keratinocyte differentiation product that is incorporated into the cornified envelope. Exp Cell Res, 217: 65–71, 1995. 74. Chikakane, K. and H. Takahashi. Measurement of skin pH and its significance in cutaneous diseases. Clin Dermatol, 13: 299–306, 1995.
75. Korting, H.C. et al. Influence of topical erythromycin preparations for acne vulgaris on skin surface pH. Clin Investig, 71: 644–648, 1993. 76. Leyden, J.J., R. Stewart, and A.M. Kligman. Updated in vivo methods for evaluating topical antimicrobial agents on human skin. J Invest Dermatol, 72: 165–170, 1979. 77. Beare, J.T. et al. The pH of the skin surface of children with seborrhoeic dermatitis compared with unaffected children. Br J Dermatol, 70: 233–241, 1958. 78. Brook, I. Microbiology of secondarily infected diaper dermatitis. Int J Dermatol, 31: 700–702, 1992. 79. Mauro, T. et al. Barrier recovery is impeded at neutral pH, independent of ionic effects: implications for extracellular lipid processing. Arch Dermatol Res, 290: 215–222, 1998. 80. Wilhelm, D., P. Elsner, and H.I. Maibach. Standardized trauma (tape stripping) in human vulvar and forearm skin. Effects on transepidermal water loss, capacitance and pH. Acta Derm Venereol, 71: 123–126, 1991. 81. Berardesca, E. et al. Alpha hydroxyacids modulate stratum corneum barrier function. Br J Dermatol, 137: 934–938, 1997. 82. Fartasch, M., J. Teal, and G.K. Menon. Mode of action of glycolic acid on human stratum corneum: ultrastructural and functional evaluation of the epidermal barrier. Arch Dermatol Res, 289: 404–409, 1997. 83. Elias, P.M., N.S. McNutt, and D.S. Friend. Membrane alterations during cornification of mammalian squamous epithelia: a freeze-fracture, tracer, and thin-section study. Anat Rec, 189: 577–594, 1977. 84. Elias, P.M. and K.R. Feingold. Coordinate regulation of epidermal differentiation and barrier homeostasis. Skin Pharmacol Appl Skin Physiol, 14 (Suppl. 1): 28–34, 2001. 85. Suzuki, N. et al. Structures, enzymatic properties, and expression of novel human and mouse secretory phospholipase A(2)s. J Biol Chem, 275: 5785–5793, 2000. 86. Tischfield, J.A. A reassessment of the low molecular weight phospholipase A2 gene family in mammals. J Biol Chem, 272: 17247–17250, 1997. 87. Hobkirk, R. Steroid sulfotransferase and steroid sulfate sulfatase: characterization and biological roles. Can J Biochem Cell Biol, 63: 1127–1144, 1985. 88. Holleran, W.M. et al. Processing of epidermal glucosylceramides is required for optimal mammalian cutaneous permeability barrier function. J Clin Invest, 91: 1656–1664, 1993. 89. Schmuth, M. et al. Permeability barrier disorder in Niemann-Pick disease: sphingomyelin- ceramide processing required for normal barrier homeostasis. J Invest Dermatol, 115: 459–466, 2000. 90. Holleran, W.M. et al. -Glucocerebrosidase activity in murine epidermis: characterization and localization in relation to differentiation. J Lipid Res, 33: 1201–1209, 1992. 91. Zettersten, E. et al. Recessive x-linked ichthyosis: role of cholesterol-sulfate accumulation in the barrier abnormality. J Invest Dermatol, 111: 784–790, 1998.
420
Handbook of Non-Invasive Methods and the Skin, Second Edition
92. Fluhr, J.W. et al. Stratum corneum acidification in neonatal skin: secretory phospholipase A2 and the NHE1 antiporter acidify neonatal rat stratum corneum. J Invest Dermatol, 122: 320–329, 2004. 93. Behne, M.J. et al. The sodium/hydrogen antiporter, NHE1, regulates stratum corneum acidification. J Biol Chem, 277: 47399–47406, 2002. 94. Bissett, D.L., J.F. McBride, and L.F. Patrick. Role of protein and calcium in stratum corneum cell cohesion. Arch Dermatol Res, 279: 184–189, 1987. 95. Chapman, S.J. et al. Lipids, proteins and corneocyte adhesion. Arch Dermatol Res, 283: 167–173, 1991. 96. Rogers, J., A. Watkinson, and C.R. Harding. Characterisation of the effects of protease inhibitors and lipids on human stratum corneum chymotrytic-like enzyme supports a role in desquamation. J Invest Dermatol, 110: 672, 1998. 97. Suzuki, Y. et al. The role of two endogenous proteases of the stratum corneum in degradation of desmoglein-1 and their reduced activity in the skin of ichthyotic patients. Br J Dermatol, 134: 460–464, 1996. 98. Suzuki, Y. et al. Detection and characterization of endogenous protease associated with desquamation of stratum corneum. Arch Dermatol Res, 285: 372–377, 1993.
99. Wessendorf, J.H. et al. Identification of a nuclear localization sequence within the structure of the human interleukin-1 alpha precursor. J Biol Chem, 268: 22100–22104, 1993. 100. Hansson, L. et al. Cloning, expression, and characterization of stratum corneum chymotryptic enzyme. A skin-specific human serine proteinase. J Biol Chem, 269: 19420–19426, 1994. 101. Eberlein-Konig, B. et al. Skin surface pH, stratum corneum hydration, trans-epidermal water loss and skin roughness related to atopic eczema and skin dryness in a population of primary school children. Acta Derm Venereol, 80: 188–191, 2000. 102. Seidenari, S. and G. Giusti. Objective assessment of the skin of children affected by atopic dermatitis: a study of pH, capacitance and TEWL in eczematous and clinically uninvolved skin. Acta Derm Venereol, 75: 429–433, 1995. 103. Tippelt, H. Zur Hautoberflächen-pH-Mesung und Alkaliresistenzprobe bei Hautgesunden und Ichthyosis vulgaris-Kranken. Dermatologica, 139: 201–210, 1969. 104. Leyden, J.J. and A.M. Kligman. The role of microorganisms in diaper dermatitis. Arch Dermatol, 114: 56–59, 1978.
pH Gradient in the Epidermis 50 The Evaluated by Serial Tape Stripping Hans Öhman Department of Dermato-Venereology, University of Linköping, Linköping, Sweden
CONTENTS 50.1 Introduction............................................................................................................................................................421 50.2 pH: Definition and Regulation ..............................................................................................................................421 50.3 Methods for pH Measuring in the Epidermis .......................................................................................................422 50.3.1 Flat pH Glass Electrode ............................................................................................................................422 50.3.2 pH-Dependent Dyes ..................................................................................................................................422 50.3.2.1 Penetrating pH Microelectrodes.................................................................................................422 50.4 The pH of the Skin................................................................................................................................................422 50.4.1 Effects of Soaps and Detergents on pH in the Epidermis........................................................................423 50.4.2 pH-Dependent Enzymes ............................................................................................................................423 50.5 Tape Stripping........................................................................................................................................................425 50.6 Recommendations..................................................................................................................................................425 References .......................................................................................................................................................................425
50.1 INTRODUCTION The skin is the barrier at the interface between the organism and its environment. The skin barrier against water loss helps us regulate water balance and is essential for terrestrial life. It also protects us from many noxious factors: mechanical, chemical, actinic, microbial, and thermal. The epidermis is continously renewed, the time being about 30 days, from the basal keratinocyte to the desquamated corneocyte. Keratinocytes are connected to each other by desmosomes, specialized junctions, where cytoskeletal proteins interact with the cell membrane. In the stratum granulosum there are two types of specialized intracellular structures: keratohyalin granules and lamellar bodies (Odland bodies). When granular cells differentiate into corneocytes, keratin filaments are aggregated by filaggrin, a protein component of the keratohyalin granulem and the lamellar bodies empty their lipid content into the intercellular space. The corneocyte has a special type of cell envelope that is formed beneath the plasma membrane. This process is catalyzed by an enzyme, transglutaminase, which polymerizes protein precursors, such as involucrin and loricrin. During the process of desquamation, the various intercellular lipids are metabolically changed and the desmosomes are degraded. This process is partly regulated by
enzymes that are pH dependent. Whereas most enzymes in the body are adapted to the interior pH, which is kept constant at 7.4, many epidermal enzymes have a pH optimum around 4 to 6.1 Accordingly, intrinsic variations in the epidermal pH could be of major importance for the regulation of enzyme activities. Over the epidermis there is a pH gradient, with a pH of about 4 to 6 on the skin surface and an internal pH of 7.4.2
50.2 PH: DEFINITION AND REGULATION The proton, H+, the nucleus of the hydrogen atom, has a high chemical reactivity. Many biological molecules can give or take protons. The activity (a) of the hydrogen ion is usually named pH, where pH = –log a H+ There is always a certain concentration of protons in diluted water solutions in the form of H3O+ in equality with H+. The pH can be used as a synonym to proton concentration. The concentration of H3O+ in blood is normally 40 nmol/l (40 × 10–9 mol/l), which gives a pH of 7.4 (9 – log 40 = 7.4). In the human body most biological molecules are affected by changes in pH. Humans have 421
422
Handbook of Non-Invasive Methods and the Skin, Second Edition
developed a finely tuned system for regulating pH within the body. There is a system of buffers that can bind or release protons and hamper changes in proton concentration. Most cells in the body are almost impermeable to protons, except for the erythrocyte, which has a high permeability to HCO3– and CO2. H+ + HCO3– ↔ H2CO3 ↔ H2O + CO2 CO2 can be eliminated from the body via the lungs. Acids can be eliminated via the kidneys. These buffering mechanisms operate less well in the epidermis due to absence of blood supply, poorer diffusion of CO2 and O2, and the lack of elimination of acids.
50.3 METHODS FOR PH MEASURING IN THE EPIDERMIS 50.3.1 FLAT PH GLASS ELECTRODE The flat glass electrode is a well-documented and commonly used instrument for pH measuring.3,4,9 There are many different producers of pH electrodes for different uses on the market, and they are of good quality. The flat pH electrode is well suited for measuring the epidermal surface both in vivo and in vitro. The measurings have a good reproducibility when using the instrument according to the manufacturers’ instructions.
50.3.2
PH-DEPENDENT
DYES
The fact that some dyes have a color that changes according to pH has been used by some authors. Wagner et al.5 have measured the pH gradient over the skin (including dermis) in vivo and in vitro (fresh and frozen excised human skin) using the Franz diffusion cell in a penetration model concerning incubation time and the pH of the acceptor solution. Their question was whether the pH across the skin might be influenced by the conditions of different in vitro test models utilized for the examination of drug delivery to the skin. They also used the Saarbruecken penetration model, where the skin itself represents the acceptor for the penetrating drug. The stratum corneum pH gradient was measured using a flat pH electrode during tape stripping. Then a specimen from the stripped area was taken out and transferred into a cryomicrotome, where it was cut into surface parallell sections, each being measured with the flat pH glass electrode. pH was also measured using two different dyes. With the same slicing method the cross sections were analyzed with a confocal laser scanning imaging system after staining with a solution of the fluoroscense carboxy-seminaphtho-rhodafluor-1 (carboxy-SNARF-1). The emission spectrum of carboxy-SNARF-1 undergoes a pH-dependent
wavelength shift, changing from acidic green to alkaline red fluoroscense. The other method used was staining the cross sections with bromthymol blue. It changes its color between pH 6 to 7.6 from acidic yellow to alkaline blue. Wagner et al.5 conclude “that the conditions of the in vitro test systems influence the pH in the skin, which for its part, will influence the diffusion of drugs and their distribution in the different skin layers.” The pH data from measuring with a flat glass electrode were in agreement with the staining methods. A problem is that carboxy-SNARF-1 and most other pH probes in the market are made for measuring pH intracellularly, with an optimum pH range of 6 to 9, and that makes the pH gradient of stratum corneum, with its range of 4 to 6, more difficult to measure with these methods.6 In addition to poor performance in the low pH range, the probe is made for the in vitro situation and is potentially toxic. Finding a probe with its optimum pH range between 4 and 7 and nontoxic would make confocal laser scanning microscopy an attractive method. 50.3.2.1 Penetrating pH Microelectrodes The microelectrode is used with good results measuring pH in the brain of rats,7 but with its very thin glass edge of 5 μ, it is too easily broken when penetrating viable human skin, specially in the in vivo situation with difficulties for the test person to sit quite still.
50.4 THE PH OF THE SKIN The finding by Schade and Marchionini that the skin surface was acid, der Säuremantel, has been confirmed by several subsequent studies.3,8,9 In different studies the pH of the skin surface usually differs between 4 and 6. Differences in pH according to skin sites,10 seasonal variations,11 and recently differences in pH over a 24-hour period have also been demonstrated.12 There has also been shown gender differences in surface pH, with males having somewhat lower pH than women1,2,5 (Figure 50.1). A higher pH, between 5 and 7, is seen in the flexures and on the forehead, and is probably due to the increased sweat excretion and microbial effects. In diseased skin, e.g., eczema, the pH is raised to 5.5 to 6, and in bacteriological or microbid reactions, up to 7 to 8.13 As soon as the skin barrier is perturbed, pH tends to become more alkaline.2 Blister fluid mostly has a pH of around 7. Schade and Marchionini,8 however, found that this is not true for vesicles in true (nonmycotic) dyshidrosis, where pH is around 5, in contrast to staphylococcal or fungal vesicles with a pH around 7 (Figure 50.2). Within the body there is a rather constant pH of around 7.4. That is, over the epidermis there is a pH gradient that can be of importance to be measured.
The pH Gradient in the Epidermis Evaluated by Serial Tape Stripping
number of strippings is not directly proportional to the depth in stratum corneum, especially not in the ichthyoses. Some investigators have used 15 to 60 strippings in normal skin to reach the glistening layer15–17; others routinely used about 100 strippings to reach this layer.1
8
Surface pH
7
50.4.1 EFFECTS OF SOAPS AND DETERGENTS ON PH IN THE EPIDERMIS
6
5 Men Women 4 0
20
40
60 80 No. of strippings
100
120
FIGURE 50.1 Change in skin surface pH as a result of tape stripping of male (n = 7) and female (n = 7) forearm skin. Mean ± SEM values. 8
7
Surface pH
423
6
5
Most detergents and soaps are alkaline, with a pH above 7. Traumiterative eczema is a well-known effect of excessive washing. Rawlings et al.18 have shown that even minor trauma of this kind to skin surface gives disturbances in cornification, leading to a xerotic skin. Stratum corneum intercellular lipids consist of a mixture of ceramides, cholesterol, and fatty acids, together with small amounts of cholesterol sulfate, glycosyl ceramides, and phospholipids. Desmosomes and, to some extent, the intercellular lipids are responsible for corneocyte adhesion. In normal stratum corneum desmosomes become more and more degraded the closer they come to skin surface. Washing with soap three times daily for one week has been shown to inhibit degradation of desmosomes in the upper layers of the stratum corneum. Increased levels of the enzyme desmoglein have also been found with repeated washing, in comparision to normal stratum corneum. Under the same provocation, the lipid bilayers in the lower layers of stratum corneum were usually normal, but progressing disorganization was seen more superficially, suggesting a loss of order in the lipid bilayers. Another difference compared to normal skin was decreased ceramide and increased fatty acid levels in the soap-induced xerotic skin.
50.4.2
4 0
2
4
6 8 10 Days after stripping
12
14
FIGURE 50.2 Change in pH during skin repair after tape stripping on day 0.
The pH gradient between the skin surface and the viable epidermis was characterized for the first time in 1994 by Öhman and Vahlquist.2 pH values recorded in a semihydrophobic milieu, such as stratum corneum, should be interpreted with caution. We do not know whether surface pH actually reflects the hydrogen ion concentrations of intercellular water, or if it represents the combined acidity of exposed corneocytes, lipids, and water-soluble compounds. The furrows on the skin surface will lead to uneven removal of corneocytes during stripping, which might skew the pH gradient.14 The
PH-DEPENDENT
ENZYMES
Most enzymes in the human body have their optimum activity around a pH of 7.4. In the epidermis some of these enzymes have changed (during evolution?) their optimum to a more suitable pH of around 5 or 6. Changes in pH in the stratum corneum caused by barrier injury or disease could affect the activity of these enzymes and thus influence normal cellular maturation and other processes. One example of a pH-dependent enzyme in epidermis is found in vitamin A metabolism. Vitamin A has effects on keratinocyte differentiation. At a subcellular level it has its highest esterifying activity in the microsomal fraction. The enzyme catalyzing the reaction, acyl CoA:retinol acyltransferase (ARAT; EC 2.3.1.76) has a pH optimum of 5.5 to 6.0, which differs from that of ARAT in other tissues, e.g., in the liver (optimum pH, 7.4). Changes in pH could possibly affect ARAT activity, and presumably the esterification of vitamin A, which could be of importance in relation to pathologic keratinocyte differentiation.19
424
Handbook of Non-Invasive Methods and the Skin, Second Edition
Another example comes from cholesterol metabolism. Epidermal cholesterol is esterified by acyl CoA:cholesterol acyltransferase (EC 2.3.1.26). When pH is reduced from 7 to 5, this enzyme more than tenfold its activity.20 In a study of acid phosphatase a main activity band in epidermis was found at pH 5.65, whereas in whole skin the activity band was at pH 6.1, similar to the band from cultured skin fibroblasts and leukocytes. Since the band from stratum corneum and epidermis was not found in the culture of skin fibroblasts and leukocytes, the authors conclude that stratum corneum and epidermis contain a tissuespecific acid phosphatase.21 Finally, a specific serine protease, stratum corneum chymotryptic enzyme, with pH optimum around 8, has been found to be responsible for cell shedding from stratum corneum, probably with a role in the degradation of intercellular cohesive structures preceeding desquamation. This protease was also shown to have significant activity at a lower pH of 5 to 6.22,23 The most pertinent finding in the present studies is the difference between the pH profiles in ichthyosis vulgaris and x-linked recessive ichthyosis skin.24 In the former condition, the pH gradient is shifted to the upper layers of stratum corneum. In x-linked recessive ichthyosis skin, on the other hand, abnormally low pH values are found throughout the stratum corneum (Figure 50.3).
7
Normal IV
pH
6
XRI
5
4 Skin surface
1/4
1/2
3/4
1/1
Fraction of total stratum corneum thickness
FIGURE 50.3 The relative position of the pH gradient in stratum corneum differs in ichthyosis vulgaris (IV), x-linked recessive ichthyosis (XRI), and normal skin. The curves were obtained by freehand adjustment of tape stripping data, compensating for differences in the total number of strippings and for variations in stripping effect.24
Given the different pH gradients in normal and ichthyotic stratum corneum, the results allow speculation not only about the pathophysiological role of the pH gradient in ichthyosis vulgaris and x-linked recessive ichthyosis, but also about the molecular origin of the so-called acid mantle of human skin. From a scrutiny of the literature, three classes of molecules emerged as the most likely source of protons in normal stratum corneum: (1) certain amino acids and filaggrin-related breakdown products, e.g., urocanic acid and pyrrolidone carboxylic acid, 25 also known as the natural moisturizing factor18,26; (2) alphahydroxy acids, such as lactic acid and butyric acid, which are naturally present in sweat and possibly other components of the skin13; and (3) acidic lipids, e.g., cholesterol sulfate and free fatty acids, which are either synthesized de novo in the upper epidermis or deposited onto the skin surface in areas with a high sebaceous and bacterial activity.15,27,28 Öhman and Vahlquist24 propose that the low surface pH on human skin is a combined effect of acidic excretion products from sweat and sebum, and hydrolytic products of filaggrin originating in the granular layer and further concentrated in the upper stratum corneum as the result of desiccation. Clearly, the lipid film on human skin, which penetrates the upper layers of the stratum disjunctum,15 cannot per se explain the low surface pH, as rinsing only marginally raises the pH29 and skin areas with high sebaceous activity (e.g., the forehead) have almost the same pH values as forearm skin.30 Likewise, since most of the cholesterol sulfate present in the lower stratum corneum will subsequently be deesterified by intercellularly located sterol sulfatase,27 it cannot contribute significantly to the acidity of the skin surface. Provided the hypothesis about the molecular origin of the pH gradient in normal stratum corneum is correct, the abnormal gradients in ichthyotic skin can easily be explained. In ichthyosis vulgaris there is a defect production of profilaggrin. The main acid derivatives of filaggrin, free fatty acids, urocanic acid, and pyrrolidon caboxylic acid, are lacking or in lower concentration, giving a relatively more alkaline pH in the stratum corneum, and hence moving the pH gradient outward. Xerosis of the skin will, however, partially counteract this event by concentrating protons near the surface. In x-linked recessive ichthyosis there is a lack of the enzyme cholesterol sulfatase, and this causes an accumulation of acid cholesterol sulfate, leading to a relatively more acidic pH, that is, an acidic shift of the pH gradient.24 These changes in pH could have an effect on the pH-dependent enzymes in stratum corneum. It is reasonable to assume that the pH gradient, created over the stratum corneum during differentiation, will affect the pH-dependent processes that are involved in the
The pH Gradient in the Epidermis Evaluated by Serial Tape Stripping
425
8
Surface pH
7
6
5 Cyanoacrylate Tape 4 0
20
40
0
2
4
60
80
100
6
8
10
120 Tape Cyanoacrylate 12
No. of strippings
FIGURE 50.4 Comparision between tape (n = 7) and cyanoacrylate resin (n = 6) stripping used to demonstrate the pH gradient and stripping efficacy.
shedding of corneocytes from the skin surface. It should be possible to test this theory experimentally, for example, by artifactually changing the pH in the stratum corneum and observing the effects on desquamation in the two types of ichthyoses.
50.5 TAPE STRIPPING Tape stripping is usually performed by pressing a piece of tape, e.g., D-squame stripping discs (Cuderm Corporation), onto the skin for a certain time and a certain pressure and then stripping away the stratum corneum consecutively until a glistening surface appears and pain develops. Different tapes take away different amounts of stratum corneum. Cyanoacrylate resin has also been used for stripping the stratum corneum with about a tenfold increase in efficacy compared to tape, but this glue is potentially irritating and contact allergy has occasionally been reported1 (Figure 50.4 and Figure 50.5). The corneocytes are loosening their desmosomes and adhesion in the upper parts of stratum corneum, which gives you an uneven stripping effect — more cells in the first tape strips you take away than deeper in the skin. In many cases, it is of importance to know the nivau in stratum corneum when you correlate to some other variabel, e.g., pH or concentration of penetrating substance. There are different methods to measure the amount of tape-stripped corneocytes. In some experiments stripping
discs are weighed before and after stripping. It is a rather time-consuming method24 (Figure 50.6). In another study to quantificate the removed corneocytes, they compared the pseudoabsorption of the corneocyte in the range of 430 nm to the protein absorption in the UV range of 278 nm and absorption at 652 nm obtained after staining of the stratum corneum proteins with trypan blue.31 In a related study they used the absorbance in the visible spectral range measured with optical spectroscopy as another method, which allows determination of the absolute mass of corneocyte aggregates on the removed tape strip.32 These authors consider these methods better than weighing the tape strips.
50.6 RECOMMENDATIONS The flat pH glass electrode is an instrument that is simple to use, has good precision, and is relatively cheap to buy. Stripping with tape is the method I recommend for taking away the stratum corneum in a slow and rather controlled way. Confocal laser microscopy is expensive in equipment, but promising and noninvasive, though still lacking the right probe for pH in the stratum corneum.
REFERENCES 1. Goldsmith L. Physiology, Biochemistry, and Molecular Biology of the Skin, 2nd ed. Oxford University Press, New York, 1991.
426
Handbook of Non-Invasive Methods and the Skin, Second Edition
A
B
C
D
FIGURE 50.5 Micrographs of skin obtained (a) before and after (b) 10, (c) 50, and (d) 100 tape strippings of male forearm skin down to the glistening surface. Stratum corneum is artifactually expanded in (a) and (b) due to maximal hydration. Original magnification, ×86. 0.12 IV Normal XRI
Weight/Strip (mg)
0.10
0.08
0.06
0.04
0.02
0.00 0
20
40
60 80 No. of strippings
100
120
FIGURE 50.6 The stripping efficacy is higher near the skin surface, but there are no overt differences between normal and ichthyotic skin. Individul values and computerized curves fitting the weight/strip in one healthy man and two patients with ichthyosis vulgaris (IV) and x-linked recessive ichthyosis (XRI), respectively. Tape discs were weighed before and after every 5 to 20 strippings of the ventral forearm.
The pH Gradient in the Epidermis Evaluated by Serial Tape Stripping
2. Öhman H, Vahlquist A. In vivo studies concerning a pH gradient in human stratum corneum and upper epidermis. Acta Derm Venereol 74:375–379, 1994. 3. Blank I. Measurement of the pH of the skin surface. J Invest Dermatol 2:67–74, 1939. 4. Jolly H, Hailey C, Netick J. pH determination of the skin. J Invest Dermatol 46:305–308, 1962. 5. Wagner H, Kostka KH, Lehr CM, Schafer UF. pH profiles in human skin, influence of two in vitro test systems for drug delivery testing. Eur J Pharm Biopharm 55:57–65, 2003. 6. Turner NG, Cullander C, Guy RH. Determination of the pH gradient across the stratum corneum. J Invest Dermatol Symp Proc 3:110–113, 1998. 7. Ekholm A. The Coupling of Ion Fluxes to Energy Homeostasis in Ischemia. Medical dissertation, Lund, Sweden, 1992. 8. Schade H, Marchionini A. Der Säuremantel der Haut. Klin Wochenschr 7:12–14, 1928–29. 9. Dickstein S, Zlotogorski A. Measurement of skin pH. Acta Derm Venereol Suppl 185:18–20, 1994. 10. Zlotogorski A. Distribution of skin surface pH on the forehead and cheek of adults. J Invest Dermatol 279:398–401, 1987. 11. Abe T, Mayuzumi J, Kikuchi N, Arai S. Seasonal variations in skin temperature, skin pH, evaporative waterloss and skin surface lipid values on human skin. Chem Pharm Bull (Tokyo) 28:387–392, 1980. 12. Yosipovitch G, Xiong GL, Haus E, Sackettlundeen L, Ashkenazi I, Maibach HI. Time-dependent variations of the skin barrier function in humans: transepidermal water loss, stratum corneum hydration, skin surface pH, and skin temperature. J Invest Dermatol 110:20–23, 1998. 13. Rothman S. pH of sweat and skin surface. In Physiology and Biochemistry of the Skin, Rothman S, Ed. University of Chicago Press, Chicago, 1954, pp. 221–232. 14. van der Molen R, Spies F, van't Noordende J, Boelsma E, Mommaas A, Koerten H. Tape stripping of human stratum corneum yields cell layers that originate from various depths because of furrows in the skin. Arch Dermatol Res 289:514–518, 1997. 15. Bommannan D, Potts R, Guy R. Examination of the stratum corneum barrier function in vivo by infrared spectroscopy. J Invest Dermatol 95:403–408, 1990. 16. Chapman S, Walsh A, Jackson S, Friedmann P. Lipids, proteins and corneocyte adhesion. Arch Dermatol Res 283:167–173, 1991. 17. Wilhem D, Elsner P, Maibach H. Standardized trauma (tape stripping) in human vulvar and forearm skin. Effects on transepidermal water loss, capacitance and pH. Acta Derm Venereol 71:123–126, 1991.
427
18. Rawlings AV, Scott IR, Harding CR, Bowser PA. Stratum corneum moisturization at the molecular level. J Invest Dermatol 103:731–740, 1994. 19. Törmä H, Vahlquist A. Vitamin A esterification in human epidermis: a relation to keratinocyte differentiation. J Invest Dermatol 94: 132–138, 1990. 20. Freinkel R, Aso K. Esterification of cholesterol in the skin. J Invest Dermatol 52:148–154, 1969. 21. Meyer J, Grundmann H, Knabenhans S. Properties of acid phosphatase in human stratum corneum. Dermatologica 180:20–24, 1990. 22. Lundström A, Egelrud T. Cell shedding from human plantar skin in vitro: evidence of its dependence on endogenous proteolysis. J Invest Dermatol 91:340–343, 1988. 23. Sondell B. Stratum Corneum Chymotryptic Enzyme, Medical dissertation 467, Umeå, 1996. 24. Öhman H, Vahlquist A. The pH gradient over the stratum corneum differs in x-linked and autosomal dominant ichthyosis: a clue to the molecular origin of the the “acid skin mantle.” J Invest Dermatol 111:674–677, 1998. 25. Warner R, Bush R, Ruebusch N. Corneocytes undergo systematic changes in element concentrations across the human inner stratum corneum. J Invest Dermatol 104:530, 1995. 26. Scott I, Harding C, Barrett J. Histidine-rich protein of the keratohyalin granules. Source of the free amino acids, urocanic acid and pyrrolidone carboxylic acid in the stratum corneum. Biochem Biophys Acta 19:110–117, 1982. 27. Long S, Wertz P, Strauss J, Downing D. Human stratum corneum polar lipids and desquamation. Arch Dermatol Res 277:284–287, 1985. 28. Elias P. Epidermal lipids, barrier function, and desquamation. J Invest Dermatol 80:44s–49s, 1983. 29. Korting H, Greiner K, Hamm G. Changes in skin pH and resident flora by washing with syntetic detergent preparations at pH 5.5 and 8.5. J Sos Cosmet Chem 42:147–58, 1991. 30. Zlotogorski A. Distribution of skin surface pH on the forehead and cheek of adults. J Invest Dermatol 279:398–401, 1987. 31. Lindemann U, Weigmann HJ, Schaefer H, Sterry W, Lademann J. Evaluation of the pseudo-absorption method to quantify human stratum corneum removed by tape stripping using protein absorption. Skin Pharmacol Appl Skin Physiol 16:228–236, 2003. 32. Weigmann HJ, Lindemann U, Antoniou C, Tsikrikas GN, Stratigos A, Katsambas A, Sterry W, Lademann J. UV/VIS absorbance allows rapid, accurate and reproducible determination of corneocytes removed by tape stripping. Skin Pharmacol Appl Skin Physiol 16:217–227, 2003.
for Visualization of Ionic 51 Techniques Gradation in Human Epidermis Mitsuhiro Denda Shiseido Research Center, Yokohama, Japan
CONTENTS 51.1 Introduction............................................................................................................................................................429 51.2 Methods..................................................................................................................................................................429 51.2.1 Chemical Indicators ...................................................................................................................................429 51.2.2 Agarose ......................................................................................................................................................429 51.2.3 Sampling ....................................................................................................................................................430 51.2.4 Process of Ion Visualization ......................................................................................................................430 51.2.5 Image of Ion Gradation in the Epidermis .................................................................................................430 References .......................................................................................................................................................................431
51.1 INTRODUCTION In the epidermis, gradation of ions such as calcium and potassium play an important role in the terminal differentiation or epidermal barrier function.1,2 Thus, observation of these ions would provide us with important information to understand epidermal homeostasis. However, image analysis of ions in the living tissue is technically difficult. The distribution of calcium and potassium has been studied by electron microscopic analysis after calcium precipitation2 or by PIXE analysis3,4 of the skin. Although these gave us important quantitative information, some other important elements, such as hydrogen and magnesium, could not be observed by these methods because of their low atomic weight. Moreover, since these methods require dehydration or fixation of the sample without destroying the native chemical composition, they require complicated processes. The secondary ion MS (SIMS)based imaging technique also can be used for observation of intercellular elements,5 but this method also requires a freeze-drying process. On the other hand, for the cell culture system, various chemical indicators for each element have been used in the cell culture system6 or even living tissue.7 Diffusion of metal ions in water is relatively slow.8 Prevention of diffusion of chemical indicators might allow us to observe the distribution of elements by these chemical probes in a tissue section. Freezing cells or small sections at a low
temperature using organic solvents may preserve the native distribution of diffusible ionic species.5 Polymer gel, such as agarose or polyacrylamide, forms a three-dimensional structure and prevents the water flow inside the structure.9 Thus, one can utilize gel for electrophoresis or in situ zymography. In the agarose gel membrane, the diffusion of the chemical indicators might be prevented and the agarose membrane shows the images of ion distribution. In this section, a new ion imaging method using the chemical indicators and agarose gel is described.10
51.2 METHODS 51.2.1 CHEMICAL INDICATORS For the method, the stability and sensitivity of the ion indicators are important. For observation of calcium, magnesium, sodium, and potassium observation, Calcium Green 1, Magnesium Green, PBFI, and Sodium Green were used. These probes were purchased from Molecular Probes (Eugene, OR). For pH imaging, Bromocresol Green was purchased from Wako (Osaka, Japan).
51.2.2 AGAROSE Mechanical stress-resistant type I-B agarose produced by Sigma (St. Louis, MO) was used. 429
430
Handbook of Non-Invasive Methods and the Skin, Second Edition
51.2.3 SAMPLING Biopsied samples should be immediately frozen in isopentane-filled metal jars, which should be kept in liquid nitrogen to prevent artifactual redistribution.5 The frozen samples are kept at –80˚C until sectioning.
51.2.4 PROCESS
OF ION
VISUALIZATION
Agarose gel (final 2%) containing each indicator was spread on a slide glass with 50 μm thickness. For calcium observation, Calcium Green 1 at a final concentration of 10 μg/ml was mixed before the formation of the membrane. For magnesium, the final concentrations of 10 μg/ml Magnesium Green and 0.2 mM EGTA were mixed together. For potassium, PBFI at a final concentration of 10 μg/ml was mixed. For sodium, Sodium Green at a final concentration of 10 μg/ml was mixed. In the case of hydrogen ion (pH) observation, the first agarose gel membrane was formed and then 20 μl of 0.01% Bromocresol Green ethanol solution was spread over the gel membrane. A frozen section, 5 to 10 μm in thickness, was put on the gel membrane and the whole picture was taken within a couple of hours, because the clear images disappeared by 12 hours after the preparation. We used an Olympus microscope system (AH3-RFC, Olympus, Tokyo, Japan) for observation of the present study. For Calcium Green 1, Magnesium Green, and Sodium Green, the wavelength of the excitation light was 546 nm. For the potassium indicator, PDF1, the wavelength of the excitation light was 334, 365 nm. For each observation, at least five sections were observed to find common features.
51.2.5 IMAGE OF ION GRADATION IN THE EPIDERMIS Images of calcium, magnesium, potassium, and sodium in the skin are shown in the Figure 51.1. These are the representatives of each observation. In the normal skin, calcium was localized in the epidermal granular layer (Figure 51.1A). Thirty minutes after tape stripping, this gradation disappeared (Figure 51.1E). Magnesium also showed the same tendency (Figure 51.1B and F). On the contrary, the concentration of potassium was the highest in the spinous layer and the lowest in the granular layer (Figure 51.1C). This gradation also disappeared after tape stripping (Figure 51.1G). Sodium showed a homogeneous distribution around the whole epidermis (Figure 51.1G), which was not affected by tape stripping. To confirm these results, we also carried out experiments using calcium or magnesium indicators using a 50 mM EDTA solution (Figure 51.2A and B). EDTA absorbed most of the fluorescence in the epidermis. This indicates that the gradation in Figure 51.1 showed the distribution of each ion. Figure 51.3 shows pictures of the spinous layer of the epidermis at a high magnification after tape stripping. Calcium (Figure 51.3A) and sodium (Figure
FIGURE 51.1 Calcium localized in the epidermal granular layer in normal skin (A, white arrow). Thirty minutes after tape stripping, the gradation disappeared (E). Magnesium also showed the same tendency (B, white arrow, before tape stripping; F, after tape stripping). On the contrary, the concentration of potassium was higher in the spinous layer (C, black arrow), and after tape stripping, this gradation disappeared (G). On the other hand, sodium showed a homogeneous distribution around the whole epidermis (D), which was not affected by tape stripping (H). Bars = 50 μm.
51.3D) distribution showed an obvious pattern. In both cases, these ions were absent in the nucleus. On the other hand, after tape stripping, the magnesium concentration was slightly higher in the nucleus (Figure 51.3B). Potassium did not show a clear distribution pattern (Figure 51.3C). A pH gradation in the skin was observed by the color gradation of Bromocresol Green. The uppermost side of the epidermis appeared yellow (acidic), and the deep side of the epidermis appeared green (neutral) (data not shown, see Denda et al.10). The ion profile has been reported to be altered in various skin diseases.4 An abnormal calcium distribution is observed in psoriatic epidermis and atopic dermatitis.4 Zinc and iron also showed on altered distribution in atopic skin. Ions might play an important role in the pathology of the skin. During the wound healing process, the distribution of magnesium and calcium in the wound fluid are altered in distribution and may activate the cell migratory response.11
Techniques for Visualization of Ionic Gradation in Human Epidermis
431
FIGURE 51.3 The epidermal spinous layer after tape stripping at a higher magnification. The distribution of calcium (A) and sodium (D) showed a heterogeneous pattern. In both cases, ions were absent in the nucleus. On the other hand, the magnesium concentration was slightly higher in the nucleus (B). Potassium did not show a clear distribution pattern (Χ). Bars = 10 μm. FIGURE 51.2 EDTA absorbed the fluorescence of both calcium (A) and magnesium (B) indicators. The same experiment using a 50 mM EDTA solution confirmed the results shown in Figure 51.1C and F. EDTA absorbed the fluorescence of both indicators. This indicates that the images in Figure 51.1 are due to each ion.
The novel method presented here might need further methodological improvement to obtain accurate quantitative information, but it would be useful to investigate the role of the ions in the skin or even other tissue.
REFERENCES 1. Lee, S.H., Elias, P.M., Proksch, E., Menon, G.K., Man, M.Q., Feingold, K.R. Calcium and potassium are important regulators of barrier homeostasis in murine epidermis. J Clin Invest 89, 530–538, 1992. 2. Menon, G.K., Elias, P.M., Lee, S.H., Feingold, K.R. Localization of calcium in murine epidermis following disruption and repair of the permeability barrier. Cell Tis Res 270, 503–512, 1992. 3. Mauro, T., Bench, G., Sidderas-Haddad, E., Feingold, K.R., Elias, P.M., Cullander, C. Acute barrier perturbation abolishes the Ca2+ and K+ gradients in murine epidermis: quantitative measurement using PIXE. J Invest Dermatol 111, 1198–1201, 1998.
4. Forslind, B., Werner-Linde, Y., Lindberg, M., Pallon, J. Elemental analysis mirrors epidermal differentiation. Acta Derm Venereol (Stockh) 79, 12–17, 1999. 5. Chandra, S., Smith, D.R., Morrison, G.H. Subcellular imaging by dynamic SIMS ion microscopy. Ann Chem, 72 104a–114a, 2000. 6. Haugland, R.P. Handbook of Fluorescent Probes and Research Chemicals. Molecular Probes, Eugene, OR, 1996. 7. Kudo, Y., Nakamura, T., Ito, E. A “macro” image analysis of fura-2 fluorescence to visualize the distribution of functional glutamate receptor subtypes in hippocampl slices. Neuro Sci 12, 412–420, 1991. 8. Robinson, R.A., Stokes, R.H. Electrolyte Solutions. Butterworths, London, 1959, pp. 513–515. 9. Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., Watson, J.D. Molecular Biology of the Cell. Gurland Publishing, New York, 1994, pp. 169–172. 10. Denda, M., Hosoi, J., Ashida, Y. Visual imaging of ion distribution in human epidermis. Biochem Biophys Res Commun 272, 134–137, 2000. 11. Grzesiak, J.J., Pierschbacher, M.D. Shifts in the concentrations of magnesium and calcium in early porcine and rat wound fluids activate the cell migratory response. J Clin Invest 95, 227–233, 1995.
52 Skin Chamber Techniques Burton Zweiman University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania
CONTENTS 52.1 Introduction............................................................................................................................................................433 52.1.1 Objective ....................................................................................................................................................434 52.2 Methodologic Principle .........................................................................................................................................434 52.2.1 Skin Blisters...............................................................................................................................................434 52.2.1.1 Induction.....................................................................................................................................434 52.2.1.2 Skin Chambers ...........................................................................................................................435 52.2.1.3 Biopsy of the Blister Base .........................................................................................................436 52.2.1.4 Associated Blood Studies...........................................................................................................436 52.3 Sources of Error.....................................................................................................................................................436 52.4 Correlation with Other Methods ...........................................................................................................................439 52.5 Recommendations..................................................................................................................................................439 Acknowledgment.............................................................................................................................................................440
52.1 INTRODUCTION Investigators have long been interested in exploring the in vivo inflammatory responses in human skin for a number of reasons.1 Important information about normal sequential biologic events, such as host responses against microbial invaders and antigens, can be obtained more feasibly than in other areas of the body, such as the lower airways.2 The clinical investigator would be interested in assessing pathogenic events in a variety of skin disorders.3 The clinical pharmacologist would value a reliable way of determining the effects of pharmacologic and other modulatory agents on both the humoral and cellular manifestations in experimentally induced inflammatory reactions. A side benefit might be an approach to measure delivery of the therapeutic agent to the skin without having to extract it from sizable portions of tissue. The intended end result of these objectives could be an enhanced overall knowledge of human inflammatory responses. Earlier approaches to skin inflammation involved assessment of the gross and histologic responses to agents applied to or injected within the skin. These approaches will be described elsewhere in the book. Although much valuable information can be obtained, as reviewed,1 one cannot obtain samples of the interstitial fluid from the dermis to determine levels of inflammatory mediators, etc. Also, it is very difficult to carry out sequential studies at the same skin site.
Therefore, attempts have been made to obtain sequential samples found in the skin by relatively noninvasive methods, that is, designs that can obtain the desired information without excessively damaging the skin.2 The major limitation to studies of dermal inflammation has been the relatively impermeable barrier posed by the epidermis, particularly the stratum corneum. Therefore, a number of approaches to removing the epidermis have been made (Table 52.1). Earlier attempts to remove the epidermis involved scraping with a scalpel blade, abrasion with a grinding cylinder, or tape stripping.3–5 A number of important observations were made about leukocyte exudative responses on cover glasses or skin chambers appended to these sites; these techniques were limited by the irregular
TABLE 52.1 Some Applications of the Skin Chamber Approach 1. 2. 3. 4. 5.
Sequential assessment of cellular and humoral events in inflammation Comparison of exudative cell responses with those in the underlying dermis Investigation of pathogenic processes in skin diseases Study of modulatory effects of systemic therapies on skin inflammation Assessment of drug delivery to inflammatory skin reactions
433
434
Handbook of Non-Invasive Methods and the Skin, Second Edition
depth of the abrasion and induction of some point bleeding. Some local discomfort and a persistent scar could occur. Therefore, other attempts at less traumatic approaches have been made. Some of these involving the induction of skin blisters will now be described.
52.1.1 OBJECTIVE This objective of this discussion is to describe (1) the several ways in which skin blisters have been induced and their use, (2) the design and use of skin collection chambers, and (3) the application of these approaches in one type of experimentally induced inflammatory reaction (immunoglobulin E (IgE) mediated) and certain skin disorders.
52.2 METHODOLOGIC PRINCIPLE 52.2.1 SKIN BLISTERS 52.2.1.1 Induction It has been recognized that induction of a skin blister with its floor at the dermal-epidermal junction would provide a valuable although quite small amount of interstitial fluid for sampling. In early studies, vesicants such as cantharides were applied to the skin surface. This was replaced by suction techniques2,6 in which blisters could be induced within 2 hours by negative pressures of about 200 mmHg or 0.3 kg/cm2 exerted through devices with openings of various diameters (up to about 15 mm). Some investigators, including our group, have found that blisters can be induced in a shorter time, 60 to 90 minutes, by a combination of relatively gentle heat and suction. The method currently used in our laboratory is: 1. A circular Plexiglas chamber apparatus (constructed especially for us) consists of an outer chamber (70 mm diameter) with two flanges (15 mm) perforated by slots for insertion of a binding strap. This outer chamber also has two outlets into which metal correcting tubes are inserted with a tight seal. Sealed within the outer chamber is an inner chamber (45 mm diameter) in which a 10-mm-diameter circular opening has been drilled in the bottom. Into the top of the inner chamber is drilled an 8-mmdiameter outlet into which a metal connecting tube is tightly sealed. This chamber apparatus is applied firmly to the skin of the forearm using Velcro-tipped binding straps (Figure 52.1). 2. The two outlets in the outer chamber are connected by appropriate tubing to the outlet and inlet ports of a thermostat-controlled (55˚C) heated water pump (Haake, Saddlebrook, NJ), so that this heated water circulates in the outer
FIGURE 52.1 Heat/suction chamber apparatus applied firmly to the forearm of a volunteer subject with a binding strap. The outlet at the top of the inner chamber is connected to a tubing from the vacuum source. The outlets in the outer chamber are connected to entrance and exit tubing attached to a heater water pump.
chamber. The outlet in the top of the inner chamber is then connected to a vacuum line. When negative pressure is conveyed to the inner chamber, suction is exerted on the opening in the bottom of the inner chamber (which is now tightly appended to the skin surface), while the heated water in the surrounding outer chamber heats the local skin surface. The transparent nature of the Plexiglas allows continuous inspection to determine when adequate blisters are formed, seen protruding into the inner chamber through the opening in the bottom. The blister induction apparatus is then removed. We can simultaneously induce two blisters on the forearm of average-size adults with little or no discomfort. Other investigators using devices employing suction only can induce at least three blisters on the forearm, depending on the diameter of the blister.2,6 3. In our lab, we can aspirate about 25 ml from individual blisters. Such small fluid samples can be analyzed for levels of inflammatory mediators if sufficiently sensitive assay systems are available. For example, histamine levels have been compared in the fluid of blisters raised over involved and uninvolved skin in certain skin disorders like urticaria.8 However, there may be a sizable nonspecific release of some mediators into these fluids due to the antecedent trauma resulting from the blister induction. For this reason, some investigators wait 24 hours after blister induction before using the site, to allow this traumatically induced mediator release to be less of a factor. This approach requires protecting the blister sites from
Skin Chamber Techniques
435
inadvertent external trauma with overlying shields. However, in our experience, it is not unusual for blisters to collapse at least partially by 24 hours. 52.2.1.2 Skin Chambers Use of the skin blisters themselves is limited by (1) the factors described above, (2) inability to obtain samples repeatedly from the same blister, and (3) inability to challenge the site of the blister repeatedly with agents like antigens. Therefore, several investigative groups have modified the skin blister technique to use collection chambers.2,6,7,9 The technique used currently by us is: 1. The blisters are inspected carefully after induction, not using any in which there is gross evidence of local bleeding. 2. The blister surface is cleansed with 70% ethanol, and the contents are aspirated and stored at –70˚C for future use when appropriate. 3. The blisters are unroofed aseptically with small forceps and scissors. The blister roof is set aside for use in epidermal cell studies, if appropriate. 4. The blister base is washed thoroughly at least twice with sterile phosphate-buffered saline (PBS) or other isotonic solution as indicated. 5. Circular collection chambers are appended to each blister base by firm crosshatched application of nonirritating tape (e.g., Durapore, 3M Company, Minneapolis, MN), so that a tight seal between the chambers and underlying skin is achieved without causing excessive pressure on underlying structures. Several variations in technique have been used by us and other investigators. a. We have used a variety of containers, all made out of various types of rigid plastic material. An example is shown in Figure 52.2. Each chamber has two rubber-sealed ports. One is used for introduction of fluids into the chamber and their subsequent removal with a tuberculin syringe and 27gauge needle; a separate 27-gauge needle is inserted through the top port just into the interior of the chamber to allow air to evacuate. We generally inject 0.3 or 0.5 ml of fluid into each chamber (depending on the protocol), looking carefully for any sign of leakage at the chamber base. b. We do not use an adhesive to adhere these chambers to the skin and have not encountered major problems during challenges as long as 9 hours if the subjects keep their arms relatively quiet. Other have used adhe-
FIGURE 52.2 A collection skin chamber of another shape appended to the skin blister base with straps of nonirritating tape. A tuberculin syringe with needle attached is inserted in one of the two ports.
sives like the cyanoacrylates Skin Bond® and Bond Fast®, reportedly allowing firm adherence of the chambers to the skin for as long as 48 hours.2,9 Dr. Charlesworth reports that occasionally prominent irritant reactions occur under a metal skin chamber adhered with Skin Bond for longer than 8 hours (personal communication). This can be reduced by placing a plastic film barrier such as Saran® (Dow Chemical) wrap between the metal chamber and the skin. c. At the end of a challenge period the chamber contents are aspirated into a tuberculin syringe with a 27-gauge needle inserted into the port (see above). The volume is measured and the fluid centrifuged (1000 rpm for 10 min at 4˚C) for separation and subsequent study of humoral and cellular components. In cases where the site is to be rechallenged, we find it very important to thoroughly wash the site with PBS or like solution to remove any residual challenge agent, mediators, etc., left on the blister base. Some investigators do this wash with the collection chamber in place. We prefer to remove the chamber and wash the blister base thoroughly before applying a fresh collection chamber. The latter action is a precaution against leaving any material on the inner surface of the first collection chamber. d. In this manner, sequential challenges with the same or different solutions can be carried out for up to 48 hours (see above). To reduce the rate of fibrin formation at the blister base, we have found it helpful to add heparin (final concentration, 10 units/ml) to all solutions
436
Handbook of Non-Invasive Methods and the Skin, Second Edition
placed in the chambers. In previous quality control studies, we had found that this concentration of heparin did not cause gross bleeding in the site or affect levels of mediators released into the overlying skin chambers. e. Depending on the requirements of the individual study, the cell-free chamber fluids are divided into aliquots, stored at –70˚C for further analysis. The cells are resuspended in PBS enumerated with an automated counter (Coulter). Aliquots are cytocentrifuged (Shandon cytocentrifuge) in replicate for differential counts, immunocytochemistry, etc., or functional studies are performed. Since the total cell yield varies considerably depending on the condition of challenge and the donor, it may be necessary to pool cells obtained from replicate challenge sites. f. At the end of the skin chamber incubation, we frequently also assess the population of inflammatory cells exuding into the blister base at that period.10,11 Sterile, well-cleaned circular cover glasses, 10 mm in diameter, are appended with gentle pressure to the blister bases for 15 to 30 minutes (depending on the study). They are then removed carefully with sterile forceps, air dried, stained with Wright’s stain (Figure 52.3), or stored at –20˚C in a wrapped container for future fixation and other types of study. We have found no impressively greater adherence of cells to cover glasses previously coated with human serum albumin before sterilization. Indeed, relatively few cells are found adhering to a second cover glass appended to the site following removal of the first cover glass after 30 min of incubation. 52.2.1.3 Biopsy of the Blister Base It has sometimes been important to assess the histologic patterns in the dermis of the blister bases following completion of the skin chamber challenge at that site. We have found that 3-mm-diameter punch biopsies at the center of the blister bases are remarkably well tolerated when performed after local infiltration with 2% lidocaine solution in a circular pattern around the periphery of the blister base. Such biopsied tissue is washed in PBS and either processed immediately or stored in appropriate fixative or carrying medium, depending on the nature of the study.11,12 Meticulous care should be taken in positioning and sectioning these biopsies since they are quite small and have no epidermal layer that would help in positioning.
FIGURE 52.3 Imprint of exuding leukocytes adhered to a cover glass appended for 30 min to the blister base of a skin chamber site following a 5-hour pollen antigen challenge in a sensitive subject (Wright’s stain, 600× magnification). Generally, over 95% of the exuding cells are granulocytes.
52.2.1.4 Associated Blood Studies It is sometimes desirable to compare patterns/events in the blood obtained during the course of the skin chamber challenge.13 Whenever feasible, we obtain the blood specimens from veins not in the region of the skin chamber (unless the latter is desired). An infusion set needle is inserted in a vein prior to initiation of the skin chamber challenge, kept patent with a heparin solution in low concentration, and capped for future removal of blood specimens if desired. The first 3.0 ml of subsequently aspirated blood is discarded to eliminate any diluting or other effects of heparin in the tubing.
52.3 SOURCES OF ERROR We and other investigators have found the skin chamber approach to be a valuable, relatively noninvasive probe of in vivo human inflammatory events in health and disease. However, there are some limitations and caveats that should be mentioned: 1. Despite attempts at making the blister induction as gentle as feasible, this process does involve some trauma leading to a variable inflammatory response by itself. This may impact more on comparisons of the inflammatory cell exudation at the control site, compared to that at the site challenged with antigen or another agonist.14 We have generally found relatively little in the way of nonspecific release of mast cell mediators like histamine and tryptase at control sites, provided that the blister bases are irrigated thoroughly before skin chambers are appended.
Skin Chamber Techniques
However, granulocyte contents may be released nonspecifically at times. 2. The skin chambers capture only those components that diffuse from the underlying dermis. This may be affected by: a. Molecular weight and charge of the compound. b. Diffusion of the released agents in other directions after release (e.g., absorption into local venules and lymphatics). c. Chemotactic/chemokinetic properties of different inflammatory cell types. For example, we find relatively few eosinophils exuding into the overlying chamber fluid even when they are relatively prominent in the underlying dermis. d. Barriers such as fibrin may act as impediments. Even with heparin in the solutions, we find that fibrin deposition is sometimes prominent in blister bases that have been exposed overnight only to the diluent solution (negative control challenges). Although much of the fibrin can be removed, it still may interfere with passage of mediators and cells from the underlying dermis. e. Inevitable dilution of any humoral agent diffusing from the underlying dermis by the volume of diluent placed in the skin chamber at the outset. Thus, the price one pays for injecting a larger volume of diluent into the skin chamber (to have more fluid for replicate analyses at the end) is lower levels of the mediators assayed. 3. The diffusion of different agonists (placed initially in the skin chamber) into the underlying dermis also may vary considerably. Although the epidermal barrier has been removed, the delivery from the skin chamber to the dermis may not be as rapid as by intradermal injection, particularly for higher-molecular-weight compounds. For example, we had found that antigen skin chamber challenge of sensitive subjects led to a more prolonged release of histamine than the very prominent but relatively short-lived histamine release induced by the nonimmunologic mast cell activator codeine.15 We wondered whether these differences could be due to the much lower molecular weight of codeine than the antigens employed. The former would then diffuse away from the site rapidly, while the antigen solution both entered and left the site more slowly. However, when sites were continuously challenged hourly with fresh antigen and codeine solutions and subsequent removal of the fluid for analysis, the patterns
437
of histamine release were the same as described above; that is, there was an initial peak of histamine release, followed by a plateau of lowerlevel histamine release at the sites repeatedly challenged with antigen. In contrast, repeated codeine challenge induced an equally prominent initial peak of histamine release, a small release in the second hour, and then no more release than at the buffer diluent challenge sites over the next 4 hours (Figure 52.4). These findings suggest that the different patterns of histamine release were due to differing biologic effects of the two agonists placed in the skin chambers, and not a difference in their retention in the dermis.16 Absorption of agonists into the underlying dermis could also be affected by a fibrin layer formed at the blister base after prolonged incubation, as noted above. This is particularly true in the case in sites of IgEmediated allergic skin reactions, where we have found increased fibrin formation.17 4. It is conceivable that some inflammatory mediators diffusing from the underlying reaction site into the skin chamber fluid over a period of hours may be metabolized to some degree by enzymes released locally by accumulating inflammatory cells such as histaminase, acetylhydrolyses, etc. Our early quality control studies indicated that exogenous histamine placed into skin chambers over blister bases was not degraded over a period of at least 1 hour. However, that environment is not quite the same as an acute inflammatory site containing many more granulocytes. We have found that much of the leukotriene B4 released into skin chambers overlying IgE-mediated reactions was omega transformed.18 This is additional evidence that the levels of particular mediators in these skin chamber fluids are underestimates of the concentrations released in the underlying reaction site. 5. Because there is increased local vascular permeability in most inflammatory reactions secondary to the release of vasoactive mediators, it is not surprising that there is a gradually increasing concentration of serum proteins in the chamber fluids after Ag challenge periods of at least 2 to 3 hours (experience in our lab and that of Dr. R. Gronneberg, Stockholm). Indeed, such plasma proteins may be the substrate for formation of the increased kallikrein levels we have found released at allergic reaction sites.19 However, the elevated serum levels of certain neutrophil and eosinophil components may indicate a contribution of serum
438
Handbook of Non-Invasive Methods and the Skin, Second Edition
Skin chamber histamine levels (ng/ml.)
100
80
60 Histamine-Ag Histamine-Cod Histamine-Buff 40
20
0 1
2
3
4
5
6
Duration of skin chamber challenge (hours)
FIGURE 52.4 A graphic illustration of typical temporal patterns of histamine release into skin chambers overlying sites of hourly blister base challenge with pollen antigen (Ag), codeine (Cod), and buffer diluent (Buff) in a very sensitive subject. Fluids were removed hourly for histamine analysis and were replaced with fresh solutions of the like substance.
protein leak, as well as release in the underlying dermis to the increased levels of proteins, like major basic protein, we have observed in the overlying skin chamber fluids.11 Thus, it is very important to get an estimate of such serum protein leaks by concomitant measurement of the chamber fluid and serum levels of other serum proteins (e.g., albumin, IgG) that one would not expect to be released within the dermis. The presence of serum proteins in the chamber fluid may affect functional measurement of certain enzymes (e.g., chymase), which are readily inhibited by certain serum proteins. Indeed, it was only because our colleague, Dr. Allen Kaplan, utilized a technique that assayed activated Hageman factor and kallikrein bound to their serum inhibitor (C1 — esterase inhibitor) that we could detect elevated levels of these proinflammatory compounds in skin chambers overlying sites of allergic reactions.19 Serum proteins present in the chamber could also act as a chemotactic agent, likely due to C5a, although such effects are generally not prominent until the serum comprises at least 50% of the total chamber fluid volume.5,6 Some investigators report that the leukocytes recovered from serum-containing chambers were deactivated for in vitro chemotaxis, compared to autologous blood leukocytes.20 However, we
have found similar in vitro responses of chamber fluid and autologous neutrophils for fMLP and activated serum, but not PAF.21 6. As in many in vivo approaches, there is some variability in sequential skin chamber responses at replicate sites in the same individual. Scheza and Forsgren6 found about 19% variability in leukocyte exudation if single sites were challenged sequentially with autologus serum; this variability was reduced to 14% of the mean of duplicate chambers studied each time point was used in the comparisons. We have generally found <10% variation in histamine release in duplicate chambers challenged with antigen in sensitive subjects at the same time. 7. There is the possibility that a prominent reaction occurring at one skin chamber site may affect the levels of mediators seen in an adjacent site if the two are placed close to one another. We have not observed this using two sites/forearms as long as the sites reasonably spaced with the control challenge sites were always distal to the site containing active agonist. However, Charlesworth reports contamination of adjacent site fluids by reactions induced by high antigen concentrations, such as 1000 PNU /ml of pollen extracts (personal communication). He also has observed less histamine release and cellular
Skin Chamber Techniques
exudation in the distal of two similar antigen challenge sites in the same arm. 8. In an attempt to reduce the early effects of blister induction on mediator release, some investigators induce the blister 24 hours before unroofing and application of the collection chamber, as described above. However, this may lead to greater leukocyte exudation at the control site than when the chambers are placed soon after blister induction.
52.4 CORRELATION WITH OTHER METHODS It is difficult to compare the skin chamber approaches just described with other in vivo methods because there is no closely paralleled technique. However, we and others have compared the extent of humoral and cellular responses in skin chamber overlying inflammatory sites with other in vivo and in vitro parameters. For example, we have found that the total amount of histamine and the number of leukocytes exuding into skin chambers overlying allergic challenge sites each correlated with the size of the gross late phase reaction seen 6 hours after intradermal injection of the same antigen in the same subject.22 There was also a correlation of the degree of skin chamber kallikrein formation with gross late reactivity. In contrast, we have found that the amount of histamine released in skin chambers overlying antigen-induced IgE-mediated reactions does not correlate precisely with the immediate wheal and flare response to intradermal antigen. This variance may be due to factors such as mast cell or basophil releasibility, end organ response to histamine, and even the gender of the subject.23,24 Scheza and Forsgren20 found no correlation between in vivo mobilization of leukocytes into chambers filled with autologous serum and the in vitro chemotaxis of blood leukocytes from the same subject to autologous blood. We have also been impressed that the degree of granulocyte exudation and granule protein release into the skin chamber does not always parallel the extent of neutrophil and eosinophil accumulation in biopsies of the underlying dermis. The latter may be due to variable degrees of leukocyte activation in such reaction sites.25
439
time (generally not more than 1 year). Although scarring is not obvious since the local dermis is preserved, one should avoid studying individuals known to form keloids. 2. At least semiquantitative with fairly good reproducibility. 3. Allows measurement of multiple mediators in the same fluid (depending on the sensitivity of the assays) and sequential assessment over a period up to 48 hours. 4. Allows measurement of levels of therapeutic agents given orally in studies of drug delivery to the skin. Can also investigate local effects of topically applied drugs. On the other hand, there are some limitations to keep in mind: 1. Modest inflammatory effects secondary to the blister induction. 2. Chamber fluid levels of mediators are indirect measures of those in the underlying dermis and are likely underestimates. 3. Effects of local enzyme action and exuding serum proteins on some of the mediators released into the chamber fluid and then awaiting aspiration. With these factors in mind, one should standardize the skin chamber approach as much as feasible, at least within the individual laboratory. Then, I feel that it would be a valuable probe for study of: 1. Certain skin diseases, e.g., chronic urticaria26 2. Therapeutic modulators, e.g., corticosteroid,27,28 antihistamines2,29 3. Molecular and cellular mechanisms underlying the pathogenesis of inflammatory reactions in health and disease The use of skin chambers in a number of investigative approaches is listed in Table 52.1. Hopefully, with technical modifications leading to improvement, the skin chamber model will be of increasing value to basic and clinical investigators.
52.5 RECOMMENDATIONS To summarize, some of the favorable aspects of the skin chamber model are: 1. Relatively noninvasive and only modestly traumatic, with little in the way of discomfort and local sequella. The blister base will heal rapidly, leaving a pigmented area for varying lengths of
ACKNOWLEDGMENT My great appreciation is extended to my laboratory colleagues who helped in the development and use of this skin chamber approach over the years. Particular gratitude goes to Mrs. Carolyn von Allmen, whose careful and thoughtful approaches have contributed to valuable modifications and productive research.
440
Handbook of Non-Invasive Methods and the Skin, Second Edition
REFERENCES 1. Zweiman, B., Mediators of allergic inflammation in the skin, Clin. Allergy, 18, 419, 1988. 2. Michel, L. and Dubertret, L., Skin models for analysis of IgE dependent reactions in vivo, in Advances in Allergology and Clinical Immunology, Godard, Ph., Bousquet, J., and Michel, F.B., Eds., Parthenon Press, Carnforth, 1992, p. 505. 3. Rebuck, J.W. and Crowley, J.H., A method of studying leukocyte functioning in vivo, Ann. N.Y. Acad. Sci., 59, 757, 1955. 4. Senn, H., Holland, J.F., and Bannerjee, T., Kinetic and comparative studies of localized leukocyte mobilization in normal man, J. Lab. Clin. Med., 74, 742, 1969. 5. Goldberg, B.S., Weston, W.L., Kohler, P.F., Harris, M.B., and Humbert, J.R., Transcutaneous leukocyte migration in vivo: cellular kinetics, platelet and C5a dependent activity, J. Invest. Dermatol., 72(5), 248–252, 1979. 6. Scheza, A. and Forsgren, A., A skin chamber technique for leukocyte migration studies: description and reproducibility, Acta Path. Microb. Immunol. Scand., C93, 25, 1985. 7. Dunksy, E.H. and Zweiman, B., The direct demonstration of histamine release in the skin using a skin chamber technique, J. Allergy Clin. Immunol., 62, 167, 1978. 8. Kaplan, A.P., Horakova, A., and Katz, S., Assessment of tissue fluid histamine levels in patients with urticaria, J. Allergy Clin. Immunol., 61, 350, 1978. 9. Charlesworth, E.N., Heard, A.F., Soter, N.A., KageySobotka, A., Norman, P.S., and Lichtenstein, L.M., Cutaneous late phase response to allergen. Mediator release and inflammatory cell infiltration, J. Clin. Invest., 83, 1529, 1989. 10. Ting, S., Zweiman, B., Lavker, R.M., and Dunsky, E.H., Histamine suppression of in vivo eosinophil accumulating and histamine release in human allergic reactions, J. Allergy Clin. Immunol., 61, 65, 1981. 11. Zweiman, B., Atkins, P.C., von Allmen, C., and Gleich, G.J., Release of eosinophil granule proteins during IgEmediated allergic skin reactions, J. Allergy Clin. Immunol., 87, 984, 1991. 12. Zweiman, B., Lavker, R.M., Presti, C., and Atkins, P.C., Comparisons of inflammatory responses in IgE-mediated and codeine-induced skin reactions, J. Allergy Clin. Immunol., 91, 963, 1993. 13. Taylor, M., Zweiman, B., Moskovitz, A., von Allmen, C., and Atkins, P.C., Platelet activating factor and leukotriene B4 induced release of lactoferrin from blood neutrophils of atopic and non-atopic individuals, J. Allergy Clin. Immunol., 86, 740, 1990. 14. Bedard, P.M., Zweiman, B., and Atkins, P.C., Quantitation by myeloperoxidase assay of neutrophil accumulation at the site of in vivo allergic reactions, J. Allergy Clin. Immunol., 3, 94, 1983.
15. Shalit, M., Schwartz, L.B., von Allmen, C., Atkins, P.C., Lavker, R.M., and Zweiman, B., Release of histamine and tryptase during continuous and interrupted cutaneous challenge with allergen in humans, J. Allergy Clin. Immunol., 86, 117, 1990. 16. Atkins, P.C., von Allmen, C., Moskovitz, A., Valenzano, M., and Zweiman, B., Fibrin formation during ongoing cutaneous allergic reactions. Comparison of responses to antigen and codeine, J. Allergy Clin. Immunol., 91, 956, 1993. 17. Atkins, P.C., Kaplan, A.P., von Allmen, C., Moskovitz, A., and Zweiman, B., Activation of the coagulation pathway during ongoing allergic cutaneous reactions in humans, J. Allergy Clin. Immunol., 89, 552, 1992. 18. Shalit, M., Valone, F.H., Atkins, P.C., Ratnoff, W.D., Goetzl, E.J., and Zweiman, B., Late appearance of phospholipid platelet activating factor and leukotriene B4 in human skin after repeated antigen challenge, J. Allergy Clin. Immunol., 83, 691, 1989. 19. Atkins, P.C., Miragliotta, G., Talbot, S.F., Zweiman, B., and Kaplan, A.P., Activation of plasma Hageman factor and kallikrein in ongoing allergic reaction in the skin, J. Immunol., 139, 2744, 1987. 20. Scheza, A. and Forsgren, A., Functional properties of polymorphonuclear leukocytes accumulated in a skin chamber, Acta Path. Microb. Immunol. Scand., C43, 31, 1985. 21. Fleekop, P.D., Atkins, P.C., von Allmen, C., Valenzano, M., Shalit, M., and Zweiman, B., Cellular inflammatory response in human allergic skin reactions, J. Allergy Clin. Immunol., 80, 140, 1987. 22. Bedard, P.M., Zweiman, B., and Atkins, P.C., Antigen induced local mediator release and cellular inflammatory responses in atopic subjects, J. Allergy Clin. Immunol., 71, 394, 1983. 23. Atkins, P.C., von Allmen, C., Valenzano, M., Olson, R., Shalit, M., and Zweiman, B., Determinants of in vivo histamine release in cutaneous allergic reactions in humans, J. Allergy Clin. Immunol., 86, 371, 1990. 24. Atkins, P.C., von Allmen, C., Valenzano, M., and Zweiman, B., The effects of gender upon allergen induced histamine release in ongoing allergic cutaneous reactions, J. Allergy Clin. Immunol., 91, 1031, 1993. 25. Zweiman, B., Kucich, U., Shalit, M., von Allmen, C., Moskovitz, A., Weinbaum, G., and Atkins, P.C., Release of lactoferrin and elastase in human allergic skin reactions, J. Immunol., 144, 3953, 1990. 26. Bedard, P.M., Brunet, C., Pelletier, G., and Hebert, J., Increased compound 48/80 induced local histamine release from nonlesional skin of patients with chronic urticaria, J. Allergy Clin. Immunol., 78, 1121, 1986. 27. Atkins, P.C., Schwartz, L.B., Adkinson, N.F., von Allmen, C., Valenzano, M., and Zweiman, B., In vivo antigen induced cutaneous mediator release: simultaneous comparisons of histamine, tryptase and prostaglandin D2 release and the effect of oral corticosteroid administration, J. Allergy Clin. Immunol., 86, 360, 1990.
Skin Chamber Techniques
28. Charlesworth, E.N., Sobotka, A., Schleimer, R.P., Norman, P.S., and Lichtenstein, L.M., Prednisone inhibits the appearance of inflammatory mediators and the influx of eosinophils and basophils associated with the cutaneous lab phase response to allergen, J. Immunol., 146, 671, 1991.
441
29. Charlesworth, E.N., Kagey-Sobotka, A., Norman, P.S., and Lichtenstein, L.M., Effects of cetirizine on mast cell mediator release and cellular traffic during the cutaneous late phase response, J. Allergy Clin. Immunol., 83, 905, 1989.
Methodology for 53 Microdialysis Sampling in the Skin Lotte Groth LEO Pharma, Ballerup, Denmark
Patricia García Ortiz and Eva Benfeldt Department of Dermatology, University of Copenhagen, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 53.1 Introduction............................................................................................................................................................443 53.2 Principle of Microdialysis .....................................................................................................................................444 53.2.1 Recovery ....................................................................................................................................................444 53.3 Microdialysis Instrumentation ...............................................................................................................................445 53.3.1 Microdialysis Probes .................................................................................................................................445 53.3.2 Perfusate.....................................................................................................................................................446 53.3.3 Microdialysis Pumps and Other Equipment .............................................................................................446 53.3.4 Analysis of the Dialysate ..........................................................................................................................446 53.4 Experimental Setup................................................................................................................................................446 53.4.1 In Vitro Setup .............................................................................................................................................446 53.4.2 Preparation for Animal Experiments.........................................................................................................446 53.4.3 Preparation for Human Experiments.........................................................................................................446 53.4.4 Insertion .....................................................................................................................................................447 53.4.5 Insertion Trauma and the Invasiveness of Microdialysis for Dermal Sampling......................................447 53.4.6 Ultrasound Scanning of Probe Depth and Skin Thickness ......................................................................447 53.5 Validation and Interpretation of Dialysate Concentrations...................................................................................448 53.5.1 Calculations of True Tissue Concentrations .............................................................................................448 53.6 Areas of Application..............................................................................................................................................449 53.6.1 Topical, Regional, or Transdermal Drug Delivery....................................................................................449 53.6.2 Inflammation, Allergology, and Physiology..............................................................................................449 53.6.3 Systemic Drug Delivery Including Target Organ Measurements.............................................................450 53.6.4 Metabolism and Homeostasis....................................................................................................................450 53.7 Advantages and Challenges of Microdialysis Methodology ................................................................................450 53.7.1 Troubleshooting .........................................................................................................................................451 53.8 Future Research Directions ...................................................................................................................................451 53.8.1 Bioequivalence Potential and Regulatory Aspects....................................................................................451 References .......................................................................................................................................................................451
53.1 INTRODUCTION Microdialysis is a technique for sampling of endogenous and exogenous substances in the extracellular space in the living tissue. The technique was originally developed in neuropharmacological sciences,1 but has since been used extensively in other tissues in animal models and human studies.2–6
The first papers concerning cutaneous microdialysis were published in 1991,7 and since then the technique has developed into a very versatile tool for skin research. Studies have investigated basic physiology and endogenous substances, as well as the pathophysiology of inflammation and allergic responses, pharmacokinetics and pharmacodynamics of topical and systemic drugs,
443
444
Handbook of Non-Invasive Methods and the Skin, Second Edition
skin barrier function and drug penetration into the skin, and more. Cutaneous microdialysis sampling in the skin provides continuous real-time monitoring of biochemical events in the tissue, and it is performed under minimally invasive conditions in comparison with traditional, invasive methods (skin biopsy, skin stripping, skin blisters) where chronological or dynamic studies are either not feasible or very difficult to conduct. The present chapter aims to give an overview of the principles of cutaneous microdialysis, the experimental setup, and the different applications in skin research. Finally, the advantages of the technique, the pitfalls and potential problems, and the future research opportunities using microdialysis sampling in the skin will be discussed.
53.2 PRINCIPLE OF MICRODIALYSIS The basic principle is to mimic the passive function of a small blood vessel by perfusion of a thin dialysis tube, which has been implanted in the tissue. A dialysis membrane is a semipermeable hollow membrane, permeable to molecules smaller than the openings in the membrane material, the so-called cutoff value (Table 53.1). When equipped with an inlet and outlet connection, this flowthrough membrane device is called a microdialysis probe. The microdialysis probe is continuously perfused with a perfusate at a very low flow rate (0.5 to 10 μl/minute). Diffusion of substances across the dialysis membrane will occur by passive diffusion, driven across the membrane by the concentration gradient. Using continuous flow through the probe, the compounds will be recovered by the membrane and sampled for further analysis. As substances can diffuse in both directions across the membrane (from the tissue and into the lumen or vice versa, if the compound is present in the perfusate), substances can be either collected from or delivered to the tissue by microdialysis.
53.2.1 RECOVERY The constant flow through the probe results in incomplete equilibration with the surrounding medium. Therefore, the
TABLE 53.1 Different Types of Microdialysis Membranes Microdialysis Membranes: Cutoff Values Low flux, 2 –20 kDa High flux, 40–100 kDa Plasmapheresis, 3000 kDa
concentration in the dialysate is not equal to the undisturbed concentration in the surrounding medium, but lower. The term relative recovery is defined as the ratio between the concentration in the dialysate (Cd) and the concentration in the medium surrounding the probe (Cm): Relative Recovery =
Cd Cm
Relative recovery is independent of concentration. The absolute recovery is defined as the total amount of drug collected in the dialysate during a defined period. As diffusion across the membrane is quantitatively equal in both directions, depending on the concentration gradient, substances can be delivered to the surrounding medium. Loss (or delivery) is the ratio of the loss in the perfusate concentration relative to the initial concentration in the perfusate:
Loss =
Cp – Cd Cp
where Cp is the concentration in the perfusate. Theoretically, recovery and loss are equal; however, for very lipophilic substances this is not always true.8 The recovery of a substance by microdialysis sampling depends on several parameters, both biological and experimental (Table 53.1 and Table 53.2). Relative recovery increases when the perfusate flow is reduced, as lower flow rates allow more time for equilibration across the dialysis membrane.9 The molecular weight, electrical charge, lipophilicity, and protein binding of the dialyzed substance will affect recovery. The blood flow in the tissue can affect the clearance, and thereby the drug
TABLE 53.2 Parameters Affecting Recovery In Vitro and In Vivo Factors Affecting Relative Recovery Molecular weight of the dialyzed compound Lipophilicity Protein binding Perfusate flow rate Probe membrane material, design, and length Perfusate Tissue Blood flow Temperature Metabolism Probe depth in dermis
Microdialysis Methodology for Sampling in the Skin
TABLE 53.3 Necessary Steps in Preparation of Human Microdialysis Experiments 1.
2.
3.
In vitro recovery and loss: linearity Ensure reproducible, stable recovery over a concentration range and time Establish that analysis is sensitive enough in the low concentration range Consider Choosing a calibrator now The effect of probe modifications introduced at a later stage (guidewire in lumen, sterilization procedures) In vivo recovery is likely to be << than in vitro recovery If studies of drug–protein binding effects or animal studies are needed Human protocol work Include pilot experiments and await analysis of pilot samples Amend setup (flow rate, probe type, perfusate composition) if necessary
concentration in the tissue, and thus also the gradient over the membrane.10,11 The ultimate proof of this is the increased dialysate concentrations found in a dead or dying rat.12,13 Although relative recovery is independent of the concentration in the medium or tissue, according to the diffusion theory, this concentration independency can sometimes be hampered by interaction between the substance investigated and the dialysis membrane and tubing (see Section 53.7.1). Thus, before starting in vivo microdialysis studies, the concentration independency of each new substance should be verified in vitro for a wide range of concentrations and for both directions across the membrane (recovery and loss). All the above-mentioned factors should be taken in to account when designing microdialysis experiments. It is important to standardize all parameters possible in order to reduce variability and optimize the recovery of the substance investigated. It is advisable to perform in vitro studies to evaluate the level of recovery and reproducibility, and small pilot studies before proper and larger animal or clinical studies are initiated (consult Table 53.3).
445
Perfusate flow
Nylon tube 3 cm accessible microdialysis fibre 0.22 mm OD
FIGURE 53.1 Principle of microdialysis sampling by a linear probe. The perfusate is pumped through the probe at a preset, low flow rate. During the passage through the membranaceous portion, which can be from 1 to 4 cm long, the diffusion of small molecules across the membrane takes place. The perfusate is now termed the dialysate.
hemodialysis cylinder. This probe type is simple to manufacture in the laboratory and therefore very inexpensive, but will require both an entrance and an exit puncture through the skin when placed in the tissue by means of a guide (introductory) cannula. In contrast, the commercially available concentric probe (Figure 53.2) requires one entrance puncture only, however, often with a guide cannula of larger dimensions. Different membrane materials with different nominal cutoff values (Table 53.1) are
53.3 MICRODIALYSIS INSTRUMENTATION 53.3.1 MICRODIALYSIS PROBES As mentioned, a dialysis membrane connected to inlet and outlet tubing is termed a microdialysis probe. There are two main types of probe design. The first is the linear probe (Figure 53.1), where the dialysis membrane is a hollow fiber, often taken from an artificial kidney, i.e., a
FIGURE 53.2 The concentric probe (commercially available). The perfusate is pumped to the tip of the probe through an inner canal and flows back in the external, circular canal, during which transport the diffusion across the membrane takes place.
446
Handbook of Non-Invasive Methods and the Skin, Second Edition
used in both types of probes, depending on the substance of interest. Most membranes have cutoff values of less than 20,000 Da, but 3000-kDa membranes exist.14 These are termed plasmapheresis probes, since the cutoff allows proteins to diffuse across the membrane. Thus, the dialysate will contain protein material in contrast to the purified dialysate obtained by regular microdialysis sampling. Attention should be paid to avoid air-exposed membrane material, during sampling, as this will lead to fluid loss by evaporation. The membranes should be biocompatible, and no interaction between the membrane material and the dialyzed compound should occur (see later). The effect on recovery when altering probe design or material, such as, e.g., employing a guidewire in the lumen,15 or sterilization procedures should be considered (Table 53.3).
technique. Very sensitive analytical methods like HPLCMS, HPLC-MS/MS, GC-FPD, or GC-MS are required.20,21 It is possible to partially compensate for inadequate analytical sensitivity by employing long sampling intervals, low flow rates, or by pooling dialysates from several probes. However, the experimental parameters are often determined by a compromise between the analytical sensitivity and the characteristics of the substance investigated. It will often be beneficial to plan and conduct studies in close collaboration with a dedicated analyst in order to optimize both the sampling methodology and the analysis of the microdialysis samples prior to in vivo experimentation (Table 53.3).
53.3.2 PERFUSATE
53.4 EXPERIMENTAL SETUP
The perfusion medium should resemble the tissue/solution surrounding the probe as much as possible. The perfusion medium used in cutaneous microdialysis is most often a Ringer solution or an isotonic saline solution with or without added glucose content. For sampling of very lipophilic compounds, the perfusate can be modified by adding albumin,16 cyclodextrins, or intralipid solution,17 and for protein-bound drugs, a protein content in the perfusate will increase recovery. Regarding the use of calibrator substances, added to the perfusate, or establishing relative recovery by loss of drug (retrodialysis by drug), refer to Section 53.5.
53.4.1 IN VITRO SETUP
53.3.3 MICRODIALYSIS PUMPS EQUIPMENT
AND
OTHER
The microdialysis pumps are essentially ultraprecise syringe drivers, and a number of commercially available pumps exist. The perfusate flow rate delivered by the pump should be checked regularly by gravimetric control of volume delivery in simple experiments conducted at different flow settings. For some purposes, a computercontrolled sampling unit can be helpful, and equipment for online analysis by, e.g., high-pressure liquid chromatography (HPLC), is also in use in some clinical settings.
53.3.4 ANALYSIS
OF THE
DIALYSATE
The analysis of the samples can be one of the biggest challenges using cutaneous microdialysis.18,19 In the dialysate, the concentration of the substance of interest can be very low and the sample volumes are small. The typical microdialysis membranes have low-molecular weight cutoff values (2 to 20 kDa), and samples will thus be free from protein with no need for sample cleanup and no enzymatic degradation of the substances collected. The limitations for microdialysis sampling experiments are often determined by the sensitivity of the analytical
The purpose of in vitro experiments is to validate the microdialysis probe and perfusate chosen by systematical investigations, establishing that recovery is concentration independent, stable, and reproducible. Flow rate and sampling intervals should be optimized for the subsequent in vivo sampling situation. Whenever possible, in vitro experiments should be conducted at 34 to 37˚C to mimic the temperature of the living tissue. Attention should be given to avoid the effect of an unstirred water layer in the medium; magnetic stirring should be employed as demonstrated necessary by Groth.13 For further steps in preparation for in vivo sampling, see Table 53.3.
53.4.2 PREPARATION
FOR
ANIMAL EXPERIMENTS
Animals will have to be anesthetized during the insertion of the probes and most often during the whole experiment. Equipment for microdialysis experiments in, e.g., hairless rats (such as temperature-controlled heating pads), is commercially available. Care should be taken to keep both core temperature and skin temperature stable. Consider the possible fluid loss during the experiment. If shaving is necessary prior to the experiment, this should be done 2 to 3 days in advance to allow for recovery of the skin barrier function.
53.4.3 PREPARATION
FOR
HUMAN EXPERIMENTS
If the volar aspect of the forearm is used (as in most studies), some male volunteers may need to shave the area to facilitate probe insertion, and again, this is best done 2 to 3 days prior to the experiment in order for the skin barrier function to normalize. The skin area should be prepared by gentle washing at the start of the experiment, the entry and exit puncture site for the guide cannula should be swabbed with ethanol,
Microdialysis Methodology for Sampling in the Skin
and the person inserting the probes should wear sterile gloves. Insertion of probes can be done following local anesthesia by s.c. injections of lidocaine (most studies), following topical anesthesia with EMLA cream22 with aid of hypnosis23 or without prior anesthesia.24 In general, EMLA anesthesia should be avoided in studies of skin barrier function or topical drug penetration due to the possible impact on barrier function and a vasoconstrictory response.
53.4.4 INSERTION For linear probes, probes are inserted using a 21-G guide cannula. The cannula is inserted in the skin through a marked entry point, forced gently through the dermis or subcutaneous tissue for the desired length, and then pushed out through a premarked exit puncture. After sliding the probe through the guide cannula from the free tip and inwards, the guide cannula is withdrawn, leaving the membranaceous part of the probe placed horizontally within the dermis (Figure 53.3). Alternatively, the probe can be placed subcutaneously. Attention should be paid to placing the probe with the entry at a higher level than the exit, securing a favorable hydrostatic pressure. This is particularly important when working with high-flux (i.e., large-pore) membranes that are very pressure sensitive.
53.4.5 INSERTION TRAUMA AND THE INVASIVENESS OF MICRODIALYSIS FOR DERMAL SAMPLING The insertion of the microdialysis probe creates a reversible trauma. The insertion causes an acute inflammatory reaction and changes the diffusion characteristics of the tissues, and this in turn may alter recovery. The trauma must thus have subsided before sampling can be started;
447
it is therefore essential to determine an appropriate equilibration period. Increased skin blood flow, erythema, and skin thickness were demonstrated in both rats and humans after insertion of a linear microdialysis probe by a 21-G guide cannula. An equilibration period of minimum 90 minutes in human skin was required to allow the skin to normalize.24 Anderson et al.25,26 found a normalization of the increased blood flow in humans 60 minutes after the insertion of a concentric probe, and a normalization of increased histamine levels after 40 minutes. Petersen et al.27 found hyperemia for up to 90 to 135 minutes after implantation of a linear probe, and suggested a 90-minute stabilization period in human skin. Krogstad et al.28 did not find complete normalization of skin perfusion after insertion of a linear probe. The use of local anesthesia has been found to reduce the vascular effects of the trauma in man in several studies.24,26,29 In the rat, the trauma effect in the skin requires an equilibration period of only 30 minutes to normalize,30,31 which is significantly shorter than human skin. In the rat, Mathy FX has showed that neither insertion of linear probes in the dermis by a 26-G (0.45-mm) guide cannula nor probe insertion by a 16-G cannula in the subcutaneous tissue will alter skin permeability, blood flow, or color.32 Histological examination of skin biopsies showed no signs of an inflammatory reaction in the tissue around the probe 8 to 10 hours after probe insertion.28 Immediately after probe insertion in rat skin, no substantial tissue changes were evident in a histology study. However, after 6 hours infiltration with lymphocytes and after 32 hours elongation of the connective tissue were observed.12 The trauma inflicted by the insertion varies with the species, type of probe, membrane material, possibly insertion depth in the dermis, and, in particular, diameter of the guide cannula. The minimum equilibration period must be considered in each new study design prior to conducting the experiment.
53.4.6 ULTRASOUND SCANNING AND SKIN THICKNESS
FIGURE 53.3 The insertion trauma, consisting of a histaminerelated wheal and flare reaction. The reaction must subside before reliable, stable sampling can be performed, since the increased capillary permeability and blood flow will alter both recoveries and drug concentrations (see text).
OF
PROBE DEPTH
At the end of the microdialysis experiment, the probe depth in the skin can be measured by 20-MHz ultrasound scanning using, e.g., the Dermascan-C (Cortex, Hadsund, Denmark), which has an accuracy of around 0.02 mm in the measurement of skin thickness (consult Chapter 56 of this handbook). It is recommended to measure skin thickness and probe depth (Figure 53.5) in three separate scans along the length of the probe in situ (near probe entry, middle, and near probe exit) and use the mean for calculations. For studies of topical drug administration, the influence of the probe depth (the distance from the surface of the skin to the microdialysis membrane inserted in the
448
Handbook of Non-Invasive Methods and the Skin, Second Edition
Microdialysis after topical drug administration
drug molecule
FIGURE 53.4 Microdialysis sampling by a linear probe, placed in the dermis. Topically applied drugs will diffuse through the epidermis to the dermis, where microdialysis sampling collects a fraction of the unbound drug concentration in the interstitial fluid. No fluid is delivered or extracted in the process.
FIGURE 53.5 Ultrasound scanning image of two microdialysis probes, placed in the dermis of the volar forearm of a human volunteer. Probes are seen as white dots. The depth of the probes, measured from the entrance echo of the epidermis, is 0.60 mm (top) and 0.53 mm (bottom), respectively.
skin) on the drug concentration sampled has been debated. Some authors have shown that more superficial probe implantation results in higher dialysate concentrations,33 and this correlation is plausible from a theoretical point of view and can be seen in Figure 53.4. However, probes are usually inserted over a narrow range of depth (0.6 to 1.0 mm) in the dermis. Furthermore, the implantation depth becomes less variable with increasing investigator experience, reducing standard deviation in implantation depth to, e.g., ±0.16 mm.34 Most studies have been unable to demonstrate the correlation between probe depth and topical drug penetration.35–38
53.5 VALIDATION AND INTERPRETATION OF DIALYSATE CONCENTRATIONS 53.5.1 CALCULATIONS OF TRUE TISSUE CONCENTRATIONS Due to the continuous perfusion of the microdialysis probe, the equilibrium between the tissue and the perfusate in the probe is never complete, with the concentration of the substance of interest always being greater in the tissue than in the dialysate.
Several methods, theoretical and experimental approaches, have been developed to estimate in vivo recovery and thereby enable quantification of the unbound extracellular concentrations from the dialysate concentrations.39–41 It is not correct to use in vitro recovery values in calculation of the extracellular concentration, since in vivo recoveries are lower than the in vitro recoveries.42 The diffusivity of substances in the living tissue is lower due to the tortuosity of the path and the limited volume fraction of the extracellular space. The three most used in vivo calibration methods are the difference method,4 the retrodialysis method,43,44 and the reference method.45–48 The difference method (or the point of no-net-flux method) is based on the net transport across the membrane being driven by the concentration gradient only. The compound of interest is added to the perfusate, and experiments are conducted with a range of different perfusate concentrations. The plotting of dialysate concentrations will identify the point (perfusate concentration) where no net diffusion across the membrane has occurred, because the concentrations inside and outside the membrane have been equal. This point represents the unbound extracellular concentration. The retrodialysis principle (also called recovery by loss) assesses the loss of the substance of interest from the perfusate and relies on the reversibility of the passive diffusion across the dialysis membrane (as determined by the concentration gradient). The net transport from the perfusate into the surrounding tissue (loss) is established experimentally and is then assumed to equal the net transport from the tissue into the perfusate (recovery). Retrodialysis can be performed before or after the actual experiment or in a separate experiment. Retrodialysis is mainly used for exogenous substances.33 The reference method (or retrodialysis by calibrator) is based on the same principle as retrodialysis, but the loss of a calibrator is used to estimate the recovery of the substance of interest.49 It is thus critical that the loss and recovery characteristics of the two substances are truly equal; therefore, the calibrator must be chemically and biologically similar to the substance of interest. Prior to the in vivo experiment, this similarity should be confirmed by in vitro investigations of both recovery and loss for the substance of interest and the proposed calibrator. In the actual in vivo experiment, the retrodialysis of the substance of interest can be performed in a separate experiment, prior to or following the experiment. Using a calibrator has the advantage of uncovering variations in probe efficacy and drug recovery during the experiment. The loss of calibrator can identify the source of variability in in vivo studies, e.g., in studies of topical drug penetration. Simonsen et al.50 found that the variability in dialysate concentrations following topical drug application was
Microdialysis Methodology for Sampling in the Skin
449
Salicylic acid concentration (μg/ml)
100,00 10,00 1,00 0,10 0,01 0,00 0
10
30
50
70
90
110 130 150 170 190
210 230
Time (min) Tape-stripped
1 % SLS
2 % SLS
Acetone treated
Unmodified
FIGURE 53.6 Pharmacokinetics of topical salicylic acid penetration in human skin, measured by microdialysis sampling in the dermis. Note the logarithmic y-axis scale. The curves show the mean salicylic acid concentration sampled by two probes in each area in 16 volunteers. The effect of barrier perturbation on topical drug penetration can be seen to be major.35
associated with the interindividual variability in skin barrier function since the variability in intradermal probe function, as determined by simultaneous loss of calibrator, was found to be low. The choice of calibration method depends on the aim and design of the microdialysis study. The methods based on the retrodialysis principle are most favorable in cutaneous microdialysis studies. Comparing two topical formulations with the same substance, it may not be necessary to transform the dialysate concentrations to extracellular dermal concentrations, since a simple comparison of the dialysate concentrations as relative values (most often the area under the curve (AUC) parameter) will provide conclusive results. However, in some instances the target tissue concentration is important, since the free drug concentration in the tissue represents the pharmacologically active fraction of the drug.
53.6 AREAS OF APPLICATION 53.6.1 TOPICAL, REGIONAL, DRUG DELIVERY
OR
TRANSDERMAL
The initial microdialysis study of percutaneous penetration was published in 1991 and concerned the penetration of ethanol applied to the skin surface.7 Since then, the research area has branched out to include transdermal medication such as the nicotine patch,36 topical vs. oral administration of the same drug, e.g., ibuprofen,51 salicylic acid,52 or psoralen,53 and topical penetration of dermatological formulations.54,55 Benfeldt et al.35 have studied the highly significant effect of barrier perturbation procedures on cutaneous penetration of salicylic acid (Figure 53.6). The effect of iontophoresis on topical drug delivery to the skin and subsequent systemic delivery has been studied
by Stagni and coworkers.56 For reviews of microdialysis in studies of topically applied drugs and substances, see Schnetz and Fartasch57 or Kreilgaard.3 In some studies, problems with no detectable topical drug in dialysates following drug application onto intact skin have been encountered.54,58 However, in the last study detectable dialysate concentrations were found following stratum corneum reduction by repetitive tape stripping of the skin surface, confirming the need for very sensitive analysis when studies of topical penetration of commercially available formulations through unmodified skin are undertaken.
53.6.2 INFLAMMATION, ALLERGOLOGY, PHYSIOLOGY
AND
Endogenous substances can be sampled in dermis using microdialysis, with the purpose of investigating different mechanisms in normal or diseased skin or in experimental models of skin inflammation. Histamine levels in the skin have been widely investigated in normal skin and diseased skin.22,59,60 Since 1992, Petersen and coworkers27,61–63 have published numerous papers on histamine release after different challenges with allergens or nonimmunological stimulation. Other interesting mediators in the wheal and flare response have been measured using microdialysis: leukotriene C4,64 leukotriene B4, prostaglandin D2,65 prostaglandin E2,66 and nitric oxide.67,68 The proinflammatory cytokine interleukine-6 (IL6) has been measured after the insertion of a microdialysis probe in the dermis.69 Schmelz and coworkers14 have studied the mechanisms of dermal neurogenic inflammation intensively, including the release of neuropeptides and plasma protein extravasation. Using large-cutoff (3000-kDa)
450
Handbook of Non-Invasive Methods and the Skin, Second Edition
plasmapheresis capillaries, it has been possible to measure plasma extravasation elicited by different stimulations of the skin.14,70,71 The release of C-GRP, Substance P, mast cell tryptase, and histamine has been sampled simultaneously with plasma extravasation.72–74 Cutaneous microdialysis of endogenous substances has considerably contributed to our understanding of some of the mechanisms in inflammatory and allergic responses,75 and this research field will undoubtedly continue to grow in the future.
53.6.3 SYSTEMIC DRUG DELIVERY INCLUDING TARGET ORGAN MEASUREMENTS Antibiotics and their distribution kinetics in the subcutaneous layer of the skin and other tissues have been extensively studied by Markus Muller and his group,76 taking the methodology from investigations in healthy volunteers to various diseases and conditions where peripheral tissue pharmacokinetics can be shown to be critically altered. For antineoplastic drugs, the penetration of drugs into tumor or metastatic tissue has been studied and compared with the drug levels in healthy tissues for, e.g., capecitabine and its metabolites.77 In patients with oral cancer, microdialysis in the tumor tissue has been used to characterize which mode of delivery (intraarterial cisplatin infusion vs. embolization) produced the most favorable cisplatin pharmacokinetics in the target tissue.78 A special feature of microdialysis is the possibility of sampling both prodrug and active drug in the same sample during drug metabolization in the body. This has been shown for the metabolization of famciclovir to penciclovir23 and for the very rapid metabolization of orally ingested acetylsalicylic acid to salicylic acid, occurring in the plasma.34
53.6.4 METABOLISM
AND
HOMEOSTASIS
Glucose metabolism is a large microdialysis research field, with initial studies of glucose metabolism and the effect of food ingestion and insulin performed in healthy volunteers,79 and later in patients with chronic diabetic foot ulcers.80 Many studies employ microdialysis methodology or modifications thereof in the development of continuous glucose monitors, portable for up to 3 weeks in patients.81 Recently, insulin has been sampled quantitatively by microdialysis in human muscle.82
53.7 ADVANTAGES AND CHALLENGES OF MICRODIALYSIS METHODOLOGY As can be seen from Table 53.4, microdialysis technology offers a number of possibilities for sampling in novel sites and for creative experimental setups with multiple sampling points.
TABLE 53.4 Advantages and Limitations of Microdialysis Advantages of Microdialysis
Limitations of Microdialysis
Highly dynamic continuous sampling High-resolution real-time sampling Both drug and metabolites in one sample Minimally invasive Multiple sites in one animal/person Sampling or delivery via the probe Purified samples (protein-free) Highly reproducible Allows simultaneous use of auxiliary techniques
Requires an initial investment in pumps Requires training or skills Insertion of probes Probe manufacturing Needs sensitive analysis Drug-specific problems Lipophilic drugs Heavily protein-bound drugs Absolute tissue levels difficult to estimate Recovery is influenced by local blood flow
Lipophilicity of the drug of interest has initially been considered a major problem in microdialysis sampling,16,54 but with the recent advances in analytical methods, the impact of reduced or very low recovery for lipophilic drugs or substances has now been reduced greatly. Indeed, the previously very inaccessible psoralens (with log P values of up to 4.5) have now been sampled reproducibly, demonstrating differentiated dermal pharmacokinetics following drug delivery via either oral administration or following topical cream application or immersion in a bath solution.53 Furthermore, in psoriatic patients, the dermal psoralen levels after oral psoralen administration have been measured in the psoriatic skin.20 For endogenous compounds, the effect of UVB light irradiation on the synthesis of 1α, 25-dihydroxyvitamin D (calcitriol) in human skin83 and on nitric oxide and prostaglandin concentrations66 in the irradiated skin has recently been demonstrated. The impact of alterations in local blood flow (around the probe; Figure 53.4) on the concentrations present in the skin is another exciting research field. The alteration can be induced by a vasoactive drug in itself, e.g., malathion,84 or manipulations of blood flow by s.c. injection of vasoconstrictors23,58 or dilatation by a topically applied glyceryl trinitrate patch on the skin overlying the probe.11 Microdialysis methodology in the skin is easily combined with other methods and has been employed in combination with scanning laser Doppler imaging to measure changes in the skin blood flow and the release of histamine within the wheal and flare.13,26 The two techniques have been shown to be powerful tools together for the investigation of the biochemical mechanisms underlying the control of the skin blood flow.11,84 Microdialysis can be conducted simultaneously with other drug sampling
Microdialysis Methodology for Sampling in the Skin
techniques23,34,85 with noninvasive bioengineering methods such as laser Doppler flowmetry, laser Doppler perfusion imaging, transepidermal water loss (TEWL), colorimetry, high-resolution ultrasound scanning,13,52 scoring of itch or pain, and many more. The technique is still undergoing further development for use in topical studies, and probe types and methodology can be optimized further. Microdialysis methodology continues to provide us with new insights into skin physiology and barrier function, into topical drug penetration and systemic absorption, and into the active skin metabolism of both topical and systemic drugs.
53.7.1 TROUBLESHOOTING A problem that can have a large impact on in vivo experiments, and which can often be identified through very simple in vitro experiments in the planning phase (Table 53.3), is nonlinearity; i.e., the recovery is not the same over a concentration range; the slope of the line drawn when the beaker concentration is plotted vs. the dialysate concentration is not a straight line. The cause of nonlinearity can be phenomena such as drug adherence to plastic vials, tubings, and probe material, as well as drug–drug self-assembly and drug instability in the solution. These can all be considered artifacts and will often recur to some extent in in vivo experiments, where the identification is difficult. Thus, it is important to measure recovery and loss of the substances of interest in vitro prior to initiating in vivo studies.
53.8 FUTURE RESEARCH DIRECTIONS We have seen the first studies that investigate microdialysis sampling simultaneously with other drug sampling methodologies.23,34,85–87 A recent study investigates the relationship between in vitro permeation methodology and ex vivo microdialysis for percutaneous penetration of topically applied drugs,88 and in another ex vivo study, the effect of formulation on the topical penetration of an herbal remedy was investigated by microdialysis.21 The topical penetration of several different components in a topical analgesic solution has also been demonstrated by microdialysis.89 Some of the many possible future directions for microdialysis research are shown in Table 53.5.
53.8.1 BIOEQUIVALENCE POTENTIAL REGULATORY ASPECTS
AND
For bioequivalence studies of topical products several methodologies are used, with the dermato-pharmacokinetic (DPK) method, consisting of repeated tape-stripping harvesting of the outermost layers of the skin, currently being recommended by the Food and Drug
451
TABLE 53.5 Future Aspects of Microdialysis Research Future Aspects of Microdialysis Research Dermal pharmacokinetics and drug metabolism of both topically and systemically administered drugs Species comparisons In vitro/in vivo correlation for studies of topical drug penetration Drug metabolism in the skin and in other target tissues, e.g., tumors Dermal blood flow — impact on drug concentration and delivery Pharmacokinetic — pharmacodynamic studies Topical and transdermal drug development Bioequivalence studies Clinical studies, including drug penetration in diseased skin
Administration (FDA). Microdialysis methodology for assessment of topical drug penetration has initially been considered promising, but in need of validation, by the regulatory authorities.90 We have subsequently seen an increasing number of both animal and human studies of topical drug penetration, conducted with low variability and of adequate size for statistical calculations and conclusions. Based on the initial bioequivalence study in humans, where lidocaine delivery from two different vehicles was compared by dermal microdialysis sampling and pharmacodynamic assessment of the pain-relieving effect of the formulations,55 the microdialysis method is currently receiving increased recognition and positive evaluation.91 Ongoing work aims at obtaining valid estimates of both the variability observed in human microdialysis studies of topical penetration and the sources thereof (Benfeldt and Shah, personal communication). The future will show us the place for microdialysis methodology in drug research and development in the preclinical phase, for regulatory purposes, and finally, most importantly, in clinical studies in patients. Whether the method is employed for drug sampling, studies of inflammation, or homeostasis, the microdialysis technique offers a unique opportunity for creative studies with real-time chronological sampling in the target tissue.
REFERENCES 1. Ungerstedt, U., Measurement of Neurotransmitter Release by Intracranial Dialysis, in Measurement of Neurotransmitter Release In Vivo, Marsden, C.A., Ed., John Wiley & Sons, New York, 1984, pp. 81–105. 2. Chaurasia, C.S., In vivo microdialysis sampling: theory and applications, Biomed. Chromatogr., 13, 317–332, 1999. 3. Kreilgaard, M., Assessment of cutaneous drug delivery using microdialysis, Adv. Drug Deliv. Rev., 54 (Suppl. 1), S99–S121, 2002.
452
Handbook of Non-Invasive Methods and the Skin, Second Edition
4. Lonnroth, P., Jansson, P.A., and Smith, U., A microdialysis method allowing characterization of intercellular water space in humans, Am. J. Physiol., 253, E228–E231, 1987. 5. Muller, M., Science, medicine, and the future: microdialysis, BMJ, 324, 588–591, 2002. 6. Stahl, M. et al., Human microdialysis, Curr. Pharm. Biotechnol., 3, 165–178, 2002. 7. Anderson, C., Andersson, T., and Molander, M., Ethanol absorption across human skin measured by in vivo microdialysis technique, Acta Derm. Venereol., 71, 389–393, 1991. 8. Groth, L. and Jorgensen, A., In vitro microdialysis of hydrophilic and lipophilic compounds, Acta Anal. Chim., 355, 75–83, 1997. 9. Benveniste, H., Brain microdialysis, J. Neurochem., 52, 1667–1679, 1989. 10. Bernards, C.M. and Kopacz, D.J., Effect of epinephrine on lidocaine clearance in vivo: a microdialysis study in humans, Anesthesiology, 91, 962–968, 1999. 11. Clough, G.F. et al., Effects of blood flow on the in vivo recovery of a small diffusible molecule by microdialysis in human skin, J. Pharmacol. Exp. Ther., 302, 681–686, 2002. 12. Ault, J.M. et al., Dermal microdialysis sampling in vivo, Pharm. Res., 11, 1631–1639, 1994. 13. Groth, L., Cutaneous microdialysis. Methodology and validation, Acta Derm. Venereol. (Stockh.), 197 (Suppl.), 1–61, 1996. 14. Schmelz, M. et al., Plasma extravasation and neuropeptide release in human skin as measured by intradermal microdialysis, Neurosci. Lett., 230, 117–120, 1997. 15. Klimowicz, A. et al., Use of an intraluminal guide wire in linear microdialysis probes: effect on recovery? Skin. Res. Technol., 10, 104–108, 2004. 16. Carneheim, C. and Stahle, L., Microdialysis of lipophilic compounds: a methodological study, Pharmacol. Toxicol., 69, 378–380, 1991. 17. Kurosaki, Y. et al., Lipo-microdialysis: a new microdialysis method for studying the pharmacokinetics of lipophilic substances, Biol. Pharm Bull., 21, 194–196, 1998. 18. Lunte, S.M. and Lunte, C.E., Microdialysis sampling for pharmacological studies: HPLC and CE analysis, Adv. Chromatogr., 36, 383–432, 1996. 19. Weiss, D.J., Lunte, C.E., and Lunte, S.M., In vivo microdialysis as a tool for monitoring pharmacokinetics, Trends Anal. Chem., 19, 606–616, 2000. 20. Leveque, N. et al., Validation of a microdialysis-gas chromatographic-mass spectrometric method to assess 8-methoxypsoralen in psoriatic patient dermis, J. Chromatogr. B Anal. Technol. Biomed. Life Sci., 780, 119–127, 2002. 21. Oberthur, C. et al., A comparative study on the skin penetration of pure tryptanthrin and tryptanthrin in Isatis tinctoria extract by dermal microdialysis coupled with isotope dilution ESI-LC-MS, Planta Med., 69, 385–389, 2003. 22. Krogstad, A.L. et al., Nerve-induced histamine release is of little importance in psoriatic skin, Br. J Dermatol., 139, 403–409, 1998.
23. Borg, N. et al., Distribution to the skin of penciclovir after oral famciclovir administration in healthy volunteers: comparison of the suction blister technique and cutaneous microdialysis, Acta Derm. Venereol., 79, 274–277, 1999. 24. Groth, L. and Serup, J., Cutaneous microdialysis in man: effects of needle insertion trauma and anaesthesia on skin perfusion, erythema and skin thickness, Acta Derm. Venereol., 78, 5–9, 1998. 25. Anderson, C., Andersson, T., and Andersson, R.G., In vivo microdialysis estimation of histamine in human skin, Skin Pharmacol., 5, 177–183, 1992. 26. Anderson, C., Andersson, T., and Wardell, K., Changes in skin circulation after insertion of a microdialysis probe visualized by laser Doppler perfusion imaging, J. Invest. Dermatol., 102, 807–811, 1994. 27. Petersen, L.J. et al., Histamine release in immediatetype hypersensitivity reactions in intact human skin measured by microdialysis. A preliminary study, Allergy, 47, 635–637, 1992. 28. Krogstad, A.L. et al., Microdialysis methodology for the measurement of dermal interstitial fluid in humans, Br. J. Dermatol., 134, 1005–1012, 1996. 29. Petersen, L.J., Measurement of histamine release in intact human skin by microdialysis technique. Clinical and experimental findings, Dan. Med. Bull., 45, 383–401, 1998. 30. Groth, L., Jorgensen, A., and Serup, J., Cutaneous microdialysis in the rat: insertion trauma and effect of anaesthesia studied by laser Doppler perfusion imaging and histamine release, Skin Pharmacol. Appl. Skin Physiol., 11, 125–132, 1998. 31. Groth, L., Jorgensen, A., and Serup, J., Cutaneous microdialysis in the rat: insertion trauma studied by ultrasound imaging, Acta Derm. Venereol., 78, 10–14, 1998. 32. Mathy, F.X. et al., In vivo tolerance assessment of skin after insertion of subcutaneous and cutaneous microdialysis probes in the rat, Skin Pharmacol. Appl. Skin Physiol., 16, 18–27, 2003. 33. Benfeldt, E. and Serup, J., Effect of barrier perturbation on cutaneous penetration of salicylic acid in hairless rats: in vivo pharmacokinetics using microdialysis and non-invasive quantification of barrier function, Arch. Dermatol. Res, 291, 517–526, 1999. 34. Benfeldt, E., Serup, J., and Menne, T., Microdialysis vs. suction blister technique for in vivo sampling of pharmacokinetics in the human dermis, Acta Derm. Venereol., 79, 338–342, 1999. 35. Benfeldt, E., Serup, J., and Menne, T., Effect of barrier perturbation on cutaneous salicylic acid penetration in human skin: in vivo pharmacokinetics using microdialysis and non-invasive quantification of barrier function, Br. J. Dermatol., 140, 739–748, 1999. 36. Hegemann, L. et al., Microdialysis in cutaneous pharmacology: kinetic analysis of transdermally delivered nicotine, J. Invest. Dermatol., 104, 839–843, 1995. 37. Muller, M. et al., Diclofenac concentrations in defined tissue layers after topical administration, Clin. Pharmacol. Ther., 62, 293–299, 1997.
Microdialysis Methodology for Sampling in the Skin
38. Simonsen, L. et al., Differentiated in vivo skin penetration of salicylic compounds in hairless rats measured by cutaneous microdialysis, Eur. J. Pharm. Sci., 21, 379–388, 2003. 39. Justice, J.B., Jr., Quantitative microdialysis of neurotransmitters, J. Neurosci. Methods, 48, 263–276, 1993. 40. Kehr, J., A survey on quantitative microdialysis: theoretical models and practical implications, J. Neurosci. Methods, 48, 251–261, 1993. 41. Parsons, L.H. and Justice, J.B., Jr., Quantitative approaches to in vivo brain microdialysis, Crit. Rev. Neurobiol., 8, 189–220, 1994. 42. Benveniste, H., Hansen, A.J., and Ottosen, N.S., Determination of brain interstitial concentrations by microdialysis, J. Neurochem., 52, 1741–1750, 1989. 43. Stahle, L., Drug distribution studies with microdialysis. I. Tissue dependent difference in recovery between caffeine and theophylline, Life Sci., 49, 1835–1842, 1991. 44. Wang, Y., Wong, S.L., and Sawchuk, R.J., Microdialysis calibration using retrodialysis and zero-net flux: application to a study of the distribution of zidovudine to rabbit cerebrospinal fluid and thalamus, Pharm. Res., 10, 1411–1419, 1993. 45. Bouw, M.R. and Hammarlund-Udenaes, M., Methodological aspects of the use of a calibrator in in vivo microdialysis: further development of the retrodialysis method, Pharm. Res., 15, 1673–1679, 1998. 46. Deguchi, Y. et al., Muscle microdialysis as a model study to relate the drug concentration in tissue interstitial fluid and dialysate, J. Pharmacobiodyn., 14, 483–492, 1991. 47. Larsson, C.I., The use of an “internal standard” for control of the recovery in microdialysis, Life Sci., 49, L73–L78, 1991. 48. Scheller, D. and Kolb, J., The internal reference technique in microdialysis: a practical approach to monitoring dialysis efficiency and to calculating tissue concentration from dialysate samples, J. Neurosci. Methods, 40, 31–38, 1991. 49. Kreilgaard, M., Dermal pharmacokinetics of microemulsion formulations determined by in vivo microdialysis, Pharm. Res., 18, 367–373, 2001. 50. Simonsen, L. et al., Differenciated in vivo skin penetration of salicylic compounds in hairless rats measured by cutaneous microdialysis, Eur. J. Pharm. Sci., 21, 379–388, 2004. 51. Tegeder, I. et al., Application of microdialysis for the determination of muscle and subcutaneous tissue concentrations after oral and topical ibuprofen administration, Clin. Pharmacol. Ther., 65, 357–368, 1999. 52. Benfeldt, E., In vivo microdialysis for the investigation of drug levels in the dermis and the effect of barrier perturbation on cutaneous drug penetration. Studies in hairless rats and human subjects, Acta Derm. Venereol. (Stockh.), 206 (Suppl.), 1–59, 1999. 53. Tegeder, I. et al., Time course of 8-methoxypsoralen concentrations in skin and plasma after topical (bath and cream) and oral administration of 8-methoxypsoralen, Clin. Pharmacol. Ther., 71, 153–161, 2002.
453
54. Benfeldt, E. and Groth, L., Feasibility of measuring lipophilic or protein-bound drugs in the dermis by in vivo microdialysis after topical or systemic drug administration, Acta Derm. Venereol., 78, 274–278, 1998. 55. Kreilgaard, M. et al., Influence of a microemulsion vehicle on cutaneous bioequivalence of a lipophilic model drug assessed by microdialysis and pharmacodynamics, Pharm. Res., 18, 593–599, 2001. 56. Stagni, G. et al., Iontophoretic current and intradermal microdialysis recovery in humans, J. Pharmacol. Toxicol. Methods, 41, 49–54, 1999. 57. Schnetz, E. and Fartasch, M., Microdialysis for the evaluation of penetration through the human skin barrier: a promising tool for future research? Eur. J. Pharm. Sci., 12, 165–174, 2001. 58. Morgan, C.J., Renwick, A.G., and Friedmann, P.S., The role of stratum corneum and dermal microvascular perfusion in penetration and tissue levels of water-soluble drugs investigated by microdialysis, Br. J. Dermatol., 148, 434–443, 2003. 59. Dvorak, M. et al., Histamine induced responses are attenuated by a cannabinoid receptor agonist in human skin, Inflamm. Res., 52, 238–245, 2003. 60. Krogstad, A.L. et al., Increased interstitial histamine concentration in the psoriatic plaque, J. Invest. Dermatol., 109, 632–635, 1997. 61. Petersen, L.J., Nielsen, H.J., and Skov, P.S., Codeineinduced histamine release in intact human skin monitored by skin microdialysis technique: comparison of intradermal injections with an atraumatic intraprobe drug delivery system, Clin. Exp. Allergy, 25, 1045–1052, 1995. 62. Petersen, L.J. et al., No release of histamine and substance P in capsaicin-induced neurogenic inflammation in intact human skin in vivo: a microdialysis study, Clin. Exp. Allergy, 27, 957–965, 1997. 63. Petersen, L.J. and Skov, P.S., Effect of terbutaline and bambuterol on immediate-type allergic skin responses and mediator release in human skin, Inflamm. Res, 52, 372–377, 2003. 64. Saarinen, J.V. et al., Release of histamine and leukotriene C4 in immediate allergic wheal reaction as measured with the microdialysis technique, Arch. Dermatol. Res., 292, 333–340, 2000. 65. Furutani, K. et al., Substance P- and antigen-induced release of leukotriene B4, prostaglandin D2 and histamine from guinea pig skin by different mechanisms in vitro, Arch. Dermatol. Res., 291, 466–473, 1999. 66. Rhodes, L.E. et al., Ultraviolet-B-induced erythema is mediated by nitric oxide and prostaglandin E2 in combination, J. Invest. Dermatol., 117, 880–885, 2001. 67. Andoh, T. and Kuraishi, Y., Quantitative determination of endogenous nitric oxide in the mouse skin in vivo by microdialysis, Eur. J. Pharmacol., 332, 279–282, 1997. 68. Clough, G.F., Bennett, A.R., and Church, M.K., Measurement of nitric oxide concentration in human skin in vivo using dermal microdialysis, Exp. Physiol, 83, 431–434, 1998.
454
Handbook of Non-Invasive Methods and the Skin, Second Edition
69. Sjogren, F., Svensson, C., and Anderson, C., Technical prerequisites for in vivo microdialysis determination of interleukin-6 in human dermis, Br. J. Dermatol., 146, 375–382, 2002. 70. Iversen, V.V. et al., Continuous measurements of plasma protein extravasation with microdialysis after various inflammatory challenges in rat and mouse skin, Am. J. Physiol. Heart Circ. Physiol., 286, H108–H112, 2004. 71. Neisius, U. et al., Prostaglandin E2 induces vasodilation and pruritus, but no protein extravasation in atopic dermatitis and controls, J. Am. Acad. Dermatol., 47, 28–32, 2002. 72. Sauerstein, K. et al., Electrically evoked neuropeptide release and neurogenic inflammation differ between rat and human skin, J. Physiol., 529, 803–810, 2000. 73. Schmelz, M. et al., Mast cell tryptase in dermal neurogenic inflammation, Clin. Exp. Allergy, 29, 695–702, 1999. 74. Weidner, C. et al., Acute effects of substance P and calcitonin gene-related peptide in human skin: a microdialysis study, J. Invest. Dermatol., 115, 1015–1020, 2000. 75. Schmelz, M. and Petersen, L.J., Neurogenic inflammation in human and rodent skin, News Physiol. Sci., 16, 33–37, 2001. 76. Muller, M. et al., Characterization of peripheral-compartment kinetics of antibiotics by in vivo microdialysis in humans, Antimicrob. Agents Chemother., 40, 2703–2709, 1996. 77. Mader, R.M. et al., Penetration of capecitabine and its metabolites into malignant and healthy tissues of patients with advanced breast cancer, Br. J. Cancer, 88, 782–787, 2003. 78. Tegeder, I. et al., Cisplatin tumor concentrations after intra-arterial cisplatin infusion or embolization in patients with oral cancer, Clin. Pharmacol. Ther., 73, 417–426, 2003. 79. Bolinder, J. et al., Microdialysis of subcutaneous adipose tissue in vivo for continuous glucose monitoring in man, Scand. J. Clin. Lab. Invest., 49, 465–474, 1989. 80. Simonsen, L. et al., Glucose metabolism in chronic diabetic foot ulcers measured in vivo using microdialysis, Clin. Physiol., 18, 355–359, 1998.
81. Lutgers, H.L. et al., Microdialysis measurement of glucose in subcutaneous adipose tissue up to three weeks in type 1 diabetic patients, Neth. J. Med., 57, 7–12, 2000. 82. Sjostrand, M., Holmang, A., and Lonnroth, P., Measurement of interstitial insulin in human muscle, Am. J. Physiol., 271, E151–E154, 1999. 83. Lehmann, B. et al., Demonstration of UVB-induced synthesis of 1 alpha,25-dihydroxyvitamin D3 (calcitriol) in human skin by microdialysis, Arch. Dermatol. Res., 295, 24–28, 2003. 84. Boutsiouki, P., Thompson, J.P., and Clough, G.F., Effects of local blood flow on the percutaneous absorption of the organophosphorus compound malathion: a microdialysis study in man, Arch. Toxicol., 75, 321–328, 2001. 85. Brunner, M. et al., Direct assessment of peripheral pharmacokinetics in humans: comparison between cantharides blister fluid sampling, in vivo microdialysis and saliva sampling, Br. J. Clin. Pharmacol., 46, 425–431, 1998. 86. Boelsma, E. et al., Microdialysis technique as a method to study the percutaneous penetration of methyl nicotinate through excised human skin, reconstructed epidermis, and human skin in vivo, Pharm. Res., 17, 141–147, 2000. 87. Cross, S.E., Anderson, C., and Roberts, M.S., Topical penetration of commercial salicylate esters and salts using human isolated skin and clinical microdialysis studies, Br. J. Clin. Pharmacol., 46, 29–35, 1998. 88. Leveque, N. et al., Comparison of Franz cells and microdialysis for assessing salicylic acid penetration through human skin, Int. J. Pharm., 269, 323–328, 2004. 89. McDonald, S. and Lunte, C., Determination of the dermal penetration of esterom components using microdialysis sampling, Pharm. Res., 20, 1827–1834, 2003. 90. Shah, V.P. et al., Bioequivalence of topical dermatological dosage forms: methods of evaluation of bioequivalence, Pharm. Res., 15, 167–171, 1998. 91. Jackson, A.J., Determination of in vivo bioequivalence, Pharm. Res., 19, 227–228, 2002.
Skin Surface Microflora
54 Sampling the Bacteria of the Skin E. Anne Eady Department of Microbiology, University of Leeds, Leeds, United Kingdom
CONTENTS 54.1 Introduction............................................................................................................................................................457 54.2 Factors Affecting the Choice and Efficacy of Sampling Methods .......................................................................457 54.3 Methods Available .................................................................................................................................................458 54.3.1 Impression/Replica Methods .....................................................................................................................459 54.3.1.1 Contact Plates.............................................................................................................................459 54.3.1.2 Pads.............................................................................................................................................459 54.3.1.3 Sellotape Stripping .....................................................................................................................459 54.3.2 Swabbing Methods ....................................................................................................................................460 54.3.3 Washing Methods ......................................................................................................................................461 54.3.3.1 The Detergent Scrub Technique.................................................................................................461 54.3.3.2 The Sterile Bag Technique.........................................................................................................462 54.3.4 Follicular Sampling Methods ....................................................................................................................462 54.3.4.1 Comedone Extractor...................................................................................................................462 54.3.4.2 Cyanoacrylate Glue ....................................................................................................................462 54.4 Correlation between the Methods: Advantages and Disadvantages of Each .......................................................463 54.5 Recommendations..................................................................................................................................................463 References .......................................................................................................................................................................465
54.1 INTRODUCTION The bacterial flora of human skin resides in the upper layers of the stratum corneum and within the infundibula of pilosebaceous follicles.1,2 These are lined with epithelium, which is continuous with the epidermis so that the microorganisms that colonize them are sequestered external to the lining of the duct. The methods discussed in this chapter fall into two groups: those that are used to sample organisms from the skin surface and those that are used to extract organisms from within pilosebaceous ducts. It should be pointed out that there is no noninvasive method that will remove bacteria from healthy pilosebaceous units. The only method that will do this requires a biopsy as the starting material.3,4 This chapter does not provide an exhaustive historical overview of bacterial sampling methods, but instead highlights and describes in some detail those methods that should be of most use in a modern skin research laboratory or for routine patient diagnosis.
54.2 FACTORS AFFECTING THE CHOICE AND EFFICACY OF SAMPLING METHODS Human skin provides a variety of different habitats for the survival and multiplication of bacteria. For example, the environment in the axilla is moist and warm and contrasts sharply with that on the surface of the arms and legs, which is relatively cool and dry. These variations are reflected in differences in the density of the bacterial flora in the different locations. As might be expected, the axilla supports the growth of far higher numbers of bacteria than are found on the limbs (excluding the hands and feet). Exposed sites such as the hands and face are likely to carry higher numbers of transient organisms than nonexposed sites. The resident flora of any given individual varies little in both numbers and types of organism from puberty to old age, when the numbers slowly decline. Perturbation of the microflora results from many factors, such as disease, trauma, therapy (not necessarily antimicrobial), antisepsis, and increased hydration. Some 457
458
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 54.1 Resident Bacteria Found at Various Anatomical Locations on Healthy Human Skin Location Organism
Forehead
Axilla
Forearm
Hands
Perineum
Toe Webs
Staphylococcus aureus Staphylococcus epidermidis and other coagulase-negative staphylococcia Micrococcus spp.b Propionibacterium acnes and other cutaneous propionibacteriac Corynebacterium spp. Brevibacterium spp. Acinetobacter spp. Anaerobic coccid
– +
(+) +
– +
(+) +
(+) +
– +
(+) +
– +
(+) (+)
(+) (+)
– +
+ +
(+) (+) – +
+ + (+) ?
(+) (+) – (+)
(+) (+) – ?
(+) (+) – ?
+ + (+) ?
Note: A plus sign indicates that the organism is carried by the majority of subjects and/or is usually present at that site in high numbers (≥ 103 cfu/cm2). The brackets indicate that the organism may be present in a majority of subjects and/or in low numbers. A minus sign indicates that the organism is not usually found at that site. a
The various species of coagulase-negative staphylococci are not evenly distributed. M. sedentarious is commonest on the foot, whereas M. luteus is more common at other sites. The latter organism often occurs in low numbers on exposed skin and may not be a true resident. c Propionibacterium avidum is most commonly found in intertriginous areas. d Otherwise referred to as Staphylococcus saccharolyticus. No systematic study of the prevalence of these organisms on human skin has been carried out. b
organisms (e.g., group A beta-hemolytic streptococci) may assume resident status for short periods, especially in disease states, but will be eliminated once the skin returns to normal. For more information about the skin flora in health and disease, refer to the excellent book edited by W.C. Noble.5 Table 54.1 summarizes the main bacterial species resident on healthy human skin and the sites in which they are commonly found. Three bacterial groups — the propionibacteria, coagulase-negative staphylococci, and aerobic coryneforms (either Corynebacterium or Brevibacterium species) — are widely distributed over the entire skin surface, although particular species may be associated with a specific habitat type. For example, Propionibacterium acnes is the most common bacterium on sebaceous gland-rich areas of skin, but coagulase-negative staphylococci and aerobic coryneforms are dominant on moist skin. When the integrity of the skin barrier is interrupted by disease or trauma, a variety of nonresident bacteria can flourish. Table 54.2 provides a summary of the variables that affect the efficacy of skin sampling methods and gives suggestions how some, but not all, of the resulting problems can be overcome.
54.3 METHODS AVAILABLE Faced with the degree of variation in numbers and types of bacteria and with the heterogeneity of the skin surface,
TABLE 54.2 Variables that Affect the Efficacy of Skin Sampling Methods Variable Uneven distribution of resident skin microflora Heterogeneity of resident bacterial flora Variation in population density from site to site and person to person Variation in surface contours of skin and presence of hairs at certain sites Damage to the skin by trauma or disease Presence of antimicrobial substance (antiseptic, disinfectant, or antibiotic)
Associated or Resulting Problems Difficult to define or obtain a representative sample Need to use a variety of selective and nonselective media to process sample Often need to serially dilute samples in order to achieve accurate quantification Use of many sampling techniques impossible, especially quantitative methods Some sampling techniques too aggressive; bacterial flora modified, pathogens likely May need to add neutralizer to diluent and/or growth medium
it is obvious that no one sampling method can be universally applicable. Since the 1950s, when interest in skin microbiology began to increase, a variety of different sampling methods have been developed and subsequently
Sampling the Bacteria of the Skin
modified, so that those available today have resulted from many years of evolution. It is convenient to divide the methods into four basic categories, which will each be discussed in turn.
54.3.1 IMPRESSION/REPLICA METHODS 54.3.1.1 Contact Plates The use of contact plates (Figure 54.1) represents the easiest way of sampling skin bacteria. Contact plates are specialized Petri dishes that are filled with any desired culture medium until the agar surface is slightly concave. They are then simply pressed firmly onto the skin to remove surface bacteria. A grid is etched onto the base of the plates to facilitate colony counting. The method is not quantitative, although it does give an estimate of the number of microcolonies at the site sampled. Because the organisms are inoculated directly onto the agar surface, replicate samples cannot be transferred to different culture media or grown under different culture conditions. Therefore, the method is of most use when a specific organism is being sought, for example, the isolation of Staphylococcus aureus from eczema lesions.6 Further examples of the use of contact plates can be found in the work of Johnston et al.,7 Brown et al.,8 and Hendley and Ashe.9
459
series of inoculating wires to transfer the organisms to the agar surface. The method is very inefficient, and only a small proportion of organisms are successfully transferred from the velvet to the recovery medium.12 Pads made from either 85% viscose and a 15% mixture of polyester and polyamide fibers or polyvinyl alcohol foam are now commercially available and represent a considerable improvement over the original material.13 Both new types of pad can be attached to an applicator and are premoistened before use in a solution consisting of 2% polysorbate 80 (Tween 80) and 0.3% lecithin in phosphate-buffered saline (PBS). A template twice the size of the applicator is held on the skin surface, and the pads are rubbed back and forth 10 times. In this way organisms lying on and below the surface should be detached. Since Raahave12 demonstrated that bacterial recoveries could be considerably improved by mechanical rinsing, the pads are no longer used directly to inoculate culture media. Instead, organisms are released from the pads in a stomacher containing 50 ml of PBS. Aliquots of buffer are removed from the stomacher and plated onto one or more selective or nonselective media as required. The ability to detach the bacteria from the pads has made it possible to obtain a quantitative estimate of bacterial numbers. 54.3.1.3 Sellotape Stripping
54.3.1.2 Pads The use of velvet pads to remove bacteria from the skin surface has the advantage over contact plates that sufficient organisms are removed to serially inoculate a number of different culture media.10,11 The pile of the fabric picks up individual bacterial cells and then acts like a
FIGURE 54.1 Contact plates from the forehead of one individual showing (a) colonies of aerobic coryneforms on fresh blood agar containing furazolidone (upper left), (b) colonies of coagulase-negative staphylococci on CLED (upper right), and (c) colonies of propionibacteria on Brian Heart Infusion containing furazolidone (lower middle). Note that the appropriate use of selective media has prevented the growth of unwanted bacteria and that the density of propionibacteria makes quantitation impossible.
Both velvet pads and contact plates only remove bacteria located on the skin surface. Sellotape stripping overcomes the problem of not reaching deeper organisms by facilitating serial sampling of successive layers of the stratum corneum, each just one cell thick.14 In this way, it was shown that the majority of aerobic bacteria reside in the upper part of the epidermis. In areas of skin containing numerous pilosebaceous follicles, some (but not all) organisms will be pulled out of the upper portion of each duct so that the numbers of bacteria may not decline in subsequent strips, as they do in areas with few pilosebaceous units. The Sellotape strips are inverted onto the surface of the culture medium and may be either left in contact with the agar surface or removed before incubation. Both procedures are associated with problems. If the tape is left in situ, the oxygen tension beneath it is reduced so that growth of aerobic and facultatively anaerobic bacteria is inhibited. If the tape is removed, not all bacteria may have been successfully transferred to the culture medium. Care must be taken to ensure that the tape that is used is sterile and does not contain any bacteriostatic or bactericidal substances. Some authors have used a combination of Sellotape stripping and contact plates for skin sampling, using the Sellotape to remove successive sheets of epidermal cells and the contact plates to remove the exposed bacteria.8,9
460
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 54.3 Advantages and Disadvantages of Recommended Sampling Methods Method Contact plates Moist swabbing
Detergent scrub technique
Comedone extraction
Cyanoacrylate glue
Advantages
Disadvantages
Easy and quick to use on intact and broken skin. Can be employed for routine patient sampling. Easy and quick to use. Can be employed for routine patient sampling. Several recovery media can be inoculated immediately (in the clinic or laboratory). Can be used in intertriginous areas and on damaged skin. Quantitative. Several recovery media can be employed.
Tedious to prepare. Limited to one recovery medium. Not quantitative. Cannot use if skin surface is uneven or hairy. Not quantitative.
Samples intrafollicular bacteria. Simple and quick procedure. Can study microflora of a single pilosebaceous unit. Quantitative. Can inoculate several recovery media. Samples intrafollicular bacteria. Can pool samples from multiple follicles or study singly. Quantitative. Can inoculate several recovery media.
54.3.2 SWABBING METHODS Methods for sampling skin bacteria based on the use of swabs generally have a poor reputation among skin microbiologists despite the fact that they have the best ratio of advantages to disadvantages of any currently available technique (see below and Table 54.3). It is the improper use of dry swabs and the associated poor recoveries of viable organisms that have relegated swabbing to a method of last resort among serious researchers. This is unfortunate, because, when properly used, moist swabs provide one of the most versatile of skin sampling techniques. It is important to stress that the use of swabs rarely provides accurate quantitative data, but there are some situations in which the method can be used semiquantitatively, for example, when few organisms are present or when the area swabbed can be accurately defined and the bacteria of interest are known to reside superficially. There is disagreement in the literature as to whether virtually all or only a small proportion of the organisms collected on a swab are released during subsequent processing. This will depend on the type of swab used and what procedure is used to transfer organisms to the culture medium. Swabs are available in a variety of materials (cotton, rayon, calcium alginate), but all types should be moistened in either phosphate-buffered saline or Williamson and Kligmans’ wash fluid (see below) before use. The area of the sample site can be standardized by holding a template onto the skin surface, but for many applications this is not necessary. The skin surface should be rubbed firmly and repeatedly for several seconds to ensure adequate removal of
Time consuming. Not suitable for routine patient sampling. Cannot be used on all skin sites or if skin is damaged. Specialized equipment needed which is not commercially available. Single sample not representative. Follicles are diseased. No similar method for normal follicles. Processing of sample tedious. Samples micromedones only. Cannot use for normal follicles. Difficult and tedious procedure.
bacteria. If required, rubbing can be carried out for a fixed time. For semiquantitative work, the swab should be transferred to 1 ml of half-strength wash fluid and decimally diluted in the same. A fixed volume (usually 100 ml) of each dilution and the undiluted sample is then plated onto one or more suitable recovery media and spread with a sterile glass spreader. The appearance of the plates following incubation is shown in Figure 54.2. Alternatively, viable counts can be determined by the method of Miles and Misra.15 For qualitative work, the swab can be used directly to inoculate one or more culture media (Figure 54.3). There are many applications in which the correct use of swabs represents the best or only possible method of skin sampling. They can be used for the collection of samples for routine clinical bacteriology of patients with skin diseases and are equally useful on intact and broken skin, even in the presence of severe structural breakdown, as in burns or ulcers. Standardized swabbing techniques can be used for research purposes and provide the only means of sampling certain skin sites, such as the toe webs and ear canal. If the swab cannot be processed immediately, it should be kept moist by immersion in wash fluid or transport medium. The viability of the collected bacteria will rapidly decline if the swab is allowed to dry out at any stage. Examples of the correct use of swabbing techniques can be found in the studies of Shaw et al.,16 Keswick et al.,17 Marshall et al.,18 McGinley et al.,19 Keyworth et al.,20 and Harkaway et al.21
Sampling the Bacteria of the Skin
FIGURE 54.2 Enumeration of coagulase-negative staphylococci by plating fixed volumes from decimal dilutions of wash fluid onto heated blood agar. Note the decreasing number of colonies with increasing dilution (from left to right). Well-separated colonies have been obtained from the 10–3 dilution and this plate (bottom left) should be used to estimate the viable bacterial count.
FIGURE 54.3 Appearance of a plate of Brian Heart Infusion inoculated with a moist swab taken from the forehead following anaerobic incubation for 7 d. The media contains 2 mg/l furazolidone to inhibit the growth of staphylococci and 5 mg/l tetracycline to select for tetracycline-resistant propionibacteria. The swab was taken and plated out by a nurse in a busy dermatology clinic.
54.3.3 WASHING METHODS 54.3.3.1 The Detergent Scrub Technique The bacterial skin sampling method that has achieved the greatest prominence and is the most widely used for research purposes is the detergent scrub technique of Williamson and Kligman.22 The method is standardized, quantitative, reproducible, and efficient (i.e., removes over 95% of the aerobic bacteria present at the sample site). Various
461
modifications and adaptations have appeared in the literature over the last 30 years, but the technique in its simplest form remains the best and most universally applicable. A metal ring is held firmly against the skin surface and 1 ml of wash fluid (0.075 M sodium phosphate buffer, pH 7.9, containing 0.1% v/v Triton-X 100) is pipetted into it. The skin surface within the ring is rubbed firmly for 1 minute with a Teflon policeman, which is lifted away from the skin every few seconds and then replaced. Care must be taken to ensure that the entire area within the ring is rubbed evenly. The wash fluid is collected into a suitable sterile container and the procedure is repeated. The two aliquots of wash fluid are pooled and decimally diluted in half-strength wash fluid. Aliquots of each dilution are then plated onto suitable recovery media, as described in Section 54.3.2. There are several features of the scrub wash technique that are worthy of further comment. The wash fluid contains a mild detergent in order to facilitate dispersal of clumps of bacteria. Various authors have studied the survival of different skin bacteria in wash fluid with inconclusive results. It is best for individual investigators to estimate for themselves the survival time in wash fluid of those organisms of special interest to them. Alternatively, it is safe to assume no significant change in viable count if the samples are processed within 1 hour of collection. In the original method, the sampling ring enclosed an area of 3.8 cm2. This area appears to have been arbitrarily chosen. Rings of any size can be used, depending upon the reason for sampling. Obviously, the smaller the diameter of the ring, the less representative the sample becomes. We use two ring sizes, 2.5 cm diameter enclosing 4.9 cm2 for routine skin microbiology and for monitoring the efficacy of antibacterial therapies for acne and eczema, and a larger ring of 3.5 cm diameter enclosing 9.6 cm2 for studies on antibiotic resistance in skin bacteria (Figure 54.4). The cutaneous microflora is more heterogeneous in terms of resistance profiles than in terms of species, so that it is necessary to remove bacteria from as large an area as possible to obtain a more representative sample of resistant strains. A smaller ring of 1.5 cm2 enclosing an area of 1.8 cm2 can be used to reveal the degree of variation in numbers of resident microorganisms at adjacent sites. The lower limit of detection of the method is from colony-forming units (cfu)/cm2 using a 2.5-cm-diameter ring and 2 ml of wash fluid. A modification of this technique has been developed for assessing the efficacy of hand wash products (see also Section 54.3.3.2).23 Each fingernail region is immersed in 7 ml of collecting fluid and scrubbed for 1 minute with an electric toothbrush. The sample site is chosen because the subungual space harbors large numbers of bacteria and is one of the most difficult sites to disinfect.
462
Handbook of Non-Invasive Methods and the Skin, Second Edition
54.3.4.1 Comedone Extractor
FIGURE 54.4 Examples of sampling rings used for the detergent scrub technique. For an explanation of when to use different-sized rings, see text.
54.3.3.2 The Sterile Bag Technique The bacterial flora of the hands consists of a much greater variety of species than is found elsewhere on the body, except the feet.18,19 Many of these will not be true residents, but transients that are picked up from the environment. The main interest in the microflora of the hand is how to get rid of it. The sterile bag technique is a simplification of the gloved hand method for assessing the efficacy of hand disinfection and provides a way of sampling the flora of the entire hand.24,25 This is obviously an example of a specialized method developed to fulfill a particular need. Essentially, the hand is enclosed in a sterile plastic bag into which is poured sterile fluid. The bag is sealed at the wrist, and the hand is rubbed vigorously for a standard length of time to release the bacteria. Aliquots of the fluid are used to estimate the viable count of bacteria using the most probable number method. Several different sampling solutions have been tested and wash fluid (see above) was found to yield the highest bacterial counts.25 It may sometimes be necessary to add neutralizers to the sampling solution to inactivate any residual disinfectant.
54.3.4 FOLLICULAR SAMPLING METHODS Acne vulgaris is a disease of pilosebaceous follicles of the face, back, and chest in which the resident bacterium, Propionibacterium acnes, plays a role in the development of inflammatory lesions.26 Researchers interested in the pathogenesis of acne need techniques that facilitate sampling of intrafollicular organisms. The methods that will be described below can be used to sample noninflamed lesions only and are not suitable for use with normal follicles or inflamed lesions. All the available data on the bacteriology of normal follicles have been obtained following their microdissection from skin biopsies.3,4
Open comedones (blackheads) can be removed nontraumatically using a comedone extractor. These are available from chemists because of their cosmetic use. The skin surface is sterilized using an isopropanol swab, and the comedone is removed by applying firm downward pressure with the extractor. The comedone is transferred to a sterile preweighed microcentrifuge tube with a sterile needle. After weighing, 200 ml of wash fluid is pipetted into the tube and the bacteria are dispersed from the comedone using a microtissue grinder. The fluid is then decimally diluted in half-strength wash fluid and plated onto one or more selective or nonselective media as required. Bacterial counts are expressed as cfu per milligram wet weight of comedonal material. The detection limit is 4 cfu/comedone. The method is unique in that it samples the entire bacterial flora of a single microhabitat. However, care must be taken to ensure that the whole of the comedone is removed from the skin because the distribution of bacteria in follicular ducts varies with depth.27 The density and composition of the microflora also vary greatly between comedones from a complete absence of viable bacteria to >106 cfu/mg of propionibacteria alone, coagulase-negative staphylococci alone, or both.28–30 Faced with such heterogeneity, it is essential to examine several comedones from each subject in order to obtain an accurate picture. This is especially important if the effects of antibacterial therapy are being studied. 54.3.4.2 Cyanoacrylate Glue Rapidly polymerizing cyanoacrylate glue was first used to remove thin sheets of stratum corneum in a similar way to Sellotape stripping.31 It was quickly realized that follicular plugs were extracted from pilosebaceous follicles as the glue was pulled away from the skin surface. In the simplest form of the method, a drop of the glue is spread over an area of skin and left to polymerize for 1 minute.32 A second drop of glue is then applied on top of the first and spread uniformly by inverting a glass slide over it and pressing down firmly. After another minute, the slide is gently removed from the skin with its adherent sheet of adhesive and follicular casts. These casts represent the contents of microcomedones and consist of a mixture of corneocytes, sebum, and microbes. They can be microdissected from the glue and pooled or homogenized individually with microtissue grinders as described above. A more standardized version of this procedure uses a sterile glass sampler of known surface area instead of a slide and a sterile Teflon ring to delineate the sample area, which is extracted twice with glue.33 The organisms are released by vortexing the end of the sampler in sterile medium in the presence of Ballotini beads. Exolift®* is a commercially * Registered trademark of Exovir, Inc., Great Neck, New York 11021.
Sampling the Bacteria of the Skin
available kit that includes a patented dermal tape and cyanoacrylate glue. It is easier to use than glass slides or samplers, but is much more expensive. Whichever procedure is followed, only follicular bacteria will be enumerated since surface organisms are sequestered between the glue and the thin sheet of stratum corneum. The main problems with the use of cyanoacrylate glue are the high frequency of incomplete takes, when the glue fails to polymerize properly over part of the sample site, and the uncertainty of removing entire follicular casts. For obvious reasons, the method should not be used near the eyes.
463
TABLE 54.4 Factors Important in Determining Choice of Sampling Method Factor Location of bacteria Type of skin Type of bacteria Choice of sample site
54.4 CORRELATION BETWEEN THE METHODS: ADVANTAGES AND DISADVANTAGES OF EACH Of the methods described above, Sellotape stripping and the use of velvet pads are now outdated and have no place in modern skin bacteriology. They will not be discussed further in this article. Several research groups have studied the correlation between different sampling methods, with various outcomes.6,13,16,20,34–36 It should be obvious that counts obtained with contact plates are consistently lower than those obtained by the other surface sampling methods because there is no dispersal step to break down aggregates of cells into smaller colony-forming units. Counts obtained by the detergent scrub technique are usually taken as the gold standard against which the efficacy of other methods is compared. No conclusive evidence has emerged from any of these studies to be able to state with certainty the relative efficacy of alternative surface sampling techniques. There is no doubt that none of the other methods are more efficient in terms of number of organisms recovered or in reproducibility than the detergent scrub technique. For many applications, the absolute number of organisms at the sample site is not of paramount importance, but the investigator must be sure that the method chosen removes a constant proportion of the organisms present in order that comparisons between samples taken from different individuals, or sites, or at different times, can be validly made. Correlations of the efficacy of follicular sampling methods would be meaningless because comedone extraction samples blackheads, whereas the use of cyanoacrylate glue samples predominantly microcomedones. In patients with numerous blackheads, these too will be removed by the glue method. Most investigators process blackheads individually, whereas microcomedones are usually pooled because of their small size. Table 54.3 summarizes the main advantages and disadvantages of the methods outlined above. Viscose and foam pads have been omitted because they possess no obvious advantages over moist swabbing of a defined area and are more cumbersome to use. Their large size (5 × 5
Efficacy of technique
Reason for sampling
Possible Alternatives/Important Considerations Surface or follicle Intact or damaged Hairy, smooth or uneven Normal flora or pathogens Selection of appropriate culture media One or several Pre-determined (e.g., lesion present) or selectable Qualitative or quantitative Number of microcolonies or individual bacteria Reproducibility Sensitivity Ease of use Research or diagnosis Comparative or noncomparative
cm) means that they are unsuitable for use on many skin sites. The sterile bag technique is the accepted method of sampling hands and the only one that can be sensibly used for estimating the efficacy of hand-washing regimens, but it is not applicable for bacterial sampling of other skin sites. When choosing a sampling method, the reader should consider which of the factors listed in Table 54.4 are the most important for a specific application. For example, it is no good selecting the detergent scrub technique if speed and simplicity are essential criteria.
54.5 RECOMMENDATIONS For ease and versatility, there is no doubt that moist swabbing is the method of choice for surface sampling. It remains the most commonly used method for routine sampling of patients with diseases, infections, or wounds of the skin. The only practical alternative is the use of contact plates. However, these have severe limitations (Table 54.3) and should only be used if a specific organism is being sought or if low numbers of bacteria (≤10 cm2) are expected and they can all be cultured on the same medium. For research purposes, and when quantitative data are required, the detergent scrub technique should be chosen unless there are overriding reasons why it cannot be used. The major limitation of this technique is that it is fairly aggressive and cannot be used on sensitive or damaged skin, although several groups have used it to sample bacteria from eczema lesions.6,37,38 With all surface sampling methods, if temporal changes in the bacterial flora are being studied, adequate time must be allowed to elapse
464
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 54.5 Choice of Growth Media for Recovery of Resident Skin Bacteria and Primary Pathogens Organism(s) Propionibacteria Coagulase-negative staphylococci
Staph. aureus
Aerobic coryneforms
Group A beta-hemolytic streptococci
Recommended Medium Brain Heart Infusion or Reinforced Clostridial agar containing 6 mg/l furazolidone Heated blood agar
Mannitol salt agar Cysteine lactose electrolyte deficient (CLED) medium Fresh blood agar containing 0.2% w/v glucose, 0.3% w/v yeast extract, 0.2% v/v Tween 80 and 6 mg/l furazolidone Fresh blood agar containing 0.0002% crystal violet or 7.5 mg/l nalidixic acid and 17 units/ml polymyxin B
Other Resident Skin Bacteria Capable of Growth None Nonlipophilic coryneforms Staph. aureus Micrococcus spp. Gram-negative rods Some coagulase-negative staphylococci Coagulase-negative staphylococci and nonlipophilic coryneforms Micrococcus spp.
None
Note: Incubated anaerobically. All other media incubated aerobically.
after sample collection to allow the bacterial flora to reestablish before a further sample can be taken from the same site. Some investigators overcome this problem by sampling from adjacent sites or from identical sites on the right and left side. Comedone extraction is the best method of sampling intrafollicular bacteria and, when the sample is correctly processed, generates data of very high quality. The limiting factor with this method is the supply of open comedones, which are only present in a minority of acne patients. Microcomedones are more common, but the use of cyanoacrylate glue is far from easy. Open and micro-
comedones, which can be obtained noninvasively, must not be used as substitutes for normal follicles, which can only be obtained following biopsy and are less frequently colonized by bacteria.39 All methods that attempt to generate accurate viable count data include a dispersal step followed by serial dilution and inoculation of recovery media. Table 54.5 lists suitable media for the isolation of different types of skin bacteria, including major pathogens. It is a waste of time using a sophisticated sampling method if subsequent processing of the sample obtained is inadequate. All media should be incubated aerobically at 37˚C for 24 to 48 h,
TABLE 54.6 Examples of Appropriate and Inappropriate Uses of Sampling Methods Reason for Sampling
Appropriate Method
Inappropriate Method
Isolation of Staph. aureus from suspected infected eczema. Detection of unknown pathogen in infected wound.
Contact plate containing selective medium. Sample several lesions. Moist swab transferred to laboratory in transport medium. Inoculate a range of selective and nonselective media immediately. Sterile bag technique, followed by most probable number method of estimating viable bacteria using a nonselective medium. Detergent scrub technique, followed by decimal dilution and plating onto antibiotic-containing and antibiotic-free medium. Sample site will affect result. Moist swabbing of many anatomical sites. Plating onto selective and nonselective media.
Dry swab used to inoculate nonselective medium only. Dry swab used to inoculate nonselective medium only.
Estimation of the efficacy of a handwashing regimen. Determination of the proportion of the resident staphylococcal flora resistant to an antibiotic. Identification of components of the aerobic skin flora of premature neonates.
Detergent scrub technique on one site.
Contact plates using antibioticcontaining and antibiotic-free medium.
Contact plates.
Sampling the Bacteria of the Skin
except for the isolation of propionibacteria. These organisms are slow-growing anaerobes, and media for their isolation must be incubated in an anaerobic jar or cabinet for 7 days. It is unlikely that there will ever be a consensus among those who study skin bacteriology as to the most suitable methods for specific tasks. As a guide, some appropriate and inappropriate uses of the methods described in this chapter are shown in Table 54.6. There are many important unanswered questions in skin microbiology, such as the role of propionibacteria in acne. It is hoped that this article has demonstrated that the cutaneous microflora is easily accessible to study and that the effort may be very rewarding. Chapter 55 by Jan Faergeman deals with sampling techniques for fungi. All of the methods described above are suitable for the enumeration of Malassezia furfur (Pityrosporum ovale), the yeast that is a member of the resident skin flora, if samples are plated out onto a selective growth medium such as that developed by Leeming and Notman.40
REFERENCES 1. Montes, L.F. and Wilborn, W.H., Location of bacterial skin flora, Br. J. Dermatol., 81 (Suppl. 1), 23, 1969. 2. Wolff, H.H. and Plewig, G., Ultrastructur der mikroflora in follikeln und comedonen, Hautarzt, 27, 432, 1976. 3. Puhvel, S.M., Reisner, R.M., and Amirian, D.A., Quantification of bacteria in isolated pilosebaceous follicles in normal skin, J. Invest. Dermatol., 65, 525, 1975. 4. Leeming, J.P., Holland, K.T., and Cunliffe, W.J., The microbial ecology of pilosebaceous units isolated from human skin, J. Gen. Microbiol., 130, 803, 1984. 5. Noble, W.C., Ed., The Skin Microflora and Microbial Skin Disease, Cambridge University Press, London, 1992. 6. Williams, R.E.A., Gibson, A.G., Aitchison, T.C., Lever, R., and Mackie, R.M., Assessment of a contact-plate sampling technique and subsequent quantitative bacterial studies in atopic dermatitis, Br. J. Dermatol., 123, 493, 1990. 7. Johnston, D.H., Fairclough, J.A., Brown, E.M., and Morris, R., Rate of bacterial recolonisation of the skin after preparation: four methods compared, Br. J. Surg., 74, 64, 1987. 8. Brown, E., Wenzel, R.P., and Hendley, J.O., Exploration of the microbial anatomy of normal human skin by using plasmid profiles of coagulase-negative staphylococci: search for the reservoir of resident skin flora, J. Infect. Dis., 160, 644, 1989. 9. Hendley, J.O. and Ashe, K.M., Effect of topical antimicrobial treatment on aerobic bacteria in the stratum corneum of human skin, Antimicrob. Agents Chemother., 35, 627, 1991. 10. Holt, R.J., Pad culture studies on skin surfaces, J. Appl. Bacteriol., 29, 625, 1966.
465
11. Gorril, R.H. and Penikett, E.J.K., New method of studying the bacterial flora of infected open wounds and burns, Lancet, 2, 370, 1957. 12. Raahave, D., Experimental evaluation of the velvet pad rinse technique as a microbial sampling method, Acta Pathol. Microbiol. Scand. Sect. B, 83, 416, 1975. 13. Hambraeus, A., Hoborn, J., and Whyte, W., Skin sampling: validation of a pad method and comparison with commonly used methods, J. Hosp. Infect., 16, 19, 1990. 14. Updegraff, D.M., A cultural method of quantitatively studying the microorganisms in the skin, J. Invest. Dermatol., 43, 129, 1964. 15. Miles, A.A. and Misra, S.S., The estimation of the bactericidal power of the blood, J. Hyg. (Cambridge), 38, 732, 1938. 16. Shaw, C.M., Smith, J.A., McBride, M.E., and Duncan, W.C., An evaluation of techniques for sampling skin flora, J. Invest. Dermatol., 54, 160, 1970. 17. Keswick, B.H., Seymour, J.L., and Milligan, M.C., Diaper area skin microflora of normal children and children with atopic dermatitis, J. Clin. Microbiol., 25, 216, 1987. 18. Marshall, J., Leeming, J.P., and Holland, K.T., The cutaneous microbiology of normal human feet, J. Appl. Bacteriol., 62, 139, 1987. 19. McGinley, K.J., Larson, E.L., and Leyden, J.J., Composition and density of microflora in the subungual space of the hand, J. Clin. Microbiol., 26, 950, 1988. 20. Keyworth, N., Millar, M.R., and Holland, K.T., Swabwash method for quantitation of cutaneous microflora, J. Clin. Microbiol., 28, 941, 1990. 21. Harkaway, K.S., McGinley, K.J., Foglia, A.N., Lee, W.L., Fried, F., Shalita, A.R., and Leyden, J.J., Antibiotic resistance patterns in coagulase-negative staphylococci after treatment with topical erythromycin, benzoyl peroxide, and combination therapy, Br. J. Dermatol., 126, 586, 1992. 22. Williamson, P. and Kligman, A.M., A new method for the quantitative investigation of cutaneous bacteria, J. Invest. Dermatol., 45, 498, 1965. 23. Mahl, M.C., New method for determination of efficacy of health care personnel hand wash products, J. Clin. Microbiol., 27, 2295, 1989. 24. Michaud, R.N., McGrath, M.B., and Gross, W.A., Application of a gloved-hand model for multiparameter measurements of skin-degerming activity, J. Clin. Microbiol., 3, 406, 1976. 25. Larson, E.L., Strom, M.S., and Evans, C.A., Analysis of three variables in sampling solutions used to assay bacteria of hands: type of solution, use of antiseptic neutralisers, and solution temperature, J. Clin. Microbiol., 12, 355, 1980. 26. Webster, G.F., Inflammatory acne, Int. J. Dermatol., 29, 313, 1990. 27. Kearney, J.N., Harnby, D., Gowland, G., and Holland, K.T., The follicular distribution and abundance of resident bacteria on human skin, J. Gen. Microbiol., 130, 797, 1984. 28. Puhvel, S.M. and Amirian, D.A., Bacterial flora of comedones, Br. J. Dermatol., 101, 543, 1979.
466
Handbook of Non-Invasive Methods and the Skin, Second Edition
29. Leeming, J.P., Ingham, E., and Cunliffe, W.J., The microbial content and C3 cleaving capacity of comedones in acne vulgaris, Acta Dermatol. Venereol. (Stockh.), 68, 468, 1988. 30. Ingham, E., Eady, E.A., Goodwin, C.E., Cove, J.H., and Cunliffe, W.J., Pro-inflammatory levels of interleukin-1 alpha-like bioactivity are present in the majority of open comedones in acne vulgaris, J. Invest. Dermatol., 98, 895, 1992. 31. Marks, R. and Dawber, R.P.R., Skin surface biopsy: an improved technique for the examination of the horny layer, Br. J. Dermatol., 84, 117, 1971. 32. Mills, O.H. and Kligman, A.M., The follicular biopsy, Dermatologica, 167, 57, 1983. 33. Holland, K.T., Roberts, C.D., Cunliffe, W.J., and Williams, M., A technique for sampling micro-organisms from the pilosebaceous ducts, J. Appl. Bacteriol., 37, 289, 1974. 34. Noble, W.C. and Somerville, D.A., Methods for examining the skin flora, in Major Problems in Dermatology, Vol. 2, Rook, A., Ed., W.B. Saunders, Philadelphia, 1974, p. 316.
35. Ayliffe, G.A.J., Babb, J.R., Bridges, K., Lilly, H.A., Lowbury, E.J.L., Varney, J., and Wilkins, M.D., Comparison of two methods for assessing the removal of total organisms and pathogens from the skin, J. Hyg. (Cambridge), 75, 259, 1975. 36. Whyte, W., Carson, W., and Hambraeus, A., Methods for calculating the efficiency of bacterial surface sampling techniques, J. Hosp. Infect., 13, 33, 1989. 37. Aly, R., Maibach, H.H., and Shinefield, H.R., Microbial flora of atopic dermatitis, Arch. Dermatol., 113, 780, 1977. 38. Nilsson, E., Henning, C., and Hjorleifsson, M.-L., Density of the microflora in hand eczema before and after topical treatment with a potent corticosteroid, J. Am. Acad. Dermatol., 15, 192, 1986. 39. Leeming, J.P., Holland, K.T., and Cunliffe, W.J., The microbial colonisation of inflamed acne vulgaris lesions, Br. J. Dermatol., 118, 203, 1988. 40. Leeming, J.P. and Notman, F.H., Improved method for isolation and enumeration of Malassezia furfur from human skin, J. Clin. Microbiol., 25, 2017, 1987.
55 Mapping the Fungi of the Skin Jan Faergemann Department of Dermatology, University of Gotenburg, Sahlgren’s Hospital, Gothenburg, Sweden
CONTENTS 55.1 Summary ................................................................................................................................................................467 55.2 Introduction............................................................................................................................................................467 55.3 The Lipophilic Yeasts ............................................................................................................................................467 55.3.1 Growth Characteristics of P. ovale............................................................................................................468 55.3.2 Techniques for Culture of P. ovale............................................................................................................468 55.3.3 Skin Distribution of P. ovale .....................................................................................................................469 References .......................................................................................................................................................................469
55.1 SUMMARY Several different fungi may occasionally be isolated from the skin. However, it is only among the yeasts that we have true residents. The lipophilic yeast Pityrosporum ovale can constantly be cultured from the skin of adults. To culture the organism, it is important to know that it is lipophilic, grows best in the temperature range of 35 to 37˚C, and is very sensitive to drying. Two culture media, one with the addition of olive oil, glycerol monostearate, and Tween 80, and the other with the addition of glycerol, glycerol monostearate, Tween 60, and cow’s milk, give the best results. Two techniques for quantitative culture are described. One is to use a modification of the Williamson–Kligman model for culturing skin bacteria. With this method, a metal ring, a glass rod, and a phosphate-buffered solution containing 0.1% Triton X-100 is used to culture P. ovale. Samples are then transferred to a sterile tube, serially diluted, and plated onto the appropriate medium. Plates are read after 6 days of incubation in plastic bags at 37˚C. With the contact plate a semiquantitative but easier technique is used. The plate is gently pressed against the skin for 15 seconds and then incubated in a plastic bag at 37˚C for 6 days. A comparative study showed that the cow’s milk medium was superior. Quantitative cultures could be used to follow the effect of antifungal treatment of various Pityrosporum-related diseases, to compare various diseases, or in other experimental situations.
55.2 INTRODUCTION With the exception of yeasts, fungi colonize the skin only occasionally. Dermatophytes may be cultured in higher
numbers in warm and humid environments, and they may therefore be present on the skin due to a contamination, instead of a colonization or real infection.1 However, colonization with dermatophytes and signs of dermatophytosis are more commonly seen among soldiers than other groups due to the presence of predisposing factors.2 Molds are not true residents of the skin, but may now and then be found on the skin due to contamination. Several yeasts can be cultured from the skin. However, even yeasts may be contaminating the skin. This is probably often so for Candida albicans. In healthy people it is not a member of the normal skin flora, but in immunocompromised patients it may be cultured from normallooking skin.3 Candida parapsilosis and Rhodotorula sp. are often found on moist, warm, but otherwise normallooking skin.1,3 Thrichosporum biegelii, the etiological agent in white piedra, may often be cultured from normallooking skin in areas such as the scrotum.
55.3 THE LIPOPHILIC YEASTS The only fungus that constantly is cultured from several regions of the skin in almost 100% is the lipophilic yeast P. ovale.4–11 P. ovale was first cultured and described in skin scales from patients with dandruff and healthy subjects by Castellani and Chalmers in 1913.12 A nonlipophilic member of the genus Pityrosporum, P. pachydermatis, was described in 1925 by Waldman.13 It is a member of the normal skin flora in animals, especially dogs, cats, and rabbits. However, it has also occasionally been cultured from human skin, e.g., in patients with pustular psoriasis.14 467
468
Handbook of Non-Invasive Methods and the Skin, Second Edition
55.3.1 GROWTH CHARACTERISTICS
OF
P.
OVALE
P. ovale is lipophilic and needs the addition of lipids to the culture medium for optimal growth.4,6,7,12,15 Our previous standard medium for isolation and continuous growth of P. ovale is a glucose–neopeptone–yeast extract medium with the addition of olive oil, Tween 80, and glycerol monostearate.7,16 However, better results for isolation have been obtained with a culture medium primarily described by Leeming et al.11 This medium contained as lipid supplements glycerol, glycereol monostearate, Tween 60, and cow’s milk.11,15 P. ovale is able to grow in the temperature range of 28 to 38˚C, but in our studies the optimal temperature is 37˚C. The yeast is very sensitive to drying, and it is therefore of importance to increase the humidity in the incubator for optimal growth. In continuous cultures and when skin scales are scraped down onto the medium, growth is visible after 2 days of incubation at 37˚C, and optimal growth is obtained after 3 to 4 days. Colonies are slightly raised, with irregular edges, white to creamy in color and 3 to 6 mm in diameter (Figure 55.1). The micromorphology of the round and oval forms of the organism is, according to the names, different, but there are now several studies showing that the two forms may change into each other,17–19 and today the majority of workers in the field believe that the different forms are only variations in the cell cycle of the same organism16–19 (Figure 55.2 and Figure 55.3).
55.3.2 TECHNIQUES
FOR
CULTURE
OF
P.
OVALE
Qualitative cultures for P. ovale can be obtained by scraping the skin with a curette down onto one of the abovementioned media.15,16 The culture plate is then incubated at 37˚C in a plastic bag and preferably in an incubator
FIGURE 55.1 Macromorphology of Pityrosporum ovale (contact plate showing semiquantitative culture).
FIGURE 55.2 Microscopic picture of the round form of Pityrosporum ovale (formerly P. orbiculare) (×400).
FIGURE 55.3 Microscopic picture of the oval form of Pityrosporum ovale (×400).
with increased humidity because P. ovale is very sensitive to drying. The plates are read after 4 d. Both of these media are very selective due to the addition of cyclohexamide, antibiotics, and lipids. However, C. albicans may sometimes be found. Because P. ovale is a member of the normal human skin flora, quantitative culture is preferable. One method is a modification of the Williamson–Kligman scrub technique for culturing skin bacteria.7,26 A stainless-steel ring 2.6 cm in internal diameter and 2.0 cm deep covering 5.5 cm2 area of the skin is used. The skin is gently rubbed with a blunt sterile glass rod and the fluid removed by a Pasteur pipette (Figure 55.4). The ring is held in place on the skin with moderate pressure from two fingers. One milliliter of sterile 0.075 M phosphate buffer, pH 7.9, containing 0.1% Triton X-100 was poured into the ring and the skin gently rubbed with the glass rod for 1 min. The fluid is removed by Pasteur pipette into a sterile glass tube. Serial dilutions are performed in phosphate-buffered saline (PBS), pH 7.4, containing 0.1% of Triton X-100. Samples (0.1 ml) from the dilutions are plated out. Plates are incubated in plastic bags at 37˚C and read after 6 d. Another semiquantitative method is to use contact plates10,11 (Figure 55.1). The plate is gently pressed against the skin for 15 s and then incubated into a plastic bag11
Mapping the Fungi of the Skin
469
FIGURE 55.4 Quantitative culture of Pityrosporum ovale using a stainless-steel ring and rubbing the skin gently with a glass rod.
or Bio-Bag CFj10 (Marion Laboratories, Kansas City, KS) at 37˚C and read after 6 d. A comparison between two contact plates showed that the plate using Leeming’s medium11 gave the best results. The results obtained with these contact plates are only semiquantitative, but the plates are very easy to use and could be used in many comparative studies. Other techniques for quantitative culture of yeasts, especially P. ovale, are mentioned in the literature but will not be mentioned here.
55.3.3 SKIN DISTRIBUTION
OF
P.
OVALE
P. ovale is a member of the normal human cutaneous flora in adults.16 However, there is great variation in its density and presence in various skin locations,5,8,9,11 in children compared to adults,9,20–23 and in normal skin compared to diseased.7,16,21 In a survey of children, from newborn to the age of 15 years, P. ovale was not cultured from normal-looking skin on the back before the age of 5 years, but was found in 93% of 15-year-old children.9 In other studies P. ovale has been cultured from normal-looking skin in infants.20,22 However, the incidence was much lower than in adults, although cultures were taken from the forehead20 or from occluded skin in the diaper area.22 The increase in colonization of normal skin with P. ovale is paralleled by the increase in sebum excretion in prepuberty and puberty. In a culture study in adults from the age of 30 to 80 years we found, using a modification of the Williamson–Kligman scrub technique for quantitative culture of P. ovale,7 a parallel between a reduction in number of cultured P. ovale and an increase in age.23 The reduction in number of cultured organisms with age may partially be explained by a reduction in lipid content of the skin in elderly individuals.23 Using direct microscopy or culture, Roberts5 found P. ovale on the normal scalp in 97% and on the chest in 92
to 100% of normal healthy adults. McGinley and coworkers24 found, by direct count, the presence of P. ovalelike organisms on practically all of 112 normal scalps. In a quantitative culture study, again using the modification of the Williamson–Kligman scrub technique, P. ovale was cultured from clinically normal skin on the chest, back, upper arm, lower leg, and dorsal aspect of the hand.8 The highest count was found on the back (mean 333/cm2) and the lowest count on the hand (mean 2/cm2). P. ovale not only is a member of the normal human cutaneous flora, but also is associated with several diseases, such as pityriasis versicolor,16–18 Pityrosporum folliculitis,16,24 seborrheic dermatitis,25 and some forms of atopic dermatitis21 and confluent and reticulate papillomatosis (Gougerot–Carteaud).16 In pityriasis versicolor P. ovale change, under the influence of predisposing factors, from the normal round or oval blastospore form to the mycelial form.16 In Pityrosporum folliculitis there is an increase in the number of P. ovale in the hair follicles, but only in a minority of patients does P. ovale change to the mycelial form.24 In seborrheic dermatitis and probably also in atopic dermatitis the number of P. ovale on the skin is the same as in healthy individuals.21,25 Several patients with seborrheic dermatitis have a slight defect in T-cell immunity and an increased amount of lipids on the skin, and the disease is probably due to an abnormal reaction to P. ovale in predisposed individuals.25 Several adult patients with a head and neck distribution of atopic dermatitis have a type I allergic reaction to P. ovale.21 There is a great variation in the number of P. ovale found in various individuals, and although the number of P. ovale is significantly higher on the skin in patients with pityriasis versicolor than in healthy subjects, the intersubject variations are very high, and therefore even a quantitative culture is of minor importance for the diagnosis of a Pityrosporum-related disease. In the other Pityrosporum-related diseases mentioned above, the number of P. ovale on the skin is the same as in normal individuals. However, quantitative cultures are still of importance in several investigations. They can be used to follow the effect of antifungal therapy, epidemiological studies, differences between various diseases, and in several other experimental situations.
REFERENCES 1. Noble, W.C., Microbiology of Human Skin, 2nd ed., LLoyd-Luke, London, 1981. 2. Taplin, D., Fungous and Bacterial Diseases in the Tropics: Final Report to the U.S. Army R and D Command, Contract DADA 17-71-C1084, 1978. 3. Odds, F.C., Candida and Candidosis, Leicester University Press, Leicester, U.K., 1979. 4. Gordon, M.A., The lipophilic mycoflora of the skin, Mycologica, 43, 524, 1951.
470
Handbook of Non-Invasive Methods and the Skin, Second Edition
5. Roberts, S.O.B., Pityrosporum orbiculare: incidence and distribution on clinically normal skin, Br. J. Dermatol., 81, 264, 1969. 6. Faergemann, J. and Bernander, S., Tinea versicolor and Pityrosporum orbiculare: a mycological investigation, Sabouraudia, 17, 171, 1979. 7. Faergemann, J., Quantitative culture of Pityrosporum orbiculare, Int. J. Dermatol., 23, 110, 1984. 8. Faergemann, J., Aly., R., and Maibach, H.I., Quantitative variations in distribution of Pityrosporum orbiculare on clinically normal skin, Acta Derm. Venereol., 63, 346, 1983. 9. Faergemann, J. and Fredriksson, T., Age incidence of Pityrosporum orbiculare on human skin, Acta Derm. Venereol., 60, 531, 1980. 10. Faergemann, J., The use of contact plates for quantitative culture of Pityrosporum orbiculare, Mykosen, 30, 298, 1987. 11. Bergbrant, I.M., Igerud, A., and Nordin, P., An improved method for quantitative culture of Malassezia furfur, Res. Microbiol., in press. 12. Castellani, A. and Chalmers, A.J., Manual of Tropical Medicine, Ballieré Cox, London, 1913. 13. Sloof, W.C., Genus Pityrosporum, in The Yeasts, 2nd ed., Lodder, J., Ed., North-Holland, Amsterdam, 1971, p. 1167. 14. Sommerville, D.A., Yeasts in a hospital for patients with skin diseases, J. Hyg. (Cambridge), 70, 667, 1972. 15. Leeming, J.P. and Notman, F.H., Improved methods for isolation and enumeration of Malassezia furfur from human skin, J. Clin. Microbiol., 25, 2017, 1987. 16. Faergemann, J., Lipophilic yeasts in skin disease, Sem. Dermatol., 4, 173, 1985.
17. Faergemann, J. and Fredriksson, T., Experimental infections in rabbits and humans with Pityrosporum orbiculare and P. ovale, J. Invest. Dermatol., 77, 314, 1981. 18. Faergemann, J., Aly, R., and Maibach, H.I., Growth and filament production of Pityrosporum orbiculare and P. ovale on human stratum corneum in vitro, Acta Derm. Venereol., 63, 388, 1983. 19. Faergemann, J., A new model for growth and filament production of Pityrosporum ovale on human stratum corneum in vitro, J. Invest. Dermatol., 92, 117, 1989. 20. Broberg, A. and Faergemann, J., Infantile seborrhoeic dermatitis and Pityrosporum ovale, Br. J. Dermatol., 120, 359, 1989. 21. Broberg, A., Faergemann, J., Johansson, S., Johansson, S.G.O., and Strannegård, I.L., Pityrosporum ovale and atopic dermatitis in children and young adults, Acta Derm. Venereol., 72, 187, 1992. 22. Ruiz-Maldonado, R., Lopez-Matinez, R., Chavarria, P., et al., Pityrosporum ovale in infantile seborrhoeic dermatitis, Pediatr. Dermatol., 6, 16, 1989. 23. Bergbrant, I.-M. and Faergemann, J., Variations of Pityrosporum orbiculare in middle-aged and elderly individuals, Acta Derm. Venereol., 68, 537, 1988. 24. Bäck, O., Faergemann, J., and Hörnqvist, R., Pityrosporum folliculitis: a common disease of the young and middle-aged, J. Am. Acad. Dermatol., 12, 56, 1985. 25. Bergbrant, I.-M., Seborrhoeic dermatitis and Pityrosporum ovale: cultural, immunological and clinical studies, Acta Derm. Venereol., Suppl., 167, 1991. 26. Williamson, P. and Kligman, A.M., A new method for the quantitative investigation of cutaneous bacteria, J. Invest. Dermatol., 45, 498, 1965.
Dermis Structure and Function Dermis Structure
Ultrasound 56 High-Frequency Examination of Skin: Introduction and Guide Jørgen Serup Department of Dermatology, Linköping University, Linköping, Sweden, and Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
Jens Keiding and Ann Fullerton LEO Pharma, Ballerup, Denmark
Monika Gniadecka and Robert Gniadecki Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 56.1 56.2 56.3 56.4 56.5 56.6 56.7 56.8
Introduction ..........................................................................................................................................................473 Physical Principles and Techniques .....................................................................................................................475 Ultrasound Velocity in Skin .................................................................................................................................475 Correlation between Ultrasonography and Histology .........................................................................................475 Correlation between Ultrasonography, Skinfold Caliper, and Radiography.......................................................476 Ultrasound Structure of Normal Skin..................................................................................................................476 Ultrasound Image Analysis ..................................................................................................................................477 Ultrasound Examination, Variables, and Practical Guidance..............................................................................478 56.8.1 Equipment, Laboratory Facility, and Examiner ....................................................................................478 56.9 Biological Variables .............................................................................................................................................480 56.10 Applications in Clinical and Experimental Dermatology ...................................................................................481 56.10.1 Inflammatory Skin Diseases ..................................................................................................................481 56.10.2 Connective Tissue Diseases of the Skin................................................................................................483 56.10.3 Cutaneous Neoplasms............................................................................................................................485 56.10.4 Leg Ulcers and the Vascular System.....................................................................................................486 56.10.5 The Nail .................................................................................................................................................486 56.11 Ultrasound Examination of Experimental Animals.............................................................................................487 References .......................................................................................................................................................................487
56.1 INTRODUCTION Ultrasound examination of the skin integument is relatively new. Since its introduction in the late 1970s, this field has, however, been in a phase of exponential growth. Computer technologies and digital imaging techniques became widely used during the same period.1 In this chapter we wish, mainly based on our own experience, to present an introduction and practical
guidance to clinicians and researchers who are starting up with the ultrasound technique. Our group has participated in the development from the very beginning with uncertainty and prototypes until the more advanced level of technical and clinical experience of today.2 The fundamental physical principle of a dermatological ultrasound scanner is the emission of high-frequency ultrasound (>10 MHz) from a transducer. The sound emission is not continuous but pulsed; i.e., the equipment 473
474
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 56.1 A-mode scan (upper half) and B-mode scan (lower half) of skin. An entrance echo is seen (left) corresponding to the epidermal surface. There is an echolucent area just underneath the epidermis. The dermis is rich in ultrasound reflections, i.e., echodense in structure. The interphase to the subcutaneous fat, which is echopoor, is well defined. The scan was obtained from a positive allergic reaction, and the band underneath the epidermis corresponded to inflammatory edema.
automatically and very rapidly switches between emission of sound and registration of the sound coming back to the same transducer (the echo) from objects being studied. The time lag between emitted and reflected sound waves depends on the physical distance between the surface of the object and the different layers of the object, which might reflect the sound. On the screen a line with peaks representing echoes from different layers, i.e., an A-mode scan is seen (Figure 56.1). The distance between peaks within the object is easily calculated when the intraobject velocity of sound is known. In B-mode scanning the transducer is automatically moved tangentially over the object and a number of A-scans are depicted and processed electronically, resulting in a cross-sectional image of the object in two dimensions, shown on the screen (Figure 56.2). In C-mode scanning a horizontal picture is depicted. The transducer is automatically moved in the horizontal level over an area of skin both along an X-axis and a Yaxis. In three-dimensional scanning a cube of data and a true three-dimensional image is obtained (Figure 56.3). In real-time scanning the scan speed allows visualization of motile structures such as arterial walls. M-mode scanning is a special procedure in which such structures and their motility pattern may be characterized. Using different modes of scanning, tissue parameters such as in vivo distance, in vivo cross-sectional area, and in vivo volume can be calculated on a purely noninvasive basis. These facilities, including documentation on a quantitative and widely observer-independent basis, make ultrasound examination an attractive and powerful tool for both research purposes and diagnostic purposes in dermatology. B-mode scanning and cross-sectional imaging is the mode of major and more widespread interest. The more advanced ultrasound techniques are mainly used in special centers of excellence. The present review will not include specialized techniques such as ultrasound microscopy and Doppler ultrasonography, but will focus on the techniques directly relevant for dermatological purposes.
z
FIGURE 56.2 The normal skin echogram as shown by B-mode scanning. The membrane between water (coupling between membrane and transducer) and gel (coupling between membrane and skin surface) is seen in the left side. The epidermis is slightly irregular in particularly corresponding to the hair follicles, which are seen as echolucent structures within the dermis and under an oblique angle. The reticular dermis (right side) creates stronger reflectance than the papillar dermis. The interphase to the echo-poor subcutaneous space (right side) is irregular.
y
x
FIGURE 56.3 Three-dimensional scan showing the bifurcation of the cubital vein, the cubital artery and the level of the dermis.
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
56.2 PHYSICAL PRINCIPLES AND TECHNIQUES The velocity of longitudinal sound waves in a tissue is determined by its elasticity and density. The acoustic characteristic impedance of a tissue is defined as the product of its density and the velocity of sound in the tissue. It is the difference in acoustic impedance between two adjacent media that determines interface echogenicity. Ultrasound reflection and refraction follow optical laws. Thus, the character of a tissue interface and the incident angle of the ultrasound beam are also important for an echo to be registered. Since the same transducer emits and records sound, the interface preferably should be perpendicular to the beam unless the interface is somewhat uneven and creates scattering of sound. It is well known from practical use of ultrasound equipment that the axial resolution is related to the center frequency of the transducer. In theory, however, resolution is not directly determined by the center frequency, but by the bandwidth (half-value range around center frequency) of the system. A thorough theoretical discussion is outside the scope of this chapter. An important consequence of these considerations is that the center frequency alone gives only limited information about the resolution of the system. With respect to resolution, a high center frequency and a bandwidth of 10 to 15 MHz are optimal for skin examination. General experience shows that in the field of dermatological ultrasonography a center frequency of 20 MHz provides a good compromise between resolution and viewing depth, and 50 MHz or higher frequencies may only be suitable for scanning of the epidermis, where a great number of other methods are easily applicable. It is often forgotten that with high frequency, the viewing field in depth becomes too small. Resolution and usefulness of a system can only be partly deduced from a list of technical specifications, and skilled use of equipment to solve real problems is the final test. It should be kept in mind that images are qualitative and open to subjective or biased evaluation unless special evaluation techniques such as digital image analysis are applied. Thus, modern principles of objective and blinded comparison should be employed whenever possible. Technical specifications to consider are: Bandwidth and center frequency Resolution (axial and lateral) Scan speed (images per second or seconds per image or real time) Swept gain (fixed or adjustable) Scanning field (B-mode or C-mode) Viewing field in depth (fixed or adjustable) Scan modes (A, B, C, or M) Measuring facilities and image analysis Image storage facilities
475
Hard-copy facilities Selection of probes and transducers Selection of display modes (color scales or split screen or zoom, etc.)
56.3 ULTRASOUND VELOCITY IN SKIN Estimates of ultrasound velocity are stratum corneum, 1550 m/s; epidermis, 1540 m/s; dermis, 1580 m/s; and subcutaneous fat, 1440 m/s.3 The average for normal fullthickness skin is 1577 m/s. Ultrasound velocity of 1580 m/s is commonly used for the calculation of total skin thickness. A study showed that ultrasound velocity of skin depended on body region (average, 1605 m/s).4 Previous studies based on oral mucosa suggested a velocity of 1518 m/sec, while studies based on human abdominal skin and porcine skin suggested a velocity of 1710 m/s.5,6 From a practical point of view, a minor deviation of ultrasound velocity from the true value of a particular location will not influence significantly the result of the thickness measurement, expressed in millimeters to one decimal point. The ultrasound velocity of the entire nail plate is 2459 m/s, and of the dorsal plate and nail matrix, 3101 and 2125 m/s, respectively.7–9
56.4 CORRELATION BETWEEN ULTRASONOGRAPHY AND HISTOLOGY Histology and electron microscopy are, with some limitations, important comparative techniques. One important difference in ultrasonography and histology is that microscopy cannot determine tissue elasticity, and in vivo elasticity is an important factor in the acoustic behavior of tissues. Histological staining of tissue specimens is a kind of desirable artifact that need not visualize significant alterations of structure demonstrable by other techniques. Thus, ultrasonography is a separate modality not directly comparable to microscopy. Some structural features are better visualized by ultrasound than histology, and vice versa. An example is the age band of the papillary dermis seen in Figure 56.4.10 In scleroderma the collagen may stain normally in histology, but it may be severely degraded in electron microscopy, and ultrasound may show an echolucent band in accordance with electron microscopy.11 Punch biopsies are of fairly limited value for comparative studies since they may undergo retraction and gross change of dimension on cutting, depending on the different circumstances.12 Nevertheless, with respect to the correlation between histological thickness, typically based on surgical biopsy, and ultrasonographic measurement of tumor thickness in malignant melanoma (Breslow thickness), a number of studies showed a remarkably high correlation between the two techniques (see Section 56.5).
476
Handbook of Non-Invasive Methods and the Skin, Second Edition
Aged skin (forearm) F 87 years
but more cumbersome way of thickness measurement is to outline the area of the block of skin on the image with the region of interest (ROI) function and divide with the width of the block.
56.6 ULTRASOUND STRUCTURE OF NORMAL SKIN
A = Subeptdermal age band
FIGURE 56.4 Low echogenic subepidermal band of aged skin of the forearm.
In vivo 20-MHz ultrasound examination does not have the resolution of histology. The advantage of ultrasound is its noninvasiveness and immediate result. A large in situ tissue block can be examined with a good presentation of the tissue microanatomy, and with free choice of body region. Thus, ultrasound is essentially a method somewhere in between subjective clinical examination and microscopy based on biopsy, depending on the problem to be evaluated. This intermediate position is also the actual situation in major and very well established fields such as abdominal and ophthalmological scanning.
56.5 CORRELATION BETWEEN ULTRASONOGRAPHY, SKINFOLD CALIPER, AND RADIOGRAPHY With skinfold caliper measurement some subcutaneous tissue is inevitably included, depending on the anatomical site, and with xeroradiography the dermis–fat interface is somewhat vague.13,14 Previous studies on the correlation of the different methods for evaluation of skin structures were based on A-mode ultrasonography only. There was generally a fair correlation between the methods, but also discrepancies in, for example, inflamed skin.15–23 There is a general trend that ultrasound shows lower thickness values than the other methods, as might be expected. There are no studies directly comparing thickness measurement of normal skin based on B-mode 20-MHz ultrasound with thickness measurement using the previous methods because researchers of today find the ultrasound method based on B-mode scanning clearly much more accurate than the previous methods. In our laboratory the reproducibility is 0.05 mm. Three A-scan lines from the top, middle, and bottom of the B-scan image are selected and the average calculated. A theoretically more precise
The epidermis and the dermis reflect ultrasound variably, but with well-defined interface echoes toward ambient air or coupling medium and toward the underlying subcutaneous fat. The normal skin echogram (Figure 56.2 and Table 56.1) was described in previous publications and in the different contributions in a recent book.24–27 Epidermal echoes may be disturbed by air contained within scales (particularly psoriatic scales) and by the keratotic material of seborrheic keratoses, which causes heavy reflections and shadows that are characteristic or even pathognomonic. The epidermis–dermis interface is obviously very uneven, but observations in psoriasis and acanthosis indicate that the ultrasound interface between epidermis and dermis is mainly determined by the top of the rete papillae.28 This is probably also the case in normal skin. Epidermis itself is low reflectant in its internal structure. By 20-MHz A-mode scanning one internal epidermal echo is seen close to the entry echo. This profile is obvious in palms and soles, where the epidermis is thick. The internal epidermal echo probably represents the water barrier zone of the skin, comparable to the internal echo found in the nail plate.7 Reflections from peaks and valleys of dermatoglyphic ridges may be seen with 50-MHz transducers.29,30 The epidermal thickness is easily measured on palms and soles, where the interface to the very low reflectant dermis of these regions can be reliably defined. Dermal echoes are, in most body regions, many and variable (Figure 56.2 and Table 56.1). They originate from the well-organized fiber network of the dermis, which is also responsible for the tensile properties of skin and the Langer lines. Affections that erode or disturb this network cause low reflectancy. Subepidermal increase of interstitial water in edema is a common cause of low reflectancy (Figure 56.1). Dermal echoes may be influenced by the distensibility state of the skin, and thus by the position of joints. Hair follicles and sebaceous glands are sometimes seen, depending on the body region (Figure 56.2). Thus, the normal and undisturbed regular fiber network of the dermis is a kind of natural contrast medium in which different pathologies can be outlined if they cause low reflectancy, or disturbance of interfaces and dimensions. Palms, soles, and, to some extent, the face and the scalp are exceptions. In these regions the fiber orientation is variable, and ultrasound reflectancy is consequently less. In neonatal skin, particularly in premature infants, the
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
477
TABLE 56.1 Skin Thickness (Full-Thickness) and Acoustic Density of the Dermis Relative to Anatomical Site Females (n = 10) Skin Thickness (mm) Forehead Cheek Neck, anterior Neck, posterior Pectral area Abdomen Back, upper Back, lower Upper arm, anterior Upper arm, posterior Forearm, extensor Forearm, flexor Hand, dorsal Palm Thigh, anterior Thigh, posterior Crus, anterior Crus, posterior Foot, dorsal Ankle Heel Sole
1.79 1.49 1.34 1.92 1.77 1.62 2.33 2.09 1.45 1.05 1.36 1.12 1.26 1.50 1.42 1.46 1.34 1.30 1.49 2.08
± 0.28 ± 0.17 ± 0.17 ± 0.40 ± 0.21 ± 0.20 ± 0.32 ± 0.34 ± 0.34 ± 0.09 ± 0.25 ± 0.19 ± 0.18 ± 0.52 ± 0.12 ± 0.16 ± 0.20 ± 0.12 ± 0.31 ± 0.38 — 1.53 ± 0.29
Males (n = 8)
Acoustic Density (a.u.) 16.5 ± 23.7 ± 31.4 ± 17.3 ± 24.2 ± 23.4 ± 14.8 ± 16.0 ± 28.0 ± 30.6 ± 27.4 ± 29.3 ± 25.2 ± 14.3 ± 30.3 ± 25.9 ± 29.0 ± 29.9 ± 27.2 ± 13.9 ± — 10.0 ±
8.6 10.7 9.3 8.9 6.8 6.7 5.9 7.8 10.8 6.2 8.5 7.1 8.4 7.8 7.4 5.0 7.5 7.3 4.9 9.4 5.9
Skin Thickness (mm) 2.19 1.83 1.61 2.09 1.92 1.88 2.62 2.21 1.53 1.21 1.42 1.31 1.50 1.48 1.59 1.51 1.42 1.34 1.74 2.01
± 0.23 ± 0.15 ± 0.38 ± 0.21 ± 0.14 ± 0.12 ± 0.44 ± 0.35 ± 0.18 ± 0.12 ± 0.14 ± 0.13 ± 0.14 ± 0.45 ± 0.21 ± 0.21 ± 0.19 ± 0.14 ± 0.26 ± 0.44 — 1.60 ± 0.28
Acoustic Density (a.u.) 8.9 ± 3.3 13.4 ± 7.10 28.0 ± 10.1 13.4 ± 4.6 22.5 ± 7.8 18.3 ± 5.9 10.1 ± 2.6 12.1 ± 4.8 22.5 ± 8.1 26.1 ± 8.0 26.6 ± 7.4 27.0 ± 5.3 23.5 ± 5.9 9.7 ± 2.7 25.6 ± 9.03 25.8 ± 4.3 27.1 ± 8.5 24.4 ± 4.5 24.5 ± 9.8 6.7 ± 2.3 — 9.0 ± 3.3
Note: Mean ± SD; a.u. = arbitrary units. Results were obtained in 18 healthy individuals (ages 24 to 41) with the Dermascan C® and the inbuild image analysis program of this equipment.26 Imaging of the dermis was difficult in the palm and the sole, and not possible on the heel. In some individuals the face might also be difficult to scan.
whole dermis is low reflectant or echo lucent.31 In early infancy the dermal echo pattern changes toward a normal or adult pattern. In aged skin a well-defined subepidermal band of low reflectancy appears10,32 on sun-exposed sites, such as the forearm (Figure 56.4). The skin becomes thin in old age, particularly on distal extremities and the dorsum of the hand, unless sun damage and repair with actinic elastosis results in thickening. Bleeding and bruises in senile skin progress in this subepidermal zone of low mechanical resistance. In advanced corticosteroid atrophy a similar alteration is seen. The subcutaneous space is normally low reflectant. Low reflectancy depends, however, on the equipment and the gain. With high gain, subcutaneous veins are seen as dark structures. Other anatomical structures, such as hypoechogenic tendons, can also be visualized by proper adjustment of the gain. On the neck, chest, and back the subcutaneous fascia is often visible. The muscle fascia is easy to define with smooth surfaces, especially toward the muscle, while it may have
attachments of retinacula toward the fat. Muscles have few internal echoes. Bone causes heavy reflection. Ultrasound measurement of in vivo distances can be no more reproducible than the actual biology. Obviously, rete papillae do not constitute a line or a plane, and the interface between dermis and subcutaneous fat is far from smooth due to attachments of subcutaneous retinacula (Figure 56.5). The anatomical thickness of the subcutaneous fatty layer is even more variable. Thus, the biology itself is so noisy that an extremely precise distance measurement can never be attained, even if the ideal scanner existed. There is a popular but obviously incorrect view that if the frequency of the ultrasound equipment is high enough, all problems of variation and precision are overcome.
56.7 ULTRASOUND IMAGE ANALYSIS S. Seidenari and A. DiNardo of Modena, Italy, developed the field of ultrasound image analysis of skin. In a series of publications they have demonstrated the utility of the
478
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 56.5 3D-ultrasound reconstruction of the skin surface (upper part) and the interphase between the reticular dermis and the subcutaneous tissue (lower part) demonstrating the irregularity of the latter.
amplitude range of particular importance, part of an image can be highlighted and characterized by a value corresponding to the number of amplitudes within that range. In the image analysis system amplitudes are represented by pixels (picture elements), which may vary from 0 to 255 in intensity. By selecting a bandwidth from 0 to 30, the hypoechogenic part of the image typical for edema and inflammation may be highlighted. In psoriasis a band between 0 and 10 appears more discriminative. By means of a band of 201 to 255 the echo-dense areas in the tissue may be selectively quantified. Textures in tissue echogram, speckle formation, and clinical relevance were reviewed in a recent paper.41 (Seidenari will give a detailed description in Chapter 57.)
56.8 ULTRASOUND EXAMINATION, VARIABLES, AND PRACTICAL GUIDANCE 56.8.1 EQUIPMENT, LABORATORY FACILITY, EXAMINER
FIGURE 56.6 The Dermascan C® originally developed in our laboratory as a prototype.2
technique for the characterization of a number of clinical conditions, including allergic and irritant contact dermatitis, corticosteroid effects, and psoriasis.33–40 The Modena group works with the Dermascan C® scanner and the Dermavision 2D® dedicated software (Cortex Technology, Hadsund, Denmark), enabling the selection of amplitude range of interest and the transformation into a binary color system (Figure 56.6). By attributing one color to a selected
AND
To perform state-of-the art echography of skin, a specialized scanner is needed (Figure 56.6). Our experience is primarily based on the Dermascan C, which was originally developed as a prototype in our laboratory and described in a separate paper.2 A center frequency of 20 MHz generally represents the best compromise between resolution and the need for a certain viewing field in depth. The resolution of the system is as previously mentioned, mainly determined by the bandwidth rather than the center frequency. Generally, 20-MHz scanners have an axial resolution in skin of 0.05 mm and a lateral resolution of 0.15 to 0.35 mm. Thus, the axial resolution is better than the lateral. The viewing field in depth is typically 15 to 25 mm. It is important that the laboratory facility is adequately equipped as a diagnostic facility for ultrasound evaluation of skin. It is typically a problem that dermatological clinics have no tradition for this type of diagnostic procedure. Obviously, the needs for space, assistance, and time must be the same as they are at other hospital departments with routine functions in diagnostic radiology or echography. The examiner must be familiar with the equipment, the way it is operated, the background literature, and must have adequate knowledge about skin structure, dermatological application, and interpretation. There is no formal training or education in dermatological echography, and the examiner has to get the insight and routine in operation on his or her own. The field is covered by a number of recent reviews.1,2,27,42–51 It is a fundamental rule that the body site, which is examined, must be kept in a fixed position during scanning. Examinations on the trunk are best performed with the patient resting in the supine position. Next to the
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
FIGURE 56.7 Specing device used in our laboratory to ensure scan location, direction and standard thickness (1 mm) of the gel layer.
FIGURE 56.8 (a) B-mode scan of forearm skin. To ensure constant axial positioning the image is oriented vertically and placed so that the probe membrane and the epidermal surface covers their respective ink marks on the monitor screen.
FIGURE 56.8 (b) B-mode scan of same examination site but under an oblique angle. Note the decrease of acoustic density of the dermis.
479
fundamental rule of the fixed positioning of the body site being examined, the significance of the scan angle and scan direction and awareness of the importance of gain setting (amplification) of the equipment, gel layer thickness, and the axial positioning are important. In image analysis of acoustic density of the skin, these prerequisites must be carefully standardized. We routinely mark the examination site and apply a spacing device to ensure scan location and direction as well as a 1-mm gel layer thickness defined by the thickness of the spacing device (Figure 56.7). We also work with two ink marks on the monitor screen (one for the probe membrane, another for the epidermal surface) to ensure axial positioning relative to the transducer. The more the scan angle deviates from the right angle, the less sound will be reflected back to the scanner, and the dermis will appear of lower acoustic density (Figure 56.8). The dermis may be measured thicker unless the change of density and delineation of the dermis–fat interface happens to compensate for the deviation from the right angle. The epidermal surface must be oriented strictly vertically on the monitor screen. The scan direction in the examination field should always be the same since the image may be influenced by differences in the fiber orientation within the dermis, as represented by the Langer lines. There are, in most body regions, systematic differences even within relatively small local distances. If a skin condition is monitored over time, the examination field must be carefully marked to permit follow-up examination at exactly the same position. There is at the moment no ultrasound phantom for 20MHz scanners available except a rubber plate with a surface texture close to the epidermal microrelief (available through Cortex Technology, Hadsund, Denmark). With the rubber phantom the gain may be adjusted to a level where the entrance echo reaches the upper margin of the monitor screen. For follow-up check of speckles and internal structure, the examiner may use his own flexor side forearm. There is a certain interinstrument variation. It may be of limited value to operate with the same gain in every situation. For the study of forearm skin we operate our two Dermascan C scanners with linear fixed gain ranging from 16 to 22 dB. In micropigs we have used 19 dB; in the rat, 25 dB; in mice, 20 dB; and in hairless guinea pigs, 23 dB (but for image analysis a gain of 18 or 19 dB). It is generally the most convenient and reproducible way to operate with a linear gain setting. When groups are studied or monitored over time, the equipment, as mentioned, must be ensured to remain constant (the fixed-gain method). However, the chosen gain setting is not optimal for any skin lesion or any body site. If scanning is performed as a diagnostic procedure in dermatological clinics to obtain a maximum of information in individual cases, whatever they might suffer from corticosteroid atrophy, scleroderma, or nodular malignant melanoma, it is far more fruitful not to operate with a
480
Handbook of Non-Invasive Methods and the Skin, Second Edition
Male Female
2.0
1.0
5
10 Region number
15
2.0
1.0
30 40 10 20 Mean density of full-thickness dermis (a.u.)
Male Female
40 30 20
FIGURE 56.10 Skin thickness vs. mean density of full thickness dermis on extremities (closed circles) and truncal skin (open circles) demonstrating the echodense structure of the thinner extremity skin. Results obtained with the Dermascan C® and the inbuild software of this equipment.26 21 different body sites were examined.
• • • • •
10 5
10 15 Region number
Extremities Trunk
20
FIGURE 56.9 (a) Skin thickness in males (closed circles) and females (open circles) vs. anatomical site (region by number, see Table 56.1). Results were obtained with the Dermascan C®.26 Females have more thin skin.
Density of full-thickness dermis (a.u.)
3.0 Skin thickness (mm)
Skin thickness (mm)
3.0
20
FIGURE 56.9 (b) Ultrasound density of full thickness dermis in males (closed circles) and females (open circles) vs. anatomical site (region by number, see Table 56.1). Results were obtained with the Dermascan C® and the inbuild software of this equipment.26 Females have more dense skin.
fixed-gain setting, but to perform the examination in real time with adjustment of the gain on the live image (the live image gain adjusted method). A pigmented seborrheic wart will, at a gain setting normally used, create a heavy shadow, but after gaining up, a normal dermis without tumor erosion is visualized, differentiating this lesion from a nodular melanoma (personal observation).
56.9 BIOLOGICAL VARIABLES Biological variables of ultrasonographic significance include variations relative to sex, age, and body site (Table 56.1 and Figure 56.9a and b and Figure 56.10), and diurnal variation related to orthostatic position as well as endogenous biorhythms. Moreover, body weight and physical activities may be influential:
• • • •
• • •
•
Females have more thin skin than males.25,26,52 Extremity skin is thinner than truncal skin.25,26,52 Facial skin is low echogenic.26 Skin of the palm and sole is echo poor or echo lucent, however, with a thick epidermis.26,29,30,52 The dermis–subcutaneous tissue interface is better defined in females than it is in males, since females have a more echo dense dermis.26 Thin skin is generally more echogenic (female > male, extremity skin > truncal skin).26 Reticular dermis is more echogenic than papillar dermis.10,27,32 Children have more thin skin.31 Senile skin is thinner, particularly on extremities; a subepidermal low-echogenic band appears.10,32,47 Thin (and more transparent) extremity skin of old people may correlate with osteoporosis.53,54 Overweight persons may have more thick skin.55 Sportsmen and persons with heavy physical activity and sun exposure may have more thick skin.56 Hot weather may be associated with cutaneous hyperemia, extravasation, and generalized cutaneous edema, which may influence ultrasound scanning significantly.140
The echogenicity of the skin shows diurnal variation (Figure 56.11 and Figure 56.12) originating from two different mechanisms.57–59 The skin becomes generally slightly more echo dense during the daytime, irrespective of orthostatic position, body site, and age. The skin on the
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
1
2
481
3
An
Fo
FIGURE 56.11 Diurnal variation of ultrasound density in an aged individual in the morning (1), 2 h after standing (2), and after 12 h (3) in the ankle region (An) and on flexor side forearm skin (Fo).57 1
2
3
An
Fo
FIGURE 56.12 Diurnal variation of ultrasound density in a young individual in the morning (1), 2 h after standing (2), and after 12 h (3) in the ankle region (An) and on flexor side forearm skin (Fo).57
legs of aged individuals, on the other hand, becomes less echogenic during daytime. This orthostatic decline in density, probably related to water accumulation, may also be observed in a minor part of the population of young individuals; however, the echogenicity of their skin remains stable if they stay supine during the whole day (Figure 56.13). The adaptation to orthostatic position mainly takes place the first 2 h after standing in the morning. The above-described variables need to be taken into account when studies including ultrasound examination
are performed. Controlling these variables is to control noise, and to go for precision on a minimum study sample.
56.10 APPLICATIONS IN CLINICAL AND EXPERIMENTAL DERMATOLOGY 56.10.1 INFLAMMATORY SKIN DISEASES Ultrasound imaging shows that the process of inflammation does not spread evenly throughout the dermis, but
482
Handbook of Non-Invasive Methods and the Skin, Second Edition
0.4
Diurnal change of LEPs
∗
0.0 ∗ ∗
∗
∗ −0.4 Ankle
Calf
Thigh
Forearm
Arm
FIGURE 56.13 Average change of low echogenic pixels during the day (mean ± 2 SE). Open bars represent a group of young individuals (n = 22, age 17–27), hatched bars a group of aged individuals (n = 22, age 75–100), and black bars a group of subjects (n = 10, age 17–23) who remained in the supine position in bed during the whole day. The LEP (low echogenic pixels, Dermavision software, Cortex Technology, amplitude range 0–30) represent differences between values in the morning and 12 h later. Asterisk indicates significant changes (p <0.05). Young individuals (open bars) showed an increase in acoustic density (decrease in LEP) during the day while aged individuals (hatched bars) showed a decrease in echo density particularly in the ankle region. Individuals who remained supine (black bars) were constant in echo density during the daytime.58
concentrates in the outermost part corresponding to the papillary dermis, where swelling and a subepidermal band is formed60 (Figure 56.1). This is a common trait in any kind of inflammation. The structural background is disturbance of the fibrillar network due to edema formation. The papillar dermis is more easily distended under the pressure of edema than the reticular dermis. The skin becomes thicker at the same time, and the skin surface may become folded, as it is seen in lichenification in atopic dermatitis, where the reticular dermis at the same time remains straight and undisturbed, according to ultrasonography.61,62 Skin thickening was previously described in positive allergic patch test reactions to the classical contact allergens (Figure 56.1), tuberculin and recall allergens,63–68 as well as irritant reactions elicited by sodium lauryl sulfate (SLS), and as a number of other primary irritants, including dimethylsulfoxide (DMSO).64,68–74 The Modena group showed, however, that ultrasound image analysis is a more powerful tool to quantify those reactions than simple thickness measurement, although thickness measurement has the advantage of being easier to do.33–40 The degree of edema formation after SLS depends on variables such
FIGURE 56.14 Wheal of urticaria with elevation and flattening of the skin surface and a decrease of echo density of the dermis.
as body site, complexion of the skin, menstrual phase, etc. Wheals of urticaria (Figure 56.14) tend to involve the dermis more deeply; however, in pseudopodia, a protrusion of the edema at the subepidermal level, where band formation normally takes place, can be seen. Thickening of wheals of urticaria does not develop linearly.75 When a wheal has reached a diameter of approximately 5 mm, the progression in thickness stops, and further development takes place laterally. At the same time, the wheal changes in shape from globoid to flat. This dynamic behavior is a manifestation of the mechanical behavior of the papillar dermis and its ability to resist the pressure of edema, as previously outlined. Effects of antihistamines can be documented.76 The Modena group also demonstrated the utility of image analysis for characterization of wheals.77 Inflammatory lesions of acne are echo poor, with deformation of the skin surface contour around the hair follicle. Interestingly, underneath atrophic scars a hyperechogenic massive, probably representing a subcutaneous scar, is quite often seen141 (Figure 56.15). This is an example of how ultrasonography can provide unexpected information not hitherto available by ordinary techniques such as histology, which cannot be practiced routinely in the face. In psoriasis, thickening and a subepidermal band are also seen (Figure 56.16), representing acanthosis and inflammation at the same time.28,39,78–82 Scales may create shadows. Pretreatment with petrolatum may be necessary. In pustular psoriasis the same type of abnormality is seen, however, focal in distribution. Skin thickness measurement by ultrasound is superior to color measurements as regards following the healing of psoriasis, as exemplified by the psoriasis plaque test.80 At the end of treatment the skin thickness normalizes while redness often persists, and this is reflected in the parameters measured.
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
FIGURE 56.15 (a) Early inflammatory lesion of acne with decrease of acoustic density of the papillar dermis and the dermis corresponding to the hair follicle.
FIGURE 56.15 (b) Advanced acne lesion with deformation of the skin contour and a pathological echodense area in the subcutaneous space (right side of image).
Special diseases, such as nodular prurigo, exhibit echographic alterations in accordance with inflammatory skin diseases, in general with focal band formation, and shadow due to scratch and hyperkeratosis.
56.10.2 CONNECTIVE TISSUE DISEASES
OF THE
SKIN
Scleroderma was one of the first applications of dermatological ultrasonography.83–86 The skin surface to bone distance over the digits, and the skin thickness on the dorsum of the hand and the forearm were measured (Figure 56.17 and Figure 56.18). Both the acrosclerotic and localized
483
FIGURE 56.15 (c) Atrophic acne scar with no sign of active inflammation. A hyperechogenic mass is seen in the subcutaneous space possibly representing the scar formation after a previous inflammatory process in the subcutaneous tissue.
(morphea) types of scleroderma, as well as scleroderma, have been extensively studied.87–93 Acroscleroderma first clears on proximal parts. It is important that monitoring includes both forearm skin thickness measurement and measurement of skin–phalanx distances. In localized scleroderma, spontaneous regression of thickening takes place and the final outcome may be thinning and pigmentation.85,86 Thickness measurement has generally turned out to be more useful than measurement of skin elasticity for the purpose of following therapy of scleroderma because the stiffening of the skin and the contractures tend to be permanent once established.85 In cases of Pasini–Pierini atrophy and in selected cases of localized scleroderma, the skin is thin from the very beginning, and there may be loss of subcutaneous fat and even bony depression. Subepidermal band formation can occasionally be seen in morphea, which can also be bullous in rare cases.11 If subepidermal band formation is observed in a patient with acroscleroderma during treatment with penicillamin, particularly over body sites exposed to pressure and accompanied with bruises, penicillamine dermatopathy or elastosis perforans serpiginosa should be suspected and the treatment taped out. It might be unethical to institute risky and expensive treatments such as photopheresis without proper monitoring and quantification to demonstrate efficacy and justify long-term treatment. With ultrasound, the thickening of the muscle fascia in fascitis and the dermal thickening and serrate fat interface corresponding to dimpling clinically in chronic graft vs. host reaction may be demonstrated.141 Dermal thinning during development of corticosteroid atrophy, typically 0.2 to 0.3 mm after 4 weeks, irrespective
484
Handbook of Non-Invasive Methods and the Skin, Second Edition
Generalized scleroderma
Healthy control
Over middle phalanx
Over proximal phalanx
FIGURE 56.16 (a) Psoriatic plaque with a subepidermal low echogenic band. Extensor aspect of forearm
Flexor aspect of forearm
FIGURE 56.16 (b) Pustular psoriasis with a subepidermal low echogenic band and areas of more advanced alteration of the echo structure.
of body site, is easy to measure following potent corticosteroid treatment, but the decline in thickness may not allow a detailed grouping of any corticosteroid.19,42,95–100 Open application of the corticosteroid is preferable since occlusive application may result in edema formation, which will interfere with the corticosteroid-induced suppression of ground substance and skin thinning. Posttreatment normalization of thickness takes place within a few weeks.42 It was recently observed that ultrasound image analysis offers an extra opportunity for detailed study of the effects of corticosteroid on dermal connective tissue, i.e., with a drop in low-amplitude echoes.142 Ultrasound is the first-line quantitative method for documentation of the corticosteroid effect using the psoriasis plaque test.80 Traumatic scars, including hypertrophic burn scars, normally are irregular and relatively echo poor due to the
1 mm
FIGURE 56.17 Ultrasound A-mode measurement of skin surface-phalanx distance and skin thickness (0–0) on the third finger and the forearm. A patient with generalized scleroderma (acrosclerosis) is compared with a healthy control. Note that the thickening becomes more advanced on the acral examination sites.
poor organization of the connective tissue fibrils of such tissues.101 Underneath atrophic acne scars a hyperechogenic massive may be found, as previously mentioned. Thermal injury and ultrasound determination of the burn–nonburn interface in the dermis was evaluated in animal as well as human experiments.101–103 The skin swells the first hour postburn, but this does not directly allow the determination of the vitality of the dermis at a certain level. Burns were not studied by state-of-the-art 20-MHz ultrasound skin scanners and image analysis. Logically, the critical question should be whether the echogenicity of the
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
FIGURE 56.18 Horizontal scanning in the front of a patient with localized scleroderma en coup de sabre. The sclerotic skin (upper part) is thickened, the subcutaneous fat is reduced, and there is a depression of the bony contour of the frontal bone. The skin of the perilesional area (lower part of the picture) appears normal.
485
FIGURE 56.19 (a) Basel cell carcinoma in the face. The tumor has not eroded through the reticular dermis and reached the subcutaneous tissue.
reticular dermis decreases significantly during the first 48 to 72 h postburn as a sign of vital reaction with inflammation. A necrotic reticular dermis is not expected to undergo this change of echo structure.
56.10.3 CUTANEOUS NEOPLASMS Cutaneous neoplasms invade and erode the connective tissue of the dermis, and they are consequently seen as echo-poor or echo-lucent structures on ultrasound (Figure 56.19). The literature on cutaneous and subcutaneous neoplasms is extensive and also conclusive with respect to the precision and usefulness of ultrasonography in preoperative evaluation and monitoring.104–131 There is a wide overlap between the internal echo structure of basal cell carcinoma, squamous cell carcinoma, and melanoma, and definite differentiation between these tumors on the basis of internal echo structure is not possible, although findings may be characteristic or typical. Moreover, benign nevi are typically echo poor as well, and distinction from dysplastic nevi by ultrasonography is uncertain or impossible. However, this distinction may also be difficult in histology and cause controversy. Ultrasonography is nevertheless helpful in the tumor clinic. Ultrasonographic tumor thickness and depth evaluation are rational prior to cryosurgery. It is important to know if a basal cell carcinoma has broken through the dermis into the subcutaneous space, and if Moh’s surgery is indicated. The studies on malignant melanoma performed with 20-MHz state-of-the-art scanners have
FIGURE 56.19 (b) Basel cell carcinoma in the face. The tumor has reached the subcutaneous space, and it also has spread laterally into the dermis at a superficial and more profound level.
FIGURE 56.19 (c) Basel cell carcinoma on the leg. Despite the size of the tumor, there is no erosion through the reticular dermis into the subcutaneous tissue.
486
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 56.20 Kaposi sarcoma with irregular internal echostructure.
FIGURE 56.21 B-scan of the skin surrounding a venous leg ulcer. A pronounced subepidermal low echogenic band is seen, and dilated vessels within the dermis may be observed.
shown a high (>90%) correlation between ultrasound measurement of melanoma thickness and histological thickness, despite the known changes of dimension of small biopsies during operation and histological precessing.12 It is useful to know whether the patient should be prepared for a small biopsy or a wide excision. The 20MHz ultrasonography will never have the resolution to directly replace histology for the final diagnosis, but ultrasonography can be helpful in the planning of the strategy of treatment, including more precise information of the patient early in the treatment phase. Distinction from pigmented seborrhoeic keratosis is easy since these create heavy shadows in ultrasonography. In clinical dermatology, however, there is no real tradition for more complicated instrumentation directly used by the clinicians to support diagnosis and planning. This is very much in contrast to other specialties, such as ophthalmology. Kaposi sarcoma has a special ultrasound structure (Figure 56.20) with a lump of separate islands of tumor tissue echo poor. The depth of the Kaposi lesions may be of interest when laser treatment is planned and the modality of laser light selected. Infiltrates of cutaneous lymphoma spread primarily in the subepidermal space as any infiltration. The subepidermal space may be enormously distended, however, with a continuous but deformed reticular dermis underneath.141 It is useful to know the thickness of the infiltrate prior to radiotherapy.
echogenic subepidermal band (Figure 56.21) appears, especially in skin surrounding the ulcer.47,57,58,59,133 This indicates a severe disorganization of structures at the level of the superficial vascular plexus of the skin, with possible consequences for the nutritional supply of the epidermis. Dilated intradermal vessels may also be seen as a pathological observation. Ultrasonography is important for characterization and delineation of lipodermatosclerosis. In lymphedema the total skin becomes thicker; however, in this condition no special band formation is observed, and there is no major tendency to leg ulceration. Patients with venous leg ulcer seem to suffer from a reduced mechanical resistance of the skin of the lower leg and distension and accumulation of edema, especially at the level of the papular dermis, with the consequences mentioned.47,57,133 Compressive bandages and stockings were shown to diminish the ultrasonographic findings described.56,57 The intradermal edema is a key factor in venous leg ulcer, and obviously, ultrasound is an advanced and powerful tool for this application. Small-size arteries such as the temporal arteria and the radial arteria can be demonstrated by ultrasound, and the diameter measured. With M-mode scanning the pulse waves can be followed. The effect of antimigraine therapy was evaluated by ultrasonographic studies of the temporal arteria, and also the degree of dilation relative to clinical symptoms was described.134–137
56.10.4 LEG ULCERS
56.10.5 THE NAIL
AND THE
VASCULAR SYSTEM
In arterial ischemia the skin of the distal leg becomes transparent and thin. In venous insufficiency and venous leg ulcer the skin becomes thicker, and a prominent low
The nail thickness is easily measured by ultrasound, particularly with the small and more convenient A-mode scanners.7,8 The correct ultrasound velocity in nail tissue
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
needs to be used.7 There is a quite strong internal echo in the nail between the dorsal plate and the matrix, and there is a weak echo in the nail bed between the underlying epidermis and the connective tissue of the distal phalanx, which is echo poor. This echo-poor zone between the nail and the phalanx bone extends proximally under the nail bed, with a relatively sharp delineation toward extensor side skin and the distal interphalangeal joint.
56.11 ULTRASOUND EXAMINATION OF EXPERIMENTAL ANIMALS Ultrasound examination may be possible in animals covered with hair, however, it is easier and more reliable in the hairless species. The researcher has to be familiar with the histology and microanatomy of the animal as a prerequisite to interpret ultrasound images. According to our experiences, animal skin is generally less echogenic than human skin. The interface between dermis and subcutaneous fat is acoustically well defined both in hairless mice and in guinea pigs and rats. It was recently shown that ultrasound image analysis is useful for quantitation of irritant reactions to SLS in hairless mice.138 Minipigs and domestic pigs have a low echogenic or echo-poor dermis, and it is often impossible to measure the thickness of the skin. In the Yucatan micropig the skin thickness is easily measured, and ultrasound is a useful tool to follow corticosteroid atrophy in experimental studies with this animal. Ultrasound imaging of the hamster flank organ was shown to correlate with histology and to be convenient for monitoring of hormone effects on this specialized sebaceous gland.139
REFERENCES 1. Perednia DA. What dermatologists should know about digital imaging. J Am Acad Dermatol 25:89–108, 1991. 2. Serup J. Ten years’ experience with high-frequency ultrasound examination of the skin: development and refinement of technique and equipment. In Ultrasound in Dermatology, P Altmeyer, S el-Gammal, K Hoffmann, Eds. Springer-Verlag, Berlin, 1992, pp. 41–54. 3. Edwards C, Payne PA. Ultrasound Velocities in Skin Components. Paper presented at the International Society for Bioengineering and the Skin: Ultrasound in Dermatology, Liege, Belgium, November 9, 1984. 4. Escoffier C, Querleux B, de Rigal J, Lévêque J-L. In vitro study of the velocity of ultrasound in the skin. Bioeng Skin 2:87–94, 1986. 5. Daly CH, Wheeler JB. The use of ultra-sonic thickness measurement in the clinical evaluation of the oral soft tissues. Int Dental J 21:418–429, 1971.
487
6. Dines KA, Sheets PW, Brink JA, Hanke CW, Condra KA, Clendenon JL, Gross SA, Smith DJ, Franklin TD. High frequency ultrasonic imaging of skin: experimental results. Ultrasonic Imaging 6:408–434, 1984. 7. Jemec GBE, Serup J. Ultrasound structure of the human nailplate. Arch Dermatol 125:643–646, 1989. 8. Finlay AJ, Western B, Edwards C. Ultrasound velocity in human fingernail and effects of hydration: validation of in vivo nail thickness measurement techniques. Br J Dermatol 123:365–373, 1990. 9. Jemec GBE, Agner T, Serup J. Transonychial water loss: relation to sex, age and nail-plate thickness. Br J Dermatol 121:443–446, 1989. 10. de Rigal J, Escoffier C, Querleux B, Faivre B, Agache P, Lévêque J-L. Assessment of aging of the human skin in vivo. Ultrasound imaging. J Invest Dermatol 93:621–625, 1989. 11. Kobayasi T, Willeberg A, Serup J, Ullman S. Morphea with blisters. Acta Derm Venereol (Stockh) 70:454–456, 1990. 12. Serup J. Punch biopsy of the skin by electric power drill: effect of speed. Dan Med Bull 32:189–191, 1985. 13. Tanner JM, Whitehouse RH. The Harpenden skinfold caliper. Am J Physiol Antropol 13:743–746, 1955. 14. Black MM. A modified method for measuring skin thickness. Br J Dermatol 81:661–666, 1969. 15. Dykes PJ, Francis AJ, Marks R. Measurement of dermal thickness with the Harpenden skinfold caliper. Arch Derm Res 256:261–263, 1976. 16. Dykes PJ, Marks R. Measurement of skin thickness: a comparison of two in vivo techniques with a conventional histiometric technique. J Invest Dermatol 69:275–278, 1977. 17. Alexander H, Miller DL. Determining skin thickness with pulsed ultrasound. J Invest Dermatol 72:17–19, 1979. 18. Tan CJ, Roberts E, Stathan B, Marks R. Reproducibility, validation and variability of dermal thickness measurement by pulsed ultrasound. Br J Dermatol 105:25–26, 1981. 19. Tan CJ, Marks R, Payne PA. Comparison of xeroradiographic and ultrasound detection of corticosteroid induced dermal thinning. J Invest Dermatol 76:126–128, 1981. 20. Newton JA, Whitaker J, Sohail S, Young MMR, Harding SM, Black MM. A comparison of pulsed ultrasound, radiography, and micrometer screw gauge in the measurement of skin thickness. Curr Med Res Opin 9:113–118, 1984. 21. Lawrence CM, Shuster S. Comparison of ultrasound and caliper measurements of normal and inflamed skin thickness. Br J Dermatol 122:195–200, 1985. 22. Stautner-Brückmann C, Reitmeier K, Gressner U, Zöllner N. Vergleichende Messing der Hautdicke des Handrückens mit Caliper und Ultraschall: erste Ergebnisse einer prospektiven Untersuchung. Blindgebund 57:67–69, 1990.
488
Handbook of Non-Invasive Methods and the Skin, Second Edition
23. Martin AD, Drinkwater DT, Clarys JP, Daniel M, Ross WD. Effects of skin thickness and skinfold compressibility on skinfold thickness measurement. Am Hum Biol 4:453–460, 1992. 24. Miyauchi S, Miki J. Normal human skin echogram. Arch Dermatol Res 275:345–348, 1983. 25. Fornage BD, Deshayes J-L. Ultrasound of normal skin. J Clin Ultrasound 14:619–622, 1986. 26. Olsen LO, Takiwaki H, Serup J. High-frequency characterization of normal skin. Skin thickness and echographic density of 22 anatomical sites. Skin Pharmacol, submitted. 27. Altmeyer P, el-Gammal S, Hoffmann K. Ultrasound in Dermatology. Springer-Verlag, Berlin, 1992. 28. Olsen LO, Serup J. High-frequency scanning for noninvasive cross-sectional imaging of psoriasis. Acta Derm Venereol (Stockh) 73:185–187, 1993. 29. Murakami S, Miki J. Human skin histology using highresolution echography. J Clin Ultrasound 17:77–82, 1989. 30. el-Gammal S, Hoffmann K, Auer T, Korten M, Altmeyer P, Höss A, Ermert H. A 50 MHz high-resolution ultrasound imaging system for dermatology. In Ultrasound in Dermatology, P. Altmeyer, S. el-Gammal, K. Hoffmann, Eds. Springer-Verlag, Berlin, 1992. 31. Petersen JR, Petersen S, Serup J. High-frequency ultrasound measurement of subcutaneous fat thickness in newborn infants. Skin Res Technol, in press. 32. Richard S, Querleux B, Bittoun J, Folivet O, Idy-Perelli I, Lachaniere O, Lévêque J-L. Characterization of the skin in vivo by high resolution magnetic resonance imaging: water behaviour and age-related effects. J Invest Dermatol 100:705–709, 1993. 33. Seidenari S, di Nardo A. A new image analysis system for the assessment of allergic patch test reactions recorded by B-scanning. Contact Derm 25:329, 1991. 34. Seidenari S, di Nardo A, Pepe P, Gianetti A. Ultrasound B-scanning with image analysis for assessment of allergic patch test reactions. Contact Derm 24:216–222, 1991. 35. Seidenari S, di Nardo A. Cutaneous reactivity to allergens at 24-h increases from the antecubital fossa to the wrist: an echographic evaluation by means of a new image analysis system. Contact Derm 26:171–176, 1992. 36. Seidenari S, di Nardo A. Echographic evaluation of corticosteroid inhibition of allergic patch test reactions. Contact Derm 26:212–213, 1992. 37. Seidenari S, di Nardo A. B-scanning evaluation of allergic reactions with binary transformation and image analysis. Acta Derm Venereol (Stockh) 175 (Suppl.):3–7, 1992. 38. Seidenari S, di Nardo A. B-scanning evaluation of irritant reactions with binary transformation and image analysis. Acta Derm Venereol (Stockh) 175 (Suppl.):9–13, 1992. 39. di Nardo A, Seidenari S, Gianetti A. B-scanning evaluation with image analysis of psoriatic skin. Exp Dermatol, in press.
40. Seidenari S, di Nardo A, Gianetti A. Assessment of topical corticosteroid activity on experimentally induced contact dermatitis: echographic evaluation with binary transformation and image analysis. Skin Pharmacol 6:85–91, 1993. 41. Thijssen JM, Oosterveld BJ. Texture in tissue echograms. Speckle or informations. J Ultrasound M 9:215–229, 1990. 42. Søndergaard J, Serup J, Tikjøb G. Ultrasonic A- and Bscanning in clinical and experimental dermatology. Acta Derm Venereol (Stockh) 120 (Suppl.):76–82, 1985. 43. Querleux B, Lévêque J-L, de Rigal J. In vivo crosssectional ultrasonic imaging of human skin. Dermatologica 177:332–337, 1988. 44. Pugliese P. Use of ultrasound in evaluation of skin care products. Cosmet Toilet 104:61–76, 1989. 45. Hoffmann K, el-Gammal S, Altmeyer P. B-scan Sonographie in der Dermatologie. Hautarzt 41:7–16, 1990. 46. Payne PA. Skin thickness measurement and cross-sectional imaging. In Cutaneous Investigation in Health and Disease: Noninvasive Methods and Instrumentation, J-L Laaevêque, Ed. Marcel Dekker, New York, 1989, pp. 183–213. 47. Serup J. High-frequency ultrasound examination of aged skin: intrinsic, actinic and gravitational aging, including new concepts of stasis dermatitis and by ulcer. In Aging Skin, Properties and Functional Changes, J-L Lévêque, Ed. Marcel Dekker, New York, 1993, pp. 69–85. 48. el-Gammal S, Auer T, Hoffmann K, Matthes U, Hammentgen R, Altmeyer P, Ermert H. High-frequency ultrasound: a noninvasive method for use in dermatology. In Noninvasive Methods for the Quantification of Skin Functions, PJ Frosch and AM Kligman, Eds. SpringerVerlag, Berlin, 1993, pp. 104–129. 49. Agner T, Serup J. Ultrasound: an update on methodology and application with special references to inflammatory reactions. In Noninvasive Methods for the Quantification of Skin Functions, PJ Frosch and AM Kligman, Eds. Springer-Verlag, Berlin, 1993. 50. Gropper CA. Diagnostic high-resolution ultrasound in dermatology. Int J Dermatol 32:243–250, 1993. 51. Fornage B, Duvic M. High-frequency sonography of the skin. U Eur Acad Derm Venereol 3:47–55, 1994. 52. Southwood WFN. The thickness of the skin. Plast Reconstr Surg 15:423–429, 1955. 53. McConkey B, Fraser GM, Bligh AS, Whiteley H. Transparent skin and osteoporosis. Lancet i:693–695, 1963. 54. Schatz H, Staudemayer T, Kligman AM. Applications in the study of human skin disorders and the response to treatment. In Ultrasound in Dermatology, P Altmeyer, S el-Gammal, K Hoffmann, Eds. Springer-Verlag, Berlin, 1992, pp. 256–263. 55. Petersen H, Agner T, Storm T. Skin thickness in patients with osteoporosis and controls quantified by ultrasound A-scan. Skin Pharmacol, submitted. 56. Lévêque J-L, Porte G, de Rigal J, Corcuff P, Francois AM, Saint Leger D. Influence of chronic skin exposure on some biophysical parameters of the human skin: an in vivo study. J Cut Aging Cosmet Dermatol 1:123–27, 1988.
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
57. Gniadecka M, Serup J, Søndergaard J. Influence of gravitational stress on skin echogenicity in young vs. aged individuals. High-frequency ultrasonography and digital image analysis: novel techniques for evaluation of dermal water distribution. Br J Dermatol, submitted. 58. Gniadecki R, Gniadecka M, Kotowski T, Serup J. Alterations of skin microcirculatory rhythmic oscillations in different positions of the lower extremity. Acta Derm Venereol (Stockh) 72:256–260, 1992. 59. Gniadecka M, Gniadecki R, Serup J, Søndergaard J. Skin mechanical properties present adaptation to man’s upright position. In vivo studies in young and aged individuals. Acta Derm Venereol (Stockh), in press. 60. Serup J. Noninvasive techniques for quantification of contact dermatitis. In Textbook of Contact Dermatitis, RJG Rycroft, T Menné, PH Frosch, C Benezra, Eds. Springer-Verlag, Berlin, 1992, pp. 323–338. 61. Serup J. Characterization of contact dermatitis and atopy using bioengineering techniques. A survey. Acta Derm Venereol (Stockh) 177 (Suppl.):14–25, 1992. 62. Hoffmann K, Schwarzt M, Dirschka T, el-Gammal S, Hoffmann A, Altmeyer P. Non-invasive evaluation of inflammation in atopic dermatitis. J Eur Acad Derm Venereol, in press. 63. Serup J, Staber B, Klemp P. Quantification of cutaneous oedema in patch test reactions by measurement of skin with high-frequency pulsed ultrasound. Contact Derm 10:88–93, 1984. 64. Serup J, Staber B. Ultrasound for assessment of allergic and irritant patch test reactions. Contact Derm 17:80–84, 1987. 65. Brazier S, Shaw S. High-frequency ultrasound measurement of patch test reactions. Contact Derm 15:199–201, 1986. 66. Beck JS, Spence VA, Lowe JG, Gibbs JH. Measurement of skin swelling in the tuberculin test by ultrasonography. J Immunol Methods 86:125–130, 1986. 67. Eun HC, Marks R. Dose-response relationships for topically applied antigens. Br J Dermatol 122:491–499, 1990. 68. Berardesca E, Maibach HI. Bioengineering and the patch test. Contact Derm 18:3–9, 1988. 69. Agner T, Serup J. Quantification of the DMSO response: a test for assessment of sensitive skin. Clin Exp Dermatol 14:214–217, 1989. 70. Agner T, Serup J, Handlos V, Bastberg W. Different skin irritation abilities of different qualities of sodium lauryl sulphate. Contact Derm 21:184–188, 1989. 71. Agner T, Serup J. Skin reactions to irritants assessed by non-invasive bioengineering methods. Contact Derm 20:352–359, 1989. 72. Agner T, Serup J. Individual and instrumental variations in irritant patch-test reactions: clinical evaluation and quantification by bioengineering methods. Clin Exp Dermatol 15:29–33, 1990. 73. Agner T, Serup J. A dose-response study of SLS irritation evaluated by different bioengineering methods. J Invest Dermatol 95:543–547, 1990. 74. Agner T, Serup J. Seasonal variation of skin resistance to irritants. Br J Dermatol 121:323–328, 1989.
489
75. Serup J. Diameter, thickness, area, and volume of skinprick histamine weals. Allergy 39:359–364, 1984. 76. Shall L, Newcombe RG, Lush M, Marks R. Doseresponse relationship between objective measures of histamine-induced wheals and dose of terfenadine. Acta Derm Venereol (Stockh) 71:199–204, 1991. 77. di Nardo A, Seidenari S. Echographic evaluation with image analysis of histamine induced weals. Skin Pharmacol, in press. 78. Serup J. Non-invasive quantification of psoriasis plaques: measurement of skin thickness with 15 MHz pulsed ultrasound. Clin Exp Dermatol 9:502–508, 1984. 79. Hermann RC, Ellis CN, Fithing DW, Ho VC, Voorhees JJ. Measurement of epidermal thickness in normal skin and psoriasis with high-frequency ultrasound. Skin Pharmacol 1:128–136, 1988. 80. Broby-Johansen U, Karlsmark T, Petersen LJ, Serup J. Ranking of the anti-psoriatic effect of various topical corticosteroids applied under a hydrocolloid dressing, skin thickness, blood-flow and colour measurements compared to clinical assessments. Clin Exp Dermatol 15:343–348, 1990. 81. Rogers S. Skin thickness in psoriasis. Clin Exp Dermatol 17:324–327, 1992. 82. Krieg PHG, Bacharach-Buhlers M, el-Gammal S, Altmeyer P. The pustule in palmoplantar psoriasis: transformed preside or mature microabscess? A threedimensional study. Dermatology 185:104–112, 1992. 83. Serup J. Quantification of acrosclerosis: measurement of skin thickness and skin-phalanx distance in females with 15 MHz pulsed ultrasound. Acta Derm Venereol (Stockh) 64:35–40, 1984. 84. Serup J. Localized scleroderma (morphea): thickness of sclerotic plaques as measured by 15 MHz pulsed ultrasound. Acta Derm Venereol (Stockh) 64:214–219, 1984. 85. Serup J. Localized scleroderma (morphea). Clinical, physiological, biochemical and ultrastructural studies with particular reference to quantification of scleroderma. Acta Derm Venereol (Stockh) 122 (Suppl.):1–61, 1986. 86. Serup J. Decreased skin thickness of pigmented spots appearing in localized scleroderma (morphea). Measurement of skin thickness by 15 MHz pulsed ultrasound. Arch Dermatol Res 276:135–137, 1984. 87. Rodman GP, Lipinski E, Luksick J. Skin thickness and collagen contact in progressive systemic sclerosis and localized scleroderma. Arthritis Rheum 22:130–140, 1979. 88. Cole GW, Handler SJ, Burnett K. The ultrasonic evaluation of skin thickness in scleroderma. J Clin Ultrasound 9:501–503, 1981. 89. Åkesson A, Forsberg L, Hedeström E, Wollheim F. Ultrasound examination of skin thickness in patients with progressive systemic sclerosis (scleroderma). Acta Radiol Diagn 27:91–94, 1986. 90. Myers SL, Cohen JS, Sheets PW, Bies JR. B-mode ultrasound evaluation of skin thickness in progressive systemic sclerosis. J Rheumatol 13:577–580, 1986.
490
Handbook of Non-Invasive Methods and the Skin, Second Edition
91. Kalis B, de Rigal J, Leonard F, Lévêque J-L, Riche O, le Corre Y, de Lacharriere O. In vivo study of scleroderma by non-invasive techniques. Br J Dermatol 122:785–791, 1990. 92. Hoffmann K, Gerbaubt U, el-Gammal S, Altmeyer P. 20-MHz B-mode ultrasound in monitoring the course of localized scleroderma (morphea). Acta Derm Venereol (Stockh) 164 (Suppl.):3–16, 1991. 93. Levy JJ, Gassmüller J, Andring H, Brenke A, AlbrechtNebe H. Darstellund der subkutanen Atrophie der Zirkumskripten Skleroderme im 20 MHz-B-scan Ultraschall. Hautarzt 44:446–451, 1993. 94. Gomez EC, Berman B, Miller DL. Ultrasonic assessment of cutaneous atrophy caused by intradermal corticosteroids. J Dermatol Oncol Surg 8:1071–1072, 1982. 95. Serup J, Holm P, Stender IM, Pichard J. Skin atrophy and telangiectasia after topical corticosteroids as measured non-invasively with high frequency ultrasound, evaporimetry and laser Doppler flowmetry. Methodological aspects including evaluations of regional differences. Bioeng Skin 3:43–58, 1987. 96. Lubach D, Grütter H, Behl M, Nagel C. Investigations on the development and regression of corticosteroidinduced thinning of the skin in various parts of the human body during and after topical application of amcinomide. Dermatologica 178:93–97, 1989. 97. Korting HC, Vielluf D, Kerscher M. 0.25% prednicarbate cream and the corresponding vehicle induce less skin atrophy than 0.1% betamethasone-17-valerate cream and 0.05% clobetasol-17-propionate cream. Eur J Clin Pharmacol 42:159–161, 1992. 98. Kerscher MJ, Korting HC. Topical glucocorticoids of the non-fluorinated double-ester type. Lack of atrophygenicity in normal skin as assessed by high-frequency ultrasound. Acta Derm Venereol (Stockh) 72:214–216, 1992. 99. Lubach D, Rath J, Kietzmann M. Steroid-induced dermal thinning: discontinuous application of clobetasol17-propionate ointment. Dermatologica 185:44–48, 1992. 100. Katz SM, Frank DH, Leopold GR, Wachtel TL. Objective measurement of hypertrophic human scar: a preliminary study of tonometry and ultrasonography. Ann Plast Surg 14:121–127, 1983. 101. Cantrell JH, Goans RE, Roswell RL. Acoustic impedance variations in burn–nonburn interfaces in porcine skin. J Acoust Soc Am 64:731–736, 1978. 102. Goans RE, Cantrell JH, Meyers FB. Ultrasonic pulseecho determination of thermal injury in deep dermal burns. Med Physics 4:259–263, 1977. 103. Bauer JA, Sauer T. Cutaneous 10 MHz ultrasound Bscan allows the quantitative assessment of burn depth. Burns 15:49–51, 1989. 104. Goldberg B. Ultrasonic evaluation of superficial masses. J Clin Ultrasound 3:90–94, 1975. 105. Rukavina B, Mohar N. An approach of ultrasound diagnostic techniques of the skin and subcutaneous tissue. Dermatologica 158:81–92, 1979.
106. Miyauchi S, Tada M, Miki J. Echographic evaluation of modular burns of the skin. J Dermatol 10:221–227, 1983. 107. Breitbart EW, Rekpenning W. Möglickkeiten und Grenzen der ultraschalldiagnostik zur in vivo Bestimmung der Invasionstiefe des maglignen Melanoms. Z Hautkr 58:975–987, 1983. 108. Kraus W, Schramm P, Hoede N. First experiences with a high-resolution ultrasonic scanner in the diagnosis of malignant melanomas. Arch Dermatol Res 275:235–238, 1983. 109. Brenner S, Ophir J, Weinraub Z. Thickness of basal cell epithelioma measured preoperatively by ultrasonography. Cutis 34:509, 1984. 110. Shafir R, Itzchak Y, Heyman Z, Azizi E, Tsur H, Hiss J. Preoperative ultrasonic measurements of the thickness of cutaneous malignant melanoma. J Ultrasound Med 3:205–208, 1984. 111. Kraus W, Nake-Elias A, Schramm P. Diagnostsische Fortschrifte bei malignen Melanomen durch die hochauflösende Real-time-sonographie. Hautarzt 36:386–392, 1985. 112. Rodriques JCF, Carlos MJ, Féria R. Squamous-cell carcinoma. Echotomographic study and treatment by chemosurgery without systematised microscopic control. Skin Cancer 2:135–140, 1987. 113. de Ascensao AC. Echotomographic study of skin tumours. Skin Cancer 2:41–45, 1987. 114. de Ascensao AC. Ultrasonography of skin tumours: basic principles and main procedures. Skin Cancer 2:5–12, 1987. 115. Schwaijhofer B, Pohl-Markl, H, Frühwald R, Stiglbauer R, Kokoschka EM. Der Diagnostische Stellenwert des Ultraschalls beim malignen Melanom. Forschr Röntgenstr 146:409–411, 1987. 116. Hughes BR, Black D, Srivastava A, Dalzid K, Marks R. Comparison of techniques for the noninvasive assessment of skin tumours. Clin Exp Dermatol 12:108–111, 1987. 117. de Ascensao AC. The importance of echography in the treatment of basal-cell carcinoma. Skin Cancer 3:237–253, 1988. 118. Breitbart EW, Müller CE, Hicks R, Vieluf D. Neue Entwichlungen der Ultraschalldiagnostik in der Dermatologie. Aktuelle Dermatol 15:57–61, 1989. 119. Edwards C, Al-Aboozi MM, Marks R. The use of Ascan ultrasound in the assessment of small skin tumours. Br J Dermatol 121:297–304, 1989. 120. Reali UM, Santucci M, Paoli G, Chiarugi C. The use of high resolution ultrasound in preoperative evaluation of cutaneous malignant melanoma thickness. Tumori 75:452–455, 1989. 121. Hoffmann K, el-Gammal S, Matthes U, Altmeyer P. Digitale 20 MHz-Sonographie der Haut in der präoperativen Diagnostik. Z Hautkr 64:851–858, 1989. 122. Nessi R, Betti R, Bencini PL, Crosti C, Blanc M, Uslenghi C. Ultrasonography of nodular and infiltrative lesions of the skin and subcutaneous tissues. J Clin Ultrasound 18:103–109, 1990.
High-Frequency Ultrasound Examination of Skin: Introduction and Guide
123. Gassenmaier G, Kieseweller F, Schell H, Zinner M. Wertigkeit der Hochauflösenden Sonographie für die Bestimmung des vertikalen Tumordurch-messers beim malignen Melanom der Haut. Hautarzt 41:360–364, 1990. 124. Hoffmann K, Stücker M, el-Gammal S, Altmeyer P. Digitale 20-MHz-Sonographie des Basalioms im bscan. Hautarzt 41:333–339, 1990. 125. Betti R, Neosi R, Blanc M, Bencini PL, Galimberti M, Crosti C, Uslenghi C. Ultrasonography of proliferative vascular lesions of the skin. J Dermatol 17:247–251, 1990. 126. Neosi R, Blanc M, Bosco M, Dameno S, Venegoni A, Betti R, Bencini PL, Crosti C, Uslenghi C. Skin ultrasound in dermatologic surgical planning. J Dermatol Surg Oncol 17:38–43, 1991. 127. Beeckman P, de Clerck S, Jong B, de Maeseneer M. The ultrasound aspect of the skin and subcutaneous fat layer in various benign and malignant breast conditions. J Belge Radiol 74:283–288, 1991. 128. Hoffman K, Jung J, el-Gammal S, Altmeyer P. Malignant melanoma in 20 MHz B-scan sonography. Case Rep Dermatol 185:49–55, 1992. 129. Nitsche N, Iro H, Hoffmann K. Ultraschalldiagnostik der ausseren Nase. Hals Nase Ohrin HNO 40:181–185, 1992. 130. Harland CC, Bamber JC, Gusterson BA, Mortimer PS. High frequency, high resolution B-scan ultrasound in the assessment of skin tumours. Br J Dermatol 128:525–532, 1993. 131. Hoffmann K, Winkler K, el-Gammal S, Altmeyer P. A wound healing model with sonographic monitoring. Clin Exp Dermatol 18:217–225, 1993.
491
132. Maurad MM, Marks R. Assessment of disease severity and outcome in the gravitational syndrome using pulsed A-scan ultrasound to measure skin thickness. Clin Exp Dermatol 15:200–205, 1990. 133. Maurad MM, Edwards C, Marks R. Skin extensibility in the gravitational syndrome. Bioeng Skin 4:199–215, 1988. 134. Nielsen T, Iversen H, Tfelt-Hansen P. Dermascan A. A high-frequency acoustic scanner for non-invasive studies of the luminal diameter of superficially situated medium-sized arteries. Cephalalgia 10 (Suppl.):72–73, 1989. 135. Iversen H, Nielsen T, Tfelt-Hansen P, Olesen J. Headache and changes in the diameter of the radial artery during 7 hours intravenous nitroglycerin infusion. Cephalalgia 10 (Suppl.): 82–83, 1989. 136. Iversen HK. N-Acetylcysteine enhances nitroglycerineinduced headache and cranial arterial responses. Clin Pharmacol Ther 52:125–133, 1992. 137. Iversen KH, Nielsen TH, Olesen J, Tfelt-Hansen P. Arterial responses during migraine headache. Lancet 336:837–839, 1990. 138. Seidenari S, Zanella C, Pepe P. Echographic evaluation of sodium lauryl sulfate (SLS)-induced irritation in mice. Contact Derm 30:41–42, 1994. 139. Combettes C, Durand-Seme V, Querleux B, Saint-Leger D, Lévêque J-L. Imaging the hamster flank organ by an ultrasonic technique: a new approach to animal tests. Br J Dermatol 121:689–699, 1989. 140. Fullerton A. Personal observation. 141. Serup J. Personal observation. 142. Serup J, Keiding J, Fullerton A, Gniadecka M, Gniadecki R. Personal observation.
B-Mode Imaging and 57 Ultrasound In Vivo Structure Analysis Stefania Seidenari Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 57.1 Introduction............................................................................................................................................................493 57.2 Ultrasound Equipment ...........................................................................................................................................494 57.2.1 The Instrument...........................................................................................................................................494 57.2.2 Technical Aspects of Echographic Recording for Image Analysis ..........................................................495 57.2.2.1 Probe-to-Skin Distance ..............................................................................................................495 57.2.2.2 The Swept-Gain Curve...............................................................................................................495 57.2.2.3 Other Parameters ........................................................................................................................495 57.3 Ultrasound Tissue Characterization.......................................................................................................................495 57.3.1 The Software for Image Analysis .............................................................................................................495 57.3.2 Present Possibilities and Perspectives .......................................................................................................496 57.3.3 Evaluation of Normal Skin........................................................................................................................497 57.3.3.1 Echographic Characteristics of Normal Volar Forearm Skin ....................................................499 57.3.4 Evaluation of Allergic Patch Test Reactions.............................................................................................499 57.3.5 Evaluation of Subclinical Allergic Patch Test Reactions..........................................................................500 57.3.6 Evaluation of the Activity of Topical Corticosteroids ..............................................................................501 57.3.7 Evaluation of Irritant Reactions ................................................................................................................501 57.3.7.1 Assessment of Sodium Lauryl Sulfate-Induced Irritation.........................................................501 57.3.7.2 Evaluation of Subclinical Irritation............................................................................................502 57.3.7.3 Evaluation of Sodium Lauryl Sulfate-Induced Irritation in Hairless Mice ..............................502 57.3.7.4 Evaluation of Other Irritant Substances.....................................................................................502 57.3.7.5 Indications for the Use of Units for Patch Tests to be Evaluated by Echography...................502 57.3.8 Evaluation of Wheals.................................................................................................................................503 57.3.9 Evaluation of Psoriatic Skin......................................................................................................................503 57.4 Conclusion .............................................................................................................................................................503 References .......................................................................................................................................................................504
57.1 INTRODUCTION The recent introduction of high-frequency ultrasound in dermatology has opened new perspectives concerning skin physiology and diagnosis of dermatological diseases. The two-dimensional sonographic methods are called B-scan procedures. A two-dimensional ultrasound image consists of single lines (A-scan lines) whose amplitudes are represented by means of a gray or color scale, which are added to form a sectional image. Like other diagnostic procedures used in dermatology, based on bioengineering methods, ultrasound B-scanning is an effective means for the objective assessment of
healthy and diseased skin. The difference between ultrasound B-mode imaging and other noninvasive techniques, providing digital displays or numbers, lies in the possibility of obtaining a pictorial representation corresponding to a cross-sectional image of the skin. The problem of looking under the surface of the skin has so far been dealt with by microscopy, which is the standard reference for morphologists. However, in respect to invasive procedures like biopsying for histological examination, an in vivo noninvasive assessment enables the evaluation of a skin site without altering the connections with the surrounding tissue, and without changing tissue orientation, tension, and thickness.1 493
494
Handbook of Non-Invasive Methods and the Skin, Second Edition
A considerable amount of data can be immediately obtained by looking at an echographic image. This information is perceived by intuition and evaluated by the observer according to his or her experience. A second phase, consisting of the computerized elaboration and quantification of data, turns the echographic procedure into an objective diagnostic means. In fact, if an immediate evaluation of an image or a comparative judgment over a short time span is conceivable without the need for instrumental elaboration, then interpretation and comparison of pictorial data showing slight variations in space and time are only possible by means of an appropriate software for image analysis.2 This chapter will provide a simplified nonmathematical description of image processing methods, which can be employed for the study of echographic pictures taken from the skin. At present, little experience on analytical methods suitable for 20-MHz ultrasound images is available. However, studies employing different image analysis procedures and different software packages are under way in many centers, in order to identify the most suitable processing equipment and techniques for analyzing highresolution ultrasound images of the skin. For a technical description of the different image analytical techniques, please refer to References 3 through 5. Image processing methods were first employed to improve various types of picture distortion and to restore images for human interpretation. One of the first applications was in correcting digitized newspaper pictures sent by submarine cable between London and New York in the early 1920s. In addition to applications in the space program, digital image processing techniques are now used for solving a variety of problems relating to methods to enhance pictorial information for human interpretation, and also for machine perception and analysis. The latter case deals with procedures for extracting information, which is suitable for computer processing, to end up with an easier interpretation or a specific application.3 A major problem regarding image processing methods applied to medicine is that all of these procedures provide information, which, although quantitative in a relative sense, is highly instrument dependent, making data collected from groups working with different equipment incompatible and incomparable.
57.2 ULTRASOUND EQUIPMENT 57.2.1 THE INSTRUMENT All the procedures that will be exemplified have been achieved on images recorded by the same instrument (Dermascan C®, Cortex Technology, Hadsund, Denmark). Because of the specificity of the information, as previously mentioned, this instrument will be briefly described.
The instrument consists of a probe with a 20-MHz transducer fixed onto an articulated self-balanced arm, a central unit for the elaboration of the data, a visualization system, and a memorizing and data storing system. The scanner displays four images per second, sufficient for easy orientation in tissue and for swept-gain regulation on the live image. The ultrasound transducers used in all the scanning heads are standard Cortex 20 MHz with a bandwidth of 15 MHz, a focal distance of 30 mm, and a 6-dB focal length of approximately 13 mm. Using this transducer, the axial system resolution is 50 mm; the lateral system resolution, 350 mm; and the typical usable tissue penetration, 10 mm. The field of view in depth can be set in four steps going from a minimum of 1.7 mm full screen to a maximum of 13.4 mm full screen. All four fields of view can be panned live to a maximum scanning depth of 30 mm. Our evaluations have been performed by employing the zoom function in the axial direction at the first magnification (at factor 2), which enables exploration of the tissue within a depth interval of 6.71 mm. The images are presented either in a 256-shade pseudocolor scale or a gray scale. The scales are coded to the A-scan amplitude. The amplitudes displayed are proportional to the recorded amplitudes up to the value 200 within the 0 to 256 scale. Above this value the amplitudes are displayed compressed. The images are computed from 224 A-scan lines recorded along the transducer movement, each A-scan having 256 sampling points. The live images can be instantly frozen and can be saved in either the system memory or permanently in a personal computer (PC). All parameter settings are saved simultaneously. Any A-scan line can be displayed, and its position marked on top of a frozen image, for closer analysis and measurement of distances. Further calculations like area and mean echo amplitude within a region of interest as defined by the user can be made directly on the scanner. The instrument has a fully documented data interface available for connection to PCs for image analysis packages. All the pictures undergoing image analysis have been recorded by the counterbalanced probe capable of scanning within a 22.4 × 22.4 mm area, with a built-in closed water path, using a water-comparable ultrasound gel as contact medium. System calibration can be carried out using a standard phantom (Cortex Technology). This consists of a selected natural rubber compound that produces an entrance echo very similar to human skin and a 1-mm-thick spacing plate, which has an open slot for the ultrasound transmission contact medium (Cogel, Comedical, Trento, Italy). The plate thickness is used as the distance between the probe membrane and the phantom when calibrating the gain, which is set to give A-scan peaks up to the maximum limit of the screen. In our instrument this adjustment is equal to 22 dB, used in all our recordings. We have tested the amplitude variations relating to the probe-to-skin
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
distance in a water tank using a plane polyvinyl chloride (PVC) plate as the reflector. By measuring the entrance echo as a function of the depth, we found that the variations in amplitude can be neglected in the interval of 1.5 to 3.5 mm in front of the nominal zero of the instrument (F = 30 mm). Accordingly, during recordings, the distance between the zero point and the skin is kept at 1.7 ± 0.2 mm. This can be achieved by making the epidermis echo coincide with a point marked on the screen frame, corresponding to a 1.7-mm distance from the zero.
57.2.2 TECHNICAL ASPECTS OF ECHOGRAPHIC RECORDING FOR IMAGE ANALYSIS Some technical aspects should be considered when performing recordings of 20-MHz B-scan pictures for image analysis. In fact, in order to obtain reproducible data, images should be recorded under the same experimental conditions. 57.2.2.1 Probe-to-Skin Distance When recording echographic images for computerized analysis, one should consider that the received ultrasound signal is influenced by attenuation due to absorption or scattering and reflection. While reflection takes place mainly at the boundaries of tissue compartments with different impedance, absorption is high when the relative amount of proteins and collagen in the tissue increases, and low if the liquid content is elevated. One result of attenuation is that the signal coming from a given tissue structure varies according to not only acoustic tissue parameters, but also the distance between the structure and the transducer.6 For this reason, the probe-to-skin distance should be kept constant. This can be achieved as previously described or by using a punched spacing plate of constant thickness to be put between the skin and the probe. The quantity of gel between the probe membrane and the skin will be standardized as well, corresponding to the amount filling the plate’s slot. 57.2.2.2 The Swept-Gain Curve The time gain compensation is used to correct for the echo attenuation of the signals coming from greater depth and can be adjusted as required. Interactive regulation of the swept-gain curve can be useful for a direct observation of the skin site under examination in order to improve the contrast and provide further information, but it eliminates the possibility of performing image analysis. In fact, if the swept-gain curve is not constant, the echoes’ amplitudes are variably modified, not depending on the characteristics of the tissue and depth alone.
495
57.2.2.3 Other Parameters Other parameters should be considered and kept constant during recording. For example, acoustic properties of the skin rely also on elasticity, tension, pressure, and temperature.6 The orientation of the collagen fibers, which strongly influences the acoustic properties of the dermis, changes under different tension forces, as shown by scanning electron microscopy.7 This observation can explain the variations in the echogenicity of the skin under different conditions of tension during ultrasound examination, which can be determined by stretching or by increasing the pressure of the probe onto the skin.8 Moreover, an increase in the water temperature in the applicator seems to be a method of improving the amplitude of the received echo signal.5
57.3 ULTRASOUND TISSUE CHARACTERIZATION Ultrasound tissue characterization can be defined as the process of describing tissue structure in terms of information, deriving from computer processing of echographic images.2 Its major goal consists of extracting image features, which may not be available to the human observer, and transforming subjective feature analysis into a quantitative process, so as to obtain objective and comparable results, which are not dependent on the diagnostician. Many acoustic parameters, such as elasticity, density, attenuation coefficient, and speed of sound, represent suitable features for tissue characterization.2 However, only image processing techniques for the evaluation of the distribution of amplitude variations of echographic pictures of the skin are available so far. A simple approach to tissue differentiation is based on the evaluation of mean graylevel values (with standard deviation) within a region of interest. A more specific procedure consists of the use of image segmentation. Segmentation is the process that subdivides an image into areas of uniform appearance. Regions or objects of interest are subsequently extracted for further processing, such as description or recognition.9 When texture characteristics corresponding to different body regions or diseases have to be defined for the first time, the segmentation attributes should be found during processing, and regions of interest must be established interactively by the user, specifying the number and location of the thresholds by trial and error.
57.3.1 THE SOFTWARE
FOR IMAGE
ANALYSIS
Software for image analysis on Dermascan C images will be briefly described. The program ascribes an arbitrary scale, ranging from 0 to 255, to the amplitude values of the echoes. It enables the interactive selection of amplitude bands, whose width and positioning on a numerical scale
496
Handbook of Non-Invasive Methods and the Skin, Second Edition
can be chosen. The choice of the band of interest is a function of the problem being considered, which consists in highlighting the picture areas formed by pixels (picture elements) characterizing structures or functional aspects of skin reactions, which should be quantitatively evidenced and assessed. In order to identify a meaningful band of interest, the recorded image can be scanned systematically, using different amplitude intervals. This can be achieved manually, by selecting and highlighting a band, corresponding to an amplitude interval of interest, on the scale appearing in the upper part of the screen. At the same time, only those parts of the image whose pixels are reflecting within the selected amplitude interval will be evidenced. After defining a region of interest, it is possible to assess quantitatively the areas in which the echoes’ amplitude is included within the selected values, by calculating its extension in square millimeters or in number of pixels.10 In Figure 57.1 through Figure 57.4 different parts of the same echographic image are enhanced by using different amplitude bands. The epidermis and the lower part of the dermis are marked with amplitude bands covering the upper part of the scale; intermediate and low amplitude intervals characterize different parts of the dermis.
57.3.2 PRESENT POSSIBILITIES
AND
therapeutic effects of drugs used for the treatment of this disease. Studies on skin tumors and other inflammatory skin diseases are under way. Most skin tumors appear as hyporeflecting areas without any particular acoustic characteristics. Attempts have been made to differentiate skin tumors by quantifying their reflectivity. The main difficulty consists in obtaining quantitative evaluations,
PERSPECTIVES
Indications for the use of image analysis on 20-MHz Bscan recordings include studies on normal skin and skin aging; determination of the intensity of allergic and irritant patch test reactions and differentiation of irritant substances; studies of the effect of steroids on the skin; and, finally, evaluation of the course of psoriasis and of the
FIGURE 57.1 Echographic aspect of dorsal forearm skin in a woman aged 86. The image is represented by the pseudocolor scale.
FIGURE 57.2 Segmentation of the image in Figure 57.1, after selection of a 0- to 30-amplitude band. Hyporeflecting parts of the dermis are highlighted.
FIGURE 57.3 Segmentation of the image in Figure 57.1, after selection of a 50- to 150-amplitude band.
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
497
FIGURE 57.5 Cheek skin in a woman aged 25.
FIGURE 57.4 Segmentation of the image in Figure 57.1, after selection of a 201- to 255-band. The entrance echo and hyperreflecting parts of the lower dermis are visible.
regardless of actual attenuation, both in front and inside the tumor itself. Other skin diseases appearing with a characteristic echographic pattern of the dermis, such as scleroderma, cannot be differentiated from normal skin on the basis of image segmentation. More elaborated image descriptors will probably be needed to characterize this and other types of tissue alterations.
57.3.3 EVALUATION
OF
FIGURE 57.6 Cheek skin in a woman aged 81.
NORMAL SKIN
Variations of normal skin according to site, age, and sex have been studied by means of numerous noninvasive methods, including A-scanning, which enables the determination of skin thickness at different sites.11–13 Elaboration of B-scan images allows the quantification of aspects, which can also be appreciated by the simple observation of the images regarding the different distribution of the echogenicity according to site, age, and sex (Figure 57.5 through Figure 57.8).14 Table 57.1 and Table 57.2 illustrate the results of a study performed on 48 men and 48 women at six different skin sites.15 Four age groups (27 to 31, 60 to 70, 70 to 80, and 80 to 90) were studied. For the evaluation of the images, two bands with low (0 to 30) and intermediate (50 to 150) amplitude values were used (Figure 57.2 and Figure 57.3). Since skin thickness variations are high according to skin site, age, and sex, echogenicity values are not expressed in absolute numbers, but as the ratio
FIGURE 57.7 Normal volar forearm skin in a young subject.
between pixel values (extension of areas) and thickness values. Low numbers in Table 57.1 correspond to a limited extension of hyporeflecting (0 to 30 band) areas, and thus to higher reflectivity in respect to other skin sites: volar and dorsal forearm skin are most echogenic, and forehead
498
Handbook of Non-Invasive Methods and the Skin, Second Edition
Volar Forearm: 0 - 30 Band Evaluation 3000 2500 2000 1500 1000 500 Men 60–70
FIGURE 57.8 Normal volar forearm skin in an elderly subject. A hypoechogenic subepidermal band is clearly visible.
TABLE 57.1 Echogenicity of the Dermis According to Skin Site, Sex, and Age Volar Dorsal Upper Lower Forehead Cheek Forearm Forearm Abdomen Back YM YW EM EW
67 48 58 41
58 41 53 33
5 5 18 14
9 10 25 19
46 49 32 27
56 63 41 34
Note: A 0- to 30-band evaluation. YM = men aged 27–31; YW = women aged 27–31; EM = men aged 60–90; EW = women aged 60–90. Numbers are obtained by dividing the mean amplitude value by the skin thickness value and by 100.
TABLE 57.2 Echogenicity of the Dermis According to Skin Site, Sex, and Age
YM YW EM EW
Forehead
Cheek
Upper Abdomen
Lower Back
15 24 16 27
19 30 18 35
29 27 33 37
16 14 27 32
Note: A 50- to 150-band evaluation. YM = men aged 27–31; YW = women aged 27–31; EM = men aged 60–90; EW = women aged 60–90. Numbers are obtained by dividing the mean amplitude value by the skin thickness value and by 100.
Women 70–80
80–90
FIGURE 57.9 Values of the extension of the hyporeflecting area at volar forearm skin in three different age groups are expressed in number of pixels (picture elements).
skin shows the lowest reflectivity. Accordingly, intermediate amplitude values (50 to 150) are not so marked on the forehead (Table 57.2). Regarding variations related to sex, we can observe that at the forehead and cheek, echogenicity of the dermis is higher in women. No significant differences can be seen on abdomen and back skin, whereas for volar and dorsal forearm skin the extension of the hyporeflecting area corresponding to the hypoechogenic band in the upper part of the dermis is higher in men. Finally, concerning modifications of the dermis with the aging process, we can see that the reflectivity of the dermis increases at the forehead, cheek, abdomen, and back in elderly subjects. Volar and dorsal forearm skin show a decrease of the dermal echogenicity due to the appearance of the hyposonic subepidermal band, which in some subjects occupies most of the dermis (Figure 57.8). Figure 57.9 represents the values of the extension of the hyporeflecting (0 to 30) area at volar forearm skin in three different age groups in elderly subjects. Whereas in men the thickness of the hyposonic band seems constant, in elderly women it increases progressively with aging. Values referring to dorsal forearm skin show the same trend. To summarize, we can say that there is a great regional variation in the behavior of ultrasound reflection of the skin according to skin site, sex, and age. Regarding skin aging, it is not possible to identify a general tendency for the variations of the acoustic properties of elderly skin: at four skin sites examined, a shift from ultrasound echoes of low intensity, characteristic of the dermis of young subjects, to intermediate reflection amplitudes is visible in the elderly. On the contrary, forearm skin generates echoes of lower intensity.16 The variations of dermal reflectivity may be correlated to structural and biochemical diversities of collagen and elastic bundles, and also to water content and to composition of the ground substance, which may alter density, homogeneousness, and spatial
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
499
organization of structural elements that are responsible for the acoustic behavior and the reflectivity to ultrasound. These changes can only be appreciated by echography and quantitatively assessed by performing segmentation of the image, which represents an innovative means in the study of skin aging.
Elbow Crease
2 cm
1
6
2 2.5 cm
57.3.3.1 Echographic Characteristics of Normal Volar Forearm Skin
0.5 cm 0.5 cm
Forearm skin is the most suitable site for experimental allergic and irritant patch testing. Yet, great regional variations in the response to irritation, medication with corticosteroids, vasodilation, and reactivity to allergens have been described.10,17–19 In fact, by performing image analysis on 20-MHz echographic pictures taken at various sites of forearm skin, great differences can be appreciated in the distribution of the reflectivity of the dermis. To carry out patch tests in a reproducible way, and to overcome intra- and interindividual variations, the same skin sites should be assessed using accurate spacing references. To this aim we use flexible rubber foils with precise measures, having holes 2.5 cm in diameter corresponding to the testing areas (Figure 57.10). Before each evaluation, the upper edge of the rubber foil is applied to the elbow crease and test sites are drawn on the skin. Thus, each successive measurement can refer to its baseline value. Table 57.3 illustrates values of pixels reflecting within a 0 to 30 amplitude interval at the five different sites indicated in Figure 57.10. Significant differences can be established between areas 6 and 7 and the other sites, and between area 3 and areas 5 to 8, pointing to the necessity of referring observations on allergic and irritant reactions to precise baseline data. Moreover, experimental irritant patch testing on forearm skin should be performed in subjects under 45 years of age, without a subepidermal band of lower echogenicity. In fact, edema, deriving from an inflammatory reaction and appearing as a hypoechogenic band or area, can be difficult to recognize on the forearm of elderly subjects (Figure 57.7 and Figure 57.8).
3
7
4 2.5 cm
0.5 cm 0.5 cm 5 2.5 cm
FIGURE 57.10 Sites and measures of testing areas.
57.3.4 EVALUATION REACTIONS
OF
ALLERGIC PATCH TEST
In clinical practice allergic reactions can be evaluated by visual and palpatory examination. However, in some circumstances, such as the monitoring of drug-induced or spontaneous variations of reactivity to contact allergens, a quantitative assessment is required, enabling an objective collection of data and their statistical evaluation. The echographic assessment of patch test reactions was first performed by Serup and coworkers,20 who evaluated doubtful and positive patch test readings by A-scan measurements of skin thickness and found that this parameter, which can be considered an expression of allergic edema, increases according to the intensity of the reaction. In a B-scan image, a positive patch test reaction appears with an increased thickness of the skin, and with a greater homogeneity and a decreased echogenicity of
TABLE 57.3 Regional Differences of Volar Forearm Skin Area
1
2
3
4
5
6
7
Mean pixel values SD No. of cases
1368.27 851.8 56
1258.71 554.82 56
1466.43 780.6 53
1440.8 792.09 55
1208.69 581.86 58
989.29 551.28 80
1015.72 497.45 79
Note: A 0- to 30-band evaluation.
500
Handbook of Non-Invasive Methods and the Skin, Second Edition
3000 2500 2000 1500 1000 500 0 Baseline
24 h
0–30 0.5% 201–255 0.5%
FIGURE 57.11 Patch-test reactions of different degrees of positivity.
48 h 0 –30 201–255
72 h 5% 5%
FIGURE 57.13 Image processing of 20-MHz B-scan recordings of 0.5 and 5% nickel sulfate pet. patch-test reactions on volar forearm skin. Segmentation after selection of a 0–30 and a 201–255 interval. Numbers express mean values of pixels reflected within the chosen amplitude bands.
increase in the number of low reflecting pixels, and a decrease in the values assessing hyperreflectivity, according to the nickel patch test concentration and the elapsing of time, was demonstrated (Figure 57.13).21
57.3.5 EVALUATION OF SUBCLINICAL ALLERGIC PATCH TEST REACTIONS
FIGURE 57.12 Same images as in Figure 57.11, after selection of a 0–30 band. The hypoechogenic part of the dermis corresponding to edema and inflammatory infiltration is highlighted.
the dermis (Figure 57.11 and Figure 57.12). At the same time, we can observe a rise in the extension of hyposonic areas, a decrease in the quantity of objects reflecting within intermediate amplitude values, and the disappearance of the hyperreflecting parts of the dermis at volar forearm skin. No difference can be noticed among reactions induced by different contact allergens. The quantification of the inflammatory response can be achieved by processing the image by segmentation using a band covering the low echo-amplitude values of the scale. Intervals like 0 to 20, 0 to 30, 0 to 40, 5 to 20, and 3 to 40 provide results that can be usefully employed in the quantification of the intensity of the reactions, correlating well with the clinical observation. In order to quantify allergic reactions of different intensities, we performed patch tests on 12 nickel-sensitized women with nickel sulfate at different concentrations (0.5, 1, 2, 3, and 5%), with a 72-hour observation period, employing a 0 to 30 and a 201 to 255 band to assess the reflectivity of the dermis. A progressive
Image segmentation enables the quantification of allergic patch test reactions that are regarded clinically as negative or doubtful. This has been demonstrated by patch testing 70 nickel-sensitized patients with 0.05% nickel sulfate, and by recording the echographic images at 24 and 72 h.22 There were 22 doubtful 72-h reactions and 12 negative 24-h responses at test sites that showed positive reactions at 72 hours; they were processed by the image analysis software. A significant increase in the extension of the hyporeflecting areas in the dermis at clinically negative and doubtful readings was clearly shown (Figure 57.14). Thus, this method can be particularly useful when evaluating substances with low sensitizing potential, or allergens contained at low concentrations in complex preparations, or subjects with a high sensitization threshold. When performing echographic quantification of allergic reactions, one should consider the differences in the reactivity to allergens on forearm skin. By patch testing 17 nickel-sensitive women at four different sites on the volar forearm, a greater response to patch tests in skin areas near the wrist crease was demonstrated (area 4 in Figure 57.15).10 This is a clear indication for the use of rotating patterns or symmetrical sites in quantitative and comparative investigations concerning the evaluation of contact allergy on volar forearm skin.
501
2500
3000
2000
2500
1500
n. pixels
n. pixels
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
1000
2000 1500 1000
500 0
500
24 h∗
Baseline
E.C.
0 Baseline
72 h∗∗
Baseline NISO4
∗ 12
Negative reactions at 24 h, which developed as positive at 72 h, ∗∗ 22 doubtful allergic reactions at 72 h
FIGURE 57.14 Echographic evaluation of doubtful and negative allergic patch-test reactions. Image analysis by the 0 to 30 band.
1400 1200
NISO4 5% pet
Clobetasol propionate
Fluocinolone acetonide
Clobetasone butyrate
h
FIGURE 57.16 Evaluation of the anti-inflammatory effects of corticosteroids. Extension of the 0–30 areas of nickel sulfate test sites treated with different preparations.
the potency of the steroid used for topical application, as evaluated by segmentation of the echographic recording and determination of the extension of the hypoechogenic area of the dermis (Figure 57.16).
57.3.7 EVALUATION
1000
OF IRRITANT
REACTIONS
57.3.7.1 Assessment of Sodium Lauryl SulfateInduced Irritation
800 600 400 200 0 Baseline
24 Area 1
Area 2
Area 3
h Area 4
FIGURE 57.15 Extension of the 0–30 area at four sites on volar forearm skin, before and after a 24-h application of a 5% nickel sulfate patch test. Mean values at test sites near the wrist crease are higher.
57.3.6 EVALUATION OF THE ACTIVITY CORTICOSTEROIDS
OF
TOPICAL
The echographic quantification of allergic patch test reactions can be employed as an evaluation method for experimental procedures specifically concerning the antiinflammatory action of corticosteroids. In fact, epicutaneous testing, which reproduces eczematous reactions in a controlled manner in sensitive subjects, seems to be an ideal procedure for studying the inhibitory activity of these drugs.23 The procedure consists of patch testing sensitized subjects with nickel sulfate 5% at different skin sites on the volar forearm, and applying medications with corticosteroids of diverse potency 16 and 40 h after the induction of the allergic reactions. The inhibition of a patch test reaction after pharmaceutical treatment is proportional to
The inflammatory component of irritant patch test reactions can be assessed by image analysis on 20-MHz Bscan recordings using the same method as for evaluating allergic responses (Figure 57.17).24 At volar forearm patch test sites, sodium lauryl sulfate (SLS)-induced inflammation appears echographically with an edema, which is more superficial than the one caused by nickel sulfate in sensitive subjects. Frequently, a subepidermal hypoechogenic band represents the first sign of irritation at the dermal level. Subsequently, hypoechogenicity of the remaining part of the dermis appears, but even if the lower 400 300 n. pixels
n. pixels
64
200 100 0 Baseline
24 h
48 h
72 h
SLS conc. 0.5%
5%
FIGURE 57.17 Echographic evaluation of SLS-induced irritation. Assessment was by a 0–30 and 201–255 band.
502
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 57.18 Echographic aspect of SLS-induced irritant reactions of different intensity.
dermis is involved, hyperreflecting areas at this location do not disappear completely, as they do at allergic patch test sites (Figure 57.18). While at positive nickel patch test sites the entry echo shows only a slight attenuation even for very intense reactions, at SLS patch test sites a clear decrease in the echogenicity of the epidermis, which is hyporeflecting or interrupted or has totally disappeared, can be appreciated at 24-h evaluations (Figure 57.19). Going on the above, it is possible to distinguish between an allergic and an SLS-induced reaction. Whereas data calculated by the 0 to 30 evaluation show a fair correlation to clinical scoring, reflectivity of the epidermis, as assessed by the 201 to 255 band, is inversely proportional to transepidermal water loss (TEWL) values.24 Thus, at SLS patch test sites, loss of superficial hypoechogenicity can be considered a visual equivalent of impaired barrier function of the skin. 57.3.7.2 Evaluation of Subclinical Irritation Image analysis on 20-MHz B-scan recordings also represents a sensitive method for evaluating subclinical irritation. In fact, SLS-treated skin areas, clinically evaluated as negative tests, show echographic modifications that are significant in comparison to normal skin values.25
FIGURE 57.19 SLS-induced reactions are represented in the upper part of the picture; nickel sulfate responses are represented in the lower part. Segmentation after selection of a band highlighting the entrance echo.
corresponding to the epidermis enables the characterization of different irritant substances from an echographic point of view. In contrast to hypoechogenicity induced by SLS, at nonanoic acid and hydrochloric acid test sites a progressive increase in the epidermal reflectivity is visible, whereas propylene glycol and sodium hydroxide do not bring about any statistical variations in comparison to baseline skin (Figure 57.20).27 57.3.7.5 Indications for the Use of Units for Patch Tests to be Evaluated by Echography When performing patch tests with irritant substances, we must bear in mind that pressure caused by the adhesion of rigid chambers to the skin can induce an edema that is not always assessable by clinical evaluation, but is echographically evident. Since the presence of edema represents a positive element in the evaluation of irritancy, it is
57.3.7.3 Evaluation of Sodium Lauryl SulfateInduced Irritation in Hairless Mice SLS-induced irritation can also be echographically assessed in hairless mice. Segmentation can be performed by using the same amplitude intervals as in humans, whereas other amplitude intervals can be employed for the determination of the dermis–hypodermis boundary.26 57.3.7.4 Evaluation of Other Irritant Substances While the inflammatory component of irritant reactions has a fairly uniform expression, superficial reflectivity
FIGURE 57.20 Echographic images of skin reactions induced by different irritant substances: SLS in the upper left corner, sodium hydroxide in the upper right, nonanoic acid in the lower left, and hydrochloric acid in the lower right. A band highlighting the entrance echo has been chosen.
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
FIGURE 57.21 Echographic aspect of histamine-induced wheals of different size.
advisable to perform patch tests with irritant substances by using units that do not exert too much pressure on the skin.28 OF
WHEALS
Besides measurement of superficial wheal extension by planimetry, and assessment of skin thickness by A-scanning, determination of the dermal hypoechogenic area by image analysis can be employed in the evaluation of skin responses to wheal-inducing substances (Figure 57.21). By measuring histamine-induced wheals, we were able to demonstrate that edema of the dermis, corresponding to the extension of the hyposonic area, increases in time according to the histamine concentration. However, for the same concentration, a more intense response with a greater wheal extension and vaster dermal hypoechogenic area was observed at proximal sites, with respect to distal ones.29
57.3.9 EVALUATION
OF
FIGURE 57.22 Echographic aspect of a psoriatic plaque: before treatment, in the upper left corner, after 7 d of treatment with dithranol in the upper right corner, and after 14 d of treatment in the lower left corner. Normal contralateral skin in the lower right corner. Images are represented after selection of a 0–30 band.
PSORIATIC SKIN
The efficacy of antipsoriatic drugs is generally evaluated clinically. However, even if well-standardized criteria are used, reproducible data are difficult to obtain, and a comparison between studies performed in different centers is seldom possible. Therefore, different noninvasive methods have been proposed, with a view to standardizing parameters of potential relevance in following the evolution of the disease.30 Echographic evaluation enables the identification of three aspects that characterize psoriatic lesions: an increase in skin thickness, the presence of a thick hyposonic subepidermal band, and an enhanced entrance echo (Figure 57.22). Since echoes coming from tissues are determined by differences in acoustic impedance, we can presume that the hyposonic band is caused by the presence of a homogeneous component, which in psoriatic lesions could be represented by inflammatory infiltrate and vasodilatation. The enhancement of the entrance echo, on the other hand, corresponds to a thick-
6 4262.90
5
Thousands
57.3.8 EVALUATION
503
4 3 2 1215.80 1
351.30
0 Normal skin
Psoriatic skin
Ps. skin after 14 d
FIGURE 57.23 Echographic assessment of psoriatic skin. Evaluation by a 0–30 band. Histograms express values of hyporeflecting areas at normal contralateral skin, psoriatic untreated skin, and psoriatic skin after 14 d of treatment.
ening and a modification of psoriatic epidermis. Figure 57.23 and Figure 57.24 show the results of processing 20MHz B-scan images performed on 10 psoriatic plaques before and after 14 d of treatment with anthralin. Both the extension of the hyposonic subepidermal band and that of the hyporeflecting dermal echo gradually and constantly decrease according to clinical observations.31 Thus, processing of 20-MHz B-scan images allows the characterization and quantification of echographic parameters, which vary during the course of the disease according to treatment.
57.4 CONCLUSION Superficiality of the skin, enabling high-resolution ultrasound, represents a unique opportunity for analyzing
504
Handbook of Non-Invasive Methods and the Skin, Second Edition
1400 1191.50
n. pixels
1200
943.10
1000 738.80
800
600 Normal skin
Psoriatic skin
Ps. skin after 14 d.
FIGURE 57.24 Echographic assessment of psoriatic skin. Evaluation by a 201–255 band. Histograms express values of hyperreflecting areas at normal contralateral skin, psoriatic untreated skin, and psoriatic skin after 14 d of treatment.
echographic images. In fact, changes in skin thickness are very slight and recording conditions can be standardized geometrically by fixing probe-to-skin distance, whereby the intensity variations caused by the focusing of the transducer will remain very small. New developments on high-resolution ultrasound image analysis methods are under way. Segmentation can represent the first step for further computer processing and description of images. In fact, echographic images of some skin conditions do not present any quantitative variation in the echoes’ amplitudes; the total area covered by pixels reflecting within a certain amplitude range are the same. However, the spatial distribution of the amplitudes differs consistently, characterizing what we call a different texture of the dermis. The aim of future computer processing is to find a way of using descriptors that highlight essential differences between objects or classes of objects, regardless of changes in factors such as location, size, and orientation: boundary descriptors such as contour length, diameter, and curvature, or regional descriptors such as area, perimeter, or compactness, could be used in the future for describing ultrasound images of the skin.
REFERENCES 1. Serup J., Ten years’ experience with high frequency ultrasound examination of the skin: development and refinement of technique and equipment, in Ultrasound in Dermatology, Altmeyer P., el Gammal S., and Hoffmann K., Eds., Springer Verlag, Berlin, 1992, p. 41. 2. Bamber J.C. and Tristam M., Diagnostic ultrasound, in The Physics of Medical Imaging, Webb S., Ed., Adam Hilger, Bristol, 1990, chap. 7. 3. Gonzalez R.C. and Wintz P., Digital Image Processing, Addison-Welsey, New York, 1987.
4. Webb S., The Physics of Medical Imaging, Adam Hilger, Bristol, 1990. 5. el Gammal S., Hoffman K., Hoss A., Hammentgen R., Altmeyer P., and Ermert H., New concepts and developments in high resolution ultrasound, in Ultrasound in Dermatology, Altmeyer P., el Gammal S., and Hoffmann K., Eds., Springer Verlag, Berlin, 1992, p. 397. 6. McDicken W.N., Ultrasound in Tissue, in Diagnostic Ultrasonics: Principles and Use of Instruments, McDicken W.N., Ed., Churchill Livingstone, Hong Kong, 1991, chap. 4. 7. Brown I.A., A-scanning electron microscopic study of the effects of uniaxial tension on human skin, Br J Dermatol, 89, 383, 1973. 8. el Gammal S., Experimental approaches and new developments in high frequency ultrasound in dermatology, Zentralbl Haut Geschlechtskr, 157, 327, 1990. 9. Gonzalez R.C. and Wintz P., Image segmentation, in Digital Image Processing, Gonzalez R.C. and Wintz P., Eds., Addison-Welsey, New York, 1987. 10. Seidenari S. and Di Nardo A., Cutaneous reactivity to allergens increases from the antecubital fossa to the wrist: an echographic evaluation by means of a new image analysis system, Contact Derm, 26, 171, 1992. 11. Berardesca E., Farinelli N., Rabbiosi G., and Maibach H.I., Skin bioengineering in the non-invasive assessment of cutaneous aging, Dermatologica, 182, 1, 1991. 12. Alexander H. and Miller D.I., Determining skin thickness with pulsed ultrasound, J Invest Dermatol, 72, 17, 1979. 13. Sondergaard J., Serup J., and Tikijob G., Ultrasonic A and B scanning in clinical and experimental dermatology, Acta Derm Venereol (Stockh), 120 (Suppl.), 76, 1985. 14. De Rigal J., Escoffier C., Querleux B., Faivre B., Agache P., and Léveque J.L., Assessment of skin aging of human skin by in vivo ultrasonic imaging, J Invest Dermatol, 93, 622, 1989. 15. Seidenari S., Pagnoni A., Lasagni C., and Giannetti A., Age-dependent variations of normal skin: image processing of 20 MHz B scan recordings, Skin Pharmacol, 5, 230, 1992. 16. Seidenari S., Pagnoni A., Di Nardo A., and Giannetti A., Echographic evaluation with image analysis of normal skin: variations according to age and sex, Skin Pharmacol, 7, 201, 1994. 17. Van der Valk P.G.M. and Maibach H.I., Potential for irritation increases from the wrist to the cubital fossa, Br J Dermatol, 121, 709, 1989. 18. Kirsch J., Gipson J.R., Darley C.R., Barth J., and Burke C.A., Forearm skin variation with the corticosteroid vasoconstrictor assay, Br J Dermatol, 106, 495, 1982. 19. Tur E., Maibach H.I., and Guy R.H., Spatial variability of vasodilatation in human forearm skin, Br J Dermatol, 113, 197, 1985. 20. Serup J., Staberg B., and Klemp P., Quantification of cutaneous oedema in patch test reactions by measurements of skin thickness with high frequency pulsed ultrasound, Contact Derm, 10, 88, 1984.
Ultrasound B-Mode Imaging and In Vivo Structure Analysis
21. Seidenari S. and Di Nardo A., B scanning evaluation of allergic reactions with binary transformation and image analysis, Acta Derm Venereol (Stockh), 175 (Suppl.), 3, 1992. 22. Seidenari S., Echographic evaluation of subclinical allergic patch test reactions, Contact Derm, 31, 146, 1994. 23. Seidenari S., Di Nardo A., and Giannetti G., Assessment of topical corticosteroid activity on experimentally induced contact dermatitis: echographic evaluation with binary transformation and image analysis, Skin Pharmacol, 6, 85, 1993. 24. Seidenari S. and Di Nardo A., B scanning evaluation of irritant reactions with binary transformation and image analysis, Acta Derm Venereol (Stockh), 175 (Suppl.), 9, 1992. 25. Seidenari S. and Belletti B., Instrumental evaluation of subclinical irritation induced by sodium lauryl sulfate, Contact Derm, 30, 175, 1994.
505
26. Seidenari S., Zanella C., and Pepe P., Echographic evaluation of sodium lauryl sulfate induced reactions in mice, Contact Derm, 30, 41, 1994. 27. Seidenari S., Di Nardo A., Schiavi E., and Pepe P., Echographic evaluation of irritant reactions induced by nonanoic acid, hydrochloric acid and sodium lauryl sulfate: a comparison with TEWL assessment, Skin Pharmacol, 5, 240, 1992. 28. Seidenari S., Turnaturi C., Motolese A., and Pepe P., Echographic evaluation of edema induced by patch test chambers, Contact Derm, 27, 331, 1992. 29. Di Nardo A. and Seidenari S., Echographic evaluation with image analysis of histamine induced wheals, Skin Pharmacol, 7, 285, 1994. 30. Berardesca E. and Maibach H.I., Non-invasive bioengineering assessment of psoriasis, Int J Dermatol, 28, 157, 1983. 31. Di Nardo A., Seidenari S., and Giannetti A., B-scanning evaluation with image analysis of psoriatic skin, Exp Dermatol, 1, 121, 1992.
Assessment of Dermal 58 Ultrasound Water and Edema In Vivo Monika Gniadecka Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 58.1 Measurement of Physiological Fluid Movement in the Skin ...............................................................................507 58.2 Measurement of Dermal Edema............................................................................................................................508 References .......................................................................................................................................................................509
Since dermatological ultrasonography was initiated in the late 1970s by Alexander and Miller,1 the method has been used for evaluation of various skin diseases. The technique is based on the difference of acoustic properties of the investigated tissue, which in turn is reflected as echogenicity scale on the ultrasound image. Increase of fluid in the dermis leads to disarrangement of collagen fibers and decrease of echogenicity. Low echogenic regions in the dermis are frequently found in water abundant tissue where inflammation and edema occur, e.g., allergic and irritant reactions,2–4 and skin tumors that are well vascularized.5,6 Scars, despite the low content of water, also show low echogenic areas because of the tightly packed collagen fibers uniformly reflecting ultrasound waves.7 High echogenic areas dominate, e.g., in psoriasis,8 when hyperkeratotic epidermis strongly reflects ultrasound, or in scleroderma,9,10 where proliferating collagen fibers strongly reflect ultrasound. Low echogenicity of the areas enriched in fluid, e.g., the cysts, has been exploited for diagnostic purposes.
Several papers employed measurement of skin echogenicity for the quantification of dermal water content and distribution. Cross-sectional images of the skin were obtained by high-frequency ultrasound, as detailed elsewhere in this textbook. Echogenicity in the regions of interest (e.g., dermal region) was calculated by dedicated image analysis pograms. Several lines of evidence firmly indicate that short-term (minutes to hours) variability in dermal echogenicity is caused by the alterations in skin water content. The most compelling data were provided by nuclear magnetic resonance spectroscopy studies showing an inverse correlation between water-specific 1H spectral peaks and dermal echogenicity11 (Figure 58.1).
58.1 MEASUREMENT OF PHYSIOLOGICAL FLUID MOVEMENT IN THE SKIN High-frequency ultrasonography is so sensitive that small changes of water influx during body position can be quantified. It appears that skin echostructure changes diurnally
E D
D
FIGURE 58.1 Nuclear magnetic resonance image of the skin in the forearm. Edema formation after histamin prick is marked with arrows. E, epidermis; D, dermis. 507
508
Handbook of Non-Invasive Methods and the Skin, Second Edition
1
2
3
a
D
b
FIGURE 58.2 Twenty-megahertz ultrasound images of the skin of the ankle in a young healthy individual (a) and an elderly healthy individual (b). 1: In the horizontal postion in the morning. 2: Two hours after person has stood up. 3: Twelve hours later. Black (low echogenic) areas reflect water-rich regions; high echogenic areas show structures strongly reflecting ultrasound, e.g., collagen fiberrich areas. In young individuals high echogenic areas increase, which reflects water removal (a).
depending on fluid movement. In healthy aged individuals there is a tendency to water accumulation in the depended parts of the body, while in young individuals the echogenicity increases, suggesting fluid depletion (Figure 58.2).12 Individuals remaining in the horizontal position do not show statisticaly significant echogenicity changes. Major redistribution of body fluids occurs during postural manipulations,13,16,17 and the skin that serves as a huge water reservoir exchanges water rapidly with the extracellular fluid volume.14,15 The dermis, rather than subcutis, most likely constitutes a rapidly exchangable fluid reservoir.16 Changes of skin water occurring during the day were further correlated with changes in skin thickness and elasticity.16 These observations suggested that elastic properties of the skin are linked to fluid content. All of the abovementioned studies demonstrated that high-frequency ultrasonography enables noninvasive real-time in vivo monitoring of water balance in the skin. The method may find application in clinical physiology for monitoring fluid intake during therapy and anesthesia.
58.2 MEASUREMENT OF DERMAL EDEMA Edema is an increase in the amount of water in the tissue. It is a common finding in the physiological conditions, i.e., premenstrual syndrome or during a prolonged supine position.18 Changes in skin water distribution were shown in patients treated with glucocorticoids.19 Pathologic edema occurs in a variety of different conditions, such as kidney disease, cardiac insufficiency, lymphodema, and venous hypertension in the postthrombotic syndrome. The
latter condition is of a special dermatologic interest because of the risk of the development of skin ulceration.20 Numerous hypotheses have been proposed to explain the link between edema and the tendency to ulceration following deep venous thrombosis. Interestingly, different types of skin edema (lymphedema, cardiac insufficiencyand venous insufficiency-related edemas) have different localizations in the skin (Figure 58.3).21 In the postthrombotic syndrome edema is found in the upper dermis. In cardiac-related edema skin water is mainly localized in the lower dermis, toward the subcutis. In lymphedema edema is uniformly distributed all over dermis. It is known that skin metabolism is most active in the papillar dermis,22 and therefore the location of edema in this region could contribute to the formation of ulceration in patients with venous leg ulcers. The dispersement character of edema localization in the dermis in lymphedema patients was confirmed by nuclear magnetic resonoance studies.23 The ultrasound method has been succesfully used to evaluate and quantify the effects of compression treatment for removal of leg edema. The effects of different compression grades exerted by compression stockings and compression bandages were studied (Figure 58.4).24–27 Grade I and II compression stockings seemed to remove edema equally effectively from the dermis. This study suggests that the patients with mixed venous and arterial ulcerations could benefit from treatment with grade I compression.25 In the same investigation it was shown that edema recurs quickly after secession of the compressive therapy. Therefore, it is crucial that patients with leg edema apply compression daily. The differences in the
Ultrasound Assessment of Dermal Water and Edema In Vivo
a
509
b
FIGURE 58.3 Twenty-megahertz ultrasound images of the skin of the ankle region in patients with lipodermatosclerosis, e.g., ulcus cruris (a) and lymphedema (b), showing different ditributions of dermal edema.
efficacy of edema removal between different types of compressive bandages were visualized.26 Leg elevation for just 3 to 4 hours resulted in ca. a 30 to 40% reduction of skin edema.28 Also in lymphedema, edema has been assessed by high-frequency ultrasonography and correlated with the clinical examination in patients with breast cancer.29 Further studies on the efficacy of compresssion for both lymphedema and edema associated with leg ulceration could help to establish more efficient compressive therapy in the future.
REFERENCES 1. Alexander, H., Miller, D. Determining skin thickness with pulsed ultrasound. J. Invest. Dermatol., 17, 72, 1979. 2. Agner, T. Non-invasive mesuring methods for the investigation of irritant patch test reactions. Acta Derm. Venereol., 173, 1, 1992. 3. Seidenari, S., DiNardo, A. Echographic evaluation with image analysis of allergic and irritant reactions. Acta Derm. Venereol., 175, 3, 1992.
4. Serup, J. Characterisation of contact dermatitis and atopy using bioengineering techniques. A survey. Acta Derm. Venereol., 177, 14, 1992. 5. Harland, C.C., Bamber, J.C., Gusterson, B.A., Mortimer, P.S. High-frequency, high-resolution B-scan ultrasound in the assessment of skin tumours. Br. J. Dermatol., 128, 525, 1993. 6. Pinto, F., Lencioni, R., Magliaro, A., Nardini, V., Armilotta, N., Bartolozzi, C. High-frequency US preoperative assessment in cutaneous malignant melanoma. Radiology, 205, 421, 1997. 7. Gniadecka, M., Danielsen, L. High-frequency ultrasound for torture inflicted lesions. Acta Derm. Venereol., 75, 375, 1995. 8. Serup, J. Non-invasive quantification of psoriasis plaques measurement of skin thickness with 15 MHz pulsed ultrasound. Clin. Exp. Dermatol., 9, 502, 1984. 9. Serup, J. Localized scleroderma: thickness of sclerotic plaques as measured by 15 MHz pulsed ultrasound. Acta Derm. Venereol., 64, 214, 1984. 10. Hofmann, K., Rochling, A, Stucker, M., el-Gammal, S., Hoffmann, A., Altmeyer, P. High-frequency sonography of skin diseases, in Handbook of Non-Invasive Methods and the Skin, Serup, J., Jemec, G.B.E., Eds. CRC Press, Boca Raton, FL, 1995, p. 269.
510
Handbook of Non-Invasive Methods and the Skin, Second Edition
a
b
FIGURE 58.4 Twenty-megahertz ultrasound image of the ankle skin in a patient with lipodermatosclerosis. a: Before treatment with class II compressive stockings. b: After 12-hour treatment. 11. Gniadecka, M., Quistorff, B. Assessment of dermal water by high-frequency ultrasound: comparative studies with nuclear magnetic resonance. Br. J. Dermatol., 135, 218, 1996. 12. Gniadecka, M., Serup, J., Sondergaard, J. Age-related diurnal changes of dermal oedema: evaluation by highfrequency ultrasound. Br. J. Dermatol., 131, 849, 1994. 13. Maw, G.J., Mackenzie, I.L., Taylor, N.A. Redistribution of body fluids during postural manipulations. Acta Phys. Scand., 155, 157 1995. 14. Guyton, A.C., Hall, J.E. The body fliud compartments: extracellular and intracellular fluids and oedema, in Textbook of Medical Physiology, 9th ed., Guyton, A.C., Hall, J.E., Eds. W.B. Saunders, Philadelphia, 1996, p. 297. 15. Bert, J.L., Reed, R.K. Pressure-volume relationship for rat dermis: comparison studies. Acta Physiol. Scand., 160, 89, 1997. 16. Tsukahara, K., Takema, Y., Moriwaki, S., Fugimura, T., Imokawa, G. Dermal fluid translocation is an important determinant of the diurnal variation in human thickness. Br. J. Dermatol., 145, 590, 2001.
17. Eisenbeiss, C., Welzel, J., Eichler, W., Klotz, K. Influence of body water distribution on skin thickness: measurements using high-frequency ultrasonography. Br. J. Dermatol., 144, 947, 2001. 18. Eisenbeiss, C., Welzel, J., Schmell, W. The influence of female sex hormones on skin thickness evaluation using 20 Mhz sonography. Br. J. Dermatol., 139, 462, 1998. 19. Heitmann, B.L., Anhoj, J., Bisgaard, A.M., Ward, L., Bisgaard, H. Changes in body water distribution during treatment with inhalated steroid in pre-school children. Ann. Hum. Biol., 31, 333, 2004. 20. Ryan, T.J. The Management of Leg Ulcers. Oxford Medical Publications, Oxford, 1987. 21. Gniadecka, M. Localization of dermal edema in lipodermatosclerosis, lymphedema, and cardiac insufficiency. High-frequency ultrasound examination of intradermal echogenicity. J. Am. Acad. Dermatol., 35, 37, 1996. 22. Zemtsov, A., Thian, C.N.G., Min, X.U.E. Human in vivo 31P spectroscopy of skin: potentially a powerful tool for non-invasive study of metabolism in cutaneous tissue. J. Dermatol. Surg. Oncol., 15, 1207, 1989. 23. Idy-Peretti, I., Bittoun, J., Alliot, F.A., Richard, S.B., Querleux, B.G., Cluzan, R.V. Lymphedematous skin and subcutis: in vivo high resolution magnetic resonance imaging evaluation. J. Invest. Dermatol., 110, 782, 1998. 24. Gniadecka, M. Dermal oedema in lipodermatosclerosis: distribution, effects of posture and compressive therapy evaluated by high-frequency ultrasonography. Acta Derm. Venereol., 75, 120, 1995. 25. Gniadecka, M., Karlsmark, T., Bertram, A. Removal of dermal edema with class I and II compression stockings in patients with lipodermatosclerosis. J. Am. Acad. Dermatol., 39, 966, 1998. 26. Gniadecka, M., Danielsen, L., Henriksen, L. Non-invasive monitoring of compression therapy: a report of three cases with venous insufficiency. Acta Derm. Venereol., 82, 460, 2002. 27. Hu, D., Phan, T.T., Cherry, G.W., Ryan, T.J. Dermal oedema assessed by high frequency ultrasound in venous leg ulcers. Br. J. Dermatol., 138, 815, 1998. 28. Xia, Z.D., Hu, D., Wilson, J.M., Cherry, G.W., Ryan, T.J. How echographic image analysis of venous oedema reveals the benefits of leg elevation. J. Wound Care, 13, 125, 2004. 29. Balzarini, A., Milella, M., Civelli, E., Sigari, C., De Conno, F. Ultrasonography of arm edema after axillary dissection for breast cancer: a preliminary study. Lymphology, 34, 152, 2001.
59 Ultrasound Assessment of Skin Aging Giovanni Pellacani, Francesca Giusti, and Stefania Seidenari Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 59.1 Introduction............................................................................................................................................................511 59.2 Methods..................................................................................................................................................................511 59.3 Thickness and Echogenicity Variations of the Skin with Age .............................................................................512 59.4 Conclusions............................................................................................................................................................513 References .......................................................................................................................................................................513
59.1 INTRODUCTION Skin aging is an uneven process characterized by epidermal and dermal disorders, accompanied by many clinical signs, such as skin dryness, color changes, loss of elasticity, and wrinkles. Whereas age-induced modification (chrono-aging) of the skin is an intrinsic process modulated by genetic, behavioral, catabolic, endocrine, and gravitational factors, chronic exposure to sunlight induces numerous relevant alterations in the various compartments of the skin (photoaging), introducing great interindividual and site-to-site variations in the aging process.1 The effects of the UV radiations on sun-exposed sites are superimposed to the morphological, biochemical, and functional changes occurring with aging, making a distinction between the two phenomena hard. With aging, both qualitative and quantitative skin changes occur.2,3 Skin wrinkling and an increase in skin surface furrows appear especially in sun-exposed sites. The epidermis becomes thinner and flatter, with modifications both in the size and shape of keratinocytes and in the rate of proliferation. During senescence, the total amount of dermal collagen decreases, leading to skin thinning. Age-related variations of collagen organization are observable, such as collagen bundles disrupted by tangled fibrils or collagen densely packed in some areas with larger intervening spaces of ground substance. Moreover, the organization of the elastic fiber network becomes compromised, leading to changes in mechanical properties of the skin.3 In actinically damaged skin, collagen bundles of the upper dermis are progressively replaced by a more homogeneously stained material that appears to take up elastic-specific stains, giving rise to the so-called elastosis.4
By means of noninvasive methods, variations in hydration, transepidermal water loss (TEWL), surface lipids, and mechanical properties of the skin in elderly subjects have been evidenced.5–10 The development of ultrasound enabled the noninvasive evaluation of the cutaneous structure and the quantification of age-dependent modifications,11 together with the effects of diurnal and hormonal changes.12,13
59.2 METHODS High-frequency ultrasound enables real-time observation of the internal structures. Since penetration depth of the ultrasound waves is inversely related to its frequency, an optimal frequency range for dermatological use is between 20 and 150 MHz. Twenty-megahertz scanners are employed when alterations of the dermis are investigated, such as during skin aging.11 B-scanning sonography enabled the production of bidimensional images representing a cross section of the skin, suitable for studying the tissues close to the body surface. When a B-mode ultrasound image of normal skin is produced, a hyporeflective band-like echo is observable between the medium and the skin, the so-called entry echo, probably generated by the change in impedance from the coupling medium and the stratum corneum. Immediately below the entry echo, the corium, rich in collagenous fibers that are the main source of its echogenicity, is to be found. The development of dedicated software based on the attribution of fictional values to the echoes’ amplitudes (i.e., Dermavision 2D, Cortex Technology, Hadsund, Denmark) permitted an objective evaluation of echographic images of the skin, improving recognition of features corresponding to tissue structures or evolutive phases of processes to be studied, and enabling the quantification of data deriving 511
512
Handbook of Non-Invasive Methods and the Skin, Second Edition
from the image and their expression as numbers, which can be used for statistical evaluation.14,15
59.3 THICKNESS AND ECHOGENICITY VARIATIONS OF THE SKIN WITH AGE Skin thickness was a widely used parameter to evaluate the influence of different factors on skin aging. Employing A-scan ultrasound, Tan et al.16 found that skin thickness on the volar forearm increased progressively up to the age of 20, and then decreased subsequently. By means of the B-scan method, skin thickness values were approximately 15% greater than A-scan measurements.17 With A-scan, the determination of the dermis–subcutis interface is based on the observation of a peak corresponding to the impedance jump between adjacent parts of the tissue, making determination of the dermis–subcutaneous tissue interface difficult, whereas B-scan measurement of skin thickness represents the mean of consecutive A-scan lines composing the whole bidimensional image. Employing the B-scan method, de Rigal et al.17 observed that skin thickness on the volar and on the dorsal aspect of the forearm increases up to 15 years of age (maturation phase), remains constant until the seventh decade of life, and diminishes thereafter (athrophy phase). The modifications in skin thickness and reflectivity appear correlated with the age-dependent degree of elastosis. In particular, the identification of a subepidermal hypoechogenic band, appearing as a relatively homogeneous, echo-lucent structure, located immediately below the entry echo, was described in the skin of the elderly on the volar and dorsal aspect of the forearm (Figure 59.1).17 Since the main source of dermal echogenicity is represented by well-arranged collagen bundles, the appearance of the subepidermal hypoechogenic band was correlated to the structural changes that occur with age. In chronologically damaged skin, collagen bundles are replaced by a more homogeneously stained material, leading to the dissolution of the regular architecture of the collagen fibers and to the deposit of a greater amount of hydrated proteoglycans and glycosaminoglycans and of unbound water.2–4,18,19 The subepidermal hypoechogenic band, which is invisible in the young, is present in most elderly subjects at the forearm and is located in the upper dermis, in some cases occupying the greatest part of the dermis thickness, progressively increasing with age, especially on sun-exposed sites (Figure 59.1).17 Evaluating the skin on the arms of cyclists in the Tour de France, Leveque et al.20 found an increase in skin thickness in areas exposed to sunlight. The influence of solar exposure in inducing skin thickening was also reported on face and neck sites by Adhoute et al.21 On the contrary, comparing the exposed and unexposed skin of the lower neck in 30 elderly women, Richard et al.22 noticed that the thickness of the
FIGURE 59.1 Twenty-megahertz ultrasonographic skin images of the dorsal aspect of the forearm (a) in a young woman (aged 21) and (b) in an elderly woman (aged 83) with marked sun exposure during the lifetime. In elderly skin on sun-exposed sites a subepidermal hypoechogenic band is frequently observable, together with skin thinning after the age of 70.
subepidermal hypoechogenic band was greater on sunexposed sites, but skin thickness was greater on unexposed ones. Aiming at the evaluation of actinic skin damage, Hoffmann et al.23 evaluated thickness and echogenicity of the subepidermal hypoechogenic band, together with skin thickness, at different skin sites characterized by intense (forehead and cheek), moderate (volar and dorsal forearms), and slight (buttocks) sun exposure. The subepidermal hypoechogenic band was observable in sun-exposed areas, with a greater thickness and lower densitometric values, which diminished with age, in intensively exposed sites, without variations in skin thickness up to 70 years of age.23 These observations were confirmed by the semiautomatic quantification of skin thickness on a large population of young and elderly subjects, evidencing sex-, site-, and age-dependent differences.24 Ultrasonographic skin variations with age can be more precisely quantified by image analysis. Selecting different intervals for image segmentation and highlighting and calculating their extension, differences in skin echogenicity between different sites, sex, and age groups were measured.25–28 The evaluation of the dermis trough the selection of different amplitude intervals for image segmentation showed that echographic modification during age is not limited to the upper dermis, characterized by the appearance of the subepidermal hypoechogenic band, but also occurs in the lower dermis, which seemed more echogenic in elderly subjects in all examined sites.25–28 Nonuniform variations in skin echogenicity were observed from childhood to adulthood, depending on the different skin areas.29 Whereas on the face and the trunk echogenicity was lower in adults than in children, a gradual increase was observed on the limbs with aging29 (Table 59.1). Site-to-site variations in skin echogenicity and thickness related to aging were reported.30,31 Skin thickness measurements of facial skin in different age groups yielded contrasting results. Denda and Takahasi30 measured skin thickness on the forehead and the cheek and
Ultrasound Assessment of Skin Aging
513
TABLE 59.1 Echogenicity Variations According to Age (0 to 30 Pixel Areas) 2–6 Years Forehead Cheek Upper abdomen Interscapular region Volar forearm Dorsal forearm Elbow crease Lower leg a
2351 2218 4031 3239 2231 3241 1817 2712
± ± ± ± ± ± ± ±
1107 1421 1252 1043 985 1446 1161 1138
25–40 Years 6267 4940 5123 5418 2365 3576 2542 2613
± ± ± ± ± ± ± ±
1824 a 1818 a 1693 a 2301 a 1102 1753 1845 2288
Significant with respect to children (p > 0.05).
observed a decrease with age, whereas Takema et al.31 found an increase on the forehead, the eye corner, the cheek, and the mouth corner. Different observations can be due to the employment of an A-scanner, which provides a unidimensional representation of skin echogenicity, and to an inaccurate specification of the evaluated skin areas. In fact, thickening with age appeared as an uneven process, showing a greater increase in skin thickness at the lateral regions of the forehead than at the central one.32 Moreover, skin thickness variations of facial skin related to aging did not show a decreasing trend, as did other skin areas. On the contrary, significant increments in skin thickness values were observed at most of the assessed facial skin sites.32 Collagen content, which is a major component of skin thickness, has been observed to decrease more rapidly than skin thickness with aging.33 Therefore, it is possible that, on actinically damaged facial skin, the decrease in collagen and ground substance content, which gradually takes place with aging, is counterbalanced by the overall rearrangement of the dermal collagen network and the accumulation of elastotic material. With regard to echogenicity, facial skin showed a poor reflectivity in both the young and the elderly, compared to other skin sites, such as the forearm, where the dermis was highly echogenic and the dermis–hypodermis boundary well outlined. As evaluated by image analysis, overall facial skin echogenicity appeared to increase with age, mainly due to an enhancement of the lower dermis echoes, rather than a decreased echogenicity of the upper dermis.32
59.4 CONCLUSIONS Age-related changes in skin structures can be evaluated and quantified by means of high-frequency ultrasound. Whereas skin thickness showed a gradual increase from birth to adulthood, only at 70 years of age does it tend to decrease, depending on skin atrophy. Skin maturation
leads to variations in the intensity of dermal echogenicity, also depending on the different skin areas and their sun exposure. Great differences in the distribution of skin reflectivity can be appreciated at different ages: in the elderly a subepidermal hypoechogenic band is frequently observable, especially on sun-exposed sites, together with an increase in lower dermis echogenicity. On the other hand, in children and young adults the skin echostructure appears more homogeneously featured.
REFERENCES 1. Kligman, L.M. and Kligman, A.M., The nature of photoaging: its prevention and repair, Photodermatology, 3, 215, 1986. 2. Smith, L., Histopathologic characteristics and ultrastructure of aging skin, Cutis, 43, 415, 1989. 3. Braverman, I.M. and Fonferko, E., Studies in cutaneous aging. I. The elastic fiber network, J. Invest. Dermatol., 78, 434, 1982. 4. Warren, R., Gartstein, V., Kligman, A.M., Montagna, W., Allendorf, R.A., and Ridder G.M., Age, sunlight, and facial skin: a histologic and quantitative study, J. Am. Acad. Dermatol., 25, 751, 1991. 5. Leveque, J.L., Porte, G., de Rigal, J., Corcuff, P., Francoi, A.M., and Saintleger, D., Influence of chronic sun exposure on some biophysical parameters of the human skin: an in vivo study, J. Cutan. Aging Cosmet. Dermatol., 1, 123, 1989. 6. Marks, R., Methods for assessment of cutaneous ageing, Int. J. Cosmet. Sci., 3, 141, 1990. 7. Berardesca, E., Farinelli, N., Rabbiosi, G., and Maibach, H.I., Skin bioengineering in the noninvasive assessment of cutaneous aging. Dermatologica, 182, 1, 1991. 8. Oikarinen, A., Aging of the skin connective tissue: how to measure the biochemical and mechanical properties of aging dermis, Photodermatol. Photoimmunol. Photomed., 10, 47, 1994. 9. Richard, S., de Rigal, J., Lacharriere, O., Berardesca, E., and Leveque, J.L., Noninvasive measurement of the effect of lifetime exposure to the sun on the aged skin, Photodermatol. Photoimmunol. Photomed., 10, 164, 1994. 10. Conti, A., Schiavi, M.E., and Seidenari, S., Capacitance, transepidermal water loss and casual level of sebum in healthy subjects in relation to site, sex and age, Int. J. Cosmet. Sci., 17, 77, 1995. 11. Altmeyer, P., Hoffmann, K., Stucker, M., Goertz, S., and el-Gammal, S., General phenomena of ultrasound in dermatology, in Ultrasound in Dermatology, Altmeyer, P., el-Gammal, S., and Hoffmann, K., Eds., SpringerVerlag, Berlin, 1992, p. 55. 12. Gniadecka, M., Serup, J., and Sondergaard, J., Agerelated diurnal changes of dermal oedema: evaluation by high frequency ultrasound, Br. J. Dermatol., 131, 849, 1994.
514
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. Callens, A., Vaillant, L., Leconte, P., Berson, M., Gall, Y., and Lorette, G., Does hormonal skin aging exist? A study of the influence of different hormone therapy regimens on the skin of postmenopausal women using noninvasive measurement techniques, Dermatology, 193, 289, 1996. 14. Gonzalez, R.C. and Wintz, P., Digital image fundamentals, in Digital Imaging Processing, Gonzalez, R.C. and Wintz, P., Eds., Addison-Wesley, Reading, MA, 1987, p. 13. 15. Seidenari, S. and Di Nardo, A., B-scanning evaluation of irritant reactions with binary transformation and image analysis, Acta Derm. Venereol., 175 (Suppl.), 9, 1992. 16. Tan, C.Y., Stathan, B., Marks, R., and Payne, P.A., Skin thickness measurement by pulsed ultrasound: its reproducibility, validation and variability, Br. J. Dermatol., 106, 657, 1982. 17. de Rigal, J., Escoffier, C., Querleux, B., Faivre, B., Agache, P., and Leveque, J.L., Assessment of aging of the human skin by in vivo ultrasonic imaging, J. Invest. Dermatol., 93, 621, 1989. 18. Oikarinen, A., Aging of the skin connective tissue: how to measure the biochemical and mechanical properties of aging dermis, Photodermatol. Photoimmunol. Photomed., 10, 47, 1994. 19. Richard, S., Querleux, B., Bittoun, J., Jolivet, O., IdyPeretti, I., Lacharriere, O., and Leveque, J.L., Characterization of the skin in vivo by high resolution magnetic resonance imaging: water behavior and age-related effects, J. Invest. Dermatol., 100, 705, 1993. 20. Leveque, J.L., Porte, G., de Rigal, J., Corcuff, P., Francoi, A.M., and Saintleger, D., Influence of chronic sun exposure on some biophysical parameters of the human skin: an in vivo study, J. Cutan. Aging Cosmet. Dermatol., 1, 123, 1989. 21. Adhoute, H., de Rigal, J., Marchand, J.P., Privat, Y., and Leveque, J.L., Influence of age and sun exposure on the biophysical properties of the human skin: an in vivo study, Photodermatol. Photoimmunol. Photomed., 9, 99, 1992.
22. Richard, S., de Rigal, J., Lacharriere, O., Berardesca, E., and Leveque, J.L., Noninvasive measurement of the effect of lifetime exposure to the sun on the aged skin, Photodermatol. Photoimmunol. Photomed., 10, 164, 1994. 23. Hoffmann, K., Dirschka, T.P., Stucker, M., el-Gammal, S., and Altmeyer, P., Assessment of actinic skin damage by 20-MHz sonography, Photodermatol. Photoimmunol. Photomed., 10, 97, 1994. 24. Lasagni, C. and Seidenari, S., Echographic assessment of age-dependent variations of skin thickness. A study on 162 subjects, Skin Res. Technol., 1, 81, 1995. 25. Seidenari, S., Pagnoni, A., Di Nardo, A., and Giannetti, A., Echographic evaluation with image analysis of normal skin: variation according to age and sex, Skin Pharmacol., 7, 201, 1995. 26. Gniadecka, M. and Jemec, G.B.E., Quantitative evaluation of chronological ageing and photoageing in vivo: studies on skin echogenicity and thickness, Br. J. Dermatol., 139, 815, 1998. 27. Gniadecka, M., Effects of ageing on dermal echogenicity, Skin Res. Technol., 7, 204, 2001. 28. Sandby-Moller, J. and Wulf, H.C., Ultrasonographic subepidermal low-echogenic band, dependence of age and body site, Skin Res. Technol., 10, 57, 2004. 29. Seidenari, S., Giusti, G., Bertoni, L., Magnoni, C., and Pellacani, G., Thickness and echogenicity of the skin in children as assessed by 20-MHz ultrasound, Dermatology, 201, 218, 2000. 30. Denda, M. and Takahasi, M., Measurement of facial skin thickness by ultrasound method, J. Soc. Cosmet. Chem. Jpn., 23, 316, 1990. 31. Takema, Y., Yorimoto, Y., Kawai, M., and Imokawa, G., Age-related changes in the elastic properties and thickness of human facial skin, Br. J. Dermatol., 131, 641, 1994. 32. Pellacani, G. and Seidenari, S., Variations in facial skin thickness and echogenicity with site and age, Acta Derm. Venereol., 79, 366, 1999. 33. Shuster, S. and Black, M.M., The influence of age and sex on skin thickness, skin collagen and density, Br. J. Dermatol., 93, 639, 1975.
Imaging of Subcutaneous 60 Ultrasound Tissue and Adjacent Structures Ximena Wortsman Servicio de Imagenologia, Hospital del Profesor, Santiago, Chile
Elisabeth A. Holm and Gregor B.E. Jemec Department of Dermatology, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark
CONTENTS 60.1 60.2 60.3
Introduction ..........................................................................................................................................................515 Technical Considerations and Instrumentation....................................................................................................516 Optimal Recommendations for Performing High-Resolution Ultrasound in Subcutaneus Tissue and Adjacent Structures............................................................................................................................517 60.4 Examination Technique of Patients .....................................................................................................................518 60.5 Ultrasound Normal Anatomy of Subcutaneous Tissue and Adjacent Structures ...............................................520 60.5.1 Normal Anatomy....................................................................................................................................520 60.6 Ultrasound Pathology...........................................................................................................................................523 60.6.1 Edema and Fluid Accumulation ............................................................................................................523 60.6.2 Lymphadenopathy ..................................................................................................................................524 60.6.3 Foreign Bodies .......................................................................................................................................524 60.7 Epidermal Disease................................................................................................................................................524 60.8 Dermal Disease ....................................................................................................................................................524 60.9 Vascular and Neural Disease ...............................................................................................................................524 60.10 Joints and Tendons ...............................................................................................................................................524 60.11 Benign Tumoral Pathology ..................................................................................................................................525 60.11.1 Cystic Lesions........................................................................................................................................525 60.11.2 Vascular and Lymphatic.........................................................................................................................526 60.11.3 Lipomatous Tumors ...............................................................................................................................527 60.11.4 Fibromatous Tumors ..............................................................................................................................527 60.11.5 Neurogenic Tumors................................................................................................................................527 60.12 Malignant Tumoral Pathology .............................................................................................................................528 60.12.1 Cutaneous Melanoma ............................................................................................................................528 60.12.2 Liposarcoma...........................................................................................................................................528 60.13 Potential Pitfalls ...................................................................................................................................................528 60.14 Indications of High-Resolution Ultrasound in the Subcutaneous and Adjacent Structures...............................529 References .......................................................................................................................................................................529
60.1 INTRODUCTION Imaging has dramatically improved in the recent years. Ultrasound is no the exception; the development of new equipment allows better resolution and visualization of more superficial structures. The images have also become easier to interpret and more intuitively understandable to
clinicians. Particularly in the study of subcutaneous tissue, this had been a great advance and has now made ultrasound the method of choice in this area. The ultrasound study of skin had been a challenge for sonographers because there was so little information available in highresolution ultrasound with the new technologies in superficial structures. 515
516
Handbook of Non-Invasive Methods and the Skin, Second Edition
60.2 TECHNICAL CONSIDERATIONS AND INSTRUMENTATION The ultrasound is the result of the sound wave displacement through the matter, and its propagation velocity is determined by the tissue physical properties. In medical equipment the sound propagation velocity is usually set at 1540 m/s, although the propagation velocity is determined by physical characteristics of tissue. This means that the propagation velocities in, e.g., muscle, renal tissue, liver, blood, water, air, and bone are all different. The measurement unit of the acoustic frequency is the hertz (1 H = 1 cycle/s). In the case of high frequencies, the units are expressed in kilohertz (1 KHz = 1000 H) or megahertz (1 MHz = 1,000,000 H).The frequencies used for diagnostic purposes go from 2 to 17 MHz in high-resolution ultrasound. The propagation velocity of sound is important for identifying the tissue and the actual measurement of distances in various tissues. With the passing of different tissues the sound loses its power and energy, and this factor is important in the determination of the efficacy of penetration at each specific tissue. High frequencies have faster attenuation than lower frequencies. Therefore, in the study of superficial layers, high frequencies (7 MHz or more) are used. The connection between the body and the ultrasonic equipment is made through transducers or probes, which are devices that convert electric energy in acoustic pulses. The pulse repetition frequency (PRF) of probes is variable. The transducers also serve as receptors of the reflected echoes, transforming differences of weak pressures caused by the reflected sound waves into electric signals for processing using piezoelectric crystals. The range of frequencies produced by a transducer is called bandwidth. Generally, at lower frequencies the transducers have higher bandwidths. The transducers can be classified according to the steer of the ultrasonic beam into mechanical or electronic. The electronic steer is used by linear array transducers, generating parallel beams of ultrasound perpendicular to the probe face, giving a rectangular image format, which is the shape of the screen representation usually seen in the superficial ultrasound studies. These lineal transducers are the most appropriate probes for the study of superficial and vascular structures. There are other kinds of transducers called curved array that give a curved image, and they are used in abdominal, pelvic, and obstetric studies. Phased array transducers produce a sectorial image and are primarily used in cardiac studies. The transducer and frequency choice are big subjects that must be considered in the realization of ultrasound studies in superficial tissues. To visualize upper-layer structures, it is recommended to use frequencies from 7.5 MHz and up, taking into account that the more superficial the structure we are going to look at, the higher the frequency we need to use. For structures located at no more that 3 cm of depth
frequencies from 7.5 to 17 MHz are the most appropriate. For the evaluation of deeper structures of 12 or 15 cm of depth, such as intrabdominal structures, frequencies of 2 to 7 MHz are appropriate, while structures at intermediate depth are examined by 5 to 7 MH (Figure 60.2). The information obtained by machines in real time is transformed in bidimensional images at gray scale. The factors that determine the image quality include the resolution and the contrast of the image. These are influenced by the width and thickness of the ultrasound beam and the differentiation of structures that are located one very close to another. Structures are described in specific terms: anechoic when a structure presents a black image in the screen, hyperechoic when it is white, and hypoechoic or echoic when it is gray or has an intermediate tone at the gray scale. Just as in other imaging techniques, the ultrasound can produce many kinds of artifacts, and some of these may be useful in the identification of specific structures, e.g., anisotropy (hypoechoic or gray area inside the fibrillar structure as a tendon or ligament that changes to hyperechoic if the angle relative to the transducer is changed by moving it over the skin). This phenomenon is produced by putting the axis obliquely to that of the ligament or tendon, comparing with the ultrasonic beam. Posterior shadowing, i.e., a posterior attenuation of sound waves (black areas), can be seen in structures with calcic components or areas that are highly cellular. Posterior enhancement implies a hyperechoic reverberation of sound (white areas) in a deep plane to a cystic structure similarly rich in additional information. Normally cystic or fluid-filled structures are anechoic (black), the solid structures are hypoechoic (different tones of gray), and the bone or calcic structures are hyperchoic (white). Among the operator-controlled variables there is the acoustic output or power that is the total energy transmitted over the entire cross-sectional area of the ultrasound beam per unit time. Increasing the transmit power will increase the amplitude of the voltage pulses across the crystals. The stronger voltage pulses will produce higheramplitude transmitted waves that will yield higher-amplitude echoes. Advantages of using a high-power setting are improved signal-to-noise ratios and better penetration. A disadvantage of using a high-power setting is the increased potential for biological effects, although there is as yet no report of cases about this. The other operator-controlled variable is the gain, which amplifies the signals from the returning echoes for display. Gain is analogous to the volume control on a radio. Increasing the volume of the radio increases the loudness of the broadcast. However, if a station is not being received adequately, increasing the volume simply makes the static noise louder. In other words, the gain or volume control cannot discriminate between signal and noise. Every electronic signal that passes through the gain circuitry is amplified, regardless
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
517
FIGURE 60.3 Doppler color nail bed.
FIGURE 60.4 Power Angio cavernous hemangioma. FIGURE 60.1 Example of a high-resolution ultrasound machine.
FIGURE 60.2 Examples of lineal probes with frequencies from 12 to 5 MHz (left) and the footprint from 15 to 7 MHz (right).
of whether it is image data, speckle, or electronic noise. Ultrasound is capable of identifying movements of elements through the Doppler effect, such as flow presence and direction of the blood. In modern equipment you can visualize this as Doppler color, which uses a color map for displaying the information based on the frequency shift’s detection of objects in movement. In this modality you can study arteries and veins and it is possible to identify thrombae or atheromatous plaques with great sensitivity (Figure 60.3). The other Doppler modality is called power Doppler or power Angio. It uses a color map for showing the power or amplitude of the Doppler signal with less noise, permitting higher effective gain settings and more sensitivity for flow detection, but does not show direction or velocity of the flow. This is, however, very useful in the study of slow flow. The use of the power Angio technique with three-dimensional imaging can
provide a rotational view of global vasculature and architecture, for example, in presurgical evaluation of tumors (Figure 60.4). Modern ultrasound high-resolution machines are big computers capable of processing over 9216 channels of information per image frame (Figure 60.1). The concept of image frame is closely related to the temporal resolution of an ultrasound unit. That is, the higher the frame rate, the more responsive the image will appear when objects in motion are imaged or when the probe is moved. Another function of the receiver is the compression of the wide range of amplitudes returning to the transducer into a range that can be displayed to the user. The ratio of the highest to the lowest amplitudes can be displayed and expressed in decibels and is referred to as the dynamic range. In clinical applications the dynamic ranges in use are over 120 dB. The modern machines can have dynamic ranges of 170 dB, and they have composition technologies of the image from different angles of view. The software often permits more sharpness, resolution, and contrast of the images, and may have extended fields of view of long corporal segments.
60.3 OPTIMAL RECOMMENDATIONS FOR PERFORMING HIGH-RESOLUTION ULTRASOUND IN SUBCUTANEUS TISSUE AND ADJACENT STRUCTURES • • • • • •
High-resolution ultrasound machine Lineal transducers with 7.5- to 17-MHz frequency Possibility of color Doppler and color power Angio Software for high-resolution imaging (e.g., compound imaging and extreme resolution) Extended field of vision Trained operator
518
Handbook of Non-Invasive Methods and the Skin, Second Edition
Epidermis Dermis
Subcutaneous Tissue
FIGURE 60.8 Anterior tibial tendon fibrillar pattern. FIGURE 60.5 Normal skin layers at the forearm.
FIGURE 60.9 Power Angio of the carotid artery at the bifurcation site.
FIGURE 60.6 Median nerve normal fascicular pattern.
FIGURE 60.10 Anterior talofibular ligament.
FIGURE 60.7 Forearm muscles’ extended field of view.
60.4 EXAMINATION TECHNIQUE OF PATIENTS Always record patient clinical history data and perform a physical examination that includes palpation of the exact location of the affected area or lesion. Ultrasound examination must include cross-sectional (transverse view) and longitudinal scans, in order to achieve multiple planes of
vision and gain an overview of the local anatomy. A gray scale is used first, and only when the presence and disposition of vessels is studied is a Doppler color or power Doppler used. At each examination, the operator must extract the relevant clinical data to enhance the diagnostic precision. This includes dynamic studies such as movements of, e.g., tendons, to verify the degree of displacement and functional state. For the study of the dermis and subcutaneous layers, frequencies of 12 to 17 MHz and possibly the use of a pad of gel or standoff pad of 0.5 to 1.0 cm are recommended. This allows a good contact between the probe
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
519
Dorsal plate A Ventral plate Matrix area Bone margin distal phalanx
FIGURE 60.11 Nail Image. Note the bilaminar structure of the dorsal and ventral plates and the matrix area.
FIGURE 60.14 Sacral subcutaneous abcess.
A
FIGURE 60.12 Subepidermoid low echogenic band (SLEB) that corresponds to the presence of elastosis in sun-exposed areas of the skin.
H
FIGURE 60.13 Hematoma adjacent to the medial gastrocnemious muscle.
and the skin for transmitting the ultrasound pulses. In some types of equipment it is possible to have small compact linear transducers with a footprint shape, initially designed for intraoperative studies, which also give good results in the imaging of the dermoepidermal area, mainly because of their high frequency. In all cases, it is necessary to use contact gel between the skin of the patient and the probe, and it is possible to get sterile gel for invasive procedures. The probes must be cleaned with special disinfectants that do not harm the transducer’s crystals.
FIGURE 60.15 Subcutaneous nonclinical suspected fluid collection in a patient with erysipela collection.
In the study of tendons, ligaments, or nerves, the dynamic movements play an important role, e.g., flexoextension of the muscle, varus or valgus of the articulations (Figure 60.5 through Figure 60.10). The movement allows for better identification of structures involved in the dynamic changes. There are many protocols for studying the muscular tendineous structures of the different articulations, as well as vascular studies for arteries and veins that are available in many ultrasound basic books, and even on the Internet at the websites related to radiology or imaging. For example, see: www.acr.org: Look in “Practice Guidelines of Ultrasound.” www.medical.philips.com: Look for education in ultrasound and vascular categories. www.gemedicalsystems.com: Look in the ultrasound segment. One of the useful things about ultrasound is the possibility to compare in real time, because there are many normal variants, and sometimes it is easy to detect abnormality based on the comparative analysis of different areas. Vascular studies are necessary to obtain spectral curves of flow that are characteristic for the vein or artery.
520
Handbook of Non-Invasive Methods and the Skin, Second Edition
Gas Abcess
FIGURE 60.16 Gluteal abcess with gas inside that corresponds to the hyperechoic zone inside the fluid.
FIGURE 60.18 Cellulitis with edema of the subcutaneous. Note the anechoic bands at the subcutaneous tissue corresponding to fluid between the fat lobules.
FIGURE 60.17 Fat focal necrosis between calipers.
The arteries present a biphasic or triphasic flow, where it is possible to distinguish the systolic and diastolic peak, and the veins have a monophasic flow that can vary with the respiratory cycle. In veins it is necessary to make valsalva maneuvers for studying venous competence, asking to the patient to cough or push, looking for the normal response, which stops the venous flow. It is important to remember that infants and children use higher frequencies than adults because of the lower distance between the focus area and the probe. In some children it is necessary to use sedation, mainly in vascular studies with Doppler, which takes more time and requires a quiet patient, because movements can produce artifacts that make it impossible to have good spectral curves or analysis. The sedations most commonly used are chloral hydrate or some kind of anxiolytic.
60.5 ULTRASOUND NORMAL ANATOMY OF SUBCUTANEOUS TISSUE AND ADJACENT STRUCTURES Ultrasound can distinguish the dermis and superficial subcutaneous layer with the use of 12- to 17-MHz probes.
FIGURE 60.19 Normal lymphonode with hyperechoic center and hypoechoic border.
With 7.5- to 12-MHz transducers we can better observe the lower subcutaneous and muscle layers. In modern equipment there are transducers that have variable frequencies, so it is easy to jump from watching superficial to deep planes by just changing the probe or type of frequency used. Several machines come with a preset setup that includes predefined types of setups calibrated for a defined organ or layer predetermined by the examiner that enhance and optimize the image. This is important in the study of lesions that initially appear superficial but which also contain a deeper element.
60.5.1 NORMAL ANATOMY The dermal layer is hyperechoic and homogeneous compared to the subcutaneous tissue, and it is located beneath the hyperechoic line of the epidermis. The subcutaneous tissue is hypoechoic and contains adipose lobules separated by hyperechoic septa, as well as vessels, nerves, lymph nodes, tendons, and ligaments (Figure 60.5). The thickness of the dermis varies depending on the segment studied, as well as the age, sex, and corporal contexture
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
521
Bullae
FIGURE 60.20 Inguinal inflammatory adenopathies with loosening of the hyperechoic center.
B
FIGURE 60.23 Bullous pemfigus with two superficial blisters on the left and with an associate ulcer on the right.
B Ulcer
FIGURE 60.21 Glass foreign body. FIGURE 60.24 Bullous pemfigus with two superficial blisters on the left and with an associate ulcer on the right.
FIGURE 60.22 Wood foreign body.
of the patient. The skin at the dorsum of the hand is thin in comparison to the skin at the sole of the feet, and the amount of vessels at the subcutaneous tissue of the front is larger than in the forearm dorsum. The vascular structures are tubular anechoic images, sometimes tortuous in their way. The nerves look like hypoechoic solid strings that in the cross-sectional images present anechoic or black dots inside corresponding to the neural fascicles (Figure 60.6).28 Today, with high frequencies, it is even possible to observe nerves as small as the digital palmaris nerves at the fingers. The hair follicles appear like parallel
FIGURE 60.25 Psoriatic plaque with wavy and thick epidermis: abnormal (left) vs. normal (right).
hypoechoic linear bands obliquely perpendicular to the skin surface. When they pass the skin surface they change their echogenicity to hyperechoic. Tendons and ligaments that go subcutaneous present a fibrillar hyperechoic homogenous pattern (Figure 60.8).29 The nail is a visible structure with good detail, and in a noninvasive way, because it is possible to distinguish the ventral and dorsal
522
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 60.26 Psoriatic hyperechoic plaque that affects only the ventral plate of the nail.
FIGURE 60.29 Thromboflebitis of a superficial vein without flow at the Doppler color study.
FIGURE 60.30 Thromboflebitis of a superficial vein without flow at the Doppler color study. FIGURE 60.27 Calcinosis deposits in a patient with dermatomiositis. Note the posterior acoustic shadowing of the calcium.
FIGURE 60.28 Discoid eritematous lupus: normal (right) vs. abnormal (left). Note the thick inflammatory changes at the dermis.
plates as hyperechoic lineal bands forming a bilaminar structure (Figure 60.11). It is possible to see the subungueal matrix as a hypoechoic region at the base of the nail.21 The epidermis looks like a hyperechoic continuous line, except at the sole of the feet, where it is a bilaminar hyperechoic structure; here the first line represents the surface of the skin and the other the dermoepidermic union. In the sun-exposed areas we can see a hypoechoic and subepidermal band called subepidermal low echogenic band (SLEB), which corresponds to the presence of elastosis and edema in the papillary dermis, which is a consistent echo-structural finding in aged and photo-
FIGURE 60.31 Normal median nerve with fascicular pattern with an area of 7 mm2.
damaged skin (Figure 60.12). The thickness of a subepidermal low echogenic band is considered to reflect the degree of cutaneous aging and may be used for the monitoring of the severity of photoaging, and the efficacy of drugs is rapidly expanding.11 For the differentiation between a venous or arterial flow the advantage of real-time ultrasound is useful. Using this method, it is possible to see how the normal veins are compressed with the transducer pressure, and also we can look for the presence of valves at the main veins, such as the femoral common vein. Compression is lost in the presence of venous thrombosis.39 In contrast, the arteries have a muscular layer, so they are not easily compressed and they have a visible pulse.
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
523
T T
FIGURE 60.32 Median nerve neuropathy with a thick and hypoechoic pattern (area, 17 mm2).
FIGURE 60.35 Extensor tendon tenosynovitis comparison between abnormal (left) and normal (right).
FIGURE 60.36 Calcic tendinitis of the Achilles tendon.
FIGURE 60.33 Normal fibrillar pattern of the tendon.
FIGURE 60.34 Comparative images — abnormal (left) and normal (right) — of a triceps tendon calcic tendinitis.
When you study arteries with Doppler color and spectral analysis, you can see their typical curves with systolic and diastolic peaks. The venous vessels have a continuous flow. For the study of vascular structures it is recommended that you use linear transducers with frequencies from 7.5 to 12 MHz, depending on the location of the veins. If they are located deeply, 7.5 MHz is appropriate; if they are superficial, use the 12-MHz frequency.
FIGURE 60.37 Cross-sectional view of the tenosynovitis flexor tendon with enlarged synovial sheath. The median nerve (MN) is displaced in this case to the periphery of the carpal tunnel by the inflammation.
60.6 ULTRASOUND PATHOLOGY 60.6.1 EDEMA
AND
FLUID ACCUMULATION
The edema of the subcutaneous tissue produces an increase of the echogenicity of the adipose tissue with filling of the spaces between the lobules with anechoic fluid. Regional contralateral areas provide control images in unilateral affections. There are different patterns of
524
Handbook of Non-Invasive Methods and the Skin, Second Edition
edema: subepidermal (lipodermatosclerosis), uniform (lymphedema), and deep dermal (heart failure).12 The edematous dermis increases its echogenicity and becomes thick, which produces a loosening of the border between dermis and subcutaneous tissue (Figure 60.18). With color Doppler or color power Angio it is possible to observe the increase of vessels in the affected areas. The subcutaneous fluid collections, such as hematomas, seromas, or abscesses, are easily visible with ultrasound; nevertheless, it is not possible to determine with accuracy the real composition (Figure 60.13 through Figure 60.16). In older hematomas it is possible to detect the presence of a fluid–fluid level.
60.6.2 LYMPHADENOPATHY The lymph nodes are seen as ovoid structures, with hypoechoic borders and a hyperechoic center (Figure 60.19). In cases with inflammatory or infiltrative diseases the lymph nodes increase their size to more than 1.0 cm in cross section and lose their hyperechoic center, becoming round, totally hypoechoic structures in which there is no difference of echogenicity between periphery and center, and the vessels are visible not only at the hilus (Figure 60.20). In the reactive lymph nodes we can see edema of the surrounding adipose tissue. It is not possible to differentiate benign from malign lymphatic structures only with ultrasound (Figure 60.52). It has been suggested that sentinel lymph nodes lose their symmetry and the ratios between the central and peripheral change, but these changes are at best suggestive of malignancy (Hahle 2003).
60.6.3 FOREIGN BODIES The presence of subcutaneous foreign bodies is one of the most useful applications of ultrasound, since they look like fragments or linear structures, mostly hyperechoic, but with different amounts of echogenicity, depending of their composition material (Figure 60.21 and Figure 60.22). Generally the glass and metal are highly hyperechoic. A surrounding hypoechoic rim and posterior shadowing or reverberation aid detection. Wood is slightly hyperechoic and may be difficult to see.4
60.7 EPIDERMAL DISEASE With ultrasound it may be possible to visualize the blisters and subclinical lesions in bullous disorders of the skin (Figure 60.23 and Figure 60.24). In psoriasis alterations in the epidermal layer at the plaque zone can be visualized. The entrance echo takes a wavy appearance instead of linear (Figure 60.25). At the nail it is possible to see hyperechoic plaques that most likely correspond to hyperkeratoses. These first involve the ventral plate. The plates
become wavy, and in later stages, the borders of the plates become lost, increasing the distance between the ventral plate and the bone margin of the distal phalanx (Figure 60.26). These findings may become one of the ways of monitoring the activity of the disease in the future.
60.8 DERMAL DISEASE The presence of subcutaneous calcinosis can be detected as hyperechoic foci that produce acoustic shadowing. Not every calcium deposit corresponds to calcinosis. You must be careful to not confuse these findings with calcified granulomas, which are seen in zones with frequent antecedents of trauma or injections. Generally in calcinosis, the deposits are bigger in size and composed by many small foci (Figure 60.27). In studies with 13-MHz ultrasound probes in localized scleroderma there are descriptions of undulations of the dermis, disorganization, loss of thickness, and thickened hyperechoic bands in the hypodermis.5 In discoid lupus, the dermis becomes hypoechoic and increases its thickness. It is possible to compare abnormal to normal areas, as well as to monitor the skin changes with treatment (Figure 60.28).
60.9 VASCULAR AND NEURAL DISEASE Thromboflebitis corresponds to thrombosis at the superficial venous system. In bidimensional studies with gray scales we see the distension of the vein and hypoechoic thrombotic material inside the lumen. There is an absence of compression, and the color Doppler with spectral analysis does not show flow inside (Figure 60.29 and Figure 60.30). The chronic valvular incompetence produces inflammatory and dystrophic changes at the skin secondary to the venous stasis of perforants and communicating superficial veins that cause the cutaneous hyperpigmentation and induration, also called lipodermatosclerosis,24 with subsequent development of venous ulcers. Median nerve affection, in carpal tunnel syndrome, is seen as the enlargement and decrease of the echogenicity of the nerve, which becomes flattened at the flexor retinaculum (Figure 60.32). In the presence of inflammation the nerves can also be affected, even presenting an increase of circulation at Doppler color in the perineural area, which suggests the presence of neuritis.
60.10 JOINTS AND TENDONS Some tendons and ligaments are superficially located, as in the wrist and ankles. Tendinitis is the inflammation of the tendons and produces enlargement and decreases the echogenicity of the tendon, respecting the fibrillar pattern. The synovial sheath is also enlarged and decreased in echogenicity when it becomes affected by inflammation
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
525
C
Flexor
FIGURE 60.38 Full-thickness tear of the supraspinatus tendon. Note the black hole that crosses the thickness of the tendon.
FIGURE 60.40 Synovial cyst. Anechoic oval image attached to a flexor tendon.
C
FIGURE 60.39 Sebaceous cyst. Anechoic round image connected with a hair follicle.
FIGURE 60.41 Oral mucoid cyst endoral cavity ultrasound of the lesion.
60.11 BENIGN TUMORAL PATHOLOGY and presents as a hypoechoic layer, sometimes with anechoic fluid around the tendons (Figure 60.33, Figure 60.35, and Figure 60.37). Not all fluids around the tendon mean pathology. There are small amounts of normal fluid around some tendons in some locations, for example, at the bicipital tendon or at the posterior tibial tendon in the inframaleolar area. When the synovial sheath is inflamed, we call it tenosynovitis. Tendinosis is a fibrillar degeneration of tendons that is common with aging and produces hypoechoic and enlarged chronic tendons. Sometimes this can have a focal manifestation, and it must not be confused with tears. Tendon rupture is seen as an interruption of the fibrillar pattern and is clinically defined as partial- or full-thickness tears, depending on the thickness compromised (Figure 60.38). Nodular tenosynovitis consists of nodular solid hypoechoic formations attached to the synovial sheath surface that really correspond to localized forms of giant cell tumors that are seen at the palmar surface of the fingers and can be very painful. They can sometimes appear as more than one solid nodule.
60.11.1 CYSTIC LESIONS Cysts are anechoic sacular formations with fluid inside. The location and presence of detritus or the appearance of the walls can be a clue for the diagnosis. For example, sebaceous cyst (Figure 60.39) presents a communication to the hair follicle that can be seen at higher frequencies like 17 MHz. Dermoid cysts (Figure 60.42) are mostly oval or round and they have detritus or sediment inside. Synovial cysts are normally around or communicated with tendons or joints (Figure 60.40). Most of them are attached to the synovial sheath and can be seen alone or related to systemic pathology as rheumatoid disease. Ganglions are also cysts related to musculotendineous structures or joints, but the difference between the synovial cyst and the ganglion is the epithelial covering. The ganglions have a mixoid origin, and the synovial cysts have a synovial origin. Mucoid cysts are seen near the nail or around the distal interphalangeal joint; they do not cause erosion and they are avascular, which is important in the differential diagnosis from a glomus tumor. The glomus tumor is
526
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 60.42 Dermoid cyst at the eyebrow. This is a typical location for this lesion. Note the echogenic debris inside.
FIGURE 60.45 Lipoma at the interoseus muscle of the hand.
FIGURE 60.46 Lipoma at the dorsum of the forearm. FIGURE 60.43 Cavernous hemangioma with big and tortuous vessels inside, visible in color Doppler.
coccyx must not be misinterpreted as a pilonidal cyst or a sinus tract.
60.11.2 VASCULAR
FIGURE 60.44 Capilar hemangioma with few vessels and hyperechoic echogenicity at the Doppler.
derived from the neuromioarterial glomus, which produces important pain with the pressure. It is located in the subungueal area and has a hypoechoic solid appearance with vessels inside. It often produces erosions of the bone margin. Mucoceles are cystic anechoic lesions that are seen at the oral mucose (Figure 60.41). Pilonidal cysts are located in the sacrococcygeal area and they have detritus inside. In children the cartilaginous
AND
LYMPHATIC
There are many kinds of hemangiomas. Cavernous hemangiomas are hypoechoic and vascular lesions that have notorious vessels at Doppler color studies (Figure 60.43). Capillary hemangiomas are hyperechoic and have few or no vascularity at Doppler color exams (Figure 60.44). It is important in these cases to use the Doppler color and define the thickness and location of the vessels. The arteriovenous hemangiomas can show abnormal communications between arteries and veins with turbulent shunts of high flow. Ultrasound could monitor the size and characteristics of the tumor in time. Lymphangiomas are composed of sequestered lymphoid tissue lined by lymphatic endothelium, and they can be cavernous with largesize lymphatic vessels or small capillary lymphatic vessels. Cystic lymphangioma or cystic hygroma is the more frequent lymphatic tumor, and it is seen as a cystic anechoic mass sometimes multiloculated and with septations inside. Angiomatosis corresponds to the diffuse infiltration of tissue by hemangiomatous or lymphangiomiomatous lesions. They are a mixture of different tissues like capillary, cavernous, and arteriovenous.
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
527
T
FIGURE 60.47 Dematofibroma at the knee.
FIGURE 60.49 Ulnar nerve schwanoma; note the emerging nerve at the right side.
FIGURE 60.50 Neuroma in a patient with a stump. FIGURE 60.48 Plantar fibromatosis with fibrotic nodule at the plantar fascia.
hose disease (plantar fibromatosis) solid hypoechoic nodules appear attached to the palmar or plantar aponeurosis.
60.11.3 LIPOMATOUS TUMORS These are the most frequents subcutaneous tumors. They have a benign nature and are composed of adipose tissue. Lipomas have a hypoechoic appearance with hyperchoic septa and sometimes adopt a fusiform or oval shape. They have sharp borders and normally are avascular (Figure 60.45 and Figure 60.46). The finding of zones of different echogenicities inside a tumor that suggest a lipoma or the appearance of vessels inside the tumor must make us suspect the presence of malignancy.
60.11.4 FIBROMATOUS TUMORS At ultrasound these tumors look like nodular hypoechoic tumors or masses, some of them with lobulations (Figure 60.47 and Figure 60.48). Because of their highly fibrotic cellularity, sometimes they produce some acoustic posterior shadowing. Generally they are avascular. In Dupuytren contracture (palmar fibromatosis) or Ledder-
60.11.5 NEUROGENIC TUMORS These tumors are derived from neural tissue and follow the anatomical ways of the neural structures. The most common tumors are the schwanomas or neurilemmomas. Schwanomas are seen as hypoechoic solid masses that can contain calcifications inside and present an eccentric location in the axis of the nerve (Figure 60.49). Neurofibromas or neurilemmomas have a spindle shape and are centrally located at the axis of the nerve. Most of them are solitary, and they do not have capsules. They also have slow-growing, painless masses. In patients with traumatic stumps, it is not infrequent to observe neuromas that can be observed as extremely painful solid subcutaneous nodules that have an afferent hypoechoic nerve (Figure 60.50). In the plantar zone, there is a painful entity called Morton’s Neuroma which is not really a neuroma. It is fibrotic hypoechoic and irregular scar tissue that is located mostly between the third and fourth metatarsal.
528
Handbook of Non-Invasive Methods and the Skin, Second Edition
it to be a liposarcoma, but sometimes it is difficult to predict because there are some undifferentiated forms, and only biopsy can give the proper result.
60.13 POTENTIAL PITFALLS
FIGURE 60.51 Melanoma nodule.
FIGURE 60.52 Metastasic adenopathy of melanoma.
60.12 MALIGNANT TUMORAL PATHOLOGY 60.12.1 CUTANEOUS MELANOMA This is one of the lesions possible to study with ultrasound. It is feasible to see the tumor itself as a hypoechoic solid nodule (Figure 60.51) and make the staging as well, looking for lymph nodes’ regional dissemination (Figure 60.52) or distance metastasis.47,48 Ultrasonographic diagnosis of melanoma metastasis has a sensitivity of 92.2% and a specificity of 98.2%, whereas diagnosis by physical examination has a sensitivity of only 51.3% and a specificity of 90.9%. So ultrasound examination was found to be highly effective and superior to physical examination for early detection of locoregional melanoma metastasis.38
60.12.2 LIPOSARCOMA This another subcutaneous tumor possible to see under ultrasound as a mass or nodule, hypoechoic and undulated. This tumor is derived from the mesenquima, and it was described in 1857. It is more frequent at the inferior extremities, mainly at the thigh. When it is possible to recognize adipose tissue inside a tumor, one can suspect
There are pitfalls that depend on the operator, and others related to the method itself. Pitfalls derived from the operator are the lack of anatomical or clinical knowledge necessary to interpret the exams, the absence of proper training in the ultrasound technique, and inadequate clinical and imaging correlation. Ultrasound is very sensitive to the operator, because it is a real-time technique, so the registry of the images depends on the expertise of the examiner. The pitfalls derived from the method itself are the result of numerous artifacts, like anisotropy of the tendons or ligaments, which we will correct making the right angle to the structures; the more parallel to the structure axis that we are, the more possible it is to achieve a better image in gray scale. In the vessels study it is important to have an angle equal to or less than 60˚, because if we have an angle near or equal to 90˚ the normal flow will not show and thrombosis may be misdiagnosed. You must be careful to not confuse the absence of venous flow with thrombosis immediately, because you must rule out the presence of a collection that can be compressing the vessels, or you must readjust your parameters, like lowering the pulse repetition frequency or the sample volume, adjusting your steer angle, etc. It is important to visualize the structures in many angles, including transverse and longitudinal scans, so we can confirm if a lesion is real or not because we can see the lesion in every projection. Compression can play a role as a helper or not, because in the venous study it is important to compress the veins when we suspect thrombosis, because one of the first things that the veins lose is elasticity, so they do not normally collapse with the pressure. In the fluid collections hard compression can hide the collection. We need to be aware of the depth of a lesion because it is important to make the adequate adjustment of focus or use the right frequency with the objective of not losing any detail or zone of the lesion. The Doppler color studies have a lot of artifacts, so the operator must be experienced to avoid misinterpretations.13 In the future there will be developments of higherfrequency transducers that will permit visualization of more superficial details, and also there will be more advances in the ultrasonographic contrast medium area, with longer media life contrasts that can have potential uses in the functions or physiopathology of the lesions and the differentiation of benign from malign tumors, because the neoplastic lesions have different vessels disposition and wash-in and wash-out curves in preliminary studies.6
Ultrasound Imaging of Subcutaneous Tissue and Adjacent Structures
60.14 INDICATIONS OF HIGHRESOLUTION ULTRASOUND IN THE SUBCUTANEOUS AND ADJACENT STRUCTURES 1. 2. 3. 4. 5. 6.
7. 8. 9. 10. 11. 12.
To determine if a lesion is real or not To find out if it is solid or cystic Exact anatomical location Depth Extension Vascularity inside or in the periphery of the lesion, what kind of vessels (arterial, venous), arteriovenous shunts, the diameter, and anatomical location of the vessels To look for associated or nonsuspected lesions To monitor treatment or the evolution of a lesion during a period in an objective way Choice of the best biopsy or punction site Perform a real-time biopsy or punction under ultrasound guidance Dynamic studies of a lesion or a normal structure Comparative studies of different sites of compromise or abnormal/normal zones
REFERENCES 1. Bakhach S, Grenier N, Berge J, et al., Color Doppler sonography of superficial capillary hemangiomas, J Radiol, 82: 1614–1619, 2001. 2. Bessoud B, Lassaud N, Koscielny S, et al., High frequency sonography and color Doppler in the management of pigmented skin lesions, Ultrasound Med Biol, 29: 875–879, 2003. 3. Bossi MC, Sanvito S, Lovati E, et al., Role of high resolution color: Doppler US of the sentinel node in patients with stage I melanoma, Radiol Med (Torino), 102: 357–362, 2001. 4. Boyse TD, Fessell DP, et al., US of soft–tissue foreign bodies and associated complications with surgical correlation, Radiographics, 21: 1251–1256, 2001. 5. Cosnes A, Anglade MC, et al., Thirteen-megahertz ultrasound probe: its role in diagnosing localized scleroderma, Br J Dermatol, 148: 724–729, 2003. 6. De Marchi A, De Petro P, Linari, A, et al., A preliminary experience in the study of soft tissue superficial masses. Color Doppler US and wash-in and wash-out curves with contrast media compared to histological results, Radiol Med (Torino), 104: 451–458, 2002. 7. Etemad-Rezai R, Peck DJ, Ultrasound guided thrombin injection of femoral artery pseudoaneurysms, Can Assoc Radiol J, 54: 118–120, 2003. 8. Feleppa EJ, Alam SK, Deng CX, Emerging ultrasound technologies for early markers of disease, Dis Markers, 18: 249–268, 2002.
529
9. Fornage B, Musculoskeletal Ultrasound, 1st ed., Churchill Livingston 1995. 10. Foster FS, Burns PN, Simpson DH, Ultrasound for the visualization and quantification of tumor microcirculation, Cancer Metastasis Rev, 190: 131–138, 200 (1-2). 11. Gniadecka M, Effects of ageing on dermal echogenicity, Skin Res, 7: 204–207, 2001. 12. Gniadecka M, Localization of dermal edema in lipodermatosclerosis, lymphedema and cardiac insufficiency. High frequency ultrasound examination of intradermal echogenicity, J Am Acad Dermatol, 35: 37–41, 1996. 13. Goldberg B, Deane CR, An Atlas of Ultrasound Color Flow Imaging, 1st ed., Martin Dunitz Ltd., 1997. 14. Gritzmann N, Hollerweger A, et al., Sonography of soft tissue masses of the neck, J Clin Ultrasound, 30(6): 356–373, 2002. 15. Harland CC, Kale SG, Jackson P, et al., Differentiation of common benign pigmented skin lesions from melanoma by high resolution ultrasound, Br Dermatol, 143: 281–289, 2000. 16. Harvey C, Pilcher J, et al., Advances in ultrasound, Clin Radiol, 57: 157–177, 2002. 17. Hughes J, Lam A, et al., Use of ultrasonography in the diagnosis of childhood pilomatrixoma, Pediatr Dermatol, 16: 341–344, 1999. 18. Iannicelli E, Rossi G, et al., Integrated imaging in peripheral nerve lesions in type 1 neurofibromatosis, Radiol Med (Torino), 103: 332–343, 2002. 19. Rajeev J, Bandhu S, Sukhpal S, et al., Sonographically guided percutaneous sclerosis using 1% polidocanol in the treatment of vascular malformations, J Clin Ultrasound, 30: 416–423, 2002. 20. Jemec GB, Hidradenitis suppurative, J Cutan Med Surg, 7: 47–56, 2003. 21. Jemec GB, Serup ultrasound structure of the human nail plate, J Arch Dermatol, 125: 643–646, 1989. 22. Jemec GB, Gniadecka M, Ulrich J, Ultrasound in dermatology. Part I. High frequency ultrasound, Eur J Dermatol, 10: 492–497, 2000. 23. Kahle B, Hoffend J, et al., Preoperative ultrasonographic identification of sentinel lymph node in patients with malignant melanoma, Cancer, 97(8): 1947–1954, 2003. 24. Kirsner RS, Pardes JB, et al., The clinical spectrum of lipodermatosclerosis, J Am Acad Dermatol, 28: 623–627, 1993. 25. Kransdorf M, Murphey M, Bralow L, Imaging of Soft Tissue Tumors, 1st ed., W.B. Saunders, 1997. 26. Lin KL, Wang HS, Chou ML, et al., Sonography for detection of spinal dermal sinus tracts, J Ultrasound Med, 21: 903–907, 2002. 27. Marcello R, Coertese F, et al., A new interventional procedure in the treatment of iatrogenic pseudoaneurysm, Radiol Med (Torino), 105: 63–68, 2003. 28. Martinoli C, Bianchi S, Derchi LE, Ultrasonography of peripheral nerves, Seminars in Ultrasound, CT, and MRI, 21: 205–213, 2000. 29. Martinoli C, Bianchi S, et al., Ultrasound of tendons and nerves, Eur Radiol, 12: 44–55, 2002. 30. Mutasim DF, Autoimmune bullous dermatoses in the elderly, Drugs Aging, 20: 663–681, 2003.
530
Handbook of Non-Invasive Methods and the Skin, Second Edition
31. Olsen DM, Rodriguez JA, et al., A prospective study of ultrasound scan-guide thrombin injection of femoral pseudoaneurysm: a trend toward minimal medication, J Vasc Surg 36: 779–782, 2002. 32. Peterson J, Kransdorf M, et al., Malignant fatty tumors: classification, clinical course, imaging appearance and treatment, Skel Radiol, 32: 493–503, 2003. 33. Quinn TJ, Jacobson JA, Craig JG, Van Holsbeeck MT, Sonography of Morton’s neuromas, AJR Am J Roentgenol, 174: 1723–1728, 2000. 34. Rallan D, Harland CC, Ultrasound in dermatology basic priciples and applications, Clin Exp Dermatol, 28: 632–638, 2003 35. Rand T, Ritschl P, Trattnig S, Breitenseher M, Imhof H, Resnick D, Zembsch A, Bindeus T, Kaderk M, Spitz S, Imaging of Bone and Soft Tissue Tumors, 1st ed., Springer, 2001. 36. Rettenbacher T, Sogner P, Springer P, et al., Schwanoma of the brachial plexus: cross sectional imaging diagnosis using CT, sonography and MR imaging, Eur Radiol, 13: 1872–1875, 2003. 37. Rumack C, Wilson S, Charboneau J.W, Diagnostic Ultrasound, 3rd ed., Mosby Elsevier, 2004. 38. Schmid-Wendtner MH, Paerscheke G, Baumert J, et al., Value of ultrasonography compared with physical examination for the detection of locoregional metastasis in patients with cutaneous melanoma, Melanoma Res, 13: 183–188, 2003. 39. Simanowski JH, Ultrasound diagnosis of venous thrombosis of the leg, Orthopade, 31: 314–316, 2002.
40. Solivetti FM, Thorel MF, Di Luca S, Role of high definition and high frequency ultrasonography in determining tumor thickness in cutaneous malignant melanoma, Radiol Med (Torino), 96: 558–561, 1998 41. Ulrich J, Voigt C, Ultrasound in dermatology. Part II. Ultrasound of regional lymphonode basins and subcutaneous tumors, Eur. J Dermatol, 11: 73–79, 2001. 42. Uren RF, Howman-Gilkes R, Thompson JF, et al., High resolution ultrasound to diagnose melanoma metastasis in patients with clinically palpable lymph nodes, Australas Radiol, 43: 148–152, 1999. 43. Van Holsbeeck M, Introcaso J, Musculoskeletal Ultrasound, 2nd ed., C.V. Mosby, 2001. 44. Voigt C, Schoengen A, et al., The role of ultrasound in detection and management of regional disease in melanoma patients, Semin Oncol, 29: 353–360, 2002. 45. Yanik B, Conkbayir I, et al., Imaging findings in Mondor’s disease, J Clin Ultrasound 31: 103–107, 2003. 46. Zagzebski J, Essentials of Ultrasound Physics, 1st ed., C.V. Mosby, 1996. 47. Nazarian et al., Malignant melanoma: impact of superficial US on management, Radiology, 199(1):273–277, 1996. 48. Blum A, et al., Ultrasound examination of regional lymph nodes significantly improves early detection of locoregional metastasis during the followup of patients with cutaneous melanoma: Results of a prospective study of 1288 patients, Cancer, 88(11): 2534–2539, 2000.
Resonance Spectroscopy of 61 Magnetic the Skin A. Zemtsov University Dermatology Center, P.C., Muncie, Indiana
CONTENTS 61.1 Introduction............................................................................................................................................................531 61.2 Biochemical Background ......................................................................................................................................531 61.3 Object and Methods...............................................................................................................................................532 61.4 Sources of Error.....................................................................................................................................................534 61.5 Recommendations..................................................................................................................................................534 References .......................................................................................................................................................................534
61.1 INTRODUCTION There are two magnetic resonance (MR) techniques used today in clinical medicine: magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS). MRI scans are two- and three-dimensional radiological pictures that are similar to radiological images produced by computer tomography (CT). MRS, on the other hand, is a nondestructive technique to noninvasively study and monitor metabolism in the human body. Human in vivo MRS and MRI utilize identical MR scanners; however, the software is different. Essentially, every standard hospitalbased MR scanner, with the purchase of additional software, can perform human in vivo MRS examinations. In 1938, Professor Rabi described the nuclear magnetic resonance phenomenon.1 Since then, MRI has been extensively used in physics, chemistry, and, in the past 10 years, in biomedical science. For their contribution to this field, Rabi (in 1944) and, later, Bloch and Purcell (in 1952) and Paul Lauterbur and Sir Peter Mansfield (in 2003) were awarded Nobel Prizes. In MRI, two- and three-dimensional images are obtained from the signal intensity of the hydrogen atom, while in MRS a spectrum is obtained that reflects the relative concentration of various metabolites. Depending on the frequency chosen, MRS can detect 31P, 1H, 13C, 19F, and other molecular concentrations. At the present time, almost all published human in vivo skin MRS experiments utilize the 31P MRS data. Zemtsov and others2–4 wrote editorials and conducted detailed studies on the subject of human in vivo skin 31P MRS, and the reader is encouraged to review these papers for additional information. Querleux et al.5,6 performed the first in vivo
skin proton MRS scans and made some data interpretation. However, skin human proton MRS has not been used yet in any clinical or cosmetic industry setting. Therefore, this chapter will be devoted to skin 31P MRS and its application in clinical medicine, cosmetic science, and basic science research.
61.2 BIOCHEMICAL BACKGROUND Adenosine triphosphate (ATP) molecules provide energy for all cellular structures. Phosphocreatine can be viewed as an energy reserve of molecules that replenish ATP reserves during periods of cellular ischemia by donating a phosphate group to ADP, as indicated by the equation below: MgADP– + PCr+2 + H+ MgATP–2 + Cr This process is regulated by cytoplasmic and mitochondrial creatine kinase (CK) enzymes. The reader interested in more detailed information is invited to review articles published by our research group as well as others.2–4,7–14 In addition to brain, cardiac, and skeletal muscle, phospocreatine (PCr) and CK are present in numerous other tissue, including skin. Lin and coworkers15 detected PCr using in vitro MRS in frog skin. Later, Cuono et al.16 found PCr in human cadaveric skin by in vitro 31P MRS experiments. Klein and Gourley17 performed in vivo animal 31P MRS and observed PCr in rodent skin. However, according to the groups of Bohning and Cowie,3,4 Zemtsov’s group was the first to perform human in vivo 31P MRS 531
532
Handbook of Non-Invasive Methods and the Skin, Second Edition
30
30
25
25 B
B
20
20
15
15
10
10
5
F
A
EDC
0
0
−5
−5 20
10
0
−10
F
5
−20
FIGURE 61.1 31P MRS of normal skin. (A) Beta, alpha, and gamma peaks of ATP molecule. (B) PCr (phosphocreatine). (C) Phosphadiester (PDE). (D) Inorganic phosphate (Pi). (E) Phosphomonoester (PME).
observations. Furthermore, Zemtsov’s group correlated 31P MRS observations by measuring the concentration of PCr and ATP in human samples by high-pressure liquid chromatography (HPLC)7 and identified and characterized CK enzymes in normal and diseased skin.2,7,8 In summary, 31P MRS measures the concentration of adenosine triphosphate (ATP), phospocreatine (PCr), inorganic phosphate (Pi), phosphomonoesters (PME), and phosphodiesters (PDE). In addition, from 31P MRS data, intracytoplasmic pH can be calculated. Furthermore, the relative amounts and ratios of PCr, Pi, and ATP are sensitive indicators of tissue oxygenation, while the relative amounts and ratios of PME and PDE reflect the cellular turnover rate.2–4,9–13 Finally, Wright et al.12 demonstrated that intracellular Mg2+ concentration can be calculated from 31P MRS data.
61.3 OBJECT AND METHODS The use of 31P MRS techniques in dermatology can be subdivided into at least five areas: 1. Plastic and dermatologic surgery: Studying pathophysiology and metabolism of skin tumors and their response to radiation therapy,
20
FIGURE 61.2 peak (E).
A
EDC
10
0
−10
−20
31
P MRS of active skin psoriasis. Note large PME
interferon, and topical chemotherapy. 31P MRS is potentially useful in developing and assessing new medical therapies to treat skin cancer, such as topical 5-Fu, imiquimod (Aldara), diclofenac sodium (Solaraze), interferons, and photodynamic therapy.2 Recently, Korean radiologists used 31P MRS to analyze and differentiate between benign and malignant breast tumors and normal breast tissue.18 Skin and breast tissue, both histologically and biochemically, are very similar and have identical 31PMR spectra.18 2. Clinical dermatology: Monitoring disease activity and comparative assessment of the effectiveness of various treatment modalities in psoriasis,14 leg ulcers,19 and mycosis fungoides.2 MRS clearly has the advantage of providing unbiased placebo-free biochemical data that shows changes (if any) that topical or systemic treatment induces in diseased skin. Zemtsov et al.14 showed clearly how 31P MRS can monitor this activity in psoriasis caused by systemic administration of methotrexate and application of topical steroids (Figure 61.2 and Figure 61.3). Furthermore, biochemical changes observed appear before any clinical improvement.
Magnetic Resonance Spectroscopy of the Skin
533
30
2. 3.
25 B 20
4.
15
5. 6.
10
7. 5
F
A
EDC
8.
0
9. −5 20
10
0
−10
−20
FIGURE 61.3 The same patient as in Figure 61.2 after systemic methotrexate administration. Note that peak E (PME) has almost completely disappeared.
3. Cosmetic science: By studying the effects of UVA radiation on normal skin, one can develop better, and compare the efficacy of, various UVA-blocking sunscreens.20 4. Dermatopharmacology: New methods for topical steroid bioequivalence to replace Stoughton–McKenzie vasoconstriction assay.21 To study percutaneous absorption rate and bioavailability of all topically applied drugs.2 5. Other applications: Studying biochemical changes if any caused by topical applications of rejuvenation creams such as retinoids, alphahydroxyl, vitamin C, and others (there are no publications on this subject). To monitor biochemical changes, if any, in cellulite.6 Below is a summary of the biochemical data of human in vivo skin 31P MRS and high-pressure liquid chromatography that can be used by future investigators. 1. Skin intracytoplasmic pH is alkaline: pH = 7.2 (Zemtsov2), pH = 7.46 ± 0.1 (Bohning et al.3), and pH 7.39 ± 0.08 (Cowie et al.4). Please note
that in most human tissue intracytoplasmic pH is neutral (pH = 7.0).2–4 PCr and CK enzymes are present exclusively in keratinocytes of the epidermis.2,7,8 CK isoenzyme distribution depends on a disease state.8 CK MM and CK BB are the major CK isoenzymes in normal skin.8 PCr, PME, and PDE are highly elevated in psoriatic tissue and CK BB is the major isoenzyme.2,8,14 Human skin has a relatively large amount of PDE, PME, and Pi.2–4,9 Human skin has a relatively high concentration of Mg2+ ions.12 There is a lower PDE concentration in black skin than in Caucasian skin.12 UVA radiation induces production of PME and causes depletion of PCr and ATP in Caucasian skin (Zemtsov and Declercq et al.20,22). Topical application of corticosteroids produces a nearly identical result based on 31PMRS in vitro data.21 As expected, healing ulcers have acidic pH and poor oxygenation as reflected by lower PCr, ATP, PCr-ATP, and PCr-Pi levels and ratios (groups of Zemtsov and Smith13,19). Hyperbaric oxygenation improves the above-mentioned parameters.19 Zemtsov and Dixon19 reported that healing ulcers have pH values in the alkali range.
Figure 61.1 to Figure 61.3 are clinical examples of human in vivo skin 31P MRS. As mentioned above, 31P MRS provides information about tissue bioenergetics, tissue turnover rate, intercellular pH, and Mg2+ ion concentration. Figure 61.1 is the 31P MRS spectrum of normal skin. One sees beta, alpha, and gamma peaks of the ATP molecule (A), phosphocreatine (B), phosphodiester (C), inorganic phosphate (D), and phosphomonoesters (E) (barely detectable in normal skin). During skin ischemia, ATP and phosphocreatine peaks disappear and the inorganic phosphate peak increases. Furthermore, intracellular pH becomes more acidic; intracellular pH is calculated on the basis of a chemical shift of inorganic phosphate relative to PCr. These changes are observed in compromised dying skin flaps and skin grafts and nonhealing chronic leg ulcers. UVA radiation causes production of phosphomonoesters and decreases in concentration of phosphocreatine (and according to Zemtsov, also decreases in concentration of ATP).20,22 However, pH remains unchanged. Phosphomonoester concentration reflects phospholipids’ biosynthesis. On the other hand, phosphodiesters are the product of membrane catabolism. Therefore, the phosphomonoester-to-phosphodiester ratio is increased in disease
534
Handbook of Non-Invasive Methods and the Skin, Second Edition
states, characterized by increased production of new cells, such as cancer and psoriasis.2–4,22 In Figure 61.2, 31P MRS skin with active psoriasis, one clearly observes a phosphomonoester peak (peak E). After systemic methotrexate administration, peak E disappears, reflecting decreased production of keratinocyte cells (Figure 61.3). Furthermore, according to Zemtsov et al.,14 clinical changes are observed days before any clinical improvement occurs. As mentioned above, there is very little, if any, data concerning 31P MRS changes occurring in skin tumors either at the baseline or associated with treatment modalities. Furthermore, there is no data whatsoever on the biochemical changes, if any, associated with topical application of rejuvenation creams. Clearly, the field of human in vivo skin 31P MRS is in its infancy.
61.4 SOURCES OF ERROR The main technical challenge of human in vivo skin is to rapidly obtain high signal-to-noise data without underlying skeletal contamination. A number of groups have been experimenting with various coils.4,9,13,22 The advantage of the Zemtsov–Declercq–Cowie approach is relatively good signal-to-noise ratio and short acquisition times. Wright, Bohning, and Spicer describe a technique that was free of skeletal muscle contamination.3,12 Their 31P MRS skin data were obtained utilizing a slotted crossover surface coil, a complex calibration phantom, and a two-power depth resolution technique (substraction algorithm technique). The obvious disadvantage of the Wright–Bohning–Spicer technique is the long acquisition time (4.5 hours), which makes this technique impractical in the clinical setting. The computer function involved in MRS spectrum reconstruction is far more complex than in any other method discussed in this book. For example, in the highfrequency digital ultrasound technique for image reconstruction the computer merely performs digital-to-analog conversion, but in the MRS techniques the computer also controls timing and duration of radio frequency pulses and the timing and intensity of magnetic field operation. Obtaining reliable, reproducible MRS data requires cooperation between the clinician or cosmetic scientist and medical physicist who specifically trained in the field of magnetic resonance spectroscopy, or at least with the radiologist who has had additional training and expertise in the field of magnetic resonance spectroscopy.
61.5 RECOMMENDATIONS The MRS of skin is still in its early pioneering stages. There are very few research groups in the world working in this area. As mentioned above, producing good, reliable
in vivo human skin MRS data requires cooperation between the physician/cosmetic industry scientists and medical physicists or radiologists with expertise in magnetic resonance spectroscopy. Furthermore, to conduct MRS experiments a research group has to be well funded. As mentioned above, MR scanning for both clinical images and spectroscopy is the same, only the software is different. At least in the U.S., since MR scanners generate clinical income, most radiology departments charge a high facility use fee. Moreover, many other clinicians, especially neurologists and cardiologists, are very interested in MRS, and at large academic centers, there is often a waiting time to reserve time on the MR scanner for in vivo human spectroscopic examination. Fortunately, most research groups working in the field of skin in vivo MRS are well funded and associated with university-based radiology departments. Magnetic resonance techniques have revolutionized the fields of physics, chemistry, and medicine; five individuals have received Nobel Prizes for their contributions to this area. Skin, brain, cardiac, and other organ in vivo MRS techniques are still in their infancy. I expect that in the future, MRS will help us better understand the pathophysiology of skin cancer, psoriasis, mycosis fungoides, and leg ulcers, and will be used in the assessment and response of these skin conditions to various new therapeutic modalities. Skin MRS has numerous and obvious applications in cosmetic science and dermatopharmacology, and one day it might be used clinically by plastic and dermatologic surgeons to assess viability of skin flaps and grafts. Meanwhile, I personally find it gratifying that the results of my original 31P MRS and high-pressure liquid chromatography studies were confirmed by the groups of Bohning, Cowie, and Smith,3,4,7–9,13 the results of a UVA study by Declercq,20,22 and leg ulcer data19 by the Smith group.13,19
REFERENCES 1. Rabi II, Zacharias JR, Millman S, Kursch P. A new method of measuring nuclear magnetic moment. Phys Rev 53: 318, 1938. 2. Zemtsov A. Human in vivo 31P magnetic resonance spectroscopy. Skin Res Technol 3: 85–87, 1997. 3. Bohning DE, Wright AC, Spicer KM. In vivo phosphorus spectroscopy of human skin. Magn Res Med 35: 186–193, 1996. 4. Cowie AG, Bastin ME, Manners DN, Hands LJ, Styles P, Radda GK. 31P NMR studies of human skin using a modified zig-zag surface coil. NMR Biomed 9: 195–200, 1996. 5. Querleaux B, Jolivet O, Bittoun J. In vivo proton magnetic resonance spectroscopy in human skin. Curr Probl Dermatol 26: 12–19, 1998.
Magnetic Resonance Spectroscopy of the Skin
6. Querleux B, Cornillon C, Jolivet O, Bittoun J. Anatomy and physiology of subcutaneous adipose tissue by in vivo magnetic resonance imaging and spectroscopy: relationship with sex and presence of cellulite. Skin Res Technol 8: 118–124, 2002. 7. Zemtsov A, Cameron GS, Stadig B, Martin J. Measurements of phosphocreatine in cutaneous tissue by high pressure liquid chromatography. Am J Med Sci 305: 8–11, 1993. 8. Zemtsov A, Cameron GS, Bradley CA, Montalvo-Lugo V, Mattioli F. Identification and activity of cytosol creatine kinase enzymes in normal and diseased skin. Am J Med Sci 308: 365–369, 1994. 9. Zemtsov A, Ng TC, Xue M. Human in vivo 31P spectroscopy of skin: potentially a powerful tool for noninvasive study of metabolism in a cutaneous tissue. J Dermatol Surg Oncol 15: 1207–1211, 1989. 10. Zemtsov A. 31P magnetic resonance spectroscopy to study noninvasively metabolism in cutaneous tissue. Plast Reconstr Surg 92: 1411–1413, 1993. 11. Zemtsov A. Letter to the editor. Magn Reson Med 27: 198–199, 1992. 12. Wright AC, Bohning DE, Spicer KM. Phosphorous metabolites in human skin and muscle obtained by phosphorous 31 magnetic resonance spectroscopy. Skin Res Technol 3: 66–72, 1996. 13. Smith DG, Mills W, Steen GR, Williams D. Levels of high energy phosphate in the dorsal skin of the foot in normal and diabetic adults: the role of 31P magnetic resonance spectroscopy and direct quantification with high pressure liquid chromatography. Foot Ankle Int 20: 258–262, 1999. 14. Zemtsov A, Dixon L, Cameron G. Human in vivo phosphorous 31 magnetic resonance spectroscopy of psoriasis. A noninvasive tool to monitor response to treatment and to study pathophysiology of the disease. J Am Acad Dermatol 30: 959–965, 1994.
535
15. Lin LE, Shporer M, Civan MM. 31P nuclear magnetic resonance analysis of perfused single frog skin. Am J Physiol 248: 177–180, 1985. 16. Cuono CB, Armitage IM, Marquetand R, Chapo GA. Nuclear magnetic resonance spectroscopy of skin: predictive correlates for clinical application. Plast Reconstr Surg 81: 1–11, 1988. 17. Klein HW, Gourley IM. Use of magnetic resonance spectroscopy in the evaluation of skin flap circulation. Ann Plast Surg 20: 547–551, 1988. 18. Park JM, Park JH. Human in vivo 31P MR spectroscopy of benign and malignant breast tumors. Kor J Radiol 2: 80–86, 2001. 19. Zemtsov A, Dixon L. Monitoring wound healing in venous stasis leg ulcers by in vivo 31P magnetic resonance spectroscopy. Skin Res Technol 1: 36–40, 1995. 20. Zemtsov A. Measurement of suberythema UVA radiation effects on skin by 31P magnetic resonance spectroscopy. Photodermatol Photoimmunol Photomed 13: 24–26, 1997. 21. Collier SW, Sardon S, Ruiz-Cabello J, Johnson WA, Colan JS, Schwartz SI. Measurements of pharmacodynamic effects of dexamethasone on epidermis by phosphorous nuclear magnetic resonance spectroscopy in vitro. J Pharm Sci 83: 1339–1344, 1994. 22. Declercq L, Perin F, Vial F, Savard S, Pentitcollin B, Beau P, Collins D, Mammone M, Marenus K, Mars K. Age-dependent response of energy metabolism of human skin to UVA exposure: an in vivo study by 31P nuclear magnetic resonance spectroscopy. Skin Res Technol 8: 125–132, 2002.
Resonance Microscopy of 62 Magnetic Normal Skin and Pigmented Skin Tumors Stephan El Gammal,1 Roland Hartwig,2 Sitke Aygen,4 T. Bauermann,4 Claudia El Gammal,1 and P. Altmeyer3 1
Dermatological Clinic of the Hospital Bethesda, Freudenberg, Germany
2
Dermatological Practice, Wuppertal, Germany
3
Dermatological Clinic of the Ruhr-University Bochum, Bochum, Germany
4
Institute for “Zentrale Analytik und Strukturanalyse” of the University Witten/Herdecke, Witten, Germany
CONTENTS 62.1 Introduction............................................................................................................................................................537 62.1.1 Non-Invasive Skin Imaging Methods........................................................................................................538 62.1.2 MR Principle..............................................................................................................................................538 62.2 Materials and Methods ..........................................................................................................................................539 62.2.1 MR Microscopy .........................................................................................................................................539 62.2.2 Spin–Echo Sequence .................................................................................................................................539 62.2.3 Longitudinal Relaxation Time (T1) ..........................................................................................................540 62.2.4 Transversal Relaxation Time (T2).............................................................................................................541 62.2.5 Voxel Reconstruction.................................................................................................................................541 62.2.6 Statistics .....................................................................................................................................................541 62.3 Results....................................................................................................................................................................541 62.3.1 Basal Cell Carcinoma ................................................................................................................................541 62.3.2 Nevocellular Nevus....................................................................................................................................542 62.3.3 Malignant Melanoma.................................................................................................................................543 62.3.4 T1 and T2 Relaxation Values ....................................................................................................................544 62.4 Discussion ..............................................................................................................................................................544 62.5 Outlook ..................................................................................................................................................................548 References .......................................................................................................................................................................548
62.1 INTRODUCTION It is a characteristic feature of the dermatological specialty that the whole spectrum of diseases, from slight irritations to malignant transformations, lies directly before ones eyes. Clinical examination by inspection and palpation thus plays a crucial role in dermatological diagnosis. Moreover, in any doubtful case, the easy accessibility of the skin allows one to take biopsies without much
discomfort to the patient. Many diagnoses can be confirmed by their typical histological picture. The diagnosis of a skin disease is therefore usually based on clinical and histological criteria. In the diagnosis of malignant melanoma, where the prognostic classification of the tumor depends on the level of invasion into the dermis [Clark et al., 1969] and the vertical tumor thickness [Breslow, 1970, 1975], preoperative evaluation of these factors is extremely helpful to 537
538
Handbook of Non-Invasive Methods and the Skin, Second Edition
determine the therapeutic excision margin and judge the risk of metastasis. Pre-therapeutic information would also be valuable in differential diagnosis improving diagnostic accuracy.
62.1.1 NON-INVASIVE SKIN IMAGING METHODS During the past several years, the development of noninvasive techniques has opened new possibilities of preoperative diagnosis. Lateral and axial tumor expansion can be determined by means of high resolution 20-MHz [Altmeyer et al., 1990; Hoffmann et al., 1990] or 50-MHz sonography [El Gammal et al., 1990]. However, this method does not always allow exact definition of the tumor margins [Altmeyer et al., 1990; El Gammal et al., 1990, 1992b; Hoffmann et al., 1990]. Our investigations show that it is especially difficult to differentiate the tumor from the concomitant inflammatory infiltrate [El Gammal et al., 1990, 1992b, 1992c, 1997]. Attempts to evaluate sonographic pictures with image analytical methods could not solve this problem [El Gammal, 1990, 1992c, 1997]. Other non-invasive imaging techniques such as computer tomography and xeroradiography require x-ray radiation and are therefore less suitable for dermatological diagnosis considering the benefit–risk ratio for the patient. A non-invasive method that has become increasingly important in the medical field throughout the past years is the magnetic resonance technique. The investigation and description of the magnetic properties of the atomic nuclei laid the base for this method. In the years 1924 through 1927 the physicist Pauli analyzed the detailed structure of the atomic nuclei (he described the anormal Zeeman effect and the exclusion principle) [Richter 1976]. Twenty years later Bloch [1946] and Purcell et al. [1946] independently discovered the nuclear spin resonance signal. The nuclear spin resonance signal provides information about the chemical composition of the investigated object. The idea to use nuclear magnetic resonance for investigations of human tissue has its origin in studies by Jackson in 1967 [Wehli, 1988]. A milestone in the history of nuclear magnetic resonance was the publication of a section of waterfilled capillaries by Lauterbur in 1973. A new imaging technique (magnetic resonance imaging, MRI) had thus been created which Lauterbur [1974] named zeugmatography and from which today’s MR tomography was developed. In MRI, different sequence protocols are used to get images of two-dimensional sections. For spin-echo images a non-selective 90° rectangular pulse of 3 to 20 μs (hardor HF-pulse) and a magnetic gradient field are combined with a slice-selective pulse of 4000 μs (so-called soft pulse) in such a way that the tissue characteristics can be assessed non-destructively in very small volumes. With the current MR tomographs for clinical applications, which use a homogeneous magnetic field between 0.23 and 2.4 Tesla and work with magnetic gradient fields of
up to 1 Gauss/cm to assess internal organs or bones, voxel (volume element) resolutions of 5 × 5 × 5 mm3 at best can be obtained [Kuhn 1990]. In the past decades MRI systems have become available which work with very high magnetic ground and gradient fields and thus allow examination of the fine structure of tissues at microscopic [Kuhn, 1990] and cellular levels [Aguayo et al., 1986]. Of particular diagnostic value is the differentiation of various tissues and pathological alterations using localized T1- and T2-measurements. Since the construction of this strong homogeneous magnetic ground field and high gradient field is technically difficult for greater probe volumes, it is possible to study small living animals in vivo with our equipment, whereas human skin must be investigated at present ex vivo [El Gammal et al., 1992a, 1993, 1995, 1996, 1997].
62.1.2 MR PRINCIPLE When entering a specimen in a homogeneous magnetic ground field, the spins of the atomic nuclei within the specimen statistically rearrange in the magnetic field in such a way that the sum vector of all spins points to the direction of the magnetization. The length of the sum vector is proportional to the strength of the magnetic field. When a second magnetic field alternating at a certain frequency is applied at a 90° angle to this ground field, the sum vector (overall magnetization vector) changes its spatial orientation. This temporal change of the magnetic sum vector perpendicular to the ground field induces a high frequency alternating current in the coil used for signal detection. The resulting curve is characteristic for the composition of the specimen. The energy previously absorbed by the specimen is now emitted, causing an exponential decrease of the received nuclear resonance signal, while the spins of the atomic nuclei relax. This relaxation process can be divided into two components, the longitudinal and the transversal relaxation. The longitudinal relaxation describes the process in which the momentum of all nuclei due to the interaction with their surrounding atoms and molecules (called grid) return to their original position parallel to the field. This phenomenon is also called spin–grid relaxation. The time it takes for 63% of the momentum of all nuclei to reach their original state in relation to the magnetic ground field is referred to as longitudinal relaxation time T1 [Longmore, 1989]. When the high frequency impulse is switched off, all nuclei first point to the same direction and process uniformly. The precession movements are, however, increasingly influenced by the inhomogeneity of the applied magnetic ground field on the one hand, and the interactions between neighboring nuclear moments on the other hand. This process is named spin–spin interaction or cross
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
Homogenous magnetic field 9.6 Tesla
MRSpectroscopy MRImaging (microscopy)
HF-sender HF-receiver 400 MHz Computer control unit
539
After MRI, the tissue was fixed in formalin for another 12 to 24 hours and paraffin-embedded. In order to obtain histological sections exactly correlating to the MRIs, the paraffin blocks were cut in the same plane as the MRIs were taken. In steps of 150 μm, several 7-μm sections were cut, stained with H&E and examined under a light microscope. This procedure allowed the choice between two sets of 150-μm distant histological sections for correlation with each 300-μm thick MRI slice.
62.2.1 MR MICROSCOPY Magnetic gradient field (X, Y, Z) 75 Gauss/cm
FIGURE 62.1 MR spectroscopy and MR microscopy. For MR microscopy, additional gradient fields are applied to obtain a spin–echo signal from small voxels (volume elements).
relaxation, the according time constant being the transversal relaxation time T2. With MR spectroscopy the entire material is measured as an integral signal assuming an equal distribution of all substances within the specimen. The use of additional magnetic gradient fields makes it possible to measure selectively within small volumes of the specimen as well. With MRI the specimen is analyzed in slices and the qualities of the received signal within the voxel (volume element) are presented as gray level modulation of one pixel (picture element) in the two-dimensional image (Figure 62.1).
62.2 MATERIALS AND METHODS Twelve nevocellular nevi, 20 basal cell carcinomas, 8 malignant melanomas, and 8 seborrheic keratoses were investigated. They were excised for therapeutic reasons from patients of the dermatological clinic of the Ruhr-University Bochum. All patients gave informed consent. Ten specimens of normal skin from various body regions were examined; the tissue was obtained from healthy skin within the security margin of excised high-risk melanomas. First the tissue was thoroughly rinsed in 0.9% NaCl solution in order to remove blood remnants, which influence the magnetic resonance signal due to the ferromagnetic effect of the hemoglobin [Wehrli, 1988]. After fixation in formalin 5% for 30 minutes, the tissue blocks were patted dry with gauze and placed in the center of a 10mm proton RF coil. Examination was performed with an MR spectroscopy unit (AM 400 WB NMR, Bruker GmbH, Rheinstetten, FRG) equipped with an advanced microimaging accessory. Each tissue block was analyzed in slices. A single slice was 300 μm thick (slice thickness = selective pulse width/gyromagnetic ratio × gradient strength).
The MR Microscopy Unit Bruker AM 400 WB NMR works with a helium-cooled supra-conducting magnetic coil of 9.4 Tesla and three orthogonal gradient coils that produce linear gradient fields of up to 75 Gauss/cm (Figure 62.1). In a 1-cm tissue block, a resolution of up to 20 μm can theoretically be achieved in the x/y plane. We used the following imaging techniques for our investigations [Kuhn, 1990]: • • •
spin–echo sequence for single and multi slices inversion–recovery sequence to determine the longitudinal relaxation time T1 multi-echo sequence to determine the transversal relaxation time T2
62.2.2 SPIN–ECHO SEQUENCE Figure 62.2 shows the sequence protocol to record spin–echo images. “x”, “y”, and “z” designate the gradient fields. The emitted high frequency pulse is referred to as Tx, while Rx represents the position of the so-called interval of signal reception in the time sequence. In the beginning, the whole specimen is stimulated by a non-selective, rectangular (90°) hard pulse. Switching on the “y”
X Y Z
90° pulse 180° pulse
Tx Rx
FIGURE 62.2 Sequence protocal of the single-slice spin–echo procedure. The non-selective 90° pulse is a rectangular “hard” pulse, the slice-selective 180° pulse is displayed as a symbol standing for different “soft”-pulse shapes (e.g., Hermite or Gauss). X, Y, and Z refer to the gradient fields. A high-frequency signal is emitted (Tx) and received (Rx) by the same coil. The strength of the Z-gradient determines the slice thickness. The axial resolution is proportional to the strength of the X-gradient.
540
Handbook of Non-Invasive Methods and the Skin, Second Edition
90° pulse
for several slices (multi-slice imaging with 4, 8, 16, or 32 slices). For the latter technique, the stimulation sequence with slice-selective pulses has been combined in such a way that overlap effects are kept to a minimum. While one already stimulated slice relaxes, another is stimulated. This allows a more efficient use of the extinction time periods that occur because the nuclei have to relax completely in longitudinal magnetization prior to stimulation. For multi-slice imaging with 8 slices, for example, the sequence 1, 3, 5, 7, 2, 4, 6, 8 is chosen in order to reduce extinctions due to overlapping of the slice profiles.
180° pulse
z'
Mo
y' x'
a
c
b
d
62.2.3 LONGITUDINAL RELAXATION TIME (T1)
FIGURE 62.3 Inversion–recovery sequence to determine T1 tissue constants: (a) balance, (b) inversion, (c) recovery, and (d) measurement. The time elapse between the 180° pulse and the 90° pulse is called inversion time. The 90° pulse turns the magnetic sum vector in transversal direction to measure its amplitude (d).
The relaxation time T1 characterizes the speed with which the longitudinal magnetization returns to its resting value after stimulation with a 180° pulse. Our region of interest was 0.01 mm. We determined the longitudinal relaxation time by the following measurement techniques. The inversion–recovery sequence uses a 180° pulse to orient the magnetization sum vector in the z direction. To measure the amplitude of this fading vector, a 90° pulse was emitted after a defined time (inversion time) turning the magnetization sum vector into the xy plane (Figure 62.3). The received signal was used for gray level modulation. By evaluating the mean gray value within the same area (corresponding to the amplitude of the magnetization sum vector in a particular direction) of eight images registered at different inversion times, a tissue-specific T1relaxation curve was plotted. In Figure 62.4 the x axis represents the different inversion times and the y axis the gray value within the same area of the tissue. As only the absolute value of the z-magnetization, but not its sign (e.g.,
gradient with a specific incrementation phase codes the magnetic resonance signal. Switching on the “z” gradient during the selective 180° soft pulse (e.g., Hermite- or Gauss-shaped pulse) causes the refocusing of the spins in only one particular slice. The “x” gradient is the read-out gradient and determines the axial resolution. The received nuclear resonance signal thus provides information about the spatial location in x and y direction through the frequency and the phase of the signal. This spin–echo sequence can be used for either a single slice (spin–echo sequence XYIMAGE.AU [Bruker, Germany] which is based on the pulse sequence of Hahn for the generation of spin–echo images [Hahn, 1950]) or 1.0 0.9 0.8
a
0.7 b
I(t)
0.6
c
0.5
d
0.4 0.3 0.2 0.1 0 t
FIGURE 62.4 Substance-specific temporal (x axis) changes of the signal amplitude of different human tissues (y axis), assessed by inversion-recovery sequence: (a) adipose tissue, (b) white matter of the brain, (c) grey matter of the brain, and (d) liquor cerebri. (Modified from Schmiedl U., Kölbel G., Griebel J. (1985) Begriffe der medizinischen Kernspintomographie, Teil 3: Die Kontrastmechanismen und der Einfluß der biologischen Parameter auf das MR-Bild. Röntgenpraxis 38: 352-356.)
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
spatial orientation), is measured, the tissue first appears darker with increasing inversion time and then becomes whiter while rising asymptotically to a fixed value. If the z-magnetization was displayed with regard to its sign, the first part of the curve in Figure 62.4 would have to be mirrored along the abscisse (e.g., Figure 62.11). The second method we used to measure the longitudinal relaxation time is the aperiodic saturation–recovery technique. Fast application of several 90° hard pulses and a slice selective 180° hermitian shaped pulse (echo time 20 ms, repetition time 100 to 5000 ms) leads to a saturation of the magnetization sum vector. To measure amplitude, the magnetization sum vector in a particular direction — like in the inversion–recovery technique — a rectangular pulse is emitted to turn the magnetization sum vector into the xy plane. The received high-frequency signal is processed in a similar way as in the inversion method. The main advantage of the saturation–recovery in comparison to the inversion–recovery procedure is that due to the saturation of the magnetization, no waiting time is needed between the successive pulses and thus the measurement time can be markedly reduced.
62.2.4 TRANSVERSAL RELAXATION TIME (T2) Like in the spin–echo imaging technique, a 180° pulse is superimposed on a previously applied 90° pulse. A spin–echo signal is emitted during refocusing. In order to obtain an increasing T2 weighting in multi-echo images, the 180° pulse is emitted repeatedly several times after the echo time. T2 values are calculated from a sequence of T2 weighted images (16 images from one section); the region of interest was 0.01 mm. T1 and T2 values could be obtained only from a part of the specimens as the technical setup for their measurement was not available from the very beginning of our study. Table 62.1 and Table 62.2 show the numbers to which the mean values in each tissue refer.
62.2.5 VOXEL RECONSTRUCTION We used a three-dimensional spin–echo sequence to aquire an isotropic 1283 data cube. The pixel resolution was between 40 μm3 and 20 μm3. The echo time was 5.5 ms, the repetition time was 60 ms. The three-dimensional spin–echo method makes it possible to shorten the echo time by the use of hard pulses. This minimizes the signal loss due to rather short T2eff relaxation times often observed in small probes. The three-dimensional visualization software enables reconstruction of sectional two-dimensional images in oblique planes. Three-dimensional image presentation provides the possibility of spacial reconstructions by extracting different structures from the data cube using an image processing workstation (X32, Bruker GmbH, Rheinstetten, FRG).
541
62.2.6 STATISTICS The U-test (Mann–Whitney–Wilcoxon) for unpaired observations was used to compare the T1 and T2 values of different tissues (respectively, epidermis, dermis, subcutis, nevocellular nevus, melanoma, basal cell carcinoma, and seborrheic keratosis). p-values of ≤0.001 were considered significant.
62.3 RESULTS As mentioned before, MRIs are non-destructive sections through the whole tissue block. Using a defined slice thickness, the MR slices could be located within the tissue bloc. After having precisely oriented the histological section plane in comparison to the MRIs, the histological sections were used to compare the morphological aspect and the visual appearance of different skin structures.
62.3.1 BASAL CELL CARCINOMA Figure 62.5a shows the MRI of a sklerodermiform basal cell carcinoma from the back of an 81-year-old patient. Figure 62.5b exhibits the corresponding histology. In the spin–echo image (Figure 62.5a), three skin regions of different texture are evident. In the upper part of the picture there is a light zone which is distinctly demarcated from a darker, more inhomogeneous, broad area in the middle part. From the upper light zone several round globules connected with it project into the darker area underneath. The lower part of the image consists of an irregular, very light region that is partly separated by dark bandlike structures. The correlating histological section demonstrates that the the regions shown in the spin–echo image correspond to the epidermis, dermis, and subcutaneous tissue. Epidermal buds of basaloid cells, which project into the papillary dermis, are apparent especially along the right upper edge of the histological picture. They correspond to the globules in the spin–echo image, which project from the light upper zone into the darker middle zone. In the lower part of the histological picture, the subcutaneous fat is visible with its typical lobular structure and connective tissue septae in between. The T1 relaxation times were determined in this basal cell carcinoma by an inversion–recovery sequence. Figure 62.6 shows the tumor at eight different inversion times. Remarkably, epidermis, dermis, and subcutaneous tissue go through the “dark point”, i.e., the intersection point of their tissue-specific T1 relaxation curve with the abscisse, at different times. In the third picture of the eight-image sequence, the subcutis is running through the dark point while the dermis and epidermis have the same lightness as in the spin–echo image. In the seventh picture, epidermis and dermis are not visible due to an amplitude of zero for the z-magnetization while the subcutis, which has
542
Handbook of Non-Invasive Methods and the Skin, Second Edition
(a)
(b)
FIGURE 62.5 Superficial basal cell carcinoma from the back of an 81-year-old man. (a) One out of 16 images from a spin–echo sequence multi-slice imaging. Due to the relatively thick slice there is superposition of basal cell carcinoma nests (arrows pointing to globular structures). (b) Correlating histology exhibits an only moderately thickened epidermis with basal cell carcinoma nests discharging into the corium.
basal cell carcinoma we found a longitudinal relaxation time of 1783 ms for the tumor tissue, 1030 ms for the corium, and 437 ms for the subcutaneous fat.
62.3.2 NEVOCELLULAR NEVUS
FIGURE 62.6 Inversion–recovery sequence with eight different inversion times to determine the longitudinal relaxation time T1 of a superficial basal cell carcinoma.
already fully relaxed, appears as a light region. The last picture of the image sequence shows all tissue components in complete relaxation. By determining the intensity of the signal in the same picture location at different inversion times, a tissue-specific relaxation curve can be plotted and the absolute T1-relaxation time calculated. In this
Figure 62.7a shows the spin–echo image of a nevocellular nevus from the back of a 25-year-old patient. The corresponding histological section is shown in Figure 62.7b. The MRI exhibits an upper wide light area with several dimple-like indentations of its upper edge (Figure 62.7a, arrowheads) sharply demarcated from a lower, dark, inhomogeneous appearing region. A cone-shaped structure with a dark loop inside projects from the upper into the lower part of the picture. The corresponding histology shows a dense aggregation of nevus cells within the thickened papillary dermis, correlating to the upper white part of the MRI. The surface of the epidermis shows several indentations. These are filled with retained horny material which is not visible in the spin–echo image due to its protone deficiency. The nevus extends into the dermis along a dilated hair follicle. The loop-shaped structure in the MRI represents a longitudinally cut hair and retained cornified material.
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
543
(a)
FIGURE 62.8 Saturation–recovery sequence (16 different saturation times) in order to determine the longitudinal relaxation time T1 of the nevocellular nevus. Since the specimen was upside down during data aquisition, the picture has been turned. Relaxation-curve values must therefore be determined from the right bottom image to the left top image.
(b)
FIGURE 62.7 Nevocellular nevus from the back of a 25-yearold man. (a) Spin–echo image from a multislice series. The nevus cell nests extended into the deeper dermis along a hair follicle as displayed also in the corresponding histological section (b).
Figure 62.8 displays the time course of the longitudinal relaxation at 16 different measurement times. The T1 times were determined using a saturation–recovery sequence. As the specimen was oriented upside-down during the measurement, the image was turned so that the relaxation curve values were determined in opposite direction from the image sequence (from the right bottom to the left top). As in the inversion technique, the different lightness levels allow calculation of the absolute T1 relaxation times of different tissue areas. In this nevocellular nevus, the values were 1365 ms for the tumor tissue, 1057 ms for the corium, and 441 ms for the subcutaneous fatty tissue lobules.
62.3.3 MALIGNANT MELANOMA Figure 62.9 shows the spin–echo image and the correlating histological picture of a superficial spreading amelanotic malignant melanoma from the scalp of a 74-year-old patient. The MRI exhibits three distinct regions. The left upper part contains a hemispheric structure which is
lighter than its surroundings and reaches almost the middle of the picture. An incomplete, sickle-shaped, dark border separates this structure from a relatively dark region extending to the right side of the picture. This area contains several irregular, light, small nests in its right part. The whole lower part of the image shows a light zone in which irregular dark thread-like bands are embedded. The corresponding histology (Figure 62.9b) shows a hemispherical tumor underneath the ulcerated epidermis on the left, which extends in irregular formations into the subcutis. The tumor is surrounded by an inflammatory infiltrate around dilated blood and lymph capillaries and by edematously swollen connective tissue. The dermis in the right side of the section is interspersed with tumor proliferates extending discontinually from the epidermis to the deep dermis. The subcutis underneath is separated by irregular fibrous septae. A cross-section through a blood vessel is apparent approximately in the middle of the subcutis. The longitudinal relaxation times for this malignant melanoma were calculated with the saturation–recovery sequence. Figure 62.10 shows the tumor at 16 different saturation times. The subcutaneous fat is the first structure to reach its final lightness in the image sequence, which means it relaxes first. The partly preserved epidermis relaxes next and then tumor conglomerates and the rest of the corium follow. The following T1 relaxation times were evaluated: 1672 ms for the tumor tissue, 1229 ms for the epidermis, 1122 ms for the corium and 427 ms for the subcutaneous fatty tissue. The measured relaxation curves
544
Handbook of Non-Invasive Methods and the Skin, Second Edition
(a)
FIGURE 62.10 Image sequence at 16 different saturation times to determine the longitudinal relaxation time T1 of a malignant melanoma.
skin tumors (seborrhoic keratosis, nevocellular nevus, basal cell carcinoma, and malignant melanoma). The T1 relaxation values were significantly different between all tissues, save for the values of malignant melanoma vs. seborrheic keratosis (Table 62.1). Concerning T2 relaxation values, they were significantly different between all tissues (Table 62.2).
62.4 DISCUSSION
(b)
FIGURE 62.9 Amelanotic malignant melanoma from the scalp of a 74-year-old man. (a) Spin–echo image from a 16-slice sequence. (b) The histology exhibits that the hemispheric structures correspond to tumor formations.
for the epidermis, the dermis, the subcutaneous fat, and the malignant melanoma are displayed in Figure 62.11. Figure 62.12 exhibits the same malignant melanoma at different transversal relaxation time periods, which were registered using a multi-echo sequence. While epidermis, tumor, and dermis are not visible after 66.7 ms (third picture of the image sequence), the subcutaneous fatty tissue is apparent even in the last image. Evaluation of the T2 relaxation times of the different tissues results in 29.75 ms for the epidermis, 23.5 ms for the dermis, 57.5 ms for the subcutaneous fat, and 29.3 ms for the malignant melanoma.
62.3.4 T1
AND
T2 RELAXATION VALUES
T1 and T2 relaxation times were measured for different skin layers (epidermis, dermis, and subcutis) and different
Application of MRI in vivo in dermatology has been possible only since special surface coils have become available [Hyde et al., 1987]. These coils have already been known in ophthalmology, where they were used to detect malignant melanomas of the uvea, for example. Querleux et al. [1988] tried to analyze the different skin layers with MRI using a magnetic ground field of 0.1 Tesla. They stated that this low magnetic ground field is not very well suited to obtain images with high resolution [Querleux et al., 1988]. Bittoun et al. used particularly strong surface coils for skin analysis [Bittoun et al., 1990]. Consistent with our results they found that the epidermis is visualized lighter than the corium in spin–echo images. The subcutaneous fat can be distinguished from other tissues by its lightness due to its short T1- and long T2-relaxation time. We used a strong magnetic ground field and a steep gradient field to obtain high-resolution MRIs. Figure 62.5, Figure 62.7, and Figure 62.9 demonstrate the very good correlation between MRI and histological picture. This comparison enables interpretation of MRIs to obtain basic tissue data of different skin structures. It has to be considered, however, that the thickness of the MR sections is 300 μm as opposed to the 7 μm thick histological sections. In MR microscopy, therefore, skin structures appear more
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
Peak no: 2 I: = I (#) (1 − EXP(−TAU/T1))
2.70
Peak no: 1 I = I (#) (1 − 2 A EXP (−TAU/T1))
1E1
1E1
9.20
545
I (#) = 82.81810 T1 = 1.22902 RSS = 33.99125
3.00 I (#) = 25.23844 A = 0.94397 T1 = 0.42682 RSS = 1.49300
0.0 6.00
−2.43 (c)
(a) Peak no: 3 I: = I (#) (1 − EXP(−TAU/T1))
Peak no: 1 I: = I (#) (1 − EXP (−TAU/T1)) 6.00
1E1
1E1
4.80
I (#) = 55.35602 T1 = 1.67260 RSS = 44.17922
I (#) = 43.42481 T1 = 1.12207 RSS = 18.54144 0.0
0.0
6.00
6.00 (b)
(d)
FIGURE 62.11 T1 relaxation curves of different skin layers: (a) epidermis, (b) corium, (c) subcutaneous fatty tissue, and (d) tumor tissue of the amelanotic malignant melanoma. The curves display different ascents, different intersection points with the x-axis, and different end values at complete relaxation.
blurred; for example, basal cell carcinoma nests are represented as globules. Moreover, artifacts cannot be fully avoided both with MR due to technical difficulties and with histology due to varying shrinkage of the tissue during the paraffin embedding and tissue processing. Our results confirm that especially the subcutaneous fatty tissue lobules can be well differentiated from adjacent structures and therefore can serve as an orientation mark in the magnetic resonance image. Earlier investigations by other research groups revealed that in computer tomograms there are no sex differences concerning the adipose tissue [Dixon, 1983] and that this tissue type does not change with age [Giloteaux and Linz, 1983]. Dooms et al. concluded that the adipose tissue would be suitable as a reference tissue in MRI and proved this assumption on a group of 78 patients [Dooms et al., 1986].
It has been relatively difficult to evaluate the epidermis in normal skin [El Gammal et al., 1992a], which is only partly due to thickness variations of the different skin layers depending on the body region. The physiological skin surface water evaporation particularly reduces the spin–echo signal in addition to the proton-deficiency of keratin. Richard et al. [1991], using a whole-body Sigma imaging system operating at 1.5 Tesla with a special surface gradient coil, were able to differentiate the epidermis of the calf in vivo as a thin, light line overlying the dark dermis. In MRIs from the heel, they found two layers of different signal intensity on the surface: an outer gray and an inner brighter layer. Their interpretation that the outer layer corresponds to the stratum corneum while the light one represents the living epidermis seems plausible, although it lacks histological verification. Salter et al.
546
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 62.1 T1 Relaxation Values are Significantly Different between All Tissues, Save for the Values of Melanoma vs. Seborrheic Keratosis
FIGURE 62.12 Multi-echo sequence with eight images to calculate the transversal relaxation time T2 of the amelanotic malignant melanoma.
[1992] investigated normal skin from the fingerpad using a 2 Tesla magnetic field combined with a 20 cm diameter gradient set. From studies of the skin at different hydration states (after immersion of the finger in water), they concluded that the living epidermis is signal intense in MRIs, whereas the stratum corneum becomes visible only when it is hydrated. To visualize the epidermis in vivo, they occluded the upper skin surface with petroleum gel to reduce significantly skin surface evaporation prior to MRI. Salter et al. [1992] concluded that the str. corneum exhibits two different layers, which appear to be different, when the skin is hydrated. We were not able to visualize these two different layers. Since we did our examinations ex vivo, it is possible that this barrier had broken down already. Similar results were obtained by Querleux et al. [1988, 1995] on the heel. The results from these in vivo investigations correlate very well with our in vitro findings. They all support the evidence that horny material, when dry, is signal poor and thus dark in spin–echo images, while the living epidermis is signal-rich. Some skin tumors increase the thickness or change the properties of the epidermis to such an extent that the epidermis can be visualized easily ex vivo. The basal cell carcinoma, a tumor that develops from the epidermis is represented as a broad light zone in the upper corium. The single basal cell tumor nests appear as globule-shaped structures connected to the epidermis and can be quite well delimited from the adjacent corium. These aggregates of tumor cells specific for a basal cell carcinoma could be exactly localized in the MRI and correlated with histological findings in recent investigations [El Gammal et al., 1992a, 1993]. Likewise, the nevocellular nevus can be precisely localized in the spin–echo image. Correlating with the histological picture it is clearly demarcated from the corium below. The expansion of the tumor along the hair follicle is represented in the MRI as cone-shaped structure and the hair follicle with the sebaceous gland can be separated from the surrounding corium. Protonepoor structures like hairs retained horny material
Tissue Type
Mean Value (ms)
Standard Deviation
n
Subcutis Dermis Epidermis Nevocellular nevus Seborrhoic keratosis Malignant melanoma Basal cell carcinoma
455.1 1087.9 1152.7 1309.0 1417.0 1445.6 1801.2
32.4 29.3 32.8 22.5 68.9 130.7 41.4
46 48 28 12 6 7 17
TABLE 62.2 T2 Relaxation Values are Significantly Different between All Tissues Tissue Type
Mean Value (ms)
Standard Deviation
n
Dermis Epidermis Seborrhoic keratosis Nevocellular nevus Basal cell carcinoma Malignant melanoma Subcutis
19.2 23.3 24.6 26.4 27.4 34.3 48.6
1.0 0.8 0.6 0.4 0.8 3.1 1.4
38 23 5 8 9 8 37
[El Gammal, 1992a, 1995, 1996] and are not visible. That the hyperkeratotic stratum corneum is not displayed has also been observed in investigations of verrucae seborrhoicae [El Gammal et al., 1992a, 1996]. The high resolution of our MRIs and the verification of structures using histology made it possible to determine the proton relaxation times precisely within the different skin layers and tumors in a region of interest of 100 × 100 μm2. By far the lowest T1 and highest T2 values were measured in the subcutis, as found already in previous studies [Bittoun et al., 1990; Richard et al., 1991]. Dooms et al. [1986] proved on a group of 78 patients the constancy of T1 and T2 values regarding the adipose tissue. They proposed to use the subcutis as a reference tissue in MRI. Richard et al. [1991] found that T2 values of the epidermis were significantly higher than of the dermis, which is consistent with our results. Divergent from our study, however, they could not establish any difference regarding T1 values. We found significant differences in relaxation times between the investigated skin tumors. Schwaighofer et al. [1989] studied malignant melanomas vs. benign pigmented skin lesions (nevi seborrheic keratoses) in vivo
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
with MRI. They observed that all skin tumors were less signal intense than fat in T1 weighted images, whereas in T2 weighted images malignant melanomas were more signal intense, benign tumors signal poorer than fat. However, as the signal intensity of structures in T1 and T2 weighted images depends on the inversion time, these results represent subjective impressions rather than quantitative data. Measurements of proton relaxation times, allowing objective evaluation of tissues, were not provided. Other researchers denied that a significant discrimination of skin tumors is possible using relaxation times [Mafee et al., 1986; Zemtsov et al., 1989]. The reasons for these difficulties in tissue differentiation are the insufficient resolution due to relatively low ground and gradient fields and the lacking histological correlation. Moreover, most researchers took sequences of only two to eight T1 or T2 weighted images to determine the relaxation times [Dooms et al., 1986; Schwaighofer et al., 1989; Richard et al., 1991]. T1 relaxation time differences between different malignant melanoma types are rather striking. El Gammal et al. [1995] found that amelanotic malignant melanoma revealed with 1650 ms a rather long longitudinal relaxation time, while melanotic malignant melanoma exhibited a T1 value of 1350 ms which lies only slightly above the values measured in nevocellular nevi. Furthermore, Table 62.1 reveals that the T1 values of malignant melanoma exhibit the largest standard deviation of all tissues examined. These findings suggest that the relaxation time is influenced by the pigment concentration of the tumor. Despite these difficulties we cannot confirm the statement of Zimmer et al. [1985] that a differentiation between malignant and benign tumors according to their relaxation times is impossible. We therefore prefer to modify this statement in that sense, that MRI can deliver some new pieces (T1 and T2 values, shape, and texture) for the puzzle making up the diagnosis. The use of up to 16 images in our study rendered greater accuracy at the cost of very long image aquisition times. The inversion–recovery experiment needed several hours. To delay autolytic processes, which cause a prolongation of the T1 and T2 time [Godd and Schmidt, 1983], we briefly fixed the specimens prior to the measurement. However, it has been described that fixation influences the relaxation times as well, causing a decrease in T1 time [Godd and Schmidt, 1983]. Apart from MRI, several other imaging methods for visualization of the skin have been recently developed. Probably the most important is high-frequency sonography, with a resolution between 40 and 200 μm. As opposed to MRI, tissue differentiation is not possible using 20 to 50 MHz ultrasound [Hoffmann et al., 1992; El Gammal et al., 1993]. In order to be visualized in sonography,
547
FIGURE 62.13 Three-dimensional spin–echo of normal skin (1283 isotropic voxel data set). Surface reconstruction of the skin sample (right) and of different skin appendages (left).
structures have to reflect a fraction of the applied energy. This is typically the case at the collagen bundles of the dermis, which appear as irregular white (echorich) spots in the sonographic image. However, inflammatory infiltrates, and various epithelial, vascular and melanocytic tumors, scar tissue, edema, etc. all appear dark (echopoor) [El Gammal et al., 1994]. The value of sonography lies in the differentiation of these processes from the surrounding dermis and determination of their size. Three-dimensional reconstruction procedures finally reveal additional information about skin structures and their orientation in space [El Gammal et al., 1992d]. Fast three-dimensional spin–echo methods can be used to register three-dimensional data cubes for voxel reconstruction. This additional information can be of particular interest to identify and separate normal skin structures, such as skin appendages, from pathological processes in the corium and subcutaneous fatty tissue using MRI. Figure 62.13 demonstrates the extraction of the outer surface of a biopsy of normal skin (right part of the image) and of skin appendages (mainly hair follicles) within the biopsy (left part of the image) from such a three-dimensional data cube using MR microscopy ex vivo. Our results show the potencies of MRI regarding the investigation of skin. We could demonstrate that the accuracy of tissue differentiation is mainly a question of resolution. With the MRI unit used in this study, we achieved a markedly higher resolution than that reported in previous investigations. However, our equipment is still far from being suitable for routine use at the moment. The main disadvantage is that only very small objects can be studied and, therefore, the investigation of human skin must be performed in vitro. In the future, the development of MR microscopy systems using strong local surface coils and gradient fields placed directly on the skin might provide an important non-invasive diagnostic tool in dermatology.
548
Handbook of Non-Invasive Methods and the Skin, Second Edition
62.5 OUTLOOK According to Zemtsov et al. [1989], an appropriate imaging technique for the diagnosis of skin tumors does not currently exist. However, Zemtsov judges MRI to be a good candidate. Perednia [1991] is more reserved about the rentability of MRI in its present stage for dermatology. Apart from the high technical requirements and the time necessary for picture aquisition as compared to the much faster histological diagnosis, the enormous costs have to be considered as well [Perednia, 1991]. On the other hand, this imaging technique has — like sonography — the advantage that no ionizing radiation is applied and thus all health risks connected with radiation are avoided [Hausser and Kalbitzer, 1989]. Roth [1984] judges the NMR technique as completely harmless to human health according to our current knowledge. We believe that MR microscopy using local surface coils and gradient fields placed directly on the skin will become an important non-invasive diagnostic tool in order to confirm the diagnosis in doubtful cases and to facilitate in difficult cases the exact preoperative staging of skin tumors.
REFERENCES Aguayo JB, Blackband SJ, Schoeniger J, Mattingly MA, Hintermann M (1986). Nuclear magnetic resonance imaging of a single cell. Nature 322: 190–191. Altmeyer P, Hoffmann K, el-Gammal S (1990). Allgemeine dermatologische Ultraschallphänomene. Hautarzt 41 Suppl 10: 124–129. Bittoun J, Saint-Jalmes H, Querleux BG, Darrasse L, Jolivet O, Idy-Peretti I, Wartski M, Richard SB, Lévèque JL (1990). In vivo high-resolution MR imaging of the skin in a whole-body system at 1,5 T. Radiology 176: 457–460. Bloch F (1946). Nuclear induction. Phys Rev 70: 460–474. Breslow A (1970). Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 172: 902–908. Breslow A (1975). Tumor thickness, level of invasion and node dissection in stage I cutaneous melanoma. Ann Surg 182: 572–575. Clark Jr. WH, From L, Bernardino EA, Mihm MC (1969). The histogenesis and biologic behavior of primary malignant melanomas of the skin. Cancer Res 29: 705–726. Dixon AK (1983). Abdominal fat assessed by computer tomography: sex difference in distribution. Clin Radiol 34: 189–191. Dooms GC, Hricak H, Margulis AR, de Geer G (1986) MR imaging of fat. Radiology 158: 51–54. El Gammal S, Aygen S, Hartwig R, Bauermann T, Hoffmann K, Altmeyer P (1992a). NMR-Mikroskopie der menschlichen Haut. H & G Z Hautkr 67: 114–121.
El Gammal S, Hartwig R, Aygen S, Bauermann T, Hoffmann K, Altmeyer P (1993). NMR-Mikroskopie und Gewebsdifferenzierung von Hauttumoren am Beispiel des Basalioms. In: Petres D and Lohrisch (Rds.) Das Basaliom und verwandte Tumoren. Springer Verlag, Berlin, pp. 99–114. El Gammal S, Hartwig R, Aygen S, Bauermann T, el Gammal C, Altmeyer P (1996). Improved resolution of magnetic resonance microscopy in examination of skin tumors. J Invest Dermatol 1287–1292. El Gammal S, Aygen S, Hartwig R, Bauermann T, el Gammal C, Hoffmann K, Altmeyer P (1997). Nevocellular nevus, malignant melanoma, basal cell carcinoma and seborrheic keratosis can be differentiated using magnetic resonance microscopy. In: Altmeyer P, Hoffmann K, Stücker M (Eds.). Skin Cancer and UV-Radiation. Springer Verlag, Heidelberg, pp. 1086–1102. El Gammal S, Hartwig R, Aygen S, Bauermann T, Hoffmann K, Altmeyer P (1995). Nuclear magnetic resonance examination of skin disorders. In: Serup J, Jemec GBE (Eds.). Handbook of Non-Invasive Methods and the Skin. CRC Press, Boca Raton, FL, pp. 305–316. El-Gammal S, Hoffmann K, Auer T, Höß A, Altmeyer P, Ermert H (1990). Computergestützte sonographische (50 MHz) Gewebsdifferenzierung von Hauttumoren. Zbl Haut Geschlkr 158: 105. El-Gammal S, Hoffmann K, Auer T, Korten M, Altmeyer P, Höß A, Ermert H (1992b). A 50-MHz high-resolution ultrasound imaging system for dermatology. In: Altmeyer P, el-Gammal S, Hoffmann K (Eds.). Ultrasound in Dermatology. Springer Verlag, Berlin, pp. 297–322. El-Gammal S, Hoffmann K, Höß A, Hammentgen R, Altmeyer P, Ermert H (1992c). New concepts and developments in high-resolution ultrasound. In: Altmeyer P, el-Gammal S, Hoffmann K (Eds.). Ultrasound in Dermatology. Springer Verlag, Berlin, pp. 399–442. El-Gammal S, Hoffmann K, Kenkmann J, Altmeyer P, Höss A, Ermert H (1992d). Principles of three-dimensional reconstructions from high-resolution ultrasound in dermatology. In: Altmeyer P, el-Gammal S, Hoffmann K (Eds.). Ultrasound in Dermatology. Springer Verlag, Berlin, pp. 355–387. El Gammal S, Hoffmann K, Stücker M, Altmeyer P (1997). Bildgebende Verfahren in der Dermatologie. Hautarzt 48: 432–450. Giloteaux S, Linz MH (1983). Histology of aging: Adipose tissues. Gerontol Geriatr Educ 4: 107–111. Grodd W, Schmitt WGH (1983). Protonenrelaxationsverhalten menschlicher und tierischer Gewebe in vitro, Änderungen bei Autolyse und Fixierung. Fortschr Röntgenstr 139: 233–240. Hahn EL (1950). Spin echoes. Phys Rev 80: 580–594. Hausser KH, Kalbitzer HR (1989). NMR für Mediziner und Biologen Strukturbestimmung, Bildgebung, In-vivo-Spektroskopie. Springer Verlag ,Berlin, p. 175. Hoffmann K, El-Gammal S, Altmeyer P (1990). B-scan Sonographie in der Dermatologie. Hautarzt 41: W7–W16. Hyde JS, Jesmanowicz A, Kneeland JB (1987). Surface coil for MR imaging of the skin. Magn Res Med 5: 456–461.
Magnetic Resonance Microscopy of Normal Skin and Pigmented Skin Tumors
Jorizzo JR, Amparo EG (1986). MR imaging of Blue Rubber Bleb Nevus Syndrome. J Comput Ass Tomogr 10: 686–688. Kuhn W (1990). NMR-Mikroskopie – Grundlagen, Grenzen und Anwendungs-möglichkeiten. Angew Chem 102: 1–20. Lauterbur PC (1973). Image formation by induced local interactions — examples employing nuclear magnetic resonance. Nature 242: 190–191. Lauterbur PC (1974). Magnetic resonance zeugmatography. Pure Appl Chem 40: 149–157. Longmore DB (1989). The principles of magnetic resonance. Brit Med Bull 45: 848–880. Mafee MF, Pegman GA, Grisolano JE, Fletcher ME, Spigos DG, Wehrli FW, Rasouli F, Capek V (1986). Malignant uveal melanoma and simulating lesions: MR imaging evaluation. Radiology 160: 773–780. Perednia DA (1991). What dermatologists should know about digital imaging. J Am Acad Dermatol 25: 89–108. Purcell EM, Torrey HC, Pound RV (1946). Resonance absorption by nuclear magnetic moments in a solid. Phys Rev 69: 37–38. Querleux B, Yassine MM, Darrasse L, Saint-Jalmes H, Sauzade M, Lévèque JL (1988). Magnetic resonance imaging of the skin. A comparison with the ultrasonic technique. Bioeng Skin 4: 1–14. Querleux B (1995). Nuclear magnetic resonance (NMR) examination of the epidermis in vivo. In: Serup J, Jemec GBE (Eds.). Handbook of Non-Invasive Methods and the Skin. CRC Press, Boca Raton, FL, pp. 133–139.
549
Richard S, Querleux B, Bittoun J, Idy-Peretti I, Jolivet O, Cermakowa E, Lévèque JL (1991). In vivo proton relaxation times analysis of skin layers by magnetic resonance imaging. J Invest Dermatol 97: 120–125. Richter S (1976). Wolfgang Pauli und die Entstehung des SpinKonzepts. Gesnerus 33: 253–270. Roth K (1984). NMR-Tomographie und -Spektroskopie in der Medizin - Eine Einführung. Springer Verlag, Berlin, p. 108. Salter DC, Hodgson RJ, Hall LD, Carpenter TA, Ablett S (1992). Moisturization processes in living human skin studied by magnetic resonance imaging microscopy. IFSCC Yokohama, Japan, pp. 587–595. Schwaighofer BW, Fruehwald FXJ, Pohl-Markl H, Neuhold A, Wicke L, Landrum WL (1989). MRI evaluation of pigmented skin tumors. Invest Radiol 24: 289–293. Schmiedl U, Kölbel G, Griebel J (1985). Begriffe der medizinischen Kernspintomographie, Teil 3: Die Kontrastmechanismen und der Einfluß der biologischen Parameter auf das MR-Bild. Röntgenpraxis 38: 352–356. Wehrli FW (1988). Principles of magnetic resonance. In: Stark DD, Bradley WG (Eds.). Magnetic Resonance Imaging. The CV Mosby Company, St. Louis, MO, pp. 3–23. Zemtsov A, Lorig R, Bergfield WF, Bailin PL, Ng TC (1989). Magnetic resonance imaging of cutaneous melanocytic lesions. J Dermatol Surg Oncol 15: 854–858. Zimmer WD, Berquist TH, McLeod RA, Sim FH, Pritchard DJ, Shives TC, Wold LE, May GR (1985). Bone tumors: Magnetic resonance imaging versus computed tomography. Radiology 155: 709-718.
63 Raman Spectroscopy of Skin Howell G.M. Edwards Chemical and Forensic Sciences, School of Pharmacy, University of Bradford, Bradford, United Kingdom
CONTENTS 63.1 Introduction............................................................................................................................................................551 63.2 Raman Spectroscopy: Appeal for Skin Characterization......................................................................................551 63.3 Implementation of Raman Spectroscopy to Skin Studies ....................................................................................552 63.3.1 Interpretation of Raman Spectroscopic Data ............................................................................................554 63.4 Case Studies...........................................................................................................................................................556 63.4.1 Water in Skin .............................................................................................................................................558 63.5 The Future..............................................................................................................................................................559 References .......................................................................................................................................................................560
63.1 INTRODUCTION Raman spectroscopy involves the interaction of monochromatic electromagnetic radiation with molecules, usually in the ultraviolet/visible/near-infrared regions from about 200 to 1300 nm; most of the incident radiation is scattered without change in wavenumber υ o (Rayleigh scattering), but a very small fraction of photons are scattered with shifted wavenumbers, υ o ± υm , where the νm are characteristic of the molecules in the spectra.1 These wavenumber shifts comprise the Raman spectrum, which, because of its origin in molecular scattering rather than radiation/absorption processes, provides vibrational spectroscopic information that is complementary to and not identical with that obtainable from infrared absorption or reflectance spectroscopy. Unlike infrared spectroscopy, Raman spectra can be generated from a wide range of wavelengths. The different origins of Raman and infrared spectra give rise to different polarizability and dipole moment changes, respectively; hence, some molecular vibrations are not active (silent) in the Raman or infrared, and others have significant intensity changes that can alter the appearance of the spectra.2 For example, the carbonyl stretching vibration, ν(C=O), is very strong in infrared absorption, but relatively weaker in Raman scattering, whereas the ν(C=C) in unsaturated species such as oleic acid is very strong in the Raman and weak in the infrared. Generally, vibrational modes that involve only a small change in dipole moment are weak in the infrared and highly polar bonds are very strong, and vibrations with no change of
dipole, such as the (S–S) stretching in keratotic biomaterials, can be Raman active only. Raman spectroscopy celebrated its 75th anniversary of discovery3 in 2003, but the first Raman spectrum of human skin was published4 only as recently as 1992. The reason for this rests at the heart of the Raman technique: until the early 1990s, most Raman spectra were generated with lasers operating in the visible region between about 400 and 650 nm, and the competition from fluorescence emission was often insurmountable, especially for biomaterials. The advent of excitation sources of lower intrinsic energy in the near infrared, particularly with Fourier transform (FT) Raman spectroscopy and Nd3+/YAG excitation at 1064 nm, and more latterly with diode lasers operating between 786 and 850 nm, created a window of opportunity for the recording of Raman spectra from sensitive biomaterials such as skin, hair, nail, and bone.5 In the area of skin research, therefore, the preferred excitation for the recording of Raman spectra is at longer wavelengths; those commonly used commercially are 785 nm (diode laser with CCD detection) and 1064 nm (Nd/YAG laser with interferometry).
63.2 RAMAN SPECTROSCOPY: APPEAL FOR SKIN CHARACTERIZATION The low Raman scattering cross sections for highly polar molecules mean that water and hydroxyl groups do not have the strongly adverse effect noted in absorption spectroscopy; this means that specimens need not be desiccated and can be examined in their natural hydration 551
552
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 63.1 Raman Band Wavenumbers Characteristic of Proteins in Various Conformations
Conformation α-Helix β-Antiparallel pleated sheet Random coil
Amide I (CO Stretching)
Amide II (CN Stretching; NH IN-PLANE BENDING)
16601650 s 16801670 s 16701665 s, br
states.6 This advantage lends itself immediately to in vivo applications; since the Raman spectrum is also sensitive to the molecular interactions between components in a complex system such as skin tissue, changes arising from disease can be monitored through the molecular compositional and conformational effects in the tissue spectra (Table 63.1). If the changes are subtle enough to be specifically recognizable for a particular disease state, they can be attributed as phenotypic markers of the disease. In this respect, therefore, the extensive literature on the Raman spectra of in vivo specimens of healthy and diseased skin tissues acts as the basis for establishing Raman spectroscopy for clinical biomedical diagnosis.7–15 A comparison of the infrared absorption spectra and the Raman spectrum of human stratum corneum is shown in Figure 63.1; the effect of water absorption on the infrared spectral quality is seen in the diminution of spectral intensity below about 1000 cm–1, a region that is important for the characterization of skin components such as lipids, natural moisturing factors, and carbon–sulfur bonds in proteins. In another context, this wavenumber region of the spectrum is critically important for the assessment of the effects of skin absorption of applied or exogenous chemicals, such as chemical enhancers for transdermal drug delivery (e.g., dimethyl sulfoxide, α-cineole). Also, in the region of 400 to 500 cm–1 in the Raman spectrum is seen the (S–S) stretching bands of the cysteine linkage in keratotic materials, such as type I collagen in skin, and these are often used as early biomarkers of skin deterioration. Figure 63.2 shows the full-scale Raman spectrum of type I collagen, indicating the detailed structural information available about protein conformation. In the application of mineral or inorganic-based ointments or pharmaceutical applications to human skin, the presence of Raman bands in the wavenumber region of 100 to 1000 cm–1 is vitally important for diagnostic purposes. Figure 63.3 shows a rather unusual application to illustrate this point: the FT Raman spectrum of 4000-yearold mummified skin of an Egyptian XIIth Dynasty burial.16 The mummy of Nekht-Ankh, which has been the subject of a forensic archaeological investigation as part of the Manchester Museum Mummy Project, yielded the
13001265 s 12401225 s 12601240 s, br
Skeletal CC Stretching 940885 m 10101000 m 960950 m, br
Raman spectrum in Figure 63.3 over the wavenumber range 300 to 1800 cm–1. No fewer than nine bands characteristic of sodium sulfate can be identified in the mummified skin, remnants of the mummification process over a period of some 60 days, which involved desiccative treatment of the cadaver with natron, a natural carbonate/bicarbonate of sodium found in Wadi Natrun, Egypt. It is interesting that no residual carbonate or bicarbonate Raman signatures are observed, but that sodium sulfate, present as a minor component in natron, is clearly present. It is particularly noteworthy that the state of preservation of the skin tissue is seen to be variable over the specimens studied; although parts are extremely well preserved, as might be expected from an important burial accomplished during the acknowledged height of Egyptian mummification skills, several regions are in a very poor state of preservation, as exemplified by the relative Raman band intensity and bandwidth changes in the spectra. The stackplotted spectral system in Figure 63.4 shows a series of poorly preserved skin specimens at the top and well-preserved specimens at the bottom, with spectra recorded between 20 and 30 different specimens and arranged according to their state of preservation spectroscopically in Figure 63.4. The specimens were identified visually, but are clearly very different spectroscopically.
63.3 IMPLEMENTATION OF RAMAN SPECTROSCOPY TO SKIN STUDIES The adoption of a novel technique for biomedical diagnostics on human skin requires several important criteria to be addressed,17 including: •
• •
A database of standard material for the characterization of normal or diseased tissue at the molecular level Sensitivity to change in the tissue — an earlywarning device for skin monitoring Application to in vivo studies — from an in vitro base
Raman Spectroscopy of Skin
553
Intensity
(a)
(b)
3500
3000
2500
2000
1500
1000
500
50
Wavenumber/cm−1
FIGURE 63.1 Comparison of infrared absorption and Raman scattering spectra of human skin in vitro. (a) FTIR spectrum. (b) FT Raman with near-infrared 1064-nm excitation. The specimen for the infrared spectrum had been desiccated, whereas that for the Raman spectrum was naturally hydrated. The different relative intensity response for the skin components in each technique should be noted. (From Williams, A.C. et al., Int. J. Pharm., 81, R11–R14, 1992. Reproduced with permission of Elsevier Science.)
• •
Information provided is novel and cannot be obtained elsewhere noninvasively The results of the analyses will be effective in clinical decision making
A critical factor in the assessment of the efficiency of a technique such as Raman spectroscopy for possible clinical adoption is the transaction from research laboratory to clinical and surgical practice. For example, it may be quite usual in purely research environments to accept the recording of Raman spectra of human skin for characterization purposes in vivo, over relatively long periods, for example, up to 30 minutes, but this is totally unacceptable for clinical applications. The technological improvements required, therefore, to reduce spectral data collection times to the order of a few seconds are necessary for the realistic appreciation and future adaptation of the technique. This has now been achieved,18,19 and prototype miniaturized mobile Raman spectrometers are available for potential
clinical uses. Time is an essential factor because of potential patient comfort and living tissue specimen movement; the Raman instrumentation typical for in vivo clinical diagnostic studies consists of a solid-state laser operating in the near infrared and a fiber-optic probe to bring the radiation onto the skin tissue and to collect the scattered radiation for analysis with a high-number aperture spectrometer.20–23 The diagnostic relevance of Raman spectroscopy applied to human skin studies, therefore, can be summarized as involving the characterization of healthy or diseased tissues; much work has been directed at psoriasis, callus, and basal cell carcinomas — all of which can be classified chemically in terms of skin protein and lipid degradation — and at the detection of extraneous materials in the skin, such as inclusions from the leakage of prosthetic devices. A sample of this work from the existing literature24–33 utilizing in vivo and in vitro comparisons will be described later.
Handbook of Non-Invasive Methods and the Skin, Second Edition
2932.42
554
512.23
409.76
643.21 621.41
829.17 852.93
758.29
1173.61 1125.50 1031.58 1003.17 957.65
1316.84 1339.88
1615.64 1550.38
2723.29
3060.31
3268.79
0.02
3576.35 3557.88 3541.10 3519.39 3502.14
0.04
1243.52
0.06
1448.34
0.08
1668.53
Intensity/arbitrary units
0.10
0.00 3500
3000
2500
2000
1500
1000
500
Wavenumber/cm−1
FIGURE 63.2 FT Raman spectrum of type I collagen showing the detailed spectroscopic features characteristic of the protein. In particular, the bands at 643 and 621 cm–1, which are characteristic of CS stretching, and the broad band at 512 cm–1, characteristic of a gauche-gauche-gauche conformation of the CSSC groups, are well represented for diagnostic applications.
63.3.1 INTERPRETATION DATA
OF
RAMAN SPECTROSCOPIC
The four parameters that are normally available from Raman spectral analysis and that are useful for the assignment of vibrational modes and characterization of the molecular structures of chemical species may be summarized as follows: υ (cm −1 ) The wavenumber position Δ υ 1/ 2 (cm −1 ) The bandwidths, usually expressed as full width at half maximum (FWHM) or as half width at half maximum (HWHM) The band intensity, which is often I (W sr–1) expressed in a relatively subjective way, such as strong, medium, or weak ρ The depolarization ratio, which may provide good supporting evidence for molecular vibrational assignments
The interplay of these factors with each other and between neighboring molecules is often complex, and the values of the wavenumber, bandwidth, band intensity, and depolarization ratio will change with the temperature of the system, molecular orientations in condensed phases, and the molecular environments. Additionally, the geometric optical effects of various sample illumination geometries and mountings in use currently, convolved with different laser excitation wavelengths and spectral data accumulation conditions, such as spectral resolution and signal-to-noise ratios, can all conspire to alter the appearance of Raman spectra obtained from one laboratory to another. This results in considerable expertise often being necessary for the interpretation of Raman spectra of complex molecular systems and the difficulties of comparing the observed spectra with Raman databases. Being a light scattering effect, Raman spectroscopy, using such a diverse range of wavelengths from the UV to the infrared (IR), can also create problems for the analytical spectroscopist working with sensitive and colored materials, because of the onset of specimen degradation, fluorescence emission, and resonance scattering effects.
Raman Spectroscopy of Skin
555
Intensity (arbitrary units)
0.06
0.04
0.02
0.00 1800
1600
1400
1200
1000
800
600
400
Wavenumber/cm−1
FIGURE 63.3 FT Raman spectra of human skin samples from the mummified body of Nekht-Ankh, Egyptian XIIth Dynasty, ca. 2000 B.C., demonstrating broad bands characteristic of protein and absence of the features arising from SS and CS band stretching. Extensive degradation of the skin has occurred, despite the excellence acknowledged to the mummification procedure; the presence of Raman bands due to sodium sulfate at 1151, 1130, 1101, 992 (very strong), 646, 633, 621, 464, and 452 cm–1. From Petersen, S. et al., J. Raman Spectrosc., 34, 375–379, 2003. Reproduced with permission of J. Wiley & Sons.)
The excitation of Raman spectra in the UV and visible blue/green spectral regions can cause problems with fluorescence generation, which may be several orders of magnitude larger than the relatively weak Raman spectral bands, and worse, may result in a spectral degeneration caused by sample absorption and decomposition. This is especially true if laser Raman microscopy is used as an analytical technique at shorter wavelengths, where great care must be taken to maintain the irradiance (W m–2) at the sample at a small level. For example, in a conventional Raman spectrometer, where the sample spot size may be 100 μm in diameter, the reduction in sampling dimension offered by Raman microscopy will provide an increase in the irradiance at the specimen of a factor of up to 103× or more, depending on the magnification of the microscope objective used. Hence, a 10-mW laser power incident on the sample for macroscopic imaging results in a power density at the specimen of 1000× more for the same laser power in the microscopic mode. For insensitive materials, the scattered Raman radiation is correspondingly greater, but biological materials often suffer at these high-power densities — and the spectral appearance is changed. The asymmetry of a complex vibrational Raman feature recorded with a specific spectral resolution may be of a different appearance under higher resolution, when several spectral features have been resolved. The applica-
tion of spectral deconvolution techniques to identify band components in poorly resolved Raman features is well illustrated for a complex biological specimen such as human skin (Figure 63.5); this provides an important topic for the biomedical diagnostic applications of modern Raman spectroscopic techniques.34 It is perhaps significant to discuss here the Raman spectroscopy of ν (–S–S–) disulfide bridging groups, which, because of their nonpolarity and symmetry, do not exhibit an IR absorption spectrum. Hence, vibrational information about this important biological structural entity is provided only from Raman spectroscopic studies. The –S–S– group has a highly significant presence in biomolecular diagnostics since it forms the bonding group between cysteine amino acid (Cys) resides, and its scission is indicative of degradation of keratotic materials, which occurs widely in diverse natural biopolymers, including skin, hair, claws, beaks, feathers, horns, hooves, reptilian scales, and shells. Although the composition of animal and human keratotics varies from about 1 to 10% cysteine, the onset of chemical or natural degradation is manifested by a decrease in intensity of the ν (–S–S–) modes in the region of 490 to 520 cm–1; the ν (–S–S–) modes are normally observed in the Raman spectrum as a rather broad, asymmetric feature, as generally there are several conformations of the C–S–S-–C moiety present. Figure
Intensity (arbitrary units)
556
Handbook of Non-Invasive Methods and the Skin, Second Edition
0.013 0.012 0.011 0.010 0.009 0.008 0.007 0.006 0.005
63.6 shows some important conformers based on the classic work of Qian and Krimm35 on the –S–S– band and cis, trans, or gauche structures; the CSSC grouping in human skin can thus be described as a tgt, ggt, or ggg type structure, where g and t are gauche and trans structures, respectively (Table 63.2).
0.004 0.003 0.002 0.001 0.000 1800 1700 1600 1500 1400 1300 1200 1100 1000
63.4 CASE STUDIES
900
Wavenumber/cm−1
Intensity (arbitrary units)
0.010
0.008
0.006
0.004
0.002
0.000 1800 1700 1600 1500 1400 1300 1200 1100 1000
900
Wavenumber/cm−1
Intensity (arbitrary units)
0.010
0.008
0.006
0.004
0.002
0.000 1800 1700 1600 1500 1400 1300 1200 1100 1000
900
Wavenumber/cm−1
FIGURE 63.4 FT Raman spectra of skin tissue from the NekhtAnkh Egyptian mummy (XIIth Dynasty, ca. 2000 B.C.). Top: Eight specimens of relatively well-preserved protein structure. Middle: Two specimens of partly degraded protein. Bottom: Three specimens of extensively degraded protein, where the αhelical protein conformation has been replaced by disordered conformations. No trace of the SS bonds is observed in any of these spectra. (From Petersen, S. et al., J. Raman Spectrosc., 34, 375–379, 2003. Reproduced with permission of John Wiley & Sons.)
Raman spectroscopic studies of skin tissue are of a relatively recent generation. The natural fluorescence of mammalian skin was an insurmountable obstacle until longerwavelength near-infrared excitation became available, generally in the 1990s. Hence, the first Raman spectrum of postmortem human skin36 was published as recently as 1992. Initially, the prime target in the investigation of human skin was the identification of key Raman spectroscopic biomarkers37,38 that could be used to characterize the skin components, and then to provide the basis for the characterization of healthy and diseased human skin tissue. During this period the effect of skin hydration, pigmentation, stratum corneum thickness, inter- and intraspecimen variation with cadaver age and sample locations, and the presence of surface hair were assessed, and also, the potential of the Raman technique for medical diagnostic work was evaluated.39–41 Important landmarks were the first studies of chemical enhancing agents for percutaneous skin absorption of drugs,42 the effect of UV radiation protective agents in skin creams,43 the effect of photoaging and chronic aging on the water and protein structures in skin,44 and the location and role of the natural moisturizing factor in skin.45 Of critical importance of the latter work was the first recording of an in vivo Raman spectrum from human skin within seconds rather than several minutes, which had been required hitherto. The Raman technique can provide novel information about the interaction of water with other materials at the molecular level, particularly relevant to water structure and hydrogen-bonding effects. Thus, it is not surprising that some of the most recent Raman work44,46,47 on human skin and related tissues has addressed the role of water in relation to lipid and protein structures. The role of water in skin has also been investigated very recently using confocal Raman microscopy for depth profiling through the stratum corneum48 and for studying the skin absorption of the penetration enhancer, dimethyl sulfoxide, in vivo as a function of depth, time, and concentration of the agent applied to the skin.49 In addition, a recent Raman study has shown that it is possible to measure in vivo spectra of carotenoids found in the skin derived from various anatomical locations as well as from healthy basal cell carcinomas, actinic keratosis, and perilesional skin. The novel information that is becoming available through the application of Raman spectroscopy will further establish
Raman Spectroscopy of Skin
557
3.69e + 08 3.32e + 08 2.95e + 08
Intensity (a.u.)
2.58e + 08 2.21e + 08 1.84e + 08 1.48e + 08 1.11e + 08 7.38e + 07 3.69e + 07 0.00e + 00 686.2
796.5
906.8
1017.1 1127.4 1237.7 1348.0 1458.3 1568.6
1678.9
1789.3
2669.2 2725.7 2782.2 2838.7 2895.2 2951.7 3008.2 3064.7 3121.2
3177.8
Wavenumber/cm−1 (a) 9.59e + 08 8.63e + 08 7.67e + 08
Intensity (a.u.)
6.71e + 08 5.76e + 08 4.80e + 08 3.84e + 08 2.88e + 08 1.92e + 08 9.59e + 07 0.00e + 00 2612.7
Wavenumber/cm−1 (b)
FIGURE 63.5 FT Raman spectra of human skin, showing spectral deconvolution techniques to assist the interpretation of the band analyses, (a) 690 to 1790 cm–1 and (b) 2610 to 3170 cm–1. The upper spectra were recorded experimentally, and the lower spectra are the component bands from which the synthetic spectrum (solid, smooth line superimposed on the experimental spectrum) has been created. (From Edwards, H.G.M. et al., J. Chem. Soc. Faraday Trans., 91, 3883–3887, 1995. Reproduced with permission of the Royal Society of Chemistry.)
558
Handbook of Non-Invasive Methods and the Skin, Second Edition
Cα' Cβ' S2
S2
Cα
Cβ'
Cα
S1
Cβ
Cα'
S1
Cβ (a) Cα' Cβ' S2
S2
Cβ'
Cα S1 Cα
Cα'
Cβ
Cβ
the database50 of key molecular vibrational bands for the detection and characterization of skin diseases, including psoriasis, atopic dermatitis, and vitiligo. Next to heart disease, cancer is one of the leading causes of death. Hence, another area of focused research is the possibility to diagnose (pre)cancerous skin tissue by Raman spectroscopy. Early detection provides the best prognosis for a patient, as in many cases curative clinical intervention (via chemotherapy) is possible when a malignancy has not yet reached the stage of an invasive tumor. The initial question that has been addressed was whether differences exist between spectra of healthy tissue and malignant tumors. Depending on the tissue of interest, the spectral changes reported result from an increase in protein content, nucleic acid components, and decreases in lipid components. These studies have shown that goodquality Raman spectra can be obtained from ex vivo tissue and, more importantly, that the Raman spectra exhibit changes.
S1
63.4.1 WATER (b) Cα' Cβ' S2
S2
Cβ'
Cα'
S1 Cα
Cβ
Cβ
S1
Cα (c)
FIGURE 63.6 Conformational arrangements of atoms in disulfide bridges for CSSC linkages in skin: (A) gauche-gauchegauche, (B) gauche-gauche-trans, (C) trans-gauche-trans. The Raman-active SS stretching wavenumbers for these conformations are given in Table 63.2. (From Edwards, H.G.M. et al., Spectrochim. Acta Part A, 54, 745–747, 1998. Reproduced with permission of Elsevier Science.)
TABLE 63.2 ν(SS) Wavenumbers and Conformation for Cystine Residue Groups, SS and CS Group S–S
C–S
Wavenumber/cm–1
Assignment
540 525 510 745–700 670–630
trans-gauche-trans gauche-gauche-trans gauche-gauche-gauche trans gauche
IN
SKIN
As an example of what can be achieved currently using Raman spectroscopic analysis of skin, the results from an automated depth-scanning confocal Raman study for the rapid in vivo determination of water concentration profiles in human skin have been described.48 This has several exciting applications in monitoring transdermal drug delivery processes and in evaluating the effect of cosmetic preparations on human skin. Water is known to play an important role in the stratum corneum barrier function. It acts as a plasticizer, keeping the stratum corneum supple and preventing cracking due to mechanical stress. Moreover, it is thought to regulate the activity of specific hydrolytic enzymes that are important for normal desquamation of corneocytes at the skin surface. It is clear, therefore, that the analysis of the water content of the stratum corneum in vivo would be of great help in addressing many biological, medical, and cosmetic research questions. The water content in tissue can be determined from the ratio of Raman intensities of the OH stretch vibration of water at 3390 cm–1 and the CH3– stretch of protein at 2935 cm–1 (Figure 63.7). For determination of water/protein ratios in skin in vivo the ratios of the integrated intensities of water (3350 to 3550 cm–1) and protein (2910 to 2965 cm–1) bands were used in order to maximize the signal-to-noise ratios and avoid overlap of the water signal with the N–H vibration of protein at 3320 cm–1. Water concentration profiles have been obtained by measuring Raman spectra at a range of depths below the skin surface with 2-μm-depth increments. The signal collection time was 5 s per spectrum. This was increased to 10 s at depths of more than about 20 μm below the skin surface, due to decreased signal intensity.
Raman Spectroscopy of Skin
559
2965
Intensity/arbitrary units
2910
ν(CH) protein
0 2600
3550
3350
ν(OH) water
2800
3000
3200
3400
3600
3800
Wavenumber/cm−1
FIGURE 63.7 Raman spectra of stratum corneum in vivo obtained at the volar aspect of the arm; the shaded areas define the spectral intervals that are used in calculations of the water content in the skin. Experimental conditions: signal collection time, 3 s laser power, 100 mW; 720-nm excitation wavelength. (From Caspers, P.J. et al., J. Invest. Dermatol., 116, 434–442, 2001. Reproduced with permission.)
63.5 THE FUTURE Raman spectroscopy provides a truly noninvasive analytical technique for studying the composition of human stratum corneum at the molecular level. From in vitro experiments a Raman database for the characterization of healthy and diseased skin was constructed, and in vivo confocal Raman spectroscopy now enables the acquisition of Raman spectra from cubic micron regions of living skin with a high spatial resolution to be made; this opens up future work for the investigation of in-depth variations in the molecular composition of human stratum corneum.
70
60
Water content (mass-%)
Figure 63.8 shows two consecutive water concentration profiles that were calculated from in vivo Raman spectra obtained at the volar aspect of the arm. The two profiles are the result of consecutive Raman scans in opposite directions, i.e., from the skin surface into the skin (o) and back out again (*). The differences between the profiles are very small. This illustrates that the influence of skin hydration caused by the laser occlusion is small and is within the time that was needed to complete this fully automated sequence (less than 3 min). Figure 63.9 shows representative in vivo water concentration profiles for normal and hydrated stratum corneum of the arm. The water concentration is low in the stratum corneum (15 to 30%) and increases sharply across the boundary between the stratum corneum and the viable epidermis (located at 10 to 15 μm below the skin surface), where it reaches a concentration of about 70%.
50
40
30
20
10
0
0
5
10
15 20 Depth (μm)
25
30
35
FIGURE 63.8 Two consecutive water concentration profiles.
The monitoring of topically applied skin-penetrating agents can be studied with time-and-depth profiling and interactions between the exogenous chemicals, and skin components such as water, lipids, and collagen can be assessed quantitatively. The operation of a confocal Raman microscope in vivo, perhaps with an optional
560
Handbook of Non-Invasive Methods and the Skin, Second Edition
Water content (mass-%)
70
B
60 50 40 30 20 10 0
0
5
10
15
20
25
30
35
40
Depth (μm)
FIGURE 63.9 Representative in vivo water concentration profiles for normal and hydrated stratum corneum of the arm.
associated imaging device, would enable a molecular picture to be assembled and correlated with visual domains in the skin, such as sebaceous-rich areas, sweat ducts, and hair follicles. The advancement of the technique to full clinical trials of diseased and healthy skin is an objective for the near future; this must involve implementation of databases for spectral recognition of skin disorders at the molecular level. Portability and reliability (robustness) of instrumentation is a necessity to accomplish this operation and must be addressed along with speed of analytical specimen throughput. A highly important sequitur is the postspectroscopic diagnostic outcome of the clinical trialling — does the implementation of this novel Raman technique provide a measurable patient benefit through speed and accuracy of the diagnostic procedures, and can earlywarning indicators of skin disorders be identified? Also, the in vivo monitoring of new drug developments and of their applications to the skin can now be accomplished alongside the normal clinical diagnostics.
REFERENCES 1. Long, D.A., Raman Spectroscopy, John Wiley & Sons, Chichester, U.K., 2002. 2. Edwards, H.G.M., Spectra-structure correlations in Raman spectroscopy, in Handbook of Vibrational Spectroscopy, Vol. 3, Chalmers, J.M. and Griffiths P.R., Eds., John Wiley & Sons, Chichester, U.K., 2002. 3. Raman, C.V. and Krishan, K.S., A new type of secondary radiation, Nature, 121, 501–502, 1928. 4. Barry, B.W., Edwards, H.G.M., and Williams, A.C., Fourier-transform Raman and infrared vibrational study of human skin: assignment of spectral bands, J. Raman Spectrosc., 23, 641, 1992.
5. Edwards, H.G.M., Raman spectroscopy: instrumentation, in Encyclopaedia of Applied Physics, Vol. 16, Trigg, G., Ed., Wiley/VCH Publ., New York, 1996, p. 1. 6. Williams, A.C., Edwards, H.G.M., and Barry, B.W., Raman spectra of human keratotic biopolymers: skin, callus, hair and nail, J. Raman Spectrosc., 25, 95–99, 1994. 7. Mizuno, A., Kitajima, H., Kawachui K., Muraishi S., and Ozaki, Y., Near-infrared Fourier-transform Raman spectroscopic study of human brain tissues and tumours, J. Raman Spectrosc., 25, 25–29, 1994. 8. Frank, C.J., McCreery, R.L., and Reed, D.C., Raman spectroscopy of normal and diseased human breast tissues, Anal. Chem., 67, 666–783 1995. 9. Feld, M.S., Manoharan, R., Salenius, J., OrensteinCarndona, J., Romer, T.J., Brennan, J.F., Dasari, R.R., and Wang, Y., Detection and characterisation of human tissue lesions with near infrared Raman spectroscopy, Proc. SPIE, 2388, 99–104 1995. 10. Liu, C.H., Das, B.B., Sha Glassman, W.L., Tang, G.C., Yoo, K.M., Zhu, H.R., Akins, D.L., Lubicz, S.S., Cleary, J., and Prudente R., Raman fluorescence, and timeresolved light scattering as optical diagnostic techniques to separate diseased and normal biomedical media, J. Photochem. Photobiol. B, 16, 187–209, 1992. 11. Mahadevan-Jansen, A., Mitchell, M.F., Ramanajam, N., Malpica, A., Thomsen, A., Utzinger, U., and RichardsKortum, R. Near-infrared Raman spectroscopy for in vitro detection of cervical precancers, Photochem. Photobiol., 68, 123–132, 1998. 12. Gniadecka, M., Wulf, H.C., Nielsen, O.F., Christensen, D.H., and Hercogova, J., Distinctive molecular abnormalities in benign and malignant skin lesions: studies by Raman spectroscopy, Photochem. Photobiol., 66, 418–423, 1997. 13. Stone, N., Stavroulaki, P., Kendall, C., Birchall, M., and Barr, H., Raman spectroscopy for early detection of laryngeal malignancy: preliminary results, Laryngoscope, 110, 1756–1763, 2000. 14. Romer, T.J., Brennan, J.F., Fitzmaurice, M., Feldstein, M.L., Deinum, G., Myles, J.L., Kramer, J.R., Lees, R.S., and Feld, M.S., Histopathology of human coronary atherosclerosis by quantifying its chemical composition with Raman spectroscopy, Circulation, 97, 878–885, 1998. 15. Salenius, J.P., Brennan, J.F., Miller, A., Wang, Y., Artez, T., Sacks, B., Dasair, R.R., and Feld, M.S., Biochemical composition of human peripheral arteries examined with near-infrared Raman spectroscopy, J. Vasc. Surg., 27, 710–719, 1998. 16. Petersen, S., Nielsen, O.F., Christensen, D.H., Edwards, H.G.M., Farwell, D.W., David, A.R., Lambert, P., Gniadecka, M., and Wulf, H.C., NIR-FT Raman spectroscopy of skin samples from the “Tomb of the Two Brothers,” Khnum-Nakht and Nekht-Anth, XIIth Dynasty Egyptian mummies (ca 2000 BC), J. Raman Spectrosc., 34, 375–379, 2003.
Raman Spectroscopy of Skin
17. Choo-Smith L.-P., Edwards, H.G.M., Endtz, H.P., Kries, J.M., Heule, F., Barr, H., Robinson, J.S., Bruining, H.A., and Puppels, G.J., Medical applications of Raman spectroscopy: from proof-of-principle to clinical implementation, Biospectroscopy, 8, 1–9, 2002. 18. Shim, M.G., Song, L.M., Marcon, N.E., and Wilson, B.C., In vivo near-infrared Raman spectroscopy: demonstration of feasibility during clinical gastrointestinal endoscopy, Photochem. Photobiol., 72, 146–150, 2000. 19. Buschman, H.P., Marple, E.T., Wach, M.L., Bennett, B., Schut, T.C., Bruining, A.V., Bruschke, A.V., van der Laarse, A., and Puppels, G.J., In vivo determination of the molecular composition of artery wall by intravascular Raman spectroscopy, Anal. Chem., 72, 3771–3775, 2000. 20. Bakker Schut, T.C., Witjes, M.J., Sterenborg, H.J., Speelman, O.C., Roodenburg, J.L., Marple, E.T., Bruining, H.A., and Puppels, G.J., In vivo detection of dysplastic tissue by Raman spectroscopy, Anal. Chem., 72, 6010–6018, 2000. 21. Shim, M.G., Wilson, B.C., Marple, E., and Wach, M., Study of fiber-optic probes for in vivo medical Raman spectroscopy, Appl. Spectrosc., 53, 619–627, 1999. 22. Caspers, P.J., Lucassen, G.W., Wolthuis, R., Bruining, H.A., and Puppels, G.J., In vitro and in vivo Raman spectroscopy of human skin, Biospectroscopy, 4, S31–S39, 1998. 23. Mahadevan-Jansen, A., Mitchell, M.F., Ramanujam, N., Utzinger, U., and Richards-Kortum, R., Development of a fiber optic probe to measure NIR Raman spectra of cervical tissue in vivo, Photochem. Photobiol., 68, 427–431, 1998. 24. Hata, T.R., Scholz, T.A., Ermakov, I.V., McClane, R.W., Khachik, F., Gellermann, W., and Pershing, L.K., Noninvasive Raman spectroscopic detection of carotenoids in human skin, J. Invest. Dermatol., 115, 441–448, 2000. 25. Schaeberle, M.D., Kalasinsky, V.F., Luke, J.L., Lewis, E.N., Levin I.W., and Treado, P.J., Raman chemical imaging: histopathology of inclusions in human breast tissue, Anal. Chem., 68, 1829–1833, 1996. 26. Kalasinsky, V.F., Johnson, F.B., and Ferwerde, R., Fourier tranform infrared and Raman microspectroscopy of materials in tissue, Cell. Mol. Biol., 44, 141–144, 1998. 27. Hanlon, E.B., Manoharan, R., Koo, T.W., Shafer, K.E., Motz, J.T., Fitzmaurice, M., Kramer, J.R., Itzkan, I., Dasari, R.R., and Feld, M.S., Prospects for in vivo Raman spectroscopy, Phys. Med. Biol., 45, R1–R59, 2000. 28. Schrader, B., Dippel, B., Fendel, S., Keller, S., Lochte, T., Riedl, M., Schulte, R., and Tatsch, E., NIR FT Raman spectroscopy: a new tool in medical diagnostics, J. Mol. Struct., 408/409, 23–31, 1997. 29. Manoharan, R., Wang, Y., and Feld, M.S., Histochemical analysis of biological tissues using Raman spectroscopy, Spectrochim. Acta Part A, 52, 215–249, 1996. 30. Mahadevan-Jansen, A. and Richards-Kortum, R., Raman spectroscopy for the detection of cancers and precancers, J. Biomed. Optics, 1, 31–70, 1996.
561
31. Bruining, H.A. and Puppels, G.J., Raman spectroscopic methods for in vitro and in vivo tissue characterisation, in Flourescent and Luminescent Probes for Biological Activity, Mason, W.T., Ed., Academic Press, London, 1999, pp. 433–455. 32. Puppels, G.J., Bakker Schut, T.C., Caspers, P.J., Wolthuis, R., van Aken, M., Buschman, H.P.J., van der Laarse, M.G., Shim, M.G., Wilson, B.C., and Bruining, H.A., In vivo Raman spectroscopy, in Handbook of Raman Spectroscopy, Lewis, I.R. and Edwards, H.G.M., Eds., Marcel Dekker, New York, 2002. 33. Schrader, B., Keller, S., Loechte, T., Fendel, S., Moore, D.S., Simon, A., and Sawatzki, J., NIR FT Raman spectroscopy in medical diagnosis, J. Mol. Struct., 348, 293–296, 1995. 34. Edwards, H.G.M., Williams, A.C., Farwell, D.W., Barry, B.W., and Rull, F., A novel spectroscopic deconvolution procedure for complex biological systems: vibrational components in the FTRS of iceman and contemporary skin, J. Chem. Soc. Faraday Trans., 91, 3883–3887, 1995. 35. Qian, W. and Krimm, S., Vibrational studies of the disulfide groups in proteins, J. Raman Spectrosc., 23, 517–523, 1992. 36. Barry, B.W., Edwards, H.G.M., and Williams, A.C., Fourier-transform Raman and infrared vibrational study of human skin: assignment of spectral bands, J. Raman Spectrosc., 23, 641–645, 1992. 37. Williams, A.C., Edwards, H.G.M., and Barry, B.W., Raman spectra of human keratotic biopolymers: skin, callus, hair and nail, J. Raman Spectrosc., 25, 95–98, 1994. 38. Edwards, H.G.M., Farwell, D.W., Williams, A.C. Barry, B.W., and Rull, F., Novel spectroscopic deconvolution procedure for complex biological systems: vibrational components in the FT-Raman spectra of iceman and contemporary skin, J. Chem. Soc. Faraday Trans., 91, 3883–3887, 1995. 39. Edwards, H.G.M., Williams, A.C., and Barry, B.W., Potential applications of FT-Raman spectroscopy for dermatological diagnostics, J. Mol. Struct., 347, 379–387, 1995. 40. Williams, A.C., Barry, B.W., Edwards, H.G.M., and Farwell, D.W., A critical comparison of some Raman spectroscopic techniques for studies of human stratum corneum, Pharm. Res., 10, 1642–1647, 1993. 41. Lawson, E.E., Anigbogu, A.N., Williams, A.C., Barry, B.W., and Edwards, H.G.M., Thermally induced molecular disorder in human stratum corneum lipids compared with a model phospholipid system: FT-Raman spectroscopy, Spectrochim. Acta A Mol. Biomol. Spectrosc., 54A, 543–558, 1998. 42. Anigbogu, A.N.C., Williams, A.C., Barry, B.W., and Edwards, H.G.M., Fourier transform Raman spectroscopy of interactions between the penetration enhancer dimethyl-sulfoxide and human stratum corneum, Int. J. Pharm., 106, 583–586, 1995.
562
Handbook of Non-Invasive Methods and the Skin, Second Edition
43. Schallreuter, K.U., Wood, J.M., Farwell, D.W., Moore, J., and Edwards, H.G.M., Oxybenzone oxidation following solar irradiation of skin: photoprotection versus antioxidant inactivation, J. Invest. Dermatol., 106, 583–586, 1996. 44. Gniadecka, M., Nielsen, O.F., Wessel, S., Heidenheim, M., Christensen, D.H., and Wulf, H.C., Water and protein structure in photoaged and chronically aged skin, J. Invest. Dermatol., 111, 1129–1133, 1998. 45. Caspers, P.J., Lucassen, G.W., Carter, E.A., Bruining, H.A., and Puppels, G.J., In vivo confocal Raman microspectroscopy of the skin: non-invasive determination of molecular concentration profiles: J. Invest. Dermatol, 116, 434–442, 2001. 46. Wessel, S., Gniadecka, M., Jemec, G.B., and Wulf, H.C., Hydration of human nails investigated by NIR-FTRaman spectroscopy, Biochim. Biophys. Acta, 1433, 210–216, 1999.
47. Gniadecka, M., Nielsen, F., Christensen, D.H., and Wulf, H.C., Structure of water, proteins, and lipids in intact human skin, hair and nail, J. Invest. Dermatol., 110, 393–398, 1998. 48. Caspers, P.J., Lucassen, G.W., Bruining, H.A., and Puppels, G.J., Automated depth-scanning confocal Raman microspectrometer for rapid in vivo determination of water concentration profiles in human skin, J. Raman Spectrosc., 31, 813–818, 2000. 49. Carter, E.A., Caspers, P.J., Edwards, H.G.M., Williams, A.C., Barry, B.W., Bruining, H.A., and Puppels, G.J., Monitoring the pentration enhancer dimethylsulfoxide in human stratum corneum in vivo by confocal Raman spectroscopy, Pharm. Res., 19, 1577–1580, 2002. 50. Wohlrab, J., Vollman, A., Wartewig, S., Marsch, W.C., and Neubert, R., Noninvasive characterisation of human stratum corneum of undiseased skin of patients with atopic dermatitis and psoriasis as studied by Fourier transform Raman spectroscopy, Biopolymers, 62, 141–146, 2001.
Mechanical Properties
64 Identification of Langer’s Lines J.C. Barbenel Bioengineering Unit, University of Strathclyde, Glasgow, Scotland
CONTENTS 64.1 Introduction............................................................................................................................................................565 64.2 Object.....................................................................................................................................................................566 64.2.1 Skin Biomechanics ....................................................................................................................................566 64.2.2 Qualitative Correlates of Langer’s Lines ..................................................................................................566 64.3 Methodological Principle ......................................................................................................................................567 64.3.1 Multiple Simple Tension Tests ..................................................................................................................567 64.3.2 Suction Chamber Methods ........................................................................................................................567 64.4 Sources of Error.....................................................................................................................................................568 64.5 Correlation with Other Methods ...........................................................................................................................568 64.6 Recommendations..................................................................................................................................................568 References .......................................................................................................................................................................568
64.1 INTRODUCTION Langer’s lines reflect systematic directional variation in the mechanical behavior of the skin. Dupuytren1 reported that a stilleto, which has a blade of circular cross section, may produce an oval wound. The phenomenon was systematically investigated by Langer,2 who punctured the skin of a series of cadavers with an awl that had a blade that was circular in cross section. The resulting wounds
were generally clefts, i.e., oval holes. Langer placed the stab wounds as closely together as possible and showed that the long axes of the wounds formed a regular pattern, now called Langer’s lines (Figure 64.1). The work has been repeated by Cox,3 who confirmed the systematic nature of the direction of the long axes of the wounds, although the direction of the lines obtained by Cox did not wholly agree with those of Langer.
FIGURE 64.1 Langer’s lines. (From Langer, K., Br. J. Plast. Surg., 31, 3, 1978. With permission.) 565
Handbook of Non-Invasive Methods and the Skin, Second Edition
There are two basic causes of the phenomenon that characterizes Langer’s lines. Even under passive, resting conditions the skin has a series of built-in internal tensions that are, in part, due to the fact that the skin acts as a container for its contents. If a circular incision is made in the skin, then the tensions produce an oval defect with major and minor axes that are larger than the diameter of the original circular incision. The circular piece of skin within the incision will shrink into an oval with axes smaller than the diameter of the incision. The long axis of the oval of the defect will generally be at right angles to the long axis of the skin circumscribed by the incision. These effects are all due to the release of the resting tension due to cutting through the skin. The skin also appears to be mechanically anisotropic even after the removal of resting tensions by excising strips of skin and testing them in vitro, as was done by Langer.1 The relationship between the anisotropic biomechanical behavior of skin and Langer’s lines makes the latter of considerable practical importance. Their direction influences the placement and orientation of surgical incisions because incisions in the direction of Langer’s lines generally gape less and produce better-quality and less conspicuous scars than incisions across Langer’s lines. The results of those methods of non-invasive testing of the mechanical properties of skin that extend the tissue may also be strongly influenced by the relationship between the test direction and the local Langer’s lines, particularly if the test loading is uniaxial. The orientation of the lines must be known if useful and meaningful comparisons are to be made between the results of the tests at similar sites on different subjects, or of the secular changes at the same site on a single subject. Unfortunately, the directions of Langer’s lines are not constant, but show significant variations between people, and may not remain constant at a single site for a specific subject. It is necessary, therefore, to develop test methods for the detection of the direction of Langer’s lines.
64.2 OBJECT The technique used by Langer is the only one that clearly and directly identifies the local directions of Langer’s lines, but it is certainly not a widely applicable method. In order to develop a non-invasive method, it is important to understand those underlying biomechanical properties of the skin that are related to Langer’s lines and to identify those features that can be measured.
64.2.1 SKIN BIOMECHANICS The skin shows nonlinear load–extension of stress–strain when subjected to simple tension4 (Figure 64.2). Langer1 himself carried out such tests on a large number of excised strips of skin, and identified the nonlinearity of the
Load/unit width
566
Extension
FIGURE 64.2 Results of uniaxial tension tests in vivo. The anisotropy of the skin appears as a difference in the terminal, linear, slope, and the extensions produced by the same load/unit width.
load–extension response, commenting that “skin does not stretch proportionally with applied load, indeed the amount of extension grows steadily smaller so that the course of progressive extension cannot be represented by a straight line but by a curve.” The nonlinearity has been repeatedly demonstrated by both in vitro5 and in vivo6 studies. The load–extension response can generally be divided into three phases. The first phase of initial extension is characterized by the high compliance of the tissues when small loads produce large extensions. The second phase represents a progressive, but nonlinear, stiffening of the tissue with increasing extension. The third, terminal, phase is one in which the stiffness of the tissues is greater and the load–extension response becomes linear. Repeated load or extension cycling produces a progressive change in the load–extension response of the skin, with the three phases of behavior becoming more marked in the second and subsequent cycles than in the initial cycle. With an increasing number of cycles the extensions obtained by the application of a specific load increases, but the difference between cycles decreases until a stable, reproducible response is obtained. This behavior is usually called preconditioning and occurs both in vivo and in vitro. It is particularly marked if the skin has been stretched in a direction other than the test direction. The skin also shows time-dependent behavior. The load required to maintain the tissue at a constant extension decreases with time, a phenomenon known as stress relaxation. Conversely, the application of a constant load produces a time-dependent increase in extension, which is known as creep. These effects imply that constant-rate testing is desirable, but rate dependence is not marked in the range of extension rates possible for in vivo tests.
64.2.2 QUALITATIVE CORRELATES OF LANGER’S LINES The load–extension curves described previously are directionally dependent (Figure 64.2), and this is usually
Identification of Langer’s Lines
discussed in connection with the directional dependence of Langer’s lines. It must be realized, however, that such in vivo comparisons are not precise because it is impossible to measure the actual direction of Langer’s lines using a penetration technique, and comparisons can only be drawn between specific non-invasive test results and what are believed to be the general form and direction of Langer’s lines at the test site. The most striking features of tests carried out in the direction of, or normal to, Langer’s lines is that within the second and third phases of the skin’s stress–strain behavior, the extension obtained at a specific load is always least in tests made in the direction of the local Langer’s lines (Figure 64.2). The slope of the third, linear, phase is rather move variable, but generally it is greatest when the skin is stretched in the direction of Langer’s lines. The slope of the initial, compliant, phase is more difficult to investigate in vivo. In vitro tests made on strips of skin aligned either along or across Langer’s lines suggested that the differences in extensibility at a specific load shown in phases 2 and 3 also occurred in the initial phase; there was, however, no significant or consistent relationship between the stiffness and direction of Langer’s lines.7
64.3 METHODOLOGICAL PRINCIPLE The relationship between the anisotropy of skin extensibility and the local orientation of Langer’s lines provides a method of detecting the latter. There are two techniques that can be applied for the purpose of detecting differences of the skin extensibility.
64.3.1 MULTIPLE SIMPLE TENSION TESTS Several devices have been described to carry out uniaxial tension tests on skin in vivo. These can be employed to make uniaxial tests in multiple directions, and the correlation between the direction of Langer’s lines and the biomechanical property of the skin previously described may be used to detect the local direction of the lines. The only major systematic study8 to validate the correlation between simple tension tests and Langer’s lines utilized a constant-speed extension device that was coupled to the skin by double-sided adhesive tape. The load–extension response was obtained for three extension tests performed in a single direction at a given site. The tests were then repeated in three other directions such that the four test axes were at 45˚ angles to their nearest neighbor. The tangent of the linear portion of the third phase of the load–extension curve in each direction was extrapolated, and the extension at which this extrapolated tangent cut the zero load axis was defined as the limit strain. The values of the limit strain in the four test directions were then used to identify the direction of minimum limit
567
strain, which was strongly correlated with the direction of Langer’s lines at the sites investigated. The technique has several drawbacks. Conducting multiple tests to obtain the complete load–extension curve, with three replicates in each direction, is extremely timeconsuming. Limitations of the number of tests to reduce the duration of the program reduce the accuracy with which the direction of the minimum limit strain can be defined. In part, the need to investigate the terminal, linear phase of extension that is required to define the limit strain increases the amount of preconditioning in other directions and the number of repetitions required. A simpler method of using uniaxial tension tests would be to measure the directional variation of extension produced by a load large enough to be well within the second phase or the start of the third phase of extension. The definition of the necessary load is not simple because uniaxial tension tests carried out in vivo produce inhomogeneous deformations9 that depend on such variables as the widths of the loading tabs and their initial separation. Stark10 investigated the angular variation of extension produced by a defined load in order to detect the orientation of Langer’s lines. He used an initial length, equivalent to the separation between the loading tabs, of 30 mm and a load of 14.2 g/mm. He found that the extension reached a relatively stable value 1.5 seconds after the application of load, and that repeated testing was not required because of the lack of preconditioning. The rapidity with which tests could be completed allowed the investigations of eight directions at each site. The published results show an impressive correlation between the direction of minimum extension and the classical direction of Langer’s lines. Stark11 produced his constant load using a simple plastic instrument similar to a pair of dividers, but containing a constant-tension coiled flat strip spring.
64.3.2 SUCTION CHAMBER METHODS The suction/inflation of a circular disc of skin with a fixed periphery has formed the basis of devices to assess the mechanical properties of the skin (see Chapter 65 and 66). Imaging the shape of an inflated disc with a fixed periphery showed that anisotropic tissue produced an inflated dome that was not axisymmetric. The directions of the axes of symmetry of the dome defined the direction of maximum and minimum extensibility of the tissue.12 The method has not been applied to skin in vivo, but has the potential to determine the direction of minimum extensibility, and hence the direction of the local Langer’s lines, from the results of a single test. The suction test devices currently used to assess skin mechanics do not image the shape of the inflated skin disc within the test apparatus, but only indicate the maximum height of the inflated dome. Hence, they cannot be used
568
Handbook of Non-Invasive Methods and the Skin, Second Edition
to investigate anisotropy. In principle, a parallel-sided slit may be used instead of a circular orifice, and variation of dome height with orientation could be used to determine the direction of least extensibility, in a way analogous to multiple simple tension tests. The technique appears, however, to offer few advantages over repeated uniaxial tension tests. It has been suggested that the suction method can be used to detect Langer’s lines if the skin is allowed to slip under the periphery of the suction device, rather than being constrained at the boundary.13 The suction device consisted of a transparent plastic cylinder that was smooth and rounded where it was applied to the skin surface in order to reduce the interface friction. The cylinder housed a dome-shaped piston that defined the shape and amount of skin sucked into the cylinder. The extent of skin slip and translation was determined by drawing a line on the skin around the circular periphery of the suction cylinder. When the suction was discontinued and the skin allowed to return to its resting configuration, the original circle became an oval. Tests were made in the abdominal midline, where both the direction of Langer’s lines and anisotropy are well established. Load–extension tests made at the test site confirmed that the long axis to the oval was in the direction of minimum skin extensibility, and hence of the local Langer’s line. Qualitative investigation of other thoracic and abdominal sites showed that the load axes of the ovals produced by the suction device coincided with the directions of Langer’s lines.
64.4 SOURCES OF ERROR Body posture may change the extension and tension in the skin, and therefore modify the directions of Langer’s lines. There may be a direct effect due to the skin being stretched, particularly close to, or overlying, mobile joints. Changes in muscle size due to shortening or contraction will also stretch the skin and alter Langer’s lines at areas distant from the joints. It is thus necessary to relate the direction of Langer’s lines to posture. A standard posture should be maintained throughout the test program if the determination of the direction of the local Langer’s lines is to precede other tests of the mechanical properties of the skin. Short-term changes in body shape due to such events as pregnancy, rapid slimming, or muscle building may produce significant alterations in Langer’s lines by stretching the overlying skin.
64.5 CORRELATION WITH OTHER METHODS The usefulness and relevance of measurements of skin extensibility previously outlined depend on the correlation
between the anisotropy of these mechanical properties and the direction of Langer’s lines. There also appears to be a correlation between the limit strain and the orientation of skin surface roughness caused by the intersecting grooves and ridges that produce the characteristics of skin surface patterns.14 The measurement of the skin surface roughness is even more complicated than the measurement of skin extensibility, and it cannot be considered a means of determining Langer’s lines.
64.6 RECOMMENDATIONS There is no commercially available equipment designed specifically to detect the local orientation of Langer’s lines. There are, however, several devices for carrying out uniaxial tension tests in vivo (see Chapter 65 and 66), which can be used to determine the load–extension behavior of the skin in multiple directions. Devices that produce a force indication can be used in the constant-force mode described by Stark,10 which is simpler than the determination of limit strain. The use of a suction device that allows skin slip appears to have considerable promise, but has not yet achieved general use and must, at present, be considered as having undergone only limited validation.
REFERENCES 1. Dupuytren, G., Theoretisch-praktische Vorlesungen uber Verletzungen durch Kriegswaffen, Vert, Berlin, 1836. 2. Langer, K., On the anatomy and physiology of the skin (from Sitzungsberichte der mathematisch-naturwissenschaftlichen Klasse der Kaiserlichen Akademie der Wissenschaften, 44, 19, 1861), Br. J. Plast. Surg., 31, 3, 1978 (translated by T. Gibson). 3. Cox, H.T., The cleavage lines of the skin, Br. J. Surg., 29, 234, 1941. 4. Kenedi, R.M., Gibson, T., Evans, J.H., and Barbenel, J.C., Tissue mechanics, Phys. Med. Biol., 20, 699, 1975. 5. Fung, Y.C., Biomechanics in Mechanical Properties of Living Tissues, Springer-Verlag, New York, 1981, chap. 7. 6. Barbenel, J.C., Skin biomechanics, in Concise Encyclopaedia of Biological and Biomedical Measurement Systems, Payne, P.A., Ed., Pergamon Press, Oxford, 1991, p. 347. 7. Wan Abas, W.A.B. and Barbenel, J.C., The response of human skin to small tensile loads in vitro, Eng. Med., 11, 43, 1982. 8. Gibson, T., Stark, H.L., and Evans, J.H., Directional variation in extensibility of human skin in vivo, J. Biomech., 2, 201, 1969. 9. Wan Abas, W.A.B. and Barbenel, J.C., Uniaxial tension test of human skin in vivo, J. Biomed. Eng., 4, 65, 1982.
Identification of Langer’s Lines
10. Stark, H.L., Directional variations in the extensibility of human skin, Br. J. Plast. Surg., 30, 105, 1977. 11. Stark, H.L., The Surgical Limits of Extension and Compression of Human Skin, Ph.D. thesis, University of Strathclyde, Glasgow, U.K., 1977. 12. Zioupos, P., Barbenel, J.C., and Fisher, J., Mechanical and optical anisotropy of bovine pericardium, Med. Biol. Eng. Comput., 30, 76, 1992.
569
13. Barbenel, J.C., A suction method for obtaining the direction of Langer’s lines, in The 8th International Symposium on Bioengineering and the Skin, 1990, p. 68. 14. Ferguson, J.M. and Barbenel, J.C., Skin surface patterns and the directional mechanical properties of the dermis, in Bioengineering and the Skin, Marks, R.M. and Payne, P., Eds., MTP Press, Lancaster, 1983, p. 833.
Chamber Method for 65 Suction Measuring Skin Mechanical Properties: The Dermaflex® Monika Gniadecka Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
Jørgen Serup Department of Dermatology, Linköping University, Linköping, Sweden and Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 65.1 Introduction............................................................................................................................................................571 65.2 Equipment and Determination of Skin Mechanical Properties ............................................................................572 65.3 Variables, Prerequisites, and Practical Guidance to Measurements .....................................................................573 65.4 Mechanical Properties of Normal Skin.................................................................................................................574 65.5 Skin Mechanical Properties in Pathology .............................................................................................................575 References .......................................................................................................................................................................576
65.1 INTRODUCTION Mechanical properties of the skin have been studied noninvasively by a number of methods, using uniaxial stretching, ballistometric techniques, torsion, indentation, and, finally, suction.1–7 The principle of the suction method is the measurement of skin elevation caused by the suction force exerted over a defined area of the skin.8–10 Measuring skin elevation in the function of the suction time or the suction force calculates various parameters of skin mechanical properties. From the physiological and pathophysiological points of view, it is crucial to determine at least two properties: stiffness (resistance to change of the shape) and elasticity (the ability to recover shape after stretch). Additional parameters may be calculated with the use of different types of equipment. From the mechanical point of view skin may be considered a five-layered structure consisting of: •
•
The stratum corneum and outer portion of the epidermis, with a fluctuating water content and continuous adaptation to environmental humidity conditions. The internal portion of the epidermis with the basement membrane zone, desmosomes, and elastic fiber, anchoring to deeper structures.
•
•
•
The papillary dermis, with a relatively loose connective tissue defining the microrelief of the skin. The reticular dermis, built up of tight connective tissue spatially organized as expressed by the Langer lines. Coarse wrinkles involve this layer. The subcutaneous space, with attachments to deeper structures, such as fascia.
Selection of the method for measuring of skin elasticity depends on the expected abnormality to be characterized, region of the body under consideration, technical specifications of the piece of equipment, in particular the measuring probe, the area of the skin that is being measured, and the type of stress (vertical, horizontal, or torsional). Methods can be divided into those exerting a proportional full-thickness strain, useful mainly for dermatological and medical applications, and those exerting a disproportional superficial strain, mainly useful for cosmetological purposes. The Dermaflex®, with its chamber diameter of 10 mm, is an example of a proportional method, whereas Cutometer® (chamber diameter of 1 mm) and twistometers are examples of disproportional methods, which highlight the influence of the outer mechanical compartment of the skin. 571
572
Handbook of Non-Invasive Methods and the Skin, Second Edition
The relevance of mechanical parameters is a complex issue. In scleroderma the correlation between stiffness and collagen excess is logic; the increased stiffness of chronically inflamed skin can also be easily understood. However, using methods based on disproportional deformation to verify effects of cosmetic creams may be more difficult to understand, since this type of strain is quite far from common practices among users, who ensure skin elasticity by hand movements parallel to the skin surface.
65.2 EQUIPMENT AND DETERMINATION OF SKIN MECHANICAL PROPERTIES The prototype of the suction machine for the investigation of skin mechanical properties was described by Grahame.10 He adopted the principle of the diaphragm method of Dick11 and Tregear,12 which was successfully used for the determination of the mechanical properties of skin explants in vitro. Later, more sophisticated electronic instruments became commercially available; the most well known are Dermaflex A (Cortex Technology, Hadsund, Denmark), developed in our laboratory, and Cutometer (Courage and Khazaka), described extensively in the laboratory of Elsner et al.9 The Dermaflex A7,25 consists of three main parts: (1) the generator of the vacuum, which is transduced to the suction probe placed directly on the skin; (2) the sensor of skin elevation inside the probe; and (3) the data elaboration and visualization system. The suction probe is equipped with a steel ring (diameter of 10 mm) that sharply demarcates an area of the skin where suction is applied (Figure 65.1). The double-adhesive ring around the steel ring additionally prevents skin creeping during suction. The elevation of the skin is determined electronically by measuring electric capacitance between skin surface and the electrode placed in the top of the suction chamber. The accuracy of this measurement has been shown to be 0.01 mm.8 The instrument allows adjustment of the vacuum strength, the length of the suction period, and the number of suction cycles. The most often used set of parameters was suction, 300 or 450 mbar; suction period, 4 or 20 seconds; and number of cycles, 5 or 6. The parameter of skin stiffness (distensibility) is a value of skin elevation or strain (in millimeters) at the end of the first suction; residual skin elevation after the release of the first suction is named resilient distension (Figure 65.2). The relative elastic retraction (RER), reflecting biological elasticity, can be calculated from the formula RER =
distensibility-resilient distention × 100% disstensibility
FIGURE 65.1 Dermaflex A® for the measurement of skin mechanical properties. (P) Suction probe of the instrument, arrow shows guard steel ring demarcating the area of measurement.
1
2
3
FIGURE 65.2 Example of skin elevation plot obtained from Dermaflex A® for determination of skin mechanical properties (a series of nine cycles): (2) distensibility, (3) resilient distension. Hysteresis (1) is observed on repeated suctions.
A 100% value represents the perfect recovery of skin shape after stretch. An additional parameter, hysteresis, represents the creeping phenomenon and is defined as the distance that the skin is stretched beyond the distensibility when the suction is repeated over the same area (Figure 65.2). Another suction instrument for assessment of skin mechanical properties is the Cutometer SEM 474 (Courage and Khazaka, Köln, FRG). The main difference from
Suction Chamber Method for Measuring Skin Mechanical Properties: The Dermaflex®
the Dermaflex A is the considerably smaller size of the suction chamber, 2 vs. 10 mm in Dermaflex A. It is conceivable that with the Cutometer one measures the mechanical properties of the epidermis, papillary dermis, and, to a lesser degree, deeper layers of the dermis and subcutis.9,13 The results of the measurements with the Dermaflex, which is equipped with a larger suction probe, are mainly influenced by the mechanical properties of the dermis. Thus, equipment with small-size suction probes would be more suitable for detection of alterations of mechanical properties due to the changes in epidermis and papillary dermis, whereas with the larger suction probes the mechanical functional status of the whole integument would be measured. The correlation of the separate parameters of skin mechanical properties with the structural elements in the skin has not been fully elucidated. According to the mechanical model, skin is a sponge-like viscoelastic structure, composed of fibers and the colloidal ground substance, covered with a relatively rigid epidermal sheet, and filled with a free-movable (Newtonian, colloid-poor) fluid.14 Evidence has been presented that elastic fibers in the dermis are principally responsible for elastic properties of the skin under minor loads. When a minor stretch of the skin takes place, elastic fibers are primarily responsible for the return of the skin to the previous shape.15,17 The loss of elastin, e.g., in cutis laxa, is connected with diminished skin elasticity.18 The relative architecture of these fibers is important. It has been shown that in normal conditions, elastic fibers in the immediate subepidermal area form a delicate framework perpendicular to the epidermal surface.19 These fibers are attached to the epidermal-dermal junction from one side, and to the layers of elastic fibers that run parallel to the surface of the skin from the other. Therefore, the structure and shape of the epidermaldermal junction may influence skin elasticity. For example, in psoriasis, where the rete ridges of the epidermis are elongated, skin elasticity is compromised.8 The importance of the structural relationship of the elastic fibers for skin elasticity can be exemplified in elastoderma, where excessive and disorganized elastin proliferation in the skin takes place with complete loss of its elastic recoil.20 In the situation of mild suction, collagen bundles unfold and do not resist skin stretching. However, when higher traction forces are applied, stiff and inelastic collagen bundles become totally stretched and counteract mechanical deformation of the skin. The quantitative deficiency of collagen and the disorganized collagen architecture result in excessive skin extensibility, laxity, and fragility to mechanical stress.21,22 Besides elastin and collagen, other fibrillar components in the skin, such as fibrillin, may be of importance in determination of skin elasticity.23 A suction chamber pressure above 50 to 100 mbar will normally represent a higher traction force, and
573
the measured distensibility represents resiliency of the collagen fiber network. An important but overlooked component in determining skin elasticity is epidermal and dermal water. Jemec et al.24 showed conclusively with the Dermaflex A that skin distensibility and hysteresis increase after epidermal moisturizing with tap water. Elasticity of the skin decreases upon epidermal hydration. Similarly, in psoriasis and histamine wheals, where inflammatory edema in the dermis is present, hysteresis of the skin increases.8,25,26 This suggests that water, which represents approximately 70% of skin mass, acts as an oil both in the epidermis and in the interstitium, increasing the distensibility and creeping, while dampening elastic retraction. Moreover, decreased water content in the skin is associated with decreased skin elasticity, exemplified by the phenomenon of inelastic skinfold, which is a reliable clinical sign of dehydration.
65.3 VARIABLES, PREREQUISITES, AND PRACTICAL GUIDANCE TO MEASUREMENTS Variables may be divided into technical variables related to the equipment, biological variables, and variables related to the laboratory, the investigator, and the lab’s environment. The manufacturer’s manual, together with the literature on the equipment, should provide sufficient technical information, including suggestions of presettings. Biological and environmental variables that need be considered are: • • • • • • • • • • • •
Sex Age Race Anatomical site, the vertical vector of skin stiffness Endocrine factors (menstrual cycle, hormonal treatment) Water balance, including treatment with diuretics General and localized diseases, including gravitational syndrome Diurnal variation of water accumulation in the dependent parts of the body Consequences of sun exposure (photodamage, PUVA photosclerosis) Seasonal variations in humidity of the stratum corneum Previous physical stress to the skin (stress memory lasts for at least 1 hour) Topical treatment
574
Handbook of Non-Invasive Methods and the Skin, Second Edition
The following points should be considered to standardize measurements: •
•
• •
•
•
•
Standardized positioning of the body region, in particular with respect to positioning of adjacent large joints, including supination/pronation. Careful selection and marking of the site to be measured; ideally adjacent sites should be measured and the mean value should be calculated. Standardized placement of the probe with respect to load and angle. Be especially aware of the importance of the zero situation immediately at the initiation of the measurement; deviation from the natural position will strongly influence the result. In contrast, once the suction has been applied, the probe can be moved vertically in the millimeter range with no influence on the final result, thus demonstrating a marginal importance of the subcutaneous septa bindings. Standardize with respect to axial orientation and Langer line position if a noncircular suction chamber is used or a method sensitive to axial orientation is applied. Observe that the surrounding skin is fixed and remains fixed during measurement, with no creeping into the suction probe (this phenomenon may manifest as a sudden increase in distention during repeated measurement). Avoid repeated measurements in the same site, at least for 1 h; the initial recording will temporary modify the viscoelastic behavior within that site.
Skin mechanics is not a static parameter, but depends heavily on a complex interaction between tissue fluid and solid components. Skin is clearly an anisotropic and viscoelastic system. Skin mechanical properties may depend on skin thickness. However, it does not seem to be justified to adjust mechanical parameters of the skin-to-skin thickness, as is typically done in traditional engineering. Precise characteristics of the skin in the measured site are more useful.
65.4 MECHANICAL PROPERTIES OF NORMAL SKIN Mechanical properties of normal human skin have been investigated in detail by Grahame,10 Cua et al.,27 Malm et al.,28 and Gniadecka et al.29 The authors found that these properties depend on the anatomical site. A vertical vector of distensibility has been described,28,29 where the skin in the acral sites (malleoli, forearms) is less distensible (more
Distensibility Young Aged
5
Elasticity
0
0
50
100
FIGURE 65.3 Vertical vector skin distensibility and elasticity in young and aged individuals in different sites of the upper and lower extremity (means with 2 SE). Distensibility is measured in millimeters (elongation after suction), biological elasticity is in percent of skin retraction after stretch. (Reproduced from Gniadecka, M., Gniadecka, R., Serup, J., and Søndergaard, J., Acta Derm. Venereol., 74, 188, 1994. With permission.)
stiff) than in the more proximal sites (thigh, arm, trunk) (Figure 65.3). In the study, where higher suction pressure has been applied,29 the similar vertical vector of elasticity (RER) has been detected, skin being less elastic in the acral sites. In addition Malm et al. detected the vertical vector of hysteresis. Acral female skin is more distensible than male skin, whereas the distensibility of the trunkal skin does not differ significantly between men and women.10 It is not known whether genetic or endocrine factors are responsible for modulation of the elasticity of the skin. It is likely, however, that female hormones play some role in modulation of skin mechanical properties, since skin elasticity in women differs throughout the menstrual cycle.30 Age is an important factor influencing mechanical properties of the skin. It is generally accepted that in humans skin elasticity (RER) diminishes with age, whereas the distensibility may remain relatively unaffected.28,39 This finding was confirmed with other methods for assessment of skin mechanical properties.31–36 In aged people it has been demonstrated with the use of the Cutometer that the indices of elasticity diminish with age in both sun-exposed and sun-protected areas.9,13,36,37 It is crucial to distinguish between the intrinsic aging within sun-protected areas, characterized by skin thinning and epidermal and dermal atrophy, and actinic aging of the sun-exposed areas, where the epidermis is hyperplastic and the dermis is thickened. Lévêque et al.38 studied the effects of chronic sun exposure on skin extensibility and elastic recovery. These authors found that in the areas exposed to long-term sun irradiation extensibility and elasticity are decreased. In another study Berardesca et al.39 found decreased skin elasticity after ultraviolet
Suction Chamber Method for Measuring Skin Mechanical Properties: The Dermaflex®
irradiation. The impairment of skin elastic properties could be prevented with topical sunscreens and reversed with topical retinoic acid.40 It has been proposed that diminished elasticity in aged skin is due to the damage, disintegration, or changes of the structure of elastic fibers41–43 that appear in both intrinsic and solar aging. Glycosaminoglycans that accumulate in the papillary dermis of the actinically damaged skin may interfere with the elastic fiber system located in this region. The effect of aging on skin stiffness has not been fully elucidated. Some authors claim that photoaging is associated with diminished extensibility,38 whereas others found no differences between aged and young individuals28,39,40 but in photodamaged and photoprotected sites. The explanation of the contradictory findings may be the different degree of photodamage of the subjects enrolled in the studies and the differential effect of UVA and UVB irradiation. Borroni et al.44,45 showed that experimental irradiation of photoprotected sites with UVA causes diminishing of both elasticity and distensibility. Our own studies showed that skin distensibility in different sites of the body of the aged individuals was not different from that in the corresponding regions in youths. However, the vertical gradient of distensibility in the extremities was weaker in the aged group.29 Decreased vertical vector of stiffness in the aged extremity skin may have serious pathophysiological consequences. Studies in tall animals, such as the giraffe, revealed that skin, fascia, and veins in the dependent areas are stiffer, to directly compensate for the increased hydrostatic pressure of blood in the standing position.46 This congenital antigravity suit can prevent venous distention and leg edema formation in the upright position, in a manner comparable to external compression. The findings of the vertical vector of skin stiffness provide evidence of a similar antigravity suit in humans.29 Impairment of this antigravity suit in aged individuals may enhance formation of the varicose veins and leg edema in the standing position in old individuals. Studies of Gniadecka et al.29 showed that distensibility and elasticity (RER) are not constant, but they fluctuate during the day. In young persons RER and distensibility increase in the afternoon (Figure 65.4). Although the mechanism of the diurnal variation of skin elasticity has not been elucidated, this phenomenon may present an adaptation to the standing position. It has been postulated that stretching of the skin enhances the time-dependent dermal accumulation of fluid, while restoration of previous shape is associated with fluid removal.14 Skin elastic forces contribute to the recovery of skin shape after stretch, and thus enhance clearance of the intercellular fluid. Therefore, even increase of elasticity of the acral skin is advantageous because strong recoil forces assist in the removal of the edematous fluid that accumulates in the extremities during the day.
575
Distensibility (mm)
2.5
Elasticity (% of retraction) Morning Evening
∗
1
0
70
∗
30
Young
0
Aged
Young
Aged
FIGURE 65.4 Diurnal changes of ankle skin distensibility and elasticisty (RER) in young vs. aged individuals. *significant p = 0.002, paired t-test. Reproduced from Gniadecka, M., Gniadecka, R., Serup, J., and Søndergaard, J., Acta Derm. Venereol., 74, 188, 1994. With permission.
TABLE 65.1 Summary of Principal Changes in Parameters of Skin Elasticity in Selected Pathological Conditions8,25,26 Condition Cutaneous edema Previous mechanical stress Psoriasis Schleroderma Sclerotic plaque Early remission Late remission
Distensibility
Hysteresis
Increased Increased
Increased Decreased
Decreased
Increased
Decreased Slightly decreased Decreased
Decreased Increased Slightly decreased
65.5 SKIN MECHANICAL PROPERTIES IN PATHOLOGY A full review of dermatological applications is outside the scope of this chapter, but a few illustrative examples of the use of Dermaflex are provided below (Table 65.1). In the acute edema in histamine wheals the distensibility and hyseresis are increased.8 The excess of water acts as oil, and the fiber network of the dermis is further extended. Repeated suction at the same measuring site results in a relatively minor strain at a higher stress. The initial maneuver brought the fibers into a relatively extended and locked position for a period of at least 1 hour. Psoriasis representing inflammation and protracted edema with a zero situation of a certain strain of fibers manifests with decreased distensibility, but increased hyseresis due to inflammatory edema. In scleroderma the distensibility is decreased in every phase,25,26 and increased skin stiffness
576
Handbook of Non-Invasive Methods and the Skin, Second Edition
is a constant feature, in contrast to thickening. In Ehlers–Danlos syndrome, in contrast, the distensibility increases, particularly on the cheek and arms.48 The subcutaneous fat and connective tissue bindings have no major influence on skin mechanics.
REFERENCES 1. Marks R, Payne PA. Bioengineering and the Skin. MTP Press Ltd., Lancaster, 1981. 2. Marks R. Mechanical properties of the skin. In Biochemistry and Physiology of the Skin, Goldsmith LA, Ed. Oxford University Press, Oxford, 1983, pp. 1237–1254. 3. Christensen MS, Hargens CW, Nacht S, Gans EH. Viscoelastic properties of intact human skin, instrumentation, hydration, effect and the contribution of the stratum corneum. J Invest Dermatol 69:282, 1977. 4. Kijd WL, Daly CH, Nansen PD. Variation in the response to mechanical stress of human soft tissues in the elderly and the young. J Prosthet Dent 32:493–500, 1974. 5. Wan Abas WAB, Barbenel JC. Uniaxial tension test of human skin in vivo. J Biomed Eng 4:65, 1982. 6. Adhoute H, Berbis P, Privat Y. Ballistometric properties of aged skin. In Aging Skin. Properties and Functional Changes, Leveque JL, Agache PG, Eds. Marcel Dekker, New York, 1993, pp. 39–48. 7. Pierard GE. Mechanical properties of aged skin: indentation and elevation experiments. In Aging Skin. Properties and Functional Changes, Leveque JL, Agache PG, Eds. Marcel Dekker, New York, 1993, pp. 49–56. 8. Serup J, Northeved A. Skin elasticity in psoriasis. In vivo measurement of tensile distensibility, hysteresis and resilient distension with a new method. Comparison with skin thickness as measured with high-frequency ultrasound. J Dermatol 12:318–324, 1985. 9. Elsner P, Wilhelm D, Maibach HI. Mechanical properties of human forearm and vulvar skin. Br J Dermatol 122:607–614, 1990. 10. Grahame R. A method for measuring human skin elasticity in vivo with observations on the effects of age, sex, and pregnancy. Clin Sci 39:223–238, 1970. 11. Dick JC. The tension and resistance of stretching of human skin and other membranes with results from a series of ormal and oedematous cases. J Physiol 112:102–113, 1951. 12. Tregear RT. Physical Function of the Skin. Academic Press, London, 1966. 13. Agache P, Monneur C, Leveque JL, de Rigal J. Mechanical properties and Young’s modulus of human skin in vivo. Arch Dermatol Res 269:221–232, 1980. 14. Oomens CWJ, van Campen DH, Grotenboer HJ. A mixture approach to the mechanics of skin. J Biomech 20:877–885, 1987. 15. Oxlund H, Manschot J, Viidik A. The role of elastin in the mechanical properties of skin. J Biomech 21:213–218, 1988.
16. Daly CH. The role of elastin in the mechanical behaviour of human skin. In Digest of the 8th ICMBE, Chicago, IL, 1969, p. 7. 17. Oxlund H. Changes in connective tissues during corticotrophin and corticosteroid treatment. Dan Med Bull 31:187–206, 1984. 18. Fitzsimmons JS, Gilbert G. Variable clinical presentations of cutis laa. Clin Genet 28:284–295, 1985. 19. Bravermann IM, Fonferko E. Studies in cutaneous aging. The elastic fiber network. J Invest Dermatol 78:434–443, 1982. 20. Kornberg RL, Hendler SS, Oikarinen A, et al. Elastoderma-disease of elastin accumulation within the skin. N Engl J Med 312:771–772, 1985. 21. Burton L. Disorders of connective tissue. In Textbook of Dermatology, Champion RH, Burton JL, Ebling FJG, Eds. Blackwell Scientific Publications, Oxford, 1992, pp. 1791–1797. 22. Silence DO, Senn A, Danks DM, et al. Genetic heterogeneity in osteogenesis imperfecta. J Med Genet 16:101–116, 1979. 23. Hollister DW, Godfrey MP, Kene DR, et al. Immunohistologic abnormalities of the microfibrillar-fiber system in the Marfan syndrome. N Engl J Med 323:152–159, 1990. 24. Jemec GBE, Jemec B, Jemec BIE, Serup J. The effect of superficial hydration on the mechanical properties of human skin in vivo: implications for plastic surgery. Plast Reconstruct Surg 85:100–103, 1990. 25. Serup J, Norheved A. Skin elasticity in localized scleroderma (morphoea). Introduction of a biaxial in vivo method, and the measurement of tensile distensibility, hysteresis and resilient distention of diseased and normal skin. J Dermatol 12:52–62, 1985. 26. Serup J. Localised scleroderma (morphoea): clinical, physiological, biochemical and ultrastructural studies with particular reference to quantitation of scleroderma. Acta Dermatol Venereol 66 (Suppl. 122): 1986. 27. Cua AB, Wilhelm KP, Maibach HI. Elastic properties of human skin: relation to age, sex, and anatomical region. Arch Dermatol Res 282:283–288, 1990. 28. Malm M, Bartling MS, Serup J. In vivo skin elasticity of twenty-two anatomical sites. The vertical gradient of skin extensibility and implications in gravitational aging. Submitted. 29. Gniadecka M, Gniadecki R, Serup J, Søndergaard J. Skin mechanical properties present adaptation to man’s upright position. In vivo studies in young and aged Individuals. Acta Dermatol Venereol (Stockh), in press. 30. Berardesca E, Gabba P, Farinelli N, Borroni G, Rabiosi G. Skin extensibility time in women. Changes in relation to sex hormones. Acta Dermatol Venereol (Stockh) 69:431–433, 1989. 31. Berardesca E, Farinelli N, Rabbiosi G, Maibach HI. Skin bioengineering in the noninvasive assessment of cutaneous aging. Dermatologica 182:1–6, 1991. 32. Pierard GE, Lapiere CM. Physiopathological variations in the mechanical properties of skin. Arch Dermatol Res 260:231–239, 1977.
Suction Chamber Method for Measuring Skin Mechanical Properties: The Dermaflex®
33. Pierard GE. A critical approach to in vivo mechanical testing of the skin. In Cutaneous Investigation in Health and Disease, Leveque JL, Ed. Marcel Dekker, New York, 1989, pp. 215–240. 34. Leveque JL, de Rigal J, Agache PG, Monneur C. Influence of ageing on the in vivo extensibility of human skin at a low stress. Arch Dermatol Res 269:127–135, 1980. 35. Dikstein S. In vivo mechanical properties of the skin measured by indentometry and levarometry. Bioeng Skin Newsl 2:23, 1979. 36. Pierard GE. Evaluation des proprietes mechanique de la peau par les methodes d’indentation et de compression. Dermatologica 168:61, 1984. 37. Pierard GE, Lapiere CM. Physiopathological variations in the mechanical properties of skin. Arch Dermatol Res 260:231–239, 1977. 38. Lévêque JL, Porte G, de Rigal J, Corcuff P, Francois AM, Saint Leger D. Influence of chronic sun exposure on some biophysical parameters of human skin: an in vivo study. J Cutan Aging Cosmet Dermatol 1:123–127, 1988/89. 39. Berardesca E, Vignoli GP, Borrni G, Rigano L, Gaspari F. Acute effects of UV rays on mechanical properties of the skin in vivo. In Proceedings of the 16th IFSCC, New York, October 1990. 40. Berardesca E, Gabba P, Farinelli N, Borroni G, Rabiosi G. In vivo tretinoin-induced changes in skin mechanical properties. Br J Dermatol 122:525–529, 1990.
577
41. Fazio MJ, Olsen DR, Uitto JJ. Skin aging: lessons from from cutis laxa and elastoderma. Cutis 43:437–444, 1989. 42. Bouissou H, Pieraggi M, Julian M, Savit T. The elastic tissue of the skin: a comparison of spontaneous and actinic (solar) aging. Int J Dermatol 27:327–335, 1988. 43. Balin AK, Pratt LA. Physiological consequences of human skin aging. Cutis 43:431–436, 1989. 44. Borroni G, Zaccone C, Vignati G, et al. Assessment of biomechanical changes induced by long-term PUVA treatment (>1000 J/sqcm) in psoriatic patients (abstract). In 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 1990, p. 59. 45. Borroni G, Vignai G, Vignoli GP, et al. PUVA-induced viscoelastic changes in the skin of psoriatic patients. Med Biol Environ 17:663–671, 1989. 46. Hargens AR, Millard RW, Pettersson K, Johansen K. Gravitational haemodynamics and oedema prevention in the giraffe. Nature 329:59–60, 1987. 47. Wickman M, Olenius M, Malm M, Jurell G, Serup M. Alterations in skin properties during rapid and slow tissue expansion for breast reconstruction. Plast Reconstr Surg 90:945–951, 1992. 48. Serup J, unpublished observations.
Chamber Method for 66 Suction Measurement of Skin Mechanics: The Cutometer® Ken-ichiro O’goshi Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 66.1 Introduction and Background ................................................................................................................................579 66.2 Measuring Principle...............................................................................................................................................579 66.3 Measuring Device and Practical Use ....................................................................................................................580 66.3.1 Measuring Conditions and Preconditioning of Individuals ......................................................................581 66.4 Physiological Variables and Normal Skin.............................................................................................................582 66.4.1 Age and Sex...............................................................................................................................................582 66.4.2 Study of Skin Disease ...............................................................................................................................582 66.4.3 Study of Product Efficacy .........................................................................................................................582 References .......................................................................................................................................................................582
66.1 INTRODUCTION AND BACKGROUND Elasticity of the skin is necessary for structural support and for interference with facticial stimuli from outside, for joint action, and for the function of the vessels and nerves. Elasticity is mainly controlled by the collagen fibers and the surrounding intercellular ground substance, which consists primarily of water and proteoglycans. Measurement of elasticity or, to be more precise, viscoelasticity is important in various conditions of healthy skin and skin diseases, as exemplified by skin aging and scleroderma. Test of the efficacy of treatments is another application. The Cutometer® (Courage + Khazaka Electronic GmbH, Cologne, Germany) has been introduced as a device that can measure the viscoelastic properties of the skin in vivo. It provides valuable information on physiological and pathological changes of human skin as well as the efficacy of topical treatment. The Cutometer is now well recognized as a commercial standard tool in dermatological and cosmetic research. The Cutometer operates similarly to the Dermaflex®, introduced before the Cutometer.
FIGURE 66.1 The probe of the Cutometer is small and convenient enough to measure skin areas that are difficult to reach.
66.2 MEASURING PRINCIPLE The Cutometer is a suction chamber method. Negative air pressure is applied to the skin surface through the probe aperture (Figure 66.1). The resultant elevation of the skin surface into the suction chamber is measured. Elevation 579
580
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 66.2 The head of the Cutometer has several aperture sizes (2, 4, 6, and 8 mm in diameter) to fit different skin sites and the different study requirements.
is measured by a noncontact optical system in the device that consists of a light transmitter and a light recipient. There are two glass prisms, which project the light from transmitter to recipient. The diminution of the infrared light beam, depending on the elevation of the skin, is measured.
66.3 MEASURING DEVICE AND PRACTICAL USE
measurement of the epidermal elasticity. The 4- or 6-mmdiameter apertures are for study of the outer skin layers, and the 8-mm-diameter aperture is for measurement of full-thickness elasticity. The probe is connected to the main unit with an air tube and an electric cable. The pressure can be adjusted between 50 and 500 mbar and can be built up immediately or gradually at a controlled rate, as decided. The suction time and relaxation time can be changed from 0.1 to 60 seconds, and the number of measurement cycles from 1 to 99. The main unit contains the pump and the evaluation electronics. Two measuring modes can be chosen, a stress–strain technique and a time–strain technique. With the stress–strain mode, the vacuum is increased over a selected period. Then, the deformation (in millimeters) is displayed as a function of negative pressure (in millibars). With the time–strain mode, which is mostly used to study viscoelastic properties of human skin, a selected vacuum is applied for the selected period. The skin deformation is shown as a function of time. The deformation curve is created from rapid deformation representing an elastic section, followed by a viscoelastic section, and finally a viscous section2 (Figure 66.3). The time–strain mode may be used with three cycles of 5-second traction under negative pressure of 500 mbar separated by 5-second relaxation periods. The skin deformation is plotted as a function of time. The resistance of the skin to be sucked up by negative pressure (firmness or distensibility) and its ability to return into its original position (elasticity or elastic relaxation) are displayed as
The Cutometer handheld probe, with a distinctive central suction head (Figure 66.2), can be chosen with different exchangeable aperture sizes (2, 4, 6, and 8 mm in diameter). The aperture can be chosen depending on the level in the skin affected by the disease or condition (Table 66.1). The 2-mm-diameter aperture is primarily used for
TABLE 66.1 Parameters Proposed by Agache et al.1 Parameters
Interpretation
Ue Uv Uf Ur Ua
Immediate deformation: extensibility Viscoelastic contribution: plasticity Final deformation: distensibility Immediate retraction Final retraction after removal of the vacuum
Ua/Uf
Gross elasticity of the skin, including viscous deformation Pure elasticity Biological elasticity The ratio between delayed and immediate deformation
Ur/Ue Ur/Uf Uv/Ue
–
–
FIGURE 66.3 Typical graphical registration of a strain–time curve on human skin using the Cutometer. A load (vacuum) of 500 mbar was applied for 5 seconds, followed by a 3-second relaxation period. (From Berndt, U. and Elsner, P., Hardware and measuring principle: the Cutometer®, in Bioengineering of the Skin: Skin Biomechanics, Elsner, P., Berardesca, E., Wilhelm, KP., and Maibach, HI., Eds., CRC Press, Boca Raton, FL, 2002.)
Suction Chamber Method for Measurement of Skin Mechanics: The Cutometer®
TABLE 66.2 Aperture Sizes (2, 4, 6, and 8 mm in Diameter) and Level in the Skin Folded or Elongated Aperture Size
Layer of the Skin
2 4 6 8
Epidermal (nonuniform) folding Outer skin folding Outer skin elongation Full skin (uniform) elongation
curves at the end of each measurement. From these curves, among the calculated values, the biological elasticity and viscoelastic/elastic ratio with Ur/Uf and Uv/Ue units, respectively, are calculated (Figure 66.2). The following parameters and nomenclature were proposed by Agache et al.1 (Table 66.2): Ue as the immediate deformation or skin extensibility; Uv as the delayed distension reflecting the viscoelastic contribution of the skin; Uf as the final skin deformation (skin distensibility); Ur as the immediate retraction; Ua as the final retraction after removal of the vacuum; Ua/Uf as the ratio of total retraction to total deformation, which is called gross elasticity of the skin, including viscous deformation; Ur/Ue closely resembles Ur/Uf and is also used as a measure of elastic recovery; Ur/Uf as the ratio of immediate retraction to the total deformation, which is called biological elasticity; and Uv/Ue as the ratio between delayed and immediate deformation, which indicates the relative contributions of the viscoelastic plus viscous and the elastic distension to the total deformation. All these parameters are functions of skin thickness. The Uv/Ue ratio increases with decreasing elasticity. High values of these ratios — maximum = 1 (100%) — indicate a high level of elasticity. Ideally, if the skin was an isotropic physical material, the experimental values should be standardized for skin thickness, which can be determination in vivo by ultrasound. Ratios of the above parameters can be taken with no measurement of skin thickness. Ratios are specific for a body site. Skin is as a biological material viscoelastic, and mechanics known from isotropic materials such as Young’s module, depending on the thickness of the sample, are not directly applicable to anisotropic skin.
66.3.1 MEASURING CONDITIONS AND PRECONDITIONING OF INDIVIDUALS Skin elasticity is a dynamic function. Skin contains about 75% water, and the viscous component is thus very important to control and standardize. Persons should be in a neutral water balance and not dehydrated, since this reduces the skin turgor. Skin turgor (from the Latin turgor, swelling) is the water tension of the skin. Turgor is known
581
to reflect the intradermal and general hydration state, as well as the fiber system of the skin, particularly agedrelated changes.3 Repeated stress and deformation of skin result in further distension or elongation, known as creaping or hysteresis. Following a distension, it may take up to an hour for the skin to regain the spontaneous state. Hysteresis of connective tissue is the mechanical change behind warming up in sports. Thus, measurement of skin elasticity with a suction cup may not be repeated on the same site for 1 to 2 hours due to mechanical recall. Applications and conditions that may cause swelling of the skin shall be avoided, such as sun exposure, skin irritation, and major physical activity. Medicines like diuretics and hormones may easily influence skin mechanics. Diurnal variation and orthostatic position may be significant. In the morning, water redistributes and accumulates in the legs shortly after standing up. The skin has an important junction as a water reservoir in the regulation of total body water. The zero situation is important to control. The joint position shall be carefully standardized since the spontaneous stretching of the skin when measurements start will influence the result. Serup4 described that the function in relation to joint motions is vital in evolution and is directly reflected in the Langer lines, which are oriented in directions that clearly respect free motion of rotator joints and hinge joints. The important mechanical skin functions can be followed as Serup described:4 • • • •
To provide a protective and supporting cover of the body To remain tense but allow free motion of joints To counteract gravitation To be soft and pliable to allow effective contact with physical surfaces as a basis for simple and complex sensory perceptions (touch, including stereological perception of objects, pain, heat, cold, and others)
The skin is far from being mechanically uniform or isotropic. It is very different from physical materials because of the woven structure with layers and fibers. The stratum corneum and epidermis are supposed to be relatively rigid, the papillar dermis quite soft and pliable, and the papillar dermis is tense and the mechanically strongest structure of the skin. Skin elasticity is obviously dependent on body region. The thin skin of extremities is due to the woven structure, more rigid than the thick skin of the trunk. Skin becomes more rigid toward the ground, which can be seen as an evolutionary adaptation to gravity, with the pretension of the skin preventing water accumulation in legs during daytime.
582
Handbook of Non-Invasive Methods and the Skin, Second Edition
66.4 PHYSIOLOGICAL VARIABLES AND NORMAL SKIN 66.4.1 AGE
AND
SEX
Age and sex affect the viscoelastic properties of skin as described by Cua et al.5 in 33 healthy subjects. Generally, Uv/Ue increased while Ur/Uf decreased with age. Responses were variable with respect to load applied. Variability within anatomical regions was also noted, but differences between the sexes were not statistically significant for most regions. Couturaud et al.6 followed the evolution of the biomechanical properties of the skin and the microdepressionary network (mDN) as a function of age with different probe diameters of the Cutometer. Measurements obtained with the Cutometer on two groups of average ages 30 and 60 years showed modifications of skin biomechanical properties as a function of age: decrease in elasticity and extension, while fatiguability increases, with age. This was independent of the area and probe diameter. Elsner et al.7 determined the mechanical properties of human genital skin with the use of the Cutometer. Uv/Ue and Ur/Uf were both significantly lower in vulvar than in forearm skin. Ur/Uf decreased significantly with load in vulvar, but not in forearm skin, whereas Uv/Ue increased, while Uv/Ue was not load dependent in either site. Uv/Ue remained constant with age in forearm and vulvar skin. In the vulva, but not in the forearm, Uv/Ue was significantly correlated with body height.
66.4.2 STUDY
OF
SKIN DISEASE
Studies were conducted in psoriasis,8 systemic sclerosis,9–11 Ehlers–Danlos syndrome,12 striae distensae,13 hypertrophic scarring,14 and diabetes mellitus.15 Also, changes of biomechanical properties following the application for ultraviolet light,16 laser treatment,17 reconstructive dermal substitutions,18 and liposuction19 have been reported.
66.4.3 STUDY
OF
PRODUCT EFFICACY
The effect of cosmetics such as emollients, moisturizers, chemical peelings, antiaging creams, etc., studied with the Cutometer were reported.20,21
REFERENCES 1. Agache PG, et al. Mechanical properties and Young’s modulus of human skin in vivo. Arch Dematol Res 269, 221, 1980. 2. Elsner P. Skin elasticity. In Bioengineering of the Skin: Methods and Instrumentation, Berardesca E, Elsner P, Wilhelm KP, Maibach HI, Eds. CRC Press, Boca Raton, FL, 1995.
3. Lockhard, RD, Ed. Living Anatomy, a Photographic Atlas of Muscles in Action and Surface Contours, 2nd ed. Faber & Faber, London, 1962. 4. Serup J. Mechanical properties of human skin: elasticity parameters and their relevance. In Bioengineering of the Skin: Skin Biomechanics, Elsner P, Berardesca E, Wilhelm KP, Maibach HI, Eds. CRC Press, Boca Raton, FL, 2002. 5. Cua AB, Wilhelm KP, Maibach HI. Elastic properties of human skin: age, sex, and anatomical region. Arch Dermatol Res 282(5), 283–286, 1990. 6. Couturaud V, Coutable J, Khaiat A. Skin biomechanical properties: in vivo evaluation of influence of age and body site by a non-invasive method. Skin Res Technol 1, 68, 1995. 7. Elsner P, Wilhelm D, Maibach HI. Mechanical properties of human forearm and vulvar skin. Br J Dermatol 122, 607, 1990. 8. Dobrev HP. In vivo study of skin mechanical properties in psoriasis vulgaris. Acta Derm Venereol (Stockh) 80, 263, 1999. 9. Dobrev HP. In vivo study of skin mechanical properties in patients with systemic sclerosis. J Am Acad Dermatol 40, 436, 1999. 10. Enomoto DN, et al. Quantification of cutaneous sclerosis with a skin elasticity meter in patients with generalized scleroderma. J Am Acad Dermatol 35, 381, 1996. 11. Nikkels-Tassoudji N, et al. Computerized evaluation of skin stiffening in scleroderma. Eur J Clin Invest 26, 457, 1996. 12. Henry F, et al. Mechanical properties of skin in EhlersDanlos syndrome, type I, II, and III. Pediatr Dermatol 13, 464, 1996. 13. Pierard GE, et al. Tensile properties of relaxed excited skin exhibiting striae distensae. J Med Eng Technol 23, 69, 1999. 14. Fong SS, Hung LK, Cheng JC. The Cutometer and ultrasonography in assessment of postburn hypertrophic scar: a preliminary study. Burns 23(Suppl. 1), 12, 1997. 15. Yoon HK, et al. Quantitative measurement of desquamation and skin elasticity in diabetic patients. Skin Res Technol 8, 250, 2002. 16. Habig J, et al. Einflu einmaliger UVA- und UVBBestrahlung auf Oberflächenbeschaffenheit und viskoelastische Eigenschaften der Haut in vivo. Hautarzt 47, 515, 1996. 17. Koch RJ, Cheng ET. Quantification of skin elasticity changes associated with pulsed carbon dioxide laser skin resurfacing. Arch Facial Plast Surg 1, 272, 1999. 18. van Zujilen PP, et al. Graft survival and effectiveness of dermal substitution in burns and reconctructive surgery in a one-stage grafting model. Plast Reconstr 106, 615, 2000. 19. Henry F, et al. Mechanical properties of skin and liposuction. Dermatol Surg 22, 566, 1996. 20. Fischer T, et al. Instrumentelle Methoden zur Bewertung der Sicherheit und Wirksamkeit von Kosmetika. Akt Dermatol 24, 243, 1998. 21. Greif C, et al. Beurteilung einer Körperlotion für trockene und empfindliche Haut. Kosmet Med 19, 24, 1998.
Chamber Method for 67 Suction Measurement of Skin Mechanics: The New Digital Version of the Cutometer André O. Barel,1 W. Courage,2 and Peter Clarys1 1 2
Faculty of Physical Education and Physiotherapy, Vrije Universiteit Brussel, Brussels, Belgium Courage+Khazaka Electronic GmbH, Köln, Germany
CONTENTS 67.1 Introduction............................................................................................................................................................583 67.2 Object of This Study .............................................................................................................................................584 67.3 Methodological Principle ......................................................................................................................................584 67.3.1 Description of the Measuring Probe .........................................................................................................584 67.3.2 Description of the Measuring Modes........................................................................................................584 67.3.3 Handling of the Probe ...............................................................................................................................585 67.3.4 Analysis of the Measuring Systems..........................................................................................................585 67.3.4.1 Strain–Time Curves....................................................................................................................585 67.3.4.2 Stress–Strain Curves...................................................................................................................586 67.3.4.3 Reproducibility of Skin Deformation Parameters and of the Modulus of Young ....................587 67.4 A Short Overview of the Results ..........................................................................................................................587 67.4.1 Single Stress–Time Curves........................................................................................................................587 67.4.2 Repetitive Stress–Time Curves..................................................................................................................588 67.4.3 Stress–Strain Curves..................................................................................................................................588 67.5 Factors Influencing the Measurements..................................................................................................................588 67.5.1 Influence of Load.......................................................................................................................................588 67.5.2 Influence of the Diameter of the Suction Device .....................................................................................588 67.5.3 Influence of the Orientation of the Probe .................................................................................................588 67.5.4 Influence of Pressure of Application on the Skin.....................................................................................589 67.5.5 Preconditioning of the Skin Surface .........................................................................................................589 67.6 Applications ...........................................................................................................................................................589 67.7 Conclusions............................................................................................................................................................589 References .......................................................................................................................................................................589
67.1 INTRODUCTION The mechanical properties of the human skin have been extensively studied in the past, most in vitro and less in vivo.1 Skin is a complex organ that, like many other biologicals, presents in a combined way the typical properties of elastic solids and viscous liquids.2 As a consequence, the mechanical properties of the skin are called viscoelastic. Typical properties of viscoelastic materials are nonlinear stress–strain properties with hysteresis (the stress–strain curves obtained on loading
will not be superposed on the curves obtained by unloading).2–5 Furthermore, the deformation of the skin is time dependent with a typical phenomenon of creep. The creep is characterized as an increasing deformation of the skin in function of time when a constant stress is applied on this material. The viscoelastic properties of the skin are due to the components of the skin: collagen fibers, elastin fibers, and cells impregnated in a ground substance of various proteoglycans and glycoproteins.6 583
584
Handbook of Non-Invasive Methods and the Skin, Second Edition
67.2 OBJECT OF THIS STUDY The general purpose consists of measuring noninvasively the biomechanical properties of the human skin in vivo by means of a simple, reliable, and reproducible biophysical technique. There have been in the past many investigations of the mechanical properties of human skin using various equipment that measure the deformation of the skin after application of uniaxial- or biaxial-oriented forces. In these studies, normally healthy and diseased skin situations must be considered. The influence of physiological factors such as age, sex, normal and actinic aging, diseases, and changes in biomechanical properties induced by various topical treatments can also be examined using mechanical measurements. Many experimental instruments and devices have been developed in research laboratories, but only very few instruments are commercially available: a torsion method (Torque Meter*) and three instruments using the suction method (Dermaflex*, Dermalab*, and Cutometer*). This report describes the new digital version of the Cutometer MPA 580 based on the suction method. The instrument measures the vertical deformation of the skin surface when the skin is pulled in the circular aperture of the measuring probe after application of a vacuum. The use of the older analog version of the Cutometer (SEM 575) has been described by many authors7–11 and will also be reviewed in Chapter 66 of this book.12
67.3 METHODOLOGICAL PRINCIPLE 67.3.1 DESCRIPTION
OF THE
MEASURING PROBE
Figure 67.1 shows a schematic representation of the measuring handheld probe, which is attached to the main apparatus with an electric cable and an air tube. Most technical information concerning the digital version of the Cutometer has been obtained from the manufacturer.13 A variable vacuum (ranging from 20 to 500 mbar) is applied on the skin surface through the opening of the probe.The resolution of applied pressure is equal to 1 mbar. The skin surface is pulled by vacuum in the aperture of the probe. The depth of the skin penetration is measured by an optical system that measures in function of skin penetration the diminution of light intensity of a green light beam (light-emitting diode, LED). Calibration of this optical system is carried out in the factory with a micrometer pin (from 0 to 3.0 mm). The skin adjacent to the opening of the probe is maintained in position by an external guard ring attached to the * Torque Meter® is a registered trademark of Dia-Stron Ltd., Andover, U.K.; Dermaflex® and Dermalab® are registered trademarks of Cortex Technology, Hadsund, Denmark; and Cutometer® is a registered trademark of Courage-Khazaka, Cologne, Germany.
Vacuum tube
Spring
Spring
Movable part of the probe
External part of the probe
Aperture of the probe Guard ring
FIGURE 67.1 Schematic view of the measuring probe of the Cutometer. Probe with a small aperture (2 mm diameter).
probe shield (external diameter of 25 mm) or can be fixed with a double-sided sticking ring. The measuring probe is applied vertically on the surface of the skin with a constant pressure by means of a spring (50 gm/cm2 or 2 × 103 N/m2). The skin deformation (in millimeters) can be measured by this new optical system (improved optical lenses) with a resolution equal to 2 μm. As a consequence, the deformation curves of the new instrument present a much smoother tracing than the old version, where the resolution was 10 μm. Above a penetration depth of 200 μm the accuracy is 3%; below a penetration depth of 100 μm there is no linearity in the optical system. The standard measuring probe has an aperture of 2 mm diameter (test area of about 3 mm2). Optional probes with apertures of 4, 6, and 8 mm are available for studying the mechanical properties of larger skin areas. With the larger measuring probes deeper layers of the skin (dermis and perhaps hypodermis) are deformed by suction. The digital probe of the MPA 580 Cutometer is connected to a central MPA 5 multiprobe apparatus. The MPA 5 apparatus is connected to a PC with a standard Windows software directing the Cutometer. The results can be displayed as curves and values (see later in this chapter). The Courage-Khazaka software allows storage of various data concerning the volunteer, date and time of experiment, skin area, external temperature and relative humidity, type of probe used, and mode of measuring technique. In addition, graphic display of the obtained experimental curves allows calculations of individual values.
67.3.2 DESCRIPTION
OF THE
MEASURING MODES
Essentially two different measuring techniques are available: stress–strain mode and strain–time mode. In the stress–strain mode the deformation of the skin (strain) is displayed as a function of the stress (load–vacuum).
585
600 0.6 0.5 0.4 0.3 0.2 0.1 0.0
Uv
500
Ur Uf
Ue
−10
−5
0
5 10 15 Time (sec)
20
25
Load P (mbar)
Deformation (mm)
Suction Chamber Method for Measurement of Skin Mechanics: The New Digital Version of the Cutometer
400 300 200 100
FIGURE 67.2 Graphical representation of a strain–time curve obtained for forearm skin with the SEM 575 Cutometer. Aperture, 2 mm; time of application, 10 sec relaxation time, 10 sec no pretension applied; load, 500 mbar. Skin deformation in millimeters.
In the strain–time mode the deformation of the skin is showed as a function of time (Figure 67.2). In both experimental modes the choice of vacuum (from 20 to 500 mbar), the duration of the measurements (from 0.1 to 320.0 sec), and the number of measurement cycles (from 1 to 10) can be preset. The apparatus offers a choice of four measuring modes. The four measuring modes are the result of different combinations of choices of application rates and release rates of vacuum on the skin. Mode 1: Measurement with constant negative pressure. Mode 2: Measurement with linear rising and falling of negative pressure. Mode 3: Measurement with first constant, then linear falling of negative pressure. Mode 4: Measurement with linear rising of negative pressure and the abrupt cessation of negative pressure. The strain vs. time curve (mode 1) is mostly used in the mechanical studies on human skin and applications in the field of dermatocosmetics. In measuring system 2, where the deformation of the skin is measured during a linear increase in vacuum from 0 to maximal 500 mbar, and subsequently during the linear decrease of vacuum, the resulting graphical display can be automatically replotted as a stress–strain curve (Figure 67.3).
67.3.3 HANDLING
OF THE
PROBE
The skin immediately adjacent to the opening of the measuring probe is normally held in position by the guard ring of the probe on application of the probe on the skin surface or by the use of a double-sided sticking ring in order to reduce some lateral displacement of the skin adjacent to the opening during the suction.
0 0.0
0.1
0.2 0.3 0.4 Deformation (mm)
0.5
0.6
FIGURE 67.3 Graphical representation of a stress–strain curve obtained for forearm skin with the SEM 575 Cutometer. Aperture, 2 mm; linear increase and decrease in vacuum, 50 mbar/sec total application time, 20 sec.
In order to minimize this lateral displacement when no sticking ring is used, it is possible to pretension the skin before the measurements are carried out. Pretension of the skin is carried out by applying a preliminary suction on the surface of the skin during a short time (about 0.1 sec) before the real vertical deformation measurements are executed.
67.3.4 ANALYSIS
OF THE
MEASURING SYSTEMS
67.3.4.1 Strain–Time Curves In the strain–time mode, which is mostly used in viscoelastic studies on human skin, the vacuum is applied for a period varying from 1 to 10 sec followed by a relaxation of 1 to 10 sec. Figure 67.2 shows the results of a typical experiment carried out on the volar part of the forearm with the old version of the Cutometer (SEM 575). The deformation parameters used in most studies in order to describe the different parts of the curve are those proposed by Aubert et al.5 and Escoffier et al.14 Ue is the immediate deformation–skin extensibility. Uv is the deviation, which reflects the viscoelastic contribution of the skin. Uf is equal to the total deviation of the skin. Ur is equal to the immediate recovery of the skin after removal of vacuum. Due to the slow return (creep) to the original state of the skin after application of a given load, the deformation Ur does not really reach a constant plateau value. A value of 0.1 second after application of suction and removal of vacuum is systematically taken as the time values for measuring Ue and Ur. Figure 67.4 shows the results of a typical experiment carried out on the skin with the new digital version of the Cutometer (MPA 580). The new deformation parameter Ua is equal to the total recovery of the skin. The residual deformation of the skin R is equal to Uf – Ua. The standard Courage-Khazaka software calculates
586
Handbook of Non-Invasive Methods and the Skin, Second Edition
E(mm)
Amplitude/Time 0.54 0.52 0.5 0.48 0.46 0.44 0.42 0.4 0.38 0.36 0.34 0.32 0.3 0.28 0.26 0.24 0.22 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0
Cutometer results 1 repetition
R0 = Uf R1 = Uf − Ua R2 = Ua/Uf R3 = last max. amplitude R4 = last min. amplitude R5 = Ur/Ue R6 = Uv/Ue R7 = Ur/Uf R8 = Ua R9 = R3 − R0 Ue
Ur
Ua
F1
Uf − Ua
Uf = R0 F0 = surface F1 = surface
0
2
1
3
4
1(s)
FIGURE 67.4 Graphical representation of a strain–time curve obtained for forearm skin with the MPA 580 Cutometer. Aperture, 2 mm; time of application, 2; sec relaxation time, 2 sec. Skin deformation in millimeters.
automatically the following parameters: Ro = Uf, R1 = Uf – Ua, R2 = Ua/Uf, R5 = Ur/Ue, R6 = Uv/Ue, R7 = Ur/Uf, and R8 = Ua. In addition, the software calculates the surfaces F0 and F1. F0 is the surface between the real curve and the value corresponding to the maximal deformation Uf when going from start of suction to stop of suction. F1 is the surface between the real recovery curve and the value corresponding to the maximal recovery going from stop of suction to stop of measurement. Both surfaces reflect the viscous part (slow deformation and slow return of the skin to the original part) of the viscoelastic properties of the skin. The deformation vs. time curves obtained for the second, third, and subsequent deformation–relaxation cycles (up to 10 cycles) are similar to the first curve. But the curves are progressively shifted vertically upward as a consequence of the slow return of the skin to the original state (Figure 67.5). In this repetitive mode the total duration of one cycle is short (1-sec suction and 1-sec recovery). The software calculates on this repetitive curve the following parameters: R3 is equal to the last maximum amplitude and R4 is equal to the last minimum amplitude. In addition, the software calculates the surface F2, which is equal to the surface between the real curve and the value corresponding to the maximal deformation R3 after 10 cycles when going from start of suction to stop of the 10 cycles, the surface F3 (surface between the two
repetitive curves), and the surface F4 (surface between the minimal deformations and the time axis). All the deformation parameters, Ue, Uf, Uv, and Ur, are dependent on skin thickness. Since there are significant differences in skin thickness between women and men, and since skin thickness varies with age, the use of these extrinsic deformation parameters for comparative studies is not adequate.14 Consequently, either intrinsic skin deformation parameters are used (deformation × skin thickness) or ratios of the extrinsic parameters are considered. The ratio Ur/Ue, ratio between immediate recovery and immediate deformation, is independent of skin thickness. This ratio is considered a biologically important factor for the characterization of elasticity of the skin.14,15 The ratio Ur/Uf, which closely resembles Ur/Ue, is also used as a measure of elastic recovery.16 67.3.4.2 Stress–Strain Curves In the stress–strain mode the total deformation of the skin is measured in function of vacuum during a linear increase of vacuum, followed by a linear decrease of suction (Figure 67.3). Generally, nonlinear curves are obtained with hysteresis. The stress–strain curve with loading will not be superposed with the unloading curve. Furthermore, the values of strain in the relaxation procedure do not return to the origin, and will return to zero values only after a long period of relaxation.
Suction Chamber Method for Measurement of Skin Mechanics: The New Digital Version of the Cutometer
587
E(mm)
Amplitude/Time 0.54 0.52 0.5 0.48 0.46 0.44 0.42 0.4 0.38 0.36 0.34 0.32 0.3 0.28 0.26 0.24 0.22 0.2 0.18 0.16 0.14 0.12 0.1 0.08 0.06 0.04 0.02 0
Cutometer surface results by repetition
F3
F4
0
1
2
3
4
5
6
7
8
9
10 t(s)
11
12
13
14
15
16
17
18
19
20
FIGURE 67.5 Graphical representation of a repetitive strain–time curve obtained for forearm skin with the MPA 580 Cutometer. Aperture, 2 mm; 10 cycles; time of application, 1 sec; relaxation time, 1 sec. Skin deformation in millimeters.
Strictly speaking, because of the nonlinearity of the stress–strain curves for human skin, the modulus of Young is not applicable. One can always define a coefficient of elasticity that corresponds to the slope of the stress to the strain curve at a given stress value.17 In the experimental ascending curves (Figure 67.3), a linear part in the stress–strain mode is clearly present between 150 and 500 mbar. From the linear part of the curves a modulus of Young (E) can be in principle calculated.1,18 For the practical calculation of the modulus of Young, a theoretical model for the deformation of the skin in the suction aperture as proposed by Barel et al.7 and Agache et al.19 can be used, allowing the calculation of the strain and the stress. In this theoretical model one assumes that the initial flat surface of the circular test area is transformed by suction in a curved surface of a segment of a sphere. 67.3.4.3 Reproducibility of Skin Deformation Parameters and of the Modulus of Young Repetitive Uf deformation measurements without skin pretension carried out on the same person show a coefficient of variability ranging from 4 to 6%, depending on the skin sites to be examined. With skin pretension the coefficient of variability is generally lower (around 4%). Similar results were obtained by other researchers when using the suction method on different skin sites.15,16 Similar repetitive measurements carried out on
the same individual in the stress–strain mode show a coefficient of variability for the modulus of Young around 10%.
67.4 A SHORT OVERVIEW OF THE RESULTS 67.4.1 SINGLE STRESS–TIME CURVES The stress–strain and strain–time curves measured on different skin sites (forearm, face, thigh, etc.) by different researchers,7,20–25 using the suction method, are in good agreement with the results obtained by other methods (tensile, torsional, elevation, and indentation).1 It is obvious that the deformation parameters (Ue, Uf, and Ur) vary in function of the load (vacuum) and the aperture of the suction probe. With the 2-mm suction probe, typical skin deformation data (Uf), ranging from 0.1 to 0.6 mm, are recorded in function of vacuum (from 100 to 500 mbar). These values of vertical skin elevation correspond typically to deformations of the epidermis and dermis, with perhaps some contribution of the hypodermis when large deformation values are measured. The values obtained for the elasticity ratio Ur/Ue are functions of anatomical sites, age, and other physiological factors. However, typical values of the elasticity ratio ranging from 0.4 to 0.9 are recorded for the different anatomical skin sites. These elasticity recovery values are
588
Handbook of Non-Invasive Methods and the Skin, Second Edition
in good agreement with the results obtained by the torsional method.14 In addition, the software calculates (Figure 67.4) the surface F0 (surface between the real curve and the value corresponding to the maximal deformation Uf when going from start of suction to stop of suction) and the surface F1 (surface between the real recovery curve and the value corresponding to the maximal recovery going from stop of suction to stop of measurement). Both surfaces reflect the viscous part (slow deformation and slow return of the skin to the original part of the viscoelastic properties of the skin) and could be considered very interesting skin parameters. These parameters are probably investigated in the cosmetic industry in order to prove the efficacy of their products, but are not published in the scientific literature.26
67.4.2 REPETITIVE STRESS–TIME CURVES Courage-Khazaka proposes in the software package to carry out 10 cycles of suction of 1 sec each (Figure 67.5). Schlangen et al.27 propose to examine the results of three longer consecutive suction release curves. The software allows calculation of the repetitive deformation vs. time curves’ different surfaces (Figure 67.5). The software calculates on this repetitive curve the following parameters: R3 is equal to the last maximum amplitude and R4 is equal to the last minimum amplitude. Starting from these values, the software calculates the surface F2 (surface between the real curve and the value corresponding to the maximal deformation R3 after 10 cycles when going from start of suction to stop of the 10 cycles), the surface F3 (surface between the two repetitive curves), and the surface F4. Once again, these surface parameters are probably investigated in the cosmetic industry in order to prove the efficacy of their products, but are not published in the scientific literature.26
67.5 FACTORS INFLUENCING THE MEASUREMENTS 67.5.1 INFLUENCE
OF
LOAD
As would logically be expected, and dependent on the opening of the suction probe, all the skin deformation parameters are nonlinearly increased in function of load (vacuum). Due to limitations of the instrument, skin deformation measurements at 400 to 500 mbar are not possible with the 8-mm suction probe. In agreement with Elsner et al.,16 the elasticity ratio and other deformation ratios are independent of load for most anatomical sites (with the exception of vulvar skin). Due to the curvature of most of the stress–strain curves, and depending on the choice of the linear part in the stress–strain curves for the calculation, the values of the modulus of Young are different and, consequently, load dependent. The values of the modulus of Young obtained in this study are of the same order of magnitude as those recently reported by Agache et al.19 In agreement with the same workers,19 the modulus of Young is more or less independent of vacuum between 150 and 500 mbar.
67.5.2 INFLUENCE OF THE DIAMETER SUCTION DEVICE
OF THE
The maximal deformation Uf of the skin increases linearly in function of the diameter of aperture of the suction device. A maximal vertical deformation of 1.2 mm can be observed with a 6-mm suction probe. It is important to mention that skin deformation measurements are not possible at vacuum loads of 400 to 500 mbar with the 8-mm suction probe.
67.4.3 STRESS–STRAIN CURVES
67.5.3 INFLUENCE PROBE
From the linear part of the stress–strain curves typical values of the modulus of Young ranging from 130 to 260 kPa were computed for the different anatomical skin sites. These in vivo values obtained by the suction method are in agreement with previous data obtained by the torsional system1,28 (modulus of Young = 42 kPa) and the suction method29,30 (modulus of Young = 129 kPa). As pointed out previously by Piérard,1 the determination of the extent of the linear part of the stress–strain curves is not always obvious and is generally rather subjectively determined by the researchers. This explains the much larger variations generally observed in the reported Young’s modulus. As a consequence, depending on the values of the applied stress and the mechanical system used in the study, the modulus of Young varies in a large range from 104 to 106 N/m2.1
The vertical deformation of the skin is recorded by the optical measuring system along a well-defined direction, which is indicated on the surface of the measuring probe. As a consequence, the eventual anisotropy in the mechanical properties of the skin could be evaluated by measuring the vertical deformation of the skin under different orientations of the probe. The mechanical properties of the forearm skin were evaluated along two perpendicular directions: parallel and perpendicular to the primary lines of the skin microtopography of the skin surface on the forearm. No significant differences were observed in our laboratory for the elasticity ratio Ur/Ue and for the modulus of Young when measurements were carried out parallel and perpendicular to the primary lines (preferential directions of the skin surface pattern) in young and middle-aged individuals.31,32
OF THE
ORIENTATION
OF THE
Suction Chamber Method for Measurement of Skin Mechanics: The New Digital Version of the Cutometer
These results seem to indicate that the suction method measures in an isotropic way the vertical deformation of the skin located at the forearm. In all experiments it is recommended to maintain the probe exactly vertical to the skin surface.
67.5.4 INFLUENCE OF PRESSURE OF APPLICATION ON THE SKIN With the spring system, the handheld probe is always applied with constant pressure on the skin surface. However, in order to reduce as much as possible small variations in pressure of application, we have found it more convenient to carry out the experiments under the following experimental setup: the skin surface to be measured is always placed in a horizontal position and the suction probe is applied with the help of a stable probe holder.
67.5.5 PRECONDITIONING
OF THE
SKIN SURFACE
The skin adjacent to the aperture of the probe is immobilized by a guard ring in order to reduce as much as possible lateral displacement of the skin toward the opening of the suction device, or can be firmly immobilized by the use of a double-sided sticking ring. The use of the sticking ring provokes an alteration of the skin surface by removal of corneocytes from the horny layer. In the strain–time mode it is possible to pretension the skin by applying a suction during a short time before the real deformation measurements are carried out (pretime setting on the instrument). It has been shown in our laboratory7 that under pretension, the values of the skin deformation parameters are more reproducible and more accurate. Measurements carried out at different skin sites (forearm, forehead, and crow’s-feet) show that with pretension the elastic recovery parameter Ur/Ue is systematically higher (typically for forearm skin in young individuals, Ur/Ue values changes from 0.82 without pretension to 0.89 with pretension). This result indicates that with preconditioning the skin regains to a greater extent the initial position after deformation.
67.6 APPLICATIONS The new digital version of the suction method is well suited, thanks to the versatility of the measurements, to study in vivo the fundamental viscoelastic properties of the dermis in normal and diseased skin. In addition to this fundamental approach of the properties of the dermis, the influence of various factors such as sex, normal and actinic aging, and anatomical skin sites can be readily evaluated by this technique. The use of the analog version of the Cutometer (SEM 575) in order to investigate the influence of aging,
589
anatomical skin sites, and of sex has been described by many authors.3,15,16,21–25,32,33 Furthermore, the efficiency of various dermatocosmetic treatments (topical cosmetic applications and treatment of various deseases) can be evaluated quantitatively by the suction method.34,35 Review articles have been published,7–11 and these topics will also be reviewed in Chapter 66 of this book.12
67.7 CONCLUSIONS Due to improvements in the optical measuring system and a new analog electronic system, the resolution (2 mm) and the accuracy (3%) are improved. The deformation curves show a continuous tracing compared to the stepwise tracing of the old version. The digital version of the Cutometer MPA allows us to measure in a simple way in vivo the viscoelastic properties of the skin. Under well-controlled experimental conditions where various parameters such as load (vacuum), aperture of the suction device, position and pressure of application of the probe, time of application and relaxation, and pretension of the skin are kept constant, reproducible and accurate stress–strain and strain–time curves are obtained that give quantitative information concerning the purely elastic and viscoelastic properties of the dermis. These parameters are used to study the properties of various anatomical skin areas in normal and diseased skin situations. Finally the influence of physiological parameters such as aging, anatomical skin sites, and the efficacy of topical dermatocosmetic treatments can be quantitatively examined by this suction method.
REFERENCES 1. Piérard, G., A critical approach to in vivo mechanical testing of the skin, in Cutaneous Investigation in Health and Disease, Noninvasive Methods and Instrumentation, Lévêque, J.L., Ed., Marcel Dekker, New York, 1989, chap. 10. 2. Larrabee, W., A finite element model of skin deformation, Laryngoscope, 96, 399, 1986. 3. Daly, C.H. and Odland, G.F., Age-related changes in the mechanical properties of human skin, J Invest Dermatol, 73, 84, 1979. 4. Vogel, H.G., Age dependence of mechanical and biochemical properties of human skin, Bioeng Skin, 3, 141, 1987. 5. Aubert, L., Anthoine, P., de Rigal, J., and Lévêque, J.L., An in vivo assessment of the biomechanical properties of human skin modifications under the influence of cosmetic products, Int J Cosmet Sci, 7, 51, 1985.
590
Handbook of Non-Invasive Methods and the Skin, Second Edition
6. Silver, F.H., Siperko, L.M., and Seehra, G.P., Mechanobiology of force transduction in dermal tissue, Skin Res Technol, 9, 3, 2003. 7. Barel, A.O., Lambrecht, R., and Clarys, P., Mechanical function of the skin: state of the art, in Skin Bioengineering Techniques and Applications in Dermatology and Cosmetology. Current Problems in Dermatology, Elsner, P., Barel, A.O., Berardesca, E., Gabard, B., and Serup, J., Eds., Karger, Basel, 1998, p. 69. 8. Barel, A.O., Clarys, P., and Gabard, B., In vivo evaluation of the hydration state of the skin: measurements and methods for claim support, in Cosmetics. Controlled Efficacy Studies and Regulation, Elsner, P., Merk, H.F., and Maibach, H.I., Eds., Springer-Verlag, Berlin, 1999, p. 57. 9. Rodrigues, L., The in vivo biomechanical testing of the skin and the cosmetological efficacy claim support: a critical overview, in Cosmetics. Controlled Efficacy Studies and Regulation, Elsner, P., Merk, H.F., and Maibach, H.I., Eds., Springer-Verlag, Berlin, 1999, p. 197. 10. Berndt, U. and Elsner, P., Hardware and measuring priciple: the Cutometer, in Bioengineering of the Skin. Skin Biomechanics, Elsner, P., Berardesca, E., Wilhelm, K.P., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 2002, p. 91. 11. Agache, P. and Varchon, D., Exploration fonctionnelle mécanique, Dans: Physiologie de la peau et explorations fonctionnelles cutanées, Agache, P., Ed., Editions Médicales Internationales, Cachan, France, 2000, p. 423. 12. O’goshi, K.I., Chapter 66, Suction Chamber Method for Measurement of Skin Mechanics: The Cutometer®, this volume. 13. Courage, W. and Khazaka, G., Hardware and software differences between the Cutometer SEM 575 and MPA 580, technical information, Courage-Khazaka Electronic GmbH, Köln, Germany, 2004. 14. Escoffier, C., de Rigal, J., Rochefort, A., Vasselet, R., Lévêque, J.L., and Agache, P., Age-related mechanical properties of human skin: an in vivo study, J Invest Dermatol, 93, 353, 1989. 15. Cua, A.B., Wilhelm, K.P., and Maibach, H.I., Elastic properties of human skin: relation to age, sex, and anatomical region, Arch Dermatol Res, 282, 283, 1990. 16. Elsner, P., Wilhelm, D., and Maibach, H.I., Mechanical properties of human forearm and vulvar skin, Br J Dermatol, 122, 607, 1990. 17. Manschot, J.F. and Brakkee, A.J., The measurement and modelling of the mechanical properties of human skin in vivo. I. The measurement. II. The Model, J Biomech, 19, 511, 1986. 18. Piérard, G.E. and Lapière, C.M., Structures et fonctions du derme et de l’hypoderme, in Précis de cosmétologie dermatologique, Pruniéras, M., Ed., Masson, Paris, 1981, chap. 2. 19. Agache, P., Varchon, D., Humbert, P., and Rochefort, A., Non-Invasive Assessment of Biaxial Young’s Modulus of Human Skin In Vivo, paper presented at the 9th International Symposium on Bioengineering and the Skin, Sendai, Japan, October 19–20, 1992.
20. Barel, A.O. and Clarys, P., Noninvasive Measurements of the Viscoelastic Properties of the Human Skin with the Suction Method, paper presented at the 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 13–16, 1990. 21. Malm, M. and Serup, J., In Vivo Skin Elasticity of Different Body Regions: The Vertical Vector, paper presented at the 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 13–16, 1990. 22. Anfossi, T., Bosio, D., and Emanuelle, G., Influence of Environment Factors on Skin Elastometric Patterns, paper presented at the 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 13–16, 1990. 23. Barbanel, J.C., A Suction Method for Determining the Direction of Langer’s Lines, paper presented at the 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 13–16, 1990. 24. Nishimura, M. and Tsuji, T., Measurements of Skin Elasticity with a New Suction Device: Relation to Age, Sex and Anatomical Regions in Normal Skin and Its Comparison with Some Diseased Skin, paper presented at the 9th International Symposium on Bioengineering and the Skin, Sendai, Japan, October 19–20, 1992. 25. Barel, A.O. and Clarys, P., In Vivo Evaluation of Skin Ageing. Relations between Visco-Elastic Properties and Skin Surface Parameters, paper presented at the 9th International Symposium on Bioengineering and the Skin, Sendai, Japan, October 19–20, 1992. 26. Courage-Khazaka, scientific information obtained from G. Khazaka, Courage-Khazaka Electronic GmbH, Köln, Germany, 2004. 27. Schlangen, L.J.M., Brokken, D., and Van Kemenade, P.M., Correlations between small aperture skin suction parameters: statistical analysis and mechanical model, Skin Res Technol, 9, 122, 2003. 28. Agache, P., Monsieur, C., Lévêque, J.L., and de Rigal, J., Mechanical properties of Young’s modulus of human skin in vivo, Arch Dermatol Res, 269, 221, 1980. 29. Hendriks, F.M., Brokken, D., Van Eemeren, J.T.W.M., Baaijens, F.P.T., and Horsten J.B.A.M., A numericalexperimental method to characterize the non-linear mechanical behaviour of human skin, Skin Res Technol, 9, 274, 2003. 30. Diridollou, S., Patat, F., Gens, F., Vaillant, L., Black, D., Lagarde, J.M., Gall, Y., and Berson, M., In vivo model of the mechanical properties of the human skin under suction, Skin Res Technol, 6, 214, 2000. 31. Van Den Eynde, A., Invloed van de leeftijd op de viscoelastische eigenschappen van de huid, B.Sc. thesis, Vrije Universiteit, Brussels, Belgium, 1990. 32. VanWonterghem, M., Mechanische eigenschappen van de menselijke huid: invloed van verschillende factoren, B.Sc. thesis, Vrije Universiteit, Brussels, Belgium, 1991. 33. Dobrev, H., Use of Cutometer to assess epidermal hydration, Skin Res Technol, 6, 239, 2000.
Suction Chamber Method for Measurement of Skin Mechanics: The New Digital Version of the Cutometer
34. Berardesca, E., Borroni, G., Borlone, R., and Rabbiosi, G., Evidence for elastic changes in aged skin revealed in an in vivo extensometric study at low loads, Bioeng Skin, 2, 261, 1986.
591
35. Sparavigna, A. and Galbiati, G., Strain-Time Curve in the Assessment of Topical Tretinoin as an Antiageing Agent, presented at the 8th International Symposium on Bioengineering and the Skin, Stresa, Italy, June 13–16, 1990.
Chamber Method for 68 Suction Measurement of Skin Mechanics: The DermaLab Gary Lee Grove, John Damia, Mary Jo Grove, and Charles Zerweck cyberDERM, Inc., Broomall, Pennsylvania
CONTENTS 68.1 Introduction............................................................................................................................................................593 68.2 Basic Description of the DermaLab Suction Cup Hardware ...............................................................................593 68.3 The DermaLab Suction Cup as a Noncomputerized Stand-Alone Device ..........................................................595 68.4 The Computerized DermaLab Suction Cup with cyberDERM Software ............................................................596 68.5 Validation Study of the Computerized DermaLab Suction Cup ..........................................................................596 68.6 Effects of Repetitive Cycles ..................................................................................................................................597 68.7 Typical Results from Studies of Human Volunteers.............................................................................................598 References .......................................................................................................................................................................599
68.1 INTRODUCTION Whether the skin feels soft, supple, compliant, firm, etc., to the touch is directly related to its mechanical properties. Probably the most widely recognized change in the mechanical properties of the skin is its age-related loss of elasticity. It is also generally appreciated that more subtle changes, such as increased skin stiffness, can provide important clinical clues for monitoring the progression of certain systemic diseases such as scleroderma. Thus, it is not surprising that over the years, a wide variety of devices have been created to noninvasively describe the mechanical properties of the skin. Despite obvious differences in design and execution, the underlying principles of all of these devices are generally much the same, i.e., load the skin surface in a standard manner and measure the resulting deformations over time. Thus, we have devices that pull, push, tug, twist, compress, wiggle, and impact the skin. In this chapter we will describe the DermaLab suction cup, which is manufactured by Cortex Technology (Hadsund, Denmark). Some aspects of an earlier version of this device have been previously covered by Serup1 and Pedersen et al.2 Since that time a number of minor improvements have been made to the basic device by Cortex Technology, and a dedicated application program has been written by cyberDERM, Inc. (Broomall, PA) that allows
the DermaLab suction cup to be computerized. Both the stand-alone and the computerized versions will be reviewed in this chapter.
68.2 BASIC DESCRIPTION OF THE DERMALAB SUCTION CUP HARDWARE The DermaLab suction cup consists of a light plastic probe (Figure 68.1) that forms a closed chamber when attached to the skin surface using double-sided sticky tape. Within the probe chamber there are two narrow beams of light that are run at different heights parallel to the skin surface and serve as elevation detectors. A computer-controlled vacuum pump is used to progressively increase the suction within the chamber. Since the time at which each of the light beams is blocked can be electronically detected, the amount of suction in kilopascals (kPa) required to lift the skin to that point can be easily determined and electronically recorded by the computer. Figure 68.2 shows a schematic diagram consisting of four panels that portray the sequence of events that occurs during a measurement procedure. When the probe is first placed on the skin, its surface will be flush across the opening of the suction chamber, as shown in Figure 68.2a. The progressive increase in suction will cause the skin to be drawn into the chamber (Figure 68.2b), and eventually 593
594
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 68.1 The lightweight plastic probe of the DermaLab suction cup.
the skin will be lifted to the point where the light beam of the lower elevation detector (level 1) will be blocked, as shown in Figure 68.2c. When this occurs, the vacuum at that point will be electronically recorded. If the pump is allowed to continue, sufficient suction will be created to additionally lift the skin to the point where the light beam of the upper elevation detector (level 2) is blocked as well, as shown in Figure 68.2d. The amount of suction required to achieve this higher level is also electronically recovered. Since the positions of the lower and upper elevation detectors are fixed by the geometry of the probe, the strain at each level is known. In the standard probe these are equivalent to the skin being stretched to 2 and 12% extensions, respectively. The suction pressure in kilopascals
To vacuum pump
Suction cup
Elevation detectors
Skin
required to lift the skin to each point provides a measure of the stress at that level. This means that the DermaLab suction cup can calculate the mechanical properties of the skin based on Hooke’s law,3 which was first formally stated by the English mathematician in 1660 as an anagram in Latin. Since no one managed to break his code, years later he translated the phrase to be “The power [sic] of any springy body is in the same proportion with the extension.” Today, Hooke’s law3 in its simplest form is generally stated that “the strain of any material is proportional to the load applied to it.” This means that if an applied tensile stress of x units will stretch a specimen of 1 unit, then a stress of 1.5x will produce an elongation of 1.5 units, a stress of 2x will produce a deformation of 2 units, and so on. The factory manual states that if certain basic assumptions are made, then the stiffness of the skin or its Young’s modulus (E) can be calculated from this stress–strain relationship as follows:
Δx = ψ ⋅ p ⋅
where x = deviation middle of surface, ψ = elasticity constant for measured object estimated from civil engineering tables, p = negative pressure, r = radius of surface measured, and s = thickness of object measured.
Elevation detectors
Skin
(c)
To vacuum pump
Elevation detectors
Skin
To vacuum pump
Suction cup
(a)
Suction cup
r4 Δp ⇒ E = 0.3125 ⋅ E ⋅ s3 Δx
(b)
FIGURE 68.2 Operating principles of the DermaLab suction cup.
Suction cup
To vacuum pump
Elevation detectors
Skin
(d)
Suction Chamber Method for Measurement of Skin Mechanics: The DermaLab
595
TABLE 68.1 Displayed Parameters When DermaLab Suction Cup Is Used as a Stand-Alone Device
ELAST RES1
RESN TIM1 TIMN
Description
Comments
The calculated elasticity modulus based on the first measurement cycle The differential vacuum necessary to elevate the skin from detection level 1 to detection level 2 on the first suction cycle
May not be appropriate for human skin measurements Very useful in describing stiffness of skin, but actual value depends upon specific probe configuration Same as RES1 Misleading due to suction curve not being linear Same as TIM1
Same as RES1 but on last suction cycle The time required for the skin to be raised to detection level 2 on the first suction cycle Same as TIM1 but on last suction cycle
In these calculations ψ is set by the manufacturer to be 0.5, r is defined by the chamber geometry as 5 mm, and s is set by the manufacturer to a standard skin thickness of 1.0 mm. We agree with Serup1 that the calculation of a Young’s modulus in the above manner is not normally considered to be appropriate for complex composite structures such as the skin. Not only is the skin highly anisotrophic and viscoelastic, but it is also composed of various layers, each of which has a different mechanical resistance. Moreover, the actual values for both (elasticity constant) and s (skin thickness) are not known and can vary considerably, depending upon the age of the patient and the anatomical region being measured. Nevertheless, we feel that this parameter must be discussed in this chapter since it is the ELAST parameter that is provided as part of the printed output of the DermaLab suction cup when used in the noncomputerized stand-alone mode.
68.3 THE DERMALAB SUCTION CUP AS A NONCOMPUTERIZED STAND-ALONE DEVICE Table 68.1 provides a listing of the various parameters that appear on the liquid crystal display of the DermaLab when used as a stand-alone instrument not interfaced to a PC. These include ELAST, which refers to the calculated elasticity modulus (E), which in our opinion should be considered to be of only limited value due to the reasons described above. We also have some reservations as to how useful the two elevation time values (TIM1 and TIMN) can be. These are the time to reach the upper detector at level 2 on the first cycle (TIM1) and the last or nth cycle (TIMN). The DermaLab suction cup does not have a vacuum tank or barometric control of the buildup of negative pressure. Instead, this device is designed with a specially constructed tube system so that the vacuum is slowly applied over a period of perhaps 30 to 60 seconds, until a plateau
is reached where the negative pressure within the chamber is maintained in equilibrium with the capacity of the pump. With stiffer skin it will take longer for this equilibrium to be achieved, which is the notion underlying the use of the elevation times TIM1 and TIMN. In the more recent version of the DermaLab suction cup, a switch has been provided to further control the flow rate of the specialized vacuum system. For most skin sites, such as the arms and legs, the switch should be set at NORMAL, which is the default setting. With skin that is very loose, such as the face and neck, it is recommended to use the REDUCED setting, which according to the factory manual reduces the rate at which the suction pressure is applied. At the time of the original paper by Serup1 the actual shape of the negative pressure curve in the chamber was not exactly known but was supposed to be nonlinear. We have been able to acquire the profile of the negative pressure curve by incorporating a digital manometer into the suction line and placing the suction cup probe on a rigid, nondeformable surface. As shown in Figure 68.3, the time required to reach equilibrium is approximately the same, 60 Suction pressure in KPa
Parameter
Normal 50 40 30 Reduced 20 10 0
0
5
10
15 20 25 30 Seconds with pump on
35
40
45
FIGURE 68.3 Negative pressure curves developed by DermaLab suction pump at either normal or reduced settings.
Handbook of Non-Invasive Methods and the Skin, Second Edition
but the level at which this occurs is considerably lower for the REDUCED setting than for the NORMAL setting. Moreover, it is quite clear that the shape of both curves is clearly nonlinear. This means that although a stiffer material will take longer to be stretched to the upper level, there is no simple relationship between that time and the stiffness of that material. In other words, a doubling in elevation time does not mean a doubling in skin stiffness. Indeed, the shape of the pressure curve clearly indicates that as time increases, the degree to which the actual stiffness is overestimated by relying on the elevation time values will increase as well. This leaves only RES1 and RESN as possible measures of stiffness that are available when the DermaLab is used as a stand-alone instrument. Both values indicate the differential vacuum necessary to elevate the skin from detection level 1 to detection level 2 in either the first or the nth and last cycle. Since these values are central to the computerized version of the DermaLab, we will deal with them in more detail in the next section.
68.4 THE COMPUTERIZED DERMALAB SUCTION CUP WITH CYBERDERM SOFTWARE As is the case with all of the DermaLab modules, the suction cup can also be configured to run in continuous mode, so that a stream of data flows via an RS-232 interface to a PC where it can be processed using specialized data acquisition programs. Actually, two data outputs are provided at a sampling rate of 20 per second, with the first being the status of the lower-level detector (ON or OFF) and the second being the negative pressure within the chamber at that point in time in kilopascals. The pressure information can be plotted in real time on a strip chart recorder so that the operator can watch the development of the suction curve in the suction chamber. By monitoring the status of the lower-level detector, the pressure developed at the point in time when the light beam of the elevation detectors is broken can easily be captured. Since the pressure pump is automatically turned off when the upper detector light beam is broken, there is no need to monitor the status of this detector, because the pressure profile will drop off immediately. Thus, it is a relatively simple matter to compute the differential vacuum necessary to elevate the skin from the lower detector beam at level 1 to the upper detector beam at level 2, which is equivalent to RES1 or RESN of the stand-alone instrument. Again, RES1 and RESN refer to values that are respectively derived from either the first or the nth and last cycle. Moreover, it is also a simple matter to provide the actual stress values in terms of pressure at both detector levels, where the strain is known due to the probe’s geometry. In the case of the standard probe, this is equiv-
alent to 2% extension at the lower level and 12% extension at the higher level. Although it would also be possible to easily extract the time required to reach the upper detector level from the time base of the pressure curve, to yield an output identical to TIM1 or TIMN of the stand-alone device, we do not think that measurements based on this parameter are easy to compare due to the nonlinear pressure curve, as previously discussed.
68.5 VALIDATION STUDY OF THE COMPUTERIZED DERMALAB SUCTION CUP We would like to emphasize that the underlying physical principle of the DermaLab suction cup is Hooke’s law.3 To determine how well the computerized device adhered to Hooke’s law,3 we used a 0.012-inch-thick latex sheet that could be stretched to different degrees before the probe was attached. By marking two reference lines 2 inches apart upon the sheet, with no tension upon it, one could achieve extensions of 25, 50, 75, and 100% by stretching these lines so that the gaps between the reference lines were 2.5, 3.0, 3.5, and 4.0 inches, respectively. The DermaLab suction cup probe could then be reattached to the latex sheet that has been stretched to various degrees. Figure 68.4 displays the results obtained for the standard probe. Note that the pressure values recorded for both the upper and lower detectors are highly correlated with the degree of stretch, which is in perfect accordance with Hooke’s law,3 with R2 being nearly 1 in both cases. It is also clear that the pressure differential between the two detectors remains constant Standard suction cup 40 35
y = 0.1665x + 20.028 R2 = 0.9986
30 25 KPa
596
20 15 10
y = 0.1452x + 3.1117 R2 = 0.987
5 0
0
10
20
30 40 50 60 70 80 Percent extension of latex sheet
90
100
FIGURE 68.4 Scatter plot of the amount of suction required to lift a reference latex sheet to block either the lower or upper level detector light beams, with the sheet stretched to various degrees. Note that results are in excellent agreement with those predicted by Hooke’s law.
Suction Chamber Method for Measurement of Skin Mechanics: The DermaLab
597
“Sceleroderma” suction cup 40
35
35
30
30
25 20 15
20
y = 0.1847x + 9.44 R2 = 0.9861
15
10
10
y = 0.1506x + 1.7517 R2 = 0.9967
5 0
y = 0.1753x + 21.175 R2 = 0.9929
25
y = 0.1554x + 11.513 R2 = 0.9936
KPa
KPa
“Facial” suction cup
40
0
10
20
30 40 50 60 70 80 Percent extension of latex sheet
90
5
100
0
0
10
20
30 40 50 60 70 80 Percent extension of latex sheet
90
100
FIGURE 68.5 The results for the scleroderma and facial probes are also in excellent agreement with those predicted by Hooke’s law.
throughout this range of extension, which justifies the use of the RES1 or RESN value as a measure of the inherent stiffness of that material, provided that we remain within the linear portion of the stress–strain curve, which is clearly true in this case. It should be pointed out that special probe configurations exist. In the scleroderma version, the upper and lower elevation detectors are moved closer to the skin surface, which allows one to measure tight skin without excessive stretching. In the facial version, the lower elevation detector is moved farther away, since the skin may be so loose and soft that it may have already moved into the chamber and blocked the lower detector before the suction is applied. As shown in Figure 68.5, regardless of the configuration, Hooke’s law3 is still obeyed. Indeed, with this type of graphic presentation it is easy to appreciate that the upper elevation detector of the scleroderma probe and the lower elevation detector of the facial probe are in approximately the same relative position above the skin surface, as shown by the fact that their plots are quite similar. Again, note that the pressure differential between the lower and upper detectors for both probes remains constant through this range of extensions, as was the case for the standard probe. Although the pressure differentials differ among the various probes, this is due to the spacing of the lower and upper elevation detectors. Indeed, rather than computing a Young’s modulus, we think it more appropriate to describe the results obtained with the DermaLab suction cup in terms of a stiffness index, which is simply delta pressure in KPa/delta distance in mm
This means that skin that is firm and taut will have a much higher stiffness index than skin that is loose and saggy. In this validation study, the material being studied with the different probe configurations was the same latex sheet. Although the positions of the lower and upper elevation detectors in the standard, scleroderma, and facial suction cup probes do differ considerably, as shown in Table 68.2, the stiffness index for the latex sheet was the same. Moreover, if we increase the thickness of the latex sheet being measured, there will be a corresponding increase in the stiffness index, as one would predict. Although the focus of this chapter is skin biomechanics, one should also realize that the DermaLab suction cup can be easily employed to measure the material properties of various elastic sheets in a meaningful manner.
68.6 EFFECTS OF REPETITIVE CYCLES With the DermaLab suction cup one can program the suction pump to do either a single on–off cycle or a series of repetitive cycles with an intervening resting time, which
TABLE 68.2 Stiffness Index of Latex Sheet Reference Standard as Measured with Three Different Suction Cup Configurations Probe Type Standard Scleroderma Facial
Mean
±
S.D.
12.07 12.44 12.18
± ± ±
0.66 1.28 0.45
598
Handbook of Non-Invasive Methods and the Skin, Second Edition
40
40 Latex
30 25 20 15 10 5
Skin
35
Level 2 Level 1 Suction pressure in KPa
Suction pressure in KPa
35
Level 2 Level 1
30 25 20 15 10 5
0
0 1
2
3 4 Number of cycles
5
1
2
3 4 Number of cycles
5
FIGURE 68.6 Effects of repetitive cycling on the amount of suction required to stretch latex or skin to block the elevation detectors at levels 1 and 2 with the DermaLab suction cup.
can be set to be from 1 to 10 seconds. The program can be set so that the device will automatically stop cycling after a predetermined number of cycles have been completed. During the validation studies we found that repetitive measurements of the latex sheet typically gave the same values for each cycle regardless of the number of cycles, as shown in Figure 68.6. In striking contrast, when measuring human volunteers we found that the suction required to lift the skin progressively decreases with each cycle, but will eventually reach an asymptote. This warming-up phenomenon is more or less obvious at different anatomical sites and in some individuals. Currently we are investigating how to best express these results and what factors are responsible for this type of behavior. Such studies can only be done with the computerized version of the device, which provides a complete data set for each and every cycle over time. With the stand-alone version only the information of the first and last cycles is provided, and these values provide only the pressure differentials (RES1 and RESN), not the individual stresses for levels 1 and 2 for each cycle.
68.7 TYPICAL RESULTS FROM STUDIES OF HUMAN VOLUNTEERS Although the probe is very lightweight (approximately 10 g), the tethered wires and pneumatic tubing, if not properly supported, will tug on the skin and alter the biomechanical properties being measured. This is extremely important when attempting to measure where the skin is lax, such as under the eye (Figure 68.7).
FIGURE 68.7 Although the probe is lightweight, it is important, especially when the skin is lax, to support the probe so that it does not tug on the skin.
Figure 68.8 summarizes the results from a small crosssectional study involving 20 healthy normal individuals, with half of them between 20 and 30 years and the other half between 50 and 60 years of age. Striking differences were found to exist at various regions of the face, especially under the eye. The facial skin of older individuals was typically less stiff than that of younger individuals in these regions, due to loss of elasticity and increased sagging. We have also found that the DermaLab suction cup provides clinically relevant data on the mechanical properties of the skin, which may help predict the severity and progression of a number of diseases, such as scleroderma. We are especially impressed with how well the DermaLab suction cup has held up under hard use in such clinical trials. A large part of its robustness stems from there not being any moving parts in the probe.
Suction Chamber Method for Measurement of Skin Mechanics: The DermaLab
REFERENCES
40 35 20–30 years
“Stiffness index”
30
50–60 years 25 20 15 10 5 0
599
Forehead
Cheek
Under eye
FIGURE 68.8 Cross-sectional survey in which the stiffness index was measured at various regions of the face of younger and older adults with a computerized DermaLab suction cup.
1. J. Serup. Hardware and measuring principles: DermaLab. In Bioengineeering of the Skin: Skin Biomechanics, P. Elsner et al, Eds. CRC Press, Boca Raton, FL, 2002 pp. 117–121. 2. L. Pedersen, B. Hansen, and G.B.E. Jemec. Mechanical properties of the skin: a comparison between two suction cup methods. Skin Res Technol 9: 111–115, 2003. 3. www.RobertHooke.com.
Measurement of 69 Twistometry Skin Elasticity Pierre G. Agache Department of Functional Dermatology, University Hospital, Besançon, France
CONTENTS 69.1 Introduction............................................................................................................................................................601 69.2 Technical and Theoretical Considerations ............................................................................................................602 69.2.1 Equipment ..................................................................................................................................................602 69.2.2 Mechanical Testing ....................................................................................................................................602 69.2.3 Mechanical Parameters..............................................................................................................................603 69.2.3.1 Elasticity Parameters ..................................................................................................................603 69.2.3.2 Viscosity Parameters ..................................................................................................................604 69.2.3.3 Skin Rheological Model.............................................................................................................604 69.2.4 Requirements for Test Validity and Correct Interpretation.......................................................................605 69.2.4.1 Geometry of the Applied System ..............................................................................................605 69.2.4.2 Absence of Pressure Perpendicular to Skin Surface .................................................................605 69.2.4.3 Attachment of Disc to the Skin .................................................................................................605 69.2.4.4 Rate of Torque Application ........................................................................................................605 69.3 Skin Mechanical Aging .........................................................................................................................................605 69.3.1 Intrinsic Aging ...........................................................................................................................................605 69.3.2 Actinic Aging.............................................................................................................................................607 69.4 Stratum Corneum and Skin Biomechanics ...........................................................................................................608 69.5 Medical Applications .............................................................................................................................................609 69.5.1 Effect of Topical Retinoic Acid.................................................................................................................609 69.5.2 Scleroderma ...............................................................................................................................................609 69.5.3 Inherited Connective Tissue Diseases .......................................................................................................610 69.6 Conclusion .............................................................................................................................................................610 Acknowledgment.............................................................................................................................................................610 References .......................................................................................................................................................................610
69.1 INTRODUCTION For the two last decades, numerous attempts have been made to non-invasively assess the human skin mechanical behavior in vivo. This was felt to be a useful way to allow a more precise follow-up of the numerous diseases or skin states characterized by an abnormal skin induration of softening and to estimate the efficacy of treatments or cosmetics. To attain this goal, devices were constructed that operate either by inducing a skin deformation and recording the resisting force or by putting a load and assessing the resulting deformation. Four directions of loads are conceivable: a vertical pressure, a vertical suction, a linear horizontal traction, and a torsion in the horizontal plane.
Mechanical stimuli perpendicular to the skin surface have the disadvantage of involving the subcutis, at least in part. This layer has wide differences in thickness or in fat content, and consequently has widespread mechanical properties. Stimuli in the horizontal plane, on the other hand, use probes stuck on the skin surface and, if the displacement is small, may be expected to involve only the epidermis and dermis. Unidirectional stresses should consider the skin mechanical anisotropy. But Langer’s lines are not easy to find, and most of them are oblique to the body or limb axis, and thereby difficult to comply with.1 Also, the sampled area is poorly delimited with such devices. All these reasons prompted some authors to use torsional devices made of a central rotating disc and a 601
602
Handbook of Non-Invasive Methods and the Skin, Second Edition
peripheral fixed ring, thus allowing a skin narrow annulus to be twisted. Vlasblom2 from Utrecht was the first to use such a device and made a theoretical study of forces and deformations implicated. Finlay3,4 from the Strathclyde group (Glasgow) extended the investigation and applied the technique to the assessment of changes in mechanical behavior with aging. Since then, the L’Oreal group (Paris), conducted by Jean-Luc Lévêque, has been using torsional devices extensively and has contributed a great deal to the development of the technique.
69.2 TECHNICAL AND THEORETICAL CONSIDERATIONS Before using or interpreting the results of a torsional test, some elements of the relationships between stress and deformation should be recalled, as well as the requirements necessary for an experiment to be mechanically valid.
69.2.1 EQUIPMENT Torsional equipment acts through a disc glued to the skin, which is rotated by a motor powered by a controlled voltage, thereby loading the peripheral skin with a torque, the value of which can be adjusted. Under this torque the skin glued under the disc moves with it, supposedly without any brake from the subcutaneous tissue. The skin around the disc is elongated in a twisting way. This is the mechanical behavior of this part of the skin that is assessed. There are two types of twists to be considered, whether there is or is not a peripheral guard ring concentrical to the disc, also glued to the skin and immobile during the disc rotation, thereby delimiting an annular area of skin submitted to elongation (Figure 69.1). When there is no guard ring,5 the skin peripheral to the disc is implicated up to an unknown distance. As it is much less extensible than the subcutis, the latter will provide for most of the twisting strain. By contrast, when there is a guard ring the shape and limits of the twisted area are known, the sliding of the skin over the subcutis is limited, and the essential part of the twisting strain takes place within the skin itself. Finlay’s4 and Jaskowski and Maceluch’s6 devices were constructed to give repeated twists at adjustable frequencies and adjustable increasing or decreasing rates. The first one was also equipped by a strain gauge, which recorded the torque generated by the skin under a constant twist and assessed force relaxation. Lévêque and De Rigal’s7 device (Figure 69.2) has a guard ring integrated to the body of the apparatus. It delivers an analogue signal directly proportional to the disc rotational angle. In a
Guard ring
Disc
Disc
A
B
FIGURE 69.1 Skin deformation produced at the periphery of a rotating disc with (A) and without (B) a guard ring.
newer version the signal is digitalized and a microprocessor both computes the main parameters and controls the measurement phases. The applied torque can be chosen between 4 and 57 mN·m, the width of the crown of skin submitted to torque is either 1, 3, or 5 mm, and the disc radius is 18 or 25 mm. The equipment is now commercially available (Dermal Torque Meter, Diastron Ltd., Andover, U.K.).
69.2.2 MECHANICAL TESTING Finlay’s4 protocol included a progressive increase in disc rotation, then a standstill to assess the skin torque relaxation, then a progressive return of the disc position to zero. He recorded both the strain and the resisting force induced in the skin during the first run and the following ones. Jaskowski and Maceluch6 used a repeated torque application up to a resonance frequency and recorded the torque
A/D conv. C M
Current supply
P D
G
Output
FIGURE 69.2 Lévêque’s twistometer diagram. (C) Rotational sensor, (M) motor, (D) disc, (G) guard ring, (P) microprocessor. Both guard ring and disc are removable.
Twistometry Measurement of Skin Elasticity
Creep
603
Recovery
L
F
Transient
θ°
Stationary
E
I α
UR
L0
UE
(a) Time 0
60s
F
FIGURE 69.3 Skin angular deformation (Θ) vs. time upon application of a constant torque. Ue: immediate deformation, Ur: immediate recovery upon torque switching off.
oscillating amplitude, the optimum frequency, and the attenuation of torque by the skin. The simple way to use torsional tests is to apply a constant torque for a couple of seconds and record the skin angular deformation (Figure 69.3). Upon torque application there is an immediate elastic deformation (Ue), followed by a creeping viscoelastic deformation (Uv). Torque suppression is associated with an immediate recovery (Ur), which is always incomplete. This is the way Lévêque and De Rigal’s8 device works. One or several runs can be programmed and preconditioned.
69.2.3 MECHANICAL PARAMETERS In vivo mechanical tests on the skin have two purposes. The first and main one is to quantitatively assess changes that are usually detectable by palpation but not measurable otherwise. The second aim is to get access to the skin intrinsic structure as far as mechanical components are concerned, and to the structure-function relationships of these components. The absolute parameters concerning the strength of skin elasticity and viscosity need to be derived from the assumptions that it is a homogeneous layer and is uniform in thickness, which is obviously wrong, but is a very rewarding approximation. 69.2.3.1 Elasticity Parameters In simple elongation tests the well-known equation of Young’s modulus (E) is a commonly used way to express the stiffness at the elastic phase: E = σ(1 – ν)/ε
(69.1)
where σ is the stress (ratio of force to the section area submitted to force), ε is the strain (elongation-to-initial sample length ratio), and ν is the ratio of relative narrowing to strain (Poisson’s ratio). In a torsional experiment the deformation is more complex because elongation is replaced by shear and is
L E
I
α
L0 (b)
FIGURE 69.4 A diagram of skin annulus deformation in torsional experiments. (a) Deformation homogeneous through thickness. (b) Deformation predominant on the skin surface. Deformation gradients are supposed linear: (L0) initial length (width of skin annulus); (L) length after elongation; (I) internal aspect of skin annulus, facing disc; (E) external aspect of skin annulus, facing guard ring, (F) force; (α) deformation angle.
rotational. The skin can be deformed through its full thickness to the same extent (Figure 69.4). This occurs when the force is high enough to act in depth while only applied on the surface, and when the skin annulus is wide enough to allow the force gradient within the skin to fully reach the deeper layers. In that case, the skin is supposedly isotropic in torsion tests and the mechanical parameter is the shear modulus μ (Lamé’s coefficient) given by the formula μ = σ(1 – ν)/α
(69.2)
where α is the deformation angle, ν Poisson’s ratio, and σ the stress as calculated by the ratio of force (torque/radius of disc) to the area submitted to torsion (skin thickness × disc perimeter). Agache et al.9 have proposed the following formula:
μ=
M 2 πe r1 r2 Θ
(69.3)
where M is the torque momentum, r1 the disc radius, r2 the guard ring inner radius, e the skin thickness, and θ the rotational angle of the disc at equilibrium. The Lamé’s coefficient is usually close to 0.4 E; therefore, it is proposed to compute E by simply dividing the above formula by 0.4. Hence,
604
Handbook of Non-Invasive Methods and the Skin, Second Edition
E=
1 M × 0.4 2 πe r1 r2 Θ
K1
(69.4) K0
η1
If only E variation is considered, the same formula gives the following equation by differentiation:10 ΔE/E = –(ΔUe/Ue + Δe/e)
(69.5)
η0
Maxwell model Kelvin voigt model
FIGURE 69.5 Bürger’s model of skin mechanical behavior, and corresponding extension (k0, k1) and viscosity (η0, η1) parameters
When the torque is low and mostly when the skin annulus is narrow, the superficial layers of the skin are more implicated than the deeper ones because the tangential force is applied on the skin surface (Figure 69.4). In that case the deformation has two components, a visible and measurable one in the superficial plane, and an invisible and immeasurable one in the vertical plane (the shear strain gradient in depth). It could be possible, from the deformation seen at the skin surface, to calculate both of these strain components and, consequently, the relevant, more complex shear coefficient and corresponding Young’s modulus equivalent. But in the present state of experimental conditions, which are submitted to large variability in the results, obtaining such precision seems unrealistic, or at least unnecessary.
This law was also validated for the recovery part of the curve, and allowed a correct determination of the immediate recovery.
69.2.3.2 Viscosity Parameters
69.2.3.3 Skin Rheological Model
The experimental protocol designed by Finlay4 includes a typical stress–relaxation step. The observed half-time of skin torque relaxation during this step is inversely related to skin viscosity for a given initial torque. Accordingly, this time would change for another torque value at the start of the relaxation step. As in an elongation test, in torsional experiments where a constant torque is applied, the deformation vs. time, as shown in Figure 69.3, can be described by the following equation (Vlasblom):
Both for better intuitive understanding of what occurs within the skin during traction and easier computing of mechanical parameters, rheological models of skin have been proposed. Wijn12 used Burger’s model for their uniaxial elongation experiments (Figure 69.5). Pichon et al. used the same model for skin torsional experiments under a constant force and using the above-quoted creep law (Equation 69.7). Accordingly the equation for the model is
U(t) = Ue + Uv (1 – e–t/τ) + At
In an experiment on six subjects with a 1-mm-wide skin annulus, they found m = 0.33 ± 0.03 and 0.335 ± 0.06 for 9 × 10–3 and 12 × 10–3 Nm torques, respectively. Increasing the width of the annulus did not change the m value, although increasing the interindividual variation. Accordingly, they proposed the following equation for the creep: ε = Uv (1 – e–t/τ) + At1/3
ε( t ) =
(69.6)
where Ue is the immediate deformation, Uv the viscous deformation (transient creep), and At a linear deformation following a longer time of torque application (stationary creep). Vlasblom showed that in the stationary creep the deformation is not proportional to time and suggested a constant term should be added to At. As this term is small compared to elastic deformation, he thought it possible to neglect it. Pichon et al.,11 using the finite differences method, undertook to determine the creep law without any influence of Ue and proposed the following equation for the creep deformation ε: (69.7)
(
)
(69.9)
The three phases of experimental strain (Ue, transient creep, and stationary creep) correspond to each of the three terms of the equation. The characteristic parameters of the springs (k0, k1) and dashpots (η0, η1) should correspond to those of special structures or arrangements within skin, e.g., the elastic resistance of elastic fibers to elongate or collagen network to deform, and the viscous resistance to displacement. All parameters of Bürger’s model can be obtained from the in vivo experiment, as follows: •
ε = ε0 Atm
σ σ σ t1 3 + 1 – e – tk1 η1 + k 0 k1 η0
(69.8)
Stress: σ = force/area submitted to stress (i.e., disc perimeter × skin thickness)
Twistometry Measurement of Skin Elasticity
•
• • •
605
Young’s modulus E (i.e., first spring strength at the casual level of first spring tension): E = k0 = σ/Ue Second spring strength: k1 = σ/Uv curvilinear (i.e., transient creep) Viscosity associated with the first spring: η0 = σ/Uv t1/3 Viscosity associated with the second spring: η1 = τk1
This model looks satisfactory, but has been used in only one published study.11 Accordingly, in the next sections the calculations relative to creep used only the second member of Equation 69.6.
69.2.4 REQUIREMENTS FOR TEST VALIDITY CORRECT INTERPRETATION
AND
Barbenel and Payne13 in a report to the International Society for Bioengineering and the Skin have presented the requirements for a torsional testing to give interpretable and reliable results. Most of them should be recalled, along with some additional warning.
T (cm. cN) A
B
330
220
A
110
B
Θ° 5
10
FIGURE 69.6 Experimental plot of Ue (Θ°) vs. applied torque (T) under 11.25 kPa pressure (A) and under 24.52 kPa pressure (B) onto skin surface. Higher pressure induced an artifactual curve where the curvilinear relationship between stress and strain was no longer visible. (From Lévêque, J-L, et al., Arch. Dermatol. Res., 269, 221–232, 1980. With permission.)
69.2.4.4 Rate of Torque Application 69.2.4.1 Geometry of the Applied System The area submitted to torque should be delimited by a guard ring in order (1) to allow the skin itself to be deformed and reduce to a minimum the skin sliding over the subcutis, and (2) to calculate the strain (i.e., deformation/initial length). The width of the twisted skin annulus should be narrow enough to prevent torsion of the subcutis. In any case, the disc diameter and inner guard ring diameter should be quoted, and also the torque value. The rotational angle θ should be small (inferior to 10˚), and the radius r of the rotating disc should be large enough to allow the displacement rθ, which is circular, to be considered as linear. 69.2.4.2 Absence of Pressure Perpendicular to Skin Surface As shown in Figure 69.6, such a pressure would change the mechanical behavior of the skin. In practice, the applied pressure, if any, should always be mentioned. 69.2.4.3 Attachment of Disc to the Skin Slipping or deformation of the attachment system should be avoided. Accordingly, the mechanical behavior of the attachment system used should be tested beforehand at the same torque as in the experiment. This was the case for Lévêque and De Rigal’s device, as tested on a steel plate,7,8 and in Finlay’s experiments.4 Anyway, the type of adhesive used should be indicated.
As skin is viscoelastic, any application of the torque at a low rate would allow both an elastic and viscous deformation to take place at the same time. A very low rate (quasi-static experiment) would reduce the viscous resistance to almost zero, and consequently assess only the elastic resistance. On the other hand, a very high deformation rate would overcome viscous resistances lower than the applied stress, and assess the elastic resistance through the measurement of immediate deformation. In any case, the rate of applied torque should be mentioned.
69.3 SKIN MECHANICAL AGING 69.3.1 INTRINSIC AGING The skin mechanical behavior over the life span, as assessed by torsional experiments in vivo, was first studied by Finlay4 in the lateral aspect of the forearm on a 4-mmwide skin annulus using a 15-mm-diameter rotating disc, inducing on the skin an 11-kPa pressure (Figure 69.6). The run consisted of progressively twisting the skin annulus (rotation of the disc 2˚ per second) up to about 20 mN·m torque, then maintaining the same twist level for 1 min before relaxing the twist down to zero at the same rate. During such a run the torque first rises up to a peak, then subsides partially while the deformation is maintained (force relaxation), then falls to a negative value; i.e., the skin would have remained deformed if it had not been forced to come back to baseline.
606
Handbook of Non-Invasive Methods and the Skin, Second Edition
The angular deformation associated with a 2 mN·m torque was felt by the author to be a good parameter of “what can be sensed by palpation.” Other chosen parameters were the ratio peak to final torque (i.e., the end of relaxation) and the time for half relaxation, indicating the relaxation intensity within 1 min, and rate, respectively. Only the angular deformation at low torque was found to significantly decrease with aging, indicating an increase in skin stiffness. Three other runs were done after the first, each one at a 1-min interval, and the same parameters pooled. Again, only the angular deformation at low torque was significantly depressed with aging. Finally, a fifth run in the reverse direction was done 1 min after the fourth run. Only the ratio peak to final torque was found significantly lower with aging (p = 0.05), indicating a decrease in force relaxation within the extended skin during this time interval. As the skin components can be grossly compared to springs and dashpots connected both in parallel and in series (Figure 69.5), the stiffer skin and less relaxing forces inside skin may denote either a weakening of springs or strengthening of dashpots (viscosity), or both. Unfortunately, the panel of subjects comprised a bulk of middle-aged adults and very few children or aged people, which may explain the small number of significant correlations found in this study. Sanders,5 using Vlasblom’s apparatus and formulas, computed the torsional modulus of the outer forearm skin in 19 healthy subjects aged 6 to 61 years. The torque was 0.83 mN·m through an 8.7-mm-diameter disc without any guard ring, so that the twisted area was kept undefined. The pressure on the skin induced by the disc was not specified. The immediate elastic deformation (Ue) was found to increase almost linearly with aging, and the delayed viscoelastic deformation (Uv) seemed to increase only beyond 40 years of age. From these data a torsional elastic modulus of 20 to 100 kPa was computed and found to decrease with aging. As the torque was very weak, only elastic fibers were supposed to be implicated. Accordingly, the decreased skin stiffness with aging was tentatively ascribed to the well-known decay of elastic fibers over the life span. However, as said previously, the absence of guard ring let the subcutis bear the major (if not the entire) part of the deformation. As this layer is loose, this explains the very low Young’s modulus observed and the increase with aging. These observed parameters were probably those of the subcutis, not of the skin. Jaskowski and Maceluch6 investigated the dynamic torsional behavior of forearm, forehead, and abdomen skin in 380 normal subjects aged 5 to 80 years. The crown of skin between disc and guarding was made to vibrate by the rotating oscillation of the disc, up to a maximum amplitude (resonance). The observed parameters were the resistance, the frequency, and the attenuation. The skin stiffness (Nm/rad) rose steadily from 20 to 60 years, then
sharply beyond 60 years, and attenuation (Nm/rad) followed the same increase. The resonance frequency and resistance slightly rose over the life span. An investigation using a torsional device similar to Findlay’s, i.e., with a guard ring, was undertaken by Lévêque’s8 and Agache’s9 groups in a series of experiments in 1980. In the two first studies on 141 subjects in the age range 3 to 89 years, they used a device made of a central disc 25 mm in diameter and a skin annulus 5 mm wide, and exerted a 12.6-kPa pressure on the skin. The protocol consisted of abruptly putting (within 15 ms) a set torque for 2 min on the volar forearm and recording the angular rotation (24 mV per degree). Two stresses were applied at some distance on the same forearm: 9 and 28.6 mN·m. Under the lower stress the immediate deformation Ue decreased by about 30% during the first two decades, kept stable until 60 years, then rose by about 10%. But the product of immediate deformation by half skinfold at the same site (as assessed by a caliper) showed a steady and significant decrease with aging. Surprisingly, at almost any age this product was lower in females than in males, indicating a higher Young’s modulus under this tensile stress, and accordingly a stiffer skin. Under the higher stress the immediate deformation Ue strongly decreased until the age of 30, then rose moderately, while the Young’s modulus, as calculated by the above-quoted formula, slightly decreased to a minimum in the 20s, then rose progressively with aging. Again, with this high torque the Young’s modulus, i.e., the skin stiffness, was greater in females in all age classes beyond 20 years. The absolute values of this parameter lie in the same range for both stresses, between 0.6 and 2.9 MPa, but they were statistically higher with the lower torque. The direction of the difference would favor a technical origin rather than a physiological one, as the higher torque was too slight to induce any damage to the skin, and anyway gave no unpleasant sensation upon application. This work was undertaken again 9 years later14 using the same torsional equipment, but owing to the newly available technique of assessing skin thickness by ultrasound high-resolution imaging (25 MHz). The measurements were done on the volar forearm set in the same position as in the previous paper, on a panel of 123 subjects. Ultrasound studies showed a thinner skin before 10 years and after 70 years of age, with nonsignificant variation in between and a thickness significantly greater (+16%) in males than in females in the whole range of ages. The skin annulus submitted to twist was reduced to 3 mm wide in order to minimize the possible implication of subcutis. Also, the applied torque values were reduced to 2.3 and 10.4 mN·m, as compared to 9 and 28.6 mN·m in the previous paper, using a central disc of 18 mm diameter instead of 25 mm. No significant change over the life span or differences between sexes were found for either immediate extension or creep. Only did the product
Twistometry Measurement of Skin Elasticity
607
1.00
Elasticity: UR/UE
0.90 0.80 0.70 0.60 0.50 5
15
25
35
45 55 Age
65
75
90
FIGURE 69.7 Skin elasticity (Ur/Ue) over the life span with 2.6 mN/m torque (dotted line) and 10.4 mN/m torque (dashed line). Bars indicate SEM. (From Escoffier C, et al., J. Invest. Dermatol., 93, 353–357, 1989. With permission.)
Ue × thickness significantly (p < 0.01) decrease by about 85% beyond 70 years of age. By contrast, the immediate recovery steadily decreased with aging and also the ratio Ur/Ue (Figure 69.7). This trend was highly significant (p < 10–4). It confirms a current observation that skin folding keeps longer as age advances. The relaxation time of the creep decreased with aging; this was only significant with the higher torque. That means a less viscous dermis with advancing age, data in accordance with the well-known decline in mucopolysaccharide content.15 The results of all above-quoted studies, as presented in Table 69.1, show an agreement for a decreased skin immediate extensibility with aging, mostly beyond 60 years, and an increase in the viscous delayed extension. As the skin thickness decreases in the 60s, the result is an increase of the elastic modulus with aging beyond this age. Only Sander’s data are discrepant with this trend. As stressed previously, the main reason seems to lie in the absence of guard ring, and this experiment was probably an assessment of the subcutis distensibility.
Other discrepancies include the force relaxation time, i.e., the ability of structural elements inside the skin to move relative to each other. As this time is reduced, as shown in Agache’s,9 Lévêque’s,16 and Escoffier’s14 experiments, their mobility would increase with aging. The reverse was observed by Finlay.4 An explanation could lie in the difference in the type of experiments. While Finlay investigated the mobility of elements during a maintained stretch (for relaxation time), the two other papers dealt with the mobility during a maintained stress (deformation time). The former would implicate the strength of an interior spring attempting to recoil against viscosity, and the latter the resistance to a deformation through associated interior spring and viscosity. Accordingly, the strength of the interior spring would diminish with aging while spring resistance to extension plus viscosity also diminishes. So the discrepancy would be only apparent, not real.
69.3.2 ACTINIC AGING A more specific study of the influence of sun exposure of skin mechanical aging was conducted by Lévêque’s group in 1988.17 This witty investigation was made on a panel of 35 professional cyclists at the end of the period of intensive training in spring and early summer in often sunny countries (Spain, Italy, France) before entering the Tour de France competition. Almost all of them wore a short-sleeved shirt, dividing the outer aspect of their arm into a sun-covered and a sun-exposed area. This was evident by the suntan limit. As UVA rays do not penetrate the skin beyond the papillary dermis, the authors used a narrower ring (1 mm wide) with an 18-mm-diameter disc, in order to restrict the shear stress to the superficial layers of the skin. A 9 mN·m torque was applied for 10 s on both exposed and covered areas on the same arm, each at 1 cm distance from the suntan limit. In exposed areas the rotation angle was reduced by 33% (p < 0.0001) and the skin
TABLE 69.1 Literature Data on Intrinsic Aging as Assessed by Torsional Tests
Author Finlay Finlay Sanders Lévêque Agache Escoffier Escoffier
Disc Diam (mm) 15 15 8.7 25 25 18 18
Torque (mN/m)
Width of Skin Annulus (mm)
2 20 0.83 9 28.6 2.3 10.4
4 4 no 5 5 3 3
Deformation with Aging Ue Range (mm)
Elastic
0.4–1.7 0.1–0.4 0.6–1.5 0.4–0.6 1.1–1.2
↓ → ↑ ↓ ↓ ↓ ↓
Viscous
↑ → ↑ ↑
Elastic Modulus Range (MPa)
0.02–0.1 0.85–2.94* 0.64–2.04* 0.16–0.36* 0.34–0.62*
Note: Ue is immediate elongation. Asterisks indicate figures not produced by the authors but computed from Equation 69.4.
608
Handbook of Non-Invasive Methods and the Skin, Second Edition
thickness increased by 22.5% (p < 0.0001). As calculated by Equation 69.5, the mean Young’s modulus moved up from 541 to 698 kPa, thus demonstrating a skin stiffening following repeated sun exposure. Also, the relative immediate recovery was reduced by 22% (p < 0.0001), denoting a substantial loss of elasticity. This work was extended by a study in three racial groups (15 blacks, 12 whites, 12 Hispanics) made on both the ventral and dorsal forearm with a 15 mN·m torque, an 18-mm-radius disc, and a free skin crown of 3 mm width.18 The torque was set abruptly, maintained for 60 s, then abruptly suppressed. While no difference between races were observed in skin thickness (as assessed by ultrasound A-scan, 15 MHz), in all subjects the dorsal forearm skin was thicker. The skin elastic modulus E on the volar forearm was the same in all groups, whereas in the dorsal forearm it was statistically lower in blacks. Also, in blacks there was no difference in E values between the two sides, while the difference was significant in other groups. Viscoelasticity was lower in the dorsal forearm, but significantly lower only in whites and Hispanics. Finally, the clinical elasticity (recovery/extensibility) was lower on the dorsal forearm except in blacks, while irrespective of groups and sites it steadily decreased with aging. In summary, the volar forearm skin had the same mechanical behavior in all races. Differences appeared on the dorsal forearm, where blacks differ by the absence of alterations related to chronic sun exposure such as an increase in Young’s modulus (i.e., stiffness) and a decrease in skin extensibility. Also a black–Hispanic–white trend was noted on the dorsal side for such alterations. The results of these two studies show that, mechanically speaking, long-lasting and repeated sun damage resembles intrinsic aging in reducing skin elasticity and increasing Young’s modulus. But the former is associated with a thicker skin, while it is the contrary for the latter. Accordingly, chronic actinic damage cannot be simply assimilated to premature aging.
69.4 STRATUM CORNEUM AND SKIN BIOMECHANICS From a mechanical standpoint, the skin is a stratified composite material whose superficial layer is the stiffest, but represents only 1/100 of the total thickness. For that reason, its role in whole skin mechanical behavior is often overlooked. Torsional experiments offered Lévêque and De Rigal7 the possibility to investigate in vivo the mechanical behavior of stratum corneum (SC). Using a 9 mN·m torque, they first studied the direct effect of pouring water on a skin crown either 1, 3, or 5 mm wide in the volar forearm of 10 volunteers. The result was a highly significant increase in the Ue parameter, 80, 40, and 15% for crown widths of 1, 3, and 5 mm, respectively. Indirect SC
hydration by 15, 30, and 120 min occlusion with a plastic sheet in six subjects caused Ue to rise progressively and significantly. The rise was also significantly steeper with a 1-mm-wide crown than with a 5-mm-wide crown. These Ue variations are inversely related to Young’s modulus ones. On the other hand, only SC mechanical properties can be modified by pouring water on the skin surface. Therefore, this experiment demonstrated that (1) SC takes an important part in the mechanical behavior of whole skin, (2) using a narrow crown of twisted skin the contribution of SC to the assessed mechanical parameters is strongly increased, and (3) the SC moisturization can be accurately assessed in vivo by torsional experiments. In the same paper, the efficacy of cosmetic formulations on SC hydration was assessed using the same technique in 13 volunteers. The significantly increased Ue found 1 h following application of either oil-in-water (O/W) or water-in-oil (W/O) creams devoid of moisturizers, or petrolatum, partially or totally subsided 1 hour later, but at a lower rate with petrolatum and with W/O emulsion. Addition of moisturizers such as 10% lactic acid or PCNa or urea had the same effect, not greater, by the first hour, and unexpectedly no sustained effect was found by the second hour following application. Only 10% glycerol had a protracted hydration effect, although less marked than with lactic acid. In a long-term study carried out on the legs of three groups of 14 volunteers, selected for showing dry skin patches, two O/W preparations containing 10% glycerol or 10% urea were applied twice daily for 3 weeks. By the first, second, and third weeks both preparations brought a significant rise in Ue, while the vehicle used as a blank did not differ from the control site. Only the effect of glycerol was still measurable (p ≤ 0.05) 1 week after cessation of treatment.7 The efficacy of other cosmetic products on SC compliance was also investigated by the L’Oreal group with the same device and experimental conditions in 10 subjects.19 The compliance parameter Ue was raised by 34 ± 0.9% 2 min following total forearm immersion in hot water (30˚C) for 3 min. Also, Ur and Ur/Ue significantly rose, by 40 ± 6 and 5%, respectively. The increased skin elasticity could be related to the rise in skin temperature together with SC hydration. In the same paper, the variations of the same parameters over 24 d of treatment are presented; each measurement was done at time intervals 24 h after the last application. All three parameters showed a steady and highly significant increase. These experiments also demonstrated the daily individual variations probably related to climatic changes (relative humidity and temperature). The dryness of facial skin was assessed on the forehead and cheeks in 55 selected subjects using both an ordinal scale (0 to 3) and torsional experiments.20 On cheeks there was a striking inverse relationship between Ue and the clinical dryness score (r = –0.66, p < 0.001).
Twistometry Measurement of Skin Elasticity
This was also significant, although less marked on the forehead (r = –0.41, p < 0.002). This body of data on the clear-cut influence of the SC hydration effect on whole skin mechanical parameters stresses the role of SC in overall skin mechanical behavior, which often had been underestimated. The conclusive evidence, i.e., modification of skin extensibility and elasticity upon SC removal, was brought again by L’Oreal’s group in the following experiment.19 After 10, 15, and 20 SC strippings with an adhesive tape stuck for 15 s under 1 kg cm–2 on forearm skin in eight volunteers, Ue, Ur, and Ur/Ue steadily and significantly rose relative to baseline. This simple experiment confirmed unequivocally that SC takes a substantial part in the skin stiffness and recovery following extension and should be accounted for as a functional as well as a structural component in any interpretation of data concerning the mechanical properties or behavior of the skin. This is the confirmation that from a mechanical standpoint, skin is a composite material made up of three layers of different thickness, stiffness, and elasticity, working in parallel.
69.5 MEDICAL APPLICATIONS 69.5.1 EFFECT
OF
TOPICAL RETINOIC ACID
Because they are supposed to take place within SC and the papillary and subpapillary dermis, the changes induced by topically applied retinoic acid (RA) might modify the mechanical behavior of the skin. Consequently, torsional devices that are stuck on the skin and put force onto the superficial layers are particularly well suited for assessing this effect. Pichon et al.11 in a group of 17 subjects evaluated the extensibility and viscosity parameters as described in Chapter 62, using both 3- and 1-mm-wide skin crowns. With the wider annulus they found a significant decrease in both types of parameters, while only extensibility parameters (k0, k1) were significantly decreased with the narrow (1-cm-wide) annulus. The difference could be ascribed to the remodeling effect of RA on both the epidermis and subpapillary dermis. The same drug was tentatively used to reduce the skin side effects of long-term systemic corticosteroid treatment in 27 kidney graft recipients.21 Each patient’s volar forearms were treated either by RA, 0.5% (or 0.025%) in O/W emulsion, or vehicle alone, at random, for 180 d. Skin thickness, as determined by an ultrasound A-scan device working at 25 MHz, was found intensively decreased at the start and significantly increased by the 60th day of treatment onward, mostly in women. Torsional tests used a 3-mm free skin crown and an 18-mm disc radius. The skin elasticity (Ur/Ue) was increased in the treated vs. control areas by a factor of 3 to 20% by the 60th day and further on, but this was only significant in women. As no experiments were done with the narrower skin crown (1
609
mm), the results are difficult to interpret in terms of location of structural change.
69.5.2 SCLERODERMA Sclerodermas are diseases characterized by induration of the lower dermis and uppermost subcutis, in either localized and well-defined plaques or bands as in morphea or diffuse as in progressive scleroderma. Increases in thickness and lower extensibility have been non-invasively documented using ultrasound echography and a suction extensometer. Kalis et al.10 confirmed these findings by using L’Oreal’s torsional device with a 3 mN·m torque applied through an 18-mm-diameter disc and a 3-mm-wide free skin annulus. The skin thickness was measured by A-scan echography. In active plaques of morphea (n = 5) the results were highly significant for both increased skin thickness (60 ± 5%) and decreased extensibility (–67 ± 11%) relative to a symmetrical healthy control area. In regressing lesions (n = 12) the skin thickness did not differ from the control area, but the extensibility was still lower, although to a lesser extent. Using Equation 69.5, the authors concluded that there was a negligible and not significant difference in the elastic modulus E in case of active plaques (ΔE = –6%), whereas in the case of regressing lesions E was significantly higher (ΔE = +49%). From these data one can infer that the excess of collagen synthesized during the sclerotic phase retains normal mechanical properties, whereas at the regressing phase the progressive dermal atrophy is associated with a stiffer collagen. Measurements in progressive systemic scleroderma were made on the volar forearm. The 11 patients were compared to 10 age-matched controls. The skin thickness was about twice that of controls (+106 ± 18%) and extensibility (Ue) was reduced by 68 ± 8%. Both these differences were highly significant. As a surprising result, E was decreased by 38%. But the skin elasticity was also reduced. Humbert et al.22 used the same device and a 15-MHz ultrasound A-scan for skin thickness measurement in a patient with morphea treated with 1,25-dihydroxyvitamin D3. After 6 months of therapy Ue rose by 211%, while skin thickness remained unchanged (1.41 and 1.35 mm, respectively). Accordingly, the Young’s modulus E decreased by 204%, a finding in accordance with clinical scoring. This result was confirmed in a series of five patients with morphea whose duration of disease had varied from 2 to 10 years.23 They received oral 1,25-(OH)-2 vit D3 (mean dose, 1.75 mg/d). In line with clinical improvement, a significant decrease in the Young’s modulus (–62%, p < 0.01) appeared after 2 to 24 months of therapy. In an open uncontrolled study Humbert et al.24 included in the same protocol 11 patients suffering from systemic scleroderma. A significant decrease in the
610
Handbook of Non-Invasive Methods and the Skin, Second Edition
Young’s modulus (–36%, p < 0.01) was also noted at the end of the study. Furthermore, these investigations demonstrated the possibility of using this device for assessing the effect of a treatment over time in a single patient.
69.5.3 INHERITED CONNECTIVE TISSUE DISEASES Torsional measurements on the volar forearm were performed on five patients with type 2 Ehlers–Danlos syndrome, using the L’Oreal device, and the results compared to those of age- and sex-matched healthy subjects (five for each patient).25 The torque was acted by a 18-mmradius disc with a free skin annulus of 3 mm width. The skin thickness was considered half of the skinfold in the same area, as measured by a caliper. While a significantly lower skinfold thickness was found in patients than in controls (1.1 ± 3 mm vs. 1.5 ± 0.2 mm, respectively), the skin extension (Ue) was significantly increased (40 ± 11 mmV vs. 25 ± 5 mmV, respectively). By combining these data, no change in the elastic modulus was found in these patients. Also, the ratio Ur/Ue (elasticity) was found to be identical to that of controls. The conclusion was that only skin thinning could account for the clinically observed extensibility in these patients. By contrast, in three cases of Marfan’s syndrome the skinfold thickness and distensibility Ue of each patient were higher than those of his or her five age- and sexmatched controls. The product Ue × thickness was significantly higher by 46 and 53% on left and right forearms, respectively, and the Young’s modulus as calculated by Equation 69.5 was decreased by 60%. On the other hand, skin elasticity was found to be slightly and nonsignificantly increased. From these data the authors suggested that Marfan’s syndrome was associated with “a true decrease in tissue stiffness.” Some years later this was confirmed by an abnormality in the fibrillin network in that disease.26
69.6 CONCLUSION Torsional experiments as used mostly with l’Oreal’s device proved a reliable and easy-to-handle tool for assessing skin mechanical behavior in current practice. As in any use of instrument, however, great care should be taken by the clinician to comply with the above-cited requirements for the validity of the experiment. Over the last 12 years this allowed significant advances to be made concerning skin physiology and in vivo mechanical properties. By contrast, only a few investigations have been made in disease, at least for two reasons. The first one is the dimension of the guard ring, which initially was 54 mm and now is reduced to 40 mm in outer diameter, thus requiring a rather large, flat area of skin, which is uncommon in pathology. The second reason is that the device, named twistometer, was a L’Oreal prototype, not
commercially available, and consequently usable by only a few groups. This is no more the case today, as the Dermal Torque Meter has come to the market. Dermatologists, whether they are interested in skin physiology or pathology, now have an easy-to-use technique to investigate one of the main skin functions, and possibly have an insight into the functioning of skin’s major components. They feel indebted to the cosmetic industry to have created the device, designed the protocol, and made the experimental studies needed to validate clinically and structurally relevant skin mechanical parameters.
ACKNOWLEDGMENT The author acknowledges Jean-Luc Lévêque for his outstanding contribution to our knowledge in skin biomechanics physiology.
REFERENCES 1. Meirson D, Goldberg LH. The influence of age and patient positioning on skin tension lines. J Dermatol Surg Oncol 19: 39–43, 1993. 2. Vlasblom DC. Skin Elasticity. Ph.D. thesis, University of Utrecht, The Netherlands, 1967. 3. Finlay B. Dynamic mechanical testing of human skin “in vivo.” J Biomech 3: 557–568, 1970. 4. Finlay B. The torsional characteristics of human skin in vivo. J Biomed Eng 6: 567–573, 1971. 5. Sanders R. Torsional elasticity of human skin in vivo. Pflügers Arch 342: 255–260, 1973. 6. Jaskowski J, Maceluch J. Nowe mozliwosci badan wlasciwosci mechanicznych skory czlowoieka. Wiad Lek 35: 1149–1155, 1982. 7. Lévêque JL, De Rigal J. In vivo measurement of the stratum corneum elasticity. Bioeng Skin 1: 13–23, 1985. 8. Lévêque JL, De Rigal J, Agache P, Monneur C. Influence of ageing on the in vivo extensibility of human skin at a low stress. Arch Dermatol Res 269: 127–135, 1980. 9. Agache P, Monneur C, Lévêque JL, De Rigal J. Mechanical properties and Young’s modulus of human skin in vivo. Arch Dermatol Res 269: 221–232, 1980. 10. Kalis B, De Rigal J, Leonard F, Lévêque JL, Riche O, Le Corre Y, De Lacharriere O. In vivo study of scleroderma by non invasive techniques. Br J Dermatol 122: 785–791, 1990. 11. Pichon E, De Rigal J, Lévêque JL. In vivo Rheological Study of the Torsional Characteristics of the Skin. Paper presented at the 8th International Symposium of Bioenginering and the Skin, Stresa, Italy, June 13–16, 1990. 12. Wijn PFF. The Alinear Viscoelastic Properties of Human Skin In Vivo for Small Deformations. Ph.D. thesis, University of Nijmegen, Holland, 1980. 13. Barbenel JC, Payne PA. In vivo mechanical testing of dermal properties. Bioeng Skin 3: 8–38, 1981.
Twistometry Measurement of Skin Elasticity
14. Escoffier C, De Rigal J, Rochefort A, Vasselet R, Lévêque JL, Agache P. Age-related mechanical properties of human skin: an in vivo study. J Invest Dermatol 93: 353–357, 1989. 15. Fleischmajer R, Perlish JS. The vascular inflammatory and fibrotic components in scleroderma skin. Monogr Pathol 24: 40–54, 1983. 16. Lévêque JL, Corcuff P, De Rigal J, Agache P. In vivo studies of the evolution of physical properties of the human skin with age. Int J Dermatol 23: 322–329, 1984. 17. Lévêque JL, Porte G, De Rigal J, Corcuff P, Francois AM, Saint Leger D. Influence of chronic sun exposure on some biophysical parameters of the human skin: an in vivo study. J Cutan Aging Cosmet Dermatol 1: 123–127, 1988/89. 18. Berardesca E, De Rigal J, Lévêque JL, Maibach HI. In vivo biophysical characterization of skin physiological differences in races. Dermatologica 182: 89–93, 1991. 19. Aubert L, Anthoine P, De Rigal J, Lévêque JL. An in vivo assessment of the biomechanical properties of human skin modifications under the influence of cosmetic products. Int J Cosmet Sci 7: 51–59, 1985. 20. Lévêque JL, Grove G, De Rigal J, Corcuff P, Kligman Am, Saint Leger D. Biophysical characterization of dry facial skin. J Soc Cosmet Chem 82: 171–177, 1987.
611
21. De Lacharriere O, Escoffier C, Gracia AM, Teillac D, Saint Leger D, Berrebi C, Debure A, Lévêque JL, Kreis H, De Prost Y. Reversal effects of topical retinoic acid on the skin of kidney transplant recipients under systemic corticotherapy. J Invest Dermatol 95: 516–522, 1990. 22. Humbert P, Dupond JL, Rochefort A, Vasselet R, Lucas A, Laurent R, Agache P. Localized scleroderma: response to 1,25-dihydroxyvitamin D3. Clin Exp Dermatol 15: 396–398, 1990. 23. Humbert P. Unpublished data. 24. Humbert P, Dupond JL, Agache P. Treatment of scleroderma with oral 1,25-dihydroxy vitamin D3. An open study. Acta Dermatol Venereol, in press. 25. Bramont C, Vasselet R, Rochefort A, Agache P. Mechanical properties of the skin in Marfan’s syndrome and Ehlers-Danlos syndrome. Bioeng Skin 4: 217–227, 1988. 26. Mollister J, Godfrey M, Sakai LY, Pyeritz RE. Immunohistologic abnormalities of the microfibrillar fiber system in the Marfan syndrome. N Engl J Med 323: 152–159, 1990.
70 Levarometry Shabtay Dikstein Unit of Cell Pharmacology, School of Pharmacy, The Hebrew University of Jerusalem, Jerusalem, Israel
Joachim Fluhr Department of Dermatology, Friedrich Schiller University, Jena, Germany
CONTENTS 70.1 70.2 70.3 70.4 70.5 70.6
Introduction............................................................................................................................................................613 The Measuring System..........................................................................................................................................613 Methodological Principle ......................................................................................................................................613 The Sensitivity and Reproducibility of the Levarometry Measurements.............................................................614 Validation ...............................................................................................................................................................614 Correlation with Other Methods ...........................................................................................................................615 70.6.1 Pull, Guard Ring, and Weight of Probe ....................................................................................................615 70.6.2 Elevations...................................................................................................................................................615 70.7 Conclusions............................................................................................................................................................615 References .......................................................................................................................................................................615
70.1 INTRODUCTION
70.3 METHODOLOGICAL PRINCIPLE
The mechanical properties of the human skin are known to change with age: the skin becomes lax and wrinkled with the loss of elasticity and turgor.1–9 A broad clinical approach to the prevention and treatment of these changes requires a non-invasive method of assessing functional consequences of structural changes in different skin compartments with age. Such a method would enable the investigator to quantify the changes in the mechanical properties of the skin with age, as well as changes due to environmental causes such as solar damage. It would also help to quantify the effects of different treatments. Levarometry is a method of evaluating skin slackness.
To measure skin slackness, we developed a device called the levarometer, 1 0 which is a modified Schade instrument11,12 with increased sensitivity. The increased sensitivity was achieved by using electronic techniques, allowing the skin to be analyzed in vivo in the low range of the stress–strain curve, shown by Daly13 to be the most sensitive parameter in differentiating between young and old skin elasticity measurements. The levarometer (Figure 70.1) operates by applying a perpendicular pull to the skin without a guard ring. A circular piece of Perspex® with a diameter of 0.5 cm is attached to the skin by double-sided adhesive tape. This tape is then attached to a counterbalanced measuring rod so that the net pressure of the system is less than 1 g/cm2. Different weights can be applied to the rod, thereby providing any desired elevating force. The elevating forces used are usually low, in the range of 5 to 40 g/cm2. The measuring rod is connected to a linear variable differential transformer (LVDT), the output of which is recorded graphically. The precision of the measuring system is 2 μm. The main problem is in ensuring that there is no friction between the measuring rod and the LVDT. The most sensitive measurement is the immediate levarometric response without measuring the creep. As a
70.2 THE MEASURING SYSTEM One of the main characteristics of wrinkled skin is that it stretches more readily than smooth skin when force is applied, i.e., wrinkled skin is more slack than smooth skin. To measure this skin slackness, one has to apply a perpendicular pull to the skin without a guard ring, thus enabling the skin to respond freely to the applied force. Levarometry works on this principle.
613
614
Handbook of Non-Invasive Methods and the Skin, Second Edition
0.25
System net pressure is less than 1 g/cm2
20–30 Years >64 Years
Electronics Different weights LVDT
Recorder
Elevation in cm
0.20 0.15 0.10 0.05
Forehead
0.00 0
Measuring area - 0.2 cm2
10
20
30
40
50
Pull g/cm2
FIGURE 70.1 Schematic view of a levarometer.
FIGURE 70.2 The dependence of levarometric measurements on age. N = 15 in each age cohort.
standard, we use a 4-g weight, since the surface area of the probe is 0.2 cm2 and the pull is 20 g/cm2 (see Section 70.5). A theoretical treatise of levarometry has been published by Lanir et al.5
When measuring the same area at different times, the variance of the measurements is less than 5.5%.14 When performing sequential measurements over the same area of 4.5 × 4.5 cm2, the deviation from the mean is 6%. This deviation is very small considering that the skin is not a homogeneous tissue.14 Topical application of a humectant (5% glycerine) on the skin did not affect the measurements. This implies that the measurements are not affected by environmental conditions or treatments that have a direct or indirect influence on the stratum corneum.14
70.5 VALIDATION Does levarometry differentiate between young and old skin? In order to answer the most interesting question for investigators in the field of skin aging, two groups, young women (20 to 30 years) and older women (65 to 80 years), were compared, with 15 healthy volunteers in each group. The measurements were performed on the forehead, and four elevating pulls were applied with 5, 10, 20, and 40 g/cm2.14 The results (Figure 70.2) show that levarometry distinguishes between the two groups. The difference was significant for all the elevation forces (p < 0.02 for 5 g/cm2 and p < 0.01 for 10, 20, and 40 g/cm2). On the basis of the forehead measurements we decided that the standard pull for all other measurements should be 20 g/cm2. This pull was chosen because it is the highest pull within the linear range. With a lower pull there is greater risk of human error, since small pulls are more difficult to handle.
Male Female
0.25 Elevation in cm
70.4 THE SENSITIVITY AND REPRODUCIBILITY OF THE LEVAROMETRY MEASUREMENTS
0.30
0.20 0.15 0.10 0.05 Forearm 0.00 0
20
40
60
80
100
Years
FIGURE 70.3 Change of standard levarometric measurement according to age for women and men. N = 22 in each age and sex cohort.
Can levarometry differentiate between male and female skin? The inner forearm skin was assessed in two groups: young subjects (20 to 30 years) and older subjects (61 years) were measured at the standard pull of 20 g/cm2.14 There were 22 women and 22 men in each age group. The inner forearm was selected as the measuring site since it is usually less exposed to natural UV, and thus extrinsic aging is less dominant as an influencing factor. The results of the forearm measurements show that levarometry differentiates between the two age groups and also between men and women (Figure 70.3). Most of the men tested in both age groups had less skin elevation than women. This difference was more marked in the older age group; in the young age group it was 40%, while in the older group it was 75% (p < 0.01). What is the difference between young and old skin by levarometry at 20 g/cm2 pull? The measurements of the female forehead skin showed that the change between the
Levarometry
615
young and old cohort is 160% (p < 0.01) (Figure 70.2). The measurement on the volar forearm skin shows a change of 60% for the young vs. old male cohort (p < 0.01), whereas in the young vs. old female cohort the change is 100% (p < 0.01) (Figure 70.3).
ring, indentometry, and ballistometry,14,18–22 suggests that levarometry without a guard ring is highly discriminating between old and young skin, and also between old male and old female skin. Levarometry is therefore suited to the study of aging skin.
70.6 CORRELATION WITH OTHER METHODS
REFERENCES
There are two research groups using similar methods to levarometry.15,16 The results are compared below.
70.6.1 PULL, GUARD RING, AND WEIGHT OF PROBE The elevating forces used by Pierard15 are 16 to 128 g/cm2, with a measuring disc diameter of 14 mm, and 64 to 512 g/cm2 for a diameter of 7 mm. Gartstein16 uses 10 g of elevating force with a measuring disc diameter of 3 mm. Pierard7 uses guard rings of different diameters. We found that if the ratio between the diameter of the measuring disc and the diameter of the guard ring is at least 6, then the ring has minimal influence on the levarometry values. Otherwise, one is measuring skin elasticity rather than skin slackness. Dikstein’s10 group did not use such a guard ring. In Pierard’s method,15 the disc is glued to the skin by cyanoacrylate, whereas Gartstein et al.16 use a vacuum. Dikstein et al. used double-sided adhesive tape instead of cyanoacrylate to avoid the difficulty of applying the cyanoacrylate only on the precise area of the measurement. It is important to note that the starting net pressure of the measuring probe on the skin should be minimal, so that it does not indent the skin before elevating it.
70.6.2 ELEVATIONS Pierard15 measured loading deformation, which is a measurement of the change 20 s after loading the force, i.e., measuring immediate levarometry with the immediate creep. Gartstein et al.16 used 0.25 s for loading time, thereby measuring immediate levarometry, allowing the same time for recovery, with a total cycle time of 0.50 s, recorded by a PC. Gartstein et al.16 also measured skin elasticity as the percentage of recovery after deformation. The reproducibility of Gartstein’s system is 4%,16 Pierard’s is lower than 8%,17 and Dikstein’s is lower than 5.5%.14
70.7 CONCLUSIONS Gartstein et al.16 found that young, undamaged skin stretches less and recovers more completely than aged or solar-damaged skin. The results from Dikstein’s group showed that young skin is less slack than aged skin (see Section 70.5). A comparison of levarometry with other mechanical methods, such as torsion, elevation with guard
1. Agache, P.G. et al., Mechanical properties and Young’s modulus of human skin in vivo, Arch Dermatol Res, 269: 221, 1980. 2. Lévêque, J.L. et al., In vivo studies of the evolution of physical properties of the human skin with age, Int J Dermatol, 23: 322, 1984. 3. Pedersen, L., B. Hansen, and G.B. Jemec, Mechanical properties of the skin: a comparison between two suction cup methods, Skin Res Technol, 9: 111, 2003. 4. Adhoute, H. et al., Influence of age and sun exposure on the biophysical properties of the human skin: an in vivo study, Photodermatol Photoimmunol Photomed, 9: 99, 1992. 5. Lanir, Y. et al., The influence of ageing on the in-vivo mechanics of the skin, Skin Pharmacol, 6: 223, 1993. 6. Oikarinen, A., Aging of the skin connective tissue: how to measure the biochemical and mechanical properties of aging dermis, Photodermatol Photoimmunol Photomed, 10: 47, 1994. 7. Gniadecka, M. et al., Skin mechanical properties present adaptation to man’s upright position. In vivo studies of young and aged individuals, Acta Derm Venereol, 74: 188, 1994. 8. Batisse, D. et al., Influence of age on the wrinkling capacities of skin, Skin Res Technol, 8: 148, 2002. 9. Aframian, V.M. and S. Dikstein, Levarometry, in Handbook of Non-Invasive Methods and the Skin, Serup, J. and Jemec, G.B.E., Eds., CRC Press, Boca Raton, FL, 1995, p. 345. 10. Dikstein, S. and A. Hartzshtark, In-vivo measurement of some elastic properties of human skin, in Bioengineering and the Skin, Marks, R. and Payne, P.A., Eds., MTA Press, England, 1979, p. 43. 11. Schade, H., Die Elasticitatsfunction des Bindegewebes und Die Intravitale Messung Ihrer Strorungen, Z Exp Pathol Ther, 11: 369, 1912. 12. Kirk, E. and S.A. Kooming, Quantitative measurements of the elastic properties of the skin and subcutaneous tissue in young and old individuals, J Gerontol, 4, 27, 1969. 13. Daly, H.C. and G.F. Ödland, Age related changes in the mechanical properties of human skin, J Invest Dermatol, 73: 84, 1979. 14. Manny, V., In-Vivo Deformation by Small Forces as a Criterion for Assessing Skin Ageing, M. Phann. thesis, Hebrew University, Jerusalem, 1989. 15. Pierard, G.E., Investigating rheological properties of skin by applying a vertical pull, Bioeng Skin Newsl, 2: 31, 1980.
616
Handbook of Non-Invasive Methods and the Skin, Second Edition
16. Warren, R. et al., Age, sunlight, and facial skin: a histologic and quantitative study, J Am Acad Dermatol, 25(Pt. 1): 751, 1991. 17. Pierard, G.E., Structure et properties mechaniques descompartiment adventitiel et reticulaire du derme, thesis for Agrege de l’Enseignement Superieur, University of Liege, Belgium, 1984. 18. Brozek, J. and W. Warren-Kinzey, Age changes in skinfold compressibility, J Gerontol, 15: 45, 1960. 19. Graham, R. and P.J.L. Holt, The influence of ageing on the in vivo elasticity of human skin, Gerontologia, 15: 121, 1969.
20. Tosti, A., F.G. Compagno, M.L. Fazzini, and S. Villardita, A ballistometer for the study of the plastoelastic properties of skin. J Invest Dermatol, 69: 315, 1977. 21. Cook, T., H. Alexander, and M.C. Cohen, An experimental method for determining the two-dimensional mechanical properties of living human skin, Med Biol Eng Comput, 15: 381, 1977. 22. Lévêque, J.-L. and P. Corcuff, In vivo studies of the evaluation of physical properties of human skin with age, Int J Dermatol, 23: 322, 1984.
71 Indentometry Shabtay Dikstein Unit of Cell Pharmacology, School of Pharmacy, The Hebrew University of Jerusalem, Jerusalem, Israel
Joachim W. Fluhr Department of Dermatology, Friedrich Schiller University, Jena, Germany
CONTENTS 71.1 Introduction............................................................................................................................................................617 71.2 Various Measuring Systems ..................................................................................................................................617 71.3 Indentometry Measurements Using Different Methods........................................................................................618 71.4 What Does Indentometry Measure?......................................................................................................................619 71.5 General Conclusions and Recommendations for Standardized Use of the Indentometry Method .....................619 References .......................................................................................................................................................................620
71.1 INTRODUCTION When asked which of various body parts are soft or hard, people generally unwittingly demonstrate the principle of indentometry. They first point to their faces, palms, etc., with the tips of their fingers or a pencil, etc. When asked what they were attempting to determine, a typical remark would be “to see how deep the pencil would go.” Industrial measurements of hardness of materials are based on a similar principle. By the same token, in order to measure skin softness, one has to apply a perpendicular force to the skin, which will then indent in response to the applied force to a degree related to its softness. The various indentometry methods are based on this principle.1
71.2 VARIOUS MEASURING SYSTEMS The first indentometer was built by Schade2 in 1912. All other indentometry measuring systems are modifications of his system. Schade’s measuring system,2 the elastometer, consists of a registering lever and a rotating drum. One end of the lever is connected to a vertical metal rod terminating in a hemispherical brass knob with an area of 50 mm2 that is applied to the skin. The rod also carries a small platform upon which a weight (ranging from 5 to 75 g) may be placed. A pen point and ink is attached to the other end of the lever, which plots the curve on paper on the rotating drum. The measuring pressure is 10 to 150 g/cm2.
The measuring system of Kirk et al.,3,4 first published in 1949, uses a similar apparatus, with the only difference being a moving lever that is traced by a celluloid point on smoked paper instead of by a pen point and ink. The standard measuring weight was 50 g, and the measuring pressure was 100 g/cm2. The measuring system of Tregear and Dirnhuber,5 first published in 1965, consists of a weighted metal rod with a measuring area of 0.1 to 5 cm2; the rod is free to slide in a vertical glass tube and can be loaded with weights ranging from 100 to 2000 g. The rod is placed on a skinfold area, the undersurface of which rests on a thin metal disc of the same diameter as the rod. The vertical position of the rod is registered by a spring-loaded dial micrometer, accurate to 2 μm. The measuring pressure is 20 to 400 g/cm2. In 1969, Robertson et al.6 first described a measuring system consisting of a measuring rod and linkage system with a low moment of inertia. The probe is a hollow aluminum rod mounted vertically on low-friction bearings. The lower, measuring end, which is applied to the skin, has a diameter of 0.5 cm. An aluminum plate at the upper end bears the weights. This probe has a simple linkage to a rotating transducer, the outer case of which rotates in an arc around a stationary inner core in response to vertical movement of the rod. The inner core contains a photosensitive element. The output of the photoresistive cell is related to the vertical displacement of the probe and is measured on a recording voltmeter. The instrument has a response time of 50 msec and can record movements 617
618
Handbook of Non-Invasive Methods and the Skin, Second Edition
of up to 1 cm with a resolution of 0.01 cm. The internal friction is approximately 1 dyn. A 50-g weight was used in all their experiments. The measuring pressure is 250 g/cm2. The measuring system of Daly et al.7,8 first published in 1974, consists of an ultrasonic transducer connected to a lead rod 4.5 mm in diameter and 20 mm long. Loads are applied to the tissue by a servo-controlled loading system. A semiconductor strain gauge bridge measures the load directly at the transducer. The load used is 5 g, and the measuring pressure is 71 g/cm2. The instrument of Pierard,9 first described in 1984, uses an aluminum rod, the force is created by a manometer, and the measurement is carried out using a comparator. The measuring pressure used ranges from 1 to 4 N per 320 mm2. The precision of the measurement is 10 μm. The measuring pressure used was 28 to 111 g/cm2. The measuring system of Dikstein et al.,10 first published in 1981, has a measuring area of 0.2 cm2 (Perspex®), connected to a light metal measuring rod that is counterbalanced so that the net pressure of the system is less than 1 g/cm2. The measuring rod can be loaded with different weights. We decided on a standard load of 2 g, since the resultant pressure is still in the linear part of the stress–strain curve. The measuring rod is connected to a linear variable differential transformer (LVDT), and the output is recorded graphically by an electronic recorder. The measuring rod is adjusted so as to just touch the skin. Once the baseline is stabilized, the weight is suddenly applied. Measuring the same area at different times using this method, the variance of the measurements is less than 6%.11 Measuring the dispersal of measurements taken immediately one after the other over an area of 4.5 × 4.5 cm2, the deviation from the mean is 10%.11 The standard measuring pressure used was 10 g/cm2.
71.3 INDENTOMETRY MEASUREMENTS USING DIFFERENT METHODS Indentometry measurements were used to assess skin softness in a variety of physiological conditions: old and young skin, male and female skin, and in various edematous conditions. The measurements of Schade2 were carried out on the leg below the knee and on the forearm below the elbow in various edematous conditions. The effect of age on the initial indentation and on elastic recovery was also studied. It was found that in edema, both the indentation and the elastic recovery decrease. With age, it is mainly the elastic recovery that changes (decreases). The measurements of Kirk et al.3,4 were carried out on the medial surface of the tibia, using a weight of 50 g. The measurements were performed in women aged 20 to 101 years and in men aged 18 to 104 years. The observa-
tions showed a definite decrease in indentation with age, and an even more marked decline in the degree of immediate resiliency (rebound — called by the author elasticity) of the skin following removal of the weight. Women’s skin in general was found to possess higher elastic properties than that of men of similar ages; the elasticity values recorded for women were usually of the same order of magnitude as those exhibited by men 10 years younger.4 Indentometry measurements were also carried out on the skin covering the lower and upper tibia. The weight used was again 50 g. The measurements were done on skin of aged (60 to 86 years) and young (18 to 22 years) volunteers. The results showed a marked difference between the two groups: the depth of indentation in the young skin was much greater than in the older age group; the immediate rebound of the tissue after removal of the weight was also much greater in the young than in the aged group.3 A force of 100 to 2000 g acting over 0.1 to 5 cm2 for up to 20 min was used by Tregear and Dirnhuber5 to compress human skin in vivo. The subjects were young white male adults. The area tested was the skin of one leg over the tibia, and the dorsal aspect of one forearm. The maximum compression produced was 0.6 mm. The time course of the compression was not exponential; there was a long-continued tail of deformation, with the compression remaining after the force had been removed. The speed of compression was increased by increasing the force. Tregear did not carry out comparative physiological studies. The measurements of Robertson et al.6 were made on the dorsum of the hand, forearm, biceps, triceps, knee, calf, and foot. Measurements were carried out on three groups of women: nonpregnant, pregnant with no clinical edema in the third trimester, and pregnant with clinical edema. In addition, measurements were performed on one pregnant woman who developed widespread edema in late pregnancy. The indentation measurements produced a similar pattern to the compressibility measurements by a modified caliper: increasing depth of indentation was accompanied by a lowered index of compressibility. Kydd et al.7 measured the anterior-medial surface of the tibia. Twenty volunteers, of both sexes, aged 8 to 86 years, participated in the study. Age did not have a dramatic effect on the response of compressive loading. In the 8 to 10 years age group, there was a tendency toward greater compression. Age was, however, shown to have its most significant effect on the recovery phase. Children aged 8 to 10 years showed almost immediate recovery, to about 97%; adults 15 to 23 years old showed 90% recovery in the same period; those aged 72 to 86 showed recovery to about 67% in 10 min, and it took 4.5 h for complete recovery.
Indentometry
The areas used for testing by Pierard9 were the volar forearm and the tibia. The effect of age and sex on these measurements was not stated. Dikstein et al.12 measured the forehead in females aged 2 to 70 years using a pressure of 10 g/cm2. In 20-year-old patients the indentation was found to be 0.043 cm, and at the age of 70 it was 0.054 cm. The elastic recovery at the age of 20 was 80.5%, compared to 65.5% at the age of 70. Thus, skin of older females has less resistance to the applied force and less elastic recovery after removing the force than does that of younger subjects. Manny,11 using the apparatus of Dikstein, measured the indentation on the forehead and the indentation of the back of the hand. The indentation of the back of the hand was measured with both open palm and closed fist. The volunteers were women aged from 20 to 30 years and over 65 years. There was no difference in the indentation between the two age groups in this experiments. However, a difference between the open-palm and closed-fist measurements could be measured, with a higher indentation of open palm. On the forehead Lanir et al.13,14 compared different age groups. These studies showed that the response to indentation loading was analyzed in reference to its glycosaminoglycan (GAG)-containing ground substance and fibers’ network microstructure. In the second study,14 these authors measured the forehead skin response in young (20 to 26 years) and old (64 to 80 years) subjects. The analysis suggests that low-load indentation and small deformation levarometry are well suited for aging studies since the skin response under these tests can be directly related to its structure and constituent properties (known to be affected by aging). However, these results suggest that levarometry is more sensitive to aging than indentometry.14 In an additional study from Hartzshtark et al.,15 topically applied pharmacological agents, which are assumed to raise the c-AMP level, decrease the low-pressure indentation value of the forehead skin of certain human volunteers. Using the same method, Robert et al.16 assessed the influence of social and work-related factors on rheological properties of aging skin. The measurements were carried out on the forehead, using a pressure of 10 g/cm2. Three different populations were studied: nuns (females), office employees (both sexes), and workers (both sexes). For the purposes of this study, subjects were defined as young with an age less than 55 years and old if they were over 55. Theses authors found that the elastic rebound (after removing the weight) decreased steadily with age. This effect was more pronounced in females. Working women lost their elasticity more rapidly than nuns, and the male workers lost their elasticity more rapidly than did the office employees. An indentometric approach was used for in vitro testing of biological material by Piruzian et al.17 They tested different passive and active mechanic load mechanisms.
619
71.4 WHAT DOES INDENTOMETRY MEASURE? The basic question is, Which skin layer is the most important for indentation measurements? In the stratum corneum, 4% lactic acid (pH 4.2) applied to the skin of four volunteers with dry skin of high pH was found to have no effect on indentation. Also, 5% glycerin, which is a strong humectant, was tried as a moisturizer on six volunteers with dry skin, without an effect on indentation.18 In an another experiment the stratum corneum was removed by stripping techniques, monitoring the surface pH to ensure that the stratum corneum was indeed removed. It was found that stripping did not affect indentation measurements.18 From those experiments it is clear that the stratum corneum does not affect indentation. In the dermis the forehead skin of three male volunteers was injected intradermally with 0.2 cm3 saline, and susequently (at least 10 d later), the same volume of saline containing 300 IU of hyaluronidase, 4 U of elastase, or 4 U of collagenase was injected. The results showed that saline decreased indentation, whereas hyaluronidase, and to a lesser degree elastase, increased indentation. Collagenase had no significant effect. The decrease in indentation caused by saline suggests, therefore, that the water moisture content in the dermis decreases indentation. Elastase, and even more so hyaluronidase, increases indentation in spite of the presence of saline. The interpretation of those experiments is that in the dermis, the state of the ground substance elastin network might be responsible for the observed changes in indentometry.18 More evidence of the effect of the ground substance on indentation was found in rat skin.5 A cross section cut through compressed rat skin showed that most of the compression occurred in the dermis. Since liquids are virtually incompressible, for the dermis to be compressed, the water must have shifted out sideways from the compressed area. The dermal fluid, the ground substance, is known to contain enough mucopolysaccharide to make it highly viscous.19 Diffusion through the dermis is greatly increased when the polysaccharides are depolymerized by hyaluronidase.20
71.5 GENERAL CONCLUSIONS AND RECOMMENDATIONS FOR STANDARDIZED USE OF THE INDENTOMETRY METHOD Indentometry is a technique for measuring skin softness. In effect, it actually measures the water status of the dermis, which is largely determined by the amount of mucopolysaccharides and their ability to bind water in the measured area. The immediate indentation measurement using indentometry was found in most studies not to discriminate sufficiently between young and old skin.4,7,12,16 The
620
Handbook of Non-Invasive Methods and the Skin, Second Edition
difference between the two age groups can be better differentiated by measuring the elastic rebound, which is the immediate rebound of the skin after unloading the force. It was found that in young skin, after unloading the force, the immediate rebound comprised a greater percentage of the initial indentation.2–4,7,11,12,16 Although some studies showed age dependence and differences between the skins of males and females with the indentometry method, it seems to us that indentometry is of most use to evaluate edematous conditions, and altered water handling of the dermis.2,6 The present data lead us to the following recommendations for a standardized use of the indentometry method: 1. In order to compare different physiologic conditions — young vs. old skin, male vs. female, and various edematous conditions — one must work within the linear area of the stress–strain curve of the skin area being measured. The standard indentation force used by Dikstein et al. is 10 g/cm2; this force is in the low linear area of the stress–strain curve of the forehead, the area on which the measurements are made.10 2. The pressure of the measuring apparatus itself should not exceed 20% of the total measuring pressure. The weight of Dikstein et al.’s apparatus creates an initial pressure of 1 g/cm2, which is 10% of the final measuring pressure. The exact initial pressure of the apparatus on the skin is not specified in most of publications. 3. During the measurement no lateral movement should occur between the skin and the apparatus. The reproducibility of the measurements should be greater than 90%.11
REFERENCES 1. Aframian, V.M. and S. Dikstein, Indentometry, in Handbook of Non-Invasive Methods and the Skin, Serup, J. and Jemec, G.B.E., Eds., CRC Press, Boca Raton, FL, 1995, p. 349. 2. Schade, H., Untersuchungen zur Organfunktion des Bindegewebes, Z. Exp. Pathol. Ther., 11: 369, 1912. 3. Kirk, E. and S.A. Kvorning, Quantitative measurements of the elastic properties of the skin and subcutaneous tissue in young and old individuals, J. Gerontol., 4: 273, 1949.
4. Kirk, J.E. and M. Chieffi, Variation with age in elasticity of skin and subcutaneous tissue in human individuals, J. Gerontol., 17: 373, 1962. 5. Tregear, R.T. and P. Dirnhuber, Viscous flow in compressed human and rat skin, J. Invest. Dermatol., 45: 119, 1965. 6. Robertson, E.G. et al., Two devices for quantifying the rate of deformation of skin and subcutaneous tissue, J. Lab. Clin. Med., 73: 594, 1969. 7. Kydd, W.L., C.H. Daly, and D. Nansen, Variation in the response to mechanical stress of human soft tissues as related to age, J. Prosthet. Dent., 32: 493, 1974. 8. Daly, C.H. and G.F. Odland, Age related changes in the mechanical properties of human skin, J. Invest. Dermatol., 73: 84, 1979. 9. Pierard, G.E., Evaluation de proprietes mecaniques dela peau pa les methodes dindentation et de compression, Dermatologica, 168: 61, 1984. 10. Dikstein, S. and A. Hartzshtark, In-vivo measurement of some elastic properties of human skin, in Bioengineering and the Skin, Marks, R. and Payne, P.A., Eds., MTA Press, England, 1981, p. 43. 11. Manny, V., In-Vivo Deformation by Small Forces as a Criterion for Assessing Skin Ageing, M. Pharm. thesis, Hebrew University, Jerusalem, 1989. 12. Dikstein, S., A. Hartzshtark, and P. Bercovici, The dependence of low-pressure indentation, slackness and surface pH on age in forehead skin of woman, J. Soc. Cosmet. Chem., 35: 221, 1984. 13. Lanir, Y. et al., In-vivo indentation of human skin, J. Biomech. Eng., 112: 63, 1990. 14. Lanir, Y. et al., Influence of ageing on the in vivo mechanics of the skin, Skin Pharmacol., 6: 223, 1993. 15. Hartzshtark, A. and S. Dikstein, The use of indentometry to study the effect of agents known to increase skin cAMP content, Experientia, 41: 378, 1985. 16. Robert, C. et al., Study of skin ageing as a function of social and professional conditions: modification of the rheological parameters measured with a noninvasive method — indentometry, Gerontology, 34: 284, 1988. 17. Piruzian, L.A. et al., [Study of biological materials based on indentometry], Izv. Akad. Nauk. SSSR Biol., 5: 769, 1990. 18. Dikstein, S. and A. Hartzshtark, What does low-pressure indentometry measure? Arztliche Kosmetologie, 13: 327, 1983. 19. Ogston, A.G. and J.E. Stanier, The physiological function of hyaluronic acid in synovial fluid, J. Physiol.,119: 244, 1953. 20. McLean, D. and C.W. Hale, Studies on diffusing factors, Biochem. J., 35: 159, 1941.
72 The Gas-Bearing Electrodynamometer C.W. Hargens Philadelphia, Pennsylvania
CONTENTS 72.1 Introduction............................................................................................................................................................621 72.2 Instrumental Application .......................................................................................................................................622 72.3 Instrumentation ......................................................................................................................................................623 72.4 Data Reduction ......................................................................................................................................................624 References .......................................................................................................................................................................625
72.1 INTRODUCTION The present chapter deals with the fundamental mechanical properties of the skin and their quantitative measurement. This is basic to an evaluation of existing dermal conditions and the efficacy of any treatment. The method is direct and not implied by inference from some other test or property, such as electrical conductivity or sonic propagation. It is the method a mechanical or materials engineer would use, the determination of stress (applied force) vs. strain (resulting deformation), to describe the characteristics influencing the performance of any structural substance. The skin is one of the most important structural substances holding the body together. We can further demonstrate that the skin’s principal strength resides in a few outer cell layers of the stratum corneum. This can be easily shown with the gas-bearing electrodynamometer (GBE) instrumentation if one examines an area where these thin cell layers are stripped away by just a few repeated applications of Scotch® tape (adhesive). A glistening layer is quickly reached, which the instrument will show has no strength of containment. In fact, such a test reveals that the mechanical modulus of the stripped area diminishes from a high value to almost nothing. This is the same phenomenon occurring in burns of the skin, which have a similar destructive effect. The term dynamometer may be somewhat foreign to the biomedical field, but technically it implies a device for measuring power, force, electrical current, or voltage. In this case, we are interested in a convenient and accurate way to measure force. A dynamometer involves the interaction of electrical and mechanical quantities, in essence a kind of transducer. Electric current in a conductor can be measured, and through its interaction with a magnetic
field, mechanical force thereby will be determined; or the reverse, measure force and derive the value of current in an electrical conductor. Usually force is determined through an arrangement of coils, magnetic fields, and a current measurement. The adjustable nature of these quantities and devices allows one great latitude in designing measuring systems to suit various purposes. Also, the precision and accuracy can be of a high order, because the calibrating forces, currents, and voltages involved can be fundamentally verified through standards. In the measurement of the skin’s mechanical properties stress–strain (force–displacement) modulus the principles just mentioned have been utilized. To further understand what is involved, it must be recognized that the skin, like other body tissues, is a viscoelastic substance; i.e., it is partly spring-like and partly viscous. Thus, the mechanical moduli one seeks to determine are significantly rate related due to the large viscous component of the modulus. In other words, the observed deformation will depend upon the speed with which a force is applied. This is why one cannot be satisfied with static measurements that were done in earlier times, for they will not explain dynamic behavior that influences our subjective perception of skin quality. The fundamental reason for there being a complex, rate-related modulus to describe skin behavior is the same as in the case of all large molecule polymers. In all of these materials there is a finite time required for a stressimposed molecular rearrangement to take place. To this phenomenon we assign the simplifying term viscosity, and it is present in soft solid substances as well as in fluids that have been studied extensively in this regard.1 Skin is composed of a well-known and complicated architecture consisting of such viscoelastic materials. As just mentioned, the main mechanical influences of this 621
622
Handbook of Non-Invasive Methods and the Skin, Second Edition
kind, as far as skin measurements are concerned, reside in the cells of the stratum corneum and the natural cements that hold the aggregate together. The GBE is sensitive enough to show that the mechanical integrity of these thin membranes is strongly influenced by small external effects such as moisture and other topically applied reagents. Returning to the matter of feasible instrumentation for skin tests, one needs equipment capable of dynamic measurement. The interesting phenomena relating to skin involve time constants on the order of 1 s. If one has sufficiently lightweight equipment, with agile components that can be quickly accelerated, these features of elasticity and viscosity can be recorded. One can apply small forces, actually a periodic forcing function such as a sine, at the 1-s rate and observe by suitable recording means the resulting deformations as a phase-displaced signal. Details of the application of these principles are explained in Section 72.2. Another point of initial consideration concerns the best direction in which to apply the test stresses in the skin to clearly enhance the desired responses. It turns out that the most sensitive indications of modulus changes in a membrane such as the stratum corneum will be obtained with the shear mode, i.e., the force vector applied directly in the plane of the surface. Techniques that indent the surface or attempt to raise a bell-shaped deformation normal to the surface with suction have been used but are far less sensitive. This is because the measurement relies upon only the smaller vector components of the principal stresses. Finally, one finds that a stress–strain characteristic taken in the shear mode gives a very adequate measurement. Plotted in real time (approximately a 1-s cycle), the diagram will be a smooth ellipse in four quadrants. It is best to keep within the linear parameters of the skin if a simple numerical modulus is to be obtained. Beyond these limits, the tissue’s response becomes very nonlinear, and analysis becomes much more complicated, involving a Fourier analysis of the individual stress and strain waveforms. The consequent interpretation of results is not worth the trouble, since the linear response involving a simple phase shift between sinusoidal stress and strain waveforms is quite adequate for the comparative evaluations one is seeking. The ellipse is of course just a display suitable for calculating the phase angle between X and Y waveforms and the modulus or dynamic spring rate of the test surface. This will be explained further in Section 72.4.
72.2 INSTRUMENTAL APPLICATION Having introduced the basic objective of the noninvasive protocol, we now present the method that has been developed over several decades and used successfully for as long.2–9 The protocol is to adhere a probe to a chosen test
FIGURE 72.1 Dynamometer attached to a facial site.
site on the skin by means of an accepted medical adhesive similar to products used in surgical procedures. The surface area contacted by the probe need be only a few square millimeters. The probe attaches in turn to the GBE, as shown in Figure 72.1. Here a facial site is being examined. The probe moves with the reciprocating action of the GBE. The amplitude of motion is usually about 1 mm, so the active area of the skin under test might be said to extend up to 10 mm beyond the attachment point, depending upon the type of skin involved and its pretension state. A more detailed view of what takes place around the probe in the plane of the skin during the stressing of the surface would seem to be the following. Stratum corneum is in itself so stiff that bodily actions are only possible because there is a certain excess of skin. That is, there are microscopic as well as macroscopic folds, creases, and ridges in its texture. As the skin moves, these function in accordion fashion to relieve the stress that would otherwise occur. Therefore, what the probe measures through the resulting strain determination, the stretch, is actually the bending of the stratum corneum within these folds, i.e., the unfolding process brought about by the stretching forces. This reality in no way diminishes the utility of the measurement, because the apparent stretchability and dynamic behavior of the skin are what one seeks to know. Regardless of the precise mechanism of dynamic response, the influence of the various skin treatments upon the measurable and subjectively sensed stiffness or dryness will become apparent from the data obtained. As the probe of the GBE applies specific forces in the plane of motion, its resulting displacement, also monitored, is controlled by the elastic and viscous moduli of the outer skin layer. It is a simple matter to display these two quantities, force and displacement, as a typical stress–strain diagram for the material.
The Gas-Bearing Electrodynamometer
A plotting device with sufficiently rapid movement, such as a storage oscilloscope, analog or digital, can display the complete elliptical diagram every second as the probe drives the skin back and forth in repeated cycles. If rapid changes in skin condition occur, as with the introduction of moisture or drying, they will appear as corresponding alterations of the diagram. For example, dryness and stiffening of the surface will appear as an elevated slope of the major axis of the ellipse; softening will cause a decrease in slope. When only a few tests are to be performed, the dynamic stress–strain plots can be measured graphically and the moduli calculated, as will be explained, from their geometry. On the other hand, in laboratories where very large testing programs continue day after day, it is best to computerize the process, taking data directly through analog-to-digital conversion of the stress and strain signals.
72.3 INSTRUMENTATION So far the GBE has been described only as to its capabilities, whereas little has been said about the mechanism itself. The probe just referred to in Section 72.2 is attached to the skin in a novel way so that the test site can be precisely preserved while the GBE is disconnected for use elsewhere. Thus, the subject can be released for other activities during the course of the test period. To facilitate this, a small plastic button serves to contact the skin, adhered by the adhesive film previously mentioned. However, the actual probe wire fits tightly into a tiny hole in the top of the button and remains thusly attached during the test. Disconnection is achieved by simply popping the probe wire out of the hole in the button. A supply of buttons and film adhesive is furnished with the instrument, along with a detailed application procedure. The GBE is visible at the top of Figure 72.1 and is mounted on a small optical bench immediately adjacent to the subject, whose head rests on the same base so as to minimize relative motion. The optical bench provides, through rack-and-pinion manual adjustments, precise but rapid positioning along the three coordinate axes. The attachment button secured to the probe wire is moved about over the subject by means of these controls until the desired test location is attained. The GBE is lowered until the tip of the probe (button) touches the skin and adheres to it. It is, of course, desirable that the subject not move for the several seconds of the test in order to avoid disturbing the recording process. This is usually not a problem, even without any particular stabilizing means for the head or other body part under study. A relaxed subject can hold quite still, even to the extent required by the rather large displacement magnification of the electronic and display systems involved. That is, 1 mm of skin displacement at the test site can correspond to 25 mm of spot movement
623
on the oscilloscope screen or plotter (a magnification of at least 25). The displacement of the probe just discussed is induced by the reciprocating armature of the GBE to which it is attached. The probe is held by a setscrew in a small plastic chuck fitted into a protruding sleeve of the armature. This chuck with probe attached can be removed as an assembly, owing to its tapered fit. This is an added convenience, as will be discovered when moving the GBE. The armature also carries a coil of fine copper wire, which moves in a strong, radial magnetic field. The field is created by a permanent magnet whose flux can be depended upon to remain constant. As the GBE’s name implies, the armature floats on a coaxial air bearing. Therefore, there is no metal-to-metal contact in the bearing, and frictional forces are reduced to those caused by the viscosity of air. This means that they are so small as to be orders of magnitude less than the skin moduli, and hence negligible. To a slightly lesser extent this is true of the hair-like leads carrying current to the coil. The force produced by the coil and magnetic field combination is in accordance with the standard electric motor equation: Force = B · l · i
(72.1)
where force is in dynes, B (magnetic flux) is in gausses, l (length of conductor) is in centimeters, and i (current) is in abamperes (1 abampere = 10 amperes). Thus, if B and l are constant, as in the dynamometer, the force is directly proportional to the current, and this fact allows one to use current to determine the force exerted by the GBE probe. Displacement (strain) is determined through another electrical device built into the GBE. This tracks the armature’s movements without exerting any force upon it. It is referred to as a linear variable differential transformer (LVDT). It consists of a primary and two secondary coils, all magnetically coupled together by a movable core. As the core moves axially with the armature through the three coils’ windings, it couples flux differentially from the primary into the two secondaries. If an electronic system is associated with the LVDT to supply current to the primary coil and rectify the combined secondaries’ output voltages (demodulate), one obtains a direct current (DC) voltage whose polarity and magnitude give an accurate record of the armature’s instantaneous position. This electronic system associated with the LVDT is referred to as a signal conditioner. The stress–strain diagram is thus created by applying a voltage corresponding to the force coil’s current to the oscilloscope’s vertical deflection system and at the same time connecting the LVDT (signal conditioner output) to the horizontal deflection system. For a viscoelastic material, whose deformation lags in time the impressed
624
Handbook of Non-Invasive Methods and the Skin, Second Edition
Stress (grams)
B/2 a
B Strain (mm)
A
FIGURE 72.2 Dynamic stress-strain diagram plotted by the GBE apparatus showing the principal measurable quantities.
sinusoidal forcing function, an ellipse is formed, as in Figure 72.2. The GBE has several important advantages for measuring the mechanical properties of soft, viscoelastic substances such as living tissue. One advantage is its ability to exert small dynamic forces and record sizeable displacements that ensue in these soft materials. Other transducers, by contrast, are very stiff, such as strain gauges and load cells, or they impose frictional elements in parallel with the test specimen. They also are usually rather bulky and inconvenient for clinical testing, or they cannot respond at high frequencies. The construction of a GBE, on the other hand, is similar to the transducer in an audio loudspeaker, but without the stiffness. For completeness it should be mentioned that a basic air-bearing dynamometer in an evolutionary form has been used at audio frequencies. However, at audio rates it operates in a slightly different fashion, because the mass of the armature, although small, and in fact negligible at the 1-Hz rate discussed earlier, would now become significant. However, it can be shown that by combining the electrical motor equation (Equation 72.1) with the corresponding generator equation involving velocity (E = Blv,) one does not need the LVDT or any other measure of displacement. Instead, it is possible to employ an electric bridge circuit to obtain precise electrical impedance analogs of the specimen’s mechanical moduli. From these electrical quantities one can calculate the actual mechanical characteristics of the material.10,11 These would combine to yield the so-called mechanical impedance. Returning to the low-frequency operation of the GBE described here, one does measure both force and displacement to produce the elliptical stress–strain diagram directly. From the diagram’s dimensions, which are calibrated in appropriate units, such as grams and millimeters, it is a simple matter to compute the moduli. The most useful modulus is the dynamic spring rate (DSR), which is the slope of the ellipse’s major axis, B/A,
as seen in Figure 72.2. As mentioned earlier, it correlates well with subjective descriptors such as dryness, softness, etc., used by trained dermatological graders.12–14 Additional information about the combined elastic and viscous parameters can be obtained by noting the openness of the elliptic loop. For example, a strictly elastic element like a steel spring would show no openness. The correct interpretation of this would be to say that there is no energy loss if the loop degenerates to a straight line; i.e., the integrated area under the curve in Figure 72.2 (the energy or work done) is the same during stretch as during return. Energy is conserved in the elastic case but not in the case involving viscous elements. There should be no mystery about the elliptical display. It is simply the Lissajous figure used to measure the phase angle between two sinusoidal waves, in this case, stress and strain. If they were shown as sine functions on two parallel time axes, one would observe the time lag or phase angle lag of the strain behind the applied stress. In our case, it is called the loss angle. In Figure 72.2 the numerical value of the loss angle is computed from the ratio of the loop opening to the total displacement. The sine of the loss angle equals this ratio: Sin θ = a/A
(72.2)
The fundamental reasons for this time lag in a viscoelastic material have been previously discussed.1 What has the loss angle to do with tissue properties in dermatological terms? It has been observed that young, healthy tissue has the least loss. This was found to be true in the case of ocular tissue when ophthalmic experiments were conducted, and it is true of skin. While speaking of the loss factor, we should interject that there is of course more to tactile sensation, self-comfort, and appearance than the dynamic shear moduli of the stratum corneum as measured by the GBE. The underlying dermal tissues are important as well, and their contribution can be conveniently measured using ballistometry. To illustrate dermal loss factor further, in a very homely analogy we are speaking about the “upholstery” of the body. That is to say, subjectively tactile sensation responds to all of the above, plus frictional effects, and there is a noticeable difference between a strictly springy surface and one with a gradual yield, even though their static stress–strain deformation coefficients may be identical. Again, note the differences between a down-filled cushion, one of foam rubber, and one packed with cotton batting. The dynamic response counts strongly.
72.4 DATA REDUCTION Graphical data reduction methods were employed when the GBE was first introduced. A Polaroid camera photographed each loop, which was later scaled and the
The Gas-Bearing Electrodynamometer
calibrated measurements tabulated. The force calibration of the GBE is straightforward and positive. First, the instrument is simply tipped up and fixed in a vertical position. The wire probe is removed from the chuck, and the dynamometer becomes a platform balance on which gram weights can be placed, a few at a time. The armature, now vertical, is balanced against gravity by means of the DC offset control of the function generator, which drives the GBE. The various balance currents introduced into the force coil of the instrument will cause a series of corresponding vertical displacements of the trace along the force axis of the oscilloscope or other plotter. The vertical amplification setting of the latter can be chosen to give a convenient scale for the force range being used in any particular experiments. Once set, of course, it should not be changed until the system is recalibrated. The displacement calibration is equally simple and positive in its overall inclusion of the system component variables. It is done with the GBE in a horizontal position and with a micrometer or other accurate scale indicating armature movement. The corresponding scale can thus be generated on the horizontal (displacement) axis of the display, again choosing an appropriate adjustment of the amplifier. When a large number of experiments must be carried out on a daily basis, a computer can be connected to the system. This will greatly facilitate recording of data, automatic calculation of DSR, and other quantities, as well as trend or statistical data summaries. The connection is made directly from the oscilloscope or recorder terminals using the signal voltages produced by the GBE system components. Each experimenter will establish his or her own protocols for equilibrating subjects’ skin when they arrive for tests. For extremely precise work, an environmental chamber has been used, although most modern laboratories have sufficiently standardized atmospheres in their general working spaces.
REFERENCES 1. Ferry, J.D., Viscoelastic Properties of Polymers, John Wiley & Sons, New York, 1961. 2. Hargens, C.W., Glaucoma and vibration tonometry, J. Franklin Inst., 207, 143, 1960.
625
3. Hargens, C.W., Viscoelastometry of biomaterials, in Proceedings of the 26th Annual Conference on Engineering in Medicine and Biology (IEEE), Minneapolis, 1973, p. 203. 4. Hargens, C.W., Instrumentation to measure the viscoelastic properties of intact human skin, in Proceedings of the 28th Annual Conference on Engineering in Medicine and Biology (IEEE), New Orleans, 1975, 17, 179. 5. Christensen, M.S., Hargens, C.W., Nacht, S., and Gans, E.H., Viscoelastic properties of intact human skin: instrumentation, hydration effects, and the contribution of the stratum corneum, J. Invest. Dermatol., 69, 282, 1977. 6. Hargens, C.W., Measurement of dynamic moduli and loss factor in viscoelastic materials using the gas bearing electrodynamometer (GBE), J. Acoust. Soc. Am., 67 (Suppl. 1), S25, 1980. 7. Hargens, C.W., The gas bearing electrodynamometer (GBE) applied to measuring mechanical changes in skin and other tissues, in Bioengineering and the Skin, Marks, R. and Payne, P.A., Eds., MTP Press, Lancaster, U.K., 1981, chap. 14. 8. Hargens, C.W., Instrumented testing of human skin in vivo, in Proceedings of the 35th Annual Conference on Engineering in Medicine and Biology (IEEE), Philadelphia, 1982, p. 12. 9. Missel, P.J., Bowman, M.J., Benzinger, M.J., and Albright, G.B., An in vitro method for skin preservation to study the influences of relative humidity and treatment on stratum corneum elasticity, Bioeng. Skin, 2, 203, 1986. 10. Hargens, C.W. and Keiper, D.A., Tonometry: challenge for electronics, in Digest of the 1961 International Conference on Medicine and Electronics (IEEE), 1961, p. 77. 11. Keiper, D.A., Dynamic mechanical properties tester for low audio and subaudio frequencies, Rev. Sci. Instrum., 33, 1181, 1962. 12. Christensen, M.S., Nacht, S., and Packman, E.W., Facial oiliness and dryness: correlation between instrumental measurements and self-assessment, J. Soc. Cosmet. Chem., 34, 241, 1983. 13. Maes, D., Short, J., Turek, B.A., and Reinstein, J.A., In vivo measuring of softness using the gas bearing electrodynamometer, Int. J. Cosmet. Sci., 5, 189, 1983. 14. Cooper, E.R., Missel, P.J., Hannon, D.P., and Albright, G.B., Mechanical properties of dry, normal and glycerol-treated skin as measured by the gas bearing electrodynamometer, J. Soc. Cosmet. Chem., 36, 335, 1985.
73 Ballistometry C.W. Hargens Philadelphia, Pennsylvania
CONTENTS 73.1 Introduction............................................................................................................................................................627 73.2 Mechanics of Ballistometry...................................................................................................................................628 73.3 Practical Ballistometry for Skin Studies ...............................................................................................................630 73.4 Data Obtained with Ballistic Measurements.........................................................................................................632 References .......................................................................................................................................................................632
73.1 INTRODUCTION The ballistometer is a neat and easy-to-use device for measuring certain properties of human skin in vivo. This statement implies the ballistometer’s usefulness, but at the same time, the expression “certain properties” is a caution. One should be very specific in all claims regarding the determination of skin properties because there are so many. One might do well to first decide which ones tell something about the subjectively sensed skin condition to be evaluated, then see which “property” is the best indicator and measure it with the correct instrument. An example of two measurements which determine entirely different skin properties concerns the role of the ballistometer in contrast to the GBE (gas-bearing electrodynamometer). The first measures certain elastic parameters below the surface, while the GBE is best for quantifying stiffness in the surface plane of the skin, the stratum corneum. Here we will speak about the ballistometer for skin studies, and it should be appreciated that this is a narrow application of the concept. Our discussion will differentiate between a strictly practical device that has served well as an experimental tool and consideration in some detail of its extended capabilities. Ballistometry, the classical use of impacting masses to measure their material properties through their interaction, is not new. Sir Isaac Newton (1642 to 1727) is credited with certain relevant philosophical propositions and experiments involving impacting bodies.1 His observations led to the conclusion that the relative velocity of two bodies after they have impacted each other is in a constant ratio to their relative velocity before impact and in the opposite direction. This constant ratio has been called the “coefficient of restitution”, usually designated by e.2
The ballistic method of investigating materials’ mechanical properties in modern times was initially applied to homogeneous, usually hard, substances such as metals.3 On the other hand, the use of the concept to examine relatively soft, viscoelastic matter is still more recent, i.e., in the middle of this century. Hollinger and Thelen4 extensively studied asphalts in the 1950s using an impacting pendulum to find the storage (elastic) and loss (viscous) moduli. The ball rebound tests of natural and synthetic rubber stocks given in handbooks show the considerable influence of temperature on polymers.3 For example, in these ballistic measurements the percent rebound of a 1.9-cm steel ball dropped 100 cm onto a natural rubber (Hevea) sample 1.9 cm thick is 45% at 20°C and 71% at 100°C. For polychloroprene (neoprene) the change was from 35 to 67% for temperature change of 20 to 100°C, respectively. Arbitrary scales have been of necessity the practice in all of these tests, and the specification of equipment details, such as the geometry of the impacting components, is essential if uniformity of results is expected. For metals the diameter of the indentation made by a small hardened steel sphere (Brinnel hardness) or the height of rebound of a small hammer (Shore scleroscope) serves as arbitrary measures of hardness. In the case of skin ballistometry the depth of penetration to particular layers of the dermis will depend to some extent upon the geometric sharpness of the impacting mass. The sharpness of course determines the instantaneous pressures exerted upon the tissue. Some standardization eventually will be helpful. The ophthalmic applications go farther back to a ballistic tonometer devised in 1930 by Vogelsang5 to assess ocular tension. The method involved photographing the rebound oscillations of a small hammer striking the cornea 627
628
Handbook of Non-Invasive Methods and the Skin, Second Edition
of the eye. Wigersma6 in 1955 apparently discovered the benefit of a lighter hammer in a method referred to as elastometry. Eventually Mamelok and Posner7 published their conviction that these measures had more to do with the mechanical properties of the cornea itself than with intraocular pressure, i.e., more in accordance with Newton’s original contention. In more recent times others have used the method to study skin properties. Here the deeper dermal structures would be influential. Tosti8 made a ballistometer for this purpose. However, one should note that the published analysis must be viewed with caution, because e is defined as an energy ratio instead of the accepted momentum ratio which Newton originally postulated. Thus, without noting this difference, one could be confused by the numerical results of the experiments. The differences result from energy being proportional to the square of the velocity, whereas momentum is mass times the first power of velocity. More specifically, the example given in the reference is of a free-falling body striking a horizontal surface. Actually a pendulum was used, and we will have more to say about pendulums. Basically it is true that kinetic energy of whatever mass system is involved should equal the starting potential energy however it is created. Then the rebound kinetic energy should equal the next potential energy peak. This reference defines e as H′/H, the ratio of rebound height, H′, to the previous starting height, H, as previously discussed. The coefficient of restitution is an important element in any analysis of the motion ensuing after the collision of two bodies. In the most elementary case it is assumed that as soon as contact begins there will be a certain period of time referred to as the “period of compression”. After maximum compression and deformation, recovery to some extent will occur during a time called the “period of restitution”; hence the name of the coefficient. These times are of course fundamentally dependent upon the period required by the materials’ molecular structures to rearrange themselves. Thus one observes the differences in behavior between those substances which are highly elastic and those that are more viscous. Experimental determination of e is traditionally done by measuring the rebound height of one object falling upon another. The expression for e is derived simply as follows:
e=
v2 – v0 v 0 – v1
(73.1)
If it can be assumed that the second body does not move, v0, their common velocity at maximum compression, equals zero, and hence we will have
e=–
v2 v1
(73.2)
In these equations v1 is the velocity of the falling mass before and v2 its velocity after impact and separation. Since these velocities are defined by the starting and final rebound heights, i.e., potential energies, one may apply the following relationships to calculate e. v1 = 2gH
and v2 = 2gh
(73.3)
Then, substituting in Equation 73.1, e=
h h or e 2 = H H
(73.4)
Thus, instrumentally one measures the height, H, from which the mass falls and the rebound height, h, to which it rises. The task of the experimenter is to find a practical way of doing this.
73.2 MECHANICS OF BALLISTOMETRY This elementary determination of the coefficient of restitution assumes that the falling mass and the surface struck are of the same material, that the surface is rigidly supported, its velocity is at all times zero, and hence the velocity of both masses at the instant of greatest compression will be equal to zero. These assumptions are not completely satisfied when ballistometry is applied to the skin. Now we must consider the details of the process further to see how valid our interpretation of the results will be. In general, when the impacting mass strikes the skin surface, several things happen. First, the elastic component of the skin begins to store some of the kinetic energy of the falling object. The subsequent release of this stored energy provides the rebound. The processes of both compression and restitution are slowed however by the viscous component whose reaction force, like a shock absorber, is proportional to velocity. It will be at a maximum at first and then decrease as indentation proceeds, energy is dissipated as viscous internal friction, and the area of contact simultaneously enlarges. The elastic component of force is not dependent upon velocity. A second deviation from simple assumptions is that all of the involved skin does not have zero velocity throughout the impact process. Instead, some parts will have been accelerated, and an exchange of momentum will occur as well as potential energy storage. This momentum initiates an acoustic wave in the skin, however rapidly attenuated, propagating in indeterminate directions. The acoustic energy will be mostly lost, although a small amount may
Ballistometry
be returned to the rebounding mass and contribute to or detract from the reaction depending upon its time phase. This is similar to a diver bouncing up and down on the end of a diving board; if he gets out of synchronism with the board, unpleasant results ensue. With skin much of the indentation energy will be lost in shear viscosity within the tissues, conversion to heat. The greatest rebound will of course occur where the ratio of elastic to viscous effect is largest. Hence the expression that the skin is “more elastic” in such cases is a fair statement and can be used to describe the observed accentuated characteristic in younger skin and for certain beneficial dermal treatments. The coefficient of restitution theoretically ranges from 0 for inelastic impact, i.e., zero after-impact velocity, to 1 for completely elastic impact (which of course never occurs). On this basis ballistometry has come to be useful for evaluating youthful and more mature skin and its response to skin treatment products. For a viscoelastic, nonlinear substance like skin, especially in compression during impact, one observes diminished times between successive rebounds. This is to be expected because the time for a mass to fall is a function of its starting height. Thus, impacts will be closer and closer in time. (See Figure 73.1 for typical rebound recordings.) One additional effect of the decreasing fall height is that the diminished impact velocity produces shallower indentation, involving essentially a different substance composed of the more superficial skin layers. This explains in part why a constant rebound ratio is not maintained. Also, it should be appreciated that, with indentation of soft substances, the area of contact increases rapidly, so that, depending upon the shape of the impacting mass, an altered stress distribution in the region occurs with depth. These last comments direct our attention to the influence of the shape of the impacting mass. This subject interested mathematicians as far back as the 19th century. Heinrich Hertz, the discoverer of wireless waves, solved the problem of impact between solid spheres and presented an extended theory of impact between a solid sphere and an elastic plate, the latter being more like our situation. These relations in mathematical form showed the dependence of the coefficient of restitution upon the material constants, Young’s modulus, density, and Poisson’s ratio.9 More recent works on this topic are the classical analyses of Timoshenko in his Theory of Elasticity in which precise relationships for stress distributions in various geometries are presented.10,11 A pendulous mass that impacts on the test surface is a better way to get control of the several ingredient parameters than by dropping weights and attempting to observe their rebounds. A compound pendulum provides a means of distributing several mass elements to increase its moment of inertia and so control the impact velocity. We can show how this is better than a simple, pivoted hammer
629
(a)
(b)
(c)
(d)
FIGURE 73.1 Sample recordings: (A and B) Reduced period with lower impact on highly elastic surface; (C) improved rebound with lower velocity impact on more viscous surface; (D) comparison of more elastic with less elastic material.
which, because of its uncontrolled impact velocity, produces only minimal results, i.e., unsatisfactory rebounds for soft, viscoelastic substances. Large rebounds are necessary to provide a high degree of discrimination among the varying viscoelastic parameters. Let us illustrate graphically the reason for controlling impact velocity when applying ballistometry to the study of viscoelastic materials. This can be shown with a graph of the force, and hence the work relationships accompanying the periods of compression and restitution. Figure 73.2 shows two plots of force acting on the indenting mass as it compresses the skin downward along the velocity axis at several speeds, one high and the other more slow. In both cases we will assume identical kinetic energy inputs to the same test site. Fundamentally it is the maximum indentation that stores the most elastic, potential, energy that will be returned as a large rebound. Identical energy input means that the total work done by the viscous and by the elastic force, i.e., the area under these curves, must be equal in both cases. It is unimportant in making our point that the exact functional relationship between these forces and the indentation of the skin cannot be known. However, the graphs show the greater apportionment of viscous energy (loss) in the highspeed case. This is of course because the viscous force is proportional to the speed, whereas the elastic force is not. Note also that the slope of the elastic force-displacement characteristic is everywhere the same as determined by the skin involved. The conclusion is that the slow-speed case stores more energy in the elastic element and to do so must mean greater indentation, as shown. Hence, the greater stored energy will occur, and the rebounds will be higher. The
630
Handbook of Non-Invasive Methods and the Skin, Second Edition
Viscous force
Viscous force
E
E O
O
XC
XC Indentation
Indentation
Elastic force
Elastic force
XC = Maximum indentation point
K
K O
O
XC
XC Indentation
Indentation
Work area (fast) OEXC + OKXC = Work area (slow) OEXC + OKXC
FIGURE 73.2 Force vs. indentation (work) diagrams of high and low velocity impact on viscous and elastic substances showing greater rebound potential for the low-velocity impact.
approximate curvature of these force characteristics is introduced to account for the varying area of contact, hence pressure, during compression and restitution. If the contact is assumed to have a conical geometry, the force will vary somewhere between a square and cubic function of the indentation.
73.3 PRACTICAL BALLISTOMETRY FOR SKIN STUDIES The qualitative measures presented in the previous section are sufficient to provide ballistometer design criteria. A practical embodiment of the ballistometer is a form of pendulum in which the pivotal angle is measured instead of mass heights. This may be done conveniently by connecting its shaft to some low-friction angle transducer such as a rotary variable differential transformer which imposes no force on the pendulum. In Reference 8 a stationary coil with its inductance varied by a moving iron core attached to the pendulum was associated with an electronic circuit to indicate angle. To directly obtain numerical values for computer entry high-resolution, digital, optical angle encoders are available. In most cases the ballistometer will operate with large angles of swing, so that some trigonometric processing will be required as
will be discussed in more detail. Still another approach is to employ an angular accelerometer whose output is integrated to give angular velocity. Programmed with a computer to deliver information precisely at the moments of impact and rebound, this arrangement can directly calculate the coefficient of restitution. Other useful functions of the computer are the establishment of a time base for release of the pendulum from its raised position and graphical display of the rebound pattern. Release of a pendulum type of moving mass will result in a certain velocity of normal impact, R dU/dt, upon the test surface, where R is the effective length of the pendulum (pivot-to-impacting point). The analysis of the pendulum’s motion must now recognize certain practical design features based upon the theory presented in the previous section. The desirability of a low-velocity, compound (physical) pendulum with a considerable moment of inertia has been discussed in terms of improved rebound. Figure 73.1 shows this improvement in measurable rebound amplitude, achieved simply by adding an adjustable counterweight to a pivoted beam assembly. Such a ballistometer is pictured in Figure 73.3 showing a bar with impacting mass and counterweight pivoted on an angle transducer.
Ballistometry
631
M1
r1
L
θ
C r2
C = Center of mass h or H
FIGURE 73.3 Ballistometer with compound pendulum and angle transducer. Impacting mass is on the right, counterweight on the left.
The angular acceleration of a compound pendulum determines its ultimate velocity. For any rotating body the angular acceleration is equal to the torque applied to it, in this case gravitational, depending upon its instantaneous angular position, divided by its moment of inertia about the pivot point. Thus the angular acceleration and hence the impact velocity can be easily adjusted by changing the positions of the impacting mass and the counterweight to alter the net gravitational torque and the moment of inertia of the system. The latter is the moment of inertia of the pivoted beam plus the contributions of the two masses. The moment of inertia of the system can be measured by timing its natural period of oscillation about the pivot and using the following relationship: 2
T I = ⎡⎢ ⎤⎥ mgL ⎣ 2π ⎦
(73.5)
I is the moment of inertia, T is the period, m is the total mass, g is gravitational acceleration, and L is the distance from the pivot to the center of mass. L can be found by balancing the beam and masses on a knife edge. The differential equation of motion for the system is d2Θ/dt2 – (mgL/I) cos Θ = 0
(73.6)
where Θ is the pivotal angle referenced to the horizontal plane of impact. Solution of the kinetic equation of motion is not absolutely necessary to the understanding, design, and use of a practical ballistometer. We can analyze the compound pendulum to see how the coefficient of restitution can be derived from energy and moment considerations. In Figure 73.4 the center of mass of the pendulum is shown at a distance L from the pivot. If the pendulum is to rotate clockwise when released, and the beam is pivoted at its midpoint, the center of mass must be to the right of
M2
FIGURE 73.4 Compound pendulum mechanics.
the pivot. Taking moments about the center of mass to find its location one obtains L=
M1 r1 – M 2 r2 M1 + M 2
(73.7)
If we can make the two masses equal, L becomes further simplified. When M1 = M2 = mg, then L=
1 (r – r ) 2 1 2
(73.8)
We will use L to calculate the effective height to which the mass of the system is raised for a given starting angle of the pendulum, Θ1, relative to the horizontal. Thus the potential energy will be obtained as PE = 2mgLsinΘ
(73.9)
This potential energy can be equated to the kinetic energy to find the velocity of impact as well as the rebound velocity, both of which are needed to find the coefficient of restitution. The kinetic energy of such a rotating system is 1/2 I (dΘ/dt)2. Solving for velocity gives v = r1 (dΘ dt ) = r1 2 mg(sin Θ) I
(73.10)
Considering this relationship for v1 where Θ1 is the starting angle of the pendulum and v2 where Θ2 is the rebound angle, the value of e may be calculated exactly as
e=
v2 = v1
sin Θ 2 sin Θ1
(73.11)
632
Handbook of Non-Invasive Methods and the Skin, Second Edition
This is the basic equation that should be used to derive the coefficient of restitution in a system that measures pendulum angles. This is in conformity with the definition of e in Equation 73.2. Equation 73.11 can be verified by showing it to be identical with Equation 73.4. This is done by substituting under the radical in Equation 73.11 the geometric expressions for sinΘ2, which is h/L, and the equivalent of sinΘ1, which is H/L. The L cancels, and one is left with the square root of h/H or Equation 73.4. We have shown how the potential energy transfer to rotational energy and back to potential is accomplished and does not alter the impact conditions. The point is made also that the angular measure is a simple way to register these processes, and the necessary trigonometric computation is not complicated. In doing this one gains control over the impact velocity which is critically important in applying ballistometry to viscoelastic materials such as the skin.
An impressive amount of data (in vivo) can be obtained with a ballistometer in a short time, particularly if the information is processed on-line via computer.12 The computer may also control a solenoid release of the ballistometer’s pendulum and start the rebound angle recording process. This may be done also with a manual electronic system. Images such as those shown in Figure 73.1 may be recorded on the screen of a storage oscilloscope whose sweep is synchronized to begin with the pendulum release. Alternatively, an X-Y plotter with good response and writing speed might be used to record raw bounce angle data. These would have to be corrected trigonometrically in accord with Equation 73.11 to obtain accurate coefficients of restitution. Experiments should be done under the same atmospherically controlled conditions as would be applied to skin studies. Subjects should be allowed to equilibrate prior to testing.
73.4 DATA OBTAINED WITH BALLISTIC MEASUREMENTS
REFERENCES
The fundamental quantity sought in ballistometrics is the coefficient of restitution. Tests on the skin as a specific extension of the measurement from its use on other materials have been reported to clearly distinguish elastic modulus differences between young and old, various body sites, as well as similar changes after pharmaceutical treatments.8,12 Ballistometry is attractive for several reasons. It is noninvasive and easy to use. No probes have to be attached to the skin. The instrument is not as expensive as for example the dynamometer. Although one cannot obtain data on the status of the stratum corneum as one does with shear measurements, it provides a practical indication of underlying tissue changes, i.e., for example, expansion from retin-A or other similar treatment. As has been pointed out, the depth of tissue responding to ballistic impact will depend upon the height to which the falling mass is raised. Thus specific skin layers will be responsive in each case. Skin structure variations between body sites and between individual subjects will of course have an effect. However, in any series of tests planned to study response to a greater or lesser depth it is advisable to select an impact velocity and not change. The moment of inertia of the pendulum, although variable, is usually not altered after the performance of the apparatus has been optimized.
1. Timoshenko, S. and Young, D.H., Engineering Mechanics, McGraw Hill, New York, 1940. 2. Loney, S.L., A Treatise on Elementary Dynamics, Cambridge University Press, New York, 1900. 3. Handbook of Chemistry and Physics, Chemical Rubber Co., Cleveland, Ohio, 1946, pp. 30, 1302, 2328. 4. Hollinger, R. and Thelen, E., Design and development of an impact tester for use with asphalt, Report R-11, National Asphalt Research Center, Franklin Institute, Philadelphia, 1956. 5. Vogelsang, K., Ueber mechanische Gewebsprüfung am Auge, Arch. f. Augenh., 108, 714, 1934. 6. Wiegersma, G., Elastometry of the eye, Am. J. Ophth., 39, 811, 1955. 7. Mamelok, A.E. and Posner, A., Measurements of corneal elasticity, Am. J. Ophth., 39, 817, 1955. 8. Tosti, A., Giovanni, C., Fazzini, M. L., and Villardita, S., A ballistometer for the study of the plasto-elastic properties of skin, J. Invest. Dermatol., 69, 315, 1977. 9. Hertz, H., J. Math. (Crelle), 92, 1881. 10. Timoshenko, S., Theory of Elasticity, McGraw Hill, New York, 1934, p. 390. 11. Lamb, H., Proc. London Math. Soc., 35, 141, 1902. 12. Fthenakis, C.G., Maes, D.H., and Smith, W.P., In vivo assessment of skin elasticity using ballistometry, J. Soc. Cosmet. Chem., 42, 211, 1991.
The Cutaneous Vasculature Skin Color and Blood Vessels
74 Colorimetry Wiete Westerhof Department of Dermatology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands
CONTENTS 74.1 Introduction............................................................................................................................................................635 74.2 Aim ........................................................................................................................................................................636 74.3 Methods..................................................................................................................................................................636 74.3.1 CIE Color System......................................................................................................................................636 74.3.2 Technical Details of the Colorimeters.......................................................................................................637 74.4 Sources of Error.....................................................................................................................................................639 74.4.1 Instruments.................................................................................................................................................639 74.4.2 Measurement Method ................................................................................................................................639 74.4.3 Reproducibility ..........................................................................................................................................639 74.5 Correlation with Other Methods ...........................................................................................................................639 74.6 Fields of Application .............................................................................................................................................641 74.6.1 Ultraviolet Radiation-Induced Erythema Measurement ...........................................................................641 74.6.2 Measurement of Erythema Due to Irritants and Contact Allergens .........................................................642 74.6.3 Measurement of the Blanching Effect of Corticosteroids28 ......................................................................642 74.6.4 Measurement of Skin Color1 .....................................................................................................................643 74.6.5 Measurement of Ultraviolet-Induced Pigmentation..................................................................................643 74.6.6 Measurement of Dose-Response Curves of Ultraviolet-Induced Erythema and Pigmentation52 ............644 74.6.7 Measurement of Bleaching Effect by Depigmenting Agents ...................................................................645 74.7 Conclusions............................................................................................................................................................645 Acknowledgments ...........................................................................................................................................................645 References .......................................................................................................................................................................645
74.1 INTRODUCTION Color information may be acquired and communicated in various ways. However in a scientific context there are requirements for consistency independent of time, distance, and language.1 Although the average person may be able to distinguish several thousands of colors, the description of visual observations by the use of general color terms is the least satisfactory in terms of precision. The range of names upon which people reliably agree is very limited and there is no simple way of using visual observations to describe the differences between colors.2 Color assessment based on comparison with sets of colored samples, known as color-order systems, can increase the range and reliability of color designations significantly.2 The Munsell color-order system3 is the oldest and perhaps the most widely recognized. Color comparisons may involve metamerism, that is: the actual physical composition of the colored lights might be
different but these are perceived as the same color. Therefore color-order systems are most reliable when used under fixed conditions of illumination to match colors of relatively flat surfaces that are uniformly pigmented. They are less appropriate for samples that are heterogeneous in surface structure. A particular limitation of color-order systems is that they only identify single colors and do not provide a way of specifying the nature or magnitude of color differences. The definitive method of measuring skin color uses the recording spectrophotometer adapted for reflectance readings. However these readings are not communicable, as they do not conform to an international standard. The limitations of visual observations may, in principle, be overcome by the instrumental evaluation of colors and color differences according to the system of color measurement established by the CIE (Commission Internationale de l’Eclairage).2,4–6 This system has become widely 635
636
Handbook of Non-Invasive Methods and the Skin, Second Edition
used with the availability of reliable reflectance spectrophotometric instrumentation that conforms to the recommendation of the CIE for the measurement of the color of surfaces.7–9 This method does not give information about the substances generating the color, but it is highly appropriate for color matching (e.g., grading of erythema and grading of melanin pigmentation, etc.). In fact, the use of the CIE tristimulus values is a concise and reliable way of approximating an actual normal color observer. In other words, by using these units one can imitate what is normally done by visual inspection but with greater reliability and reproducibility than a given human observer could achieve.
Yxy color space y
0.8
0.6
0.4
74.2 AIM The objective of this chapter is to present qualitative and quantitative measurement methods of color of the skin, mainly erythema and melanin pigmentation. This is done with a tristimulus computer-controlled color analyzer which measures reflected object color in an accurate and reproducible way utilizing CIELAB (CIE 1976 L*a*b*) color space values. These CIE color space parameters are proposed for the unambiguous communication of skin-color information that relates directly to visual observation of clinical importance or scientific interest.
74.3 METHODS 74.3.1 CIE COLOR SYSTEM The perceived color of objects depends on: (1) the nature of the illuminating light, (2) its modification by interaction with the object, and (3) the characteristics of the observer response. The CIE system defines these conditions as follows: (1) The relative spectral energy distributions of various illuminants, known as CIE standard illuminants, are specified and available as published tables,2,4–9 (2) the modification of an illuminant by interaction with the object is measured with a reflectance spectrophotometer having an optical configuration that conforms CIE recommendations,4 and provides a visible spectrum expressed as the fractions of incident light intensity reflected in the wavelength range 400–700 nm; (3) the nature of human color vision has been quantified for the purpose of color measurement in terms of three color-matching functions x, y, and z (Figure 74.1). Three are required because color vision has been found to be trichromatic: a single perceived color may be regarded as resulting from the effect of three separate stimuli on the visual cortex.6 Their numerical values are available as published tables and are known collectively as a CIE standard observer.2,4,5,9 They may be regarded simply as a numerical description of average human color vision. From a practical viewpoint
0.2
0
0.2
0.4
0.6
x
FIGURE 74.1 X, Y, and Z color coordinates.
their use has been made more convenient by incorporation of the tabulated values within the software provided with color-measuring reflectance spectrophotometers. Similarly, the tabulated values of the relative spectral-energy distributions of various CIE illuminants are provided within the software because they do not exist as actual physical sources of light within instrumentation. Colors are measured in terms of their tristimulus values X, Y, and Z by combining a selected table of illuminant data, the measured values of reflectance, and a selected table of color-matching functions with three summations, each having the form Σ ER x– = X At selected intervals in the wavelength range 400 to 700 nm the relative energy (E) of the chosen illuminant is multiplied by the fraction reflected (R) and the numerical value of the standard observer (x or y or z). A wavelength interval of 10 nm, which requires 31 terms in each summation, gives adequate precision for most purposes. Modern color-measurement instrumentation normally incorporates microcomputer hardware and software so that the spectral measurements and subsequent calculations are integrated so as to produce a copy of the results within a few seconds. However the tristimulus values of colors are difficult to relate to the experience of seeing them. In addition, in any study involving comparisons, contrasts, or changes, tristimulus values do not directly enable
Colorimetry
637
L∗= 100 A ΔE A’
+b∗ h0
B −a∗
+a∗
−b∗
L∗ = 0
FIGURE 74.2 Color space; CIELAB.
measurement of the difference between two colors. This concern has now been overcome by using the tristimulus values to calculate the CIE 1976 L*a*b* (CIELAB) color space values.2,4,7,9 The mathematical manipulations that convert tristimulus values to CIELAB color space values enable colors to be regarded as existing in an approximately uniform three-dimensional space in which each particular color has a unique location defined in terms of its cartesian coordinates with respect to the axes L*, a*, and b* as shown in Figure 74.2. Modern computer-controlled reflectance spectrophotometers provide for automatic calculation of CIELAB values from the spectral data they produce. The measured value of a color has been recommended by the CIE as the psychometric correlate of the visually perceived color attribute of “lightness”,4,7 to which the descriptive terms assigned might include the words “light”, “dark”, etc. In other words, L* would measure the change along a gray scale from black to white that visually varied in perceptually uniform manner. The L* scale, which ranges from 0 for a theoretical black to 100 for white, corresponds to the notion of the value attribute in the Munsell color system. The a* and b* coordinates may be conceptually related to Hering’s opponent color theory,8 which was based on the proposition that the retina of the eye contains opponent color channels that distinguish colors according to their red-vs.-green and yellow-vs.-blue attributes. In CIELAB space they are more useful when converted into polar coordinates. This facilitates the definition of a hue angle, h* = arctan (b*/a*) (Figure 74.2), which is recommended by the CIE as the psychometric correlate of the visually perceived attribute of hue (e.g., red, orange, yellow, etc.).4,8 Measured hue angles make the use of visually assigned hue terms unnecessary, although it is simple and
often convenient to relate them in a general way. The general angular position of some of the main generic hues are shown in Figure 74.1. CIE hue angle corresponds conceptually to the attribute of hue in the Munsell colororder system, but no simple relationship has been found between measured hue-angle values and Munsell hue designations.10 Colors for which both a* and b* are zero, and therefore lie on the L* axis, are termed achromatic and would be perceived as gray, white, or black. The visually perceived color attribute of “saturation”, which might be described by the use of the terms “weak”, “strong”, etc., may be measured in terms of its distance away from the L* axis in the a*b* plane. This is the length of the line C in the diagram. It is termed the CIE(1976)a*b* chroma and is calculated using coordinate geometry as C = [(a*)2 + (b*)2]1/2 It corresponds conceptually to the attribute of chroma in the Munsell system but the measured values do not relate in any simple way to Munsell designations.10 Thus the use of CIELAB coordinates enables measurement of the three attributes of a color by which it is visually distinguished. CIELAB space is not only more convenient than tristimulus values with respect to its conceptual relationship to the actual experience of seeing colors but it has the important advantage of providing a means of measuring the differences between any two colors.2,8 Their color difference (E) is calculated, using coordinate geometry, as the length of the line joining their coordinate positions: E = [(L*)2 + (a*)2 + (b*)2]1/2 In some circumstances the differences between two colors may also be considered in terms of differences in hue angle and/or chroma. Modern computer-controlled reflectance spectrophotometers used for color measurement often include the software necessary for automatic calculation of color differences from two sets of spectral data. For more details the reader is referred to the appendix of the excellent paper by Weatherall and Coombs.1
74.3.2 TECHNICAL DETAILS
OF THE
COLORIMETERS
Several colorimeters have been manufactured for medical and scientific use. Here we discuss three of the most commonly used instruments. 1. Labscan 6000® (Hunter Associates Inc., USA) scanning reflectance visible spectrophotometer has 0° illumination and 45° viewing geometry with the specular component excluded. The instrument is calibrated with a supplied white standard traceable to the National Bureau of
638
Handbook of Non-Invasive Methods and the Skin, Second Edition
Standard’s perfect white diffuser. The spectrophotometer is controlled by an IBM-XT microcomputer, which performs all color calculations from the digitized spectral data by means of a menu-driven set of programs supplied with the instrument. The skin is measured over a 30-mm diameter open circular port with a 26-mm illuminated area in the horizontal upper surface of the sensor module. Reflectance spectra over the wavelength range of 400 to 700 nm, requiring about 3 s for measurement, can be obtained. Tabulated data for CIE illuminant D65 and the CIE 1964 10′ standard observer are selected under software control and combined with the spectral data at 10-nm intervals to compute the CIELAB L*a*b* values. The latter two are then further converted to CIELAB color space and chroma. The instrumental setup is not suitable for field conditions and routine clinical applications because of its volume and weight. 2. The Minolta Chromameter CR 200® (Osaka, Japan) is a lightweight and compact tristimulus color analyzer for measuring reflected object color (Figure 74.3). Utilizing high-sensitivity silicon photocells, filtered to match CIE Standard Observer Response, the Chromameter CR 200® assures good accuracy and reproducibility (measuring error 1%). Readings, taken through the measuring head, are processed by a built-in microcomputer and then presented digitally on a liquid-crystal display. The measuring head
FIGURE 74.3 Chromameter.
also contains its own standard light source: a high-power xenon arc lamp which provides diffuse illumination from a controlled angle for vertical viewing and constant, even lighting on the object. The illumination chamber of the Minolta equipment is cylindrical, with a conical opening pointing toward the skin surface. These two parts are separated by a diffusing plate with an inner circular aperture 4 mm in diameter. The outer aperture is 11 mm, and the diameter of the estimated measuring area is 8 mm. The photo receiver is centered in the top of the chamber, and the tristimulus optical analysis unit with six silicon photocells is located directly in the probe unit (in the CR series), and in previous equipment in the main body. The meter’s precise double-beam feedback system, using six photocells, detects any slight deviations in the xenon light’s spectral distribution, and the microcomputer compensates for them, thus ensuring the utmost accuracy in measurements. The measuring area is circular and 8 mm in diameter. The Chromameter offers five measuring modes: two chromaticity-measuring modes — Yxy and L*a*b* — and three colordeviation measuring modes — ± (Yxy1 ± (L*a*b*) and Eab). The weight of the Minolta CR-200 is 1951 g. 3. The Dr. Lange Micro Color® (Dr. Bruno Lange GmbH, Dusseldorf, FRG) is computerized with a number of facilities. L, a*, and b* values are presented on a digital display. It is based on illumination with a xenon flash light. In the Lange equipment the chamber is spherical with a circular aperture, 5 mm in diameter, directed toward the skin surface. Measuring and reference photo receivers are mounted on the top of this sphere, but eccentrically aligned by 6° and 8°. Optical fibers lead the signal to the main body containing six silicon photocells for tristimulus analysis (red, green, blue) of both signal and reference light, according to DIN 5033. The Lange equipment has a short cable (600 mm) between probe and main body. The weight of the Lange equipment is 6 kg. It works at a slightly slower rate than the Minolta equipment. The technical reproducibility of the two colorimeters as given by manufacturers is the same, i.e., 0.15 E* to white. Both are calibrated against their respective white calibration tiles before use. For the practical handling of the colorimeters the reader is referred to the instruction manuals of the respective apparatus.
Colorimetry
639
74.4 SOURCES OF ERROR
12
74.4.1 INSTRUMENTS
74.4.2 MEASUREMENT METHOD The skin is not a homogeneous surface structure. It differs in primary, secondary, and tertiary skinfolds depending on the site of the body. Also the greasiness and humidity of the skin differ from site to site. This will affect the glossiness of the skin. Similarly the hairs, blood vessels, blemishes spots, and scars may determine the color aspects of the skin. The reflection from the skin is influenced by all the above factors leading to incorrect color measurement and significant site-related variations in reflectance. It is therefore imperative to carefully select the skin sites to be measured and to take the measurements in duplicate or triplicate. If erythema is to be measured it has to be borne in mind that this is a function of blood vessel size and flow. These parameters can be influenced among other things, by compression. Therefore the measuring head with the aperture must not be pressed too hard against the skin. The measuring head must also be kept in a perpendicular fashion to the skin surface. Otherwise differences in reflectance of the light source will occur.
74.4.3 REPRODUCIBILITY Another source of error lies in the method of provoking the color change. In UV-induced erythema a 3-plus visual grading with edema may be less red due to capillary compression and extracellular fluid accumulation. The time lapses of UVA and UVB induction is different. Therefore the moment of measurement is critical. If
10
8
Δa*
It is not possible to go into much technical detail within the framework of this chapter. A source of error could be the illumination. A light source comparable to the sun does not exist. Most of the artificial light sources emit incontinuous bands out of the whole light spectrum. This may give rise to metamerism. The filters used in this tristimulus chromameter are not exactly monochromatic. Especially at the periphery of the wave band of absorbtion, deviations in the filtering capacity exist that can lead to incorrect color measurement. Also the liquid or glass fiber light guide can contribute to inhomogeneous loss in light spectrum constituents. The angle of illumination of the object being measured, in our case the skin, determines the degree of specular reflection which negatively influences the color analysis of the reflected light. The lack of standardization in the production of different colorimeters will lead to incomparable results of measurement. One given instrument is usually very reproducible in measurements of the same object.5
6
4
2
29.4
47.0
75.2
117.5
UV-B dosis in mJ. cm−2
FIGURE 74.4 UV sensitivity of the skin related to body site. Diagram of mean Δ a* vs. irradiation doses given to 5 body regions of 8 volunteers: •-• right scapular back, - right lumbar back, - distal arm, - medial arm, ▫-▫ proximal arm.
follow-up measurements are required from one and the same measuring site it is necessary that care is taken to meticulously place the aperture over exactly the same place as chosen for the previous measurements. As was already explained a few milimeters away from the measuring site the characteristics of the skin may be completely different. We have devised measuring bracelets that allowed us to exactly measure the UV-irradiated sites of the skin of the lower arm for the development of erythema and pigmentation (Figure 74.4). For other sites of the body similar appliances could be devised. These devices help to increase the reproducibility of the measuring technique. In Section 74.6, Fields of Application, the methodology of inducing color changes or measuring color changes over time need to be standardized in order to be able to make chromametric measurements in a reproducible way.
74.5 CORRELATION WITH OTHER METHODS Reflectance chromameters have not been compared with spectrophotometers. However Weatherall et al.1 used a spectrophotometer, which converted the reflected wavelengths taken at 10-nm intervals into CIELAB values. Serup et al.11 evaluated two commercially available colorimeters (Minolta chromameter and Dr. Lange microcolor) and compared them with laser-Doppler flowmetry,
640
Handbook of Non-Invasive Methods and the Skin, Second Edition
which is widely used for quantification of erythema and cutaneous inflammation. Irritant reactions after application of sodium lauryl sulfate (SLS) were studied. Evaluations of technical reproducibility using white tiles and a red standard simulating erythema are presented. Repeated measurements (n = 10) were performed with the two colorimeters on their respective white calibration tiles and on a red color standard simulating moderate erythema.12 The difference between the two standard tiles was measured by the Lange equipment. The linearity of the colorimeters was examined using different color scales from the color-standard book.12 Red (erythematous scale) was registered as mainly parallel a*-axis curves both in the lower end, relevant for erythema measurement, as well as in the upper end. The Lange equipment showed a steeper increase in values in comparison with the Minolta equipment. The Lange equipment appeared more sensitive for determination of yellow, with a steeper increase in values on the b* axis. The equipment determined blue with parallel curves. However, the Lange equipment measured at a substantially lower level, i.e., about 40 U lower on the b* axis. Green was also determined at a lower level, on the a* axis, by the Lange equipment, especially at high intensities. In the three axes only a small part of the large dynamic range of the colorimeters was relevant for skin color and erythema measurement. Color measurement by colorimeters using the CIE system was useful for the characterization of erythema, confirming previous studies.3,4 In this paper dealing with light-induced erythema a more detailed evaluation of erythema was performed. Serup et al.11 concentrated on the technical aspects of color measurement. Laser-Doppler flowmetry proved useful for the characterization of erythema as shown already, with a positive correlation to clinical scoring and colorimetry (a*-axis movement toward red).15,16 The colorimetry and the flowmeter probably measure somewhat different features of cutaneous circulation, i.e., the colorimeters mainly measures the capillary accumulation of blood and the flowmeter measures the total blood flow, mainly determined by the arteriolar tone.17 Thus, the methods are, to some extent, complementary in the evaluation of erythema. The technical reproducibility of the two colorimeters was very good with low coefficients of variation, although higher for a* then for the other axes. The variation in color in the group of subjects studied was remarkably low with respect to the L value, and limited with respect to the a* value. The high coefficient of variation of b*-axis values recorded with the Dr. Lange equipment was partly attributable to the low mean values. Standard deviations might indicate a similar reproducibility of colorimeters in control skin and erythematous skin in contrast to laser-Doppler flowmetry, which showed a five times higher standard deviation in erythematous skin.
This is probably due to the small area of skin illuminated by the laser light in contrast to the colorimeters, which operate with a far larger area and thus provide a better overall assessment of the circulation. Thus, with the laser-Doppler it is of special importance that results are obtained through averaging a number of recordings of the irritant reaction studied. In contrast, a single colorimeter measurement may suffice, depending on the situation or the purpose. It is well known that laser-Doppler flowmetry, recording a dynamic situation, is sensitive to factors such as noise and talking. Colorimetry is not affected by such environmental sources of variation, making the method more suitable for routine work. The two colorimeters were essentially equally suited as tools for quantification of erythema. However they do not give identical values, and the technical part of the study concluded that they record some differences on the different color axes. Thus, a simple transformation factor cannot be applied. The CIE color system has the advantage of being internationally accepted, and equipment using this system is available. Modifications were developed by Munsell and more recently by a Swedish group, the latter system being named NCS (natural color system).18 This system was suitable for assessment of skin color during Argon-laser treatment of port wine stains. However, for the wider exchange of information it is advantageous if a uniform system is used, although experts may not find it ideal in every situation. It should not be forgotten that inflammatory responses start with vasodilatation, while edema formation takes over in advanced reactions, compressing the vasculature. Laser-Doppler flowmetry and high-frequency ultrasound measurement of skin thickness of histamine wheals, irritant reactions after application of SLS and allergic patchtest reactions show that grading of reactions relative to vasodilatation is only possible in light and moderately severe reactions.16,19,22 However, ultrasound examination also shows the development of edema within irritant reactions is less in comparison with allergic reactions. Thus, in studying irritancy, colorimetry is likely to be suitable for the grading of a relevant spectrum of reactions, which previous studies also demonstrated.13,14 In conclusion, Serup et al.,11 found that colorimetry using the CIE system is reproducible and useful for quantification of erythema, even under busy laboratory conditions. In group studies (winter) using untanned skin, the influence of melanin pigmentation appears negligible. However seasonal and regional differences in skin color need to be studied in more detail. The relevance and potential areas of application in experimental and clinical dermatology are many since this color system takes into account the actual color perception of the human eye, which spectrophotometric recording of light absorbency and reflectance does not do.
Colorimetry
74.6 FIELDS OF APPLICATION The application of reflectance chromameters in clinical practice and dermatological research is wide and still expanding. The most important applications reported in the literature are measurement of: 1. 2. 3. 4. 5. 6.
UV radiation-induced erythema erythema due to irritants and contact allergens the blanching effect of corticosteroids skin color UV-induced pigmentation dose-response curves of UV-induced erythema and pigmentation 7. the bleaching effect of depigmenting agents
641
determined with the aid of a Zeiss M4 QIII® monochromator and an EG&G 550® radiometer at a distance of 25 cm from the xenon-arc source. The UVB output was measured with an EG&G 550® radiometer fitted with a calibrated UV-enhanced silicon detector probe, on top of which a Scott UVB filter was mounted. This filter matches the erythematogenic action spectrum of normal skin at the long wave side as presented by Berger.25 The monitored UVB output of the solar simulator per second was 2.35 mW/cm2. After 24 h all irradiated sites were visually scored by two independent investigators according to the following erythema grading scale: 0 = no perceptible erythema /2 = slight or partial erythema 1 = minimal erythema with defined borders (MED) 11/2 = three arbitrarily increased steps of erythema formation without edema or vesiculation 2 (2 + erythema in the commonly used scale) 21/2 3 = erythema with induration and blisters 1
74.6.1 ULTRAVIOLET RADIATION-INDUCED ERYTHEMA MEASUREMENT The assessment of sensitivity of human skin to UV radiation is important to photochemotherapy, diagnosis of photodermatosis, photoaging, photocarcinogenesis, and photoprotection. As an example we use the method described by Westerhof et al.5 Eight healthy volunteers (four males, four females, 22 to 26 years old) with skin type II and III were irradiated on seven different regions of the body with a solar simulator. The regions were chosen following the reports of Farr and Diffey23 and Olson et al.24 On the ventral side of the left forearm, the three sites were: proximal near the elbow, distal near the wrist, and in between. On the subject’s back the four sites chosen were related to the vertebral level: upper dorsal (T3) — a spot to the left and a spot to the right of the spine, and for lumbar (L2) — a spot to the left and a spot to the right of the spine at 1 cm from the midline. Care was taken to select the spots in such a way they were without terminal hairs and without nevi or scars, etc. The test areas were then exposed to a series of four simulated solar radiation dosed increasing in delivered energy dose by a factor 22/3 of the foregoing dose. The radiation dose was chosen in such a way that one or two test spots (out of four) received a dose equal to one minimal erythema dose (MED). The solar simulator consisted of a 1000-W xenon arc in a lamphouse fitted with a quartz collimating lens (f/0.7) and a water filter with quartz windows. A dichroic (‘cold type’) mirror which reflects most of the radiation below 500 mm was placed into the beam to minimize the infrared radiation further. A liquid light guide conducted the radiation from the exit port of the lamp optics to the volunteer’s skin. An applicator that housed a quartz lens was attached to the proximal end of the light guide and produced a uniform beam of radiation, 10 mm in diameter, on the skin. The stimulated solar emission curve was
Thereafter the erythema was measured with the Chromameter II Reflectance.® The same site was measured twice on the L*a*b* mode. Of the three chromaticity values L*a*b*, the a* value is the best index for the erythema measurement. This is to be expected since a* stands for chromaticity coordinates ranging from perceived red to green in the three-dimensional color space diagram (CIE 1976). After measuring the erythematous spot, a piece of nonirradiated adjacent skin was measured for comparison. The difference between the two gives the erythematous color aspects of the skin (a*). The Chromameter was always held perpendicular to the skin surface, hardly touching it. The aperture of the measuring unit is fitted with an applicator so that the cutaneous capillaries are not compressed. Angle errors in the vertical measuring position of the instrument’s head did not show any influence on the a* and b* chromaticity coordinates and gave a slight deviation of L* (coefficient of variation 6, 8%; n = 20). The differences between duplicate readings showed a SD of 0.30 in case of L* in the range 57.0 to 67.9; a SD of 0.14 in case of a* in the range 6.5 to 20.7 and a SD of 0.21 in case of b* in the range 11.8 to 22.2. The L* value decreased when erythema developed, indicating some skin darkening but to a relatively smaller extent than the increase of a*. The b* values did not change significantly in our erythema measurements. Westerhof et al.5 tried to relate the visual ratings (0, 1/2, 1, 11/2, 2, 21/2, 3) to the Chromameter measurements.
642
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 74.5 Bracelet to make follow-up measurement with chromameter possible. The measuring head exactly fits into the holes of the bracelet, which can be applied to the lower arm in only one way as a notch should exactly overlay an ink marking made on the skin.
In Figure 74.2 the mean visual grading is plotted as a function of a*. It is easy to see that in the range of 1 to 11 there is a linear relation between visual grading and the Chromameter reading. Above a Chromameter reading of 11 a saturation of the visual grading can be seen. Westerhof et al.5 were able to confirm the findings of Olson et al.24 and Farr and Diffey.23 The sensitivity to erythema formation increases when one compares the distal part of the flexor side of the forearm with the proximal part near the elbow while keeping the energy dose of UVB radiation constant (Figure 74.5). The same is true on the subject’s back when comparing the thoracal region (T3) with the lumbar region (L2). It is also remarkable that the back is more sensitive to UV radiation than the flexor side of the forearm.
74.6.2 MEASUREMENT OF ERYTHEMA DUE TO IRRITANTS AND CONTACT ALLERGENS In dermatology and the cosmetic sciences tristimulus colorimetric measurements of erythema offer an advantage in quantifying skin reactions to allergic reactions evoked by patch tests and irritation responses. We studied the role of melanin in inflammatory reactions by provoking erythema with the potent irritant anthralin in vitiliginous skin versus normal pigmented skin.26 The locations chosen for anthralin application were the trunk, arms, and legs under the condition that normal control skin was to be as close as possible to lesional skin. These precautions were taken because of the known existence of a site-dependent inflammatory reaction to anthralin.27 Anthralin was used in concentrations of 0.1, 0.5, 1.0, and 5.0% dispensed in a cream containing 1% salicylic acid as a stabilizer; the vehiculum served as a control. The serial applications of the four concentrations of anthralin were done with
patch-test chambers. The patch-test chambers were fixed with hypoallergic cloth tape. The patch-test chambers were removed after 24 h and the irritation response was read after the next 24 h. For evaluation of the irritation response the following semiquantitative scale was used: 0, no erythema; 1, erythema (+, slightly visible erythema, ++, moderate erythema, +++, intense erythema, ++++, extreme erythema); 2, erythema with induration; and 3, erythema with induration and blisters. To avoid an optical illusion, a white paper was used to cover the test area so that the surrounding reference area was always the same. For an accurate quantitative evaluation of the anthralin-induced erythema we used the Chromameter Reflectance II® (Minolta). The control substance did not induce any inflammatory reaction. The erythematous reactions with 0.1% anthralin were weak, those with 0.5 and 1.0% tended to be stronger, and those with 5.0% showed a very intense erythema. The irritation response never reached beyond the margins of the test chamber and at the time of reading, neither induration nor blisters were observed. With the visual grading the irritation response was considered to be more intense in normal skin than in vitiliginous skin. In contrast to the visual estimations the chromameter readings indicated that the changes in redness due to anthralin were more pronounced in vitiliginous skin than in normal pigmented skin. The staining produced by anthralin appeared to be regular and of the same intensity when vitiliginous and neighboring pigmented test sites were compared. This fact was confirmed by colorimetric measurements. The relation between the degree of pigmentation and the irritation response to anthralin could only be evaluated making use of a sensitive colorimetric method. Visual estimation of the erythema can be misleading as the redness observed in the pigmented skin is known to be comprised of genuine erythema and a red component of the complex brown color. The human eye cannot discriminate these two different sources of redness. Colorimetric measurements can help to overcome this difficulty. From this study it was concluded that melanin pigment had an inhibitory effect on the anthralin-induced irritation response probably by scavenging free radicals and reactive oxygen species.
74.6.3 MEASUREMENT OF THE BLANCHING EFFECT OF CORTICOSTEROIDS28 The vasoconstrictor assay 29 (human skin-blanching assay) is a well-established method for ranking topical steroids. Cornell and Stoughton30,31 gave a rank for U.S. formulations with seven orders of potency. In Europe, the current classification for topical corticosteroids has four ranks,32,33 with group I being the most potent. In this assay the visual assessment of blanching generally gives good
Colorimetry
643
results, but under certain circumstances is not a very satisfactory method, e.g., nonoccluded tests for weak or moderately potent formulations. Many studies using quantitative instrumental methods have attempted to improve the objectivity of this test. Some have completely failed34,36 and those that succeeded were too cumbersome and time consuming for general use.37,39 The only study using a colorimetric system40 was not convincing and this method was abandoned. Queillo-Roussel et al.28 used the Minolta tristimulus colorimeter to quantify the blanching effect of topical corticosteroids in a nonoccluded vasoconstriction test. To investigate the influence of time on variations in colorimetric parameters, an initial series of measurements was performed on day one on six predetermined sites on the ventral surface of the forearm of six healthy volunteers every 2 h over a 12-h period. The colorimetric values were shown to be site related but hourly variations occurred with similar profiles for all sites. On day two, four topical corticosteroid creams, representative of their potency group, as well as a cream base were applied in a randomized double-blind manner on five predetermined sites. Visual gradings and colorimetric measurements were carried out every 2 h over the following 12-h diurnal period and were continued on day three. The colorimetric parameters L* (luminance) and a* (color hue ranging from green (–) to red (+)) gave a rank order correlated to corticosteroid potency that showed superior discrimination compared to simple visual grading. In this study L* was a more discriminative parameter than a*.
74.6.4 MEASUREMENT
OF
SKIN COLOR1
Reflectance colorimetric methods have been widely applied in studies of the chromatic characteristics of skin in the context of anthropologic studies of human population genetics.41–43 CIELAB color space parameters also have potential cosmetic applications, such as the design of prosthetic devices and the use in color matching of skin grafts in plastic surgery (e.g., after skin tumor excision). For complete color measurement it has been shown that filter colorimetry can give results that differ systematically from those based on scanning spectrophotometry.2,44 Evaluation of the ventral upper arm skin color in volunteers of various ethnic origin were measured by applying reflectance spectroscopy with CIELAB color space parameters. The volunteers classified themselves as being of European, Chinese, Indian, or Polynesian ethnicity, or of mixed race. Measurements of the visible reflectance spectrum of skin and the calculation of CIELAB values provided a practical numerical basis for quantifying the perceived color of human skin. The excess of data in a full spectrum can be reduced to a set of color space parameters that relate directly to the appearance of the color as it would be clinically observed and without
making any assumptions about the nature of the pigments involved. However the same spectral data could be used if there was a requirement beyond the simple numerical specification of appearance as in the determination of the concentration of chromophores in physiologic studies.
74.6.5 MEASUREMENT OF ULTRAVIOLET-INDUCED PIGMENTATION Tanning is a cardinal response of human skin to UV radiation. A spectrophotometric technique for measuring changes in the rate of transition between the first erythematous response and delayed pigment formation has important applications for evaluating the efficacy of sunscreens and potential pigment enhancers (e.g., psoralens with UVA and tyrosine).44,46,47 Unfortunately, the clinical usefulness of early spectrophotometers for measuring this color transition of skin has been limited by technical problems including the stability of the instrument light source, the need for external references, and the use of multiple filters to obtain the exact skin color.48,49 Furthermore, the wide spectrum of human skin colors has made comparative treatment studies with these devices virtually impossible since there is no means of recording baseline data on an individual subject for later paired comparison.50 With a handheld tristimulus colorimeter these problems have been overcome. Natural skin tones can be stored in the colorimeter memory from a subject for direct comparison to the solar exposed skin color. The capability of the tristimulus colorimeter to simultaneously evaluate the hue and saturation of skin color affords an improved opportunity to quantitate the transition from cutaneous erythema to tanning. Seitz et al.51 tested areas on the back, which were UV irradiated at the exposed sites previously estimated to produce 1, 1.5, and 2 MED for each individual. 24 h later erythema and tanning were scored by a dermatologist. The subjects received repeated UV exposures once every 48 h for 14 d. The Minolta chromameter showed subtle, continuous transition between the primary erythamous response and delayed tanning of the skin. The initial changes were below the visual threshold for detection as seen in the dermatologist’s scores and in the instrument validation responses. Linear regressions comparing the dermatologist’s scores with the meter values indicated that erythema to UVA and UVB exposure was perceived as a pure red shift in skin color, darkening in intensity with continued exposure until tanning started to appear. In contrast, tanning was perceived as an intensifying yellow hue superimposed on the existing erythema. However, since the human eye integrates all visual stimuli, any admixtures of color or other variables, e.g., skin blemishes, background skin tone, quality of lighting, etc., would partially explain the variability in the dermatologist’s erythema
644
Handbook of Non-Invasive Methods and the Skin, Second Edition
scores. These results also emphasize the difficulty in visually distinguishing between erythema and tanning.
74.6.6 MEASUREMENT OF DOSE-RESPONSE CURVES OF ULTRAVIOLET-INDUCED ERYTHEMA AND PIGMENTATION52 The assessment of sensitivity of human skin to UV radiation is important to photochemotherapy, photodermatoses, photoaging, photocarcinogenesis, and photoprotection. Today time, the most frequently used means of attempting to predict UV sensitivity, without phototesting, is classification of the persons’s skin type. This Working Classification of Sun Reactive Skin Types, as introduced by Fitzpatrick, is based on the history of an individual’s tendency to sunburn and to tan, along with some racial parameters. When more objective determination of sensitivity of the skin to UV radiation is desired, is determined experimentally. A number of investigators have sought to compare and correlate the various means of predicting and measuring UV sensitivity. Olson et al.35 reported that the MED correlated well with melanosome size, quantity, density, and distribution in various skin colors. Shono et al.36 also found a good correlation between MED and skin color, but a less clear one between the minimal melanogenic dose (MMD) and either skin color or MED. Sayre et al.37 studied the erythema action spectrum of 30 Caucasians and concluded that several factors, among which was constitutional skin color, had no significant effect on MED. Haake et al.,38 studying a population consisting of Fitzpatrick’s types I to IV, found no correlation between the UV sensitivity and color of the skin. It is inconclusive from these studies whether or not the MED correlates with skin types or skin color. Most researchers dealing with photo testing recognize that, within a so-called skin type, there is a larger interindividual variation in UV sensitivity, as assessed by MED measurements. Since a value such as the MED is merely a point on a dose-response curve, it can be expected that the full curve will yield much more information than does any arbitrarily chosen single point such as the MED. In studies of the responses of human skin to UV radiation, it is convenient to have a means of estimating the reactivity of an individual’s skin without performing photo testing. The most frequently used means at present is classification of skin into one of the six skin reactive types, based primarily on the history of previous responses of the skin to sunlight. This has proved to be a convenient method because it requires no measurements and can be done rapidly. When measurement of skin responses to UV radiation is performed, the usual method is the measurement of a MED. The MED is only an estimate of the amount of UV radiation required to produce detectable erythema and does not reveal the incremental increases in
erythema with increasing UV doses. Dose-response data for erythema would more accurately measure UV responses of human skin,23 but obtaining such data has been difficult. The availability of a sophisticated chromameter interfaced with a computer has now made possible the easy, objective measurement of erythema and pigmentation responses to UV radiation and the obtaining of doseresponse data. Westerhof et al.,52 performed comprehensive photo testing, including the determination of MED and MMD values and the measurement of dose responses for erythema and pigmentation, on a large number of volunteers and attempted to assess the value of the skin type and objectively measure skin color in predicting the skin sensitivity to UV radiation. Westerhof et al.,52 confirmed the results of Stern and Momtaz53 and those of Azizi et al.,54 who found a greater than threefold difference between the highest and the lowest MED values in each skin type, reflecting the large variation in individual MED values. The conclusion from this study52 is that the skin type does not correlate well with the MED. The reason for this may be that the skin type does not accurately estimate UV sensitivity or that the MED is not a sensitive measurement of UV responses of human skin. The lack of a close relationship of skin type with the dose-response curves for erythema and pigmentation would point to the skin type as an inadequate predictor of UV responses. With the reflectance measurement (Y) obtained with the chromameter prior to photo testing, we were able to objectively estimate skin color. The constitutional skin color did not correlate well with the skin type and neither with measured MED or MMD values. However, we found the objectively measured skin color to be the best predictor of the dose-response measurements for both erythema and for pigmentation. In lightly pigmented skin, the doseresponse curves were steep, whereas in darkly pigmented skin the curves were much flatter. From these studies52 it appears that dose-response data best measures the sensitivity of human skin to UV radiation. Because the technology to obtain dose-response information is not widely available, MED measurements will continue to be the most convenient means of measuring the response to UV radiation. However, for sophisticated photo testing in the future, such as in measuring the protection provided by sunscreens, the use of dose-response determinations might better reflect changes in UV sensitivity. The dose-response curves for erythema, which can be easily measured 24 h after irradiation of the skin, should become the standard for modern photo testing. In summary, skin color is a more valid predictor than skin type for the measurement of UV sensitivity, which is best expressed by dose-response curves. Although skin typing will continue to be used because of its convenience, one must be aware that it has severe limitations as a predictor of UV sensitivity. Perhaps if a means of easily
Colorimetry
645
estimating constitutional skin color at the site of anticipated UV exposure becomes readily available, this will prove a more valid means of predicting UV sensitivity.
74.6.7 MEASUREMENT OF BLEACHING EFFECT DEPIGMENTING AGENTS
BY
There is a need to show efficacy in the treatment of hyperpigmentary disorders such as melasma, lentigo solaris, cafe au lait spots, postinflammatory hyperpigmentation, etc. Many products have been advocated; few of them work after months of application, some have serious side effects such as depigmentation or even more severe hyperpigmentation. Duteil and Ortonne55 aimed to assess the activity of 20% azelaic acid cream in light-induced skin pigmentation in subjects. There were five test zones, all located on the middle of the back: two were treated with azelaic acid cream, two others with the vehicle, and one was left untreated. Each product was applied twice daily, 5 d a week, for 4 weeks on one zone, and for 5 weeks on the other. In the middle of the fourth week, the test zones were exposed to UVB + UVA + visible light, with a total of three times the minimal erythema dose distributed progressively over 3 consecutive days. For 7 and 10 d after the last irradiation, the induced photopigmentation was assessed by colorimetric and visual means. Compared with its vehicle, the azelaic acid cream had neither a depigmenting effect nor a preventive effect on the lightacquired skin pigmentation. Moreover, interrupting or continuing azelaic acid treatment after skin irradiation had no influence on the resulting pigmentation.
74.7 CONCLUSIONS As already stated, the complexity of the erythematogenic response in the optically complex multilayered system of the skin is difficult to approximate by a formula. Erythema indices based on such equations are bound to be oversimplifications, so that the computed mathematical interpretations might lead to inappropriate correlation with the degree of erythema. The advantage of measuring the erythema of the skin by using the CIE system of tristimulus values is that one does not need to make any assumptions about the processes underlying erythema formation. In clinical situations, e.g., when only the degree of sensitivity of the skin to UV radiation is required, one can simply estimate the degree of erythema while comparing it to normal neighboring skin. Therefore, especially in the early stages of erythema formation, it is not necessary to account for the different translucent layers of the skin, in particular the pigments in each of these layers and the degree of light scattering due to differences in turbidity. These other qualities are leveled out when comparing erythematous skin with normal skin. For scientific
investigations, however, colorimetric measurements are indispensible. Colorimeters using the CIE system allow for quantitative comparison of erythema and pigmentation in individuals and between individuals in a way that is consistent with visual judgements, but with greater reliability and reproducibility than would be possible by human observers. In this way, reliable dose response curves can be constructed. Colorimeters able to quantify reflected colors from the skin using the CIE system appear to be precise, quick, and handy in their operation.
ACKNOWLEDGMENTS This work was supported by the Dutch Pigment Cell Foundation. The constructive advise of Professor Oscar Estevez Uscanga and Dr. Henk E. Menke greatly increased the readability of this chapter.
REFERENCES 1. Weatherall IL, Coombs BD. Skin color measurements in terms of CIELAB color space values. J Invest Dermatol 99:468–473, 1992. 2. Billmeyer FW, Saltzman M. Principles of Color Technology, 2nd ed., Wiley-Interscience, New York, 1981. 3. Feather JW, Ryatt KS, Dawson JB, Cotteril JA, Barker DJ, Ellis DJ. Reflectance spectrophotometric quantification of skin color changes induced by topical corticosteroid preparations. Br J Dermatol 106:437–444, 1982. 4. Babulak SW, Rhein LD, Scala DD, Simion FA, Grove GL. Quantitation of erythema in a soap chamber test using the Minolta Chroma (Reflectance) Meter: comparison of instrumental results with visual assessments. J Soc Cosmet Chem 37:475–479, 1986. 5. Westerhof W, van Hasselt BAAM, Kammeyer A. Quantification of UV-induced erythema with a portable computer controlled chromameter, Photodermatology 3:310–314, 1986. 6. Hacham H, Freeman SE, Gange RW, Maytum DJ, Sutherland JC, Sutherland BM. Do pyridine dimer yields correlate with erythema inducation in human skin irradiated in situ with ultraviolet light (275–365m)? Photobiology 53(4):559–563, 1991. 7. Hunter RS, Harold RW. The Measurement of Appearance, 2nd ed., Wiley-Interscience, New York, 1987. 8. Marchesini R, Brambilla M, Clemente C, Maniezzo M, Sichirollo AE, Testori A, Venturoli DR, Cascinelli N. In vivo spectrophotometric evaluation of non-neoplastic skin pigmented lesions. I. Reflectance measurements. Photochem Photobiol 53(1):77–84, 1991. 9. Feather JW, Hajizadeh-Saffiar M, Leslie G, Dawson JB. A portable scanning spectrophotometer using visible wavelengths for the rapid measurements of skin pigments. Phys Med Biol 34:807–820, 1989. 10. Kollias N, Bager A. Quantitative assessment of UVinduced pigmentation and erythema. Photodermatology 5:53–60, 1988.
646
Handbook of Non-Invasive Methods and the Skin, Second Edition
11. Serup J, Agner T. Colorimetric quantification of erythema-a comparison of two colorimeters (Lange Micro Color and Minolta Chroma Meter CR-200) with a clinical scoring scheme and Laser-Doppler flowmetry. Clin Exp Dermatol 15:267–272, 1990. 12. Kollias N, Bager A. Spectroscopic characteristics of human melanin in vivo, J Invest Dermatol 85:38–42, 1985. 13. Babulak SW, Rhein LD, Scala DD, Simion A, Grove GL. Quantitation of erythema in a soap chamber test using the Minolta Chroma (reflectance) Meter: comparison of instrumental results with visual assessments. J Soc Cosmet Chem 37:475–479, 1986. 14. Westerhof W, van Hasselt BAAM, Kammeyer A. Quantification of UV-induced erythema with a portable computer controlled chromameter. Photodermatology 3:310–314, 1986. 15. Nilsson GE, Wahlberg JE. Assessment of skin irritancy in man by laser Doppler flowmetry. Contact Dermatitis 8:401–406, 1982. 16. Staberg B, Serup J. Allergic and irritant reactions evaluated by laser Doppler flowmetry. Contact Dermatitis 18:40–45, 1988. 17. Engelhart M, Kristensen JK. Evaluation of cutaneous blood flow responses by 133 Xenon washout and a laser Doppler flowmeter. J Invest Dermatol 80:12–15, 1983. 18. Malm M, Jurell G, Tonnquist G. Natural color system — a new method for evaluating skin colors during Argon laser treatment of port-wine stain. Ann Plast Surg 20:317–321, 1988. 19. Serup J. Quantification of wheal reactions with laser Doppler flowmetry. Allergy 40:223–237, 1985. 20. Staberg B, Klemp P, Serup J. Patch test responses evaluated by cutaneous blood flow measurements. Arch Dermatol 120:741–743, 1984. 21. Serup J, Staberg B, Klemp P. Quantification of cutaneous oedema in patch test reactions by measurement of skin thickness with high-frequency pulsed ultrasound. Contact Dermatitis 10:88–93, 1984. 22. Serup J, Staberg B. Ultrasound dor assessment of allergic and irritant patch test reactions. Contact Dermatitis 17:80–84, 1987. 23. Farr PM, Diffey BL. Quantitative studies on cutaneous erythema induced by ultra violet radiation. Br J Dermatol 111:673–682, 1984. 24. Olson RL, Sayre RM, Ecerett MA. Effect of anatomic location and time on ultraviolet erythema. Arch Derm 93:211–215, 1966. 25. Berger DS. The sunburn ultravioletmeter: design and performance. Photochem Photobiol 24:587–593, 1976. 26. Westerhof W, Buehre Y, Pavel S, Bos JD, Das PK, Krieg S, Siddiqui AH. Increases anthralin irritation response in vitiliginous skin. Arch Dermatol Res 281:52–56, 1989. 27. Dawson JB, Barker DJ, Ellis DJ, et al. A theoretical and experimental study of light absorption and scattering by in vivo skin. Phys Med Biol 25:695–209, 1980.
28. Queillo-Roussel C, Poncet M, Schaefer H. Quantification of skin color changes induced by topical corticosteroid preparations using Minolta Chroma Meter. Br J Dermatol 124:264–270, 1991. 29. Wan S, Parrish JA, Jeaniche KF. Quantitative evaluation of ultraviolet induces erythema. Photochem Photobiol 37:643–648, 1983. 30. Cornell RC, Stoughton RB. Correlation of the vasoconstriction assay and clinical activity in psoriasis. Arch Dermatol 121:63–7, 1985. 31. Robertson AR. The CIE 1976 color difference formulas. Color Research and Application 1977;2:7–11. 32. Polano MK, August PL. In: Polano MK (ed.). Topical Skin Therapeutics. Churchill Livingstone, Edinburgh, 1984, p. 101. 34. Fitzpatrick TB, Pathak MA, Parrish JA. Protection of human skin against the effects of the sunburn ultraviolet (290–320nm). In: Fitzpatrick TB et al. (eds). Sunlight and Man-Normal and Abnormal Photobiological Responses. University of Tokyo Press, Tokyo, 1974, p. 751. 35. Olson RL, Gaylor J, Everett MA. Skin Color, malonin and erythema. Arch Dermatol 108:541–544, 1973. 36. Shono S, Imura M, Ota M, Ono S, Toda K. Relationship of skin color, UVB-induced erythema and melanogenesis. J Invest Dermatol 84:265–267, 1985. 37. Sayre RM, Desrocher DL, Wilson CJ, Marlowe EL. Skin type, minimal erythema dose (MED), and sunlight acclimatization. J Am Acad Dermatol 5:429–443, 1981. 38. Haake N, Buhles N, Altmeyer P. Sensitivity of human skin to UV-light-practicability and limits of clinical diagnostics. Z Hautkr 62:1505–1509, 1987. 39. Berger DS. Design of a solar simulator. J Invest Dermatol 53:192–199, 1969. 40. Pathak MA, Fitzpatrick TB, Greiter F, Kraus EW. Preventive treatment of sunburn, dermatoheliosis, and skin cancer with sunprotective agents. In: FitzpatricK TB et al. (eds). Dermatology in General Medicine. McGrawHill, New York, 1987. 41. Little MA, Wolf ME. Skin and hair reflectance of women with red hair. Ann Human Biol 8(3):231–241, 1981. 42. Clarke P, Stark AE, Walsh RJ. A twin study of skin reflectance. Ann Human Biol 8:529–541, 1981. 43. Towne B, Hulse FS. Generational change in skin color variation among Habbani Yemini Jews. Human Biol 62(1):85–100, 1990. 44. Stevenson JM, Weatherall IL, Littlejohn RP, Seman DL. A comparison of two different instruments for measuring venison CIELAB values and color assessment by trained panel. N Zeal J Agr Res 34:207–211, 1991. 45. Jarnecke-Munster H. Uber die Zusammenhange der am Hauteigenstoffwechsel beteiligten Aminosauren, ins besondere Histiden und Tyrosin. Arch Dermatol Syph 180:290–293, 1940. 46. Cripps DJ. Natural and artificial photoprotection. J Invest Dermatol 76:154–157, 1981. 47. Shigeaki S, Imura M, Ota M. The relationship of skin color, spectrophotometric technique. J Invest Dermatol 84:265–267, 1985.
Colorimetry
48. Buckley WR, Grum F. Measurement of skin color, spectrophotometric technique. J Soc Cosmet Chem 15:79–85, 1964. 49. Farrington D, Imbrie JD. Comparison between visual grading and reflectance measurements of erythema produced by sunlight. Br J Dermatol 111:295–304, 1984. 50. Wasserman HP. The colour of human skin. Dermatologica 143:166–173, 1971. 51. Seitz JC, Whitmore CG. Measurement of erythema and tanning responses in human skin using a tri-stimulus colorimeter. Dermatologica 177:70–75, 1988. 52. Westerhof W, Estevez-Uscanga O, Meens J, Kammeyer A, Durocq M, Cairo I. The relation between constitutional skin color and photosensitivity estimated from UV-induced erythema and pigmentation dose response curves. J Invest Dermatol 94:812–816, 1990.
647
53. Stern RS, Momatz K. Skin typing for assessment of skin cancer risk and acute response to UVB and oral methoxalen photochemotherapy. Arch Dermatol 120: 869–872, 1984. 54. Azizi E, Lusky A, Kushelevsky AP, Shewach-Millet M. Skin type, hair color, and freckles are predictors of decreased minimal erythema ultraviolet radiations dose. J Am Acad Dermatol 19:32–38, 1988. 55. Duteil L, Ortonne JP. Colorimetric assessments of the effects of azelaic on light-induced skin pigmentation. Photodermatol Photoimmunol Photomed 9:67–71, 1992.
Color Measurement 75 Quasi-L*a*b* from Digital Images Hirotsugu Takiwaki Department of Dermatology, The University of Tokushima School of Medicine, Tokushima, Japan
CONTENTS 75.1 Introduction............................................................................................................................................................649 75.2 Objective and Methodological Principle...............................................................................................................649 75.3 Sources of Error.....................................................................................................................................................650 75.4 Correlation with Other Methods ...........................................................................................................................650 75.5 Recommendation ...................................................................................................................................................651 References .......................................................................................................................................................................651
75.1 INTRODUCTION In this era of digital images, the evolutional and rapid progress of charge coupled devices (CCD) and computers has enabled us to easily obtain and retouch high-resolution digital images. As every picture element (pixel) in digital images has color information, i.e. the brightness intensity in the red, green, and blue (RGB) channels, it is only natural that attempts have been made to quantify skin color by using digital images of the skin. In addition, unlike with reflectance instruments such as colorimeters or reflectance spectrometers, not only information about skin color itself, but also about color distribution can be obtained with digital images. However, color information obtained from digital images is inevitably influenced by the properties of CCD, the characteristics of light sources, and various other conditions under which images are taken. Therefore, if exact or “absolute” color data from digital images are needed, preliminary processing is necessary in order to convert raw data into a more reliable and standardized color scale. In this chapter, some simple methods to obtain values that approximate CIE-L*a*b* (quasi-L*a*b*) are described.
75.2 OBJECTIVE AND METHODOLOGICAL PRINCIPLE Objective: To quantify normal or lesional skin color in the L*a*b* color space by processing RGB brightness data of digital images of the skin. As described in other chapters, L* is the coordinate for the intensity of brightness, and a* and b* are the chromaticity coordinates.
Instruments: Any instruments for obtaining digital images, such as digital cameras and videomicroscopes equipped with a CCD camera, can be used. Computers and computer software for color image analysis, such as Image J (NIH freeware created by Wayne Rasband)1 and Adobe Photoshop, are also needed. Method: As the color of an object always depends on the characteristics of the illumination, digital images of the skin should always be taken under the same conditions. When images are taken with a digital camera, manual rather than automatic operation under specified conditions is recommended. If skin color is to be compared in a series of images, the distance between the camera and objects should be kept strictly constant. It is also advisable to place color chips with known color values close to the objective skin area and obtain images of the two together for color reference. The most commonly used method for estimating quasi-L*a*b* values of pixel(s) in digital images is to convert the brightness values in RGB channels to the corresponding L*, a*, and b* values. This conversion must be preceded by the preliminary formulation of regression curve equations based on an examination of the relationship between RGB and CIE-L*a*b* values of the same standard objects, such as color chips or color charts.2 In order to obtain these regression curves, it is also necessary to examine, with the aid of statistical software for multiregression analysis, the correlation between RGB brightness values of the object image and its CIE-L*a*b* values obtained with reflectance instruments. The histogram function of image analysis software can be used when the distribution and average of brightness data on a region of 649
650
Handbook of Non-Invasive Methods and the Skin, Second Edition
interest (ROI) are required for each channel (stack) of RGB. An alternative and simpler method may be to substitute RGB brightness values for the terms of XYZ in the formula of CIE-L*a*b* derivation3 as follows: Quasi-L* = 116Rg1/3 – 16 Quasi-a* = 500(Rr1/3 – Rg1/3) Quasi-b* = 200(Rg1/3 – Rb1/3) Rr,g,b = (BSr,g,b – BBr,g,b) / (BWr,g,b – BBr,g,b) where BS is the mean brightness value of the skin image, BW is that of the same size area in the image of a white standard, and BB is the background brightness. Results obtained for human and dog skin in this way show good linear correlation with CIE-L*a*b* values.3,4 Finally, an experimental technique has been developed which employs a multiband camera and special algorithm using Wiener estimation for estimating the spectral reflectance of each individual image pixel.5 This techniques is now undergoing improvements. Although not commercialized yet, it is hoped that this revolutionary method may provide not only more accurate color data but also information on optical properties of the object simply by taking its picture with a special digital camera.
the differences in the angle from the illumination source. To determine the presence of these artificial or unavoidable factors, preliminary evaluation of color images with the aid of a phantom model, such as a mannequin,2 is recommended. Images that are to be analyzed should in principle be saved in memory media or computers as TIFF files because JPEG compression may alter the fine configuration of skin surface structures or skin eruptions and therefore may influence the color values.
75.4 CORRELATION WITH OTHER METHODS When skin color is evaluated, quasi-L*a*b* values, obtained by using both multi-regression analytic method (m-L*a*b*) and “pseudo” CIE-L*a*b* formula method (p-L*a*b*), reportedly show good to excellent linear correlation with CIE-L*a*b* values measured with a MINOLTA Chromameter CR-300TM.2,3 The respective correlation coefficients between the m-L*a*b* and CIEL*a*b* values (n=187) were 0.897 (L*), 0.815 (a*) and 0.817 (b*), and those between the p-L*a*b* and CIEL*a*b* values (n=89) 0.973(L*), 0.913 (a*) and 0.837 (b*) (Figure 75.1). However, these results may depend on the instrument actually used, as the properties of the CCD
75.3 SOURCES OF ERROR Even if digital images are taken with the same camera or video equipment, the following conditions should always be kept constant in order to obtain accurate values and comparable data: 1. No direct sunlight from windows 2. No or dim external illumination 3. Same brightness intensity of flash bulb or builtin illumination 4. Same distance between camera and object 5. Same camera settings. Reference color chips, such as white standard or skin color standard, should be photographed at the same time to calibrate the illumination or check camera settings. Insufficient light diffusion due to uneven illumination may result in shiny reflections and shadows on the object surface, which in turn may cause artificial errors. If videomicroscopes are used, ROI should be located at the center of the image because darkening usually occurs in the peripheral areas of the image. In the face or extremities, anatomical curvature often causes unevenness of brightness because of variations in distance from the camera or
FIGURE 75.1 Relationship between corresponding coordinates of CIE- and quasi-L*a*b* systems. (From Takiwaki, H. et al., Skin Res. Technol. 3, 42-44, 1994. With permission.)
Quasi-L*a*b* Color Measurement from Digital Images
used cannot be expected to be identical. Therefore, it should be confirmed for each image-analytic quasiL*a*b* system that it shows good linear correlation with CIE-L*a*b* system. It should also be kept in mind that CIE-L*a*b* values of the skin depend on the instrument used.6 This phenomenon seems to result in part from the relation between the area of the measuring-head aperture and light scattering characteristics (Rayleigh scattering) of the dermis.7 Therefore, quasi-L*a*b* values thus obtained can also be expected to be influenced by the reflectance instruments that are used as reference.
75.5 RECOMMENDATION Quasi-L*a*b* evaluation from skin images appears to be the most accurate method for measuring the standardized color of very small skin areas or lesions, such as moles, ephelides, cherry angiomas and tiny pustules, that are far smaller than the area of the measuring head aperture of reflectance instruments. This method is also suitable for color evaluation of the whole area of the irregularly demarcated regions, as these areas can be selected with software for photo retouching,4 as well as for eroded or ulcerated skin lesions that are difficult to measure with a contact-type instrument.
651
REFERENCES 1. Image J – Image Processing and Analysis in Java, http://rsb.info.nih.gov/ij/ (Acc 2004-01-12) 2. Miyamoto, K., Takiwaki, H., Hillebrand, G. G., and Arase, S., Development of a digital imaging system for objective measurement of hyperpigmented spots on the face, Skin Res. Technol., 8, 227–235, 2002. 3. Takiwaki, H., Miyamoto, H., and Ahsan, K., A simple method to estimate CIE-L*a*b* values of the skin from its videomicroscopic image, Skin Res. Technol., 3, 42–44, 1997. 4. Boysen, L., Serup, J., Sorensen, P., and Kristensen, F., Use of chromametry and didital photography for objective measurement of skin color in clinically normal dogs, Am. J. Vet. Res., 63, 559–564, 2002. 5. Miyake, Y., and Yokoyama, Y., Obtaining and reproduction of accurate color images based on human perception, Proc. SPIE., 3300, 190–197, 1998. 6. Serup, J. and Agner, T., Colorimetric quantification of erythema –a comparison of two colorimeters (Lange Micro Color and Minolta Chromameter CR-200) with a clinical scoring scheme and laser-Doppler flowmetry, Clin. Exp. Dermatol., 15, 267–272, 1990. 7. Takiwaki, H., Miyaoka, Y., Skrebova, N., Kohno, H., and Arase, S., Skin reflectance-spectra and colour-value dependency on measuring-head aperture area in ordinary reflectance spectrophotometry and tristimulus colorimetry, Skin Res. Technol., 8, 94–97, 2002.
Color Calibration for 76 Practical Dermatoscopic Images Constantino Grana, Giovanni Pellacani, and Stefania Seidenari University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 76.1 The Importance of Color Calibration in Dermoscopy..........................................................................................653 76.2 Instruments Characteristics and Calibration .........................................................................................................654 76.2.1 Analysis of the Video Camera’s Physical Properties................................................................................654 76.2.2 Illumination and Border Defects Correction.............................................................................................655 76.2.3 Assessment of γ .........................................................................................................................................656 76.2.4 Conversion from the Instrument’s RGB to XYZ......................................................................................657 76.2.5 Conversion from XYZ to a Known and Standard Color Space...............................................................658 76.3 An Example of Multi-Instrument Calibration.......................................................................................................659 76.4 Conclusions............................................................................................................................................................661 References .......................................................................................................................................................................663
76.1 THE IMPORTANCE OF COLOR CALIBRATION IN DERMOSCOPY Surface microscopy, which employs incident light magnification systems associated to the epiluminescence technique or polarized light, improves diagnostic accuracy with respect to simple clinical observation, especially for difficult-to-diagnose lesions [1,2]. The assessment of colors is essential for melanoma (MM) diagnosis, both for pattern analysis on dermoscopic images [3], and when employing semiquantitative methods [4–6]. Malignant lesions frequently show more than 3 colors, whereas in nevi, 3 or fewer than 3 colors are usually observed [7]. Moreover, different colors prevail in different lesion types. In order to overcome subjectivity and variability in the interpretation of dermoscopic images, several image analysis programs, also enabling color description by means of numerical data calculated from different color channels, have been recently introduced as a possible support to clinical diagnosis [8–15]. Recently, we described an image analysis method for the evaluation of colors in melanocytic lesion images based on an approach which shares some similarities with the human perception of colors [16,17]. Since color analysis has proved to be an important factor in the diagnostic process of dermatoscopic images, color degradation could negatively influence the diagnostic ability of the clinician [18]. Not much investigation
has been conducted on the interchange of dermatoscopic images and on the differences in color reproduction between different instruments or units of the same instruments. Even if this evaluation may seem insignificant, since images appear more or less similar in successive comparisons, the problem becomes apparent in automatic analysis, that is computer based color measure and characterization. Thus, the algorithms employed by different image analysis programs are strictly applicable only to pigmented skin lesions acquired with the same instrument and technique, and are not adaptable to images generated by different tools, sometimes also employing different acquisition methods [19]. For example if an instrument allows manual light intensity tuning, so that images can be adjusted by sight to appear bright enough, any comparison on the dark to light variation (for instance in the search for dark areas) is influenced by this setting [20]. If the light intensity tuning is continuously modified for every image, some difficulties could arise, leading to the question whether it is possible to use those images in an automatic framework or not. Most of the work in computerized dermatoscopy deals with the use of colors, but few studies even attempt to acquire knowledge on the color space under examination, usually only referring to an unspecified RGB color space [21,22]. Unfortunately, in the digital dermoscopy area, color standardization is very seldom taken into account, even if working in conjunction with image analysis 653
654
Handbook of Non-Invasive Methods and the Skin, Second Edition
programs. Most references are made to a generic RGB space that is implicitly assumed to be the RGB color space provided by their instrument. For instance, Vannoorenberghe et al. [21] describe a system that relies on learned probabilities to detect lesion contours directly in the RGB color space, Faziloglou et al. [22] use color histograms to detect differences between melanoma and nevus colors: to get rid of color influences they subtract the average skin color from each lesion pixel, without however considering gamma or contrast differences. In Gerger’s study [23] a completely different approach was followed, but unspecified RGB statistics were still present, disallowing any possible systematic reproduction of the published results. It is natural to assume that color calibration technologies and methods are still being studied or are as yet unknown in the dermatologic field, but this is not the case, since for example Koriki et al. [24] illustrated an on-skin lipstick color measurement system with full calibration, and in a work by Miyamoto et al. [25] a quick but effective conversion was estimated from RGB to L*a*b*, by means of a third order multiple regression analysis. The authors were also careful in referring to the obtained values as quasi-L*a*b* to stress the fact that this was an approximation of the full calibration approach. Setaro and Sparavigna [26] obtained a simple calibration by using a standard reference based on three colors and then adjusting the images comparing the measured values of the marker to the image colors by simple difference. Even this first order model is reported to produce good reproducibility. A systematic work for color measurement from video camera in dermatology was presented by Herbin et al. [27], but one of the most detailed and precise methods was provided by Haeghen et al. [28]. In this study a full calibration approach was followed and the problem of obtaining standardized images was successfully resolved. However, the final choice of sRGB color space for image exchange and handling led to the production of low contrast images and loss of details, owing to the limited area of the above mentioned color space, when applied to pigmented skin lesions acquired by digital dermoscopy. It is time for literature dealing with color images to cope with the problem of calibration, in order to generate reproducible data. In this chapter we provide a detailed and practical description of a calibration framework: since many instruments show different illumination structures, first of all illumination correction is explored; secondly an easy camera gamma estimation technique is described; thirdly, the use of specific color spaces to avoid low contrast effects due to the 8 bit quantization of color channels is described. Mention is also made to camera temperature effects over the color image reading.
76.2 INSTRUMENTS CHARACTERISTICS AND CALIBRATION The calibration techniques here described are applicable to every video camera system, and take into account problems that are very likely found in every dermatoscopic setup. For the sake of precision we will refer in detail to the specific case of an epiluminescence microscope for dermatology (FotoFinder, TeachScreen software GmbH, Bad Birnbach, Germany), that consists of a probe, comprising a CCD-chip color video camera with an integrated handle and optics for epiluminescence microscopy, a processing unit and a color monitor. Optics are set in a removable conic structure with a cylindrical transparent spacer and contact plate at the end, and with 6 bright white LEDs, positioned at the bottom of the structure, for constant illumination of the viewing area. According to the size of the lesion to be examined, 20- to 70-fold magnifications can be employed. When the probe is applied onto the skin area for imaging, a magnified color picture appears on the monitor. Focus is automatically adjusted by an electronically powered autofocus system. The S-video signal is digitized at the PAL standard resolution and digital images can be stored and processed at any time. The digitized images offer a spatial resolution of 768 × 576 pixels and 16 million colors. For the epiluminescence observation, a drop of contact medium, such as alcohol in water solution, is applied between the contact plane and the skin, enabling the recognition of subsurface structures, within and outside the lesion. The instrument is easy to handle and acquisition of the images is quick and simple. Furthermore it is one of the most widely diffused digital videomicroscopes, with over 750 pieces sold in Europe by the end of 2003.
76.2.1 ANALYSIS OF THE VIDEO CAMERA’S PHYSICAL PROPERTIES An often underrated problem is the stability of the readings taken from the digital camera. It is often assumed that, though having a signal noise component, camera readings tend to be constant over time. Unfortunately, this is not the case with many commercial camera based systems, as the FotoFinder equipment, and we assume that the same problem is present in many other devices. A first glance at the instrument’s behavior during time can be seen in Figure 76.1, in which a neutral gray patch of GretagMacbeth™ ColorChecker, precisely number 22 (xyY = 0.310,0.316,19.8), was acquired every 5 minutes from the start up of the video camera, without any change in light or room temperature. Average values taken on a selected window of the image are represented. It is clear that the camera has a so called “warm up” period during which it is practically impossible to get a stable reading of the object under examination; measurements are stable
Practical Color Calibration for Dermatoscopic Images
655
180 170 160 150
MeanR MeanG MeanB
140 130 120 110 100 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.1 (See color insert following page 678.) Mean RGB values of patch 22 in successive acquisitions (5 min distance).
180 170 160 150
MeanR MeanG MeanB
140 130 120 110 100 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.2 (See color insert.) Mean RGB values of patch 22 in successive acquisitions (1 min distance).
180 160 140 120 MeanR MeanG MeanB
100 80 60 40 20 0 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.3 (See color insert.) Mean RGB values of patch 22 in successive acquisitions (1 min distance with pauses every six acquisitions).
after 60 minutes. However, the frequency of acquisitions also influences the mean R, G and B values. In fact acquiring the same target every 30 seconds leads to a difference of about 10 levels per channel (Figure 76.2). This may be explained by a heating of the video camera during the acquisition due to the motorized motion of the zoom (that is reset at every acquisition) and of the auto focus. To obtain an acceptably stable reading over time, we had to insert pauses between series of acquisitions, leaving the instrument time to return to its normal conditions. Figure 76.3 illustrates a series of acquisitions of patch 22, taken at 1 minute time spans. After every 6 captures, a 10 minute pause was taken. This leads to a variation from the first to the last acquisition of less than 3 gray levels on
average, which is more than satisfactory for our aim. This procedure reproduces the routine visit of pigmented skin lesion clinic patients who usually require more than one acquisition per visit for follow up or documentation purposes.
76.2.2 ILLUMINATION CORRECTION
AND
BORDER DEFECTS
Color calibration process begins with a first step to correct the irregular illumination of the instruments. The basic assumption here is that if we are imaging a uniform reference surface, we should obtain an almost constant reading for the whole image. This is not always so because of
656
Handbook of Non-Invasive Methods and the Skin, Second Edition
the instrument’s characteristics, but we can measure the deviation from uniformity to invert it, thus obtaining a correction map for all the acquired images. A masking process must also occur to get rid of pixels that do not convey any data, such as top or bottom lines that are always black because of the frame grabber settings, or the black ring that some instruments present at lower magnification levels (20-fold for FotoFinder). The filter values are computed separately for each color channel, so if we call R(x, y) the value of the red channel, the value of the filter in that point is:
( )
FR x, y =
( )
R x, y − M R MR
where MR is the most represented value in a selected window of the image. This value can be computed from the histogram of the channel, that is a function of the channel value equal to the number of pixels with that value. MR can be obtained as the bin index with maximum value:
Gray Scale as a reference for our calibration. After computing the MRi for each step of the gray scale, and discarding those squares whose values clip at 255, we evaluate FR as: 19
( )
FR x, y = min
j ∈[1,19 ]
()
{( x, y ) : R ( x, y ) = i} ()
M R = arg max hR i i∈[ 0 ,255 ]
The window selection should be taken in a well lit area to uniform the image to the area that is best illuminated. This enables also MR to be used as an indicator of channel saturation. In fact one should be careful to avoid getting clipped values during the calibration phase, since these could disrupt the final result. In our application, since we were using the FotoFinder equipment, which has a semicircular set of photo emitting LEDs, we chose the rectangular window from (200,350) to (567,550). Moreover, this filter assessment should be carried out so that it is independent of the specific surface selected and the resulting acquisition. Thus, we used the Kodak
i R
j R
i =1
which is the median of the measured values FRi . The same process is applied to the green and blue channels, obtaining FG and FB. Each image I is then filtered for each channel using the equation:
˜I ( x, y ) = R
I R ( x, y ) 1 + FR ( x, y )
obtaining the filtered image I.˜
76.2.3 ASSESSMENT hR i = #
∑ F ( x, y ) − F ( x, y )
OF
γ
After obtaining light compensated images, the next step is to estimate the non-linear relation between the luminance factor, also known as CIE tristimulus value Y, and the digital values provided by the camera. Y is linearly related to incident light and is a standard of light measurement. Moreover, given fixed light source and geometry, it should be linearly related to the reflectance R of the surface, that can be estimated from the optical density OD as R = 10 2 −OD . Power 2 is used to scale R between 0% and 100% as it is commonly expressed. Measuring the Kodak Gray Scale with a Minolta CL-100 calibrated colorimeter, we verified that the declared optical density steps of 0.1 were indeed linearly related to Y (Figure 76.4).
100 90 80 70 60 50 40
Colorimeter Y
30 20 10 0 0
10
20
30
40
50
60
70
80
90
100
Declared reflectance
FIGURE 76.4 (See color insert.) Comparison of declared and measured reflectance of the KODAK Gray Scale.
Practical Color Calibration for Dermatoscopic Images
657
1 0,9 0,8 0,7 0,6
dr dg db
0,5 0,4 0,3 0,2 0,1 0 0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0,45
0,5
Colorimeter measured ratio Y/Yn
FIGURE 76.5 (See color insert.) Normalized RGB values of the video camera for gamma estimation.
Measuring the Kodak Gray Scale with the FotoFinder we verified the presence of gamma correction of RGB values, that is a power relation between the digital measures and the known Y values (Figure 76.5). To give an estimate of this relation we compared the normalized values of Yˆ = Y/Yn and dr = R/255, dg = R/255, db = R/255, with the following equation: y = ax γ + b and estimated the three parameters (a, b, γ) separately for the three channels. The aim was to use the N measured values di (with reference to the currently considered color channel) and the corresponding Yˆi to obtain the triplet: N
( a, b, γ ) = arg min ∑ Yˆ − ( ad a ,b , γ
i
γ i
)
+b .
i =1
The assessment was conducted by exhaustive search in a ±0.5 space around an initial starting point (1.0, 0.5, 0.5) with 0.1 steps. After finding the triplet producing the minimum value, the search was repeated at finer steps around that point, narrowing the search space progressively, up to the desired precision. If the solution was at the edge of the defined space, the search was repeated without narrowing the area, just moving the center to the best point found. Table 76.1 shows the values obtained for one of our instruments. It is interesting to note that the instrument
TABLE 76.1 Example of Estimated Gamma Parameters
R G B
a
b
1.5653 1.6465 1.8273
-0.4500 -0.4600 -0.4930
0.3275 0.3375 0.3434
presents a sort of constant offset on the Y value, probably correlated with a black level setting of the camera. The gamma value is significantly lower than typical values for video cameras, and the main difference is a different gain for each color channel, mainly enhancing the blue one. We wish to point out that these measures depend highly on the camera hardware settings, and are provided only as an example, not as “standard” figures. Following this estimation, inversion of the power relation to obtain triplets of RGB values that can be linearly transformed by matrix multiplication into XYZ triplets is simple.
76.2.4 CONVERSION FROM THE INSTRUMENT’S RGB TO XYZ Which matrix should be used to convert RGB to XYZ? Many matrices can be found in text books or on web pages, however none of these should be used. The conversion we are looking for should be assessed from the behavior of our specific instrument, which can only be observed by means of a reference object, such as the GretagMacbeth ColorChecker Color Rendition Chart (usually called the ColorChecker), a target that is often used in television broadcasting in order to evaluate the color accuracy of TV cameras. This target is produced using painted papers; therefore, it is not the ideal surface for skin reference, but its matte surfaces reduce problems regarding reflexes and the squares are large enough for quick positioning of the instrument. The ColorChecker should be handled carefully since it cannot be cleaned and the instrument’s head, used to compress the skin, could ruin the uniformity of the squares. This target was also measured with the Minolta CL-100 and resulting values were found to be quite consistent with tabulated data, supplied with the target. Every ColorChecker square was acquired and filtered with the previously estimated filter, then the average value of pixels from a central rectangle, whose sides were half the width and half the height of each image, was computed as the measured value for the patch. MX, the maximum histogram value (where X is the channel), was evaluated to reveal patches that couldn’t be correctly imaged by the
658
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 76.6 (See color insert.) Declared sRGB values of the ColorChecker and corresponding values as measured by the instrument.
instrument. Unlike other works, we did not attempt to calibrate the instrument so that it could correctly describe all the colors in the ColorChecker, on the contrary, we only chose to evaluate the settings that the “standard” clinician would use. Thus we discarded all patches resulting in MX = 0 or MX = 255. In our experiments one patch alone (the white one) couldn’t be used because of the extremely narrow range of the blue channel that saturated to 255 (Figure 76.6). Following a common procedure for optimization, we took the M valid triplets of RGB values, their corresponding XYZ ones and searched for the matrix A that gave the “best” transform: ⎡ Xi ⎤ ⎡ a11 ⎢ ⎥ ⎢ ⎢ Yi ⎥ ≅ ⎢ a21 ⎢ Z i ⎥ ⎢ a31 ⎣ ⎦ ⎣
⎛ R⎞ Θ 9 ⎜ G ⎟ = R G B RG GB BR R 2 G 2 B 2 ⎜ ⎟ ⎜⎝ B ⎟⎠
a13 ⎤ ⎡ Ri ⎤ ⎥⎢ ⎥ a23 ⎥ ⎢Gi ⎥ a333 ⎥⎦ ⎢⎣ Bi ⎥⎦
a12 a22 a32
(
for all the patches. This means that we solved the linear system given by: ⎡ ⎢R ⎢ i ⎢0 ⎢ ⎢0 ⎢⎣
... Gi
Bi
0
0
0
0
0
0
0
Ri
Gi
Bi
0
0
0
0
0
0
0
Ri
Gi
...
we used the Singular Value Decomposition (SVD) which presents the optimal solution, since it gives the minimum Euclidean difference between the known XYZ values and those assessed from RGB ones. We could have searched for other objectives, for instance to minimize the difference measured in the CIE L*a*b* color space, ΔE or CIE94, [29]. Haeghen et al. have shown experimentally that no evident benefit was obtained by doing so. Unfortunately the relation between RGB and XYZ values is not always well described by a linear transform, so to also cope with slightly more complex relations, we used a non-linear operator that included the covariance terms:
⎤ ⎡ a11 ⎤ ⎥ 0 ⎢ a12 ⎥ ⎥⎢ . ⎥ 0 ⎥ ⎢ .. ⎥ ⎥⎢ ⎥ Bi ⎥ ⎢ a32 ⎥ ⎥⎦ ⎢⎣ a33 ⎥⎦
⎡ ... ⎤ ⎢X ⎥ ⎢ i⎥ = ⎢ Yi ⎥ ⎢ ⎥ ⎢ Zi ⎥ ⎢⎣ ... ⎥⎦ In most cases, this can be solved exactly for three patches, leading to a perfect conversion for the three colors but a bad one for the others. For a resolution with more patches
)
T
This operator was described by Haeghen and provided the best accordance with the ColorChecker data. Thus, after applying this operator to the data, matrix A becomes 3 rows by 9 columns, but the optimization technique is the same.
76.2.5 CONVERSION FROM XYZ TO A KNOWN AND STANDARD COLOR SPACE At this point we have obtained an assessment of XYZ color coordinates for every measured pixel. The aim is to transform these values into another known color space enabling visualization and a simpler storage. Haeghen et al. [28] chose the sRGB color space because it is based on the phosphors used in many modern CRT-based display devices, including computer monitors. This means that an image stored in sRGB does not have to be converted before being displayed, and should look fairly realistic on a computer monitor. sRGB has a white point of 6500 K color temperature or D65, which means that the color produced by combining the full output of each color
Practical Color Calibration for Dermatoscopic Images
channel on an output device is the same as that of a blackbody at 6500 K. It has been standardized, so the meaningful exchange of images is also possible. sRGB tristimulus values have a known relationship to CIE XYZ tristimulus ones:
659
⎡ R ⎤ ⎡ 2.352 ⎢ ⎥ ⎢ ⎢G ⎥ = ⎢-0.731 ⎢ B ⎥ ⎢-0.098 ⎣ ⎦ ⎣
-0.802 1.358 -0.187
-0.329 ⎤ ⎡ X ⎤ ⎥⎢ ⎥ 0..392 ⎥ ⎢ Y ⎥ 1.273 ⎥⎦ ⎢⎣ Z ⎥⎦ D 65
and the gamma conversion is given by: ⎡ R ⎤ ⎡ 3.2406 ⎢ ⎥ ⎢ ⎢G ⎥ = ⎢-0.9689 ⎢ B ⎥ ⎢ 0.0557 ⎣ ⎦ ⎣
-1.5372 1..8758 -0.2040
-0.4986 ⎤ ⎡ X ⎤ ⎥⎢ ⎥ 0.0415 ⎥ ⎢ Y ⎥ 1.0570 ⎥⎦ ⎢⎣ Z ⎥⎦
X ′ = 1.46 ⋅ X (
1.0 1.85
D 65
After this linear transformation the same known gamma correction is applied to each color channel, as in the following equation: ⎧⎪ 12.92 ⋅ X X′ = ⎨ (1.0 2.4) − 0.055 ⎩⎪1.055 ⋅ X
) − 0.46
X ≤ 0.0031308 else
A problem arises in color space conversions, due to quantization to be applied to the values. Since this results in the subdivision of a “light” range into a fixed number of steps (without considering the effect of gamma correction on the steps’ characteristics). Indeed, if we are interested in a limited area of the sRGB color space, we lose much color detail that is set aside to describe color space regions that our images will never use. This color loss is clearly observed in dermatoscopic images, which become less clear (lower contrast and dynamic range) when converted into an sRGB color space. For the above reasons, we decided to describe our images by a new color space, especially conceived for our instrument for daily use. The color settings usually provided by the instrument are satisfactory, so we just searched for a formal description of the color transformation from XYZ back to our instrument’s RGB. This was performed by converting a large number of colors randomly taken from real lesions to XYZ using the calibration procedure previously described and then finding the best transformation enabling us to go back to the original colors. This is not the same as simply inverting the first transformation, since real images are used instead of the ColorChecker. The color space obtained is extracted from an average characterization produced by many images, thus it is safely applicable to dermatoscopic images, since it is designed to achieve a better use of the color representation in the spectrum area occupied by this kind of images. Its known relation with XYZ enables a simple conversion into any other chosen color space, and the practice of viewing the images on an sRGB calibrated computer monitor (a common setting available in most modern monitors), allows a common evaluation of images, even if obtained from different sources. Our proposed conversion uses the following linear conversion:
The choice for the gamma values was made by averaging those of the three channels. This conversion produces not too saturated images and allows simple comparisons between different units.
76.3 AN EXAMPLE OF MULTIINSTRUMENT CALIBRATION For testing purposes, we decided to verify the effectiveness of the calibration procedure on two different FotoFinder units without carrying out a preliminary hardware calibration, i.e., attempting to subjectively adjust images. We thus decided to set the instruments at their default configuration (factory settings), choosing an Iris setting of -25 in order to avoid skin color saturation in light skinned patients. Moreover, with this setting, all colors from the GretagMacbeth ColorChecker (GMCC) can be included, ensuring a good fitting of the data with the least squares method. To verify the effects of the resulting calibration, we used a Home Made ColorChecker (HMCC), consisting of different squares of colored paper. An example acquisition of corresponding patches is provided in Figure 76.7. The calibration procedure was performed starting from the light compensation filter, going on to the gamma functions assessment and finally to the RGB to XYZ matrix computation. From the XYZ space we employed the known conversion to get CIE L*a*b* color values and from these we measured the Euclidean distances (E). Results are summarized in Table 76.2 and show a fairly good outcome, which is strongly influenced by the errors observed on the yellow and orange patches. These errors are caused by extremely low values on the blue channel that cannot be corrected by the estimated matrix. Looking directly at the RGB color space, we can compare the Euclidean difference before and after correction (Table 76.3), and this enables us to see how different values provided by different units of the same instrument can be. For a better understanding of reported values it is useful to remember that the maximum difference can be 255 3 ≅ 442 . The comparison shows that even on a surface that is not ideal (colored paper obviously has a different color response compared to the Color Checker and the skin), differences obtained with the HMCC were
660
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 76.7 (See color insert.) Comparison of the visual aspect of the same color patches acquired with two differen equipments.
6 times lower and those obtained with the GMCC 10 times lower. It is interesting to note that results are obviously better for the surface used for calibration, so for a real comparison, another target is needed. This is an important consideration, for a serious validation of results. Finally, we had to test the results on real world images, i.e. skin lesions. The problem here was that we couldn’t just compare the average values of the images, because this wouldn’t be sufficiently accurate (images are not uniform surfaces). So we decided to provide an initial measure distinguishing the skin from the lesion, then we tried to compare the distribution of colors in the RGB color space. To this aim the 16 million colors RGB cube was divided into uniformly distributed sub-areas, allowing each channel to have values from 0 to 7, i.e., a 3 bit per channel representation, that leads to 512 possible colors. A three-dimensional color histogram was produced as before using
h ( r, g, b ) = # {( x, y ) | R ( x, y ) = r ∧ G ( x, y ) = g ∧ B ( x, y ) = b} then a common histogram comparison technique, called histogram intersection [30], was used to provide a metric to assess color similarity. Even if this technique tends to underrate similarity [31], it is perfectly suitable to assess a measure of color similarity improvement, especially if we know the compared objects are the same. The similarity measure is simply given by: 7
(
)
HI h1, h2 =
∑ min { h (r, g, b ), h (r, g, b )} 1
2
r , g ,b = 0
The value is computed over the normalized versions of the two histograms (each value of the histogram is divided
Practical Color Calibration for Dermatoscopic Images
TABLE 76.2 Euclidean Distance in the CIEL*a*b* Color Space between Corresponding Patches Measured with the Two Different Instruments Patch Number 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Mean Maximum
GMCC
HMCC
0,73 2,21 1,27 3,29 1,45 0,60 21,71 1,35 1,41 6,61 1,52 13,73 6,10 2,35 10,55 10,54 0,73 1,39 4,31 1,16 1,24 2,35 1,11 1,85 4,15 21,71
11,59 25,45 2,75 37,42 2,07 2,20 3,70 3,31 2,49 1,47 2,64 2,08 3,97 1,88 2,69 12,73 3,65 3,86 10,17 30,44 31,89 14,56 13,55 6,61 9,71 37,42
by the image size), so that HI is limited between 0 (no correspondence) and 1 (perfect histogram match). Table 76.4 reports the resulting comparisons and in Figure 76.8 and Figure 76.9 it is possible to see the result on the images that gave the best (lesion 4) and worst (lesion 1) results.
76.4 CONCLUSIONS Future developments of dermoscopy will deal with the progress of image analysis systems, tied to automatic diagnosis, and with tele-dermatology, to obtain a remote diagnostic consultation. In both cases, color calibration is of fundamental importance for two reasons: to make the developed algorithms applicable on various instruments, i.e. diagnostic center independent, and on the other hand to allow expert evaluators to have reproducible diagnoses, not biased by color degradation [32,33]. We presented a complete workflow for dermatologic image calibration, taking into account some practical, but
661
TABLE 76.3 Euclidean Distances in the Instruments RGB Color Space between Corresponding Patches Measured with the Two Different Instruments, before and after Calibration GMCC
HMCC
Uncalibrated Calibrated Uncalibrated Calibrated 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Mean Maximum
57,29 71,32 77,12 60,72 75,55 66,82 69,97 85,41 76,24 61,15 77,93 76,48 56,09 80,38 58,06 59,39 81,17 66,33 62,41 72,95 68,75 79,14 59,99 46,93 68,65 85,41
2,92 5,56 6,10 3,65 8,06 4,23 5,97 6,34 6,39 9,15 5,40 10,28 6,52 7,59 12,71 4,07 4,43 2,81 0,02 1,25 7,18 13,31 4,21 5,34 5,98 13,31
76,64 57,17 56,60 71,00 62,58 56,10 76,23 70,71 82,38 78,25 61,27 51,08 65,94 83,82 72,00 40,47 73,52 68,11 74,57 60,20 70,12 63,87 47,17 71,75 66,31 83,82
29,17 9,68 3,43 15,70 10,24 2,24 12,89 13,08 14,21 6,40 6,78 5,68 0,25 13,01 10,14 3,33 5,94 6,23 19,25 12,51 15,50 6,37 11,06 16,61 10,40 29,17
very important issues such as camera temperature effects, illumination correction, easy camera gamma estimation and a specific color space generation. The system is simple, and after calibration, allows the user to continue using his own software and algorithms, but with a much higher informative content. Corrected images should be handled carefully to ensure conveyance of the final selected color space to whoever receives them, to enable their use in color calibrated contexts. At the time we didn’t explore the possibility of including the color space description into an ICC specification, but this could be the next step for a commercially suitable product. The analysis of smaller gamut color spaces (to better describe interest zones) is still a matter of debate in standardized color management documents. Encouraging the widespread use of color calibration in this field, will not only improve the quality of dermatoscopic digital libraries, but will also open the way to teleconsulting, remote analysis and result comparisons
662
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 76.4 Euclidean Distances and Histogram Intersections in the Instruments RGB Color Space between Corresponding Lesions Measured with the Two Different Instruments, before and after Calibration Lesion Uncalibrated 1 2 3 4 5 6 Mean Maximum
82,96 75,22 100,86 66,73 68,66 83,78 79,70 100,86
Skin
Calibrated 15,78 4,61 37,61 8,19 5,96 17,33 14,91 37,61
Uncalibrated 121,92 113,45 133,30 102,16 114,19 113,44 116,41 133,30
Histogram Calibrated 21,02 3,69 25,50 7,55 5,76 3,26 11,13 25,50
Uncalibrated
Calibrated
15,05% 13,94% 8,19% 23,14% 6,76% 0,11% 11,20% 23,14%
61,63% 82,78% 64,44% 90,56% 80,37% 77,30% 76,18% 90,56%
FIGURE 76.8 (See color insert.) Worst calibration result: first line shows the original images, second line shows the results of lighting correction, third line shows the calibrated images.
Practical Color Calibration for Dermatoscopic Images
663
FIGURE 76.9 (See color insert.) Best calibration result: first line shows the original images, second line shows the results of lighting correction, third line shows the calibrated images.
between different computer algorithms, in order to stimulate joint development or knowledge sharing.
REFERENCES 1. Kenet R.O., Kang S., Kenet B.J., Fitzpatrick T.B., Sober A.J., and Barnhill R.L., Clinical diagnosis of pigmented lesions using digital epiluminescence microscopy, Arch. Dermatol., 129, 157, 1993. 2. Bufounta M.L., Beauchet A., Aegerter P., and Saiag P., Is dermoscopy (epiluminescence microscopy) useful for the diagnosis of melanoma? Arch. Dermatol., 137, 1343, 2001. 3. Pehamberger H., Steiner A., and Wolff K., In vivo epiluminescence microscopy of pigmented skin lesions. I. Pattern analysis of pigmented skin lesions, J. Am. Acad. Dermatol., 17, 571, 1987.
4. Nachbar F., Stolz W., Merkle T., Cognetta A.B., Vogt T., Landthaler M., Bilek P., Braun-Falco O., and Plewig G., The ABCD rule of dermatoscopy, J. Am. Acad. Dermatol., 30, 551, 1994. 5. Menzies S.W., Ingvar C., and McCarthy W.H., A sensitivity and specificity analysis of the surface microscopy features of invasive melanoma, Melanoma. Res., 6, 55, 1996. 6. Argenziano G., Fabbrocini G., Carli P., De Giorgi V., Sammarco E., and Delfino M., Epiluminescence microscopy for the diagnosis of doubtful melanocytic skin lesions. Comparison of the ABCD rule of dermoscopy and a new 7-point checklist based on pattern analysis, Arch. Dermatol., 134, 1563, 1998. 7. Mac Kie R.M., Fleming C., Mc Mahon A.D., and Jarret P., The use of the dermatoscope to identify early melanoma using the three-colour test, Brit. J. Dermatol., 146, 481, 2002.
664
Handbook of Non-Invasive Methods and the Skin, Second Edition
8. Cascinelli N., Ferrario M., Bufalino R., Zurrida S., Galimberti V., Mascheroni L., Bartoli C., and Clemente C., Results obtained by using a computerized image analysis system designed as an aid to diagnosis of cutaneous melanoma, Melanoma Res., 2, 163, 1992. 9. Schindewolf T., Stolz W., Albert R., Abmayr W., and Harms H., Classification of melanocytic lesions with color and texture analysis using digital image processing, Analyt. Quant. Cytol. Histol., 15, 1, 1993. 10. Green A.C., Martin N.G., Pfitzner J., O’Rourke M., and Knight N., Computer image analysis in the diagnosis of melanoma, J. Am. Acad. Dermatol., 31, 958, 1994. 11. Cucchiara R., Grana C., Seidenari S., and Pellacani G., Exploiting color and topological features for region segmentation with recursive fuzzy c-means, Machine Graphics and Vision, 11, 169, 2002. 12. Gutkowicz-Krusin D., Elbaum M., Szwaykowski P., and Kopf A.W., Can early malignant melanoma be differentiated from atypical melanocytic nevus by in vivo techniques? Part II. Automatic machine vision classification, Skin Res. Technol., 3, 15, 1997. 13. Seidenari S., Pellacani G., and Pepe P., Digital videomicroscopy improves diagnostic accuracy for melanoma, J. Am. Acad. Dermatol., 39, 175, 1998. 14. Pellacani G., Martini M., and Seidenari S., Digital videomicroscopy with image analysis and automatic classification as an aid for diagnosis of Spitz nevus, Skin Res. Technol., 5, 266, 1999. 15. Seidenari S., Pellacani G., and Giannetti A., Digital videomicroscopy and image analysis with automatic classification for detection of thin melanoma, Melanoma Res., 9, 163, 1999. 16. Seidenari S., Pellacani G., Grana C., Computer description of colours in dermoscopic melanocytic lesion images reproducing clinical assessment, Br. J. Dermatol., 149, 523, 2003. 17. Pellacani G., Grana C., and Seidenari S., Automated description of colours in polarized-light surface microscopy images of melanocytic lesions, Melanoma Res., 14, 125, 2004. 18. Seidenari S., Pellacani G., Righi E., and Di Nardo A., Is JPEG-compression of videomicroscopic images compatible with telediagnosis? Comparison between diagnostic performance and pattern recognition on uncompressed TIFF images and JPEG compressed ones, Telem. J. E-Health, 10, 294, 2004. 19. Pellacani G., and Seidenari S., Comparison between morphological parameters in pigmented skin lesion images acquired by means of epiluminescence surface microscopy and polarized light videomicroscopy, Clin. Dermatol., 20, 222, 2002.
20. Pellacani G., Grana C., Cucchiara R., and Seidenari S., Automated extraction and description of dark areas in surface microscopy melanocytic lesion images, Dermatology, 208, 21, 2004. 21. Vannoorenberghe P., Colot O., and De Brucq D., Dempster-Shafer’s Theory as an aid to Color Information Processing Application to Melanoma Detection in Dermatology, Proceedings of the Int. Conf. Image Analysis and Processing, 774, 1999. 22. Faziloglou Y., Stanley R.J., Moss R.H., Van Stoecker W., and McLean R.P., Colour histogram analysis for melanoma discrimination in clinical images, Skin Res. Technol., 9, 147, 2003. 23. Gerger A., Stolz W., Pompl R., and Smolle J., Automated epiluminecence microscopy-tissue counter analysis using CART and 1-NN in the diagnosis of Melanoma, Skin Res. Technol., 9, 101, 2003. 24. Korichi R., Provost R., Heusèle C., and Schnebert S., Quantitative assessment of properties of make-up products by video imaging: application to lipsticks, Skin Res. Technol., 6, 222, 2000. 25. Miyamoto K., Takiwaki H., Hillebrand G.G., and Arase S., Development of a digital imaging system for objective measurement of hyperpigmented spots on the face, Skin Res. Technol., 8, 227, 2002. 26. Setaro M., and Sparavigna A., Quantification of erythema using digital camera and computer-based colour image analysis: a multicentre study, Skin Res. Technol., 8, 84, 2002. 27. Herbin M., Venot A., Devaux J.Y., and Piette C., Color Quantitation Through Image Processing in Dermatology, IEEE T. Med. Imaging., 9, 262, 1990. 28. Haeghen Y.V., Naeyaert J.M.A.D., Lemahieu I., and Philips W., An Imaging System with Calibrated Color Image Acquisition for Use in Dermatology, IEEE T. Med. Imaging., 19, 722, 2000. 29. Berns R.S., Billmeyer and Saltzman’s Principles of ColorTechnology, 3rd ed. Wiley-Interscience, New York, 2000. 30. Swain M.J., and Ballard D.H., Color Indexing, Int. J. Comput. Vision., 7, 11, 1991. 31. Rubner Y., Tomasi C., and Guibas L.J., A Metric for Distributions with Applications to Image Databases, Proceedings of the 1998 IEEE Int. Conf. Comput. Vision, 59, 1998. 32. Seidenari S., Pellacani G., and Grana C., Computer description of colours in dermoscopic melanocytic lesion images reproducing clinical assessment, Br. J. Dermatol., 149, 523, 2003. 33. Pellacani G., Grana C., Cucchiara R., and Seidenari S., Automated extraction and description of dark areas in surface microscopy melanocytic lesion images, Dermatology, 208, 21, 2004.
of Erythema and 77 Measurement Melanin Indices Hirotsugu Takiwaki Department of Dermatology, The University of Tokushima School of Medicine, Tokushima, Japan
CONTENTS 77.1 Introduction............................................................................................................................................................665 77.2 Object and Methodological Principle....................................................................................................................665 77.2.1 Theoretical Aspect .....................................................................................................................................665 77.2.2 Instruments.................................................................................................................................................667 77.3 Sources of Error.....................................................................................................................................................667 77.4 Correlation with Other Methods ...........................................................................................................................668 77.5 Clinical and Experimental Applications................................................................................................................669 77.6 Recommendation ...................................................................................................................................................670 References .......................................................................................................................................................................670
77.1 INTRODUCTION
77.2.1 THEORETICAL ASPECT
Erythema, also referred to as hemoglobin (Hb), and melanin indices are the indicators that quantify the intensity of erythema and pigmentation, respectively, and are derived from reflectance data of the skin at specific wavelengths. Unlike color coordinates, such as CIE-L*a*b*,1 these indices are designed to show quantities that correlate linearly with the amounts of hemoglobin and melanin in the skin. Therefore, they can be handled as genuine physical quantities. Although these indices were first devised in the 1980s to extract information about the amounts of Hb and melanin from reflectance spectra of the skin,2 portable types of specialized instruments have been developed and are now commercially available and widely used in the fields of dermatology, cosmetic science, and pharmacology.
The erythema and melanin indices are both based on the reflectance of an object in a selected band of the spectrum. In order to understand the significance of the erythema and melanin indices, it is helpful to use a simplified, multilayered skin model. It is composed of three layers, with the uppermost layer containing only melanin, the middle layer only hemoglobin, and the bottom layer neither chromophore (Figure 77.1). These layers represent the epidermis, the plexus of blood vessels in the upper dermis, and the dermis below them, respectively. In this model, we assume that when each layer is placed separately on an ideal black background, there is no regular reflection from the surface, the diffuse reflectance of each of the upper two layers is nearly zero, and that of the bottom layer is high. According to Dawson et al.2 and Diffey et al.,3 the total reflectance, R, of this model at a given wavelength can be roughly expressed as
77.2 OBJECT AND METHODOLOGICAL PRINCIPLE The quantity of hemoglobin, which corresponds directly to the extent of erythema, is expressed as the erythema index or hemoglobin index, and that of melanin as the melanin index. The objective of the erythema index is to quantitatively evaluate anemic or hyperemic color changes of the skin, and that of the melanin index to assess hypoor hyperpigmented skin conditions.
R = I/I0 ≈ TM2 TH2 RD
(77.1)
where I0 is the intensity of incident light and I that of reflected light, TM is the transmittance of the uppermost layer and TH that of the middle layer, and RD is the diffuse reflectance of the bottom layer. By taking the logarithm to base 10 of the inverse reflectance (LIR), from Equation 77.1 we get 665
666
Handbook of Non-Invasive Methods and the Skin, Second Edition
80 Absorbance (a.u.)
Epidermis
Superficial plexus
Dermis
60 40 20 0 450
(a)
Transmittance
Melanin layer
TM
Hemoglobin layer
TH Diffuse reflectance RD
Chromophore free backing (b)
FIGURE 77.1 (a) Schematic structure of the skin. (b) Optical skin model of three-layered structure with an outer melanin layer, an inner hemoglobin layer, and a backing representing chromophore-free dermis.
log(1/R) ≈ 2log(1/TM) + 2log(1/TH) + log(1/RD) (77.2) Since it is assumed that the scattering within the upper two layers is extremely minor, the Beer–Lambert law of radiation absorption can be applied when the concentration of melanin and hemoglobin is low. Equation 77.2 can then converted to Log(1/R) ≈ dMEMCM + dHEHCH + log(1/RD) (77.3) where dM is the thickness of the melanin layer, EM the coefficient determined by the absorbance of melanin at a given wavelength, and CM the concentration of melanin, respectively, while dH, EH, and CH are their counterparts in the hemoglobin layer. Log (1/R) is referred to as the apparent absorbance, as this value is regarded as a double of the ordinary absorbance (= log (1/transmittance)). By substituting A for log (1/R), m for dMEM, h for dHEH, and d for log (1/RD), Equation 77.3 can be written as A ≈ mCM + hλCH + dλ
550 600 Wavelength (nm)
650
FIGURE 77.2 Spectra of apparent absorbance of oxyhemoglobin (●), deoxyhemoglobin (), and melanin (x). (Modified from Hagisawa, S. and Ferguson-Pell, M., Ikakikaigaku, 64, 299, 1994. With permission.)
Reflected light (I)
Incident light (Io)
500
(77.4)
where Δ equals the given wavelength. If we select two different wavelengths, 1and 2, and designate the respective apparent absorbance at these wavelengths as A1 and A2, we get A1 – A2 = ΔA ≈ (m1 – m2) CM + (h1 – h2) CH (77.5) since the human dermis has similar reflectance values in wavelengths of visible light. Figure 77.2 shows the absorbance spectra of reduced or oxygenated Hb and that of melanin. If we select a set of λ1 and λ2 for which m1 – m2 is nearly zero and h1 – h2 is substantially high, then A (h1 – h2) CH
(77.6)
A is therefore suitable for use as the erythema index, because this value is likely to be proportional to CH. Similarly, if we select a set of 1 and 2 for which h1 – h2 is nearly zero and m1 – m2 is substantially high, A is likely to be proportional to Cm, which therefore is suitable for use as the melanin index. In practice, a wavelength or narrowband wavelengths corresponding to green (around 550 nm) and red (around 650 nm) light are selected as λ1 and λ2, respectively.3–7 For the melanin index, a wavelength or narrowband wavelengths corresponding to red light (e.g., 620 and 720 nm)4,8 are selected as both λ1 and λ2 for some systems, while one instrument (Mexameter) uses red and near-infrared lights.7 Another instrument (DermaSpectrometer) uses only one narrowband wavelength centered at 670 nm instead of two separate wavelengths, as hλ is nearly zero in this wavelength.6 In this case, Equation 77.4 is written as A ≈ mλ CM + dλ
(77.7)
in which it is of note that dλ is not negligible, so that the index value is not zero, but appears to be high even if the skin contains no melanin. Most instruments and systems
Measurement of Erythema and Melanin Indices
use the definition of erythema and melanin indices outlined here. More complex definitions are also used, however, to produce a more accurate erythema (or melanin) index that is less influenced by factors other than Hb (or melanin) when a full-band reflectance spectrophotometer is used for derivation of the indices.2,9 The principle of index derivation, however, is basically similar to that underlying the absorbance of the multilayered model.
77.2.2 INSTRUMENTS
Erythema and melanin indices can be roughly assessed from digital images of the skin captured with a digital camera or videomicroscope.10 In this case, a white color chip should be included in the image. This chip is used for the white standard to calibrate brightness of illumination and to calculate the averaged reflectance of the region of interest selected in the image. By substituting apparent absorbance data in the red and green channels for those measured at λ1 and λ2, an erythema index and a DermaSpectrometer-type melanin index can be derived.
77.3 SOURCES OF ERROR Although these instruments are convenient for obtaining indices, there are some critical points that examiners should keep in mind. First, the erythema index may be affected by hyperpigmentation.4,10 Equation 77.5 indicates that the influence of melanin cannot be neglected if CM is high, which is the case with hyperpigmented skin. Figure 77.4 shows the relationship between the erythema index and the melanin index measured with DermaSpectrometer at the sites with various degrees of pigmentation induced by ultraviolet (UV) B irradiation in five Japanese subjects 14 days before the measurement. This result clearly indicates that the erythema index increases in an apparently linear fashion as the melanin index increases. Comparison of erythema indices between two sites showing quite different degrees of pigmentation should therefore be avoided. Second, the melanin index may be affected by the oxygen saturation level of hemoglobin. Since reduced hemoglobin absorbs more red light than oxyhemoglobin (Figure 77.2), the term hCH in Equation 77.4 cannot be disregarded when the concentration of reduced hemoglobin is relatively high, even if a wavelength corresponding to red light is selected. This implies that the melanin index apparently increases under static or cyanotic conditions.
Erythema index
When a full-band reflectance spectrometer is used, erythema and melanin indices can be calculated based on the various formulae that have been published.2–9 However, most of the reflectance spectrometers widely used in dermatology and cosmetology seem to be of the type that can measure reflectance at every 5 to 10 nm within the visible wavelength of light. If reflectance at the desired wavelength cannot be obtained, it is then necessary to calculate it by means of interpolation using data available for neighboring wavelengths. The most popular, commercially available instruments for measuring erythema and melanin indices seem to be the DermaSpectrometer (Cortex Technology, Hadsund, Denmark) and Mexameter MX16® and MX18® (Courage and Khazaka Electronic GmbH, Cologne, Germany) (Figure 77.3). The DermaSpectrometer is a handheld instrument equipped with two light-emitting diodes (LEDs) that emit green light centered at 568 nm and red light at 655 nm. The Mexameter is a portable instrument with three LEDs that emit green, red, and near-infrared light with respective centered wavelengths of 568, 660, and 880 nm. Both instruments detect reflected light at these narrowband wavelengths with photodiodes and use built-in microcomputers to calculate erythema and melanin indices with Equations 77.5 and 77.6.
667
FIGURE 77.3 Instruments for measuring erythema and melanin indices. From left to right, the Mexameter MX18, the reflectance spectrophotometer Minolta CM-2002, and the DermaSpectrometer. (From Kohno, H. and Takiwaki, H., Dermatol. Pract., 14, 78, 2002. With permission.)
20 18 16 14 12 10 8 6 4 2 0
25
30
35 40 Melanin index
45
50
FIGURE 77.4 Dependence of erythema index on melanin index. Both indices were measured with the DermaSpectrometer at test sites showing various degrees of pigmentation induced by UVB irradiation of different doses 14 d before the measurement.
668
Handbook of Non-Invasive Methods and the Skin, Second Edition
In fact, reduced hemoglobin easily increases during arm 20
500
Eds
Emx
400
15
300 10 200 5
100 E
0 0 60
10
20
30
40
50
E
0 0 500
Mds
50
10
20
30
40
50
Mmx
400
40 300 30 200 20 100
10 0
M 0
2
4
6
8
10
M 0
0
2
4
6
8
10
FIGURE 77.5 Correlations between the erythema index by Dawson et al.2 (E) measured with the reflectance spectrophotometer CM2002 and two types of erythema indices measured with the DermaSpectrometer (Eds) and Mexameter MX18 (Emx). Correlations between melanin indices (M, Mmx, and Mds) are also shown. Measurements were made at four different anatomical sites of 15 healthy Japanese subjects. (Redrawn from Kohno, H. and Takiwaki, H., Dermatol. Pract., 14, 78, 2002. With permission.)
lowering, which results in a paradoxical increase in the melanin index.11 These errors result inevitably from the definition of the erythema and melanin indices. In addition, errors due to incorrect techniques may occur when the measuring head is applied with too much pressure, or when the measurement area to be measured is too hairy. Measurement in direct sunlight should be avoided as well.
77.4 CORRELATION WITH OTHER METHODS Moderate to excellent linear correlations have been found between erythema indices of normal and erythematous skin measured with different instruments or systems.12 Our measurements at four anatomical locations of the normal skin of 15 Japanese volunteers showed that the correlation coefficients between Dawson’s erythema index2 and erythema indices measured with the DermaSpectrometer (DS) and Mexameter (MX) were 0.77 and 0.87, respectively (Figure 77.5).13 Clarys et al.12 reported a correlation
of 0.81 for the erythema indices determined with DS and MX. Unlike the erythema index, the definition of the melanin index differs considerably, depending on the equipment. Nevertheless, the correlation coefficients between Dawson’s melanin index2 and the melanin indices measured with DS and MX were 0.82 and 0.94, respectively, according to our evaluation (Figure 77.5).13 However, Clarys et al.12 reported the correlation for the melanin indices determined with DS and MX to be rather low (0.53). They also mentioned that the melanin index of the Mexameter MX16 was less sensitive; however, the manufacturer has developed an allegedly improved version, the MX18. As mentioned in Section 77.2, it should be noted that the melanin index of DS inevitably shows far higher values than that of other systems, even if no melanin is present in the measured skin, because the influence of the dermal factor on the melanin index has not been excluded. This raised-bottom value of the DermaSpectrometer melanin index corresponds to the value for vitiliginous skin. Our assessments of the color of normal skin and of psoriasis lesions measured with the DermaSpectrometer
Measurement of Erythema and Melanin Indices
669
ΔLDF (arbitray unit)
a∗ 30 25 20 15 10 : n = 230
5 0
10
20 Erythema index
: n = 50
–4
–2
Arm elevation
40
30
1 0.8 0.6 0.4 0.2 0 –0.2 –0.4 –0.6 –0.8 –1 –6
FIGURE 77.6 Correlation between erythema index and a* measured at 23 anatomical sites of 10 healthy Caucasian subjects (·) and that in 50 plaques of psoriasis of 10 patients (). Correlation coefficient r = 0.92 and r = 0.91, respectively. Regression lines are shown for each group. The DermaSpectrometer (erythema index) and Minolta Chromameter CR-200 (a*) were used. (Modified from Takiwaki, H. et al., Skin Pharmacol., 7, 217, 1994. With permission.)
80
0 2 ΔErythema index
4
6
8
Arm lowering
FIGURE 77.8 Relationship between the change in erythema index and that in laser Doppler blood flow following arm elevation () and lowering (●), where the change is expressed by the difference (Δ) between the value at heart and that in each position.
The correlation between the erythema index and laser Doppler flow (LDF) seems to depend upon what is the cause of erythema. The erythema index correlates positively with LDF in acute inflammation, such as seen in irritant or allergic patch test reaction15 and UV-induced erythema, whereas LDF decreases despite an increase in the erythema index during arm lowering (Figure 77.8).11 This is most likely caused by the fact that the erythema index reflects only the blood volume, while LDF reflects the movement of blood cells.16
L∗
70
77.5 CLINICAL AND EXPERIMENTAL APPLICATIONS
60
In comparison with tristimulus colorimetry, the instruments and methods discussed here have the following advantages:
50
FIGURE 77.7 Correlation between melanin index and L* measured at 23 sites of 10 Caucasians: r = –0.56, p < 0.001. (From Takiwaki, H. et al., Skin Pharmacol., 7, 217, 1994. With permission.)
1. The intensities of erythema and of pigmentation can be separately quantified. 2. The erythema index is likely to have a linear relationship with the content of red blood cells in the upper dermis, and the melanin index with that of melanin in the epidermis, unless the extent of erythema or pigmentation is intense.
and Minolta Chromameter CR-200®6,14 showed a strong linear correlation between the erythema index and a*, representing the red-green axis in the CIE-L*a*b* space (Figure 77.6). The melanin index also correlated negatively with L* values representing brightness (Figure 77.7). This correlation, however, does not mean that the melanin index is equivalent to L*, because L* is mainly determined by the reflectance of green light, and the melanin index by that of red light. The melanin index is more specific for the degree of pigmentation than L*, which is also affected by the degree of erythema.
These instruments are not suitable for measuring changes in color due to other chromophores, such as jaundice. Moreover, these indices may be inadequate for the quantification of the hemoglobin and melanin located deeply in the dermis in such lesions as cavernous hemangioma, subcutaneous hemorrhage, and nevus of Ota, because their spectral reflectance is affected by the scattering effect of the dermis, and differs from that of the usual erythema and pigmentation. Therefore, direct numerical comparison of index values should be avoided for skin diseases or conditions in which pigments are
40 10
20
30 Melanin index
40
50
670
Handbook of Non-Invasive Methods and the Skin, Second Edition
present at different depths (e.g., port-wine stain and cavernous hemangioma). Measurement of the erythema index on black skin was found to be unsuccessful.17 Various quantitative in vivo studies have been carried out by using these advantages of erythema and melanin indices. This is especially true for studies of various aspects of inflammatory skin response to ultraviolet light, such as the exact relationship between dose and response, time course of the response, regional or constitutional variations in sensitivity, dependency on wavelength, suppressive effects of anti-inflammatory agents, and the relationship between UV-induced erythema and postinflammatory pigmentation.18–24 The intensity and time course of skin response to irritants or allergens have also been assessed quantitatively.15,25,26 Furthermore, these instruments and methods can be readily applied to other assessments, such as blanching or the vasoconstrictive action of corticosteroids, evaluation of the severity of inflammatory skin diseases,14,27 and monitoring of the efficacy of topical treatments, chemical peeling, and laser surgery for various kinds of nevi and hemangiomas.28–31
77.6 RECOMMENDATION Considering how the erythema and melanin indices are defined, the following is recommended in order to make the best use of these instruments and methods: 1. Subjects should take a standardized position for each series of measurement, especially if it is performed in the extremities. 2. The measuring head should be placed very softly and perpendicularly onto the skin. Holding it too long at a test site should be avoided. Much or incorrect pressure makes the site anemic or congested. Measurements should not be performed under direct sunlight. 3. Avoid comparing erythema indices between two sites where the levels of pigmentation (melanin index) are considerably different. If obtainable, empirical corrections for the indices4,10 are desirable. 4. Skin color shows spontaneous diurnal variations32 and regional differences.6 When the intensity of skin test reaction is successively examined in a wide area, the data for the control should be obtained in the normal skin adjacent to the test site at each measurement.
REFERENCES 1. Robertson, A., The CIE 1976 color-difference formulae, Color Res. Appl., 2, 7, 1977.
2. Dawson, J.B., Barker, D.J., Ellis, D.J., Grassam, E., Cotterill, J.A., Fisher, G.W., and Feather, J.W., A theoretical and experimental study of light absorption and scattering by in vivo skin, Phys. Med. Biol., 25, 695, 1980. 3. Diffey, B.L., Oliver, R.J., and Farr, P.M., A portable instrument for quantifying erythema induced by ultraviolet radiation, Br. J. Dermatol., 111, 663, 1984. 4. Feather, J.W., Ellis, D.J., and Leslie, G., A portable reflectometer for rapid quantification of cutaneous haemoglobin and melanin, Phys. Med. Biol., 33, 711, 1988. 5. Pearse, A.D., Edwards, C., Hill, S., and Marks, R., Portable erythema meter and its application to use in human skin, Int. J. Cosmet. Sci., 12, 63, 1990. 6. Takiwaki, H., Overgaard, L., and Serup, J., Comparison of narrow-band reflectance spectrophotometric and tristimulus colorimetric measurement of skin color: 23 anatomical sites evaluated by the DermaSpectrometer and the Chromameter CR-200, Skin Pharmacol., 7, 217, 1994. 7. Edwards, C., The Mexameter MX 16™, in Bioengineering of the Skin: Methods and Instrumentation, Berardesca, E., Elsner, P., Wilhelm, K.P., and Maibach, H.I., Eds., CRC Press, Boca Raton, FL, 1995, p. 127. 8. Kollias, N. and Baqer, A., Spectroscopic characteristics of human melanin in vivo, J. Invest. Dermatol., 85, 38, 1985. 9. Feather, J.W., Hajizadeh-Saffer, M., Leslie, G., and Dawson, J.B., A portable scanning reflectance spectrophotometer using visible wavelengths for the rapid measurement of skin pigments, Phys. Med. Biol., 34, 807, 1989. 10. Takiwaki, H., Shirai, S., Kanno, Y., Watanabe, Y., and Arase, S., Quantification of erythema and pigmentation using a videomicroscope and a computer, Br. J. Dermatol., 131, 85, 1994. 11. Takiwaki, H. and Serup, J., Variation in color and blood flow of the forearm skin during orthostatic maneuver, Skin Pharmacol., 7, 226, 1994. 12. Clarys, P., Aleweaters, K., Lambrecht, R., and Barel, A.O., Skin color measurements: comparison between three instruments: the Chromameter, the DermaSpectrometer, and the Mexameter, Skin Res. Technol., 6, 230, 2000. 13. Kohno, H. and Takiwaki, H., Erythema and melanin index meter, Dermatol. Pract., 14, 78, 2002 (in Japanese). 14. Takiwaki, H. and Serup, J., Measurement of color parameters of psoriatic plaques by narrow-band reflectance spectrometry and tristimulus colorimetry, Skin Pharmacol., 7, 145, 1994. 15. Gawkrodger, D.J., McDonagh, A.J.G., and Wright, A.L., Quantification of allergic and irritant patch test reactions using laser-Doppler flowmetry and erythema index, Contact Derm., 24, 172, 1991. 16. Bircher, A., Boer, E.M.D., Agner, T., Wahlberg, J.E., and Serup, J., Guidelines for measurement of cutaneous blood flow by laser Doppler flowmetry. A report from the standardization group of European Society of Contact Dermatitis, Contact Derm., 30, 65, 1994.
Measurement of Erythema and Melanin Indices
17. Takiwaki, H., personal communication, 1991. 18. Farr, P.M. and Diffey, B.L., Quantitative studies on cutaneous erythema induced by ultraviolet radiation, Br. J. Dermatol., 111, 673, 1984. 19. Farr, P.M. and Diffey, B.L., The erythemal response of human skin to ultraviolet radiation, Br. J. Dermatol., 113, 65, 1985. 20. Farr, P.M., Besab, J.E., and Diffey, B.L., The time course of UVB and UVC erythema, J. Invest. Dermatol., 91, 454, 1988. 21. Farr, P.M. and Diffey, B.L., A quantitative study of the effect of topical indomethacin on cutaneous erythema induced by UVB and UVC radiation, Br. J. Dermatol., 115, 453, 1986. 22. Takiwaki, H., Shirai, S., Kohno, H., Soh, H., and Arase, S., The degrees of UVB-induced erythema and pigmentation correlate linearly and are reduced in a parallel manner by topical anti-inflammatory agents, J. Invest. Dermatol., 103, 642, 1994. 23. Park, S.B., Huh, C.H., Choe, Y.B., and Youn, J.I., Time course of ultraviolet-induced skin reactions evaluated by two different reflectance spectrophotometers: DermaSpectrometer and Minolta spectrophotometer, Photodermatol. Photoimmunol. Photomed., 18, 23, 2002. 24. Damian, D.L., Halliday, G.M., and Barnetson, R.S., Prediction of minimal erythema dose with a reflectance melanin meter, Br. J. Dermatol., 136, 714, 1997. 25. Held, E. and Agner, T., Comparison between 2 test models in evaluating the effect of a moisturizer on irritated human skin, Contact Derm., 40, 261, 1999.
671
26. Fluhr, J.W., Kuss, O., Diepgen, T., Lazzerini, S., Pelosi, A., Gloor, M., and Barardesca, E., Testing for irritation with a multifactorial approach: comparison of eight noninvasive measuring techniques on five different irritation types, Br. J. Dermatol., 145, 696, 2001. 27. Suh, D.H., Kwon, T.E., Kim, S.D., Park, S.B., Kwon, O.S., Eun, H.C., and Youn, J.I., Changes of skin blood flow and color on lesional and control sites during PUVA therapy for psoriasis, J. Am. Acad. Dermatol., 44, 987, 2001. 28. Gladstone, H.B., Nguyen, S.L., Williams, R., Ottomeyer, T., Wortzman, M., Jeffers, M., and Moy, R.L., Efficacy of hydroquinone cream (USP 4%) used alone or in combination with salicylic acid peels in improving photodamage on the neck and upper chest, Dermatol. Surg., 26, 333, 2000. 29. Hurley, M.E., Guevara, I.L., Gonzales, R.M., and Pandya, A.G., Efficacy of glycolic acid peels in the treatment of melasma, Arch. Dermatol., 138, 1578, 2002. 30. Miyamoto, H., Takiwaki, H., Yamano, M., Ahsan, K., and Nakanishi, H., Color analysis of nevus of Ota for evaluation of treatment with a Q-switched alexandrite laser, Skin Res. Technol., 3, 45, 1997. 31. Manuskiatti, W., Sivayathorn, A., Leelaudomlipi, P., and Fitzpatrik, R.E., Treatment of acquired bilateral nevus of Ota-like macules (Hori’s nevus) using a combination of scanned carbon dioxide laser followed by Q-switched ruby laser, J. Am. Acad. Dermatol., 48, 584, 2003. 32. Queille-Roussel, C., Poncet, M., and Schaffer, H., Quantification of skin-colour changes induced by topical corticosteroid preparation using Minolta Chroma Meter, Br. J. Dermatol., 124, 264, 1991.
78 Dynamic Capillaroscopy H.S. Yu,1 C.H. Lee,2 and C.H. Chang3 1
Department of Dermatology, College of Medicine, National Taiwan University Hospital, Taipei, Taiwan 2Department of Dermatology, College of Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan 3Department of Dermatology, College of Medicine, Tzu-Chi Medical University, Hualien, Taiwan
CONTENTS 78.1 Introduction............................................................................................................................................................673 78.2 Methodological Principle ......................................................................................................................................673 78.2.1 Computerized Laser Capillary Microscopy ..............................................................................................673 78.2.2 Dynamic Capillaroscopy with Load: Pressure, Cold Provocation, and Iontophoresis ............................674 78.3 Application.............................................................................................................................................................674 78.3.1 Tetralogy of Fallot .....................................................................................................................................674 78.3.2 Raynaud’s Syndrome and Connective Tissue Diseases............................................................................675 78.3.3 Diabetes Mellitus .......................................................................................................................................675 78.3.4 Arteriosclerosis ..........................................................................................................................................675 References .......................................................................................................................................................................676
78.1 INTRODUCTION Cutaneous microcirculation can be divided into thermoregulatory shunt vessels and nutritive skin capillaries. Laser Doppler flowmetry, transcutaneous oxygen pressure measurement, and capillary microscopy are known noninvasive methods for assessing cutaneous microcirculation. Flux in nonnutritional shunt vessels dominates the signal recording of the laser Doppler flowmetry. Transcutaneous oxygen pressure primarily reflects the function of the nutritive skin capillaries. Presently, capillary microcopy is the best choice for studying the nutritional status of a certain skin area, and it is the only method that allows direct visualization of the nutritive dermal vessels in vivo.1,2 In 1964, Zimmer and Demis3 demonstrated a microscope–television system for studying dynamic blood flow in human skin capillaries. Bollinger and coworkers4 further refined this method and adapted a frame-to-frame analysis to measure both blood flow velocity and vessel diameters in nail-fold capillaries. The application of the cross-correlation technique5 for measuring the velocity of blood cells in the capillaries greatly improved this measurement. Thereafter, capillary microscopy has been coupled with a videophotometric system and used with software to analyze the capillary blood cell velocity (CBV).6 This technique makes it possible to noninvasively study human skin capillaries under physiological and pathophysiological conditions. Applications
include evaluation of the dynamic microcirculatory status in peripheral vascular disorders, including arterial occlusive disease, connective tissue disease, and diabetes mellitus. The purpose of this chapter is to introduce recent advances in dynamic capillaroscopy in the study of diseases, namely, tetralogy of Fallot, Raynaud’s syndrome, atherosclerosis, and diabetes mellitus (DM); a focus on clinical application will be presented.
78.2 METHODOLOGICAL PRINCIPLE 78.2.1 COMPUTERIZED LASER CAPILLARY MICROSCOPY During the past few years, a new type of computerized capillary microscope has been developed (Figure 78.1). A low-power near-infrared laser is focused to a 10-microndiameter beam that can be positioned onto a single capillary. The built-in high-resolution charge coupled device (CCD) camera allows continuous monitoring of the capillary position on either the computer screen or a separate video monitor. The laser beam is reflected by moving blood cells at the focal point. The frequency of the reflected beam is Doppler shifted. The detected shift is directly proportional to the velocity of the reflecting blood cells. Coupled with software for analysis, digital image 673
674
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 78.1 Computerized laser capillary microscope.
change and rest CBV (rCBV) in capillaries. However, in some pathological conditions, physical load is necessary to display the abnormality in capillaries. A 1-min arterial occlusion and cold provocation tests are useful methods for this purpose. An arterial occlusion in of digital arteries results in a reactive hyperemia, which indicates the myogenic response of the precapillary sphincter. The important parameters for postocculsive reactive hyperemia response include peak CBV (pCBV), time to pCBV (tpCBV), and the percent increase of CBV above rCBV during postocculsive reactive hyperemia (%PRH). The cold provocation test5–15 has been used for studying the disturbances of skin microvascular reactivity in different types of Raynaud’s syndrome. Using dynamic capillaroscopy and iontophoresis, Serne et al.7 demonstrated that systemic hyperinsulinemia in skin induces recruitment of capillaries, augments vasodilatation, and influences vasomotion in diabetic patients.
78.3 APPLICATION 78.3.1 TETRALOGY
FIGURE 78.2 Normal hairpin-like capillary loops in regular arrangement.
output, and an improved optical and electronic system, it provides an easy, stable, direct, and accurate measurement of cutaneous capillary dynamics. For CBV measurement, capillaries in the nail-fold area are suitable for this purpose because they lie parallel to the skin surface and can be visualized rather nicely in their full length, resembling a hairpin (Figure 78.2). A finger or toe is placed on the investigation plate, and a small bracket is allowed to lightly touch the distal end of the nail. A drop of immersion oil is applied to make the nail fold transparent and increase refraction. When capillary microscopy is performed using a video recording system, it permits realtime imaging of cutaneous capillaries and retrospective analysis of capillary dynamics. The whole process is performed automatically. Observation parameters consist of recording of capillary density; capillary morphology, including caliber, length, shape, and tortuosity; intercapillary distance; and CBV in afferent limb, apex, and efferent limb of the capillary loops.
OF
FALLOT
Tetralogy of Fallot (TF) is recognized as the most common congenital cyanotic cardiac malformation, consisting of a ventricular septal defect, pulmonary stenosis, aortic override, and right ventricular hypertrophy. Patients with TF usually suffer from generalized cyanosis, clubbing digits, and a high risk of thrombosis due to secondary polycythemia. Cutaneous microcirculation was studied, and we found that the nail-fold capillaries in TF patients became dilated, tortuous, and branched with an increase of total length and vascular area (Figure 78.3). The degree of dilation and vascularity was closely related to the hemoglobin (Hb) concentration. CBV declined with the increase of hematocrit (Hct), and a significant reduction was noted when Hb >19 g/dl. TF patients are good models for studying the effects of long-term hypoxia. Capillary dilation and vascularity increase are compensatory
78.2.2 DYNAMIC CAPILLAROSCOPY WITH LOAD: PRESSURE, COLD PROVOCATION, AND IONTOPHORESIS Computerized capillary microscopy is the most sensitive noninvasive method for evaluation of both morphological
FIGURE 78.3 Extremely dilated, torturous capillary loops observed in a patient of tetralogy of Fallot with hemoglobin of 22 g/dl.
Dynamic Capillaroscopy
675
reactions and are reversible. We can use capillary microscopy to evaluate the compensatory status of TF patients before operation and dynamic changes after operation.8
78.3.2 RAYNAUD’S SYNDROME TISSUE DISEASES
AND
CONNECTIVE
Raynaud’s syndrome is the paroxysmal constriction of small arteries of the extremities, usually precipitated by cold. When exposed to low temperatures, the digits become white (ischemic), then blue (cyanotic), and finally red (hyperemia).9 Over 90% of the cases are primary Raynaud’s syndrome. The principal clinical challenge is to distinguish idiopathic cases of primary Raynaud’s syndrome from secondary ones due to underlying disease of connective tissues, obstructive arterial disease, blood dyscrasia, drug toxicity, or artery injury. Maricq et al.10,11 first described the abnormality of capillary morphology in connective tissue disease, especially in scleroderma. It is a qualitative method to differentiate primary from secondary Raynaud’s syndrome. A quantitative morphological analysis of capillary microscopy was developed by Lefford and Edward.12 The pattern of capillary morphology consists of enlarged and deformed capillaries with dilation of both limbs of the loop, which is often engorged with blood (sausage loop). Marked disorganization of the loop is observed. Loss of capillaries produces many avascular areas and the disruption of the orderly appearance of the capillary bed. Capillary morphology of primary Raynaud’s disease is approximately normal, but the CBV decreased markedly after cold exposure.13 On the other hand, abnormal morphological changes were noted in secondary Raynaud’s syndrome, including an increase in intercapillary distance and ratio of torturous capillary loops, and a decrease in capillary loops, numbers, and total length.14,15 In addition, there is a significant dilation of afferent and efferent limbs and apex.16 The degree of abnormal morphological changes in systemic sclerosis (Figure 78.4) is more severe than lupus erythematosus. Enlarged, giant capillaries, a
FIGURE 78.4 In progressive systemic sclerosis (PSS), a continuous blood stream in capillary is replaced by a slow blood flow and a granular pattern (arrow) that reflects an aggregation of erythrocytes in capillaries.
reduced numbers of capillaries, severe avascularity, and hemorrhage were most commonly seen in systemic sclerosis than in other connective tissue diseases and normal controls.17 Patients with Raynaud’s phenomenon with avascularity or a mean of more than two megacapillaries per digit are likely to develop a scleroderma spectrum disorder.18 Dynamic capillaroscopy reveals its accuracy in differentiating systemic sclerosis from primary Raynaud’s syndrome and lupus erythematosus, as well as the clinical progressions of underlying diseases.
78.3.3 DIABETES MELLITUS Microcirculation is known to be disturbed in many organs of diabetic patients. A previous study revealed that prevalence of coiled and slightly enlarged microvessels is significantly increased in long-term diabetics at the nail folds of fingers19 and toes.20 In diabetes without retinopathy, no significant changes of nail-fold capillaries could be detected. In diabetes with background retinopathy, they showed dilatation and tortuosity, whereas in diabetes with proliferative retinopathy, they showed loss of loop and retardation of blood flow (Figure 78.5). It was also found that rCBV did not differ significantly in patients with diabetes compared to controls.21 However, dynamic capillaroscopy with a 1-min digital arterial occlusion test can detect early impairment of cutaneous microcirculation in DM.22 In nonretinopathy diabetes, functional impairments of capillary circulation include decreased rCBV, pCBV, and prolonged tpCBV. The degree of tortouosity of capillaries and impairment in pCBV and tpCBV of capillary circulation are significantly correlated with the gravity of retinopathy in diabetic patients.23 Dynamic capillaroscopy used in concert with ophthalmoscopy can facilitate a comprehensive examination of vasculopathy in DM.
78.3.4 ARTERIOSCLEROSIS Morphological changes of capillaries appear with the progression of peripheral vascular disease. Jorneskog et al.24
FIGURE 78.5 In diabetes with proliferative retinopathy, in addition to tortuosity, there is a loss of loop.
676
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 78.6 In arteriosclerosis, there are many dilated and tortuous loops of capillaries.
utilized dynamic capillaroscopy to measure capillary blood cell velocity during rest and following a 1-min arterial occlusion at the toe base. The skin microvascular reactivity was impaired in both diabetic and nondiabetic patients with peripheral vascular disorders.24 In patients with arteriosclerosis, the morphology analysis demonstrated increased numbers of tortuous loops of capillaries in affected toes (Figure 78.6). The rCBV of both toes and fingers was within normal range. However, there was a significant deterioration in the parameters of postocclusive reactive hyperemia response, i.e., a decrease in pCBV and an increase in tpCBV. These results suggest that dynamic capillaroscopy is a sensitive method for evaluating the cutaneous nutritive status in arteriosclerosis.25 Dynamic capillaroscopy provides a new approach to the early detection of circulatory disturbances resulting from different mechanisms such as collagen vascular diseases, diabetes mellitus, arteriosclerosis, and perhaps even some types of heart disease.
REFERENCES 1. Fagrell B. Vital capillary microscopy: a clinical method for studying changes of the nutritional skin capillaries in legs with arteriosclerosis obliterans. Scan J Clin Lab Invest, Suppl. 133, 1973. 2. Fagrell B. Advances in microcirculation network evaluation: an update. Int J Microcirc 15:34–40, 1995. 3. Zimmer JG, Demis DJ. The study of the physiology and pharmacology of the human cutaneous microcirculation by capillary microscopy and television cinematography. Angiology 15:232–235, 1964. 4. Bollinger A, Butti P, Barras JP, Trachsler H, Siegenthaler W. Red blood cell velocity in nailfold capillaries of man measured by a television microscopy technique. Microvasc Res 7:61–72, 1974. 5. Intaglietta M, Silverman NR, Tompkins WR. Capillary flow velocity measurements in vivo and in situ by television methods. Microvasc Res 10:165–179, 1975. 6. Fagrell B, Eriksson SE, Malmstrom S, Sjolund A. Computerized data analysis of capillary blood cell velocity. Int J Microcirc Clin Exp 7:276, 1988.
7. Serne EH, IJzerman RG, Gans RO, Nijveldt R, De Vries G, Evertz R, Donker AJ, Stehouwer CD. Direct evidence for insulin-induced capillary recruitment in skin of healthy subjects during physiological hyperinsulinemia. Diabetes 51:1515–1522, 2002. 8. Chang CH, Yu HS. Study of cutaneous microcirculation in tetralogy of Fallot. Microsvas Res 51:59–68, 1996. 9. Coffman JD. Rayanud’s phenomenon. Hypertension 17:593–602, 1991. 10. Maricq HR, LeRoy EC. Patterns of finger capillary abnormalities in connective tissue disease by wide-field microscopy. Arthritis Rheum 16:619–628, 1973. 11. Maricq HR, LeRoy EC, Dangelo WA, Medsger TA, Rodnan GP, Sharp GC, Wolfe JF. Diagnostic potential of in vivo capillary microscopy in scleroderma and related disorders. Arthritis Rheum 23:183–189, 1980. 12. Lefford F, Edward JCW. Nailfold capillary microscopy in connective tissue disease: a quantitative morphological analysis. Ann Rheum Dis 45:741–749, 1986. 13. Jacohs MJHM, Breslau PJ, Slaaf DW, Reneman RS, Lemmens JAJ. Nomenclature of Raynaud’s phenomenon: a capillary microscopic and hemorheologic study. Surgery 101:136–145, 1987. 14. Ohtsuka T, Ishikawa H. Graphic analysis of nailfold capillary in patients with collagen disease, especially in those with systemic scleorosis. Jpn J Clin Dermatol 45:637–644, 1991. 15. Caspary L, Schmees C, Schoetensack I, Hartung K, Stannat S, Deicher H, Creutig A, Alexander K. Alternations of the nailfold capillary morphology associated with Raynaud’s phenomenon in patients with systemic lupus erythematosus. Rheumatology 18:559–566, 1991. 16. Liu CG, Su W, Luo Y. Changes in cutaneous microcirculation, hemorheology and platelet aggregation function in dermatomyositis. J Dermatol Sci 2:346–352, 1991. 17. Kabasakal Y, Elvins DM, Ring EF, McHugh NJ. Quantitative nailfold capillaroscopy findings in a population with connective tissue disease and in normal healthy controls. Ann Rheum Dis 55:507–512, 1996. 18. Zufferey P, Depairon M. Prognostic significance of nailfold capillary microscopy in patients with Raynaud’s phenomenon and scleroderma-pattern abnormalities. A six-year follow-up study. Clin Rheumatol 11:536–541, 1992. 19. Rouen LR, Terry EN, Doft BH, Clauss RH, Redisch W. Classification and measurement of surface microvessels in man. Microvasc Res 4:285–292, 1972. 20. Chazan BI, Balodimos MC, Larine RI, Koncz L. Capillaries of the nailfold in diabetes mellitus. Microvasc Res 2:504–507, 1970. 21. Tooke JE, Lins PE, Ostergren J, Fagrell B. Skin microvascualr autoregulatoy response in type I diabetes: the influence of duration and control. Int J Microcirc Clin Exp 4:249–256, 1985. 22. Tooke JE, Ostergern J, Lins PE, Fagrell B. Skin microvascular blood flow control in long duration diabetes with and without complications. Diabetes Res 5:187–192, 1987.
Dynamic Capillaroscopy
23. Chang CH, Tsai RK, Wu WC, Kuo SL, Yu HS. Use of dynamic capillaroscopy for studying cutaneous microcirculation in patients with diabetes mellitus. Microvasc Res 53:121–127, 1997. 24. Jorneskog G, Brismar K, Fagrell B. Skin capillary circulation is more impaired in the toes of diabetic than non-diabetic patients with peripheral vascular disease. Diabet Med 12:36–41, 1995.
677
25. Yu HS, Chang CH, Chen GS, Yang SA, Yu CL. Study of dynamic microcirculatory problems in ‘blackfoot disease’: emphasizing its differences from arteriosclerosis. J Biomed Sci 2:183–188, 1995.
and 79 Capillaroscopy Videocapillaroscopy Assessment of Skin Microcirculation: Dermatological and Cosmetic Approaches Philippe Humbert,1 Jean-Marie Sainthillier,1 Sophie Mac-Mary,1 Adeline Petitjean,1 Pierre Creidi,1 and Tijani Gharbi2 1 2
Laboratoire de Biologie et Cutanée d’Ingénierie, Besançon France Laboratoire d’Optique P.M. Duffieux, University of Franche-Comté, Besançon France
CONTENTS 79.1 79.2 79.3 79.4 79.5 79.6 79.7 79.8 79.9 79.10
Anatomy of the Skin Microcirculation................................................................................................................679 Capillaroscopy......................................................................................................................................................679 Periungual Capillaroscopy ...................................................................................................................................680 Videocapillaroscopy .............................................................................................................................................680 Capillaries Morphology .......................................................................................................................................681 Hypertension Assessment.....................................................................................................................................682 Venous Insufficiency ............................................................................................................................................682 Age-Related Changes of the Cutaneous Microcirculation..................................................................................683 Pharmacological Inhibition of the Dermal Microcirculation ..............................................................................683 Quantitative Assessment.......................................................................................................................................684 79.10.1 Image Processing Techniques................................................................................................................684 79.10.2 Capillaries Detection..............................................................................................................................684 79.10.3 Geometrical Capillary Network Analysis..............................................................................................684 79.11 Cosmetical Example.............................................................................................................................................685 79.11.1 Couperose and Erythrosis Assessment ..................................................................................................685 79.12 Conclusion............................................................................................................................................................685 References .......................................................................................................................................................................686
79.1 ANATOMY OF THE SKIN MICROCIRCULATION The skin microcirculation is organized as two horizontal plexuses, one located 1 to 1.5 mm below the skin surface, and the other at the dermal-subcutaneous junction that comprises collecting veins. It consists in arterioles, which may divide into many capillary loops at the level of the papillary layer of the skin. Indeed, arterial capillaries rise to form the dermal papillary loops at this level.1,2 Then, capillaries converge into collecting systems of the venous plexus. The vascular network of the skin varies considerably from one area to another.3,4
79.2 CAPILLAROSCOPY Among the different techniques available5 to study skin microcirculation (Table 79.1), human skin capillaroscopy, a specialized form of intravital microscopy, is the only method that allows direct visualization of the capillary network in vivo. Its principle is easy. After skin transparency has been enhanced by a drop of oil, an optical magnifying system allows visualization of its vascular network directly through the skin. The optical devices used to examine the cutaneous capillaries in vivo are most of the time the light microscope and stereomicroscope,5 but also the videocapillaroscope. 679
680
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 79.1 Non-Invasive Bioengineering Techniques Used to Study Skin Microcirculation Skin temperature measurements Capillaroscopy Dynamic capillaroscopy Dynamic capillaroscopy with dye Laser Doppler flowmetry Isotope techniques (133xenon) Transcutaneous measurement of partial oxygen pressure Capillary pressure Photopulse plethysmography Infrared thermography Colorimetry
79.3 PERIUNGUAL CAPILLAROSCOPY Nail-fold capillaroscopy (Figure 79.1) is usually performed in order to search some capillary deformations, characterizing the presence of a pathological situation.6,7 In this area, capillaries lie in a horizontal plane, so that a large part of their loops can be observed. Their aspect is particular. They look like hairpins with a diameter of 8 to 15 μm. Each loop is parallel to the other one, and oriented to the extremity of the finger (Figure 79.2A). Some particular aspects of the capillary loops allow to detect some systemic conditions, such as progressive systemic sclerosis and lupus erythematosus (Figure 79.2B). Nail-fold capillaroscopy has a good sensitivity and a good specificity. Therefore, morphological abnormalities must systematically be looked for on all accessible fingers.
79.4 VIDEOCAPILLAROSCOPY
(a)
(b)
FIGURE 79.2 Nail-fold capillaroscopy images. (a) Normal aspect. (b) Dilated capillaries with heterogenous distribution.
Contact videomicroscopy systems appeared recently in the industry for nondestructive control processes. They require epi-illumination of the skin surface8 and image transmission to a video camera via the optics of a microscope. The sensor located at one end of a flexible cord easily allows investigation on the whole tegument’s surface. Optical fibers convey illuminating light, which is provided by the handheld probe. The images are then visualized on a screen and picture digitalization is thus performed. Image numerization can be done directly during the examination; therefore, the data quality can be controlled instantly. The video imaging systems (Scopeman®, FORT®, Microvision®, Microwatcher) consist of a video signal control unit and a mini-CCD camera. A manually adjusted
Videocapillaroscopy devices (Figure 79.3) tend to replace the more conventional capillaroscopy instruments.
FIGURE 79.1 Capillaroscopy equipment for the nail-fold analysis.
FIGURE 79.3 Contact videocapillaroscopy with monitoring on TV.
Capillaroscopy and Videocapillaroscopy Assessment of Skin Microcirculation
681
TABLE 79.2 Different Architectural Frameworks of Skin Capillary Network Parallel arrangement and regular meshes network
Parallel arrangement with irregular meshes network
Perpendicular arrangement and regular dot line
Perpendicular arrangement and irregular dot line
Special pattern with parallel arrangement
Forehead Cheekbone region Cheek Chin Internal surface of arms
Trunk (anterior and posterior aspects) Breast Arms (external surface) Legs (internal and external surfaces)
Fingertip Eminentia tenar Eminencia hipotenar Tip of toes
Palm of hands Back of the hand and the foot Nipple
Finger nail fold Labial mucosa
From Miniati, B. et al., Ital. J. Anat. Embryol., 106, 233–238, 2001.
focusing system coupled with the camera head allows the obtaining of a sharp image of the capillary network. Magnification ranges from ×100 to ×1000. A magnification higher than ×600 enables visualization of blood cells inside the capillary. Venous congestion increases the number of capillaries detected.
79.5 CAPILLARIES MORPHOLOGY Capillaroscopy does not visualize the capillary wall, but only the blood red cells that cast the vessel. Thus, only functional capillaries can be detected. In most skin body
areas, the vascular morphology showed differences (Table 79.2). On the forehead and crow’s-foot, line and network forms dominate; on the dorsum of the hand, the dot and comma ruled5; while on the inner forearm both types were basically equal. A normal architectural framework shows two main patterns, a parallel and a perpendicular arrangement of capillary loops with respect to the skin surface (Figure 79.4). Capillary loops with a parallel arrangement form a vascular network with meshes that may or may not be regular. In most skin body areas9,10 capillaries are perpendicular to the skin surface, so that only the summit can be
A
B
C
D
FIGURE 79.4 Vascular morphology examples in different skin body areas: (A) scalp, (B) forehead, (C) cheek, (D) forearm.
682
Handbook of Non-Invasive Methods and the Skin, Second Edition
79.6 HYPERTENSION ASSESSMENT TABLE 79.3 Pathological or Cosmetical Usefulness of Videocapillaroscopy Peripheral arterial obliterative disorders12 Venous insufficiency13 Diabetes mellitus14 Hypertension15 Psoriasis16
Aging17
Specific morphological changes Rarefaction and dilatation of the capillary loops Tortuous and dilated capillaries Capillary rarefaction Grossly dilated and tortuous capillaries; many more perfused capillaries Reduced dermal papillary loops
Effects of topical cosmetics or chemical agents
seen. It looks like a point or a comma. The caliber of capillaries varies from 15 to 20 μm in regions with parallel arrangement of capillary loops. The capillary density ranges from 14 to 30 capillary loops per square millimeter in skin regions where capillary loops are arranged perpendicularly to the skin surface.4 The availability of videomicroscopes together with advances in our knowledge about the importance of microcirculation in the pathogenesis of trophic complications of arterial and venous insufficiencies explains the present development of capillaroscopy on the skin. The determination of morphological or dynamic changes in the microcirculation belongs to the non-invasive techniques of the biometrological domain. In pathology, numerous conditions can be better examined with videocapillaroscopy (Table 79.3): peripheral arterial obliterative disorders, venous disorders such as venous insufficiency, diabetes mellitus, hypertension, couperosa (Figure 79.5A), and psoriasis (Figure 79.5B).
(a)
Capillary network analysis on the volar aspect of the forearm or on the fingers could be of interest in hypertension. Indeed, it was suggested that microvascular rarefaction represents an important mechanism in primary hypertension. Capillary rarefaction has been described in various tissues from patients with essential hypertension. The introduction of intravital videomicroscopy allowed the discovery of a 15 to 20% reduction in the capillary density of the nail-fold skin, and intravital fluorescein angiography found a 20% reduction in capillary density in the forearm skin of hypertensive subjects compared with normotensive subjects.11 Some other conditions, such as peripheral arterial obliterative disorders, venous disorders like venous insufficiency, diabetes mellitus, and psoriasis (Table 79.3), are candidates for microcirculatory evaluation.
79.7 VENOUS INSUFFICIENCY Videocapillaroscopy is mainly performed on the instep in venous insufficiency, and on the back of the foot (first intermetatarsian space and toes) in arterial insufficiency. Venous insufficiency is characterized by the decrease of the capillaries density, the widening of the dermal papilla, the size of which becomes heterogeneous, and the contours marked by the hemosiderine deposits of ochre dermatitis. The decrease of the capillaries density is partially compensated by their increased length, inducing an increased number of meanders, which can even take the shape of a glomerular cluster in the most severe forms (lipodermatosclerosis, white atrophy). As soon as the first trophic troubles occur, the venules are not visible any longer.18–20
(b)
FIGURE 79.5 Peripheral arterial obliterative disorders. (a) Couperosa (×50 = 27 mm2). (b) Psoriasis with dilated and tortuous capillaries (×200 = 1.73 mm2).
Capillaroscopy and Videocapillaroscopy Assessment of Skin Microcirculation
683
TABLE 79.4 Number of Capillary Loops in Five Age Groups on Different Examined Sites (n = 20) Sites
Group 1
Group 2
Group 3
Group 4
Group 5
Age range (years) (Mean age) Hand Arm Postauricular
20–29 (25.3) 48.5 23.3 12.6
30–39 (34.5) 41.4 23.1 16.5
40–49 (44.6) 49 26.8 13.3
50–59 (53.5) 38 26.8 15.5
60–69 (64.4) 28.5 20.1 2.6
Source: Zhu, W., Assessment of Cutaneous Microvasculature in Aging and Photoaging: A Videocapillaroscopic Study on Caucasian Women, personal data.
Fagrell has described three classes of increasing severity and validated their discriminating value for the local trophic prognosis:21,22 A, capillary dilatation; B, edema impairing capillary visualization; and C, absence of visible capillary (prenecrosis stage).
79.8 AGE-RELATED CHANGES OF THE CUTANEOUS MICROCIRCULATION A significant decrease of cutaneous capillary loop density is observed with age (Table 79.4). The microvasculature is regularly formed in young skin, with many orderly arranged capillary loops (dots) and some horizontal vessels (lines) (Figure 79.6A).23 It becomes thicker, twisted, and irregular in older skins (Figure 79.6B), with horizontal vessels that appear tortuous, elongated, disorganized, and dilated (W. Zhu, Assessment of Cutaneous Microvasculature in Aging and Photoaging: A Videocapillaroscopic Study on Caucasian Women, personal data). Thus, the parallel vasculature can be more easily observed with aging. This result accords with that of biopsy specimens
(a)
observed by light or electric microscope. As the epidermis becomes thinner, the transparency of the skin increases, facilitating the observation of the papillary vascular plexus. And as the vasculature expands and thickens, some of the microvasculature usually difficult to observe and the deeper vasculature can also be examined. Results also indicate that photodamaged vessels correspond to an increased formation of new vessels.24
79.9 PHARMACOLOGICAL INHIBITION OF THE DERMAL MICROCIRCULATION A pharmacological agent such as neosynephrin is able to induce dermal capillary density reduction after topical application.25,26 In this case, videocapillaroscopy is a reliable tool to quantify in vivo the capillary loop density in the dermis before and after the application of the solution. This pharmacological model could be useful to evaluate the effects of different cosmetic or drug agents able to inhibit the dermal adrenergic response in dermis.
(b)
FIGURE 79.6 Capillary patterns observed by videocapillaroscopy with young (a) and elderly subject (b) on the hand.
684
Handbook of Non-Invasive Methods and the Skin, Second Edition
79.10 QUANTITATIVE ASSESSMENT
A
Videocapillaroscopy is mainly used for the morphological information it provides, and it is different from indirect techniques (laser Doppler, transcutaneous PO2) also because it is free from interpretation artifacts, a considerable advantage in a field as complex as microcirculation. However, the absence of quantification and the subjective and operator-dependent character of capillaroscopy show its limits.
79.10.1 IMAGE PROCESSING TECHNIQUES Quantitative capillaroscopy is now possible due to the development of computerized systems, and series of advanced image processing methods have been developed.
79.10.2 CAPILLARIES DETECTION Finding capillary loops automatically in an image is a difficult yet important first step in order to achieve microcirculation analysis. An automatic counting of capillaries is now available27,28 with a detection rate of 82%. This detection system was based on a principal component analysis (PCA) and was associated with a retinally connected neural network. This filter system is capable of real-time processing, recognizes capillaries anywhere in an image, and operates successfully under a wide range of lighting and noisy conditions. It is assumed that the study of microcirculation must include all dynamic and cooperative processes between the capillaries. Indeed, the geometric complexity of the capillary structure can lead to heterogeneities in oxygen delivery. If a finite tissue element is not within the proximity of a capillary site, necrosis could develop in that tissue area.
79.10.3 GEOMETRICAL CAPILLARY NETWORK ANALYSIS For characterizing capillary ensembles, the statistical and geometrical properties of the network need to be explored.29 In fact, this network forms a map that is based on adjacent relationships among the capillaries. These relationships30,31 can be quantified by a distance parameter (distance between one capillary and all its neighbors) and a surface parameter (an influence area corresponding to a surface around the capillary). Zhong et al.29 described for the first time a method to construct the capillary network by using image processing with Delaunay triangulation and a Voronoï diagram. Delaunay triangulation32 was implemented to obtain the nearest neighbor for each capillary (Figure 79.7A). This representation permitted calculation of the minimal, maximal, or mean distance between capillaries. The Voronoi diagram33 was used to determine a surface parameter that
B
FIGURE 79.7 Delaunay triangulation (A) and Voronoï diagrams (B): 67 capilllaries, on the scalp (×200 = 1.73 mm2).
could be considered an oxygen diffusion area (Figure 79.7B). For each parameter, the distribution (thresholding or normalization) was analyzed in order to eliminate extreme values, which correspond to artifacts.34 We noticed that without these values, the proximity parameters distribution was normal. These algorithms have been implemented in a software platform (Capilab Toolbox®) providing functions and interactive tools for analyzing pictures of the skin. This graphical environment integrates complete statistical and geometrical computing, visualization, and image enhancement algorithms. The analysis could be improved by taking into account morphologic parameters (shape, roundness) of each capillary (Figure 79.8). Indeed, we do not know whether a detected capillary is physiologically active, or if an interaction exists between patterns with different shapes or sizes. The mathematical morphology could be a mean to focus the analysis on a specific group of capillaries, with the same properties, or to construct a network with shapebased relations.
Capillaroscopy and Videocapillaroscopy Assessment of Skin Microcirculation
A
B
C
D
685
FIGURE 79.8 Intermediate images showing the morphologic image processing technique.
79.11 COSMETICAL EXAMPLE 79.11.1 COUPEROSE
AND
ERYTHROSIS ASSESSMENT
Rosacea is a frequent disease that occurs mostly in women. Rosacea is heralded, around the age of 20 years, by intermittent facial erythema and by the gradual development of permanent erythema (erythrosis) with telangiectasia (couperose). Image processing techniques based on neural network algorithms can be used again to detect and quantify by a surface parameter the erythema in the pictures. A color neural network-based method was developed (based on an RGB and a lab code) that is capable of real-time processings, increasing the quality of videocapillaroscope images and minimizing the disturbance of artifacts.35 This system can selectively recognize regions where the couperose is heavy or light. The algorithms (implemented in Capilab Toolbox) can process any kind of color images of the skin (Figure 79.9).
79.12 CONCLUSION The determination of morphological or dynamic changes in the cutaneous microcirculation belongs to the non-invasive techniques of the biometrological domain. Every capillary modification due to topical cosmetic products or chemical agents can then be observed. In pathology,
numerous conditions can be better examined with this system. Capillaroscopy is now used routinely in different hospitals and companies. Although its employment requires some experience, this technique brings direct information on the capillary network morphology. It differs from the indirect methods used to explore microcirculation, such as laser Doppler or transcutaneous partial oxygen. The techniques to visualize skin capillaries enable getting an irreplaceable approach of the physiology and physiopathology of the skin capillary circulation. Compared to other heavy research methods, traditional capillaroscopy techniques, much simpler and cheaper, have shown their usefulness in the detection of connectivity microangiopathies and vascular acrosyndromes. Combined with the potential of numerical image analysis, they will probably extend their application fields to the assessment of the influence of arterial and venous diseases on the skin’s nutritional circulation. Capiflow®, Capiflow AB, Kista, Sweden FORT® Imaging Systems, Curno BG, Italy Scopeman ® Moritex, P.M. Industries Ltd., Cambridge, U.K. Microvision® MV 2100, Finlay Microvision Co. Ltd., Warwickshire, U.K.
686
Handbook of Non-Invasive Methods and the Skin, Second Edition
A
B
C
D
FIGURE 79.9 Examples of couperose image processing.
Microwatcher Model VS-10®, Mitsubishi Kasei Corp., Tokyo, Japan Capilab Toolbox®, LIBC, Besançon, France
REFERENCES 1. Rhodin JAG. Anatomy of the microcirculation. In Microcirculation: Current Physiologic, Medical and Surgical Concepts, Effros RM, Schmid-Schöbein H, Ditzel J, Eds. Academic Press, New York, 1981, pp. 11–7. 2. Braverman IM. The cutaneous microcirculation. J Invest Dermatol 5: 3–9, 2000. 3. Bongard O, Bounameaux H. Clinical investigation of skin microcirculation. Dermatology 186: 6–11, 1993. 4. Miniati B, Macchi C, Molino Lova R, et al. Descriptive and morphometric anatomy of the architectural framework of microcirculation: a videocapillaroscopic study on healthy adult subjects. Ital J Anat Embryol 106: 233–238, 2001. 5. Carpentier PH. Méthodes d’exploration vasculaire chez l’homme: microcirculation et veines. Therapie 54: 369–374, 1999. 6. Hu Q, Mahler F. New system for image analysis in nailfold capillaroscopy. Microcirculation 6: 227–235, 1999.
7. Allen PD, Taylor CJ, Herrick AL, Moore T. Image Analysis of Nailfold Capillary Patterns. Medical Image Understanding and Analysis 1998. http://www.isbe.man.ac.uk/~pa/NailFold.html. 8. Jairo J, Monari M. Human capillaroscopy by light emitting diode epi-illumination. Microvasc Res 59: 172–175, 2000. 9. Li L, Mac-Mary S, Sainthillier JM, et al. Cutaneous facial vascular network cartography. Ann Dermatol Venereol 129: 1172, 2002 (abstract). 10. Li L, Mac-Mary S, Sainthillier JM, et al. Vascular network cartography of the body skin. Ann Dermatol Venereol 129: 1173, 2002 (abstract). 11. Prasad A, Dunnill GS, Mortimer PS, MacGregor GA. Capillary rarefaction in the forearm skin in essential hypertension. J Hypertens 13: 265–268, 1995. 12. Hern S, Mortimer PS. Visualization of dermal blood vessels: capillaroscopy. Clin Exp Dermatol 24: 473–478, 1999. 13. Fagrell B. Advances in microcirculation network evaluation: an update. Int J Microcirc 15 (Suppl. 1): 34–40, 1995. 14. Chang CH, Tsai RK, Wu WC, et al. Use of dynamic capillaroscopy for studying cutaneous microcirculation in patients with diabetes mellitus. Microvasc Res 53: 121–127, 1997. 15. Serne EH, Gans RO, ter Maaten JC, et al. Impaired skin capillary recruitment in essential hypertension is caused by both functional and structural capillary rarefaction. Hypertension 38: 238–242, 2001.
Capillaroscopy and Videocapillaroscopy Assessment of Skin Microcirculation
16. Bull RH, Bates DO, Mortimer PS. Intravital video-capillaroscopy for the study of the microcirculation in psoriasis. Br J Dermatol 126: 436–445, 1992. 17. Kelly RI, Pearse R, Bull RH, et al. The effects of aging on the cutaneous microvasculature. J Am Acad Dermatol 33: 749–756, 1995. 18. Fagrell B. Vital microscopy and the pathophysiology of deep venous insufficiency. Int Angiol 14: 18–22, 1995. 19. Franzeck UK, Bollinger A, Huch R, Huch A. Transcutaneous oxygen tension and capillary morphologic characteristics and density in patients with chronic venous incompetence. Circulation 70: 806–811, 1984. 20. Stucker M, Schobe MC, Hoffmann K, Schultz-Ehrenburg U. Cutaneous microcirculation in skin lesions associated with chronic venous insufficiency. Dermatol Surg 21: 877–882, 1995. 21. Fagrell B, Hermansson IL, Karlander SG, Ostergren J. Vital capillary microscopy for assessment of skin viability and microangiopathy in patients with diabetes mellitus. Acta Med Scand, Suppl. 687: 25–28, 1984. 22. Bollinger A, Fagrell B. Clinical Capillaroscopy. Hogrefe et Huber, Toronto, 1990. 23. Li L, Mary S, Sainthillier JM, Degouy A, Gharbi T, De Lacharriere O, Humbert P. Changes of cutaneous microcirculation in the different anatomic sites with aging in women. China J Microcirc 2: 43–45, 2004. 24. Toyoda M, Nakamura M, Luo Y, et al. Ultrastructural characterization of microvasculature in photoaging. J Dermatol Sci 27: S32–S41, 2001. 25. Degouy A, Creidi P, Sainthillier JM, et al. In vivo transcutaneous capillaroscopy: assessment of dermal capillary density decrease after topical pharmacological agent applications. Ann Dermatol Venereol 129: 414, 2002 (abstract).
687
26. Sainthillier JM, Creidi P, Degouy A, et al. Topical application of a manganese gluconate preparation inhibits the effects of neosynephrin on the cutaneous microcirculation. Ann Dermatol Venereol 129: 459, 2002 (abstract). 27. Sainthillier JM, Bonnans V, Degouy A, et al. Application des réseaux neuronaux pour le traitement et l’analyse des images en bio-ingénierie cutanée. In Actualités en Ingénierie Cutanée, Vol. 3, Perrenoud D, Gabard B, Eds. Eska, Paris, 2003, pp. 117–124. 28. Sainthillier JM, Gharbi T, Muret P, Humbert P. Pattern recognition of skin capillary network by means of neural algorithms. Skin Res Technol, in press. 29. Zhong J, Asker CL, Salerud EG. Imaging, image processing and pattern analysis of skin capillary ensembles. Skin Res Technol 6: 45–57, 2000. 30. Robert JM, Toussaint GT. Computational geometry and facility location. In Proceedings of the International Conference on Operations Research and Management Science, Manila, The Philippines, 1990, pp. B1–B19. 31. De Berg M, Kreweld M, Overmars M, Schwarzkopf O. Computational Geometry: Algorithms and Applications. Springer-Verlag, Berlin, 2000. 32. Fortune S. Voronoi diagrams and Delaunay triangulations. In Computing in Euclidean Geometry, Lecture Notes Series on Computing, Vol. 4, Ding-Zhu D, Hwang F, Eds. World Scientific, Singapore, 1995, pp. 225–265. 33. Fortune S. Sweepline algorithms for Voronoi diagrams. Algorithma 2: 153–174, 1987. 34. Sainthillier JM, Degouy A, Gharbi T, et al. Geometrical capillary network analysis. Skin Res Technol 9: 312–320, 2003. 35. Degouy A, Sainthillier JM, Mac-Mary S, et al. Evaluation de l’activité clinique et vidéocapillaroscopique d’une formulation cosmétique sur la microcirculation cutanée des couperoses légères à modérées. Nouv Dermatol 22: 554–556, 2003.
Blood Flow, Vasomotion, and Vascular Functions
Doppler Measurement of Skin 80 Laser Blood Flux: Variation and Validation Andreas J. Bircher Department of Dermatology, University Hospital, Basel, Switzerland
CONTENTS 80.1 Introduction............................................................................................................................................................691 80.2 Anatomical and Physiological Factors..................................................................................................................692 80.2.1 Skin Microvasculature ...............................................................................................................................692 80.2.2 Cutaneous Blood Flow ..............................................................................................................................692 80.3 Methodological Principles: Technical Aspects .....................................................................................................692 80.4 Factors Influencing Cutaneous Blood Flow..........................................................................................................693 80.4.1 Subject-Related Variables ..........................................................................................................................693 80.4.2 Instrument-Related Variables.....................................................................................................................693 80.5 Correlation with Other Methods ...........................................................................................................................694 80.5.1 133Xenon Clearance....................................................................................................................................694 80.5.2 Venous Occlusion Plethysmography .........................................................................................................694 80.5.3 Photoplethysmography ..............................................................................................................................695 80.5.4 Thermographic Methods............................................................................................................................695 80.5.5 Other Lasers...............................................................................................................................................695 80.6 Conclusions............................................................................................................................................................695 References .......................................................................................................................................................................695
80.1 INTRODUCTION The cutaneous microcirculation plays an outstanding role in physiologic processes and is also involved in many pathologic conditions. Therefore, there is considerable interest to objectivate changes in the cutaneous blood flow (CBF) in physiologic conditions and pathologic disorders as well as to pharmacological stimuli. Blood flow has been measured on virtually all animal and human organ surfaces, such as the nasal and gastrointestinal mucosa, in the eye, and on bone and muscular tissue. The skin, however, remains the most readily accessible organ, and therefore a vast amount of literature on numerous aspects of CBF has been published in the last years. Several methods to identify and quantitate CBF fluctuations have been developed. Among these are more direct methods such as photoplethysmography, venous occlusion plethysmography, and 133xenon (133Xe) clearance, as well as indirect techniques such as the determination of skin temperature and transcutaneous pO2. A newer noninvasive technology is laser Doppler velocimetry or flowmetry (LDF), which is based on optical princi-
ples and which allows a more direct measurement of the cutaneous microcirculation. LDF has some advantages over the other methods in that it allows noninvasive, continuous recording of CBF in relatively superficial skin layers on virtually any reachable skin or mucous membrane surface. The application of the Doppler phenomenon and the use of the unique properties of laser light to detect the motion of macromolecules were suggested by Cummins et al.21 and first used in a biological system by Riva et al.22 Later the approach was improved and applied to human retinal vessels by Tanaka et al.23 Stern et al.9 was the first to determine blood flow in the intact cutaneous microvasculature of humans.1 Several instruments that were usable also in clinical settings were then developed. The first, the laser Doppler velocimeter, was designed and later improved by Holloway and Watkins.2 Subsequently, a modified version, the LDF, was developed by Nilsson and colleagues.3 Both instruments use basically the same principles based on light-bearing spectroscopy with some technical variations to improve the yield of the measurements. Detailed descriptions of the theoretical aspects of 691
692
Handbook of Non-Invasive Methods and the Skin, Second Edition
LDF and the available instruments that have later been designed have been published.1
80.2 ANATOMICAL AND PHYSIOLOGICAL FACTORS 80.2.1 SKIN MICROVASCULATURE The human epidermis contains no vessels and is therefore nourished by diffusion from the dermis. Its thickness, which is dependent on the anatomical site, varies from 50 to 200 mm. In the dermis, supplied by vessels from the subcutaneous tissue, there is a complex vascular network to perform the particular tasks, such as thermoregulation, nutrition, and metabolism. In the upper dermis a superficial plexus is present with a special capillary network extending into the dermal papillae, the so-called capillary loops. They have a mean length of approximately 0.2 to 0.4 mm and each supplies an average of 0.04 to 0.27 mm2 to the skin surface. A deeper plexus is situated at the dermal-subcutaneous border. The microcirculatory blood flow is regulated by smooth muscle cells that are mainly located in the walls of the small arteries and arterioles and, to a lesser extent, in venules and small veins. The true capillaries are free of smooth muscle cells; however, the pericyte cells located in the capillary vessel walls may induce to some extent capillary contractions. Arteriovenous anastomoses are present in some locations, particularly in the face and in the acral areas. They play an outstanding role in the thermoregulation of the organism.4
80.2.2 CUTANEOUS BLOOD FLOW Blood flow or flux implies the movement of blood, which is a heterogeneic mixture of liquid, soluble, and cellular elements, through the preformed channels of the arterial and venous system, which is linked by the above-mentioned capillary network or by arteriovenous anastomoses. Blood flow is far from being constant; on the contrary, it is dynamically regulated to fulfill the requirements of the organism. Dependent on the measuring technique, its resolution, and the investigated tissue volume, any method of measurement expresses blood flow only in relation to the type and localization of the examined vessels. In optical methods, further variables that influence the measurements are the structure of the skin surface, the skin thickness, and the pigmentation of the epidermis. Due to the small measuring area and the anatomical architecture of the dermal microvasculature, considerable variations in blood flow are present.5 Therefore, the regional variations of CBF have a considerable magnitude and have to be taken into account when using such measuring methods. Most of the current laser Doppler devices use helium–neon lasers with a wavelength l = 632.8 nm and small probes. It is estimated that the tissue volume in
which CBF is measured with such devices has an approximate surface area of 1 mm2, extending to an estimated depth of 1 mm, resulting in a theoretical total measured tissue volume of 1 mm3. The penetration depth of laser radiation, however, is rather variable and dependent on the above-mentioned factors. The incident laser light is absorbed, scattered, and only to a small extent reflected by the skin tissue structures. Stationary tissue, such as fibers, macromolecules, and vessel walls, scatter and reflect the incident radiation at the same frequency. Typically the largest fraction of the diffusely reflected laser light waves comes from stationary tissues. Tissue components such as red blood cells moving with a mean estimated speed of 1 mm/sec reflect the light with a shifted frequency (optical Doppler effect). The blood flow signal measured by laser Doppler instruments is an indicator of the cutaneous perfusion; however, due to the complex structure of the skin texture and somewhat random orientation of the cutaneous vessels, only semiquantitative, relative measurements of CBF can be made6 with these techniques. Commonly the CBF results are given in millivolts or in arbitrary units. Laser Doppler measurements of CBF from normal relaxed subjects have three major characteristics (Figure 80.1): (A) pulsatile flow synchronized with the cardiac cycle, (B) vasomotor waves of lower frequency (approximately 4 to 6 per minute), and (C) skin blood flow, i.e., the deviation from the instrument baseline.7
80.3 METHODOLOGICAL PRINCIPLES: TECHNICAL ASPECTS In laser Doppler measurements, a laser light source and the Doppler effect are used to generate an output proportional to the red blood cell movement through the skin vessels under investigation. Prerequisites for this technique are the characteristics of laser radiation, such as mV 300
A
B
C
250 200 150 100 50 0
FIGURE 80.1 Schematic response pattern in LDF measured blood flow: (A) pulsatile flow synchronized with the cardiac cycle, (B) vasomotor waves of lower frequency, and (C) relative blood flow, i.e., the deviation from the instrument baseline. (Redrawn from Karanfilian, R. et al., Am. Surg., 50, 641, 1984. With permission.)
Laser Doppler Measurement of Skin Blood Flux: Variation and Validation
693
80.4 FACTORS INFLUENCING CUTANEOUS BLOOD FLOW 80.4.1 SUBJECT-RELATED VARIABLES
A
B
C
FIGURE 80.2 Schematic presentation of laser-light interaction with tissue components. Transmitted light is reflected and shifted in frequency by (A) moving erythrocytes or (B) reflected without frequency shift from stationary tissue or (C) does not reach the receiving probe fiber.
monochromaticity and spatial and temporal coherence. In typical laser Doppler instruments, a 2- or 5-mW-powered helium–neon laser (l = 632.8 nm) is usually the light source. Other instruments make use of an infrared laser (l = 780 nm). The laser radiation is guided through an optical fiber to the probe head. The probes are usually held in contact with the skin by a double-sided adhesive ringshaped tape. The emitted radiation enters the skin and is reflected by stationary and moving tissue components (Figure 80.2). Stationary tissue scatters and reflects the incident radiation at the same frequency. Red blood cells moving with a certain speed reflect the light with a shifted frequency, the Doppler effect. Thus, the light waves returning to the instrument are composed of two components: the frequency-modulated spectrally broadened light, which is directly related to the number of erythrocytes times their velocity, and the nonshifted fraction, which has been reflected from nonmoving tissue. The reflected light is transmitted through one or more receiving optical fibers, and the nonshifted reference and the Doppler-shifted beam are then mixed on a photodetector and processed by optical heterodyning. The generated beat frequency is then converted into an electrical output. The signal is usually measured as a fluctuating voltage, expressed in millivolts. This flow signal measured by laser Doppler instruments is an indicator of cutaneous perfusion; however, due to the complex structure of the skin texture and the somewhat random orientation of the cutaneous microcirculation, only semiquantitative relative measurements of CBF are obtained.1
Recently, a position paper on guidelines for LDF has been published by the standardization group of the European Contact Dermatitis Society.8 In this paper variables influencing CBF have been extensively reviewed. A broad spectrum of aspects may influence the measurements of CBF. Such variables include individual, interindividual, environmental, and technical factors. The latter variables are discussed in Bircher et al.8 Individualrelated variables include age, gender, and ethnic background. Selectable individual-related variables are the anatomical location of the measurements, the skin temperature, the subject’s position, physical and mental activities, previous consumption of food, beverages, drugs, and nicotine, the menstrual cycle, possibly some laboratory values, and to some extent the considerable temporal variation of repeated measurements. Environmental variables include air convection, ambient temperature, and probably air humidity (Table 80.1).
80.4.2 INSTRUMENT-RELATED VARIABLES In the above-mentioned position paper guidelines for LDF have also been published.8 Therefore, these guidelines will only be briefly summarized here. Due to the wide range of instruments available that work with several techniques and principles, a universal standard procedure cannot be developed. However, some general recommendations for validation can be given. The instrument should be used in accordance with the recommendations of the manufacturer, particularly with regard to instrument setup and safety precautions. The operating procedure of every laboratory should be clearly stated, and a standardized procedure that includes validation under the laboratory’s own specific conditions is proposed. Some recommendations concerning the adjustments of the zero levels are briefly outlined. The instrument zero is the value obtained when the probe is held against a white surface, e.g., white porcelain. It may slightly deviate from the electrical zero of the instrument. The biological zero, which is higher than the instrumental zero, is obtained in an anatomical site, e.g., the forearm, under arterial occlusion. In the same procedure the dynamic range of the blood flow can be measured. This basic standard reactive hyperemia experiment should be performed in at least three healthy adults. These recommendations should facilitate the comparison of measurements obtained at different research facilities with different laser Doppler instruments. Naturally, the individual and environmental variables described should be controlled and mentioned. Subjects should be managed with regard to smoking, food and drug
694
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 80.1 Individual and Environmental Variables Affecting Cutaneous Blood Flow Variable Age Gender Menstrual cycle Ethnic background Subject position Anatomical location Skin temperature Laboratory hematologic values Temporal (same day) Temporal (day to day) Physical activity Mental activity Food and beverages Systemic drugs Topical drugs Nicotine Ambient temperature
Influence on Blood Flow Mostly minor Minor or none Minor or none None or minor Major Considerable variation Major None or minor Minor Considerable Considerable Considerable Considerable Considerable Major Major Major
Remarks Dependent on location — — Reflection of pigmented skin Orthostatic dependence Also age related Also environment related Only when major pathology is present — — Short influence Short influence e.g., caffeine, alcohol Vasoactive compounds Vasoactive compounds Vasoconstriction —
Modified from Bircher, A.J. et al., Contact Dermatitis, 30, 65, 1994. With permission.
intake, and physical and mental stress. When small probes are used, several blood flow determinations, which may be averaged, are recommended. The study environment should be controlled with respect to temperature, air movements, and other factors. Instrumental factors to be controlled include warming up of the instrument and particular precautions with the optical fiber and the application pressure of the probe.
80.5 CORRELATION WITH OTHER METHODS 80.5.1 133
133 ENON
X
CLEARANCE
Xe clearance was compared to LDF in ultraviolet-stimulated CBF in forearm skin.2,9 In both studies 133Xe was delivered by injection. To minimize the spatial and temporal variation in LDF measurements, the average of 5 and 10 measuring sites, respectively, was determined. In these experiments a linear relationship (r = 0.9) between the two methods was found. In a study using an atraumatic delivery technique for 133Xe in four subjects, fingertip and finger web blood flow was examined.10 With both methods, parallel changes of CBF in skin without arteriovenous anastomoses were observed; however, in areas with shunt vessels, no relation was present. Also, in healthy skin and in uninvolved skin of psoriatic patients11 a poor correlation between LDF and xenon washout was observed. A better correlation was obtained in psoriatic lesions, leading to the conclusion that although LDF allows a rough estimate of CBF at high perfusion rates, it is not reliable in areas
with low blood flow. A comparison of LDF and 133Xe washout in skin lesions of localized and generalized morphea showed a good correlation between the methods. The sclerotic plaques and the perilesional inflammatory lilac rings had a significantly greater CBF than normal skin. Older plaques had higher CBF than early lesions. However, in the most advanced burnt out plaques, CBF was not different from that in normal skin.12 Very similar results were obtained in another study of scleroderma.13 Again, in ultraviolet B-induced erythema of the forearm a significant correlation between 133Xe washout and LDF measurements of CBF was found. Surprisingly, LDF measurements showed a fourfold greater increase over 133Xe determinations. The importance of the use of the biological zero instead of the instrumental electrical zero was also emphasized.
80.5.2 VENOUS OCCLUSION PLETHYSMOGRAPHY Venous occlusion or strain gauge plethysmography is an established method to determine total blood flow, particularly in extremities. In thermally stimulated CBF by direct heating of the whole body surface14 a good correlation between CBF measured by LDF and total forearm blood flow determined by occlusion plethysmography (r = 0.94 to 0.98) was found. However, the correlation varied considerably between and within subjects. A nonlinear relationship between occlusion plethysmography and LDFCBF measurements, stimulated by exercise and heat stress, was reported, suggesting that the two methods measure different blood flow parameters.15 In a comparison of
Laser Doppler Measurement of Skin Blood Flux: Variation and Validation
LDF, occlusion plethysmography, and thermal clearance, considerable variability in the results was found, which was explained on the basis of the different skin volumes and types of blood flow measured by the methods.16
80.5.3 PHOTOPLETHYSMOGRAPHY The resting blood flow levels as a function of the anatomic position were determined in 52 body sites in 10 subjects17 by photoplethysmography and LDF. Photoplethysmographic and LDF measurements agreed well in the areas with low CBF, but differed considerably in highly perfused areas. This was interpreted to be the consequence of the different parameters measured by the two methods (blood volume in photoplethysmography, volume times velocity in LDF).
80.5.4 THERMOGRAPHIC METHODS Several thermal methods that have been compared to LDF were found to give acceptable correlations. Thermal measurements, however, were usually slower in the speed of the response and more dependent on ambient temperature.18 Also, upon an intravenous injection of naftidrofuryl, similar blood flow values, as measured by thermal conductivity and LDF, were observed. The skin temperature also decreased over time, however, with a temporally delayed response.19
80.5.5 OTHER LASERS In a study investigating histamine-induced CBF a helium–neon (l = 633 nm) and an infrared (l = 780 nm) laser were compared.20 Very similar results of CBF changes were found with the two lasers, although the infrared laser radiation has a greater penetration. This implies that probably only superficial vessels were mainly affected by the histamine effect.
80.6 CONCLUSIONS Laser Doppler flowmetry requires a sophisticated technology; however, the instrument’s handling is simple. It allows noninvasive measurements of skin blood flow on virtually any area of the body and provides a continuous recording of the actual, mostly superficial, blood flow. It is particularly suited to measure and follow stimulated blood flow. Major fields of interest and application have been the determination of the qualitative and quantitative effects of systemically and topically applied vasoactive drugs, the study of inflammatory mediators, the investigation of allergic and irritant skin reactions, particularly in combination with other noninvasive bioengineering methods, and the evaluation of vascular phenomena in skin and other diseases, and also effects of therapeutic interventions on skin disorders.
695
Because the instrument is easy to use, some precautions should be taken when performing measurements. One major limitation is the rather small measured tissue volume, which may be overcome by future technical improvements, e.g., of scanning probes, as described in the next chapter. A variety of factors, including the instrument, the subject, and the environment, influence the measurements. Because of the high variability of skin blood flow, such as inter- and intraindividual subject variability and environmental parameters, these factors have to be taken into consideration in the planning and realization of experiments.
REFERENCES 1. Shepherd, A. and Öberg, P., Laser-Doppler Blood Flowmetry, Kluwer Academic Publishers, Boston, 1990. 2. Holloway, G. and Watkins, D., Laser Doppler measurement of cutaneous blood flow, J. Invest. Dermatol., 69, 306, 1977. 3. Nilsson, G., Tenland, T., and Öberg, P., A new instrument for continuous measurement of tissue blood flow by light bearing spectroscopy, IEEE Trans. Biomed. Eng., 27, 12, 1980. 4. Tenland, T., On Laser Doppler Flowmetry. Thesis, Linköping University, Linköping, Sweden, 1982. 5. Braverman, I., Keh, A., and Goldminz, D., Correlation of laser Doppler wave patterns with underlying microvascular anatomy, J. Invest. Dermatol., 95, 283, 1990. 6. Holloway, G., Laser Doppler measurement of cutaneous blood flow, in Non-invasive Physiological Measurements, Rolfe, P., Ed., Academic Press, New York, 1983, p. 219. 7. Karanfilian, R., Lynch, T., Lee, B., Long, J., and Hobson, I.R., The assessment of skin blood flow in peripheral vascular disease by laser Doppler velocimetry, Am. Surg., 50, 641, 1984. 8. Bircher, A.J., de Boer, E., Agner, T., Wahlberg, J., and Serup, J., Guidelines for the measurement of cutaneous blood flow by laser Doppler flowmetry, Contact Derm., in press. 9. Stern, M., Lappe, D., Bowen, P., Chimosky, J., Holloway, G., Keiser, H., and Bowman, R., Continuous measurement of tissue blood flow by laser Doppler spectroscopy, Am. J. Physiol., 232, H441, 1977. 10. Engelhart, M. and Kristensen, J., Evaluation of cutaneous blood flow responses by 133 xenon washout and laser Doppler flowmeter, J. Invest. Dermatol., 80, 12, 1983. 11. Klemp, P. and Staberg, B., The effect of antipsoriatic treatment on cutaneous blood flow in psoriasis measured by 133xenon washout method and laser Doppler velocimetry, J. Invest. Dermatol., 85, 259, 1985. 12. Serup, J. and Kristensen, J., Blood flow of morphoea plaques as measured by laser Doppler flowmetry, Arch. Dermatol. Res., 276, 322, 1984.
696
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. De Lacharrière, O. and Kalis, B., Measurement of cutaneous microcirculation in dermatology and dermatopharmacology, in Cutaneous Investigation in Health and Disease, Leveque, J., Ed., Marcel Dekker, New York, 1989, pp. 385–420. 14. Johnson, J., Taylor, W., Shepherd, A., and Park, M., Laser Doppler measurement of skin blood flow, comparison with plethysmography, J. Appl. Physiol., 56, 798, 1984. 15. Smolander, J. and Kolari, P., Laser Doppler and plethysmographic skin blood flow during exercise and acute heat stress in the sauna, Eur. J. Appl. Physiol., 54, 371, 1985. 16. Saumet, J., Dittmar, A., and Leftheriotis, G., Non-invasive measurement of skin blood flow: comparison between plethysmography, laser-Doppler flowmeter and heat thermal clearance method, Int. J. Microcirc. Clin. Exp., 5, 73, 1986.
17. Tur, E., Tur, M., Maibach, H.I., and Guy, R.H., Basal perfusion of the cutaneous microcirculation: measurements as a function of anatomic position, J. Invest. Dermatol., 81, 442, 1983. 18. Johnson, J., The cutaneous circulation, in Laser-Doppler Blood Flowmetry, Shepherd, A. and Öberg, P., Eds., Kluwer Academic Publishers, Boston, 1990, pp. 121–139. 19. Dittmar, A., Skin thermal conductivity, in Cutaneous Investigation in Health and Disease, Leveque, J., Ed., Marcel Dekker, New York, 1989, pp. 323–358. 20. Coulsen, M., Hayes, N., and Foreman, J., Comparison of infrared and helium-neon lasers in the measurement of blood flow in human skin by the laser Doppler technique, Skin Pharmacol., 5, 81, 1992. 21. Cummins et al. 22. Riva et al. 23. Tanaka et al.
of Periodic Fluctuations 81 Examination in Cutaneous Blood Flow Robert Gniadecki, Monika Gniadecka, and Jørgen Serup Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 81.1 Physiology of Vasomotion.....................................................................................................................................697 81.2 Methods of Detection and Analysis of Vasomotion .............................................................................................698 81.2.1 Intravital Microscopy.................................................................................................................................698 81.2.2 Laser Doppler Technique ..........................................................................................................................699 81.2.3 Analysis of Waveform Patterns .................................................................................................................700 81.2.3.1 Fast Fourier Transform...............................................................................................................700 81.2.3.2 Prony Spectral Line Estimation .................................................................................................702 81.2.3.3 Autoregressive Modeling ...........................................................................................................702 81.3 Vasomotion in Pathologic Conditions ...................................................................................................................702 81.3.1 Increased Venous Pressure, Chronic Venous Insufficiency, and the Postthrombotic Syndrome .............702 81.3.2 Arterial Insufficiency .................................................................................................................................703 81.3.3 Sickle Cell Disease ....................................................................................................................................703 References .......................................................................................................................................................................704
81.1 PHYSIOLOGY OF VASOMOTION Small arteries and arterioles in the skin and the subcutaneous tissue exhibit vasomotion, i.e., rhythmic changes of vessel diameter due to a series of contractions and relaxations. The principal features of vasomotion have been described by Nicoll and Webb,1,2 who observed rhythmic changes of diameter of small vessels in bat wing in vivo. These authors reported that vasomotion induced variations of blood blow in the vessels (flow motion). Since that time vasomotion has been directly observed in various animal and human tissues: in the skin (hamster skinfold window preparation3 and cheek pouch4), nail-fold capillaries in man,5 skeletal muscle,6 testicle,7,8 conjunctiva,9 retina,10 brain,11 and heart.12 The origin of vasomotion has been studied mainly in the rabbit tenuissimus muscle model.6 It has been found that vasomotion is elicited by the rhythmic activity of smooth muscle pacemaker cells in which the spontaneous depolarization occurs.13–15 These cells are located in cushion-like thickening of vessel wall near the branching points and are supposed to provide the original trigger for vasomotion, which is eventually propagated downward to the larger arterioles (Figure 81.1).16–18 The frequency of vasomotion changes abruptly at bifurcation points and
gradually increases in the downstream direction, from 0.5 to 18 cycles per minute (cpm; 60 cpm = 1 Hz) in the larger arteries to 9 to 21 cpm in terminal arterioles. This downstream propagation of contractions and dilations causes superposition of waves, and the final vasomotion pattern in the distal elements of the vascular tree is the composite effect of signals that originate at various branching points in the microvasculature.18 The major role of arteriolar vasomotion and cyclic oscillations of blood flow in the microcirculation is probably the enhancement of blood passage in the capillaries. Poiseuille’s law, which describes flow of the fluid in the cylinder-like vessel, states that conductivity of the vessel is proportional to the fourth power of its radius. Therefore, the vessel of the oscillating diameter will be much more conductive than the vessel of the same mean but constant diameter.19,20 Additionally, vasomotion plays an important role in the control of vascular resistance. Secomb et al.19 and Secomb and Gross21 predicted that the increase of vascular resistance to four times the initial value would be impossible without the participation of vasomotion, because in this instance the vascular diameter must have been controlled within very tight and unrealistic limits, and the diameter of the vessel would be smaller than the critical minimum diameter for passage of red blood cells. 697
698
Handbook of Non-Invasive Methods and the Skin, Second Edition
Epidermis
1 min
Capillaries
Fluxmotion
Superficial arteriolar plexus
Vasomotion 5–20 cpm
Arterioles Arterioles
Vasomotion 2–5 cpm
Bifurcational thickening (smooth muscle cells) vasomotion pacemaker
FIGURE 81.1 Hypothetical origin of vasomotion in human cutaneous microcirculation. (Based on References 13–18, 75.) Vasomotion is triggered by pacemaker smooth muscle cells located at arterial bifurcations. The dominant vasomotor frequency (5–10 cpm) is generated at the origin of ascending arterioles, whereas vasomotion in the larger arteries has lower frequency. The complex pulsating flow in the superficial arteriolar plexus and the capillaries is a result of superposing of rhythms from several pacemakers.
Therefore, vasomotion enables red blood cell transfer in the conditions of increased vascular resistance. Other functions of vasomotion are listed in Table 81.1.
81.2 METHODS OF DETECTION AND ANALYSIS OF VASOMOTION The ideal method for the examination of cutaneous vasomotion would be the direct observation of the changes of the diameter of the arterioles in the skin. Since this cannot be easily accomplished in humans, methods have been developed that detect temporal variations of blood flow
TABLE 81.1 Role of Vasomotion in Microcirculation Function Enhancement of vessel conductivity Control of vascular resistance Improvement of oxygen delivery to the issues Maintenance of blood fluidity Removal of tissue edema Stimulation of lymph flow Detachment of cells adherent to endothelium
Ref. 19 21 19 91 92, 100 93, 101 90
(flow motion) or red blood cell flux* (flux motion) that appear secondarily to arteriolar vasomotion. Flow motion in skin capillaries can be analyzed with dynamic capillaroscopy, whereas the laser Doppler technique is used for the analysis of flux motion in the microvasculature.
81.2.1 INTRAVITAL MICROSCOPY Intravital dynamic capillaroscopy is based on obtaining a magnified image of skin capillaries and dynamic recording of red blood cell velocity in these vessels. The most often used system has been developed by Bollinger et al.5 and Butti et al.22 and further improved by Fagrell et al.23,24 In this system vessels are visualized with an in vivo microscope and the image is recorded with the aid of a closedcircuit television.25 Nail-fold capillaries (in the finger or toe) have been commonly investigated, but some studies were also done in a titanium chamber system.26 The alterations of the velocity of red cells are analyzed with video densitometric techniques with correlation of the photometric signals.27 Wayland and Johnson28 developed a dualslit method that is based upon the measurement of the delay between the two signals elicited by the same configuration of red blood cells as they pass two photosensors separated by a fixed distance. More recently, Intaglietta et * Flux is the product of the red blood cell concentration and speed.
Examination of Periodic Fluctuations in Cutaneous Blood Flow
al.29 described a video dual-window technique for measuring of blood velocity. The video signal passes through two independent square areas (windows) fitted in size to the capillary width. The velocity of blood is determined by measuring the intrawindow transit time. Simultaneously with blood velocity other parameters may be recorded, such as arterial pulsations in the finger, respirations, ECG, etc.23 However, the real-time online determination of blood cell velocity with these systems is complicated, especially when high velocities are to be measured. Slaaf et al.30 proposed an easy-to-operate system based on a three-stage prism grating technique. A microscopic image of a microvessel is projected on a grating of alternating transparent and opaque lines, and the light that passes through the grating is focused on a photosensor via the transfer lens. Moving erythrocytes modulate the intensity of light that is recorded by the photosensor. This allows the online determination of the velocity and the direction of flow in the capillary. Direct observation of blood flow velocity changes in human nail-fold capillaries revealed spontaneous fluctuations, in most cases at the frequency of 6 to 10 cpm, that were not related to normal respiration.5,23,31 These fluctuations of flow velocity are considered to be a result of arteriolar vasomotion31–33 and are correlated with changes of pressure of blood in a capillary, increased pressure being associated with the most rapid speed of flow.34 The principal drawback of examining vasomotion by direct observation of blood vessel is that capillary microscopy is restricted to nail-fold capillaries. Examination of vasomotion in other regions requires implantation of a titanium chamber and cannot be considered fully noninvasive.
81.2.2 LASER DOPPLER TECHNIQUE The laser Doppler method has recently become an attractive alternative to intravital dynamic capillaroscopy for studying cutaneous vasomotion. The laser Doppler technique enables simple, real-time monitoring of relative changes of red blood cell flux in the cutaneous microvascular bed in any region of the body. The pioneering laser Doppler studies of Holloway and Watkins,35 Tenland et al.,36 and Salerud et al.37 demonstrated spontaneous oscillations of blood flux (flux motion) in normal human skin in resting conditions. The average frequency of the oscillations recorded by these authors was 8.6 cpm, and this value was comparable with rhythmic variations of capillary blood pressure,34,38 capillary blood velocity,5,39,40 and vasomotion.3 Oscillations of flux could not be suppressed by a proximal nervous blockade, implying a local, myogenic mechanism. It therefore has been concluded that the flux motion recorded with the laser Doppler technique reflects changes of microcirculatory blood flow due to arteriolar vasomotion.37
699
It is not fully understood what actually generates the laser Doppler signal in the skin,32,41–44 and there is much debate to what extent flux motion recorded with laser Doppler technique reflects flow motion in the microvasculature due to arteriolar vasomotion. Tooke et al.43 compared the laser Doppler method with video dynamic capillaroscopy and found that with both techniques it was possible to record oscillations of 4 to 6 cpm. When the Pearson product momentum correlation coefficient, which predicts the relationship between two sets of paired data, was calculated, a linear relationship between the laser Doppler signal and red blood cell velocity was seen in 11 of 14 sets of records. In 8 of the 14 recordings the best correlation of flux motion and flow motion was found at no time delay. Spectral analysis showed that in 14 of 16 recordings the flow motion pattern had similar frequency spectra to the flux motion. The amplitudes of flow oscillations were significantly higher than the amplitudes of flux motion. Therefore, although laser Doppler measures blood flux not only in the nutritive capillaries, but also in deeper elements of the skin vascular tree,42–45 this is a useful tool for evaluation of cutaneous vasomotion. The issue of the correlation of the origin of laser Doppler signal with the microvascular segments of human skin has been further investigated by Braverman et al.46 They found that with a commonly used laser Doppler probe the maximum amplitude of rhythmic oscillations could be obtained when the probe was placed directly over an ascending elastic arteriole and its immediate branches. On the other hand, when the probe was moved to the site between ascending arterioles, the nonpulsatile laser Doppler signal of the flow flux value was recorded. The dependency of the laser Doppler flux pattern on the position of the probe on the skin may explain low reproducibility in the detection of flux motion among different authors. In some studies cyclic changes of cutaneous flux could not be detected in normal conditions in humans.47,48 Because of these difficulties, methods for amplification of vasomotion have been devised. Wilkin47 observed that cutaneous flux motion is provoked in the early phase of postocclusive reactive hyperemia. The hyperemia response has been obtained after occlusion of the brachial artery with a sphygmomanometer cuff for 6 min. After release of the occlusion, a prompt increase of flux is seen and blood flux oscillations can be recorded from the forearm skin during the return of blood flux to normal values (Figure 81.2). The average period of these oscillations is 9.6 ± 0.3 s (SE), in accordance with the reported rhythmic variations of blood flow observed microscopically in human nail-fold capillaries23 and flux motion in the forehead37 and leg49 obtained in the normal resting conditions. It is conceivable that amplified laser Doppler flux oscillations are due to the local synchronicity of oscillatory blood flow in a group of cutaneous capillaries in the period of hyperemia. The synchronicity is
700
Handbook of Non-Invasive Methods and the Skin, Second Edition
a
b
c
TABLE 81.2 Modulation of Oscillatory Blood Flow in Cutaneous Circulation Oscillatory Blood Flow Amplitude
Frequency
Dependent position (increased venous pressure) Post-thrombotic syndrome
↓ ↓
0 ↓
↓ ↓
Arterial insufficiency Sickle cell disease Diabetes mellitus General anesthesia Postocclusive reactive hyperemia Topical application of: Propionaldehyde Nitroglycerine Carbon dioxide Thermal challenge Hyperventillation Smoking Hypertension Aging
0 ↑ ↓ ↓ ↑
↓ ND ↑ ↑ 0 ↓ 0 ND
70a, 66, 67 68 69, 71 56, 73 48, 77 94b, 95c, 96 51, 97 47, 48, 51
↑ ↑ ↑ ↑ ↑ 0 ↑ ↓
ND ND ND 0 0 ↓ 0 0
50 51 53 52 54 26 98d 99
Flux
Condition
Time
FIGURE 81.2 Enhancement of cutaneous blood flow oscillations in the early phase of postocclusive hyperemia. (a) Baseline flux, (b) arterial occlusion, and (c) reactive hyperemia. Note prominent oscillations of blood flux on the descending arm of hyperemic response (c).
limited to a small area of the skin. Loss of synchronicity has been reported in the sites only 0.5 to 2 cm apart, implying a local origin of the oscillations.37,43,47 Besides postocclusive hyperemia, cutaneous flux oscillations can be induced by other stimuli that provoke a local increase in skin blood flow (Table 81.2). Wilkin50 applied topically a 5 M aqueous solution of propionaldehyde and reproducibly a tenfold increase of cutaneous blood flux. During the recovery to the resting flux level marked oscillations of flux could be detected. A similar hyperemic response and augmentation of flux motion in the recovery phase could be induced with a topical 2.2 × 10–2 M nitroglycerine.51 Kastrup et al.52 reported that during local skin heating to 42˚C flux motion is induced in the posthyperemic phase. The oscillations, with a mean frequency of 6.9 cpm (range, 5.2 to 10.4) were not suppressed during local or central nervous blockade, with lidocaine or trimethaphan campsylate, respectively. Some authors were also able to induce vasomotion by a local application of carbon dioxide53 and during hyperventilation.54
81.2.3 ANALYSIS
OF
WAVEFORM PATTERNS
Cyclic oscillations in the blood flow or flux tracings may be analyzed by manual determination of wave frequency, amplitude, and prevalence. It has been realized that the frequency and amplitude of flux motion waves are heterogeneous and may be further subdivided into discrete components. Kastrup et al.52 divided flux motion into two constituents: high-frequency regular oscillations (mean, 6.8 cpm) of the nonneurogenic (probably myogenic) origin and irregular low-frequency oscillations (mean, 1.5 cpm) that were caused by periodic changes in the autonomic tone. Similarly, Scheffler and Rieger,55 Seifert et al.,56 and Bongard and Fagrell57 found two categories of
Ref. 49 57
Note: ↓-decreased; ↑-increased; 0-no change; ND-not done. a
In legs with edema, vasomotion restored after edema removal. Based on the thermal clearance method (detection of arteriovenous anastomotic blood flow). c Based on venous occlusion plethysmography. d Studies in experimental animals. b
oscillations: large waves of large amplitude and low frequency (<10 cpm) and small waves of low amplitude and a frequency higher than 15 cpm. The prevalence of the small waves in the laser Doppler signal has been reported to be between 5 and 50%. For the objective analysis of vasomotion spectral methods have been recently introduced. The most widely used techniques, fast Fourier transform, autoregressive algorithms, and Prony spectral line estimation (PSLE), will be briefly reviewed in Sections 81.2.3.1 to 81.2.3.3. 81.2.3.1 Fast Fourier Transform Fourier transform isolates the repetitive harmonic (sine and cosine) parts of the time-domain data and calculates the frequency and power of each repetitive component.58 Rather than the classical Fourier transform, which is extremely computational intensive, its fast version is usually used. The requirement for the fast Fourier transform (FFT) is that the number of sample points in a
Examination of Periodic Fluctuations in Cutaneous Blood Flow
time-domain signal must be a power of 2. An FFT of an oscillatory tracing gives two sets of frequency domain data: real and imaginary. The first set shows the values of component sine waves at different frequencies, while the latter shows cosine harmonic components. For example, if the original data are a simple sine wave with a frequency of 1 Hz and an average amplitude of 10 U, the FFT will give a value of 10 at frequencies +1 Hz (real) and –1 Hz (imaginary). The power of the component frequencies is usually shown in a power spectrum. Each power spectrum value is the sum of the squares of the values of the real and imaginary FFT values (in our example, 102 + 102 = 200 U at 1 Hz). Another presentation of FFT is the power spectral density graph. Each value for power spectral density is calculated as power spectrum value/frequency bandwidth per point, and thus the area under the power spectral density curve will be in units of power. Power spectral density allows for easier comparison of the data independently on the sample length and rate of recording. The most prominent limitation of FFT is the resolution of the spectral components of similar frequency. A second limitation is due to the windowing of the data that occurs when processing with FFT. In practice, one deals always with finite sets of time-domain data, and this introduces error to Fourier analysis, since FFT tries to find repetitive
701
signals from the edges of the data block (a leakage phenomenon).58 Due to the leakage phenomenon, the power of the main lobe “leaks” to the side lobes, distorting other spectral responses that are present so that weak components can be masked by higher sidelobes from stronger spectral responses. Leakage may be substantially reduced by mathematical modification of the input data (windowing59,60). Such transformed data give cleaner power spectra, however, at the expense of reduced frequency resolution. Data windowing should be principally used when the analysis of a weak repetitive component, which may be easily obscured by leakage, is required. The typical laser Doppler recording with the visible cyclic oscillations and the result of its FFT analysis are shown in Figure 81.3. Usually, besides the heartbeat component at about 1 Hz (60 cpm), two or three low-frequency spectral components are obtained at approximately 2.5, 5, and 12 cpm, the first two being the most prominent. It is conceivable that the very low frequency components (2.5 cpm) arise in the larger arterioles, whereas the higherfrequency components are produced in the ascending arterioles and subcapillary arteriolar plexus in the skin. The origin of high-frequency components (>10 cpm) that are sometimes seen in PSD may reflect either vasomotor
40
20
0 6.24
6.32
6.40
6.48
6.56
7.4
7.12
7.20
7.28
7.36
7.44
7.52
8.0
8.8
(a)
2 × 102
1 2
3
1 × 102
0 × 102 0.000
0.312
0.625
0.938
1.2
(b)
FIGURE 81.3 Fast Fourier transform (FFT) analysis of cutaneous flux-motion in humans. (a) Laser Doppler recording taken from the lower extremity (horizontal position), x axis = time (min:sec), y axis = units of flux. (b) FFT analysis — PSD graph. x axis = frequency (Hz), y axis = power of harmonic components. Note two distinct vasomotion-specific peaks at 2.34 (1), and 4.68 (2) cpm, in addition to a smaller peak at 61 cpm (3), characteristic for pulse-related flux changes in the skin.
702
Handbook of Non-Invasive Methods and the Skin, Second Edition
activity or, alternatively, the ventilatory effect on venous return, respiratory sinus arrhythmia, or periodic blood pressure waves — all were shown to reside in the 9.6- to 15-cpm frequency band.61 81.2.3.2 Prony Spectral Line Estimation The PSLE technique has been proposed for analysis of vasomotion by Colantuoni et al.18 and Meyer and Intaglietta.62 The basic algorithm has been devised by Prony63 and later by Burkhardt.64 A finite block of time-domain data is modeled as the sum of finite numbers of nonharmonically related sinusoids in white (uncorrelated) noise, and the original waveform pattern is reconstructed by the iterative procedure. A first order of size approximations contains one sinusoid, and the correlation coefficient is computed at each increment of order size. If the correlation coefficient is larger than the previous order, then this solution is kept and the previous is discarded. Thus, a maximal correlation will be seen at a specific order.65 If the correlation is greater than the predetermined level (usually the correlation coefficient >0.8 to 0.95), a solution has been found. In certain circumstances PSLE is more powerful than FFT, because the resolution is not dependent upon the data length.58 The drawback of the PSLE method is computational complexity and the necessity to determine the order.58 Moreover, the PSLE graph may be difficult to interpret for an uninitiated individual due to a relatively large number of spectral lines. 81.2.3.3 Autoregressive Modeling Autoregressive PSD estimation employs the autocorrelation method to find the repetitive components in the input data.58 Autoregression gives good results for strong sinusoidal components; however, the power of amplitude of the components cannot be calculated.58 Other limitations of autoregression methods involve the degrading effect of observation noise, the presence of multiple spurious peaks, and the shifting of main frequencies toward higher values at low signal-to-noise ratio.58 The resolution of autoregressive modeling is often better than that of FFT. Moreover, this procedure is not dependent on the number of samples, and the results are not affected by windowing.
81.3 VASOMOTION IN PATHOLOGIC CONDITIONS Vasomotion is easily modulated in a variety of physiologic and pathologic conditions (Table 81.2). Changes of vasomotion in arterial insufficiency, venous hypertension, and sickle cell disease were investigated in some detail.
81.3.1 INCREASED VENOUS PRESSURE, CHRONIC VENOUS INSUFFICIENCY, AND THE POSTTHROMBOTIC SYNDROME The influence of the increased venous pressure in the lower extremity, caused either experimentally by leg lowering or by the pathologic process of venous thrombosis and valve injury (a postthrombotic syndrome), has been recently investigated by several groups with laser Doppler flowmetry.49,57,66–69 Increased leg venous pressure led to postural vasoconstriction and decrease of amplitude of vasomotion49,57 (Figure 81.4). It is not clear whether increased venous pressure causes changes of the frequency of vasomotion. Bongard and Fagrell57 found a decrease of vasomotion frequency from 3.5 to 2.1 cpm, whereas the Fourier spectral analysis performed by Gniadecki et al.49 did not reveal significant changes of the frequency of the dominant harmonic component of the laser Doppler flux signal. The postural attenuation of vasomotion could be suppressed by the application of leg compression,49 a finding providing further proof that increased venous pressure and venous distention are responsible for the suppression of vasomotion. The vasomotion patterns in patients with chronic venous insufficiency are not consistent. Gniadecka et al.70 reported that in these patients vasomotion could not be detected in the region of lipodermatosclerosis either in the horizontal or in the dependent position of the lower extremity. Removal of the leg edema, which coexists with chronic venous insufficiency, with long-term (4 weeks) compressive therapy restored normal cutaneous vasomotion. Therefore, it is likely that limb edema itself inhibits vasomotion in lipodermatosclerosis. Similar findings were reported by Pekanmäki et al.66 These authors were able to restore vasomotion (5 cpm; range, 2 to 8 cpm) in the patients with lipodermatosclerosis by a single treatment with intermittent pneumatic compression (Ventipress®, Lemi, Finland). However, in this study, the position of the patients during laser Doppler measurement has not been reported and leg edema has not been assessed. Therefore, it is difficult to conclude whether the restoration of vasomotion due to intermittent pneumatic compression of the leg could be attributed to the removal of leg edema or to other mechanisms. Decrease of the amplitude of blood flux oscillations in the postthrombotic syndrome has also been observed by Belcaro et al.68 In a marked contrast, Chittenden et al.69 and Cheatle et al.71 reported an enhancement of vasomotion in patients with chronic venous insufficiency in the areas of lipodermatosclerosis. These authors recorded statistically significant higher baseline laser Doppler flux, higher vasomotion frequency (3.2 vs. 1.92 cpm in the control), and higher vasomotion amplitude (8.5 vs. 1.2 mV, control) in the lipodermatosclerosis. Increased frequency and amplitude of vasomotion were unlikely to
Examination of Periodic Fluctuations in Cutaneous Blood Flow
induced by stiff sickle erythrocytes blocking blood flow in the microvessels (see Sections 81.3.2 and 81.3.3).
40 H
D
81.3.2 ARTERIAL INSUFFICIENCY
20 F1 F2 0 0.0
703
2.40
5.20
8.0
10.40
13.20
16.0
18.40 21
(a) 6 × 102
4 × 102
2 × 102
0 × 102 0.000
0.312
0.625
0.
(b) 3 × 102
2 × 102
1 × 102
0 × 102 0.000
0.312
0.625
0.
(c)
FIGURE 81.4 Effects of the increased venous pressure, caused by leg lowering, on cutaneous blood flux oscillations: (a) original laser Doppler recordings (H = horizontal position, D = dependent position), (b) power spectral densities (FFT) of blood flux in the horizontal position, and (c) in the dependent position. Note the marked decrease of power amplitudes in (c).
be attributed to high flux because increasing of flux with pilocarpine did not induce changes of vasomotion. These results are difficult to compare with the studies showing decreased vasomotion in chronic venous insufficiency, because the authors did not assess leg edema in their patients. In the absence of edema, vasomotion in lipodermatosclerosis and chronic venous insufficiency could be enhanced secondarily, due to capillary plugging by white blood cells,72 a mechanism similar to that seen in sickle cell disease, where cutaneous vasomotion is
The effects of leg ischemia due to peripheral atherosclerosis on vasomotion were studied by Yanar et al.73 and Seifert et al.56 Cutaneous vasomotion in the second and third toe were recorded with laser Doppler in four groups of patients with different stages of the disease (Table 81.3). The authors found increased prevalence of the high-frequency vasomotion laser Doppler waves proportional to the severity of the peripheral ischemia (Table 81.3). Increased vasomotion frequency was directly linked to peripheral hypoperfusion since the prevalence of highfrequency waves decreased after successful restoration of peripheral circulation by means of angioplasty or thrombolysis.74 The pathophysiologic basis of the changes of vasomotion in peripheral ischemia was studied by Bertuglia et al.75 in a hamster skinfold window model. These authors divided skin microvasculature according to Strahler classification,75,76 so that order 0 was capillaries and order 4 the largest skin arterioles. In the baseline situation the dominant vasomotion frequency was found in order 1 arterioles (4 to 18 cpm; mean, 9.1 ± 3.9 cpm) and decreased with the increasing arteriole order, a finding in accordance with earlier observations of Colantuoni et al. 18 However, during experimentally evoked tissue hypoxia caused by the inspiration of an 8 and 11% O2 mixture, the frequency of vasomotion increased in all arteriolar branches and the dominant frequency was generated in order 3 arterioles (25.5 ± 4.5 cpm at 8% O2 vs. 3.4 ± 1.8 cpm in the control). Increased frequency of vasomotion was accompanied by a decreased mean and effective vessel diameters and reduced capillary blood flow. It is probable that the laser Doppler findings of increased flux oscillation frequency during limb ischemia in humans could be explained by the increased activity of vasomotion pacemaker in larger cutaneous arteries. Such a phenomenon is likely to reduce resistance in the skin microcirculatory network and ensure adequate blood supply to the tissue.
81.3.3 SICKLE CELL DISEASE Cutaneous vasomotion in the patients with sickle cell disease has been investigated with laser Doppler fluxmetry by Rodgers et al.48 and Gniadecka et al.77 These authors found prominent oscillations of blood flux of the period 7 to 10 sec and peak-to-trough magnitudes about half the mean flow. The oscillations were apparently associated with the presence of the pathological hemoglobin S, since in two patients studied by Rodgers et al.48 blood transfusion resulted in the diminution or disappearance of the rhythmic variations in blood flux.
704
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 81.3 Fluxmotion in Patients with Peripheral Ischemia High Frequency Wavesa Degree of Limb Ischemia No (control) Light Moderate Severe a b
Ankle/Arm Pressure Ratioa 1.22 0.81 0.62 0.38
(0.2) (0.1) (0.2) (0.1)
Walking Distance
Frequency (Hz)
Amplitudeb
Prevelance (%)
— >200 m <200 m Rest pain
26 23.0 (5.3) 19.9 (3.9) 22.5 (4.0)
0.24 0.27 (0.1) 0.22 (0.1) 0.19 (0.1)
8 33 75 92
Mean with (SE). Laser Doppler blood flux units.
Modified from Bollinger, A., Hoffman, U., Seifert, H., in Progress in Applied Microcirculation, Vol. 15, Intaglietta, M., Ed., Basel, Karger, 1989, 87–92. With permission.
The origin of enhanced vasomotion is not clear. One possibility is that sickle cells, which are stiffer and more resistant to deformation than normal erythrocytes,78,79 become trapped at the entrance to capillaries, as shown in animal models and in human retina.80–84 This phenomenon may lead to the local increase of intravascular pressure85 that stimulates contraction of smooth muscle in the precapillary sphincters and myogenic oscillation pattern in the arterioles.86–88 Alternatively, cyclic skin hypoperfusion and hypoxia caused by the plugging of capillaries with sickle cells may stimulate vasomotion, as shown in animal models by Zweifach and Lipowsky89 and Bertuglia et al.75 Mechanical stimulation of the endothelium by sickle cells and release of vasoactive mediators cannot be excluded.90 As discussed above, vasomotion significantly improves tissue perfusion. Therefore, it is likely that enhancement of vasomotion in sickle cell disease is the compensatory mechanism to upregulate nutritive blood flow in the skin. The finding that, unlikely in the normal situation, vasomotion does not disappear after postural vasoconstriction caused by the lowering of the leg77 (Figure 81.4) further supports the hypothesis for the compensatory role of vasomotion in sickle cell disease.
REFERENCES 1. Nicoll PA, Webb RL. Blood circulation in the subcutaneous tissue of the living bat’s wing. Ann NY Acad Sci 46: 697–709, 1946. 2. Nicoll PA, Webb RL. Vascular patterns and active vasomotion as determiners of flow through minute vessels. Angiology 6: 291–310, 1955. 3. Colantuoni A, Bertuglia S, Intaglietta M. Quantification of rhythmic diameter changes in arterial microcirculation. Am J Physiol 246: H508–H517, 1984. 4. Duling BR, Gore RW, Dacey RC, Damon DN. Methods for isolation, cannulation and in vitro study of single microvessels. J Appl Physiol 241: H108–H116, 1981.
5. Bollinger A, Butti P, Barras JP, Traschler H, Siegenthaler W. Red blood cell velocity in nail fold capillaries of man measured by a television microscopy technique. Microvasc Res 7: 61–72, 1974. 6. Lindbom L, Tuma RF, Arfors KE. Blood flow in the rabbit tenuissimus muscle. Influence of preparative procedures for intravital microscopic observations. Acta Physiol Scand 114: 121–127, 1982. 7. Collin O, Bergh A, Damber JE, Widmark A. Control of testicular vasomotion by testosterone and tubular factors in rats. J Reprod Fertil 97: 115–121, 1993. 8. Damber JE, Maddocks S, Widmark A, Bergh A. Testicular blood flow and vasomotion can be maintained by testosterone in Leydig cell-depleted rats. Int J Androl 15: 385–393, 1992. 9. Bachir D, Maurel A, Portos JL, Galacteros F. Comparative evaluation of laser Doppler flux metering, bulbar conjunctival angioscopy, and nail fold capillaroscopy in sickle cell disease. Microvasc Res 45: 20–32, 1993. 10. Braun RD, Linsenmeier RA, Yancey CM. Spontaneous fluctuations in oxygen tension in the cat retina. Microvasc Res 44: 73–84, 1992. 11. Morita Tsuzuki Y, Bouksela E, Hardebo JE. Vasomotion in the rat cerebral microcirculation recorded by laser Doppler flowmetry. Acta Physiol Scand 146: 431–439, 1992. 12. Iversen PO. Evidence for long-term fluctuations in regional blood flow within the rabbit left ventricle. Acta Physiol Scand 146: 329–339, 1992. 13. Casteels R, Droogmans G, Himpens B. Excitation-contraction coupling in vascular smooth muscle cells and perivascular nerve stimulation. J Cardiovasc Pharmacol 6 (Suppl.): S9–S12, 1985. 14. Mulvany MJ. Functional characteristics of vascular smooth muscle. In Progress in Applied Microcirculation, Vol. 3. Karger, Basel, 1983, pp. 4–18. 15. Colantuoni A, Bertuglia S, Intaglietta M. The effects of alpha- or beta-adrenergic receptor agonists and antagonists and calcium entry blockers on the spontaneous vasomotion. Microvasc Res 28: 143–158, 1984.
Examination of Periodic Fluctuations in Cutaneous Blood Flow
16. Meyer JU, Lindblom L, Intaglietta M. Coordinated diameter oscillations at arteriolar bifurcations in skeletal muscle. Am J Physiol 253: H568–H573, 1987. 17. Meyer JU, Borgström P, Intaglietta M. Is vasomotion due to microvascular pacemaker cells? In Vasomotion and Flow Modulation in the Microcirculation, Progress in Applied Microcirculation, Vol. 5, Intaglietta M, Ed. Basel, Karger, 1989, pp. 41–48. 18. Colantuoni A, Bertuglia S, Intaglietta M. Variation of rhythmic diameter changes at the arterial microvascular bifurcations. Pflügers Arch 403: 289–295, 1985. 19. Secomb TW, Intaglietta M, Gross JF. Effects of vasomotion on microcirculatory mass transport. In Vasomotion and Flow Modulation in the Microcirculation, Progress in Applied Microcirculation, Vol. 15, Intaglietta M, Ed. Basel, Karger, 1989, pp. 49–61. 20. Wilkin JK. Poiseuille, periodicity, and perfusion: rhythmic oscillatory vasomotion in the skin. J Invest Dermatol 93: 113S–118S, 1989. 21. Secomb TW, Gross JF. Flow of red blood cells in narrow capillaries. Role of membrane tension. Int J Microcirc Clin Exp 2: 229–240, 1983. 22. Butti P, Intaglietta M, Reimann H, Hollinger CH, Bollinger A, Anliker M. Capillary red blood cell velocity measurements in human nail fold by videodensitometric method. Microvasc Res 10: 1–8, 1975. 23. Fagrell B, Frontek A, Intaglietta M. A microscope-television system for studying flow velocity in human skin capillaries. Am J Physiol 233: H318–H321, 1977. 24. Fagrell B, Frontek A, Intaglietta M. A microscope-television system for studying flow velocity in human skin capillaries. Am J Physiol 246: H508–H517, 1984. 25. Bloch EH. A quantitative study of the haemodynamics in the living microvascular system. Am J Anat 110: 125–153, 1962. 26. Asano M, Bränamark P-I. Cardiovascular and microvascular responses to smoking in man. Adv Microcirc 3: 125–158, 1970. 27. Wood E, Strum RE, Sanders JJ. Data processing in cardiovascular physiology with particular reference to roentgen videodensitometry. Mayo Clin Proc 39: 849–865, 1964. 28. Wayland H, Johnson PC. Erythrocyte velocity measurement in microvessels by two-slit photometric method. J Appl Physiol 22: 333–337, 1967. 29. Intaglietta M, Silverman NR, Tompkins WR. Capillary flow velocity measurements in vivo and in situ by television methods. Microvasc Res 10: 165–179, 1975. 30. Slaaf DW, Rood JPSM, Tangelder GJ, Jeurens TJM, Alewijnse R, Reneman RS, Arts T. A bidirectional optical (BDO) three-stage prism grating system for on-line measurement of red blood cell velocity in microvessels. Microvasc Res 22: 110–122, 1981. 31. Fagrell B. Capillary dynamics in man. In Vasomotion and Quantitative Capillaroscopy, Progress in Applied Microcirculation, Vol. 3, Messmer, Hammersen, Eds. Karger, Basel, 1983, pp. 119–130.
705
32. Fagrell B. Microcirculation in the skin. In Physiology and Pharmacology of the Microcirculation, Vol. 2, Mortillaro NA, Ed. Academic Press, New York, 1984, pp. 133–180. 33. Funk W, Endrich B, Messmer K, Intaglietta M. Spontaneous arteriolar vasomotion as a determinant of peripheral vascular resistance. Int J Microcirc Clin Exp 2: 11–25, 1983. 34. Mahler F, Muheim MH, Intaglietta M, Bollinger A, Anliker M. Blood pressure fluctuations in human nailfold capillaries. Am J Physiol 236: H888–H893, 1979. 35. Holloway GA, Watkins DW. Laser Doppler measurement of cutaneous blood flow. J Invest Dermatol 69: 306–309, 1977. 36. Tenland T, Salerud EG, Nilsson GE, Öberg PÄ. Spatial and temporal variations in human skin blood flow. Int J Microcirc Clin Exp 2: 81–90, 1983. 37. Salerud EG, Tenland T, Nilsson GE, Oberg PA. Rhythmical variations in human skin blood flow. Int J Microcirc Clin Exp 2: 91–102, 1983. 38. Wiederhielm CA, Weston BV. Microvascular, lymphatic and tissue pressures in the unanesthesized mammal. Am J Physiol 225: 992–996, 1973. 39. Fagrell B, Intgalietta M, Östergren J. Relative hematocrit in human skin capillaries and its relationship to capillary flow velocity. Microvasc Res 20: 327–335, 1980. 40. Fagrell B, Intgalietta M, Tsai AM, Östergren J. Combination of laser Doppler flowmetry and capillary microscopy for evaluating the dynamics of skin microcirculation. Prog Appl Microcirc 11: 125–138, 1986. 41. Caspary L, Creutzig A, Alexander K. Biological zero in laser Doppler fluxmetry. Int J Microcirc Clin Exp 7: 367–371, 1988. 42. Nilsson GE, Tenland T, Öberg PÄ. Evaluation of a laser Doppler flowmeter for measurement of tissue blood flow. IEEE Trans Biomed Eng 27: 597–604, 1980. 43. Tooke JE, Östergren J, Fagrell B. Synchronous assessment of human skin microcirculation by laser Doppler flowmetry and dynamic capillaroscopy. Int J Microcirc Clin Exp 2: 277–284, 1983. 44. Svensson H, Jönsson BA. Laser Doppler flowmetry during hyperaemic reaction in the skin. Int J Microcirc Clin Exp 7: 87–96, 1987. 45. Hales JRS, Westerman RA, Roberts RGD, Fawcett AA, Stephens FRN. Evidence for Laser Doppler Discrimination between Skin AVA and Capillary Perfusion. Paper presented at the European Laser Doppler Users Group (ELDUG), plenary session, London, 1992. 46. Braverman IM, Keh A, Goldminz D. Correlation of laser Doppler wave patterns with underlying microvascular anatomy. J Invest Dermatol 95: 283–286, 1990. 47. Wilkin JK. Periodic cutaneous blood flow during postocclusive reactive hyperaemia. Am J Physiol 250: H767–H768, 1986. 48. Rodgers GP, Schechter AN, Noguchi CT, Klein HG, Nienhuis AW, Bonner RT. Periodic microcirculatory flow in patients with sickle-cell disease. N Engl J Med 311: 1534–1538, 1984.
706
Handbook of Non-Invasive Methods and the Skin, Second Edition
49. Gniadecki R, Gniadecka M, Kotowski T, Serup J. Alterations of skin microcirculatory rhythmic oscillations in different positions of the lower extremity. Acta Derm Venereol (Stockh) 72: 259–260, 1992. 50. Wilkin JK. Periodic cutaneous blood flow during aldehyde-provoked hyperemia. Microvasc Res 35: 287–294, 1988. 51. Wilkin JK. Vasomotion in the cutaneous circulation. In Vasomotion and Flow Modulation in the Microcirculation, Progress in Applied Microcirculation, Vol. 15, Intaglietta M, Ed. Basel, Karger, 1989, pp. 62–74. 52. Kastrup J, Bülow J, Lassen NA. Vasomotion in human skin before and after local heating recorded with laser Doppler flowmetry. A method for induction of vasomotion. Int J Microcirc Clin Exp 8: 205–215, 1989. 53. Erdl R, Schnizer WE, Schöps K. Untersuchungen zur Wirkungsweise von CO2-bädern. Messungen an der Mikrocirkulation der Haut mittels eines Laser-DopplerFlowmeters. Herz/Kreisl 18: 387–391, 1986. 54. Smits TM, Aarnoudse JJ, Geerdink JJ, Zijlstra WG. Hyperventilation-induced changes in periodic oscillations in forehead skin blood flow measured by laser Doppler flowmetry. Int J Microcirc Clin Exp 6: 149–159, 1987. 55. Scheffler A, Rieger H. Signalverlaufmuster des LaserDoppler-Fluxes bei peripherer arterieller Verschlusskrankheit und ihre Beziehung zum systolischen Knöcheldruckindex. In Proceedings of the 12th Annual Meeting of the Schweizerischen Gesellschaft für Microzirkulation, Bern, 1988, pp. 73–77. 56. Seifert H, Jäger K, Bollinger A. Analysis of flowmotion by laser Doppler technique in patients with peripheral arterial occlusive disease. Int J Microcirc Clin Exp 7: 223–236, 1988. 57. Bongard O, Fagrell B. Variations in laser Doppler flux and flow motion patterns in the dorsal skin of the human foot. Microvasc Res 39: 212–219, 1990. 58. Kay MS, Marple SL. Spectrum analysis. A modern perspective. Proc IEEE 69: 1380–1419, 1981. 59. Harris FJ. On the use of windows for harmonic analysis with the discrete Fourier transform. Proc IEEE 66: 51–83, 1978. 60. Nuttall AH. Some windows with very good sidelobe behavior. IEEE Trans Acoust Speech Signal Process 29: 84–89, 1981. 61. Schmid-Schönbein H, Ziege S. Attractors and QuasiAttractors and the Assessment of Fluctuations in LaserDoppler Signals by Spectral Analysis. Paper presented at the European Laser Doppler Users Group (ELDUG), plenary session, London, 1992. 62. Meyer JU, Intaglietta M. Measurement of the dynamics of arteriolar diameter. Ann Biomed Eng 14: 109–117, 1986. 63. Prony GRB. Essai experimental et analytique, etc. Paris J Ecole Polytech, 1: 24–76, 1795. 64. Burkhardt P. Modification of the Prony Spectral Line Estimator with Applications to Vasomotion. Ph.D. dissertation, University of California, San Diego, 1983. 65. Hildebrand FF. Introduction to Numerical Analysis. McGraw-Hill, New York, 1956.
66. Pekanmäki K, Kolari PJ, Kiistala U. Laser Doppler vasomotion among patients with post-thrombotic venous insufficiency: effect of intermittent pneumatic compression. Vasa 20: 394–397, 1991. 67. Rowell LB. Human Circulation: Regulation during Physical Stress. Oxford University Press, Oxford, 1986, p. 416. 68. Belcaro C, Rulo A, Vaskedis S, Williams MA, Nicolaides AN. Combined evaluation of postphlebitic limbs by laser Doppler flowmetry and transcutaneous PO2/PCO2 measurements. Vasa 17: 259–261, 1988. 69. Chittenden SJ, Shami SK, Cheatle TR, Scurr JH, Coleridge-Smith PD. Vasomotion in the leg skin of patients with chronic venous insufficiency. Vasa 21: 138–142, 1992. 70. Gniadecka M, Gniadecki R, Serup J. Vasomotion, Posture, and Leg Ulceration. Paper presented at the 2nd Annual Meeting of the European Tissue Repair Society, Malmö, Sweden, August 26–28, 1992. 71. Cheatle TR, Shami SK, Stibe E, Coleridge-Smith PD, Scurr JH. Vasomotion in venous disease. J R Soc Med 84: 261–269, 1991. 72. Coleridge-Smith PD, Thomas P, Scurr JH, Dormandy JA. Causes of venous ulceration: a new hypothesis. Br Med J 296: 1726–1727, 1988. 73. Yanar A, Hoffmann U, Geiger M, Franzeck UK, Bollinger A. Laser-Doppler-Fluxmotion bei peripherer arterieller Verschlubkrankheit (PAVK). Vasa, Suppl. 8: 48–50, 1987. 74. Bollinger A, Hoffmann U, Seifert H. Flux motion in peripheral ischemia. In Vasomotion and Flow Modulation in the Microcirculation, Progress in Applied Microcirculation, Vol. 15, Intaglietta M, Ed. Basel, Karger, 1989, pp. 87–92. 75. Bertuglia S, Colantuoni A, Coppini G, Intaglietta M. Hypoxia- or hyperoxia-induced changes in arteriolar vasomotion in skeletal muscle microcirculation. Am J Physiol 260: H362–H372, 1991. 76. Ellsworth ML, Liu A, Dawant B, Popel AS, Pittman RN. Analysis of vascular pattern and dimension in arteriolar networks of the retractor muscle in young hamsters. Microvasc Res 34: 168–183, 1987. 77. Gniadecka M, Gniadecki R, Serup J, Saaondergaard J. Microvascular reactions to postural changes in patients with sickle cell disease. Acta Derm Venereol, in press. 78. Kaul DK, Fabry ME, Windisch P, Baez S, Nagel RL. Erythrocytes in sickle cell anemia are heterogeneous in their rheological and hemodynamic characteristics. J Clin Invest 72: 22–31, 1983. 79. Chien S. Rheology of sickle cells and erythrocyte content. Blood Cells 3: 283–303, 1977. 80. Nagpal KC, Goldberg MF, Rabb MF. Ocular manifestations of sickle hemoglobinopathies. Surv Ophtalmol 21: 391–411, 1977. 81. Lipowsky HH, Usami S, Chien S. Human SS red cell rheological behavior in the microcirculation of cremaster muscle. Blood Cells 8: 113–126, 1982. 82. Goldberg MF. Natural history of untreated proliferative sickle retinopathy. Arch Ophtalmol 85: 428–437, 1971.
Examination of Periodic Fluctuations in Cutaneous Blood Flow
83. Serjeant GR. The Clinical Features of Sickle Cell Disease. Elsevier, New York, 1974, pp. 208–219. 84. Klug PP, Lessin LS. Microvascular blood flow of sickled erythrocytes: a dynamic morphologic study. Blood Cells 3: 263–272, 1977. 85. Bonner RF, Rodgers GP, Schechter AN. Laser-Doppler measurements (LDV) of skin blood flow and number density of flowing RBCs in sickle cell patients. Int J Microcirc Clin Exp 3: 432, 1984. 86. Bayliss WM. On the local reactions of the arterial wall to changes of internal pressure. J Physiol (London) 28: 220–231, 1902. 87. Folkow B. Intravascular pressure as a factor regulating the tone of the small vessels. Acta Physiol Scand 17: 289–310, 1949. 88. Patterson GC. The role of intravascular pressure in the causation of reactive hyperaemia in the human forearm. Clin Sci 15: 17–25, 1956. 89. Zweifach BW, Lipowsky HH. Pressure flow relations in blood and lymph microcirculation. In Handbook of Physiology, Sec. 2, The Cardiovascular System, Vol. 4, Part 1, Ranking EM, Michel CC, Eds. American Physiological Society, Bethesda, MD, 1984, pp. 251–307. 90. Chien S. Rheology of sickle cells and the microcirculation. N Engl J Med 311: 1567–1569, 1984. 91. Schmid-Schönbein H, Klitzman B, Johnson PC. Vasomotion and blood rheology: maintenance of blood fluidity in the microvessels by rhythmic vasomotion. Bibl Anat 20: 138–143, 1981. 92. Papenfuss HD, Gross JF. Vasomotion and transvascular exchange of fluid and plasma proteins. Microcirc Endothel Lymphatics 2: 577–596, 1985.
707
93. Intaglietta M, Gross JF. Vasomotion, tissue fluid flow and the formation of lymph. Int J Microcirc Clin Exp 1: 55–56, 1982. 94. Corcoran JS, Yudkin JS. Loss of spontaneous variability of fingertip anastomotic blood flow in diabetic autonomic neuropathy. Clin Sci 72: 557–562, 1987. 95. Christensen NJ. Spontaneous variations in resting blood flow, postischemic peak flow and vibratory perception in the feet of diabetics. Diabetologia 5: 171–178, 1969. 96. Le Dévéhat C, Khodabandehlou T, Testu N. Effects of Buflomedil* on Cutaneous Blood Flow, Vasomotion, and Transcutaneous Oxygen Pressure in Diabetic Patients in an Open Study. Paper presented at the European Laser Doppler Users Group (ELDUG), plenary session, London, 1992. 97. Colantuoni A, Bertuglia S, Intaglietta M. Effects of anaesthesia on the spontaneous activity of the microvasculature. Int J Microcirc Clin Exp 3: 13–28, 1984. 98. Boegehold MA. Enhanced arteriolar vasomotion in rats with chronic salt-induced hypertension. Microcirc Res 45: 83–94, 1983. 99. Weiss M, Milman B, Rosén B, Eisenstein Z, Zim Lichman R. Analysis of the diminished skin perfusion in elderly people by laser Doppler flowmetry. Age Ageing 21: 237–241, 1992. 100. Intaglietta M. Vasomotor activity, time-dependent fluid exchange and tissue pressure. Microvasc Res 21: 153–164, 1981. 101. Skalak TC, Schmid-Schönbein GW, Zweifach BW. New morphological evidence for a mechanism of lymph formation in skeletal muscle. Microvasc Res 28: 95–112, 1984.
Doppler Flowmetry: Principles 82 Laser of Technology and Clinical Applications Gianni Belcaro and Andrew N. Nicolaides Irvine Laboratory for Cardiovascular Investigation and Research, St. Mary’s Hospital Medical School, London, United Kingdom
CONTENTS 82.1 82.2 82.3 82.4
Introduction............................................................................................................................................................709 Theory of Laser Doppler Flowmetry ....................................................................................................................709 Terminology...........................................................................................................................................................710 Calibration..............................................................................................................................................................710 82.4.1 The Zero LDF Calibration and the Biological Zero.................................................................................710 82.5 Problems of LDF in Clinical Practice...................................................................................................................711 82.5.1 Depth of Measurements and Volume in the Skin .....................................................................................711 82.6 Vasomotion ............................................................................................................................................................712 82.7 New Technology ....................................................................................................................................................713 82.8 Conclusions............................................................................................................................................................713 References .......................................................................................................................................................................714
82.1 INTRODUCTION Noninvasive optical methods to evaluate skin flow have been used for many years. The most used method is photoplethysmography (PPG), which records variations in skin flow by the evaluation of the absorption characteristics of the skin that are related to its blood flow content. By PPG qualitative data may be obtained, but quantitative data are difficult to obtain and standardize. Fluctuations in venous volume, due to postural changes or motion artifacts, may alter the signal and make the interpretation of skin flow variations difficult. Therefore, clinical applications of PPG in the assessment of arterial or venous disease have never been defined with standard and conclusive diagnostic methods of evaluation. The use of PPG has been mainly limited to complementary methods able to evaluate some aspects of vascular disease, e.g., the pulsatility of skin flow in Raynaud’s disease or the refilling time after venous emptying following an exercise test. Furthermore, PPG evaluates skin flow for a variable depth (2 to 6 mm), including in this layer different skin circulatory elements without separating them from the more superficial, nutritional skin flow. Also, the PPG probes are
not easily usable to assess flow in cavities (such as intestine) or parenchymatous organs (liver, speen, muscle, etc.) Laser Doppler flowmetry (LDF), a theoretically comparable method, has been developed in the last 15 years from a pure scientific instrument to a research tool, and in the last few years to a clinical diagnostic technique to assess tissue viability and perfusion.1–4 The fundamental principles of LDF as applied to the measurements of blood flow have been described in detail in several publications.4–8,15 Measurement of blood velocity in single, large vessels and the measurement of skin perfusion can both be obtained using LDF. However, the technology required for these two applications is completely different, and in this chapter we refer only to the measurement of microvascular perfusion.
82.2 THEORY OF LASER DOPPLER FLOWMETRY A detailed technical description of the method is beyond the aim of this chapter. However, some simple concepts are needed to understand the method, its applications, and possible limitations. Most commercially available laser 709
710
Handbook of Non-Invasive Methods and the Skin, Second Edition
Doppler instruments utilize helium–neon gas or gallium–aluminum arsenide elements to produce a weak laser beam with low tissue penetration. This type of laser light does not alter the tissues under evaluation or produce an increase in tissue temperature. Most tissues (i.e., skin) are relatively opaque, as they contain matters that refract (scatter) light in various and random directions. Excluding blood within large vessels, blood itself constitutes only a small fraction of tissue volume, and most light scattering is due to small particles and stationary tissue elements. Only moving parts in the sample volume (blood cells) will cause a Doppler shift in the light frequency. The scattering angles and red cell velocities are variable, and they can be determined only in a statistical sense.22 Therefore, the LDF signal is a stochastic representation of the number of cells in the sample volume multiplied by their velocities. The technology of LFL using coherent laser light overcomes some problems observed with other optical and nonoptical methods used to record skin blood perfusion. The helium–neon laser, which emits a red light and is used in many LDF instruments, detects very small changes in the wavelength of the laser light as a result of red blood cell movements, well below the resolution of the optical spectroscope.1,2 The frequency distribution of the signal is defined by computerized spectral analysis of the output that produces a power spectrum with a separation between the noise and the signal due to blood cell motion. Some components of the LDF signal are due to external biological or instrumental elements (e.g., vibration) and some to internal (mainly electronic noise) factors. The LDF photodetector signal contains all the Doppler frequencies arising from the laser interaction with moving particles in the tissue and static components. Elaborate electronic processing and filtering are needed to transform the LDF signal into a physiologically reproducible, meaningful, and useful parameter so that the LDF output varies linearly with the blood flow within the sample volume. Considering the LDF output, as seen above, the total power in the signal depends on the number of the moving particles producing a laser Doppler shift of frequency. Therefore, an LD flowmeter must be capable of measuring the mean frequency shift in the signal. Some LD flowmeters digitize the signal and analyze it with a fast Fourier transform, from which the mean frequency shift can be measured. Other systems employ analog signal processing circuitry. Electronic components are needed to normalize the signal and to compensate for noise. In most equipment the noise level remains relatively stable, and theoretically it can be extracted from the LDF output by subtracting a constant offset. Different technical solutions have been applied in low noise lasers or laser diodes.4,5,13,14 This simple preliminary outline of LDF technology may be useful to comprehend the basic concepts. More specific technical information and details on the structure of LD flowmeters can be found in other reports.1,2,5,6
82.3 TERMINOLOGY The output from the LDF is referred to as flux instead of flow. This may cause some confusion. The difference in the terminology is explained by an example reported by Almond.2 Replacing, by hypothesis, blood with saline, which does not contain particles generating Doppler shift, a method that measures volume flux (flow) of the fluid, such as venous occlusion plethysmography, would give approximately the same value of flow. In these conditions an LD flowmeter would give a zero output, as there are no scattering particles to produce a Doppler shift. However, it is reasonable to suppose that in most physiological and clinical situations. even with a variable relationship — due to the fact that blood cells are not homogeneously distributed — LDF volume flux and flow have a good correlation.
82.4 CALIBRATION Many studies demonstrate a good correlation between LDF measurements and other blood flow measurements,2,13,23 while some other studies have shown an irregular correlation between LDF and isotopic methods — e.g., xenon clearance,7 where arteriovenous shunts are present8 or in the bone.10 The major problem at this stage is that there is no gold standard against which LDF can be compared, particularly for skin flow. Also, for the medical operator and vascular technologist, it is still confusing to consider that any single LDF instrument gives its own values, and it is difficult to compare results obtained with an instrument to results obtained with a different flowmeter. Therefore, relative changes (i.e., before and after treatment, after thermal or postural changes, or after arterial occlusion to cause reactive hyperemia) are clinically better accepted than absolute flux measurements. The validity of a universal calibration does not appear possible at the moment, since in different tissues several factors (e.g., light penetration, diffusion, and reflection) make the calibration relevant only to that single tissue. Even with these limitations LDF has many positive aspects that have progressively enlarged its fields of application from experimental and clinical physiology to clinical practice.
82.4.1 THE ZERO LDF CALIBRATION BIOLOGICAL ZERO
AND THE
Flux zero calibration is obtained by positioning the probe against a white surface. An instrumental zero is different from the biological zero, which can be obtained from the skin9–20 by complete occlusion of the arterial supply (Figure 82.1). When blood flow in the skin is completely abolished, the LDF signal decreases to 20 to 50% of the
Laser Doppler Flowmetry: Principles of Technology and Clinical Applications
Arterial occulsion
Biological zero
Instrumental zero
FIGURE 82.1 The concept of the instrumental zero (obtained placing the probe against a white, neutral surface) and the biological zero obtained with arterial occlusion.
tissue flux measurement. The relative ratio between the normal resting flux state and the occlusion state values is variable from region to region and also varies in different organs. It has been observed that in a fingertip the ratio is 5 to 7%, and that in closely related areas the biological zeroes are very similar.20 The origin and physiology of the biological zero are not clearly understood. Studies on excised tissue have shown an elevated baseline from the instrumental zero even several hours after excision. The elevated biological zero disappeared after a few days. Freezing tissues immediately abolish the biological zero baseline, lowering the values to a level very similar to the instrumental baseline.20 In situations of inflammation of human skin it has been shown that the biological zero is increased up to approximately 50 to 70%.9 The clinical implications of these findings are still not clear and further studies in this field are essential. However, in the practical clinical evaluation of limbs, particularly in low-perfusion states, it is recommended that each measurement should be associated with an estimation of the zero baseline and of the biological zero.
82.5 PROBLEMS OF LDF IN CLINICAL PRACTICE LDF is noninvasive and does not interfere with the microcirculation when measuring local blood flow.22 LDF is also particularly useful, as it produces a continuous output that can be used for prolonged monitoring of tissue viability — e.g., after plastic surgery, to record skin flow perfusion during sleep or in newborns, etc. — a technique that is impossible to obtain with any other noninvasive or invasive technique. LDF monitoring is also relatively stable, reproducible, and easy to learn to apply. The technology
711
of clinically usable LDF instruments is continuously improving. Multiple-channel systems capable of measuring distant tissue areas at the same time are now produced by different manufacturers. LDF imagers20 and multiwavelength systems2 are currently being developed and may be used for more accurate and interesting physiological clinical applications. LDF has been extensively used in clinical physiological evaluations and in clinical practice in humans. The technique is easy, noninvasive, and has become very popular, but at times the interpretation of clinical results has been uncritical.9 LDF, as seen above, measures minute particle motion. In living tissue such particles include mainly blood cells, which will constitute the major component of the LDF output. However, it is obvious that in situations of altered microcirculation or alteration of the blood hematocrit (i.e., in inflammation or leukemia), the number of particles can change, determining a different output. The LDF signal can be considered a stochastic representation of the motion of all particles in the sample volume. This measurement cannot be a physiologically correctly defined flow, although it has been shown that LDF values in certain tissues and conditions are proportional and closely related to flow9 measured with isotopic methods. Also, in the skin the precise sampling volume is not always easy to determine and volume flow cannot be measured. Therefore, the LDF output signal recorded has been defined flux, but possibly the most correct expression in clinical application is relative perfusion units.9 Other terms used to express output signal, such as volt, millivolt, and arbitrary units, need to be unified. This appears to be difficult at the moment, as there is no dialogue concerning a common standard among manufacturers. However, when referring to flow in certain contexts and considering the above limitation, the term flow may be used to express a concept that is more familiar to most physicians.
82.5.1 DEPTH OF MEASUREMENTS THE SKIN
AND
VOLUME
IN
The sensitivity of the reflected LDF measurements decreases exponentially with the distance from the probe.22 It has been estimated that the theoretical average depth of sampling in the skin is 0.14 mm, but in artificial models it has been estimated that it is 1.5 mm. However, these values have little importance in clinical applications. There is also evidence that in intestinal models the measuring depth can be greater than 6 mm and that by placing a mirror on the opposite intestinal wall the signal output can be increased from 85 to 100%. The above observations indicate that the measuring depth is not a fixed value for the skin, but most probably a continuous variable, as in other tissues.
712
Handbook of Non-Invasive Methods and the Skin, Second Edition
Capillary flow 10%
LD LD Subpapillary flow 90%
Spacer
NF
FIGURE 82.2 The sampling volume of a standard laser Doppler (LD) flowmeter includes the more superficial nutritional flow (NF) and the deep thermoregulatory capillary layer.
As reported by Fagrell,9 the microcirculation — particularly skin microcirculation — can be considered divided in a small, more superficial, thin layer characterized by nutritional capillaries and in a deeper layer with thermoregulatory vessels. Figure 82.2 shows that the nutritional capillaries are the most superficial ones (0.1 to 0.05 mm from the skin surface) and in normal conditions supply only a very small amount of skin blood (5 to 10%). In the subpapillary, mostly thermoregulatory, bed (0.05 to 2.0 mm in depth) the most common (95%) vessels are venules, with only a small portion of arterioles. In this bed at least 95% of skin blood flow can be responsible for the LDF output, while in the nutritional bed only a small percentage of flow (5 to 10%) contributes to the global output (Figure 82.3). Therefore, LDF measurements in the skin measuring 1 to 2 mm of skin depth record a signal in which the thermoregulatory flow component is predominant (usually >89%), and the nutritional flow is a small component. New microprobes and new LDF flowmeters may reduce the measuring depth and selectively evaluate the most relevant superficial nutritional capillaries, but so far clinically relevant data are not available. Using 1-mm-thick plastic spacers with the same reflectance properties and light transmission characteristics of the skin, it is theoretically possible to evaluate the most superficial skin flux. However, the more distal LDF signal seems to contain a high proportion of noise components, and it appears to be weaker. Early results from our study in which skin spacers are used24 indicate that there is a significant important difference in skin flux values and in flux responses, i.e., venoarteriolar response, flow increase with local temperature increase, etc. This
FIGURE 82.3 The percentage of nutritional flow in the LDF tracing is possibly less than 10% as most of the signal component is due to the non-nutritional subpapillary flow.
difference is noted both with and without the spacers, but with a different, not correlatable extent in low-perfusion microangiopathy — peripheral vascular disease, hypertension associated with vasospasm, and Raynaud’s disease and phenomenon. However, in venous disease and diabetic microangiopathy (these conditions can be defined as highperfusion microangiopathy) the reflex responses and skin flux measurements are reduced with the spacers, but comparable and parallel to measurements obtained without them. This possibly indicates that the differentiation between the nutritional and the thermoregulatory component is more important in states of low-perfusion microangiopathy. In low-perfusion microangiopathy the increase in skin flux as measured by LDF, i.e., following administration of a vasoactive drug or superficial revascularization, may only reflect an increase in the thermoregulatory, nonnutritional flow. This often could be irrelevant in the evaluation of skin flow changes due to the treatment, and produce misleading conclusions in relation to the healing or development of skin necrosis.
82.6 VASOMOTION As LDF measures flux continuously, flux motion can also be easily evaluated. During LDF monitoring in normal conditions the skin vessels in the microcirculation fill with blood rhythmically as a consequence of pressure and flow changes due to cardiac action respiration and vasomotion. The LDF output therefore shows a continuous variation and the sample volume varies continuously. Flux motion patterns are markedly changed in patients with peripheral vascular disease, who frequently present a high-frequency flux motion component. The prevalence of high-frequency motion waves is significantly increased in low-perfusion states, and it is proportionally more evident with increased levels of ischemia. The relationship between the presence of high-frequency flux motion waves at the forefoot has been evaluated by Hoffman et al.11 before and after percutaneous transluminal angioplasty. In successfully treated patients, a significant decrease in high-frequency flux waves was observed after angioplasty. However, the
Laser Doppler Flowmetry: Principles of Technology and Clinical Applications
713
Scanner
Laser Processor Detector
Computer
Plotter
Averaging Profile Subtraction Single values
FIGURE 82.4 A diagram of the LDF scanning system.
prevalence of such waves after angioplasty indicated that high-frequency waves are related to severe chronic ischemia. Different patterns of vasomotor waves were detected by Hoffman et al.11 in different conditions of perfusion. Flux motion in normals was characterized by low-frequency and pulsatile flux waves. Occasionally, additional high-frequency wave components appeared in the recording. Flux motion patterns, in severe ischemia, showed almost no pulsatile flux waves, whereas highfrequency waves were frequently observed in more severe ischemia. In severe cases of ischemia, no flux motion was observed. By contrast, patients with intermittent claudication showed variable patterns (excluding the last pattern with no vasomotion). Therefore, using frequency analysis, it appears possible to qualitatively differentiate degrees of ischemia. However, while there is little doubt that the alteration of vasomotion in low- or high-perfusion states is clinically relevant to indicate microcirculatory disturbances, no definite clinically usable answer has been provided about the analysis of vasomotion. Sophisticated frequency analysis, not commercially available with the instrumentation but usable in postprocessing with the signal output, is needed to extract clinically meaningful data from the analysis and limits so far the application of the method. There is, however, little doubt that qualitative analysis of vasomotion may offer new and interesting concepts to evaluate microcirculatory disturbances.
82.7 NEW TECHNOLOGY LD imaging is a new method of evaluating skin perfusion (other tissue can be also studied). A diagram of the system is shown in Figure 82.4. The multiple skin sampling produces a color perfusion image of the tissue under evaluation and a perfusion profile (Figure 82.5). The color-coded perfusion scale clearly indicates absence of perfusion in the second finger from the bottom in the image due to occlusion with a rubber band (dark blue). The system has been developed very recently, and some clinical application of the system appears very promising.
82.8 CONCLUSIONS The use of LDF is progressively increasing in physiological, pharmacological, and practical clinical evaluation of vascular disease. The monitoring of the effects of treatment on the microcirculation appears to be one of the most promising fields of application of LDF. The technical development of LDF combined with extensive clinical research application and frequency analysis systems will make this method one of the future’s most interesting noninvasive fields of investigation in vascular disease. At this stage, there are some problems concerning standardization of the different systems and calibration problems. It would be useful to define a common standard so that measurements in different centers and different instruments could be comparable, if not universal.
714
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 82.5 An example of LDF scanning. The low flow in the second finger from below is due to occlusion with a rubber band and is indicated in the color-coded perfusion scan by the dark blue color.
REFERENCES 1. Almond NE, Jones DP, Bowcock SA, and Cooke ED. A laser Doppler blood flowmeter used to detect thermal entrainment in normal persons and patients with Raynaud’s phenomenon. In Practical Aspects of Skin Blood Flow Measurement, Spence VA, Sheldon CD, Eds. Biological Engineering Society, London, 1985, p. 31. 2. Almond NE. Laser Doppler flowmetry: instrumentation theory and practice. In: Belcaro G, Hoffman U, Nicolaides AN, Bollinger A, Eds. Med-Orion Publishing Co., in press. 3. Almond NE, Jones DP, and Cooke ED. Noninvasive measurement of the human peripheral circulation: relationship between laser Doppler flowmeter and photoplethysmograph signals from the finger. Angiology 39: 819, 1988. 4. Boggett D, Blond J, and Rolfe P. Laser Doppler measurement of blood flow in skin tissue. J Biomed Eng 7: 225, 1985. 5. Bonner RF and Nossal R. Model for laser Doppler measurements of blood flow in tissue. Appl Opt 20: 2097, 1981. 6. Bonner RF, Clem TR, Bowen PD, and Bowman RL. Laser Doppler continuous realtime monitor of pulsatile and mean blood flow in tissue microcirculation. In Scattering Techniques Applied to Supramolecular and Nonequilibrium Systems, Nato Advanced Study Institute, Series B, Physics, Chen S, Chu B, Nossal R, Eds. Plenum Press, New York, 1981, p. 685. 7. Borgos JA. TSI’s LDV blood flowmeter. In Laser Doppler Blood Flowmetry, Shepherd A, Oberg P, Eds. Kluwer Academic Publ., Boston, 1990, p. 73.
8. Engelhart M and Kristensen JK. Evaluation of cutaneous blood flow responses by 133Xenon washout and a laser Doppler flowmeter. J Invest Dermatol 80: 12, 1983. 9. Fagrell B. Problems using laser Doppler on the skin in clinical practice. In Belcaro G, Hoffman U, Nicolaides AN, Bollinger A, Eds. Med-Orion Publishing Co., in press. 10. Hellem S, Jacobsson LS, Nilsson GE, and Lewis DH. Measurement of microvascular blood flow in cancellous bone using laser Doppler flowmetry and 133Xe clearance. Int J Oral Surg 12: 165, 1983. 11. Hoffman U, Seifert H, Baider E, and Bollinger A. Skin blood flux in peripheral arterial occlusive disease. In Laser-Doppler Flowmetry Experimental and Clinical Applications, Belcaro G, Bollinger A, Franzeck U, Hoffman U, Nicolaides AN, Eds. Med-Orion Publishing Co., in press. 12. Holloway GA and Watkins DW. Laser Doppler measurement of cutaneous blood flow. J Invest Dermatol 69: 306, 1977. 13. Johnson JM, Taylor WF, Shepherd AP, and Park MK. Laser Doppler measurement of skin blood flow: comparison with plethysmography. J Appl Physiol Respirat Environ Exercise Physiol 56: 798, 1984. 14. Nilsson GE, Tenlan T, and Oberg PA. Evaluation of a laser Doppler flowmeter for measurement of tissue blood flow. IEEE Trans Biomed Eng 27: 597, 1980. 15. Obeid AN, Boggett DM, Barnett NJ, Dougherty G, and Rolfe P. Depth discrimination in laser Doppler skin blood flow measurement using different lasers. Med Biol Eng Comput 26: 415, 1988. 16. Obeid AN, Barnett NJ, Dougherty G, and Ward G. A critical review of laser Doppler flowmetry. J Med Eng Technol 14: 178, 1990.
Laser Doppler Flowmetry: Principles of Technology and Clinical Applications
17. Riva C, Ross B, and Benedek GB. Laser Doppler measurement of blood flow in capillary tubes and retinal arteries. Invest Ophthalmol 11: 936, 1972. 18. Stern MD. In vivo evaluation of microcirculation by coherent light scattering. Nature 254: 56, 1975. 19. Stern MD, Lappe DL, Bowen BD, Chimosky JE, Holloway GA, Keiser HR, and Bowman RL. Continuous measurement of tissue blood flow by laser Doppler spectroscopy. Am J Physiol 232: H441, 1977. 20. Tonneson KH and Pederson LJ. Laser Doppler flowmetry: problems with calibration. In Belcaro G, Hoffman U, Nicolaides AN, Bollinger A, Eds. Med-Orion Publishing Co., in press.
715
21. Wardell K, Jakobbson A, and Nilsson GE. A Laser Doppler Imager for Microcirculatory Studies. Paper presented at the 1st European Laser Doppler Users Meeting, Oxford, March 1991. 22. Weis GH, Nossal R, and Bonne RF. Statistics of penetration depth of photons re-emitted from irradiation tissue. J Mod Opt 36: 349, 1989. 23. Winsor T, Haumschild DJ, Winsor DW, Wang Y, and Luong TN. Clinical application of laser Doppler flowmetry for measurement of cutaneous circulation in health and disease. Angiology 10: 727, 1987. 24. Belcaro G and Nicolaides AN. Article in preparation.
83 Laser Doppler Imaging of Skin Karin Wårdell Department of Biomedical Engineering, Linköping University, Linköping, Sweden
CONTENTS 83.1 Introduction............................................................................................................................................................717 83.2 Object.....................................................................................................................................................................717 83.3 Methodological Principle ......................................................................................................................................718 83.3.1 Operating Principle and Instrumentation ..................................................................................................718 83.3.2 Data Analysis .............................................................................................................................................718 83.3.3 Design of a Measurement Procedure ........................................................................................................719 83.4 Sources of Error.....................................................................................................................................................720 83.4.1 Temporal Changes in Perfusion and Movement Artifacts........................................................................720 83.4.2 Distance, Angular, and Reflection Errors..................................................................................................720 83.5 Correlation with Other Methods ...........................................................................................................................721 Acknowledgment.............................................................................................................................................................721 References .......................................................................................................................................................................721
83.1 INTRODUCTION Skin blood flow is an important parameter to record and assess in a large number of clinical and experimental settings. Peripheral vascular disease often manifests itself as a disturbance in cutaneous microcirculation, while an early result of agents irritating the skin is an elevation in its perfusion. Measurement of skin blood flow, therefore, constitutes an important diagnostic procedure in the vascular laboratory as well as in the evaluation of consumer products and potential skin irritants. Methods for measuring skin blood flow should preferably be noninvasive, analytical, versatile, easy to use, and cost-effective. The noninvasiveness should, if possible, also be extended to imply that the measuring device does not have to be in physical contact with the tissue, because even the weakest external stimuli may disturb the flow conditions of the microvascular network under study. The methods should be analytical in the sense that the recorded blood flow signals may be stored for later analysis by means of dedicated software packages. Versatility implies that the same device should be applicable to studies of microvascular perfusion under clinical as well as laboratory conditions. Ease of use and cost-effectiveness are important features, especially in clinical situations where the measurements need to be done on a routine basis. Very few of the methods described in the literature over the years fulfill all these requirements. Laser Doppler
flowmetry, laser Doppler perfusion monitoring (LDPM),1 and laser Doppler perfusion imaging (LDPI)2 influence the microvascular network only to a minimal extent during measurements and are possible to apply to studies of many tissues of the body. Since these technologies are also easy to use and supported by analytical software packages, they have become increasingly important for studies of microcirculation both in the laboratory and in clinical settings.
83.2 OBJECT Skin blood flow generally possesses both substantial temporal and spatial variations. The temporal variations can be rhythmic in nature or show a more fluctuating and stochastic pattern.3 Depending on the architecture of the underlying microvascular network, skin blood flow shows a characteristic granular speckle pattern, and large variations in perfusion can be demonstrated even at adjacent skin sites.4 In addition, the blood flow in the skin is generally known to be compartmentalized. The superficial capillaries are perfused by slow-speed red cells that supply the tissue with oxygen and nutritive substances and remove waste metabolites. Deeper-lying arteriovenous anastomoses take an active part in body temperature regulation, while small arteries and veins constitute routes for the supply and drainage of blood. Taking all this into account, the ideal method for assessing skin perfusion should be able to capture both 717
718
Handbook of Non-Invasive Methods and the Skin, Second Edition
Scanner head Output signal
Laser beam Backscattered light
Signal processing
Computer
Light interaction with tissue Two-dimensional flow map
FIGURE 83.1 Operating principle of the LDPI system.
the temporal and spatial variations in skin blood flow as well as have the potential to separate the signals generated by the different compartments of the microvascular network. Conventional LDPM can readily track fast changes in perfusion, but the small measuring volume prevents assessment of the spatial variability in skin blood flow. In order to overcome this limitation, LDPI was developed in the late 1980s. With this method it is possible to create two-dimensional flow maps of a specific tissue and to visualize the spatial variation of its perfusion.
83.3 METHODOLOGICAL PRINCIPLE 83.3.1 OPERATING PRINCIPLE INSTRUMENTATION
AND
The laser Doppler perfusion imager is a data acquisition and analysis system that generates color-coded images of the tissue perfusion. Two different systems have been presented in the literature and are also available on the market today. One of these systems utilizes a continuously moving laser beam,5 and the other one a stepwise scanning beam.2 The latter will be referred to and described in more detail in the following sections. The optical scanner guides a low-power laser beam (1 mW, 670 nm) to the tissue surface. At each measurement point, the laser light interacts with the skin and its microvasculature. After interaction with the moving red blood cells, the light becomes spectrally broadened due to the Doppler effect. A fraction of this Doppler-broadened light is backscattered and detected by a photodetector positioned in the scanner head. The instantaneous light intensity is converted into an electrical signal and processed to form an output value proportional to the perfusion, defined as the product of the average blood cell speed
and concentration. A detailed description of the laser Doppler theory is found in Nilsson et al.6 For each measurement position, the perfusion value is stored in the computer memory for further signal processing, image generation, and data analysis. In addition to images, data can be captured in the duplex mode.7 Duplex LDPI makes possible the recording of time traces over a small measurement area, including up to 16 measurement sites. With this feature, a combination of spatial and temporal components of the microvascular blood flow can be studied. A schematic overview of the operating principle is presented in Figure 83.1.
83.3.2 DATA ANALYSIS During a measurement procedure, the total light intensity (TLI) and the Doppler components are sampled and stored separately. The TLI value is used by the system to automatically determine whether the light level is adequate enough for recording of a Doppler value and to display a photographic gray-scale image of the measurement area. In practice, this arrangement makes discrimination between the object and the background possible (assuming that a light-absorbing material is used as the background). During a measurement procedure each captured perfusion value is immediately color coded and an image is continuously generated on the monitor. To convert the set of perfusion values into more quantitative parameters, data analysis functions such as perfusion value profile generation, selection of regions of interest (ROIs) for statistical calculation, and image subtraction are incorporated as an integral part of the system. Figure 83.2 exemplifies this by a recording from a skin tumor.
Laser Doppler Imaging of Skin
719
2
FIGURE 83.2 LDPI recording from a skin tumor. Left perfusion image and right photographic image of the tumor. (a)
(b)
(c)
0.98–7.10 V
0–10 V
0–5 V
was applied. The same image is presented with three different fixed color scales, namely, the actual minimum and maximum values (0.89 to 7.10 V) of this image (A), the entire range of the system (0 to 10 V) (B), as well as the range 0 to 5 V (C). In D through F the corresponding interpolated images are displayed. Images recorded by the LDPI system may also be converted to standard formats such as TIFF and ASCII for further image analysis and statistical calculations in software packages of individual preference.
83.3.3 DESIGN (d)
(e)
(f )
FIGURE 83.3 The same image presented in three individual color ranges (0.89 to 7.10 V, 0 to 10 V, and 0 to 5 V) (A, B, and C, respectively). The corresponding interpolated images are presented below the corresponding color ranges (D to F).
Figure 83.3 demonstrates a recording performed on the dorsal side of the hand 15 min after a vasodilatating cream
OF A
MEASUREMENT PROCEDURE
Before a measurement is initiated it is advisable to establish a well-organized protocol. Such a protocol simplifies the analysis and evaluation, which often take place at several points in time following image capturing. The test subject should be sitting or laying down in a comfortable position during the entire measurement procedure. Since the ambient temperature is known to have a substantial influence on skin blood flow, it is recommended that this be recorded. When imaging of the skin perfusion of an
720
Handbook of Non-Invasive Methods and the Skin, Second Edition
extremity is to be performed, this extremity should preferably be attached to, e.g., a sandbag or an arm or leg rest with tape or Velcro® straps, in order to keep the tissue in the same position throughout the scanning procedure. Treat the skin area under investigation carefully, and avoid scratching or applying unintentional pressure or mechanical stimuli, since even the slightest stimulation may substantially influence the perfusion. In order to be able to differentiate the backscattered light as either tissue or background, the object should be placed on a light-absorbing material. To simplify the identification and orientation of the perfusion image, it is useful to mark at least two to three reference points on the skin. If the measurement situation allows, use either a black or green ink pen, dark paper, or tape that can be removed after the scanning procedure is completed. When the laser beam hits the dark background or markers, the light is absorbed instead of being backscattered to the detector. The markers will appear as nonperfusion pixels in the image. The spatial resolution of the captured image depends on the distance between the scanner head and the tissue surface as well as on the system parameter resolution, which can be set by the operator in the software. An image-capturing procedure starts by positioning the lower side of the scanner head in parallel with the tissue surface of interest. By keeping the scanner head parallel to the tissue surface, geometrical distortion of the image is reduced to a minimum. Some distortion, however, is inevitable if imaging is performed on a surface with sharp curvatures. When all parameters are set, the extension of the tissue area to be imaged may be marked by moving the beam along its boundaries. This procedure indicates the area to be imaged and allows the operator to adjust the scanner head or any of the parameters before a measurement is started. To reduce optical interference with the laser light to a minimum, direct ambient light should be avoided. Excessive ambient light levels may render the light-absorbing reference markers placed on the skin invisible in the image. Recordings of dark lesions or pigmented skin perfusion may require a reduced distance between the scanner head and tissue in order to increase the backscattered light level to above the threshold level for tissue–background discrimination. In summary, the most important steps in preparing a successful image of skin tissue are: 1. Use a well-designed protocol. 2. Position the test subject such that gross movement artifacts can be avoided. 3. Use light-absorbing material as the background. 4. Mark reference points on the skin surface. 5. Position the scanner head in parallel with the skin surface.
6. Mark a measurement area and adjust the scanner head if necessary. 7. Protect the scanner head from ambient light. Guidelines for visualization of cutaneous blood flow by laser Doppler imaging have been made available through the European Union 5th Framework Program, denoted Hirelado.8
83.4 SOURCES OF ERROR 83.4.1 TEMPORAL CHANGES IN PERFUSION MOVEMENT ARTIFACTS
AND
In LDPI it is assumed that the perfusion of the tissue to be imaged is stationary and does not show any temporal variations during the time required to capture an image. If this is not the case, temporal variations will manifest themselves as a false spatial heterogeneity in tissue perfusion. Rhythmical variations in tissue perfusion such as vasomotion or a temporary reduction in flow, due to, e.g., taking a deep breath, generally show up as isolated stripes of perfusion values different from the perfusion in adjacent spots. Isolated gross movements of the tissue will likewise appear as stripes of falsely elevated and sometimes saturated values. To avoid such artifacts, it is important that the test subject sit as still as possible during the scanning procedure, or in the case of animal experiments, to apply anesthesia to prevent muscle contraction and shivering. Rhythmical movements of the whole tissue caused by breathing may give rise to a periodic pattern in the image, which is partly caused by tissue movement artifacts and partly by respiratory-related changes in tissue perfusion. The respiratory tissue movement artifacts may be particularly substantial in small animals, and a careful selection of skin areas that are only minimally affected by these movements is recommended.
83.4.2 DISTANCE, ANGULAR, ERRORS
AND
REFLECTION
The system has been designed to be virtually independent of the distance between the skin surface and the scanner head. The dependence of the angle between the light beam incident on the detector surface and a line perpendicular to the detector surface for the calculated perfusion value is automatically compensated for by the system for each measurement site. Surfaces with a pronounced curvature, however, have a tendency to scatter away the light due to pure surface reflection from the object, at least at the image boundaries. A small fraction of the light beam is scattered directly on the surface of the skin due to a mismatch in the refractive indexes of air (n = 1) and tissue (n = 1.5). This surface reflection is, however, generally
Laser Doppler Imaging of Skin
limited to about 5% in normal skin. If the skin is covered by an optical semitransparent material, images of the underlying tissue can still be captured, but the sensitivity is generally reduced due to scattering and absorption in the material or reflections from the surface. Imaging of skin surfaces immersed in water, where the water has been thermostated to a well-defined temperature, can be performed. A geometrical distortion due to the differences in the refractive indexes of water and air is, however, inevitable. Care must also be taken to ensure that the water surface is kept still. If it is not, the light reflected from the water surface may include Doppler components that are difficult to separate from perfusion-generated Doppler shifts in the final perfusion image.
83.5 CORRELATION WITH OTHER METHODS LDPM and LDPI use a similar signal processing algorithm for the calculation of the perfusion value. LDPM, however, is intended for continuous recordings at a single site, whereas LDPI records the perfusion within a specific tissue area. Combining the two methods therefore facilitates studies of both the temporal and spatial variations in skin perfusion. Thus, it is not surprising that Sefalian et al.9 as well as Harrison and colleagues10 have been able to demonstrate a good agreement between the two methods when studying skin perfusion. This agreement was demonstrated despite the spatial variations that exist in normal skin perfusion between adjacent skin areas. However, if the LDPM probe is moved to an adjacent site, the slope of the LDPI/LDPM curve can be expected to change due to the heterogeneity in skin perfusion. This has been demonstrated by several studies.3,9 By the use of biopsy, and a special probe holder for the generation of LDPM topographic maps, Braverman et al.11 correlated different LDPM perfusion patterns with vessel type. Furthermore, this LDPM mapping technique has been compared to LDPI with the modification of using a longer sampling time at each measurement site than in the ordinary setup.12 Perfusion values recorded from ventral forearm skin areas with consistent high and low perfusion coincided when measurements were made by both systems. Harrison and coworkers10 also compared LDPI with thermographic mapping of the skin and found that the regional temperature profile resembled the perfusion values, but the extreme heterogeneity picked up as large pixel-to-pixel variations by LDPI could not be demonstrated by thermography. Thermography, however, records a parameter (skin temperature) that is only indirectly related to skin perfusion, while LDPI directly senses the speed and concentration of the blood cells in the microvascular network. Diverging results may therefore be expected, especially when measurements are made at the
721
fingertips or over ulcers with a high evaporative water loss that significantly reduces the tissue temperature. Since its introduction, LDPI has been used in a wide range of skin applications. Some examples are characterization of the cutaneous axon reflex,13 irritant and allergy patch testing,14,15 investigation of perfusion pattern in basal cell carcinoma treated by photodynamic therapy,16 burn treatment,17 evaluation of UVB reactions in phototesting,18 and study of perfusion in leg ulcers.19
ACKNOWLEDGMENT The author thanks Michail Ilias, Ph.D., at the Department of Biomedical Engineering for invaluable help by preparing the manuscript and illustrations.
REFERENCES 1. Nilsson, G.E., Tenland, T., and Öberg, P.Å., Evaluation of a laser Doppler flowmeter for measurement for tissue blood flow, IEEE Trans Biomed Eng, 27, 597, 1980. 2. Wårdell, K., Jakobsson, A., and Nilsson, G.E., Laser Doppler perfusion imaging by dynamic light scattering, IEEE Trans Biomed Eng, 40, 309, 1993. 3. Tenland, T., Salerud, G., and Nilsson, G.E., Spatial and temporal variations in human skin blood flow, Int J Microcirc Clin Exp, 2, 81, 1983. 4. Braverman, I.M. and Schechner, J.S., Contour mapping of the cutaneous microvasculature by computerized laser Doppler velocimetry, J Invest Dermatol, 97, 1013, 1991. 5. Essex, T.J. and Byrne, P.O., A laser Doppler scanner for imaging blood flow in skin, J Biomed Eng, 13, 189, 1991. 6. Nilsson, G.E., Salerud, E.G., Strömberg, T., and Wårdell, K., Laser Doppler monitoring and imaging techniques, in Biomedical Photonics Handbook, VoDinh, T., Ed., CRC Press , Boca Raton, FL, 2003, chap. 15, pp. 1–24. 7. Wårdell, K. and Nilsson, G.E., Duplex laser Doppler perfusion imaging, Microvasc Res, 52, 171, 1996. 8. Fullerton, A., Stücker, M., Wilhelm, K.-P., Wårdell, K., Anderson, C., Fischer, T., Nilsson, G.E., and Serup, J., Guidelines for visualisation of cutaneous blood flow by laser Doppler imaging, Contact Derm, 46, 129, 2002. 9. Seifalian, A.M., Stansby, G., Jackson, A., Howell, K., and Hamilton, G., Comparison of laser Doppler perfusion imaging, laser Doppler flowmetry, and the thermographic imaging for the assessment of blood flow in human skin, Eur J Vasc Surg, 1993. 10. Harrison, D.K., Abbot, N.C., Swanson Beck, J., and McCollum, P.T., A preliminary assessment of laser Doppler perfusion imaging in human skin using the tuberculin reaction as a model, Physiol Meas, 14, 241, 1993. 11. Braverman, M.I., Keh, A., and Goldmintz, D., Correlation of laser Doppler wave patterns with underlying microvascular anatomy, J Invest Dermatol, 3, 283, 1990.
722
Handbook of Non-Invasive Methods and the Skin, Second Edition
12. Wårdell, K., Braverman, I.M., Silverman, D.G., and Nilsson, G.E., Spatial heterogeity in normal skin perfusion recorded with laser Doppler imaging and flowmetry, Microvasc Res, 48, 26, 1994. 13. Wårdell, K., Naver, H.K., Nilsson, G.E., and Wallin, B.G., The cutaneous vascular axon reflex in humans characterized by laser Doppler perfusion imaging, J Physiol, 460, 185, 1993. 14. Fullerton, A., Benfeldt, E., Petersen, J.R., Jensen, S.B., and Serup, J., The calcipotriol dose-irritation relationship, 48 hour occlusive testing in healthy volunteers using Finn chambers, Br J Dermatol,138, 259, 1998. 15. Bjarnason, B. and Fischer, T., Objective assessment of nickel sulfate patch test reactions with laser Doppler perfusion imaging, Contact Derm, 39, 112, 1998.
16. Enejder, A.M., Klinteberg, C., Wang, I., AnderssonEngels, S., Bendsoe, N., Svanberg, S., and Svanberg, K., Blood perfusion studies on basal cell carcinomas in conjunction with photodynamic therapy and cryotherapy employing laser-Doppler perfusion imaging, Acta Derm Venereol, 80, 19, 2000. 17. Kloppenberg, F.W., Beerthuizen, G.I., and ten Duis, H.J., Perfusion of burn wounds assessed by laser doppler imaging is related to burn depth and healing time, Burns, 27, 359, 2001. 18. Ilias, M.A., Wårdell, K., Falk, M., and Anderson, C., Phototesting based on a divergent beam: a study on normal subjects, Photodermatol Photoimmunol Photomed, 17, 189, 2001. 19. Gschwandtner, M.E., Ambrozy, E., Maric, S., Willfort, A., Schneider, B., Böhler, K., Gaggl, U., and Ehringer, H., Microcirculation is similar in ischemic and venous ulcers, Microvasc Res, 62, 226, 2001.
Heat Wash-In and Heat Wash-Out 84 The Technique for Quantitative, NonInvasive Measurement of Cutaneous Blood Flow Rate Per Sejrsen and Mette Midttun Department of Medical Physiology, The Panum Institute, University of Copenhagen, Copenhagen, Denmark
CONTENTS 84.1 Introduction............................................................................................................................................................723 84.2 Object.....................................................................................................................................................................723 84.3 Methodological Principles .....................................................................................................................................724 84.3.1 Physical Principles.....................................................................................................................................724 84.3.2 The Measuring Probe ................................................................................................................................724 84.3.3 Registration and Data Management ..........................................................................................................724 84.3.4 The Wash-In, Wash-Out Model.................................................................................................................724 84.3.5 Calculation of Blood Flow Rates ..............................................................................................................725 84.3.6 Loss of Heat to the Surrounding Air and the Surrounding Tissue...........................................................725 84.4 Sources of Error.....................................................................................................................................................726 84.5 Correlation with Other Methods ...........................................................................................................................727 84.6 Experimental and Clinical Applications................................................................................................................728 84.6.1 Experimental Studies .................................................................................................................................728 84.6.2 Clinical Studies..........................................................................................................................................729 84.7 Recommendations..................................................................................................................................................730 Acknowledgment.............................................................................................................................................................730 References .......................................................................................................................................................................730
84.1 INTRODUCTION Measurement of blood flow rate in cutaneous tissue using heat as an indicator has the serious problem that heat diffuses about 100 times faster than gases. Therefore, heat has the possibility to escape by other routes than transport with the blood stream. The consequence is that the wash-in rates of heat after cooling will be dominated by the fast diffusion transport of heat into the measuring area from the surrounding air and tissue. After heating, the registered wash-out rate will be dominated by the fast diffusion transport of heat from the measuring area to the surrounding air and tissue. To minimize these problems, the measuring probe has been constructed with a cap that is thermostatically controlled to keep the same temperature as the measuring disc
in contact with the skin surface, in order to eliminate the temperature gradient between the measuring disc and the surrounding air and tissue. The cap is located to the outside of the probe directed against the surrounding air. It is in contact with the surface of the cutaneous tissue by a metal ring placed in the circumference of the central measuring disc at a distance of 2 mm.1,2
84.2 OBJECT The purpose of the present chapter is to describe the heat wash-in and the heat wash-out technique for noninvasive, quantitative measurement of blood flow rates in cutaneous tissue in areas without as well as with arteriovenous anastomoses. 723
724
Handbook of Non-Invasive Methods and the Skin, Second Edition
84.3 METHODOLOGICAL PRINCIPLES 84.3.1 PHYSICAL PRINCIPLES Heat has a diffusion coefficient of 10–3 cm2·s–1. This is, as mentioned above, about 100 times faster than for gases. The consequence is that heat has the possibility to escape from the measuring area by diffusion to the surrounding air and the surrounding tissue during the measuring period, with an elimination rate that is much faster than the elimination caused by the blood flow rate. Therefore, it is necessary to minimize the heat uptake from or the heat loss to the surrounding air and tissue, in order to obtain the heat wash-in rate or the heat wash-out rate separately due to the blood flow rate in the tissue area under study. The probe is constructed with a thermostatically controlled cap set to keep the same temperature as the central metal measuring disc by heating or cooling delivered by a Peltier element located in the cap. A thermistor is placed in the cap for registration of the temperature in the cap. The cap is in contact with the skin surface by a metal ring located in the circumference of the measuring disc at a distance of 2 mm from the central disc. By this construction, the thermostatically controlled cap will eliminate the temperature gradient between the measuring site and the surrounding air and the surrounding tissue area. During the primary heating or cooling a metal bridge is placed to give a thermic contact between the cap with the Peltier elements and the central measuring disc. After the primary period of cooling or heating to a steady-state temperature level the metal bridge is redrawn to the cap and the temperature of the cap is set to follow the temperature of the measuring disc.1,2
84.3.2 THE MEASURING PROBE The heat wash-in and heat wash-out probe are presented in Figure 84.1; 20a is the central measuring disc containing a thermistor 45 and 15 is the thermostated cap with thermistor 40. The thermal metal bridge is 35, with the part 35a, which can be brought in contact with the measuring disc by rotation of the finger screw, which is the top of the bridge. The values 25 and 30 are the Peltier elements for cooling or heating, 10 and 50 are cooling ribs for exchange of heat with the surrounding air, and 20b and 55 are made of a plastic material.1,2
84.3.3 REGISTRATION
AND
DATA MANAGEMENT
A 2 to 5°C change in temperature is introduced into the measuring disc and the tissue in contact with the probe by heating or cooling with the Peltier elements located in the cap of the probe via the thermal metal bridge in contact position.
35
10
50 25
30 40 55
15
35a 45 20a
20b
10 mm
FIGURE 84.1 Drawing of the heat wash-in, heat wash-out probe for measuring cutaneous blood flow rates.2 For details, see text.
When a steady-state temperature is obtained in the measuring disc, after a few minutes the thermic bridge is redrawn to the cap and the heating or cooling is interrupted. The cap is thereafter set to follow the temperature of the central measuring disc. The temperature change per time unit of the measuring disc is registered in 5-s periods. The temperature change is registered until the baseline temperature is obtained. The baseline temperature was measured in the steady-state temperature period before heating or cooling, and again after the end of heat washout or heat wash-in. The registered temperature values are corrected for the baseline temperature. A fine adjustment of the baseline temperature value can be made manually in 0.01°C intervals if necessary due to an eventual small change in baseline temperature. The temperature wash-in or wash-out curves corrected for the baseline temperature level follow after a few seconds a monoexponential wash-in or wash-out function. This monoexponential function is caused by the supply or elimination of heat by the blood flowing through the tissue. The initial phase of a few seconds duration is due to a temperature equilibration between the probe and the tissue. The slope of the following monoexponential function is equal to the blood flow rate in the cutaneous tissue. A heat wash-out curve measured on the forearm in a normal human subject is seen in Figure 84.2.2
84.3.4 THE WASH-IN, WASH-OUT MODEL The wash-out system is in analogy with an electric model, where Coulon = joule, voltage = temperature, capacitans = joule/degree Celsius, and ampere = joule/time = watt. Ohm = temperature/watt = 1/(flow rate) CD = heat in the measuring disc RI = temperature fall or increase per time unit due to the initial heat equilibration
The Heat Wash-In and Heat Wash-Out Technique
log T – Tb
10
1
0.1 0
2
4 6 Minutes
8
10
FIGURE 84.2 Heat wash-out from the forearm in a normal human subject after an initial heating of the measuring disc of the probe and the underlying tissue to 43°C.2 The results are plotted as the measured temperature minus the baseline temperature on a logarithmic ordinate scale against the time in minutes on a linear scale on the abscissa.
CC = heat in the skin tissue RF = temperature fall or increase per time unit due to blood perfusion in the cutaneous tissue See Figure 84.3. The temperature, U, of the measuring disc in contact with the skin surface is then U = U0(a1·e–t/T1 + b1· e–t/T2) Analogous electrical model Coulon = joule Voltage = temperature Capacitans = joule/degree Ampere = joule/time Ohm = temperature/watt = 1/flow CD = heat in the probe
RI = initial temp. fall
CC
CC = heat in the tissue
RF = temp. fall due to perfusion
FIGURE 84.3 A presentation of an electrical analog model. For details, see text.
725
where t is the time variable in the wash-in or wash-out process. After a curve resolution the resulting two components have the intercepts a1 and b1 and the rate constants 1/T1 and 1/T2 for the initial and following components, respectively. Just after cessation of the heating an equilibration of heat will take place between the measuring disc and the tissue area under study. This is a fast process conditioned on the very high diffusion coefficient of heat in the tissue and the metal disc. It is observed that this initial fast component has a duration of a few seconds, with a maximum of 10 to 20 s. It can be assumed that CD, the amount of heat in the measuring disc, is small in comparison to CC, the amount of heat in the cutaneous tissue area under study. Furthermore, it can be assumed that RI, the temperature fall per time unit in the cutaneous tissue area in the initial phase due to the diffusional equilibration of heat, is very fast and brief compared to RF, the following monoexponential temperature fall due to the heat wash-out by the blood perfusion. The rate constant of this function, 1/T2, is a direct expression of the elimination rate caused by the blood flow rate in the cutaneous tissue: T2 = RF·(CD + CC)
84.3.5 CALCULATION
OF
BLOOD FLOW RATES
Blood flow rates can be calculated from the rate constant of the monoexponential function (Figure 84.3) using the equation of Kety,5 f = ln2·(T1/2)–1·lambda·100 (ml·(100 g·min)–1), where ln2 is the natural logarithm to 2, lambda is the tissue-to-blood partition coefficient for heat in milliliters per gram of tissue, and T1/2 is the half-time of the monoexponential wash-in or wash-out function. The tissue-to-blood partition coefficient lambda for heat is 0.954 ml·g–1, indicating that the heat capacity for tissue and blood, respectively, is almost equal. For simplification, a lambda value of 1 ml·g–1 can be used.
84.3.6 LOSS AND
HEAT TO THE SURROUNDING AIR THE SURROUNDING TISSUE OF
Measurement of heat elimination during blood flow cessation was performed with a cuff on the upper arm. Furthermore, a lead shield was placed distally to the probe in order to produce a pressure of 40 mmHg to eliminate the blood flow in the venous rete on the forearm. When the pressure in the cuff was above the systolic blood pressure (200 mmHg), the diffusional, non-blood-flow-dependent combined loss of heat to the surrounding air and surrounding tissue was measured. The results given as ml·(100 g·min)–1 for measurements performed on the forearm were an average of 0.9 in seven measurements, range 0.5 to 1.5, and on the pulp
726
Handbook of Non-Invasive Methods and the Skin, Second Edition
of the thumb the results were an average of 2.9 in six measurements, range 2.7 to 3.1.2 Similar measurements with the probe placed on a block of styrene foam gave the average result of 5.2 in six measurements, range 5.1 to 5.3.2 As a consequence of the low values obtained in these measurements, a correction for this error will be unnecessary in most situations. However, as mentioned above, the heat loss in the area under study can be measured during blood flow cessation and a correction performed if necessary.
2. 3.
4.
84.4 SOURCES OF ERROR The basic assumptions of the heat wash-in, heat wash-out method are identical to those of the 133Xe wash-out method.3–5 The assumptions are the following:
5.
70
60 Blood flow rate, heat (ml⋅(100g⋅min)–1)
1. Blood flow rate shall be constant during the registration, and no trauma at the introduction of the indicator must occur; i.e., a steady-state blood flow rate shall be present during the measurement. 2. The indicator must be stable; i.e., it must not be metabolized or bound in the tissue. 3. The indicator must not leave or be supplied to the system by routes other than the flowing blood. 4. Equilibrium for the indicator between the tissue and the flowing blood shall be obtained during the passage of the blood through the tissue. 5. The system shall be linear; i.e., a doubling of the indicator input to the system shall result in a doubling of the response. 6. The system shall be homogeneous. 7. Registration of the slope of the wash-out curve shall be estimated correctly. 8. The partition coefficient between tissue and blood, lambda, shall be known for the indicator used.
20 min after heating with the probe to a given temperature level by the procedure used.1,14 The indicator heat is stable. The heat production in the tissue is relatively very small and constant. The supply of heat or elimination of heat by routes other than the blood is minimized by the use of a thermostatically controlled cap. This is seen from the very low wash-in or wash-out rate values measured during blood flow cessation.2 The fulfillment of the assumption concerning equilibrium for the indicator between tissue and blood during the passage of blood through the tissue is presumably obtained. This is seen from the monoexponential wash-in and wash-out of heat, and from the correlation of the results of the heat wash-out with the 133Xe wash-out method in simultaneously performed experiments on the human forearm. In this region, only nutritive capillary blood flow is present. A correlation diagram is presented in Figure 84.4.1,14 The system has been shown to be linear in experiments on the pulp of the thumb at various temperature levels from 38 to 45°C (Figure 84.6).1,14
50
40
30
20 Subject 1 10
The fulfillment of the above-listed assumptions for the heat wash-in, heat wash-out method is made probable by the following conditions:
0
Subject 2
0
10
20
30
40
50
60
70
Blood flow rate, 133Xe (ml⋅(100g⋅min)–1)
1. The subjects were kept in termoneutral surroundings, and the indicator was introduced to the tissue area under study with heating or cooling by the probe to a steady-state temperature level which was obtained for some minutes. By the 133Xe wash-out method,3,4 it was shown that the blood flow rate remained constant for about
FIGURE 84.4 A comparison of measurements with the 133Xe wash-out method and the heat wash-out method.1,14 The measurements were performed simultaneously and in the same area on the forearm in two normal human subjects. During the measurements blood flow cessation was performed in the venous rete on the forearm by compression with a lead shield. For details, see text.
The Heat Wash-In and Heat Wash-Out Technique
6. The wash-in and wash-out curve for heat shows monoexponential functions for the cutaneous tissue over more than two decades. 7. To make an accurate registration of the slope of the curve, the wash-in and wash-out functions have been followed over a range of about two decades. 8. The partition coefficient, lambda, between tissue and blood for heat was calculated as the ratio of the data for heat capacitance of the tissue and blood, respectively. As the basic assumptions of the heat wash-in and heat wash-out method seem to be fulfilled for all practical purposes, and furthermore, because the results were found closely correlated to results obtained by the 133Xe washout method in simultaneous experiments,3,4 the heat washin, heat wash-out method can be considered a quantitative measure of blood flow rate in the cutaneous tissue. By taking the average value of a heat wash-in and a heat wash-out measurement in the same tissue area and with the same temperature change from the baseline temperature, it is presumably possible to get a blood flow rate equal to that in undisturbed conditions. In areas with arteriovenous anastomoses, as in the thumb pulp, the heat wash-in, heat wash-out method will measure the sum of the nutritive blood flow rate in the capillaries and that in the arteriovenous anastomoses. The blood flow rate in the capillaries in these regions can be measured by the 133Xe wash-out method.3,4 By subtraction of the 133Xe results from the heat wash-in or heat wash-
727
out results, it is possible to get the blood flow rate separately in the arteriovenous anastomoses.
84.5 CORRELATION WITH OTHER METHODS A comparative study has been performed as a simultaneous study on the same cutaneous area on the forearm with the heat wash-out method and the 133Xe wash-out method in two normal human subjects (Figure 84.4).1,14 The equation for the regression line is y = 2.5 + 0.968·x, and the correlation coefficient is 0.986 in the temperature interval from 37 to 45°C. With a Bland Altman6 statistical treatment for assessing agreement between results obtained with the two methods (Figure 84.5), the mean difference is 0.86, with SEM = ±0.27. The SD = 1.22 and the 95% confidence interval is –1.58 to 3.30, all values given in ml•(100 g•min)–1. Despite this, the limits of agreement (–1.58 and 3.30) are small enough to be confident, so that the heat wash-out method can be used in place of the 133Xe washout method for clinical purposes with use of a lambda value of 1.00 (Figure 84.5). If the correct partition coefficient lambda for heat, 0.954 ml·g–1, is used in this statistical treatment, it gives the following values: mean difference = 0.01, SEM = ±0.32, SD = 1.50, 95% confidence interval = –0.31 to 0.33, and the limits of agreement = –2.99 to 3.01. These results shows only a little closer agreement between the results of the two methods than does the use of lambda = 1.0.
Difference in blood flow rate ml⋅(100g⋅min)–1
4 +2 SD 3 +1 SD
2
1
Mean
0 –1 SD –1 –2 SD –2
10
70 Average blood flow rate of the two methods ml⋅(100g⋅min)–1
FIGURE 84.5 A Bland Altman plot6 of the result in Figure 84.4. For details, see text.
728
Handbook of Non-Invasive Methods and the Skin, Second Edition
160 Blood flow rate (ml⋅(100g⋅min)–1)
1
140 120 100 80 60 40 20 0
0.1 0
2
4
6
8 10 Time (min)
12
14
36
16
FIGURE 84.6 Heat wash-out measurements from the pulp of the thumb in a normal human subject after heating the measuring disc of the probe and the underlying tissue to various temperatures.1,14 The curves presented are heat wash-out after initial heating to temperatures in the interval 38 to 45°C, with steps of 1°C. The results are plotted after subtraction of the baseline temperature on a logarithmic ordinate scale with a linear timescale on the abscissa. The crosses indicate the start and end of the calculated regression lines.
84.6 EXPERIMENTAL AND CLINICAL APPLICATIONS 84.6.1 EXPERIMENTAL STUDIES Blood flow rate in the pulp of the thumb measured by the heat wash-out method in a normal subject at various temperatures from 38 to 45°C is shown in Figure 84.6.1,14 The wash-out rates are almost the same in the eight measurements, as the subject is in warm conditions. The blood flow rate is in warm conditions dominated by the blood flow in the open arteriovenous anastomoses. A study with measurements on the pulp of the thumb in two normal human subjects has been performed to compare results obtained by the heat wash-out method, the 133Xe wash-out method on the skinfold between the thumb and the forefinger, and venous occlusion plethysmography on the distal part of the thumb. The results from one of the subjects are shown in Figure 84.7.1,14 The measurements were done at various temperatures from 36 to 45°C. The heat wash-out method shows much higher values than the two other methods. Figure 84.8 shows blood flow rate measured by the heat wash-out method and the 133Xe wash-out method in the thumb pulp in a normal human subject exercising on an ergometer bicycle.7,14 After 2 min of rest follows 7 min of exercise with a moderate load, and finally again resting conditions. Blood flow rate measured by the 133Xe washout method in the cutaneous capillaries does not change during the exercise period. The heat wash-out method
37
38
39 40 41 42 Temperature (°C)
43
44
45
FIGURE 84.7 Blood flow rates in the pulp of the thumb in a normal human subject measured in the temperature interval from 36 to 45°C by the 133Xe wash-out method (triangles), by venous occlusion plethysmography (squares), and by the heat wash-out method (dots).1,14
200
Blood flow rate (ml⋅(100g⋅min)–1)
ΔT = T – Tb (°C, log scale)
10
150
100
50 Rest
Exercise
Rest
0 0
5
10
15
Minutes
FIGURE 84.8 Blood flow rates in the pulp of the thumb measured in a normal human subject with the heat wash-out method. The blood flow rate in the skinfold between the thumb and the forefinger was measured by the 133Xe wash-out method. The measurements were done during rest for 2 min, exercise for 5.5 min, and finally rest for 3.5 min. The cutaneous capillary blood flow rate measured by the 133Xe wash-out method is shown by the lower line.8,14
shows the combined blood flow rate in the arteriovenous shunt vessels and the capillaries. It is demonstrated that the blood flow rate in the shunt vessels is reduced in the
The Heat Wash-In and Heat Wash-Out Technique
729
120
100 Series 1
100
Series 2
ml⋅(100g⋅min)–1
60
40
20
0 h.l.
80 60 40 20
45 cm elev. h.l. 70 cm elev. Position in or above heart level
h.l.
Series 1: Measurements on the pulp of the thumb at various positions. Series 2: Measurements on the pulp of the thumb with the other hand in ice-water.
FIGURE 84.9 Blood flow rate in the pulp of the thumb in a normal human subject during ortostatic maneuvers. Measurements were performed with the hand at the heart level, 45 cm above the heart level, at heart level, 70 cm above the heart level, and finally at the heart level. The lower line shows blood flow rates with the contralateral hand in ice water at heart level, 45 cm above heart level, and at heart level.2
initial 4.5 min of the exercise period. This is followed by an increase presumably triggered by the center of thermoregulation in hypothalamus in order to eliminate heat from the body. Figure 84.9 demonstrates the reduction in blood flow rate in the pulp of the thumb when the position of the hand is changed from heart level to positions 45 and 70 cm above heart level, respectively.2 Furthermore, the near blood flow cessation in the pulp of the thumb, as an effect of placing the contralateral hand in ice water, is demonstrated.2 Figure 84.10 shows the blood flow rate in the ear lobe and the nose in two normal subjects.8,14 By subtraction of the blood flow rate measured by the 133Xe wash-out method from that measured by the heat wash-out method, the blood flow rate in the arteriovenous shunt vessels has been obtained. Heat wash-out measurements on the pulp of the thumb and the first toe showed in 30 subjects (from 3 to 93 years old) a decrease with age.9,14 This is in accordance with the decrease in metabolic rate during the lifetime.
84.6.2 CLINICAL STUDIES The results of clinical examinations of blood flow rate in the pulp of the first toe during ortostatic maneuvers in claudicants, patients with critical ischemia, and normals are shown in Figure 84.11.10,14 Blood flow rate is low in patients with critical ischemia both at the heart level and in a position 50 cm below the heart level. In claudicants
0
Els
Elr
Nls
Nlr
Ells
Ellr
Nlls
Nllr
FIGURE 84.10 Average blood flow rates in the ear lobe, E, on the side of the nose, N, upon sitting, s, and in the recumbent position, r, in two normal subjects, n = 5. The results of measurements with the heat wash-out method are the total height of the columns.10,14 The hashed parts of the columns denote blood flow rate in the capillaries as measured by the 133Xe wash-out method. The white part of the columns denotes blood flow rate in the arteriovenous anastomoses obtained by subtraction of the 133Xe values from the heat wash-out values.
70
60
50 ml⋅(100g⋅min)–1
ml⋅(100g⋅min)–1
80
40
30
20
Norm. Claud. SL Claud. AS Crit. isch. SL Crit. isch. AS
10
0 +50 cm above
0 Heart level
–50 cm below
FIGURE 84.11 The median blood flow rate in the pulp of the first toe at ortostatic maneuvers in normal subjects (crosses); in claudicants, asymptomatic side (open squares); in claudicants, symptomatic side (filled squares); in patients with critical ischemia, asymptomatic side (open circles); and in patients with critical ischemia, symptomatic side (filled circles). The positions were 50 cm above heart level, at heart level, and 50 cm below heart level.11,14
730
Handbook of Non-Invasive Methods and the Skin, Second Edition
blood flow rate is low at the heart level, but it increases with a factor of 1.7, to a normal level, in the dependent position 50 cm below the heart level. Above the heart level the blood flow rate is reduced in both normals and claudicants. After surgical revascularization claudicants showed normal blood flow rate responses to ortostatic maneuvers.11,14 In patients with nonspecific aneurysmal disease of the infrarenal aorta the blood flow rate measured by the heat wash-out method on the pulp of the first toe showed values equal to those found in normal subjects. However, the blood flow rate in the subcutaneous adipose tissue on the forefoot, measured by the 133Xe wash-out method, was in these patients about three times higher than in normals at heart level and during ortostatic maneuvers.12 The results can be interpreted as a general degeneration of the elastic fibers in the arterial vessel walls, reducing the vascular resistence. Blood flow rate in the cutaneous tissue on the thorax was measured by the heat wash-in method in the initial phase just after placing the probe on the skin surface.13 By this procedure the blood flow to the area under study will increase the temperature until a steady-state level. By subtraction of the steady-state temperature from the registered temperatures a monoexponential wash-in function is obtained, and the rate constant of this function yields the blood flow rate in the cutaneous tissue. In this study on six subjects, it was shown that blood flow rate increased during thoracic epidural blockade from 13.6 (range, 10.6 to 14.6 ml•(100 g•min)–1) to 18.4 (range, 13.9 to 24.5 ml • (100 g • min) –1 ) (p < 0.05). During unblocked conditions, local heating to 40°C gave an increase in blood flow rate in five of the six patients, and the result was 27.9 on average (range, 20.8 to 34.6 ml•(100 g•min)–1) (p = not significant). Thus, heating to 40°C yields a blood flow rate that is not significantly different from that obtained during blockade. In a recent study blood flow rate in cutaneous tissue was measured by the heat wash-out method on the forehead in patients with partial obstruction in the common carotic artery. The measurements were done in an area over the medial part of the eyebrow. In this cutaneous area blood is supplied from the internal carotic artery. During entarterectomia it was shown that the blood flow to this region was increased by 18.6% after opening of the external carotic artery and by 81.4% after opening the internal carotic artery. Thus, it seems as if an indirect measure of blood flow rate to the brain can be obtained by using this cutaneous region supplied by the internal carotic artery.
84.7 RECOMMENDATIONS The heat wash-in, heat wash-out method, which is an atraumatic and quantitative method for measurement of cutaneous blood flow rate, seems to present good condi-
tions for use in experimental physiology and pharmacology and in clinical studies. Another possibility is to use the method for calibration of other methods giving qualitative measuring results. This is in accordance with the use of the 133Xe wash-out method for this purpose. The heat wash-in, heat wash-out method has the advantage, in comparison with the 133Xe wash-out method, to have no expense to buy the indicator 133Xe, with intervals of about 2 weeks. Furthermore, the time interval required for a measurement is much shorter, about 5 to 20 min with the heat wash-in, heat wash-out method, compared to about 25 to 90 min with the 133Xe wash-out method. The heat wash-in, heat wash-out method has a great potential for clinical use within the following fields: arterial and venous diseases in the legs before and after treatment, valuation of level for amputation, white fingers, burn and cold injuries, wound healing, skin diseases, and diabetes. Another field is validation of the sympathetic tonus in patients with heart diseases, in supervision of prematures, in patients during surgery, and in intensive care. Furthermore, the possibility for measuring blood flow rate in arteriovenous anastomoses is of special importance in the temperature regulation investigation.
ACKNOWLEDGMENT The authors acknowledge Jens D. Hove, M.D., Ph.D., MA. Phys., for his contribution by creating the analogous electrical model for the heat wash-in, heat wash-out technique.
REFERENCES 1. Midttun, M., Sejrsen, P., and Colding-Jørgensen, M., Heat-washout: a new method for measuring cutaneous blood flow rate in areas with and without arteriovenous anastomoses, Clin Physiol, 16, 259, 1996. 2 Sejrsen, P. and Midttun, M., A Method and an Apparatus for Measuring Flow Rates, International Publication WO 01/43629 A1, published under the Patent Cooperation Treaty (PCT), and U.S. Patent Application 20030139676 A1, http://appft1.uspto.gov/netacgi/nphParser. 3. Sejrsen, P., Measurement of cutaneous blood flow by freely diffusible radioactive isotopes. Methodological studies of the washout of krypton-85 and xenon-133 from the cutaneous tissue in man, Dan Med Bull, Suppl. 18, 9, 1971. 4. Sejrsen, P., The 133xenon wash-out technique for quantitative measurement of cutaneous and subcutaneous blood flow rates, chap. 17.5, ibid. 5. Kety, S.S., Measurement of regional circulation by the local clearance of radioactive sodium, Am Heart J, 38, 321, 1949.
The Heat Wash-In and Heat Wash-Out Technique
6. Bland, J.M. and Altman, D.G., Statistical methods for assessing agreement between two methods of clinical measurement, Lancet, 1, 307, 1986. 7. Midttun, M. and Sejrsen, P., Cutaneous blood flow rate in areas with and without arteriovenous anastomoses during exercise, Scand J Med Sci Sports, 8, 84. 8. Midttun, M. and Sejrsen, P., Blood flow rate in arteriovenous anastomoses and capillaries in thumb, first toe, ear lobe, and nose, Clin Physiol, 16, 275, 1996. 9. Midttun, M., Blood flow rate in arteriovenous anastomoses: from the cradle to the grave, Clin Physiol, 20, 5, 360, 2000. 10. Midttun, M., Sejrsen, P., and Paaske, W.P., Blood flow rate during orthostatic pressure changes in the pulp skin of the first toe, Eur J Vasc Endovasc Surg, 13, 278, 1997.
731
11. Midttun, M., Sejrsen, P., and Paaske, W.P., Peripheral blood flow rates and microvascular responses to orthostatic pressure changes in claudicants before and after revascularisation, Eur J Vasc Endovasc Surg, 17, 225, 1999. 12. Midttun, M., Sejrsen, P., and Paaske, W.P., Is non-specific aneurysmal disease of the infrarenal aorta also a peripheral microvascular disease? Eur J Vasc Endovasc Surg, 19, 625, 2000. 13. Nygård, E., Sejrsen, P., and Kofoed, K.F., Thoracic sympatholysis with epidural blockade assessed by quantitative measurement of cutaneous blood flow, Acta Anaest Scand, 46, 1037, 2002. 14. Midttun, M., Heat-washout: a new method for measuring cutaneous blood flow in areas with and without arteriovenous anastomoses. Physiological and patophysiological examinations, Dan Med Bull, Suppl., 2004.
85
The 133Xenon Wash-Out Technique for Quantitative Measurement of Cutaneous and Subcutaneous Blood Flow Rates Per Sejrsen Department of Medical Physiology, The Panum Institute, University of Copenhagen, Copenhagen, Denmark
CONTENTS 85.1 Introduction............................................................................................................................................................733 85.2 Object.....................................................................................................................................................................733 85.3 Methodological Principles .....................................................................................................................................733 85.3.1 Physical Principles.....................................................................................................................................733 85.3.2 Atraumatic Local Labeling........................................................................................................................733 85.3.3 Registration and Data Management ..........................................................................................................734 85.3.4 The Washout Model...................................................................................................................................735 85.3.5 Calculation of Blood Flow Rates ..............................................................................................................738 85.3.6 Loss of 133Xe from the Skin Surface.........................................................................................................738 85.4 Sources of Error.....................................................................................................................................................739 85.5 Correlation with Other Method.............................................................................................................................739 85.6 Recommendations..................................................................................................................................................739 References .......................................................................................................................................................................740
85.1 INTRODUCTION Measurement of blood flow rates in cutaneous and subcutaneous tissues is of interest in human physiology, pathophysiology, and in control of the therapeutic effect. It is especially of interest in the understanding of the distribution of cardiac output to the skin during test, orthostatic maneuvers, and dynamic exercise, and in thermoregulation. Most of the methods developed for this purpose have been qualitative in nature. The introduction of the 133Xe washout method after epicutaneous labeling has made it possible to measure the cutaneous and subcutaneous blood flow rates quantitatively during atraumatic conditions.1
85.2 OBJECT The purpose of the present chapter is to describe the measurement of cutaneous and subcutaneous blood flow
rates by the washout of 133Xe after atraumatic local epicutaneous labeling using external residue detection.
85.3 METHODOLOGICAL PRINCIPLES 85.3.1 PHYSICAL PRINCIPLES 133
Xe is a radioactive inert gas isotope with a physical half-life of 5.3 days. The radiations emitted by disintegration of 133Xe are x-ray and emission. By a NaI (T1) scintillation detector coupled to a γ spectrometer, it is possible to register the 133Xe γ emission of 81 keV, with an incidence of 35.5%, and the x-ray of about 40 keV, with an incidence of 64.5%, by setting the window to include these two energy peaks. The distance between the 133Xe deposit and the detector shall be kept constant throughout the total period of registration to measure the relative washout rate. The collimation shall be so wide that registration is obtained from the total 133Xe depot area — also when this expands by diffusion to the surrounding tissue area. A 733
734
Handbook of Non-Invasive Methods and the Skin, Second Edition
Mylar membrane
Seen from the side
Adhesive area
Cutis
Mylar membrane Seen from above
Adhesive area
5 cm
FIGURE 85.1 Technique for epicutaneous application of Physiol., 24, 570, 1968. With permission.)
133
Xe gas or
suitable distance between deposit and detector is 15 to 20 cm. This distance will minimize the effect on the counting efficiency of the expansion of the 133Xe deposit by diffusion, which will increase the distance between the detector and the labeled area. Another detector suitable for registration of 133Xe activity is a cadmium telluride (chloride) detector (Cd Te (Cl)).2 As this detector type can be fixed to the region by adhesive plaster, keeping the counting geometry almost constant, it is a portable solution of the registration. It is important to note that the labeled area, by the short distance used with this detector type, shall be either so small that 133Xe cannot leave the counting area by diffusion or so large that a constant concentration is present in an area somewhat larger than the area of registration. The short counting distance used by this detector type has been from 1 to 20 mm, which makes it very important to correct for the elimination rate due to expansion by diffusion, if present. This can be done by subtracting the elimination rate measured during blood flow cessation from that obtained with intact blood flow. As 133Xe is a gas, it is freely diffusible in the tissues, and equilibrium between tissue and blood is obtained for 133Xe during the passage of blood through the tissue. This has been shown in experiments on semi-isolated, autoperfused gastrocnemius muscles in cats, where the 133Xe washout method was compared to the directly measured outflow rate of blood.3 On this basis, 133Xe is a suitable indicator for measurement of blood flow rates, as the 133Xe washout rate is proportional to blood flow rate.
85.3.2 ATRAUMATIC LOCAL LABELING Atraumatic local labeling of the tissue with 133Xe can be done by application of 133Xe gas or a 133Xe in saline
133
Xe dissolved in isotonic saline. (From Sejrsen, P., J. Appl.
solution on the skin surface for a few minutes, e.g., 3 min.4 In practice this is performed by the following technique. A deposit of 133Xe is placed on the skin surface in a chamber formed by the skin surface and a circular gastight Mylar® membrane 3 to 8 cm in diameter and 20 mm thick. The membrane is attached to the skin surface by a ringshaped, 0.7- to 1.5-cm-wide, adhesive membrane with adhesive material on both sides (Figure 85.1).1 The dimension of the central chamber will then be from 1.6 to 5 cm in diameter. A thin injection needle is placed between the Mylar and the adhesive membrane, leading from the outside into the chamber. From a syringe it is then possible to introduce a deposit of 133Xe gas or 133Xe in isotonic saline solution into the chamber. After labeling by diffusion from the deposit on the skin surface into the skin for about 3 min, the deposit is redrawn to the syringe. The membrane with adhesive ring, needle, and syringe is removed, the region is dabbed with a piece of soft tissue paper, and the surplus of 133Xe is blown away.
85.3.3 REGISTRATION
AND
DATA MANAGEMENT
The registration of the 133Xe activity is then performed as a residue detection by external counting in time intervals of, e.g., 20 sec (from 1 sec to 1 min), dependent on the purpose of the measurement and the level of activity. The data obtained are then plotted in a semilogarithmic diagram after subtraction of the background activity. The x axis is time in a linear scale, and the y axis is the activity in a logarithmic scale. The washout of 133Xe after atraumatic local labeling of a skin area shows a biexponential course (Figure 85.2). This is due to diffusion of 133Xe from the cutaneous venous blood out through the walls of the venous vessels during the passage of this bloodstream through veins located in
The 133Xenon Wash-Out Technique for Quantitative Measurement
Application
Counts/minute
105
104
0
30
60 Minutes
90
120
FIGURE 85.2 Washout curve after epicutaneous labeling with 133Xe for 3 min on the lateral side of crus. Solid circles show the registered activity with time. The open circles show the result of graphic curve resolution. The mathematical expression of the two exponentials obtained by the graphic curved resolution are presented in Equation 85.8. The washout curve separate for the cutaneous tissue is constructed by drawing a line parallel to the straight line through the open circles from the top of the registered curve. By subtracting the values given by this line from those of the registered curve the separate curve for the subcutaneous tissue is constructed, here denoted by crosses. The mathematical expressions of these separate washout curves for the two tissues are given in Equation 85.7. (From Sejrsen, P., Circ. Res., 25, 215, 1969. With permission.)
the subcutaneous tissue. 133Xe has about 10 times higher solubility in subcutaneous adipose tissue than in blood. This has the effect that the subcutaneous tissue acts as a zinc for 133Xe, resulting in an accumulation of 133Xe in the subcutaneous tissue with time. This is illustrated in Figure 85.3a and b, showing the distribution of 133Xe in the tissue after atraumatic local labeling by autoradiographic technique. After 2 min of in vivo washout the 133Xe is located almost exclusively in the cutaneous tissue, and after 70 min almost exclusively in the subcutaneous tissue. The result of the very limited transport by diffusion alone without blood flow after 70 min is shown in Figure 85.3c.5 A diffusion of 133Xe directly from cutaneous to subcutaneous tissue over the contact area between these two tissues thus seems of lesser importance than the above-mentioned transport by convection with venous blood flow combined with diffusion out through the venous vessel walls and into the surrounding subcutaneous tissue. A transport in the opposite direction from subcutaneous tissue to cutaneous tissue later in the washout process seems to be negligible due to the following reasons. The very high solubility of 133Xe in subcutaneous tissue compared to that in blood and cutaneous tissue will counteract an exchange by diffusion from the subcutaneous tissue. The
735
higher linear velocity of blood in the arterial vessels, and the lower contact area between blood and tissue in these vessels compared to that of the venous vessels, will also minimize the exchange between the subcutaneous and cutaneous tissues. The biexponential washout of 133Xe is on the abovementioned conditions a combined washout curve including an initial, fast washout component from the cutaneous tissue and an accumulation in the subcutaneous tissue, followed by a washout from this tissue. The accumulation in the subcutaneous tissue is determined by the convective transport with the cutaneous venous blood. Thus, the registered curve contains only two washout rate constants — the cutaneous and the subcutaneous — and by a graphic curve resolution these two rates can be obtained (Figure 85.2).5 In a special region of the skinfold between the extended thumb and the forefinger it was possible to measure the washout of 133Xe separately from cutaneous tissue. This was done after atraumatic local labeling with 133Xe gas of the region and a shielding of the rest of the hand by a 3-mm-thick lead shield. By registration of the activity from the distal, unshielded 3 to 4 mm of the skinfold, being solely cutaneous tissue, a monoexponential washout curve was obtained over 3.5 decades (Figure 85.4). A similar result has been obtained for a skinfold raised on the back of the hand. A monoexponential washout of 133Xe from subcutaneous tissue has also been observed. After an atraumatic local labeling of subcutaneous fatty tissue in an autoperfused inguinal fat pad preparation in cats, a monoexponential washout was followed over 2.5 h (Figure 85.5).5
85.3.4 THE WASHOUT MODEL On the basis of these observations of monoexponential washout of 133Xe from cutaneous and subcutaneous tissues the following combined in-series and in-parallel washout model is described.5 The model assumes that under steadystate conditions a constant fraction, E, of the 133Xe in the cutaneous venous blood is extracted, as it passes through the subcutaneous tissue, due to the 10-fold higher solubility of 133Xe in this tissue than in blood. The model consists of two homogeneous compartments symbolized by C and S for the cutaneous and subcutaneous compartments, respectively. The input to the system is in the form of an impulse into C. With the initial amount of 1 U, the output from C is divided into two fractions: (1) the extracted fraction, E, and (2) the transmitted fraction, 1 – E. The extracted fraction, E, of the output from C is reaching the subcutaneous compartment, and thus C is in-series with S for this fraction. The complementary fraction, 1 – E, is transmitted solely through the vascular volume by the flowing blood. The transport
736
Handbook of Non-Invasive Methods and the Skin, Second Edition
C S
(a)
C S
(b)
C S
(c)
FIGURE 85.3 Radioautograms of cutaneous (C) and subcutaneous (S) tissues. The tissue boundaries are illustrated in the schematic drawings to the right, (a and b) were taken after epicutaneous labeling with 133Xe gas for 3 min. (a) was taken after 2 min of in vivo washout, (b) after 70 min. (c) is the distribution of 133Xe in the tissue after intracutaneous injection of 0.1 ml of 133Xe in isotonic saline taken 70 min later demonstrating the minimum of exchange by diffusion between the cutaneous and the subcutaneous tissue without blood flow. The exposure time of the film emulsion was 20 h in (a), and 120 h in (b) and (c). (From Sejrsen, P., Circ. Res., 25, 215, 1969. With permission.)
of this fraction is thus an output from C, which is inparallel with the output from S (Figure 85.6). The total amount 1 is initially located in compartment C, and the cumulative output from C at time t is called H1. The amount retained in compartment C at time t, Rc, can then be written as Rc = 1 – H1. It is observed that compartment C is a well-mixed compartment with an exponential washout. The expression Rc = 1 – H1 = e–kc·t
(85.1)
can then be written for compartment C, where kc is the elimination rate constant for C.
At time zero there is no indicator in compartment S. The rate of input of tracer to S is a constant fraction, E, of the output rate from C. This can be obtained by differentiation of H1 with change of the sign. The input rate to S, Is, is then Is = E · kc · e–kc·t
(85.2)
It is assumed that the output from S follows a monoexponential function, which can be described by ks·e–ks·t for 1 U of indicator. Corresponding to the input rate to S, given by Equation 85.2, the output, Os, from S is the input rate, Is, convoluted by the impulse response for S:
The 133Xenon Wash-Out Technique for Quantitative Measurement
737
t
105
Os =
Application
∫E⋅k 0
⋅ e – kc⋅τ ⋅ k s ⋅ e – ks⋅( t – τ ) dτ
E ⋅ kc ⋅ ks ⋅ e – ks⋅t – e – kc⋅t kc – ks
(
Os = 104
c
(85.3)
)
(85.4)
A combined expression for the amount of indicator contained in C plus S, Rc plus Rs is called Q(t):
Counts/minute
Q(t) = Rc + Rs
(85.5)
By inserting Equations 85.1, 85.2, and 85.4 in Equation 85.5, the following expressions are obtained:
103
t
t
∫
∫
Q(t ) = R c + I s dt – O s dt 0
(85.6)
0
⎤ ⎡ E – kc Q(t ) = e – kc⋅t + ⎢ ⋅ e – ks⋅t – e – kc⋅t ⎥ k k – s ⎦ ⎣ c
102
(
)
(85.7)
⎛ E ⋅ k c ⎞ – kc⋅t E ⋅ kc ⋅e + ⋅ e – ks⋅t (85.8) Q(t ) = ⎜1 – k c – k s ⎟⎠ kc – ks ⎝ 101 0
30
60 Minutes
FIGURE 85.4 Washout curve separate from cutaneous tissue after epicutaneous labeling with 133Xe gas in 3 min. (From Sejrsen, P., Circ. Res., 25, 215, 1969. With permission.)
Counts/minute
104
Thus, it is possible by graphic curve resolution to determine the rate constants for the cutaneous and subcutaneous components. Furthermore, it is possible to calculate the E fraction from the intercepts of the two curves and their rate constants. The rate constants are the blood flow rate-to-partition coefficient ratios for the two tissues. On average, E was observed to be 0.50 in 10 washout experiments on the lateral side of the lower leg.
Application
103
102 0
30
60
90
120
150
Minutes
FIGURE 85.5 Washout curve separate from cutaneous tissue after local labeling with Res., 25, 215, 1969. With permission.)
133
Xe gas in 1 min. (From Sejrsen, P., Circ.
738
Handbook of Non-Invasive Methods and the Skin, Second Edition
S E C 1-E
FIGURE 85.6 Schematic diagram of the washout model. (C) is the cutaneous and (S) the subcutaneous component. (E) is the extracted fraction, which is washed out via the subcutaneous tissue. (1 – E) is the complementary fraction, which is washed out from the cutaneous tissue and transmitted by convection through the vascular volume. (From Sejrsen, P., Circ. Res., 25, 215, 1969. With permission.)
85.3.5 CALCULATION
OF
BLOOD FLOW RATES
Blood flow rates can be calculated from the rate constants using the equation introduced by Kety:6,7 f = ln2 · T1/2–1 · λ · 100 (ml · (100 g · min)–1)(85.9) where ln2 is the natural logarithm to 2 and l is the tissueto-blood partition coefficient for 133Xe in milliliters per gram (0.7 for cutaneous tissue and 10 for subcutaneous tissue).1 t1/2 is the half-time of the two monoexponential components as obtained by graphic curve resolution for the two tissues. The factor 100 is introduced to give the results per 100 g of tissue, which is the conventional term.
85.3.6 LOSS
OF 133XE FROM THE
SKIN SURFACE
The question concerning loss of 133Xe by diffusion out through the intact skin surface during the washout period has been elucidated by experiments with blood flow cessation. This was done by a cuff placed on the upper arm
60000
Counts/minute
Application
and inflated to a pressure of 230 mmHg, a pressure chosen well above the systolic pressure. Under this condition the very slow elimination rate is solely due to a loss of 133Xe out through the epidermal membrane (Figure 85.7).4 This has been demonstrated by placing a gastight Mylar membrane, 20 μm thick, over the deposit area with a drop of water interposed. After this gastight sealing of the surface the curve has an almost horizontal course with a decline equal to the physical decay. Without this gastight sealing of the surface the observed very slow elimination rate is about 1 to 2% of that observed during intact blood flow at normal, thermoneutral conditions (Figure 85.7). However, under sweating conditions, it can account for as much as 20 to 25% of the measured elimination rate of 133Xe from a cutaneous deposit, as measured just after the end of the labeling period. In such situations a correction can be performed to give the cutaneous blood flow rate. This can be done by subtraction of the elimination rate measured during blood flow cessation from that obtained for cutaneous tissue during intact blood flow after curve resolution. Thus, the elimination rates are in the order of 10–4 –1 min due to the physical decay of 133Xe, 10–3 min–1 due to diffusional loss of 133Xe out through the intact epidermal membrane, 10–2 min–1 due to profuse sweating, and from 0.07 to 0.7 min–1 due to cutaneous blood flow, corresponding to blood flow rates of about 6 to 50 ml · (100 g · min)–1 at normal conditions and at local heating to 45°C, respectively.5,8 On this basis the errors due to physical decay and loss of 133Xe out through the intact epidermal membrane are considered negligible when calculating blood flow rate from 133Xe washout curves. Blood flow rates in subcutaneous tissue are at normal conditions and during heating of the skin surface to 45°C measured to about 3 and 50 ml · (100 g · min)–1, respectively.1,8
Tourniquet
k = 0.0007 min−1 −1
k = 0.0430 min
30000
Environmental temperature 20°C
15000 0
10
20
30 Minutes
k = 0.1160 min−1
40
50
FIGURE 85.7 Washout of 133Xe after epicutaneous labeling with and without bloodflow cessation. (From Sejrsen, P., Circ. Res., 25, 215, 1969. With permission.)
The 133Xenon Wash-Out Technique for Quantitative Measurement
85.4 SOURCES OF ERROR The loss of 133Xe out through the skin surface is, as mentioned, negligible with an intact epidermal membrane, but correction can be necessary during sweat secretion. The gastight epidermal membrane can be removed by 40 to 50 times of stripping with adhesive plaster. This procedure removes the dry part of the epidermis, stratum corneum, which normally has a water content of about 4%. By this procedure the basal, living cell layers in stratum germinativum are left in situ. After 40 or 50 times of stripping with adhesive plaster the loss of 133Xe from the skin surface will increase about 30 to 40 times.9 Epidermal desquamation or denudation due to pathological processes can therefore give rise to a severe loss of 133Xe out through the skin surface. However, by placing a 20-mm-thick Mylar membrane on the skin with a drop of water interposed, it is possible to reestablish a gastight sealing of the surface. Steady state of blood flow is an assumption for the method. This is the reason for a demand of constant thermal conditions during a measurement. Alsom traumatic influence has to be avoided. The reason for not using a labeling technique with intracutaneous injection of 133Xe dissolved in isotonic saline is just the trauma effected by the injection, leading to hyperemia in the following 10 to 30 min. An uptake of 133Xe in rubber and plastic materials has been observed. When a cadmium telluride detector mounted with a rubber cap is placed in contact with the labeled skin area, the 133Xe uptake in the rubber invalidates the method. The use of standard values for the tissue-toblood partition coefficient, λ, for cutaneous and subcutaneous tissues are presumably acceptable in many regions. However, in regions with a thin subcutaneous layer, the lipid contents can be reduced, and a lower value has to be used for this tissue. In such a region it is possible to make an estimate of the relative contents of lipid, water, and protein in a tissue biopsy, and from these values and the corresponding values for the blood, to calculate a λ value for the tissue in question.
85.5 CORRELATION WITH OTHER METHODS Other methods have been employed in attempts to measure cutaneous blood flow rates. 85Krypton washout has been used with registration of the β activity by a Geiger–Müller tube after an intra-arterial injection of 85Kr dissolved in isotonic saline.10 This method is invalidated by diffusion processes in the tissue due to the short halfvalue thickness in the tissue of the emitted β radiation, only 0.25 mm.11 Other radioactive isotopes have been used, such as 24Na,6,7,12 131iodine,13 125I-antipyrine, and 131I-antipyrine,14
739
given as local injections with the isotopes dissolved in isotonic saline. The problems have been the trauma of injection, and that equilibrium between tissue and the flowing blood cannot be obtained for most of these indicators, as they are not freely diffusible in the tissues. Helium uptake through the skin has been used on the extremities. This method underestimates the cutaneous blood flow rate due to the existence of the epidermal diffusion barrier to gases.15,16 In accordance with this valuation, values obtained by this method have been low, about 3 to 4 ml · (100 g · min)–1). Heat conduction has been employed as a qualitative measure of cutaneous blood flow rate.17,18 The loss of heat to the surroundings invalidates this method as a quantitative method. From the amount of heat dissipated from the skin, a blood flow rate in the order of magnitude 2 to 10 ml(100 gmin)–1 seems likely.19,20 Venous occlusion plethysmography can only give a rough evaluation of the cutaneous blood flow rate, as it is based on measurements of blood flow rates before and after iontophoresis of epinephrine into the skin combined with complicated subtraction procedures. These procedures are necessary to correct for blood flow in subcutaneous and muscle tissue.21,22 By using values of the ratio between the weight of the cutaneous and subcutaneous tissues and the blood flow rate in subcutaneous tissue, a rough estimate gives a cutaneous blood flow rate in the order of 6 to 9 ml · (100 g · min)–1. Measurements with the laser Doppler technique are determined by both the velocity and the contents of the red blood cells in the tissue, causing this method to be of a semiquantitative nature. The change in blood cell contents in the vessels during orthostatic maneuvers and heat stress limits use of this method.23
85.6 RECOMMENDATIONS It is recommended to make control experiments with blood flow cessation to exclude loss of 133Xe from the skin surface or change in counting geometry during the registration. By this type of measurement it is possible to estimate a rate constant for loss of 133Xe to the surrounding air, and thereby to make a correction for this non-bloodflow-dependent elimination. As the 133Xe washout method with graphic curve resolution is based on the assumption of a steady-state blood flow rate, it is important to maintain a constant temperature in the body and in the surroundings during the registration. Also, the body and the region under study shall be kept in a constant position during the measurement to obtain steady-state conditions. Changes in blood flow rate in cutaneous tissue during the measurement can be registered quantitatively by the 133Xe washout method in the skinfold on the hand. A similar possibility is present for the subcutaneous tissue in the later part of the washout curve, which is separate
740
Handbook of Non-Invasive Methods and the Skin, Second Edition
from this tissue. It is important to use a sufficiently high 133Xe activity and to follow the washout for a sufficiently long time (1.5 to 2 h) to get an acceptably low standard deviation of the subcutaneous washout rate. This is a necessary basis for a reasonable graphic curve resolution and determination of the cutaneous washout rate.
REFERENCES 1. Sejrsen, P., Measurement of cutaneous blood flow by freely diffusible radioactive indicators, Dan Med Bull, Suppl. 18, 1, 1971. 2. Bojsen, J., Staberg, B., and Kølendorf, K., Subcutaneous measurements of 133Xe disappearance with portable CdTe(Cl) detectors: elimination of interference from combined convection and diffusion, Clin Physiol, 4, 309, 1984. 3. Sejrsen, P. and Tønnesen, K.H., Inert gas diffusion method for measurement of blood flow using saturation techniques: comparison with directly measured blood flow in isolated gastrocnemius muscle of the cat, Circ Res, 22, 679, 1968. 4. Sejrsen, P., Atraumatic local labelling of skin by inert gas: epicutaneous application of xenon-133, J Appl Physiol, 24, 570, 1968. 5. Sejrsen, P., Blood flow in cutaneous tissue in man studied by washout of xenon-133, Circ Res, 25, 215, 1969. 6. Kety, S.S., Quantitative measurement of regional circulation by the clearance of radioactive sodium, Am J Med Sci, 215, 352, 1948. 7. Kety, S.S., Measurement of regional circulation by the clearance of radioactive sodium, Am Heart J, 38, 321, 1949. 8. Jaszczak, P. and Sejrsen, P., Determination of skin blood flow by 133Xe washout and by heat flux from a heated tc-Po2 electrode, Acta Anaest Scand, 28, 482, 1984. 9. Sejrsen, P., Epidermal diffusion barrier to xenon-133 in man and studies of clearance of xenon-133 by sweat, J Appl Physiol, 24, 211, 1968. 10. Jacobsson, S., Studies of the blood circulation in lymphoedematous limbs, Scand J Plast Reconstruct Surg, Suppl. 3, 1, 1967.
11. Sejrsen, P., Diffusion processes invalidating the intraarterial krypton-85 beta particle clearance method for measurement of skin blood flow in man, Circ Res, 21, 281, 1967. 12. Braithwaite, F., Farmer, F.T., and Herbert, F.I., Observations on the vascular channels of tubed pedicles using radioactive sodium, III, Br J Plast Surg, 4, 38, 1951. 13. Alpert, J.S. and Coffman, J.D., Effect of intravenous epinephrine on skeletal muscle, skin, and subcutaneous blood flow, Am J Physiol, 216, 156, 1969. 14. Kövamees, A., Skin blood flow in obliterative arterial disease of the leg, Acta Chir Scand, Suppl. 397, 1, 1968. 15. Behnke, A.R. and Willmon, T.L., Cutaneous diffusion of helium in relation to peripheral blood flow and the absorption of atmospheric nitrogen through the skin, Am J Physiol, 131, 627, 1940/41. 16. Klocke, R.A., Gurtner, G.H., and Farhi, L.E., Gas transfer across the skin in man, J Appl Physiol, 18, 311, 1963. 17. Hensel, H., Messkopf zur Durchblutungsregistrierung an Oberflauachen, Arch Physiol, 268, 604, 1959. 18. Golenhofen, K., Die Hautdurchblutung des Menschen; Möglichkeiten zur Objektivierung von Hautreaktionen, Fette, Seifen Anstrichmittel, 3, 177, 1968. 19. Hardy, J.D. and Soderstrom, G.F., Heat loss from the nude body and peripheral blood flow at temperatures of 22 to 35C, J Nutr, 16, 493, 1938. 20. Stewart, H.J. and Evans, W.F., The peripheral blood flow under basal conditions in normal male subjects in the third decade, Am Heart J, 26, 67, 1943. 21. Cooper, K.E., Edholm, O.G., and Mottram, R.E., Blood flow in the skin and muscle of the human forearm, J Physiol (London), 128, 258, 1955. 22. Kontos, H.A., Richardson, D.W., and Patterson, J.L., Blood flow and metabolism of forearm muscle in man at rest and during sustained contraction, Am J Physiol, 211, 869, 1966. 23. Klemp, P. and Staberg, B., The effect of antipsoriatic treatment on cutaneous blood flow in psoriasis measured by 133Xe washout method and laser Doppler velocimetry, J Invest Dermatol, 85, 259, 1986.
86 Evaluation of Lymph Flow P.S. Mortimer St. George’s and Royal Marsden Hospitals, London, United Kingdom
CONTENTS 86.1 Introduction............................................................................................................................................................741 86.2 Background ............................................................................................................................................................742 86.2.1 Lymphangiography ....................................................................................................................................742 86.2.1.1 Fluorescence Microlymphangiography12 ...................................................................................742 86.2.1.2 Indirect Lymphography13............................................................................................................742 86.2.2 Lymphoscintigraphy ..................................................................................................................................742 86.2.2.1 Historical Perspective.................................................................................................................742 86.2.2.2 Lymph Transport Kinetics..........................................................................................................742 86.2.3 Skin Lymph Flow ......................................................................................................................................743 86.3 Object.....................................................................................................................................................................743 86.3.1 Radiolabeled Tracers .................................................................................................................................743 86.3.2 Procedure ...................................................................................................................................................743 86.3.3 Analysis of Data ........................................................................................................................................743 86.3.4 Reliability and Reproducibility .................................................................................................................744 86.3.4.1 Animal Studies ...........................................................................................................................744 86.3.4.2 Human Studies ...........................................................................................................................744 86.3.4.3 Studies in Pathological Skin ......................................................................................................745 86.4 Sources of Error.....................................................................................................................................................747 86.4.1 Tracer Migration ........................................................................................................................................747 86.4.2 Blood Clearance ........................................................................................................................................747 86.4.3 Injection Trauma........................................................................................................................................747 86.4.4 Injection Depth ..........................................................................................................................................748 86.4.5 Volume of Distribution ..............................................................................................................................748 86.4.6 Extrinsic Forces .........................................................................................................................................748 86.5 Correlation with Other Methods ...........................................................................................................................749 86.6 Recommendations..................................................................................................................................................749 References .......................................................................................................................................................................749
86.1 INTRODUCTION Few techniques exist for the functional assessment of skin lymphatics; yet, it is in this area of microcirculation research that most questions relating to the role of lymphatics in disease remain unanswered. The lymphatic vessels provide an important “limb” to the microcirculation of the skin and, together with the blood vessels, cater for the constant recirculation of protein- and lymph-borne cells, e.g., Langerhans cells1 and T lymphocytes. It is the essential function of the lymphatic system to return to the vascular compartment extravascu-
lar protein molecules, colloids, and particulate matter too large to reenter the blood capillaries directly.2 The rate at which labeled protein molecules or colloids are removed from the interstitial tissues has been regarded as an index of lymphatic function.3–5 Measurement of skin and subcutaneous lymph flow has employed the same principle of isotope clearance as measurement of skin and subcutaneous blood flow.6,7 However, the interpretation of the clearance of tracers from the skin in disease states appears difficult and unreliable.8,9 741
742
Handbook of Non-Invasive Methods and the Skin, Second Edition
86.2 BACKGROUND 86.2.1 LYMPHANGIOGRAPHY In vivo visualization of lymphatic vessels (lymphangiography) using x-ray contrast medium10 remains the gold standard for lymphatic vessel abnormalities. The technique, however, is invasive, difficult to perform, and provides only anatomical detail with no functional information. Only subcutaneous lymphatics as large, or larger than, collectors can be opacified, except in pathological circumstances when dermal backflow occurs and small skin lymphatics become visible. Intravital dyes, e.g., patent blue, used to delineate subcutaneous lymphatics prior to direct cannulation for x-ray lymphography, can be used to visualize dermal lymphatics but not capillaries.11 The results, however, are transitory. Two new methods of lymphangiography have been developed in recent years. 86.2.1.1 Fluorescence Microlymphangiography12 This technique enables the superficial lymph capillary network of the skin to be seen under the vital microscope by means of fluorescing macromolecules (FITC-Dextran, Sigma) injected subepidermally and cleared exclusively by lymphatics. Information regarding the morphology of lymphatic capillaries and precollectors (initial lymphatics) and the extent of tracer propagation within the dermal lymphatic network can be recorded on video for analysis. 86.2.1.2 Indirect Lymphography13 Indirect lymphography employs water-soluble nonionic xray contrast media that can be administered via an interstitial injection without recourse to direct access to lymphatics. Iotralan® or Iotasol® (Schering AG, Berlin) is infused by a motor pump into the skin; 2 to 3 ml injected intradermally leads to considerable local skin distention and is not without discomfort. Dermal and subcutaneous collecting lymphatics can be visualized by x-ray using the mammography film method. In the presence of incompetent valves and dermal backflow, initial lymphatics can also be seen. All lymphangiographic methods are limited in their ability to evaluate lymph flow, as the techniques are essentially for demonstrating the anatomy of lymph vessels.
86.2.2 LYMPHOSCINTIGRAPHY The development of lymphoscintigraphy was aimed mainly at imaging the lymphatic system and in particular the lymph nodes.14,15 Lymphoscintigraphy has proved more useful in the determination of lymph flow. Because of the close interrelationship between lymph formation
and flow, lymphoscintigraphy theoretically provides a much more comprehensive and functional examination of lymph drainage than does x-ray lymphography. 86.2.2.1 Historical Perspective The earliest studies involved measurement of lymph flow by external counting following the subcutaneous injection of 131I-labeled plasma protein into the hind limb of healthy dogs.16 The author, using an external scintillation counting technique, concluded incorrectly that the major route of the removal from the tissue spaces of crystalloid and protein molecules is via the blood capillary bed. Taylor et al.,3 using 131I-human serum albumin (HSA), concluded that the behavior of radioactive proteins injected subcutaneously was consistent with removal by the lymphatic route and that the rate of removal was slower in patients with lymphedema than in normal subjects. Hollander et al.4 studied patients with and without edema by 131I-HSA clearance from the subcutaneous tissue and concluded that the rate of removal was significantly reduced in edema caused by lymphatic obstruction, but significantly increased in edema caused by venous obstruction, congestive cardiac failure, or hypoproteinemia. Similar results were obtained by Sage et al.17 using radioactive gold (198Au), but the absorbed dose of radioactivity was unacceptably high. Emmett et al.18 stated that 131I-protein clearance studies are of little value for the initial assessment of individual patients with swollen limbs but may well prove to be the most sensitive method for evaluating the response to treatment. Although further studies on lymphatic tracer clearance were performed, their usefulness for clinical measurement was of some doubt. 86.2.2.2 Lymph Transport Kinetics Lymphoscintigraphy has proved to be a sensitive and specific method for the study of lymph transport kinetics19,20 and useful for repeat examinations.21 Computerization has allowed data on depot clearance, colloid transit, and nodal uptake to be correlated. Investigation times have been reduced owing to extrapolation of data collected over 1 to 2 hours. 86.2.2.2.1 Lymph Node Uptake Lymph node uptake has proved to be a more reliable measurement than tracer clearance.22 Most studies have employed a subcutaneous injection. Intradermal administration of tracer seems to encourage increased migration of tracer, possibly due to higher interstitial pressures in the dermis than the subcutis. 86.2.2.2.2 Transit Times The speed of passage of tracer along main lymphatic vessels immediately following the administration of tracer can be measured. Such transit times indicate patency of
Evaluation of Lymph Flow
lymph drainage routes as well as being an indirect measurement of lymph flow. 86.2.2.2.3 Fractional Removal Rate (Depot Clearance) The amount of an injected tracer deposited in a tissue principally decreases along an exponential curve, the slope of which is expressed as a clearance constant.23 The tracer must be freely diffusible through the tissue, and to equate with lymphatic clearance, the tracer must be removed solely through the lymphatic route.
86.2.3 SKIN LYMPH FLOW Measurement of lymphatic function specifically in the skin by the disappearance rate of 131I albumin from the dermis was first reported in 19708; 50 patients were investigated and radioactivity declined exponentially, giving a linear plot over 50 hours. Radioactivity fell more rapidly during the first 4 hours, giving an initial curve on a semilogarithmic plot. Results were reproducible and clearance rates varied according to the injection site. The only further study of skin lymph flow examined albumin clearance from psoriatic skin.9 Again, small quantities (0.1 ml) of 131I-HSA containing 10 mCi of radioactivity were administered intradermally, and the radioactivity of each depot was measured by sodium iodide scintillation detectors. Half clearance times were calculated by regression analysis by the least squares method. Clearance was shown to be monoexponential and increased in involved psoriatic skin, indicating increased lymph drainage.
86.3 OBJECT If interstitial protein clearance is the essential function of the lymphatic system, can skin lymph flow be measured reliably using the principle of isotope clearance and give meaningful results? A measure of lymphatic function is the efficiency of protein removal from the tissues. Lymph flow is considered here as equivalent to the movement of protein and accompanying fluid for the purpose of lymph drainage. (Flow refers to bulk transport per unit volume of tissue per unit time. Strictly, it is not possible to measure absolute lymph flow in vivo.) Skin lymph flow relies on several interdependent steps, which include lymph formation, its entry into lymph capillaries followed by transit through noncontractile lymphatics, and then propulsion by subcutaneous contractile lymphatics. Movement of solid matter need not necessarily relate to that of fluid, and lymph flow is driven by many extrinsic forces. Colloidal (or protein) clearance from the dermis does involve every step of skin lymph flow and is theoretically the ideal test. Perhaps a more appropriate
743
term would be skin lymph drainage instead of lymph flow. For this reason, lymph flow was expressed as a half clearance time (t1/2) and calculated from the slope of the exponential clearance curve.
86.3.1 RADIOLABELED TRACERS Macromolecules and colloids of a certain size are transported exclusively in the lymph. The ideal tracer is one that migrates freely from the injection site through the tissues and away in the lymphatics. It was discovered that the optimal colloid was one with a particle size of a few nanometers with a small dispersion around that value.24 Too large a particle resulted in poor absorption from the injection depot, and too small a particle risked blood clearance. 99mTc-labeled agents offer significant advantages in terms of radiation dose and energy characteristics. Much lower doses of radioactivity are possible with external scintillation counting, but image formation using scintiscanner or gamma camera demands higher radioactivity. In a study25 comparing 131I-HSA, 198Au, and 99mTccolloid (TCK 17, Cis), t1/2 values for 198Au were extremely long and variable. The clearance of 99mTc-colloid was slightly faster, but not significantly so, than 131I-has, and more consistent (Figure 86.1). To reduce leakage, the needle was inserted obliquely through the skin. Tracer was injected slowly with minimal force. The needle entry site was wiped once with a cotton wool swab after withdrawal of the needle.
86.3.2 PROCEDURE Injections of tracer were made into the dermis at a superficial level (subepidermal). An injection volume of 0.03 ml was used using a 30-gauge needle. Albumin may be more physiological, but the main function of the lymphatics is the removal of macromolecules, including exogenous colloids, from the interstitium. A well-collimated sodium iodide detector was positioned over the injection site with the collimator surface 10 mm from the skin. Additional detectors can be used to monitor other injection sites when paired studies are to be performed or to monitor uptake in the regional lymph nodes. Each detector was connected to a dual-channel interface analyzer operating in the multiscaler mode to give a digital output of radioactive counts with time. Counts were integrated over periods of 10 to 50 seconds and data recorded on a computer for a total time of 30 min. Observations up to periods of 90 min were not found to give any greater consistency.
86.3.3 ANALYSIS
OF
DATA
Changes in radioactive counts were plotted as the percentage of the maximum count rate against time. The resulting clearance curves were analyzed using a single exponential equation.25 The data points were fitted to this equation
744
Handbook of Non-Invasive Methods and the Skin, Second Edition
5
Half clearance times - T½ (hr)
4
3
2
1
0 Pig 1
Pig 2
Pig 1
99MTc-colloid
Pig 2 1311-HSA
99MTC-Colloid
1311-HSA
Cocktail
FIGURE 86.1 Variations in t1/2 values for the clearance of 99mTc-colloid with 131I-HSA (●) when injected as single tracers or together as a cocktail into the flank skin of the pig. The individual results of separated measurements on two pigs are given along with the mean ± SE for each set of results.
using a nonlinear least squares method. The rate of removal was corrected for the decay of the isotope (99mTc = 6 hours) and lymph flow expressed as a half clearance time (t1/2) in hours.
86.3.4 RELIABILITY
AND
REPRODUCIBILITY
86.3.4.1 Animal Studies The ICT proved to be a reliable and reproducible method when performed as a group test under controlled laboratory conditions in anesthetized large white pigs.25 Data were found to fit best to a monoexponential equation with good correlation coefficients (>0.86), indicating that the clearance was a mono- rather than biexponential function. This is in keeping with the previously published human work.8,9 Long investigation times are not only impractical, but fluctuations in lymph flow may be expected, particularly from movement in an unanesthetized animal. Differences were observed in lymph flow between the two pigs studied (Figure 86.1) and between skin sites within each animal. Differences in lymph vessel density and distribution and in tissue compliance were the likely explanations. Age was not a major factor in these studies, but has been recently shown to result in a decline in limb lymph drainage in humans.26
86.3.4.2 Human Studies As in the controlled studies performed in pig skin,25 similar studies performed in human skin revealed essentially monoexponential clearance.27 There was, however, very little consistency of t1/2 in repeat basal studies (basal = without interference, e.g., lymph flow enhancement), and serious doubts must be raised regarding the value of single lymph flow determinations at rest (basal). Fractional removal rates were not significantly different in pretibial skin compared with the skin of the thigh or foot. The wide range of t1/2 values witnessed in normal human skin differed from the reproducible results seen under controlled conditions in pig skin. This was considered to indicate real differences in lymph flow rather than technical error, particularly as results varied in repeat studies in the same subject on consecutive days, whereas right and left legs showed similarity when examined together. Lymph flow at rest is slow and subject to instant fluctuations, and its measurement is clearly prone to error unless carried out under strictly controlled conditions. All components of lymph movement depend upon changes in local tissue and hydrostatic pressure. These changes are produced by external compression,28 muscular activity,29,30 skin surface massage,31,32 passive movements,33 and local arterial pulsation.34 Movement of macromolecules from
Evaluation of Lymph Flow
745
1.0 0.9 0.8
Half clearance - T½ (hr)
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 Control
Massage
Pig 1
Control
Massage
Pig 2
FIGURE 86.2 The effect of local massage on the half clearance time for 99mTc-colloid from skin on the flank of pigs. The individual results are given for separate measurements on two pigs (●, unmassaged sites; , massaged sites) along with the mean values ± SE.
interstitium toward, into, and through peripheral lymphatics would appear to be a predominantly passive process dependent on many extrinsic forces rather than an active process generated by the lymphatic itself. A study of lymph flow enhancement comparing vibration with local massage demonstrated greater colloid clearance from massage. The response to vibration was disappointing. A possible explanation for this failure would be that the rapid movement of vibration was too fast to allow adequate lymph vessel filling. Local massage significantly enhanced colloid clearance in both normal pig25 (Figure 86.2) and human skin (Figure 86.3). Massage performed some distance away in the leg from the injection depot according to the principle of manual lymphatic drainage35 invoked a significant increase in clearance superior to that generated by pneumatic compression therapy (Talley Medical Equipment Ltd.) (Figure 86.4). The proposed theory is that such massage has a milking or siphoning effect on distal lymph. Stimulation of the intrinsic contractility of the main lymphatic collecting vessels in the limb would pump lymph proximally, so generating a pressure gradient that draws lymph from peripheral lymphatics, including in the skin. The lack of efficacy from pneumatic compression therapy was a surprise as the 10-chamber inflatable garment produced a pressure wave moving repeatedly up the limb that
was far stronger than the massage. These machines are widely used for the treatment of lymphedema, but may do little to mobilize protein via lymphatics.36 Comparison of colloid clearance from normal skin on the dorsum of the foot with the same site in a lymphedematous leg showed no difference under basal conditions with the subject supine. Clearance of 99mTc-colloid from the dermis, as a measure of skin lymph flow, could only differentiate normal subjects from patients with lymphedema by the response to massage. Single lymph flow determinations using 99mTc-colloid clearance from the dermis over a short investigation time are therefore only meaningful when attempts are made to enhance lymph flow and so test lymph transport capacity. Massage by stimulating lymph flow exposes the deficiency in lymph transport, which examination at rest may miss. Only then can lymphatic insufficiency be distinguished from normal function. 86.3.4.3 Studies in Pathological Skin Because edema and connective tissue changes are wellknown sequelae of lymphatic damage, and because such changes occur following radiation to the skin, lymph flow studies using the ICT (99mTc-colloid) were undertaken in pig skin following single doses of x-rays.37 Paired estimates of lymphatic clearance were performed in irradiated
746
Handbook of Non-Invasive Methods and the Skin, Second Edition
10
8
T½ (hrs)
6
4
2
0 Basal
Vibration
Basal
Stimuloval
Basal
No vibration/ stimuloval
FIGURE 86.3 The clearance of 99mTc-colloid from normal skin of the lower leg expressed as t1/2, in normal subjects at rest (basal), and following a period of lymph flow enhancement (vibration or surface massage).
35
30
T½ (hrs)
25
20
15
10
5
0 Basal
PT101
Basal
Manual massage
FIGURE 86.4 The clearance of 99mTc-colloid, expressed as t1/2, from the skin of the dorsum of the foot during rest, basal, (0 to 30 min post-injection) and following a period (30 to 60 min) of lymph flow enhancement with either pneumatic compression (PT101) or manual massage.
and unirradiated sites on the flank skin of anesthetized large white pigs at 3, 6, 9, 12, 26, 39, 52, 64, and 78 weeks after a single dose of 18 Gy of x-rays. The results demonstrated good consistency of results relative to site and time examined. The results demonstrated two waves of
impaired lymphatic clearance with time, which correlates well with the gross morphological changes observed (Figure 86.5). It was concluded that impaired lymphatic drainage probably contributes to the gross and histological changes observed in the skin following x-irradiation. The
Evaluation of Lymph Flow
747
6
Half clearance time - T½ (hr)
5
4
3
2
1
0
0
12
24
39 54 Time after irradiation (weeks)
66
78
FIGURE 86.5 Time-related changes in the clearance of 99mTc-colloid from the dermis of the flank of the pig following irradiation with a dose of 18Gy of x-rays. Results are expressed as the mean half clearance items (± SE) for irradiated (●) and nonirradiated () fields. The shaded area represents the mean t1/2 value (± SE) in normal skin. Error bars indicated.
study demonstrated the value of the ICT for lymph flow measurements even in pathological skin, providing experimental conditions are well controlled.
86.4 SOURCES OF ERROR The advantage of the ICT is that it utilizes the principal function of the lymphatic, namely, the removal of protein or colloid from the interstitium, and examines it dynamically. As such, it explores the capacity of the lymphatic system to absorb material from the interstitial space and transport it to the regional node. This corresponds to the route taken by protein leaked from blood capillaries and subsequently drained from the interstitium by the lymphatics. Of the whole circulating plasma protein pool, 50 to 100% leaves the circulation daily. It is the lymphatic system that maintains this “extravascular circulation of plasma proteins.”38
86.4.1 TRACER MIGRATION Flow can be calculated from clearance provided that (1) the tracer leaves by only one route and (2) it reaches an instant diffusion equilibrium between lymph and tissue. The rate-limiting step for measuring lymph flow is almost certainly the poor migration of tracers from the injected site. Injected proteins behave differently from native
plasma proteins. When an isotope in its colloidal form is injected into the tissues, approximately 90% is precipitated on the local tissue proteins. Only a small proportion is attached to the mobile proteins and to phagocytes, and so taken up by the lymphatics.39 This obviously limits the measurable quantity of 99mTc-colloid available for clearance. In the studies described, an average of 15% of colloid injected was absorbed by lymphatics in the first 30 min.
86.4.2 BLOOD CLEARANCE Isotope clearance, as a measure of lymph flow, has been criticized because of the risk of significant amounts of radioactivity escaping by the bloodstream, thus invalidating clearance values interpreted as lymph flow. This problem was investigated by comparing blood clearance of 99m Tc-colloid with its total (lymphatic) clearance. 25 Results revealed that the percentage of tracer cleared by the bloodstream was never more than 1.5% of total clearance. Lymphoscintigraphic studies confirmed that blood clearance was negligible.
86.4.3 INJECTION TRAUMA Injection into the skin is obviously traumatic and nonphysiological. The insertion of a needle will undoubtedly on
748
Handbook of Non-Invasive Methods and the Skin, Second Edition
10 9
Half clearance times - T½ (hr)
8 7 6 5 4 3 2 1 0 S/E
S/C
S/E
D/D
Depth of injection
FIGURE 86.6 Variations in the t1/2 values for the clearance of 99mTc-colloid from the normal flank of the pig following injections at different depths. Injections were subepidermal (S/E, ) subcutaneous (S/C, ●), and deep-dermal (D/D, ). The individual resuls of separate measurements are given along with the mean ± SE for each set of results..
occasions disrupt a lymph capillary mesh 600 to 1000 mm wide. Disruption of blood capillaries is more likely, but bleeding at the entry point can be minimized if care is taken. Obvious bleeding should lead to termination of the measurements. Laceration of the lymph capillary network effectively results in direct injection into lymphatics with increased uptake and filling of the vessels.11 Nevertheless, this clearly does not prevent continuation of satisfactory clearance, and with good injection techniques the disturbance is both minimal and consistent.40
86.4.4 INJECTION DEPTH Studies of injection depth in pig skin demonstrated the importance of an accurate injection for dependable results.25 Subepidermal localization of the tracer produced faster and more consistent clearance (Figure 86.6). Similar results were observed in blood flow studies using the same technique,7 where clearance rates correlated with local vascular density. The much denser network of lymphatic capillaries existing in the subpapillary region of pig and human skin41 provides a greater surface area for absorption and would satisfactorily explain the faster subepidermal clearance of colloid than deep dermal and subcutaneous sites.
86.4.5 VOLUME
OF
DISTRIBUTION
Consideration must also be given to injection pressure and volume when interpreting results. A change in injection volume did not significantly influence t1/2.25 Clearance rates were slower with a larger volume. However, the clearance rate (KT) is only proportional to lymph flow (FL) when the volume of distribution (Vi) remains the same. Therefore, no change in t1/2 despite an increase in volume suggests an increased lymph flow. This is possibly the most serious source of error when interpreting clearance rates as lymph flow. Clearly, therefore, it is not possible to compare directly clearance from normal skin with lymphedematous skin because Vi remains unknown.
86.4.6 EXTRINSIC FORCES The differences in the consistency of lymph flow measurements performed in pig skin compared with the wide variation seen in human skin were considered a reflection of the influences of extrinsic forces. By controlling through the laboratory conditions for ambient temperature active and passive movements and pulse rate, these problems were largely overcome. This is not easily possible in human studies unless the extrinsic forces are specifically used to enhance lymph flow. Only by increasing lymph flow in response to standardized stimuli, e.g.,
Evaluation of Lymph Flow
massage, could impaired lymph drainage from the skin be detected.25
86.5 CORRELATION WITH OTHER METHODS The scintillation detector system with simultaneous measurement of two comparative sites, or depot clearance and nodal uptake, provides a portable and low-radiation method for the functional assessment of peripheral lymphatics in humans. The relative, simple, inexpensive equipment permits the technique to be used at the bedside. This has benefits in centers without gamma camera or whole-body scintillation scanners. An increased number of external detectors connected in series at intervals along the lymph drainage route from the injection site could be used to improve the sensitivity of the technique, particularly in relation to speed of lymph movement. Small detectors strapped to the limb in a method similar to that of pressure transducers would be a possibility. Isotope clearance is the only method currently available that provides objective and dynamic information on skin lymph flow. Most clinical methods that examine the lymphatic system — histology and electron microscopy, lymphangiography, and lymphoscintigraphy — either focus on large lymphatic vessels outside the skin or provide static, anatomical, or structural detail. Lymphoscintigraphic studies supported the results of the external detector studies and demonstrated that clearance of approximately 99% of the tracer was lymphatic. Blood clearance and tracer diffusion were negligible in normal skin. Results using external scintillation detectors were in broad agreement with the published findings using a gamma camera, although most examinations employed a subcutaneous injection of tracer. Gamma camera studies demand a 20-fold greater dose of radioactivity, but nevertheless still fall within category I for radiation risk. External detector studies permit multiple repeat lymph flow measurements with safety. Gamma camera measurements of isotope clearance from the injection site have been considered unreliable, but Pecking et al.21,42 in extensive studies has shown significant differences between lymphedema and normal limbs, according to clearance, with good reproducibility.
86.6 RECOMMENDATIONS Physiological and clinical measurement of the microcirculation is important for understanding the dynamic changes that occur with pathology. So often functional questions are answered incorrectly by extrapolation of data from static studies. The ICT has the major advantage that it utilizes the principal function of the lymphatic,
749
namely, the removal of protein or colloid from the interstitium, and examines it dynamically. Skin lymph flow can only be reliably measured when conditions are controlled. For lymph flow at rest, this means controlling for all extrinsic influences as well as intersite and intersubject variations. Extrinsic factors such as massage strongly influence lymph flow. Greater sensitivity in detecting lymphatic insufficiency may be achieved if a standardized stimulus to lymph flow is administered. The response in clearance to the lymph flow enhancement will be the best indicator of lymph drainage abnormalities.
REFERENCES 1. Silberberg-Sinakin, I., Thorbecke, G.J., Baer, R.L., Rosenthal, S.A., and Berezowsky, V., Antigen bearing Langerhans cells in skin, dermal lymphatics and in lymph nodes, Cell. Immunol., 25, 137, 1976. 2. Drinker, C.K. and Field, M.E., The protein content of mammalian lymph and the relation of lymph to tissue fluid, Am. J. Physiol., 97, 32, 1931. 3. Taylor, G.W., Kinmonth, J.B., Rollinson, E., and Rotblat, J., Lymphatic circulation studied with radioactive plasma protein, Br. Med. J., 1, 133, 1957. 4. Hollander, W., Reilly, P., and Burrows, B.A., Lymphatic flow in human subjects as indicated by the disappearance of 131I-labelled albumin from the subcutaneous tissue, J. Clin. Invest., 40, 222, 1961. 5. Fernandez, M.J., Davies, W.T., Owen, G.M., and Tyler, A., Lymphatic flow in humans as indicated by the clearance of 125I-labelled albumin from the subcutaneous tissue of the leg, J. Surg. Res., 35, 101, 1983. 6. Engelhart, M. and Kristensen, J.K., Evaluation of cutaneous blood flow responses by 133-xenon washout and a laser Doppler flowmeter, J. Invest. Dermatol., 80, 12, 1983. 7. Young, C.M. and Hopewell, J.W., The evaluation of an isotope clearance technique in the dermis of pig skin: a correlation of functional and morphology parameters, Microvasc. Res., 20, 182, 1980. 8. Ellis, J.P., Marks, R., and Perry, B.J., Lymphatic function: the disappearance rate of 131I albumin from the dermis, Br. J. Dermatol., 82, 593, 1970. 9. Staberg, B., Klemp, P., Aasted, M., Worm, A.M., and Lund, P., Lymphatic albumin clearance from psoriatic skin, J. Am. Acad. Dermatol., 9, 857, 1983. 10. Kinmonth, J.B., Lymphangiography in man, Clin. Sci., 11, 13, 1952. 11. Hudack, S.S. and McMaster, P.D., Lymphatic participation in human cutaneous phenomena, J. Exp. Med., 57, 751, 1933. 12. Bollinger, A., Jager, K., Sgier, F., and Seglias, J., Fluorescence microlymphography, Circulation, 64, 1195, 1981. 13. Partsch, H., Wenzel-Hora, B., and Urbank, A., Differential diagnosis of lymphoedema after indirect lymphography with Iotasul, Lymphology, 16, 12, 1983.
750
Handbook of Non-Invasive Methods and the Skin, Second Edition
14. Battezzati, M. and Donini, I., The use of radioisotopes in the study of the physiopathology of the lymphatic system, J. Cardiovasc. Surg., 5, 691, 1964. 15. Anghileri, L.J., Lymph nodes distribution of several radio colloids: migration ability through the tissues, J. Nucl. Biol. Med., 11, 180, 1967. 16. Jepson, R.P., Simeone, F.A., and Dobyns, B.M., Removal from skin of plasma protein labelled with radioactive iodine, Am. J. Physiol., 175, 443, 1953. 17. Sage, H.H., Sinha, B.K., Kizilay, D., and Toulon, R., Radioactive colloidal gold measurements of lymph flow and functional patterns of lymphatics and lymph nodes in the extremities, J. Nucl. Med., 5, 626, 1964. 18. Emmett, A.J., Barron, J.N., and Veall, N., The use of 131I-albumin tissue clearance measurements and other physiological tests for the clinical assessment of patients with lymphoedema, Br. J. Plast. Surg., 20, 1, 1967. 19. Kleinhaus, E., Baumeister, R., Hahn, D., Siuda, S., Bull, U., and Moser, R., Evaluation of transport kinetics in lymphoscintigraphy. Follow-up study in patients with transplanted lymphatic vessels, Eur. J. Nucl. Med., 10, 349, 1985. 20. Stewart, G., Gaunt, J., Croft, D.N., and Browse, N.L., Isotope lymphography: a new method of investigating the role of lymphatics, Br. J. Surg., 72, 906, 1985. 21. Pecking, A., Cluzan, R., Desprez-Curely, J.P., and Guerin, P., Functional study of the limb lymphatic system, Phlebology, 1, 129, 1986. 22. Mostbeck, A., Kahn, P., and Partsch, H., Quantitative lymphography in lymphoedema, in The Initial Lymphatics, Bollinger, A., Partsch, H., and Wolf, J.H., Eds., Thieme Verlag, Stuttgart, 1985, p. 123. 23. Kety, S.S., Measurement of regional circulation by the clearance of radioactive sodium, Am. Heart J., 38, 321, 1949. 24. Strand, S.E. and Persson, B.R., Quantitative lymphoscintigraphy. I. Basic concepts for optimal uptake of radiocolloids in the parastemal lymph nodes of rabbits, J. Nucl. Med., 20, 1038, 1979. 25. Mortimer, P.S., Simmonds, R., Rezvani, M., Robbins, M., Hopewell, J.W., and Ryan, T.J., The measurement of skin lymph flow by isotope clearance: reliability, reproducibility, injection dynamics, and the effect of massage, J. Invest. Dermatol., 95, 677, 1990. 26. Bull, R.H., Gane, J., Evans, J., Joseph, A., and Mortimer, P.S., Abnormal lymph drainage in patients with chronic venous leg ulceration, J. Am. Acad. Dermatol., in press. 27. Mortimer, P.S., Measurement of Skin Lymph Flow by an Isotope Clearance Technique, M.D. thesis, University of London, London, 1990.
28. Miller, G.E. and Seale, J.L., The mechanics of terminal lymph flow, J. Biomech. Eng., 107, 376, 1985. 29. Yoffey, J.M. and Courtice, F.G., Lymphatics, Lymph and Lymphomyeloid Complex, Academic Press, New York, 1970. 30. Barnes, J.M. and Trueta, J., Absorption of bacteria, toxins and snake venoms from the tissues, Lancet, I, 623, 1941. 31. Calnan, J.S., Pflug, J.J., Reis, N.D., and Taylor, L.M., Lymphatic pressures and the flow of lymph, Br. J. Plast. Surg., 23, 305, 1970. 32. Olszewski, W.L., Peripheral Lymph: Formation and Immune Function, CRC Press, Boca Raton, FL, 1985. 33. Jacobsson, S., Lymph flow from the lower leg in man, Acta Chir. Scand., 133, 79, 1967. 34. Parsons, R.J. and McMaster, P.D., The effect of the pulse upon the formation and flow of lymph, J. Exp. Med., 68, 353, 1938. 35. Stijns, H.J. and Leduc, A., The contribution of physical therapy in the treatment of lymphoedema, in Lymphoedema, Clodius, L., Ed., Thieme, Stuttgart, 1977, p. 27. 36. Partsch, H., Mostbeck, A., and Leitner, G., Experimentelle untersuchungen zur wirkung einer druckwellenmassage (lymphapress) bein lymphodem, Lymphologie, V, 35, 1981. 37. Mortimer, P.S., Simmonds, R.H., Rezvani, M., Robbins, M.E., Ryan, T.J., and Hopewell, J.W., Time related changes in lymphatic clearance in pig skin after a single dose of 18Gy of X rays, Br. J. Radiol., 64, 1140, 1991. 38. Mayerson, H.S., The physiologic importance of lymph, in Handbook of Physiology, Vol. 2, Hamilton, W.F. and Dows, P.G., Eds., Waverly, Baltimore, 1963, p. 1035. 39. Haagensen, C.D., Methods of study of the lymphatic system, in The Lymphatics in Cancer, Haagensen, C.D., Ed., W.B. Saunders, Philadelphia, 1972, p. 14. 40. Courtice, F.C., Lymph and plasma proteins: barriers to their movement throughout the extracellular fluid, Lymphology, 4, 9, 1971. 41. Mortimer, P.S., Jones, R.L., and Ryan, T.J., Human skin lymphatics: regional variation and relationship to elastin, in Immunology and Haematology Research: Progress in Lymphology, Vol. 2, Heim, L., Ed., Immunology Research Foundation, Inc., Nearburgh, 1984, p. 59. 42. Pecking, A., Cluzan, R., Desprez-Curely, J.P., and Guerin, P., Indirect lymphoscintigraphy in patients with limb oedema, Phlebology, 1, 211, 1986.
Temperature and Thermoregulation
and Handheld Devices for 87 Sensors Surface Measurement of Skin Temperature Roderick A. Thomas Snell International, Tyn-Y-Coed, Pontardulais, Swansea, United Kingdom
CONTENTS 87.1 87.2 87.3 87.4 87.5
Introduction............................................................................................................................................................753 Sensors and Handheld Devices for Surface Measurement of Skin Temperature ................................................754 Temperature Measurement ....................................................................................................................................754 Measurement Location ..........................................................................................................................................754 Contact Temperature Measurement.......................................................................................................................754 87.5.1 Thermocouples...........................................................................................................................................754 87.5.2 Thermometers ............................................................................................................................................754 87.6 Noncontact Temperature Measurement.................................................................................................................756 87.6.1 Infrared Theory..........................................................................................................................................756 87.6.2 Infrared Thermometers ..............................................................................................................................756 87.6.3 Fixed Monitoring Systems ........................................................................................................................757 87.6.4 Infrared Radiometers .................................................................................................................................757 87.6.5 Mechanical Scanning and Focal Plane Arrays..........................................................................................758 87.6.6 Detector Arrays..........................................................................................................................................760 87.7 Detectors ................................................................................................................................................................760 87.7.1 Photon Detectors........................................................................................................................................760 87.7.2 Short-Wave and Long-Wave Photon Detectors.........................................................................................761 87.7.3 Photovoltaic Detectors...............................................................................................................................761 87.7.4 Photoconductive Detectors ........................................................................................................................762 87.7.5 Quantum Well Infrared Photodetectors.....................................................................................................762 87.7.6 Thermal Detectors .....................................................................................................................................763 87.8 Emerging Technology............................................................................................................................................764 87.8.1 Computer Systems .....................................................................................................................................764 87.8.2 Expert Systems ..........................................................................................................................................764 87.9 The Future..............................................................................................................................................................764 Useful Terms ...................................................................................................................................................................765 Acknowledgments ...........................................................................................................................................................767 References .......................................................................................................................................................................767
87.1 INTRODUCTION With the current technological advancements associated with infrared (IR) thermography resulting in the development of a number of different thermal imaging devices
(radiometers), the operation of which is dictated by the type of detector used, this section sets out to introduce the main infrared systems and the operation of detectors therein.
753
754
Handbook of Non-Invasive Methods and the Skin, Second Edition
87.2 SENSORS AND HANDHELD DEVICES FOR SURFACE MEASUREMENT OF SKIN TEMPERATURE The establishment, development, and consequential success of a medical infrared thermographic (MIT) intervention is primarily based on the understanding of the following: 1. Problem/condition to be monitored 2. Setup and correct operation of infrared system 3. Appropriate conditions during the monitoring process 4. Evaluation of activity and development of standards and protocol In 2, setup and correct operation of infrared system, an understanding of the various methods of skin temperature is useful. Also adopting the correct training is critical to the efficacy of monitoring.
87.3 TEMPERATURE MEASUREMENT Traditional temperature measurement of the internal body and the external skin is an important medical practice. A thermal balance at a temperature of 37°C (±0.75°C) must be maintained for optimum cellular function. Physiological thermoregulation is a complex central and peripheral interaction that attempts to balance body heat against loss. For example, heat production is a result of motor activity, shivering, and metabolic thermoregulation, while heat loss/transfer includes conduction, convection, radiation, and evaporation: • • • •
Conduction through objects by direct contact Convection through air or liquid as contact medium Radiation through space without contact Evaporation through liquid, then to air
Body heat is dynamic, moving across tissue boundaries. Heat transfer occurs when there is a difference in heat content of adjacent areas. The sum of differences between one area and another is known as a gradient. Therefore, what is actually measured, as skin temperature, is energy in motion in search of equilibrium from warmer to cooler. Temperature measurement of skin is captured using a number of different devices, as described in the next section. However, skin temperature is often captured at a moment in time and, for example, with a thermometer positioned in such a way that it can capture the transfer of heat from one space to another — at the same time minimizing the effects of the surrounding environment. An infrared ear thermometer, for example, measures radiated heat in the form of infrared energy to the space around it.
Cold junction T1 + ΔV
Hot junction T2
T + ΔT
ΔV = αΔT
Where α = Seebeck coefficient
FIGURE 87.1 Seebeck effect.
87.4 MEASUREMENT LOCATION The location of the temperature measurement device is dependent upon the condition being monitored. There are 13 different locations on the human body where clinical temperature is measured. Some are internal and some external: oral cavity, rectum, axilla, ear, tympanic membrane, nasopharynx, inguinal area, forehead, esophagus, pulmonary artery, bladder, vagina, and the great toe. Table 87.1 illustrates advantages and disadvantages of various devices and locations.
87.5 CONTACT TEMPERATURE MEASUREMENT 87.5.1 THERMOCOUPLES Thermocouples function due to the Seebeck effect, which is a combination of the Peltier effect, by which a small voltage exists at the junction of two unlike metals, and a second effect credited to Lord Kelvin, which produces a small voltage along a conductor in a temperature gradient. The Seebeck effect is an emf ΔV generated due to a temperature difference ΔT (Figure 87.1). Both effects are proportional to the temperatures involved. The total voltage produced in a circuit, including a number of thermocouples, is zero, as there is no temperature difference around the loop. Thus, two thermocouples are normally employed. One is maintained at a reference temperature (e.g., the freezing point of water), and the other acts as the thermometer. The sensitivities of common thermocouples range from 6.5 to 80 μV/dC, with accuracies from 0.25 to 1%. Several thermocouples can be arranged in series to form a thermopile to increase the sensitivity. The advantages of thermocouples are their relatively fast response (down to 1 ms), small size (down to 12 μm diameter), ease of fabrication, and long-term stability. Their disadvantages are small output voltage, low sensitivity, and need for reference temperature. Small thermocouples can be inserted into catheters and hypodermic needles. Some clinical electronic thermometers employ thermocouples.
87.5.2 THERMOMETERS There are a number of devices available; some are more common than others:
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
755
TABLE 87.1 Advantages and Disadvantages of Traditional Temperature Measurement Devices Site
Devices Used
Advantages
Oral cavity
Glass mercury Electronic predictive Phase change (dot matrix)
Easy access Familiar Minimally invasive
Rectum
Glass mercury Electronic predictive Phase change (dot matrix) Deep rectal probe Glass mercury Electronic predictive Phase change (dot matrix) Electronic sensor
Preferred by MDs
Axilla
Disadvantages Affected by eating, drinking, etc. Temperature varies within oral cavity Hard to keep thermometer in place Site records highest temp in body; lags behind other core sites when temp is changing rapidly
Ear
Infrared ear thermometer
Tympanic membrane Nasopharynx
Contact electronic sensor
Easy access Familiar Minimally invasive Preferred by American Academy of Pediatrics for use in infants Easy access Familiar Minimally invasive Two sites available Reflective of brain temperature Reflective of brain temperature
Electronic sensor
Reflective of brain temperature
Affected by breathing Invasive and uncomfortable
Glass mercury Electronic predictive Phase change (dot matrix) Electronic sensor Phase change (dot matrix) Electronic sensor
Easy access in infants and small children
Skin temperature Requires leg to be drawn up against abdomen Normal range not well documented Skin temperature Affected by environment
Esophagus
Electronic sensor
Pulmonary artery Bladder
Electronic sensor
Reflects temperature of body core Reflects temperature of body core Reflects temperature of body core
Groin
Forehead
Electronic sensor
Reflects skin temperature Not always a good indicator of core temperature Must be held in place Takes long time to reach equilibrium
Easy access Noninvasive
Vagina
Glass mercury Electronic
Preferred for basal temperature
Great toe
Electronic sensor
Easy access Noninvasive Can be informative if used with core temperature
Uses Most common site in adults and children over 5 Often requested by MDs as the most accurate site for core temperature Most common site in children under 5 Sometimes used during surgery
Requires thorough training and attention to technique
Commonly used in hospitals and clinics
Invasive and uncomfortable
Used during anesthesia Used during anesthesia, but not very common Used in infants and neonates
Temperature varies according to depth of probe placement Affected by temperature of infused fluids Affected by amount of throughput Lags behind other core body temperature sites Invasive
Peripheral skin temperature very remote from body core
Used during surgery for rough monitoring Used during anesthesia Used in surgery and critical care Used in surgery and emergency or critical care Usually used by women tracking their fertility Used during surgery to reflect peripheral circulation and temperature
From www.graduateresearch.com/thermometry/sites.htm.
•
•
Liquid-in-glass thermometer, which relies upon the expansion of a liquid or solid as the temperature rises. Mercury in glass is best known, and there are a number of variations, notably the maximum reading clinical thermometer. Digital thermometers.
• • • •
Electronic sensor/thermometers. Chemical thermometers. Dot matrix or phase change thermometers. Radiation thermometers, infrared ear thermometers.
756
Handbook of Non-Invasive Methods and the Skin, Second Edition
87.6 NONCONTACT TEMPERATURE MEASUREMENT There are essentially three types of infrared thermographic temperature measurement equipment:4 spot thermometers (sometimes known as single point devices), fixed monitoring (sometimes known as line scanners), and portable infrared thermal imaging cameras (including radiometers). The progressing development of high-quality imaging optics and novel IR detectors has substantially improved the thermal and spatial resolution of thermal imagers, resulting in excellent-quality images with quantitative information. The basis for accurate quantitative measurements is reliability, repeatability, and comparability of data. In the case of temperature measurements, as a consequence of Europeanization and globalization, the trend toward generally acknowledged quality assurance in accordance with ISO/IEC 17025 and the worldwide equivalence of measurements require traceability to the SI unit of temperature, the Kelvin, according to the International Temperature Scale of 1990 (ITS-90).
where E = energy (J) h = Planck’s constant (6.625 × 10–34 Js) ƒ = frequency (Hz) c = speed of light (m/sec) λ = wavelength (m) Planck derived (empirically) a formula to relate the radiated power spectral density from a blackbody radiating into cold space at any temperature. Bλ(T) is the energy in Joules emitted per second per unit wavelength from 1 m2 of a perfect blackbody at a temperature T (Kelvin):
( )
Bλ T =
2 πhc 2 / λ 5 e hc / λkT − 1
(87.2)
where h = Planck’s constant (6.625 × 10–34 Js) k = Boltzmann’s constant (1.3804 × 10–23 J/K) λ = wavelength (m) c = speed of light (3 × 108 m/s) T = temperature (K)
87.6.1 INFRARED THEORY In most objects, at a temperature above absolute zero (0 Kelvin or –273.16°C), every atom and every molecule vibrate. According to the laws of electrodynamics, a moving electric charge is associated with a variable electric field that in turn produces an alternating magnetic field. This vibration produces an electromagnetic wave that radiates from the body at the speed of light. A blackbody is defined as an object that absorbs all radiation that impinges on it at any wavelength. Kirchoff’s law states that any body that is capable of absorbing all radiation is equally capable of the emission of radiation. An example of a blackbody radiator is a lightproof box with a small hole with the following characteristics: •
•
Any radiation that enters the hole is scattered and absorbed by continued reflection from within the box, resulting in almost no energy escaping. The box when heated becomes a cavity radiator that radiates energy, the characteristics of which are determined only by the temperature.
Planck proposed that the frequency of the radiation emitted, and hence the energy, should be quantized to specific values determined by the frequency:
E = hf =
hc [J ] λ
(87.1)
Even though the radiation energies are quantized, there are so many that they continue into the microwave band. The most probable frequency is determined by equating to zero the first derivative (with respect to λ) of Planck’s equation:
λm =
2898 [μm ] T
(87.3)
This is known as Wein’s law. It states that the higher the temperature, the shorter the radiated wavelength. There is a known relationship between the surface temperature of an object and its radiant power. This principle makes it possible to measure the temperature of a body without physical contact with it. Medical thermography is a technique whereby the temperature distribution of the surface of the body is mapped within a few tenths of a Kelvin. The human skin approximates to within 1% of a blackbody radiator, and so a radiation thermometer can accurately detect the temperature of the skin.
87.6.2 INFRARED THERMOMETERS These devices are designed to yield the average amount of infrared energy (often an average temperature) over a small area, often referred to as a single point. The size being sensed depends on the design of the instrument’s optics and the distance from the surface to be measured (Figure 87.2).
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
0.6 cm @ 0.61 m
2.22 cm @2m
87.6.4 INFRARED RADIOMETERS
5.55 cm @5m
S
0.24 in @ 2 ft
0.87 in @ 6.6 ft
757
2.18 in @ 16.4 ft
D
FIGURE 87.2 Distance-to-spot size (D:S) ratio.
The optical system of an infrared thermometer collects the infrared energy from a circular measurement spot and focuses it on the detector. Optical resolution is defined by the ratio of the distance from the instrument to the object compared to the size of the spot being measured (D:S ratio). The larger the ratio’s number, the better the instrument’s resolution, and the smaller the spot size that can be measured. The laser sighting included in some instruments only helps to aim at the measured spot. Examples of some infrared thermometers are illustrated in Figure 87.3.
87.6.3 FIXED MONITORING SYSTEMS Fixed monitoring systems consist of scanning using a thermal image camera capable of producing surface temperature profiles of objects, producing a picture made up of hundreds of points across the detector’s surface. Line scanners provide a single-dimensional view or line of comparative radiation. There are also examples of thermal imaging in surgery.1,2 Recently, increased awareness in this field due to development associated with the charge coupled device (CCD) camera, has greatly reduced in detector physical size and now extends vision accurately through the smallest of openings, with the added capabilities of recording and trending. These charge coupled devices also form the basis of electronic optical systems for noncontact visual measurement and inspection in medical applications.
FIGURE 87.3 Examples of modern infrared thermometers.
Infrared thermal imagers are instruments for detecting, measuring, and recording the thermal emissions from a surface without contacting it, at a safe distance and at varying speeds. The term radiometer is generally, though not always, applied to devices that measure infrared radiation. Figure 87.4 illustrates a number of these devices, many of which resemble conventional video camcorders. There are currently a myriad of infrared radiometers available operating within different wavelengths of the infrared spectrum, dependent primarily on the type of detector used (Figure 87.5). These devices can for the benefit of this section be classified as follows: • • •
Long-wave infrared (LWIR; 7.5 to 14 μm) Mid-wave infrared (MWIR; 3 to 5 μm) Short-wave infrared (SWIR; 0.9 to 1.7 μm)
The optimum wavelength of an infrared radiometer is therefore determined by the wavelength distribution of the emitted radiation and type of detector. Another consideration is the transparency of the atmosphere to the transmission of infrared radiation between radiometer and object. At particular wavelengths there is a lack of radiative transparency within segments of the infrared spectrum.3 There are high levels of infrared transparency, at 3 to 5 μm (with poor transmission at 4.2 μm, due to carbon dioxide absorption) and 7.5 to 14 μm. This seems to coincide as the two common wavelengths of a number of different radiometers, although not exclusively. For thermal measurements over short ranges, such as those found in the examining room, laboratory, or operating theater, it is possible to work outside these wavelengths. As well as two common wavelengths there are two types of infrared radiometer: mechanical scanning and focal plane arrays (FPAs). Mechanical scanners are slowly being superseded by focal plane arrays for a number of reasons, partly highlighted in Table 87.2. Plassman and Jones5 have identified that there are currently a number of possible deficiences surrounding some thermal imaging cameras, which need to be
758
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 87.4 Examples of modern infrared radiometers. (From Snell Infrared.)
Gamma rays 0.1 A
X-rays 1A
Ultraviolet
1U A 100 A 0.1μ
Radio EHF SHF UHF VHF HF
Infrared 1μ
10 μ 100 μ 0.1 cm
1 cm 10 cm 1 m
MF
LF
VLF
10 m 100 m 1 km 10 km 100 km Wavelength
Infrared 0.4
0.6
0.8
1
1.5
Measurement 2
3
Region 4
6
8
10
15
20
30 Wavelength μm
FIGURE 87.5 Infrared measurement region. (From Raytek.)
monitored especially when used in medical applications. Summarized these are:
• •
• •
•
Specifications — offset ±2°C even after manufacturer calibration. Stability • Short term: drift after internal calibration possible. • Medium term: even after switch-on time specified by manufacturer. • Long term: over several hours/days. Range — fluctuation of around ±1°C possible within the “human” range.
Uniformity — optical limitations typically ±0.5°C. Scene — flooding effect ±0.2°C.
87.6.5 MECHANICAL SCANNING ARRAYS
AND
FOCAL PLANE
The design of thermal imaging cameras has moved away from mechanical scanning technology to focal plane arrays. All radiometers have a detector array or array detectors, and optics to form an image on the detector (Figure 87.6). The term staring array is also associated with FPAs, but refers specifically to the use of array detectors, each of which looks at one point of the total image (Figure 87.7).
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
759
TABLE 87.2 Radiometer Types: Advantages and Disadvantages
Advantages
Disadvantages
Mechanical Scanner (Cooled Detector)
Focal Plane Array (Cooled Detector)
Microbolometer (Uncooled Detector)
Uses a few relatively cheap detectors Cooler tends to be more reliable Excellent measurement accuracy Detector needs cooling Requires electric power for operating mechanical scanner Not very portable
Less weight due to a reduction in mechanical parts (no mechanical scanner) — more portable, lighter, reliable with improved battery life Excellent resolution (dependent on number of pixels) Generally has fast frame rates, 30–60 Hz Complex expensive detector arrays with multiplexing electronics Detector needs cooling Start-up times; cooler has to work hard to maintain lower temperatures because of the larger mass of detector material
Mechanical simplicity Improved reliability of camera — no cooling required
Internal temperature stability Compensation required for internal thermal noise Fill factors can be as low as 40%
From Thomas, R.A., Handbook on Thermography, Coxmoor Publishers, Oxford, 1999.
Lens
Filter
Detector
Electronics
Outputs
Object
FIGURE 87.6 Simplified layout of a radiometer.
Detector
Lens
Operator controls Video display
Video processing
FIGURE 87.7 Focal plane array. (From FLIR.)
The advantages and disadvantages between these systems are reviewed below: •
•
Scanning systems generally provide single- or two-dimensional images and often are fixed in particular locations. Focal plane arrays (staring systems) are predominantly used in industry and medicine as portable IR camcorders with cooled or
uncooled detectors, most providing radiometric capabilities. Recent technology has resulted in staring systems being used in fixed industrial locations providing three-dimensional image technology. As mentioned previously, there exist a plethora of thermographic cameras, some offering similar performance characteristics. More recently the uncooled
Handbook of Non-Invasive Methods and the Skin, Second Edition
microbolometer detector has become a popular option in industry. To appreciate the implications of these devices, Table 87.2 attempts to illustrate some of the general advantages and disadvantages of a mechanical scanner and focal plane array with and without cooling.
87.6.6 DETECTOR ARRAYS Detector arrays generally have the advantage of no mechanical moving parts. The spatial resolution, and ultimately picture quality, is determined by the number of pixels within the detector array. For example, there are currently two formats commonly employed: 256 × 256, providing 65,536 detectors, and 320 × 240 (320 columns × 240 rows = 76,800 detectors). More recently, a large 512 × 512 pixel platinum silicide (PtSi) focal plane array has been fabricated with CCD/CMOS technology with high performance and good pixel-to-pixel uniformity. Typical pitches between pixels are in the range 20 to 50 μm. Lack of uniformity of the detector elements across the array can affect performance. Individual pixel response characteristics differ considerably across the array. Therefore, pixel correction is required prior to final camera image. This amounts to calibrating each individual pixel, by exposing the array to calibrated surfaces of known temperature.
87.7 DETECTORS In all thermographic systems the type of detector used to convert the incident radiation into a meaningful signal ultimately shapes the functionality of the infrared camera. The actual detector used will depend on the wavelengths to be detected. Thermal sensitivity/resolution or noiseequivalent temperature difference (NETD) is considered to be one of the most useful measurement parameters in medical thermography, particularly with reference to measuring small temperature differences on the skin surface. NETD is defined as the temperature difference that will produce a signal-to-noise ratio of unity. It is a measure of the performance of detector and processing electronics. Grenn measured the NETD for first- and second-generation infrared cameras (mechanical scanners and focal plane arrays) from the analog video outputs with wide field-of-view optics (Figure 87.8). From Figure 87.8 it can be seen that the new detectors reveal a tenfold increase in sensitivity. There is no single figure of merit that will measure the quality of an infrared image, but NETD is widely adopted. The infrared detector can be divided into two groups known as thermal detectors and photon (quantum) detectors. From Figure 87.9 it can be seen that each detector has a different response profile compared to the operating wavelength.
0.35 0.30 Min: max NETD (K)
760
0.25 0.20 0.15 0.10 0.05 0.00 Serial scanner PtSI Uncooled QWIP
InSb
HdCdTe
Camera technology
FIGURE 87.8 Various minimum and maximum noise-equivalent temperature differences. (From Grenn, M.W., Recent advances in portable infrared imaging systems, Proc. 18th Intl. Conf. IEEE Eng. Med. Biol. Soc., Amsterdam, Paper 1091, 1996.)
Different detectors each have a maximum response, but at different wavelengths; the question often asked is: Should I use a short-wave or long-wave camera? Figure 87.9 illustrates that each detector can achieve an optimum performance level (albeit at different wavelengths), and provided the selected camera meets the general requirements concerning picture quality, thermal sensitivity, and accuracy, then cost is probably the determining factor. As mentioned previously, atmospheric absorption of infrared energy generally limits the useful band of infrared detectors; this is why the predominantly used thermal imaging systems operate at 3 to 5 μm and 7.5 to 14 μm, respectively (Figure 87.9).
87.7.1 PHOTON DETECTORS Photon detectors (sometimes known as quantum detectors) convert radiation directly to an electrical signal. In Figure 87.10 the detector and cooler are combined. For example, the absorption of long-wave radiation results directly in some specific quantum event, such as photoelectric emission of electrons from a surface, or electronic interband transitions in semiconductor materials. The output of photon detectors is governed by the rate of absorption of photons and not directly on the photon energy. It is usual to cool the detector down to cryogenic temperatures (77 K, liquid nitrogen; 4 K, liquid helium) to reduce any excessive dark current, resulting in improvedperformance larger detectors, but with smaller response times. Photon detectors exhibit the following characteristics: • • •
Need cooling Very sensitive Very stable
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
761
System response curves
1 0.9
PtSi
0.8
Response
0.7 0.6
Microbolometer
0.5 0.4
GaAs Qwip
0.3 0.2 0.1 0.0
3
4
5
6
7
8
9 10 11 12 Wavelength (μm)
13
14
15
16
17
18
FIGURE 87.9 Typical infrared detector response curves. (From FLIR.)
FIGURE 87.10 Example photon detector with sterling cooler (SC1000). IR photon radiation
Electrical current flow
Electrons
Upper band
Energy gap
Electron movement – lower band Power supply
FIGURE 87.11 Photon detector operation.
When infrared photon radiation falls on the detector, electrons can move freely in the conductive band and contribute to the electrical signal (Figure 87.11). At room temperature all the electrons have thermal movement jumping up and down. The higher the temperature, the more jumping around, resulting in a large number of electrons jumping from the lower to higher bands, contributing to a large signal current called noise current. If the detector is cooled down, the movement of the electrons gets much lower, with just a few of sufficient
energy to jump the gap. The result is low-noise current — significantly reduced unwanted noise current. To derive a suitable signal from the detector and to help the electrons jump the gap, infrared photon radiation is sent onto the detector cell. The photons penetrate the detector material and hit the electrons, which are sent over the gap to the upper region, where they contribute to the electrical signal. This signal will be proportional to the number of photons with enough energy hitting the detector. The energy of the photons needs to be high enough to get the electrons over the gap.
87.7.2 SHORT-WAVE DETECTORS
AND
LONG-WAVE PHOTON
Short-wave photon detectors generally have higher energies and can easily kick electrons over the gap. Therefore, the gap can be made wider; this means that the detector does not need to be cooled down as much as in the long wave, allowing the photons to jump around quite freely because they want to jump the gap anyway (Figure 87.12). A common detector material used in short wave is platinum silicide (PtSi), and quantum wells in the long wave. There are two types of photon detectors characterized by the response caused with interaction to photons of radiation: photovoltaic (generate a potential difference across a junction) and photoconductive (generate free carriers in a semiconductor that in turn increases the conductivity).
87.7.3 PHOTOVOLTAIC DETECTORS As infrared radiation passes near a junction, it is absorbed and gives an electron enough energy to reach the conductive band. This generates an electron in the conduction band and leaves behind a hole that can also
762
Handbook of Non-Invasive Methods and the Skin, Second Edition
Photons require greater energy than the energy in the gap
Photons require greater energy than the energy in the gap
Free electrons
Free electrons Energy gap
Energy gap Trapped electrons Longwave
Trapped electrons Shortwave
FIGURE 87.12 Energy gap in photon detectors. Responsivity
8.75 μm Peak With grating Without grating
6
7
9 10 8 Wavelength (μm)
11
FIGURE 87.13 Photomicrograph of a QWIP detector array with grating (SC3000) and its spectral response curve. (From FLIR.)
contribute to conduction. This process is sometimes referred to as kicking the electron into the conduction band or creating an electron–hole pair. The presence of an electron–hole pair changes the current–voltage relationship of the diode. That change can be monitored to provide infrared detection. Photovoltaic devices need an internal potential barrier with built-in electric field in order to separate photogenerated electron–hole pairs. Such potential barriers can be created by the use of Scottky barriers. Whereas the current–voltage characteristics of photoconductive devices are symmetric with respect to polarity of the applied voltage, photovoltaic devices exhibit rectifying behavior. Typically, photovoltaic detector materials are: • • •
Platinum silicide (PtSi) Mercury cadmium telluride (MCT or HgCdTe) Indium antimonide (InSb)
87.7.4 PHOTOCONDUCTIVE DETECTORS The operation of photoconductive detectors is based on the photogeneration of charge carriers (electron–hole pairs). These charge carriers increase the conductivity of the device material. Typically photoconductive detectors materials are:
• • • • •
Mercury cadmium telluride (MCT) Indium antimonide (InSb) Lead sulphide (PbS) Lead selenide (PbSe) Quantum well infrared photodetector (QWIP)
In most cases photon detectors are cooled to cryogenic temperatures, with the exception of thermoelectric cooling, where 200 K seems to be sufficient, as in 3 to 5 μm MCT detectors.
87.7.5 QUANTUM WELL INFRARED PHOTODETECTORS One of the more recent detectors to appear particularly useful is the quantum well infrared photodetector (QWIP). These devices consist of quantum wells in semiconductor material, where resultant electronic levels can be tailored to absorb radiation in the 3- to 20-μm-wavelength region. Special grating structures are necessary in order to achieve a high quantum efficiency of the detector (Figure 87.13). Some of the characteristics associated with QWIPs are: • • •
Thermal sensitivity, 20 mK at 30°C Spatial resolution, 1.1 mrad Real-time 14-bit digital output
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
763
TABLE 87.3 Characteristics of Three Different Detector Materials QWIP (AlGaAs/GaAs/InGaAs) Wavelength NETD Uniformity Optics Cooling Operability
• •
MWIR/LWIR <20 mK Good Si/Ge 70 K High
MCT
InSb
MWIR/LWIR 20 mK Bad Si/Ge 77 K Good
MWIR 20 mK Good Si 77 K Good
4 kU
Data acquisition up to 900 Hz 320 × 240 pixels (76,800 detectors), FPA
tors suitable for focal plane array operation, where the latter two properties are less critical. The general trend seems to be that microbolometers are utilized more and more, particularly with reference to industry. The uncooled microbolometer array has an absorptive, temperature-sensitive element that is thermally linked to a temperature reference. (Figure 87.14 illustrates an individual detector within the array.) As infrared radiation is absorbed the temperature of the element changes (increases). Measuring the temperature changes provides the incident power. Each detector is freestanding in the air, supported only by two thin legs at each side (Figure 87.15). Some of the characteristics of microbolometers include:
87.7.6 THERMAL DETECTORS In contrast to photon detectors, the operation of thermal detectors relies on a two-step process. The absorption of IR radiation in these detectors raises the temperature of the device, which in turn changes some temperaturedependent parameter, for example, electrical conductivity. There are a number of thermal detectors:
•
• •
Microbolometer: Detects change in resistance; very stable in gain; DC coupled; radiometric. Ferroelectric and pyroelectric: Detect change in capacitor charge; AC coupled; need chopper; difficult to get radiometric.
• •
The major advantage of thermal detectors is that they can operate at room temperature. However, the sensitivity is lower and the response time longer (several milliseconds) than for photon detectors. This makes thermal detec-
• •
Have very good linearity and gain stability Radiation from the internal parts is 10 times higher than object radiation Need an accurate system for keeping track of internal radiation Eliminate effects of internal radiation by an advanced automatic temperature compensation system Based on accurate temperature sensors and a temperature reference Use neural network algorithms
IR radiation
Vanadium oxide (Vox) resistor Supporting leg
A/D readout circuits CMOS substrate
FIGURE 87.15 Cross-section of microbolometer detector.
× 2.30
FIGURE 87.14 Scanning electron micrograph of detector element.
QWIP FPAs require cooling using miniature sterling coolers, normally operating between 70 and 75 K. Table 87.3 illustrates the characteristics of three different detector materials.
•
10 μm
50 μm pitch 0.5 μm 2.5 μm
764
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 87.4 Key Features of a Portable Thermographic System Ergonomics: Portability and ease of use (size, weight, battery life, trailing cables and tripods) Wavelength (long or short) Scanning speed Storage capacity (data recording) Resolution (thermal or spatial) Stability (offset drift) Range (offset, not constant or linear) Uniformity (offset variations due to optical limits) Scene (offset variations due to flooding) Reference temperature Accuracy (bias, offset) Integration time Detector (cooled or uncooled) Screen display (size) Functionality: Measurement features, recording features, analysis tools and control
As a guide, the general features associated with portable thermographic radiometers are shown in Table 87.4.
system is dependent on a knowledge base partly consisting of manually input patient data combined with historical computerized data relating to condition or the result of particular tests. This knowledge based can be used to: • • • • •
An expert system may be defined as a system that stores known expertise on a computer and can make it available to a large number of expert users. It is designed to explain its reasoning in an easy and understandable way. Its most distinguished feature is its growth on an evolutionary basis, improving its expertise as it grows. An expert system uses established facts and design rules to represent expertise. An expert system can also minimize the following: •
87.8 EMERGING TECHNOLOGY 87.8.1 COMPUTER SYSTEMS As technology advances, the emphasis on computerized systems continues to grow alongside detector, optics, and signal processing. There are many reasons for this, including enhanced image quality, faster rate of capture, improved accuracy and repeatability, and simplicity of operation. But perhaps the greatest challenge in the future will be in the validation of results. To this end, the database and archive capability will play a major part in a successful thermographic operation that can stand up to scrutiny. Ring et al. at the University of Glamorgan have realized this potential and developed over recent years a searchable full-text archive of infrared papers (www.medimaging.org).
87.8.2 EXPERT SYSTEMS The added capability of software systems combing with powerful archives will allow the development of a patientbased predictive expert system. Prof. Francis Ring at the University of Glamorgan, U.K., has established a database and archive within the Department of Computing spanning over 30 years. More recently, with Dr. Peter Plassman, an Infrared Atlas of Normals (IAN) is being developed. It is a database of normal human skin temperature distribution extracted from a large number of patients. These images are then “morphed” into a standard shape and various statistical parameters are then calculated. The expert
Identify a trend or anomaly in collected information Identify frequency of monitoring Establish alarm levels Diagnose a condition based on a large amount of data Identify severity of condition
•
A human expert can forget exact types of information (this problem increases with age). A human expert on leaving or retiring from his post will take with him his knowledge; to prevent this, the information is retained at the knowledge base.
87.9 THE FUTURE It is interesting to note that at a recent conference hosted in Korea, Prof. Dr. Jung-Yul Park (2003) suggested that the application of thermal imaging in medicine was experiencing a resurgence — most of the reasoning relates to image processing, standardization, and protocol of images and camera technology — and reported the following reasons: • • •
•
•
Standardized protocol for image acquisition Better understanding of clinical and physiologic mechanisms of various disorders and pain Refined digital still imaging and real-time acquisition of images using mobile sensor and camera Powerful image processing: • 640 × 480 spatial resolution (or better): 1 mrad (1 × 1 mm at 1-m range) • Thermal resolution: 0.08˚C (or better), 8- to 12-bit digitization Variety of techniques to augment diagnostic accuracy and image intensification • Dynamic
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
• • • • •
• Functional • Image fusion techniques Multiple connections for multiple viewings Standardized transfer and retrieve protocol (DICOM) Variety of sensors Better scientific proofs with modern equipment and techniques Less expensive equipment and smaller size
USEFUL TERMS Absolute zero — The temperature (0 Kelvin) of an object defined by the theoretical condition where the object has zero energy. Accuracy — Maximum deviation, expressed in temperature units, or as a percentage of the temperature reading, or as a percentage of the full-scale temperature value, or as a percentage of the target temperature, indicating the difference between a temperature reading given by an instrument under ideal operating conditions and the temperature of a calibration source (per the ASTM standard test method E 1256-88). Ambient operating range — Range of the ambient temperature conditions over which the thermometer is designed to operate. Ambient temperature — The room temperature or temperature surrounding the instrument. Ambient temperature compensation (TAMB) — S e e reflected energy compensation. ASTM — Abbreviation for American Society for Testing and Materials. Atmospheric windows — The infrared spectral bands in which the atmosphere best transmits radiant energy. Two predominant windows are located at 2 to 5 μm and 8 to 14 μm. Background temperature — Temperature behind and surrounding the target, as viewed from the instrument. Blackbody — A perfect emitter; an object that absorbs the entire radiant energy incident on it at all wavelengths and reflects and transmits none. A surface with emissivity of unity (1.00). °C (Celsius) — Temperature scale based on 0° as the freezing point of water and 100° as the vaporization point of water, at standard pressure. °C =
°F – 32 1.8
Calibration — A methodical measurement procedure to determine all the parameters significantly affecting an instrument’s performance. Calibration source — A source (blackbody, hot plate, etc.) of known and traceable temperature and emissivity.
765
Usually NIST traceable in the U.S., with other recognized standards available for international customers. Colored body — See non-gray body. D:S — Distance-to-size ratio. See optical resolution. Detector — A transducer that produces a voltage or current proportional to the IR energy incident upon it. DIN — Deutsches Institut für Normung, the German standard for many instrumentation products. Display resolution — The level of precision to which a temperature value can be displayed, usually expressed in degrees or tenths of degrees. Drift — The change in instrument indication over a long period, not caused by external influences on the device (per the ASTM standard test method E 1256-88). EMC — Electromagnetic compatibility, the resistance to electrical signal disturbances within IR thermometers. EMI/RFI noise — Electromagnetic interference/radio frequency interference; may cause disturbances to electrical signals within IR thermometers. Devices most commonly cause EMI and RFI noise by switching motors (air conditioners, power tools, refrigeration systems, etc.). Emissivity — The ratio of infrared energy radiated by an object at a given temperature and spectral band to the energy emitted by a perfect radiator (blackbody) at the same temperature and spectral band. The emissivity of a perfect blackbody is unity (1.00). ˚F (Fahrenheit) — Temperature scale where °F = (°C × 1.8) + 32 = °R – 459.67. Far field — A measured distance substantially greater than the focus distance of the instrument; typically greater than 10 times the focus distance. Field of view (FOV) — The region, at the target, measured by the IR thermometer. Typically presented by giving the spot diameter as a function of distance from the instrument. Also presented as the angular size of the spot at the focus point. See optical resolution. Focus point (or distance) — The distance from the instrument where the optical resolution is greatest. Gray body — A radiating object whose emissivity is in constant ratio (not unity) at all wavelengths to that of a blackbody at the same temperature, and does not transmit infrared energy. Hertz (Hz) — Units in which frequency is expressed. Synonymous with cycles per second. Infrared (IR)— The portion of the electromagnetic spectrum extending from the far red, visible at approximately 0.75 μm, out to 1000 μm. However, because of instrument design considerations and the atmospheric windows, most infrared measurements are made between 0.75 and 20 μm. Infrared thermometer — An instrument that converts incoming IR radiation from a spot on a target surface to a measurement value that can be related to the temperature of that spot.
766
Handbook of Non-Invasive Methods and the Skin, Second Edition
K (Kelvin) — The unit of absolute or thermodynamic temperature scale where 0 K is absolute zero and 273.15 K is equal to 0°C. There is no degree symbol used with the Kelvin scale, and K = °C + 273.15. Laser — Single or dual lasers are used in some units for aiming or locating the optimum temperature measurement point. LOC — Location. Units with the data-logging feature store data in numbered locations, which can be recalled and reviewed on the display when necessary. Minimum spot size — The smallest spot an instrument can accurately measure. NETD — Noise-equivalent temperature difference. Peak-to-peak system electrical noise normally measured at the output (display or analog), expressed in °F or °C. NIST traceability — Calibration in accordance with and against standards traceable to NIST (National Institute of Standards and Technology, U.S.). Traceability to NIST is a means of ensuring that reference standards remain valid and their calibration remains current. Non-gray body — A radiating object that is partly transparent to infrared (transmits infrared energy at certain wavelengths); also called colored bodies. Glass and plastic films are examples of non-gray bodies. Optical pyrometer — A system that, by comparing a source whose temperature is to be measured to a standardized source of illumination (usually compared to the human eye), determines the temperature of the former source. Optical resolution — The distance-to-size ratio (D:S) of the IR measurement spot, where the distance is usually defined at the focus distance, and the size is defined by the diameter of the IR energy spot at the focus (typically at the 90% IR energy spot diameter). Optical resolution may also be specified for the far field by using values of far-field distance and spot size. Pyroelectric detector — Infrared detector that behaves as a current source with an output proportional to the rate of change of the incident IR energy. Radiation thermometer — A device that calculates an object’s temperature (given a known emissivity) from measurement of either visible or infrared radiation from that object. Reflectance — The ratio of the radiant energy reflected off a surface to that incident on the surface. For a gray body this is equal to unity minus emittance; for a perfect mirror this approaches unity; and for a blackbody the reflectance is zero. Reflected temperature compensation (RTC) — C o r rection feature used to achieve greater accuracy when, due to a high uniform background temperature, IR energy is reflected off the target into the instrument. If the background temperature is known, the instrument reading can be corrected by using this feature. Targets that have low emissivities will reflect energy from nearby
objects, which may result in inaccurate readings. Sometimes objects near the target (machines, furnaces, or other heat sources) have a temperature much higher than that of the target. In these situations it is necessary to compensate for the reflected energy from those objects. (RTC has no effect if the emissivity is 1.0.) Relative humidity — The ratio, expressed as a percent, of the amount of water vapor actually present in a sample of air to the greatest amount of water vapor possible at the same temperature. Repeatability — The degree to which a single instrument gives the same reading on the same object over successive measures under the same ambient and target conditions (per the ASTM standard test method E 125688). Resolution — See temperature resolution or optical resolution. Response time — A measure of an instrument’s change of output corresponding to an instantaneous change in target temperature, generally expressed in milliseconds, for 95% of full-scale temperature indication (per the ASTM standard test method E 1256-88). The specification for Raytek instruments also includes the average time required for software computations. Scatter — See size of source effect. Size of source effect — An undesirable increase in temperature reading caused by IR energy outside the spot reaching the detector. The effect is most pronounced when the target is much larger than the field of view. Spectral response — The wavelength region in which the IR thermometer is sensitive. Spot — The diameter of the area on the target where the temperature determination is made. The spot is defined by the circular aperture at the target, which allows typically 90% of the IR energy to be collected by the instrument, as compared with the 100% spot diameter, which is defined by the IR energy collected from a very large target. The actual size and distance to the target for the 100% spot diameter is specified in the calibration procedure for each instrument. Target — The object upon which the temperature determination is made. Temperature — A degree of hotness or coldness of an object measurable by a specific scale, where heat is defined as thermal energy in transit and flows from objects of higher temperature to objects of lower temperature. Temperature resolution — The minimum simulated or actual change in target temperature that gives a usable change in output or indication (per the ASTM standard test method E 1256-88). Thermal shock — A short-term error in accuracy caused by a transient ambient temperature change. The instrument recovers from its accuracy error when it comes back into equilibrium with the new ambient conditions.
Sensors and Handheld Devices for Surface Measurement of Skin Temperature
Time constant — The time it takes for a sensing element to respond to 63.2% of a step change at the target. Transfer standard — A precision radiometric measurement instrument with NIST traceable calibration in the U.S. (with other recognized standards available for international customers) used to calibrate radiation reference sources. Transmittance — The ratio of IR radiant energy transmitted through an object to the total IR energy received by the object for any given spectral range; the sum of emittance, reflectance, and transmittance is unity. Warm-up time — Time, after turn-on, until the instrument will function within specified repeatability (per the ASTM standard test method E 1256-88).
ACKNOWLEDGMENTS The author thanks the following organizations: Snellinfrared, Land Infrared, Raytek, and FLIR.
767
REFERENCES 1. Campbell, P.A. et al., Real-time thermography during energized vessel sealing and dissection, Surg. Endosc., 17, 1645–1645, 2003. 2. Gibson, B.J., Thermal Imaging during Laser Surgery, Ph.D. thesis, University of Glasgow, Scotland, 1995. 3. Runciman, H.N., Thermal Imaging, CRC Press, Boca Raton, FL, 2000. 4. Thomas, R.A., Handbook on Thermography, Coxmoor Publishers, Oxford, 1999. 5. Plassman, P. and Jones, C., Reliability of quantitative measurements in medical thermography, 8th Congress of the Polish Association of Thermology, 19–20 March, 2005, Zakopane, Poland.
88 Thermal Imaging of Skin Temperature E.F.J. Ring Medical Imaging Research Group, School of Computing, University of Glamorgan, Pontypridd, United Kingdom
CONTENTS 88.1 Historical Background ...........................................................................................................................................769 88.2 Human Body Temperature and the Skin...............................................................................................................770 88.2.1 Mean Skin Temperature ............................................................................................................................771 88.3 Skin Temperature Measurement............................................................................................................................772 88.3.1 Conduction.................................................................................................................................................772 88.3.2 Convection .................................................................................................................................................772 88.3.3 Radiation ....................................................................................................................................................772 88.3.4 General Principles......................................................................................................................................772 88.3.5 Skin Temperature Measured by Conduction.............................................................................................772 88.3.5.1 Thermocouples ...........................................................................................................................773 88.3.5.2 Thermistors.................................................................................................................................773 88.3.5.3 Active Contact Thermometry.....................................................................................................773 88.3.5.4 Liquid Crystal Contact Temperature Measurement...................................................................774 88.4 Principles of Thermal Radiation ...........................................................................................................................775 88.4.1 Skin Emissivity ..........................................................................................................................................775 88.4.2 Radiative Heat Transfer from the Skin .....................................................................................................777 88.4.3 Thermal Imaging Systems.........................................................................................................................777 88.4.4 The Examination Room.............................................................................................................................777 88.4.4.1 Image Capture ............................................................................................................................777 88.4.4.2 Image Display ............................................................................................................................778 88.4.4.3 Image Analysis ...........................................................................................................................778 88.4.4.4 Reference IR Images..................................................................................................................778 88.5 Applications of Thermal Imaging .........................................................................................................................778 88.5.1 Thermal Patterns of the Skin.....................................................................................................................778 88.5.2 Increased Skin Temperature ......................................................................................................................778 88.5.3 Thermal Symmetry ....................................................................................................................................780 88.5.4 Decreased Skin Temperature .....................................................................................................................780 88.5.4.1 Raynaud’s Phenomenon .............................................................................................................780 88.5.4.2 Neurological Dysfunction ..........................................................................................................781 88.6 Conclusion .............................................................................................................................................................782 References .......................................................................................................................................................................782
88.1 HISTORICAL BACKGROUND The association between temperature and disease can be traced to the earliest records in history. Fever was observed as a natural phenomenon, which could be detected by touching the skin. It was also claimed that if wet mud was applied to the skin, rapid drying out in a localized area might indicate the presence of an underlying disease. The oldest known record may be that of a papyrus, dating back to the 17th century B.C. This alludes
to practices that could be as old as 29 B.C., when suppurating wounds were described as hot, and constantly issuing heat detectable by the human hand. It is now known that human fingers can only discriminate approximately 1°C under certain favorable conditions. Throughout the 17th century A.D., the thermometer evolved as an indicator of temperature level. While the application to medicine had previously received little attention, a few pioneers had experimented with the Galileo 769
770
Handbook of Non-Invasive Methods and the Skin, Second Edition
thermoscope. The Florentine glass blowers were ingenious in their variety of designs and had produced a density thermometer, which could be tied to the body surface. Beads, suspended in a liquid within the sealed bulb of the instrument, were observed to rise or fall at given levels of heat. Considerable progress was made by Carl Wunderlich, who introduced a clinical thermometer for axillary temperature in 1871. Working systematically on his sick patients in Leipzig, he recorded the course of temperature changes on paper with over 10,000 observations.1 The hitherto suspicion of temperature observation that was independent of touch and clinical judgment was finally laid. Over 100 years later, medical practitioners throughout the world regard temperature measurement to be a first-line clinical observation. Attempts to measure skin temperature per se had not, however, been so successful. Progress ultimately came from the work of Seebeck, who in 1821 established the principles of thermoelectricity. Four years later, he constructed the thermocouple by the fusion of two dissimilar metals, which produced a current relative to the temperature of the coupled elements. It was many years later, closer to the turn of the 19th century, that medical and physiological researchers began to study temperature of the skin of man and animals. Multiple thermocouples and (later) chart recorders were used, but their data were difficult to interpret. Thermopiles with increased sensitivity were found to be of value for noncontact skin thermometry as late as the mid-20th century. Important progress was made in 1934 by J.D. Hardy, an American physiologist, who determined that human skin behaves as a near blackbody radiator, with an emissivity close to unity.2 In the late 1950s, thermal imaging devices that had been developed during World War II were declassified in Europe and the U.S. These systems produced an image of skin temperature, which opened up a new era in the functional study of the skin and human thermoregulation. The original pioneering work in thermal radiation can be attributed to Sir William Herschel in 1800.3 He detected the presence of heating rays beyond the visible red of the spectrum in a classic experiment with a prism and a set of thermometers (Figure 88.1). His son, John Herschel, in 1840, substantiating his father’s observations, created the first thermal image, using a suspension of carbon in alcohol and a lens placed in a beam of natural sunlight.4 He introduced the word thermogram (Figure 88.2). It is interesting to note that John Herschel’s closest and lifelong friend was Charles Babbage, a founding father of the computer. Modern thermograms are produced with the aid of computer false color, and resulting temperature measurements are the ultimate product of three famous men who worked together. The development of thermal imaging from the original military systems has been dramatic. The advent of solidstate circuits, high-definition oscilloscope and liquid
FIGURE 88.1 Sir William Herschel.
FIGURE 88.2 John Herschel.
crystal displays, and finally small image processing computers has resulted in reliable compact cameras for passive infrared (IR) measurement. It is therefore possible to study skin temperature distribution from many aspects of health and disease using accurate noncontact methods.
88.2 HUMAN BODY TEMPERATURE AND THE SKIN The skin plays a vital role in the ecology of man. As the human body is homeothermic, it constantly self-regulates to maintain a steady core temperature. The skin functions as a thermal interface between the body and its environment. Skin temperature is therefore influenced by both internal and external conditions. The skin itself produces
Thermal Imaging of Skin Temperature
a very small quantity of heat, and this is evenly spread over the total surface. In most situations, a temperature gradient can be measured over the skin surface; the temperature is usually maximal at the head and minimal at the lower extremities. This difference in temperature, which becomes increasingly marked as ambient temperature falls, is the result of a complex mechanism based on the balance required by the body for thermal regulation. This may be overridden by specific local demands, particularly in the extremities. The hyperemic flush in the fingers after handling a cold substance is an example of this. Blood convection is the main warming agent of the skin, transferring heat from the core. Localized changes in blood flow can therefore induce changes in skin temperature. Some heat is transferred to the skin by conduction from the underlying organs and tissues. The skin temperature at a given site is therefore dependent on (1) the vascular supply to the area (at that time) and (2) the thermal conductivity of the subcutaneous structures. In certain pathological conditions, increased metabolism (for example, of a tumor) may generate extra heat to a localized area. The degree to which each area of the skin surface will exchange heat with the environment partly depends on skin anatomy. The forehead is different from the intermedial aspect of the thighs, where the surfaces are radiating to one another. Body position will also modify the thermal distribution of the skin, by increasing exposure to some areas and decreasing others (Figure 88.3). Heat loss can be increased when the subject is lying on a bed or other flat surface. If that surface becomes an open mesh, increased heat exchange with the ambient will occur. Conversely, a subject in the crouching position will lose less heat than in the normal sitting position on a chair. The postural position should therefore be standardized for all skin temperature measurements on a given subject, even when measurement is confined to one specific area.5
0.84
0.66
FIGURE 88.3 The effect of posture on the ratio of radiating surface of the skin.
771
Since heat is also generated by exercise, skin temperature is usually measured after a period of rest, to minimize the effects of physical activity. Food and drink are usually quoted as having an effect on skin temperature. In practice, a moderate meal taken without alcohol or a hot drink may have only transient effects. Alcohol flushing, particularly of the face and extremities, can be marked for 1 hour or more. In certain physiological states, a hot drink, if taken quickly and in sufficient quantity, can also have a measurable effect on these areas. Thermal neutrality of the human body occurs when the subject is at rest and thermoregulatory mechanisms are inactive. Basal heat production and heat loss are then equal. This state can usually be achieved at an ambient temperature of 30˚C at low humidity, <30%, and air movement of less than 0.5 ms–1. At all other temperatures, the skin will be at a different thermal state for a given site in a given ambient temperature. In temperate countries, skin temperature is often measured in a cold environment. If the temperature is too low, <17°C, vasoconstriction occurs and shivering may be induced. At higher temperatures, e.g., >25°C, sweating commences, thus creating surface heat loss by evaporation. The optimal temperature must therefore lie between these observable points. They differ with climatic and racial variation and may change in a given subject due to acclimatization. In Northern Europe, most subjects are comfortable for skin temperature measurement in an environment of 20 to 22°C when partially clothed. Higher temperatures are adopted for thermal comfort levels in warmer climates.
88.2.1 MEAN SKIN TEMPERATURE Physiologists have for some time debated the question of how to obtain a meaningful single value representing the state of thermal balance in a subject. This is a theoretical concept that has many inaccuracies. Mean skin temperature is usually based on a series of skin temperature measurements. There is little agreement over the optimal sites selected. To relate skin temperature variation to the underlying structures, a coefficient is used for each measurement site. A large, simple anatomic surface, such as the abdomen, requires a single measurement with a high coefficient. The extremities, which are more complex, require a greater number of measurements with a lower coefficient. Mitchell and Wyndham6 have reviewed some of the different methods. They concluded that each of the published methods, though different, could be used for most applications. Later studies, using techniques such as infrared thermography, have shown that relatively few sites can be used — upper abdomen, medial forearm, sacroiliac area, and posterior knee — for a reasonable estimation of mean skin temperature.7 Infrared thermal imaging of the body surface can now be used to calculate mean body surface temperature with improved
772
Handbook of Non-Invasive Methods and the Skin, Second Edition
certainty. The whole of the anterior, lateral, and posterior surface of the body can be imaged in single views with the use of wide-angle IR-transmitting lenses.
88.3 SKIN TEMPERATURE MEASUREMENT Skin temperature is often required to assess the change induced by an external or internal stimulus within a given area. It may also be required as a comparison with a similar contralateral or reference point. There are three fundamental methods of heat transfer, all of which are used in the study of body temperature.
88.3.1 CONDUCTION The earliest sensors applied to the skin were able to demonstrate changes in fluid density by heat conduction. Fluid thermometers, and later more popular thermistors and thermocouples, placed on the skin, record by conducted heat. The latter are well suited to long-term temperature recordings from one site over several minutes to hours. Temperature distribution of a small area may be demonstrated with liquid crystal sheets, which change color according to temperature.
88.3.2 CONVECTION Schlieren photography is a technique used to study the heat flow from the body surface. The technique is used by physiologists to study heat loss in high convection currents, e.g., industrial machines, helicopters, and arctic survival studies.
88.3.3 RADIATION The skin is a highly efficient emitter of infrared and microwave radiation. Radiation thermometers and infrared imaging systems provide high-speed two-dimensional temperature recordings. Microwave radiometers may be used for temperature measurement from deeper tissues. Most thermal radiation systems are noncontact, and therefore do not interfere with the skin itself.
88.3.4 GENERAL PRINCIPLES Any form of temperature measurement can affect the heat exchange between the skin and its surroundings. In contact thermometry, the intrinsic heat of the sensor can influence the skin. In radiation detection, the sensor must be distanced from the skin to avoid a direct thermal influence. In practice, all temperature measurement techniques should be so designed that the influences introduced by the method are smaller than the acceptable levels of error. There are particular problems with surface contact
sensors, for differences in pressure on the skin affect not only the thickness and position of the tissue layers, but also the pattern of local blood flow. Invasive techniques, e.g., needle thermocouples, are even more disturbing, with changes created by the release of histamines, etc., by hematoma or by healing tissues with modified thermal properties. A reproducible technique is essential for all measuring procedures, particularly for temperature. It is important that all parameters that can be altered should be noted and repeated for each measurement. The investigator should know the effect that changing parameters would have on the technique and results. The errors introduced by the detector or instrument should also be defined before conclusions are drawn from the measurement data. These technical conditions may appear to be obvious, but are frequently overlooked. As the magnitude of change decreases, so the requirements for the scientific methodology become more critical. In many cases, skin temperature changes may be small, so that good technique is of great importance.8
88.3.5 SKIN TEMPERATURE MEASURED CONDUCTION
BY
Human skin is a dynamic organ, which means that temperature measurement by any contact method is not always a simple process, or as predictable as measuring an inanimate object such as metal. Contact thermometry requires that the detector and the skin surface should equilibrate, usually from the warm skin to the cooler detector. Passive systems using thermistor and thermocouple probes work in this way. The measurement probe thus cools the skin on contact until an intermediate temperature is reached between that of the detector and the skin surface. In most cases, the mass of the detector is very much less than that of the skin, so the resulting temperature measured is close to the initial skin temperature prior to contact. The temperature–time function of the equilibrating process is exponential and not linear. The final stable temperature is reached very slowly; therefore, the exact endpoint cannot be precisely defined. Most instrument response times are given as a fraction or percentage, i.e., 90% of the total response. Some thermocouples designed for skin temperature measurement are actually calibrated in water. The response time in water will be less than that of surface contact with the skin. It is generally considered that thermistors are more suited to long-term temperature measurements at a given site. Thermocouples with fast response times are preferred for short-term local use. A thermally conducting paste may be required as a coupling medium, to improve stability of long-term recording (e.g., intensive care monitoring).
Thermal Imaging of Skin Temperature
88.3.5.1 Thermocouples Certain metals produce a thermoelectric voltage difference when brought together in a simple circuit. Two junctions can be used, one making contact with a reference temperature, the other with the unknown (e.g., skin). When the voltage is amplified, the temperature difference can be accurately shown. However, the relationship between thermocouple voltage and temperature is not linear. For this reason, the selection of suitable metals for a specific temperature range is limited. The most commonly used are iron, copper, nickel, and a copper–nickel alloy, constantan. Copper–constantan produces a voltage charge of 42 μV/°C within the body range of 30 to 40°C. Solid-state amplifiers and digital displays have greatly reduced the size and improved the convenience of these devices.
773
smaller the probe, the more rapidly it will equilibrate with the skin. Calibration of contact thermometers used on the skin should be made by surface coupling rather than immersion. The most suitable method will be against radiation measurement of a surface at known emissivity.9 It should also be noted that measurements made on the living human skin surface will be subject to the microclimate around the area to be measured. The ambient temperature and the air movement within the room in turn will affect the measurement. The manufacture of semiconductors has reached an advanced stage. It is now possible to have inexpensive disposable sensors for clinical use. One Swedish instrument for oral temperature uses disposable thermistors in a paper strip that can be clipped into a power-and-display handle (Figure 88.4).
88.3.5.2 Thermistors These detectors depend on the electrical resistance or conductance changes of certain semiconductors with temperature. It is now possible to obtain very small sensors that can be made exchangeable. Thermistors may be formed in a variety of shapes and sizes, and are often manufactured for specific probes to be used in body cavities, as well as on the skin surface. The effect of aging on these detectors can produce sizeable errors. It is therefore important that stability tests are made against a known temperature standard at regular intervals. Resistance thermometers require an external power supply so that the current and voltage changes can be measured. There are two categories of resistance thermometry sensors: metal resistors and semiconductor resistors. Metal resistors, normally platinum or nickel wire, have a constant high resistance and prove stable over long periods. The electrical resistance of metals rises in proportion to the square of the temperature. Semiconductors (thermistors) are often inhomogeneous in chemical composition. They are therefore characterized by an exponential dependence of resistance on temperature. Most of the sensors used in the medical range show reduced resistance with increasing temperature. Very small bead resistors of 0.2 mm upward can be constructed. They are often fused to a platinum–indium lead and protected by a glass cover. There is necessarily a flow of current, which may have a thermal effect on the probe. For this reason, they are not recommended for longterm monitoring of temperature. The response time of a thermometer may be important for particular applications. The adjustment of heat between the skin and the sensor is reflected in an exponential change of temperature with time. The time constant involved depends on the conditions of the measurement, which may not be defined by a manufacturer, but the
88.3.5.3 Active Contact Thermometry The general principles described above refer to the passive measurement of skin temperature. It is possible to learn more of the skin’s thermal behavior with a heated probe. These probes, for which there are several designs, can give an approximation of skin microcapillary blood flow, based on thermal conductivity. The probe usually consists of an annular array of thermocouples around a circular heater, which raises the skin temperature by several degrees. The system is first allowed to stabilize with the skin, after which the heater current is applied. A new steady state is reached after a few minutes, and the difference in temperature between the first and second steady states is recorded. In principle, the more blood perfusion present beneath the probe, the more heat is conducted away, resulting in a small temperature difference between steady states. The converse is true with higher temperature differences, showing lower thermal conductivity (Figure 88.5). There are a number of problems with the technique, which should be carefully defined. The heater and
FIGURE 88.4 Oral thermometer with disposable thermocouple.
774
Handbook of Non-Invasive Methods and the Skin, Second Edition
2nd steady state
°C
Thermocouple
Normal Δt 9°c Heater
High perfusion
Δt 4°c
1st steady state
20 mins
FIGURE 88.5 Principle of thermal clearance, where temperature increase beneath the probe is measured between the first steady-state (unheated) and second (heated) stages.
sensors must be as far apart as possible (normally 1 cm) to avoid lateral diffusion of heat. The site chosen must not be varied for comparative studies. Britton et al.10,11 have shown through a mathematical model that both fat and blood flow are major factors that alter the thermal characteristics of the skin for thermal clearance techniques. The diameter of the probe appears to be one of the major determinants in the estimated depth of this particular measurement. A probe design based on a ceramic potentiometer base has been published by Ring et al.12 (Figure 88.6a and b). Recent interest in tcPO2 (transcutaneous oxygen tension) of the skin has resulted in a wider use of this technique on different anatomical sites. The probe includes a heater to dilate the superficial capillaries and a thermocouple to act as a regulating sensor. It is designed to measure the heat required to raise the skin to a fixed temperature of 40˚C in order to assess oxygen tension. In principle, this information can also be used to indicate the thermal clearance of the sample area of skin.
Thermocouple
FIGURE 88.6 (a) Thermal clearance probe with an annular heater and a thermocouple bridge circuit.
Δ IRT −
−
±
I.R.T. +
27
13
8
6
No. Subjects
Control
+
++
+++
u.v. Erythema
4
3 ΔT °C
0
2
1
Forearm
88.3.5.4 Liquid Crystal Contact Temperature Measurement
FIGURE 88.6 (b) Results obtained using thermal clearance on a group of subjects treated with a standard UV lamp exposure, some of whom developed erythema. The thermographic changes were only observed in those subjects showing a marked erythema. This also illustrates that skin color (reddening) does not necessarily indicate an increase in skin temperature.
Cholesteric liquid crystals are substances that exhibit some liquid properties, e.g., low viscosity, combined with other properties of crystals. Their optical properties are temperature dependent, hence their application for thermography. Other forms of liquid crystals exist, such as nematic and sinetic forms, which have different physical properties. The characteristic optical properties of the cholesterics are explained by their helical structure, circular dichroism, and rotating power. When an incident beam of light is directed at a thin film of liquid crystals, double polarization occurs. Linearly polarized light with a specific wavelength is reflected, while another component is transmitted to the next layer. Each molecular layer
reflects a part of the incident white light with the same specific wavelength. When that beam is within the spectrum of visible light, the liquid crystals appear to be colored. The actual wavelength is related to the pitch of the helix representing the structure of the molecule. As temperature increases, the pitch decreases because the molecular layers move toward each other. On cooling, the layers move apart, thus changing the wavelength again. Within the visible spectrum, the cooler colors appear at the longest wavelengths, i.e., red. At higher temperatures, shorter wavelengths, green and blue, are visible. Liquid crystal formulations for medical use are
Thermal Imaging of Skin Temperature
prepared by mixing several esters of cholesterol. The range of temperature of the given mixture is dependent on the proportions of the component esters. Preparing a mixture of a specified temperature range is difficult and therefore expensive.13 Little use is now made of this technique, which is largely unsuitable for clinical application. The convenience and reliability of radiation detection methods have generally superseded the contact liquid crystal techniques for thermal detection.
88.4 PRINCIPLES OF THERMAL RADIATION Infrared radiation is a particular form of electromagnetic energy. It occurs beyond the visible spectrum at wavelengths from 1 μm (near infrared) to 15 μm (in the far infrared). From 15 μm to 1 mm an extreme infrared band occurs, before the radio wavelengths at extra high frequency lead into the UHF and VHP bands, commonly used in sound and television broadcasting. Infrared or thermal radiation is generated by the motion of charged atomic particles in any material whose temperature is above absolute zero, i.e., 0 K or –273°C. In addition to the natural properties of waves, electromagnetic radiation also behaves in the manner of a stream of particles or photons. The amount of energy an object radiates is dependent on two factors, temperature, T, and emissivity, e. Emissivity is an expression of the ratio of the rate of energy emitted to that emitted from an ideal blackbody or perfect radiator at the same temperature. An important principle was established by Stefan in 1879, who found that the total power emitted by a blackbody of emissivity 1.0 is proportional to the fourth power of temperature. In practice, surfaces usually emit and absorb less than an ideal black surface. Emissive power or emittance of a surface at temperature T is defined as Me = εσeT4(W/cm2) where σ represents Stefan–Boltzmann’s constant and has the value of 5.67 × 10–8 Wm2K–4. If the emissivity of a surface is known, the radiant power is a measure of its temperature. When the spectra emitted by a blackbody at different temperatures are recorded, the spectral radiant emittance at each wavelength interval can be seen to pass through a maximum, which moves to shorter wavelengths as the temperature of the body is increased. When the temperature of the body reaches 1000 K, the radiant emittance in the visible part of the spectrum becomes luminous, with color changes according to the mixtures of wavelengths present. For example, a hot white or yellow iron bar taken
775
from a furnace will be at a temperature higher than that if it is red in color. In a theoretical perfect radiator the assumption is made that the surface does not impede the emission of energy from any part of the electromagnetic spectrum. However, a perfect radiator not only emits its total energy without impedance, but also absorbs energy with the same freedom. This theoretical situation does not exist for real bodies or surfaces. The characteristics of the surface do actually exert some degree of influence over the absorption and emission of radiant energy. The emissive properties of a substance can be strongly dependent on wavelength, so that a high-emissivity substance at one wavelength may be lower at another. The human body is a living radiator, with the skin as a dynamic interface between the internal body organs and the environment surrounding the subject. Heat produced by the body must be lost into the environment. While some heat may be exchanged by respiration and excretion, the majority of heat exchange is through the skin. One of the highly organized functions of living skin is to have efficient radiative properties in the infrared region, with a high emissivity. This property is unaffected by color or pigmentation of the skin in the visible range of the spectrum14 (Figure 88.7).
88.4.1 SKIN EMISSIVITY An American physiologist, J.D. Hardy, showed by experiment that living skin on a normal subject will exhibit an average emissivity over the whole spectrum of 0.97 to 0.98.15 A number of workers have reexamined this phenomenon and have given general agreement. It is important that the emissivity of skin should be known, for without such data it is not possible to determine skin temperature by the radiation emitted from its surface. There are many advantages of radiation thermometry as a means of measuring skin temperature. One major advantage is that the sensor does not make contact with the skin, which as a living organ is very reactive to physical contact. Investigations into skin emissivity have been conducted by a number of scientists. These include Buttner,16 Mitchel et al.,17 Watmough and Oliver,18 and Stekettee.19,20 The results of these studies are in approximate agreement, but show many inconsistencies that result in a range of emissivity values of some 8%. In all these studies, infrared radiation was measured in a thermal steady state. The conditions used provided for a heat flow across the surface of the skin, due to the temperature difference between the skin and the ambient space. When both temperatures are equal, the reflection and emission at the surface have the same energy spectrum. As a result, it is
776
Handbook of Non-Invasive Methods and the Skin, Second Edition
Infrared thermography
Microwave thermometry
Visible light
X-rays Ultraviolet radiation
LR. Photography
Red glow 1000 K
Thermal radiation from skin 300 K = 27°C 10−4 3 .105
10−3 3.104
10−2 3.103
Wavelength
0.1
1
10
3 .102
30
3 GHz 0.3 Frequency
cm 100
FIGURE 88.7 A comparative representation of the thermal radiation from the skin, which is a fraction of the intensity of heat emitted by a red glowing object at 1000 K. The relative spectra used for IR thermography and microwave thermometry are shown.
not possible to discriminate between these functions, and therefore to accurately measure emission per se. When a steady-state flux does exist across the skin, a temperature gradient is built up in the tissues. In this case, it is difficult to measure the exact temperature of the skin with a contact thermometer. In reflectance measurement, in a steady state, incident radiation causes a change in heat flow across the skin. This has the resulting effect that the actual temperature of the skin will be influenced. Even when these effects are minimal, they can cause a significant effect on the calculated skin emissivity. When the emissivity value is close to unity, as in the case of human skin, the small difference between radiant energy from the skin and that from a blackbody at the same temperature must be determined. Togawa in 198521 proposed a measurement method for determining skin emissivity in a transient state. The method used is a zero-heat-flow thermometer described by Fox and Solman in 197122 to achieve isothermal conditions in the skin. Radiation from the skin was measured immediately after the probe was removed from the surface. This technique presented a problem in that an accurate calibration was needed between the heat flow thermometer and the radiation detector. It also suffered from the disadvantage that some 15 minutes was required to achieve isothermal conditions beneath the heat flow probe. The improved technique by Togawa in 198923 addressed these problems very effectively. In this later method, they measured the infrared radiation emitted by the skin at a stepwise increment in ambient radiation temperature. As a result, they were able to remove the necessity for absolute calibration between the contact and noncontact devices. At the same time, the measurements could be
made in 1 minute only. Togawa was thus able to show that skin emissivity could be calculated from the radiometric measurements before and after switching the ambient radiation, as long as the temperature of the latter was known. The new device used two conical shades to create a specific ambient over the test skin surface. The radiometer sensed the infrared energy emitted from the skin through an aperture in the peak of the cone. A mechanical device allowed the two cones, one of which was heated to 40°C, to be alternately moved into position between the radiometer and the skin surface within 0.5 second. The radiometer, a HgCdTe (77K) detector with a uniform sensitivity within the 8- to 14-μm range, was sampled every 100 msec. By applying this device to different areas of the human body surface in male and female subjects, an average emissivity value was obtained. Neither the different sites nor the male and female differences were found to be statistically different. Their observed emissivities remained in a range between 0.961 and 0.981, with an overall average and standard deviation of 0.971 ± 0.005. Togawa has discussed in detail why these results differ from some of the earlier published data. The advantages of the Togawa method are improved accuracy of measurement and rapid acquisition of data, making it highly adaptive for clinical studies in pathological conditions. Little is known of the possible changes, if any, in emissivity of human skin in certain disease states. It is not unreasonable to question the effects of some disruptive dermatological conditions on the physical properties of the skin surface. Ulcerated skin lesions are clearly abnormal, once the intact surface of the epidermis is destroyed.
Thermal Imaging of Skin Temperature
88.4.2 RADIATIVE HEAT TRANSFER
FROM THE
777
SKIN
The human body is a thermal organism and is described as homeothermic or self-temperature regulating. Changes in the skin occur to regulate the superficial blood flow necessary to balance the heat flow to and from the ambient. This is necessary to maintain a specific core temperature around the vital organs of the head and thorax. The skin thus forms a dynamic interface between the body and its (normally) cooler environment. As such, it is the principal source of distributing excess heat into the immediate surroundings of the body. It does this by radiating heat to form a microclimate or thin layer of warmth over the body surface. Ideal conditions for measuring skin temperature occur when the skin is in a thermal steady state. In clinical practice some compromise is usually adopted. Light in the near infrared up to 6 μm is partially reflected by skin tissues. The slight differences in reflectance of skin of different pigmentation and color in the near infrared become negligible in the far infrared. However, human skin is moderately transparent to the infrared wavelengths between 0.7 and 1.5 μm down to a depth of 2 mm.16 Veins near the skin surface absorb infrared radiation and may appear dark in reflected light. This can be demonstrated by viewing superficial veins beneath the skin surface through red glass or taking infrared photographs by reflected red light. Some thermal radiation passes through the superficial horny layer of the skin. Infrared temperature measurements of the skin are therefore not entirely dependent on the actual surface temperature, but also on the immediate subsurface of the skin.
88.4.3 THERMAL IMAGING SYSTEMS Electro-optical systems for passive noncontact imaging of temperature are now highly developed. The current technology used is based on the detection of infrared radiation emitted by the skin. The detectors commonly used cadmium mercury telluride, or indium antimonide, which generates an electrical signal that is amplified and displayed. A scanning optical arrangement allows the rapid construction of a two-dimensional image.24 In its simplest form this may be displayed as white = hot, black = cool, or vice versa. Thermal imaging systems for medical use are normally equipped with basic controls over a temperature range of 2 to 40°C and may be calibrated by an internal or external thermal reference. Temperature measurement facilities vary, but may include selected spot temperatures, a temperature difference between two or more spots or regions of the image. Isotherms, which link points at the same temperature in the electronic image, and selectable rectangular or irregular regions of interest are also standard. Calculations from this region may be
for maximum, minimum, or mean temperature or a defined index calculated by a chosen formula. Computer image processing is now almost universal, with provision for color coding of temperature (color isotherms). A conventional color thermogram will use a spectral color range, where blue is cold and red-white is hot. Intermediary temperatures are shown in shades of green, yellow, orange, etc.
88.4.4 THE EXAMINATION ROOM Modern thermal imagers are small and portable, and therefore usable in a clinic or ward. However, unless good environmental conditions can be provided, the image obtained will be of limited value. Ideally, the ambient temperature must be controlled, and not subject to variation, or achieved by high air speeds. Direct sunlight on the subject must be avoided. The patient should, where possible, be positioned in a standard and reproducible way, at least 15 cm from the nearest wall, to avoid reflection of heat. Humidity should be optimal at around 45% relative humidity (RH), since high humidity has a marked effect on increasing sweat function, which will affect the resultant image. Recommendations for the ideal conditions were described in 197925 and 1984.26 A good quantitative and reproducible technique is only achieved when the instrumentation is of proven stability and when the physiological conditions are also correct for the patient. Examination of inflammatory areas of the skin requires a cool environment — 20°C in Europe — whereas neurological and sympathetically mediated reactions should be tested in a warmer ambient temperature to avoid vasoconstriction; 22 to 24°C is used in Europe, while higher temperatures have been used in Japan and warmer countries. Preparation of the patient is important. The areas to be examined should be unclothed. A period of 15 minutes is often used to achieve some form of equilibrium with the environment. The thermal contrast on the skin, temperatures measured over the body surface from head to foot, will be dependent on ambient temperature and time of exposure.27 These conditions should be determined and standardized for any given investigation using thermal imaging techniques. Each stage of the imaging procedure needs to be standardized to a repeatable protocol. 88.4.4.1 Image Capture In clinical thermography the patient is positioned at a short distance from the imaging camera, often between 30 and 1500 cm. The camera should be mounted on a parallaxfree stand. A studio photographic stand is more suitable than a tripod, which is not easily reproducible in position and angle. If a specially adapted room is available, it
778
Handbook of Non-Invasive Methods and the Skin, Second Edition
should have a low stage approximately 15 cm high and large enough for a patient to stand or sit for examination of the lower extremities. This also enables the camera to be kept in a position that is parrallel to the ground for standardization. A stand with a pillar height of 2 to 2.5 m is ideal and enables all views of the human body to be aquired. Ceiling-mounted stands, as used in radiology, can also be used, but are expensive to purchase new. The camera should be fully stabilized before image capture begins; this can take up to several hours, depending on the type of system used. Some of the modern uncooled detector systems can be left running continously, and will then provide extremely stable temperature measurements during clinical use. 88.4.4.2 Image Display A primary infrared image is normally displayed as a graytone image with white as hot and black as cold. In this mode, good vascular or skin hair patterns are readily observed. False color images, which are commonly used in clinical applications, should be based on the rainbow spectrum, where red is hot and blue is cold. Industrial software for thermography frequently displays images in an iron mode, where white is the hottest, then yellow and red in descending order. Color displays in Doppler and nuclear medicine normally adopt the spectral rainbow, and it is therefore less confusing to use the same convention. 88.4.4.3 Image Analysis Once the images have been acquired, they are usually labeled and stored on digital media. Analysis of the image may be obtained by extracting a line scan across the most useful plane of the image or, more commonly, from a selected region of interest. Recent research has shown that the most reproducible regions of interest are those that are selected according to a code based on anatomical description for each coordinate of the box, circle, or irregular shape used. Typically, the mean, maximum, minimum, and standard deviation of temperature data are obtained for each region selected. 88.4.4.4 Reference IR Images Although this technique has been available for some 40 years, there is still a lack of reference data from the normal human body in terms of skin temperature. Most publications in the literature have been devoted to pathological and abnormal changes in temperature patterns. A European initiative to construct a normal reference atlas is in progress to address this.
88.5 APPLICATIONS OF THERMAL IMAGING 88.5.1 THERMAL PATTERNS
OF THE
SKIN
The human body is subject to a wide range of biological variants, many of which influence skin temperature. However, practical experience with infrared thermography shows that under given conditions, human skin temperature will exhibit a predictable pattern from which pathological effects can be measured (Figure 88.8). There is a definite relationship between skin perfusion and temperature.28 In general, well-perfused tissues are warm and will be less subject to temperature reduction in a cooler ambient. Conversely, tissues with low blood perfusion will lose heat more dramatically in a colder ambient. In the extremities, sympathetic control of the vascular system may be triggered by external stimuli, which may be physical (hot or cold) or chemical (vasoactive drugs). Inflammation with increased local perfusion may be present on or below the skin. The congested vascular network may show little or no direct response to humoral or neurological influence. A number of publications are available with useful clinical reference data, including Dermatological Thermography,30 The Thermal Image in Medicine and Biology,31 and A Casebook of Infrared Imaging in Clinical Medicine,32 published in 2003.
88.5.2 INCREASED SKIN TEMPERATURE There are many practical applications of this natural phenomenon in clinical medicine that can be monitored by infrared imaging. Inflammation has been recognized for centuries by heat, redness, pain, swelling, and loss of function.32 Thermal imaging is a reliable technique for monitoring local heat, and therefore identifying the extent, degree, and response to treatment.33–36 Under strictly controlled conditions, inflammatory lesions such as arthritis may be measured and the efficacy of anti-inflammatory drug treatment or surgery (e.g., synovectomy) can be quantified.37,38 Unlike the many subjective parameters used to assess arthritis severity, thermal imaging will reveal the temperature independent of local pain. Therefore, analgesic doses of some agents may improve function, but only true anti-inflammatory doses of a drug will produce a reduction in temperature over the affected area.39 This observation made in 197440 has been used extensively in clinical trials of new nonsteroidal compounds, steroids, and second-line treatments (penicillamine, methotrexate, colloidal gold therapy, etc.). A thermal index that is based on the distribution of isotherms and the mean temperature of a defined region has been used to compare known analgesic and anti-inflammatory agents. Different analogs of prednisolone have been tested on groups of inflamed joints to compare plasma drug levels with local and systemic effects.41 The thermal index has also been
Thermal Imaging of Skin Temperature
779
°C °C 36 34 32 30 28 26 24
e
36 34 32 30 28 26 24
e
FIGURE 88.8 Diagrammatic representation of the temperature distribution of a subject in a 20°C ambient after 20 minutes standing. 6
p<0.0005
5 Thermal index (tibia)
used to demonstrate dose–response curves of antiinflammatory drugs in groups of patients and in experimental arthritis in animals.38 It is one of the few objective tests of inflammation that can be equally applied to animals and man. Increased skin temperature has been quantitatively monitored in joint infection42 and in Paget’s disease of bone.43,44 Certain anatomical sites where bone is close to the skin surface, e.g., forehead and tibiae, may show large temperature increases in active disease. Figure 88.9 illustrates the thermal indices measured over active painful Paget’s disease of the tibia compared with nonpainful disease and controls. Increasing temperature often precedes increased bone pain and vice versa. Temperature may be shown to fall during calcitonin and bisphosphonate therapy, rebounding when the treatment is discontinued and the disease is still active. Studies on erythema and urticarial eruptions have shown that quite large skin temperature increases can be measured. However, the area of visible skin color change may be smaller than the surrounding skin temperature increase.45 A comprehensive collection of thermograms relating to dermatological problems is given by Stuttgen and Flesch.47 Melanomas have been studied by a number of investigators.47–49 These authors have shown that a high-temperature flare with melanoma is frequently associated with malignancy. Ippolito et al.51 studied 10 cases of nodular melanoma (2.1 to 4.5 mm) and investigated the hyperthermal halo using an immunohistochemical technique with mono-
4 3 2 1
Activity at other sites
0 Controls
Tibiae
Tibiae: Severe bone pain
Paget’s disease
FIGURE 88.9 Thermal index recorded over the tibiae of patients with Paget’s disease of the bone. Increased heat is associated with bone pain.
clonal antibodies. In the more invasive melanomas, >2.49 mm, there was a prevalence of T8 cells and inversion of the T4/T8 ratio. This group has also studied a larger number of cases of melanomas with dynamic thermography to study the potential of thermal stimulation in diagnostic investigation. Topical agents of many kinds may be monitored for temperature change. Vasodilators such as prostaglandin and nicotinic acid-based products that cause an increase in skin temperature may be studied. In one study a nicotinate product was found to be an effective medium for improved absorption through the skin of methyl and ethyl
780
Handbook of Non-Invasive Methods and the Skin, Second Edition
C. Thermal gradient °C 5
+
−
5
10
Control
n 150
Raynauds phenomenon
144
Connective tissue
122
Diseases
C.T.G. = (ROI1A − ROI2A) + (ROI1B − ROI2B) 2 1 ROI
ROI
FIGURE 88.10 Scale of values obtained from the dorsal hands of patients with Raynaud’s phenomenon. A standard cold provocation of 1-minute dry immersion in water at 20°C is follwed for a fixed time to measure the temperature gradient at recovery. The more severe the condition, the more negative this index, shown by the connective tissue disease group.
salicylates. Topical anti-inflammatory compounds have also been successfully tested in clinical trials using quantitative thermal imaging to assess their local potency.
88.5.3 THERMAL SYMMETRY The human body when in good health normally has a symmetrical temperature distribution from one side to the other. Studies on normal skin temperature distribution have established that the side-to-side temperature differences are normally less than 04°C, and that the mean temperature differences at the extremities can be even less. However, all studies related to assessment of temperature differences must be performed in a stable and controlled ambient temperature.
88.5.4 DECREASED SKIN TEMPERATURE A reduction in skin temperature may occur in a number of conditions where there is decrease in blood perfusion. On the extremities, this is readily detectable in a unilateral condition, where comparison with the unaffected arm or leg can readily be made. In certain situations, e.g., venous ulceration, the hypothermic area may be localized, revealing the extent of decreased skin perfusion surrounding the lesion. It may, however, be necessary to apply a thermal or chemical stress to provoke a temperature difference. This is commonly applied to the measurement of severity of Raynaud’s phenomenon, where recovery from a brief immersion of the hands in water is significantly delayed. Direct skin cooling by topical application as a gel or aerosol is easily demonstrated with thermal imaging. Products designed for pain relief, e.g., in sports injuries, may be tested for their duration of effect in comparison with placebo volatile substances, or direct ice pack treatment.
procedure is for the patient to rest in a chair at 22 to 23°C ambient temperature for at least 10 minutes. A thermogram of the dorsal surface of the hands is then recorded. This is followed by a 1-minute 20°C water bath immersion of both hands up to the wrists. The hands are kept dry by large disposable plastic gloves (or bags). At the end of the stress period, the hands are allowed to rest for 10 minutes and are then rethermogrammed. Measurement of the mean temperature of two separate regions, fingers to MCP joints and MCP joints to wrist, can be used to indicate the mean temperature gradient (strictly mean temperature difference) from the wrist to fingertips. This gradient records a positive value when reactive hyperemia occurs. In Raynaud’s phenomenon, a negative gradient before cold stress becomes even more negative after stress (Figure 88.10). The degree of negativity is a measure of severity.52 Some centers have modified this method, either by varying the temperature of the water bath, e.g., 10 to 15°C, or in continuously monitoring fingertip temperatures during recovery. The conclusions are largely similar, but results are only comparable when the same strict protocol has been followed. While absolute temperatures over the hand may be influenced by external climatic conditions, the
88.5.4.1 Raynaud’s Phenomenon Thermal imaging is a very efficient technique for noncontact assessment of vasospasticity.51 The generally accepted
FIGURE 88.11 Finger macrothermogram showing cooling from the evaporation of sweat from the skin.
Thermal Imaging of Skin Temperature
781
35.0
FIGURE 88.12 Thermogram of hands in a patient with a neurological problem affecting the right hand, resulting in cold fingers.
stress test overcomes external influences and is a consistent and objective indicator of Raynaud’s phenomenon. The stress test has also been used to test for neurological effects from cervical sympathectomy, brachial plexus injury, stroke, hemiplegia, and chronic regional pain syndrome (Figure 88.12). Localized cold fingers may also occur in hand-arm syndrome (vibration white finger) as a result of damage to the peripheral vessels. It can be shown that by exposing the affected hand to contact with a vibrating plate, ischemia in the affected extremities will occur. Normal hands recover to normal temperature distribution in a short time, unlike the affected fingers or areas, which have a delayed recovery to baseline temperatures.
In the absence of disease or injury and under controlled conditions, the human body shows a high level of thermal symmetry. Independent studies on control subjects have shown that in many cases this may be of the order of 0.3˚C, from left to right over much of the trunk and limbs.53,54 Subjects who are regularly involved in sports may show increased temperature differences due to higher perfusion of the dominant arm or leg. However, in certain cases where peripheral nerve injury or congestion has occurred, a decrease in temperature at the extremity may be evident. This hypothermic region does not necessarily fall into a dermatome distribution, but will sometimes be quite characteristic.55 Critics of this application of thermography rightly point out that this is not a perfect sign for nerve root compression or injury. However, it has been used effectively in conjunction with other investigations, such as computed tomography (CT), magnetic resonance imaging (MRI), EMG, etc., and can be repeated more frequently to monitor progress. In chronic regional pain syndrome, decreased temperatures on the affected limb can be marked — up to 5 or 6°C in ambient temperatures of 22°C.56 Here also, a very localized hypothermia can be demonstrated, which cannot be attributed to simply disuse of the limb. Thermal imaging does offer objective evidence of such peripheral skin temperature changes that cannot be influenced by the patient (Figure 88.13). For this reason, it is a good procedure to use when investigating possible psychosomatic complaints, where normal temperature distribution may preclude a genuine pathology.
R
R
23.0°C
88.5.4.2 Neurological Dysfunction
33.0
25.0°C
35.0
FIGURE 88.13 Thermograms of the feet of a patient with complex regional pain syndrome affecting the left foot before and 6 weeks after successful treatment when thermal symmetry is restored.
782
Handbook of Non-Invasive Methods and the Skin, Second Edition
In posttrauma follow-up, thermal imaging may reveal an abnormal response in the suspect lesion. Where relevant (e.g., injury to one or more digits), cold stress tests can reveal an abnormal reaction that was not indicated by the resting thermogram. Thermal imaging has also been used successfully to confirm the level of limb amputation prior to surgery.57 A poorly vascularized region, showing cold on a thermogram, is likely to result in delayed or impaired healing of scar tissues. A warm, well-perfused area is more likely to heal successfully after surgery. Limited use of the technique has also been made in assessing burn injuries and subsequent skin grafting. A thermogram may give early indications of graft rejection by persistent low skin temperature.
88.6 CONCLUSION Thermal imaging, particularly infrared thermography, is a highly developed technology. Its use in medicine has proved controversial in some areas. Nevertheless, it is indisputably the most efficient means of studying skin temperature since the birth of medicine. Used correctly, under controlled conditions, it can far outweigh the information that is so universally accepted from the clinical or oral thermometer. Since these modern technologies have become available, the clinical knowledge of skin temperature patterns has been extended and clarified. The ability to image discrete cooling effects of sweat pore secretion in real time is one of many new areas for study by thermography. Unfortunately, the highest-resolution thermal imaging is expensive at the present time. Nevertheless, the new technical possibilities with high-resolution thermal imaging and prospects for lowcost portable thermal imaging cameras are proving of value in medical research.58 With the improved knowledge of their potential and better understanding of thermal physiology, they are more likely to find a valuable role in medicine than when the first systems were used in the early 1960s. Unlike the early days of thermal imaging, it is possible to extend the observation of thermal patterns on the skin to fast and efficient two-dimensional measurement of skin temperature. When you can measure what you are speaking about, and express it in numbers, you know something about it. But when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meagre kind. It may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the stage of science. — Lord Kelvin (1824–1907)
REFERENCES 1. Wunderlich, C., On the Temperature in Disease: A Manual of Medical Thermometry, Bathhurst Woodman, W., Trans., New Sydenham Society, London, 1972. 2. Hardy, J.D., The radiation of heat from the human body. III. The human skin as a black body radiator, J. Clin. Invest., 13, 615, 1934. 3. Herschel, J.F.W., Investigation of the power of the prismatic colours to heat and illuminate objects, Philos. Trans., 90, 255, 1800. 4. Herschel, J.F.W., On the action of the rays of the solar spectrum on preparation of silver and other substances. Note 1 on the distribution of the calorific rays in the solar system, Philos. Trans., 51, 1840. 5. Houdas, Y. and Ring, E.F.J., Man and his environment, in Human Body Temperature: Its Measurement and Regulaion, Plenum Press, New York, 1982, chap. 4. 6. Mitchell, D. and Wyndham, C.H., Comparison of weighting formulas for calculating mean skin temperature, J. Appl. Physiol., 26, 616, 1969. 7. Houdas, Y., Sauvage, A., Bonaventure, M., and Guieu, J.D., Control of heat exchange: an alternative concept for temperature regulation, in Regulation and Control in Physiological Systems, Iberall, A.S. and Guyton, A.C., Eds., International Federation for Automatic Control, Pittsburgh, 1973, p. 217. 8. Schonecht, G., Electrical contact thermometry, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1985, p. 35. 9. Magdesburg, H., Regulations for the certification of contact thermometers, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1985, p. 203. 10. Britton, N.F., Barker, J.R., and Ring, E.F.J., An assessment of the thermal clearance method for measuring perfusion, in Recent Advances in Thermology, Ring, E.F.J. and Phillips, B., Eds., Plenum Press, New York, 1984, p. 327. 11. Britton, N.F., Barker, J.R., and Ring, E.F.J., A mathematical model for a thermal clearance probe, IMA J. Math. Appl. Med. Biol., 1, 95, 1984. 12. Ring, E.F.J., Watson, C., and Barker, J., Infrared thermography and thermal clearance of the skin, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1985, p. 133. 13. Flesch, U., Techniques for liquid crystal thermography in medicine, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1985, p. 45. 14. Ring, E.F.J., Skin temperature measurement, Bioeng. Skin, 2, 15, 1986. 15. Houdas, Y. and Ring, E.F.J., Man and his environment, in Human Body Temperature, Plenum Press, New York, 1984, chap. 4, p. 57. 16. Hardy, J.D., The radiation of heat from the human body, J. Clin. Invest., 13, 539, 1934.
Thermal Imaging of Skin Temperature
17. Mitchel, D., Wyndham, C.H., and Hodgson, T., Emissivity and transmittance of excised human skin in its human emission, J. Appl. Physiol., 23, 390, 1967. 18. Watmough, D.J. and Oliver, R., Emissivity of human skin in-vivo between 2.0μ and 5.4μ measured at normal incidence, Nature, 218, 885, 1968. 19. Stekettee, J., Spectral emissivity of skin and pericardium, Phys. Med. Biol., 18, 686, 1973. 20. Stekettee, J., The influence of cosmetics and ointments on the spectral emissivity of skin, Phys. Med. Biol., 21, 920, 1976. 21. Togawa, T., Skin emissivity measurement using unsteady state immediately after removal of a zeroheat-flow thermometer probe, in Proceedings of the 14th International Congress of Medical and Biological Engineering, ESPOO, 1985, p. 1016. 22. Fox, R.H. and Solman, S.J., A new technique for monitoring the deep body temperature in man from the intact skin surface, J. Physiol., 212, 8, 1971. 23. Togawa, T., Non-contact skin emissivity: measurement from reflectance using a step change in ambient radiation temperature, Clin. Phys. Physiol. Meas., 10, 39, 1989. 24. Putley, E.H., The development of thermal imaging systems, in Recent Advances in Medical Thermology, Ring, E.F.J. and Phillips, B., Eds., Plenum Press, New York, 1984, p. 151. 25. Lloyd, J.M., Thermal imaging system types, in Thermal Imaging Systems, Plenum Press, New York, 1975, chap. 8. 26. Anglo-Dutch Thermographic Society, Thermography in locomotor diseases: recommended procedure, Ear. J. Rheumatol. Inflammation, 2, 299, 1979. 27. Ring, E.F.J., Engel, J.M., and Page Thomas, D.P.,Thermological methods in clinical pharmacology, Int. J. Clin. Pharmacol. Ther. Toxicol., 22, 20, 1984. 28. Ring, E.F.J. and Ammer, K., The technique of infrared imaging in medicine, Thermol. Int., 10, 7, 2000. 29. Houdas, Y. and Ring, E.F.J., Temperature distribution, in Human Body Temperature, Its Measurement and Regulation, Plenum Press, New York, 1982, p. 96. 30. Stuttgen, G. and Flesch, U., Dermatological Thermography (English and German text versions), Applied Thermology Series, Verlag Chemie, Weinheim, Germany, 1985. 31. Ammer, K. and Ring, E.F.J., The Thermal Image in Medicine and Biology, Uhlen Verlag, Vienna, 1995. 32. Jung, A., Zuber, J., and Ring, E.F.J., Eds., A Casebook of Infrared Imaging in Clinical Medicine, Medpress, Warsaw, 2003. 33. Ring, E.F.J. and Ammer, K., Infrared thermal imaging in rheumatic diseases: a bibliographic review, Thermol. Int., 11, 161, 2001. 34. Collins, A.J. and Ring, E.F.J., Measurement of inflammation in man and animals by radiometry, Br. J. Pharmacol., 44, 145, 1972.
783
35. Bacon, P.A., Collins, A.J., Ring, F.J., and Cosh, J.A., Thermography in the assessment of inflammatory arthritis, Clin. Rheum. Dis., 2, 51, 1976. 36. Ring, E.F.J., Thermographic and scintigraphic examination of the early phases of inflammatory disease, Scand. J. Rheumatol., Suppl. 65, 77, 1987. 37. Engel, J.M., Thermographische Diagnostik, Rheumatol. Aktuel Rheumatol., 4, 25, 1979. 38. Bacon, P.A., Ring, E.F.J., and Collins, A.J., Thermography in the assessment of antirheumatic agents, in Rheumatoid Arthritis, Gordon, J.I. and Hazleman, B.I., Eds., Elsevier, Amsterdam, 1977, p. 105. 39. Ring, E.F.J., Thermal imaging and therapeutic drugs, in Biomedical Thermology, Gautherie, M., Ed., Alan R. Liss, New York, 1982, p. 463. 40. Bacon, P.A. and Ring, E.F.J., Thermal imaging in assessment of drugs in rheumatology, in Recent Advances in Medical Thermology, Ring, E.F.J. and Phillips, B., Eds., Plenum Press, New York, 1984. 41. Ring, E.F.J., Collins, A.J., Bacon, P.A., and Cosh, J.A., Quantitation of thermography in arthritis using multi-isothermal analysis. II. Effect of nonsteroidal anti-inflammatory therapy on the thermographic index, Ann. Rheum. Dis., 33, 353, 1974. 42. Bird, H.A., Ring, E.F.J., and Bacon, P.A., A thermographic and clinical comparison of three intraarticular steroid preparations in rheumatoid arthritis, Ann. Rheum. Dis., 38, 36, 1979. 43. Bird, H.A. and Ring, E.F.J., Thermography and radiology in the localisation of injection, Rheumatol. Rehab., 17, 103, 1978. 44. Ring, E.F.J. and Davies, J., Thermal monitoring of Paget’s disease of bone, Thermology, 3, 167, 1990. 45. Ring, E.F.J., Quantitative thermal imaging, Clin. Phys. Physiol. Meas., 11 (Suppl. A), 87, 1990. 46. Stuttgen, G., Dermatology and thermography, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1984, p. 257. 47. Stuttgen, G. and Flesch, U., Dermatological Thermography, Verlag Chemie, Weinheim, Germany, 1985. 48. Hessler, C. and Battaglino, G., Prognostic Thermographique des Melanomes Malins Cutanes, Med. Medicale, 142, 115, 1977. 49. Amalric, R., Altschuler, C., Giraud, D., Thomassin, L., and Spitalier, J.M., The value of infrared thermography in the assessment of malignant melanomas of the skin, in Recent Advances in Medical Thermology, Ring, E.F.J. and Phillips, B., Eds., Pleanum Press, New York, 1984, p. 623. 50. Michel, V., Hornstein, O.P., and Schroenberger, A., Infrared thermography in malignant melanoma, Haurtartzt, 36, 83, 1985. 51. Ippolito, F., Di Carlo, A., and Leone, G., Hyperthermic halo in malignant melanoma: immunohistochemical study, Thermol. Osterreich, 2/S, 19, 1992.
784
Handbook of Non-Invasive Methods and the Skin, Second Edition
52. European Association of Thermology Report, Raynaud’s phenomenon assessment by thermography, Thermology, 3, 69, 1988. 53. Ring, E.F.J. and Elvins, D.M., Quantification of thermal images, J. Photogr. Sci., 37, 164, 1989. 54. Uematsu, S., Symmetry of skin temperature comparing one side of the body to the other, Thermology, 1, 4,1985. 55. Uematsu, S., Edwin, D.H., and Jankel, W.R., Quantification of thermal asymmetry. 1. Normal values and reproducibility, J. Neurosurg., 69, 552, 1988. 56. Green, J., Leon Barth, C.A., Mickey, S.T., and Dieter, J., Efficacy of neurodiagnostic studies in patients with lumbosacral and single-leg pain of sciatic distribution of 90 days or more, Pain Dig., 2, 213, 1992.
57. Will, R.K., Ring, E.F.J., Clarke, A.K., and Maddison, P.J., Infrared thermography: what is its place in rheumatology in the 1990’s? Br. J. Rheumatol., 31, 337, 1992. 58. Spence, V.A., Walker, W.F., Troup, I.M., and Murdoch,G., Amputation of the ischaemic limb: selection of the optimum site by thermography, Angiology, 32, 155, 1981. 59. Ring, E.F.J., Video thermal imaging, in Thermological Methods, Engel, J.M., Flesch, U., and Stuttgen, G., Eds., Verlag Chemie, Weinheim, Germany, 1985, p. 101.
Neural Supply
89 Assessment of Cutaneous Pain Lars Arendt-Nielsen Center for Sensory-Motor Interaction, Laboratory for Experimental Pain Research, Aalborg University, Aalborg, Denmark
CONTENTS 89.1 The Ideal Cutaneous Stimulator............................................................................................................................788 89.2 Induction of Cutaneous Pain .................................................................................................................................790 89.2.1 Electrical Stimulation ................................................................................................................................790 89.2.2 Mechanical Stimulation .............................................................................................................................791 89.2.2.1 Tactile .........................................................................................................................................791 89.2.2.2 Impact .........................................................................................................................................791 89.2.2.3 Pressure.......................................................................................................................................791 89.2.3 Chemical Stimulation ................................................................................................................................792 89.2.4 Cold Stimulation........................................................................................................................................792 89.2.5 Heat Stimulation ........................................................................................................................................792 89.2.5.1 Heat-Responding Cutaneous Nociceptors .................................................................................792 89.2.5.2 Thermodes ..................................................................................................................................793 89.2.5.3 Focused Light .............................................................................................................................793 89.2.5.4 Lasers..........................................................................................................................................793 89.3 Assessment of Cutaneous Pain..............................................................................................................................794 89.3.1 Psychophysical Methods ...........................................................................................................................795 89.3.2 Electrophysiological Methods ...................................................................................................................795 89.3.2.1 Evoked Potentials .......................................................................................................................795 89.3.2.2 Nociceptive Withdrawal Reflex..................................................................................................795 89.4 Induction and Assessment of Cutaneous Hyperalgesia and Allodynia ................................................................796 89.4.1 Induction of Cutaneous Hyperalgesia .......................................................................................................796 89.4.1.1 Burn Injury .................................................................................................................................796 89.4.1.2 Cold ............................................................................................................................................796 89.4.1.3 Capsaicin ....................................................................................................................................796 89.4.1.4 Mustard Oil ................................................................................................................................796 89.4.1.5 Mechanical Trauma ....................................................................................................................796 89.4.1.6 UV Radiation..............................................................................................................................797 89.4.2 Assessment of Cutaneous Hyperalgesia....................................................................................................797 89.5 Summary ................................................................................................................................................................797 References .......................................................................................................................................................................797
It has always been the dream of researchers and clinicians to have an objective noninvasive measure of pain — this is not possible and most likely will never be. Pain is a multidimensional unpleasant sensory and emotional experience and cannot as such be represented or described by a single parameter or number. However, different possibilities exist to assess quantitatively and noninvasively various aspects of this complex sensory experience of pain. By measuring different aspects of the pain experi-
ence and by a combination of the various measures, more can be learned about which mechanisms/pathways are impaired and which are functioning normally. The ultimate goal of pain assessment procedures is to obtain a better understanding of mechanisms involved in pain transduction, transmission, and perception under normal and pathophysiological conditions. Such quantitative and differentiated information on pain provides fundamental knowledge, is of diagnostic value, may help in targeting 787
788
Handbook of Non-Invasive Methods and the Skin, Second Edition
treatment, and can be used to evaluate quantitatively the progression of diseases and for the evaluation of treatment efficacies/outcome. A major advantage of the quantitative assessment is that the various pain measures can be followed over time and provide the researcher or clinician with quantitative data. The aim of this chapter is to describe the different noninvasive techniques to activate and assess cutaneous pain and pain mechanisms, and to discuss how to use these techniques to study the excitability of the cutaneous nociceptors (pain receptors) and the interaction between nociceptors and the biochemical environment in the skin (normal and pathological conditions). Noninvasive assessment of cutaneous pain can be divided into three main categories: • •
•
Assessment of ongoing clinical pain (e.g., trauma or inflammation) Assessment of experimentally evoked cutaneous pain for diagnosis and monitoring in the clinic Assessment of experimentally evoked cutaneous pain for basic studies in healthy volunteers under normal conditions and conditions with experimentally induced cutaneous hyperexcitability (hyperalgesia and allodynia)
Assessment of pain intensity, quality, and location may be useful in many clinical situations, but may not provide detailed information about the actual pathways or pathophysiological mechanisms involved. An extended and more advanced approach is to impose some additional
and standardized cutaneous painful stimuli to the patient and evaluate the response to such standardized provocations. This quantitative and experimental approach is termed quantitative sensory testing (QST) or human experimental pain research. By measuring different aspects of the pain experience and by combination of the various measures, more can be learned about which mechanisms/pathways are inhibited/facilitated and which are functioning normally. As normal and pathophysiological pain systems often respond differently to standardized experimental stimuli, it is important to have human experimental pathophysiological models available. Using such models, healthy volunteers can transiently be turned into patients, and hence simulate the pathophysiological conditions seen in real patients. Experimental investigations of such mechanisms can increase the knowledge associated with skin diseases such as inflammatory pain. The different models and possibilities to assess cutaneous pain in experimental and clinical studies are sketched in Figure 89.1.
89.1 THE IDEAL CUTANEOUS STIMULATOR Standardized activation/stimulation of the nociceptive and non-nociceptive pathways is important for quantitative sensory testing in basic experiments or for diagnostic purposes. The advantage of experimental assessment of cutaneous pain is the possibility of applying exactly defined and
Experimental
Clinical
Volunteers
Patients Ongoing
Evoked
Normal
Hyperexcitability
Evoked
Experimental hyperalgesia
Pain
Pain
Modulation
Assessment
Modulation
Assessment Assessment
Clinical measures
Evoked
Clinical and experimental measures
Experimental measures
Assessment Experimental measures
FIGURE 89.1 A sketch of the various conditions where cutaneous pain can be assessed noninvasively. Ongoing clinical pain can be assessed by using various scales (left). In patients, it is possible to evoke additional pain by standardized activation of cutaneous nociceptors and assessment of the responses quantitatively. This provides a possibility to evaluate the excitability of the nociceptors in, e.g., affected and nonaffected skin areas. The right panel shows how experimental cutaneous painful stimuli can be applied to healthy volunteers under normal and simulated pathophysiological conditions (experimentally induced peripheral and central hyperexcitability by topical application of, e.g., capsaicin) and how the responses should be evaluated quantitatively.
Assessment of Cutaneous Pain
precisely controlled stimuli to induce cutaneous pain. There are meanwhile various experimental methods of pain induction. As pain is a multidimensional perception, the reaction to a single standardized experimental stimulus of a given modality can obviously only represent a very limited fraction of the entire pain experience. Therefore, it is mandatory to combine different cutaneous stimulation and assessment approaches to gain advanced differentiated information about the nociceptive system under normal and pathophysiological conditions. A misleading, but still common belief is that the various stimulation methods are interchangeable. When selecting the method of noninvasive pain induction, the various types of experimental pain cannot simply be assumed to be equivalent. It is rather advisable either to examine pain perception using several pain induction methods or to deduce explicitly the appropriate pain induction method from the question under investigation. The main advantages of an experimental pain stimulation approach are: • •
•
•
Stimulus intensity, duration, and modality are controlled and can be repeated over time. Differentiated responses to different cutaneous stimulus modalities can be assessed (multimodal sensory testing approach). The physiological and psychophysical responses can be assessed quantitatively and compared over time. Pain sensitivity can be compared quantitatively between various normal/affected regions.
Ideally, an experimental cutaneous pain stimulus should have the following characteristics: • • • • • • •
Noninvasive and produce no tissue damage. Specific: Activate pain receptors selectively. Sensitive: Be able to measure pain within a range that is physiologically relevant. Measurable in physical units and can establish a relation between stimulus and pain intensity. Variable from zero to maximal tolerable levels. Reproducible and repeatable with no change in the response over time. Applicable and easy to apply and control in laboratory and clinical settings.
Unfortunately, the ideal stimulator does not exist, and hence the most adequate for a particular study should be selected according to pros and cons. Stimuli can have phasic (short-lasting, milliseconds to a few seconds) or tonic (long-lasting, many seconds to minutes) properties. The most frequently used cutaneous pain induction methods are based on mechanical, thermal,
789
TABLE 89.1 Cutaneous Pain Induction Modalities and Methods Electrical Transcutaneous Intracutaneous Thermal Heat Radiant (laser, light, infrared) Contact thermode Cold Cold pressor test Contact thermode Evaporation of gas Menthol Mechanical Brushing/stroking (in case of allodynia) Pinprick Pinch Impact stimuli Pressure Chemical Capsaicin Mustard oil Melittin Bradykinin, serotonin, substance P, and other algogenic substances
electrical, or chemical stimuli, and the various cutaneous pain induction techniques are summarized in Table 89.1. Most of the phasic stimuli can be applied (1) as a single stimulus or as a series of repeated stimuli (to evoke temporal summation, i.e., pain increase over time) and (2) as a stimulus activating a small or large/multiple skin area(s) for the study of spatial summation (pain increase for increased area of stimulation). It should be realized that activation of a nociceptor causes a cascade of events (Figure 89.2 and Figure 89.3). First, the nociceptor membrane (the free nerve ending) has to activate by the given natural stimulus — this stimulus causes a depolarization of the membrane and a membrane potential occurs. If this membrane potential exceeds a given threshold, action potentials are generated (Figure 89.1). Generally, the firing frequency of these action potentials correlates with the stimulus intensity and hence the pain intensity perceived. The action potentials not only travel toward the spinal cord, but also spread antidromically to the nearby peripheral nerve endings of the same nerve fiber branches (Figure 89.3). Together with the receptor activation, a variety of vasoactive peptides (e.g., SP, CGRP) are released from the nerve ending, leading to vasodilatation and plasma protein extravasation.1 The peptides synthesized in the dorsal root ganglion are transported to the nerve endings by axonal transport. These peptides are important for the maintenance of many diseases, and, for example, if the nerve is transected (e.g., by an accident), a psoriasis
Handbook of Non-Invasive Methods and the Skin, Second Edition
Stim. intensity
790
Stimulus at receptor
Depolarisation
Threshold
Depolarisation
Time
Action potential
Receptor potential
Time
Potential in afferent nerve
Time
FIGURE 89.2 Nociceptor activation. A natural cutaneous stimulus of increased intensity causes an increase in membrane potential measured within the nociceptor (free nerve ending). When the receptor potential exceeds a given threshold, action potentials are generated and they travel along the axon and its branches. When the stimulus intensity increases and the receptor potential increases, the firing frequency increases. In general, higher firing frequency causes stronger pain.
Firin
g
Stimulus
Nerve ending
Antidromic activation
Action potentials
DRG
Axo n tran al spor t
Release of peptides
FIGURE 89.3 A given nociceptive stimulus elicits action potentials from the nociceptor (free nerve ending) and causes local release of peptides from the nerve ending. The potentials travel toward the spinal cord and spread antidromically to the nearby peripheral nerve endings of the same nerve fiber branches. This antidromic transmission causes release of more vasoactive peptide from the adjacent free nerve endings and the neurogenic spread is initiated. This explains why an erythema may spread around a site of injury.
plaque can disappear or a rheumatic joint will be less swollen, although the disease manifestations remain the same on the intact side. The released peptides change the environment for the nociceptors and may eventually sensitize the nociceptor and lead to cutaneous hyperalgesia.
of these fibers in humans and quantify the human responses to painful stimuli.4,5 The different stimulation techniques (Table 89.1) will briefly be described together with their advantages and disadvantages.
89.2.1 ELECTRICAL STIMULATION 89.2 INDUCTION OF CUTANEOUS PAIN The pioneering studies by von Frey2 and Goldscheider3 showed that pain could be elicited if specific stimulus modalities and configurations were applied to the skin. Part of the peripheral, physiological correlate of pain came from discoveries in 1936 when Zottermans made the first electrophysiological recordings from the pain fiber in cats. This finding intensified the interest to assess the function
The application of cutaneous electrical stimuli has had a long tradition in experimental pain research and is still widely used. This form of pain induction has frequently been criticized as a nonphysiological technique as the stimuli bypass the receptor transduction and depolarize the afferent fibers directly.6,7 Furthermore, fiber activation occurs artificially by circumventing the nociceptors. As the thick myelinated mechanosensitive afferents are activated at lower stimulus intensities than unmyelinated
Assessment of Cutaneous Pain
fibers, pain cannot be elicited without concurrent activation of a tactile sensation. The electrical stimulation of the tooth pulp was developed as an alternative method for a more selective stimulation of the nociceptive nerve fibers. For cutaneous electrical stimulation the challenge is to minimize the spread of current to adjacent structures. The current, and hence spread, can be minimized by using intracutaneous or ring electrodes. Both constant current and constant voltage have been used for stimulation, but today most often constant-current stimulation is used. Furthermore, the artificial qualities of cutaneous pain and the resulting feelings of fear led to objections to the electrical methods of pain induction. However, Rollman8 pointed out that the strangeness of the pain perception could be desirable in certain experiments, which deal with the influence of certain emotions and cognitions. In recent years, the electrical stimulation devices have been computer controlled. This provides the possibility to deliver preprogrammed series of stimuli with different intensities, duration, and frequencies, which is a great advantage. However, this has caused new problems to arise because, up to now, hardly two studies have been carried out with the same stimuli parameters. The reliability of the electrical pain perception tests is regarded as good.9
89.2.2 MECHANICAL STIMULATION Mechanical painful stimulation can be divided into tactile stimulation and algometry (pressure). 89.2.2.1 Tactile For quantitative testing of tactile sensibility, the von Frey nylon filaments are easy to use. They constitute a series of filaments of varying thickness, calibrated according to the force required to make them bend. The von Frey hairs were originally made by horsehairs of different diameter, but are now replaced by graded nylon filaments eventually attached to a vibrating solenoid.10,11 The von Frey hairs have different bending forces, and depending on the diameter of the hair, the tactile receptors (Aβ) or nociceptive receptors (Aδ and C) can be activated. One method to assess tactile sensation is to apply the hairs in an ascending and descending order of magnitude and to record both the appearance and disappearance threshold. In neuropathic pain, tactile sensibility as measured by von Frey hairs may be reduced in affected skin areas.12 This is typically a finding that may be overseen in a routine neurological examination, where testing for tactile sensibility with a cotton swab may only give a sensation of hyperesthesia (in fact, allodynia to light mechanical stimulation), masking an eventual reduction in tactile sensibility. Another way of assessing the tactile threshold is to determine the
791
value of the bending force of the filament, which is detected 50% of the time it is applied to the skin. With increasing bending force, the von Frey hairs will excite skin nociceptors and may be used to determine tactile pain detection thresholds. The nylon filaments have been used for determination of cutaneous allodynia/hyperalgesia to punctate stimuli in human experimental models.13 The von Frey hairs may also be employed for mapping of areas of secondary hyperalgesia to punctate stimuli (due to central sensitization) in experimental models13 of pain or in a clinical context.14 It has been shown recently that secondary hyperalgesia to punctate stimuli is mediated by conduction in A-nociceptive fibers,15 in contrast to the Aβ-fiber-mediated secondary hyperalgesia to light brush. 89.2.2.2 Impact High-speed impact cutaneous stimulation has been used to activate cutaneous nociceptors where the velocity of the projectile determines the painfulness of the cutaneous impact.16,17 89.2.2.3 Pressure The pressure stimulus is regarded a natural pain stimulus, which induces sensations familiar from everyday experience. A variety of techniques have been developed for pressure algometry. This method is not suitable for specific activation of cutaneous nociceptors, and the main application has been to assess tenderness in myofascial tissues18–20 and joints.21 When using the variable-pressure algometer, a probe is pressed against an area of the skin applying increasing pressure at a constant rate (e.g., 1 kg/s or 50 kPa/s) until the subjective criterion aimed at (e.g., the pain detection or tolerance threshold) is reached, and the current pressure is measured with a spring mechanism or with a more sophisticated electronic pressure transducer.22 When using a constant pressure algometer, for example, the Forgione–Barber pressure dolorimeter, a constant, potentially painful pressure is exerted onto an area of the skin above a bony process until the subjective criterion aimed at is reached and the stimulation time is recorded.23 The pressure pain thresholds vary substantially between skin regions, and methodological studies are necessary for each new location examined.24,25 In some of the commercially available pressure algometers, the pressure application rate can be monitored. This is important for the reliability of the results. However, algometers are needed where the rate and peak pressure can be predefined and automatically controlled.
792
Handbook of Non-Invasive Methods and the Skin, Second Edition
89.2.3 CHEMICAL STIMULATION The chemical pain stimulation, in which algogenic substances are applied or injected into the skin (for example, histamine, bradykinin, capsaicin), is considered hardly suitable for the examination of pain perception as a standard tool due to the fact that it cannot be adequately standardized or precisely controlled. If, however, local skin or muscle hyperalgesia and its treatment or the effect of topical analgesics is of special interest, this form of chemical stimulation may be the method of choice.6,7 A completely different form of chemical pain stimulation was developed by Kobal.26 His method involved activating chemosensitive nociceptors in the nasal mucous tissue via short puffs of CO2 gas. The possibility of exactly controlling the onset of this stimulus even allows for measuring of evoked brain potentials. If the interest is focused on the trigeminal pain system, this chemonasal pain stimulation may be an interesting alternative to other pain induction methods.
89.2.4 COLD STIMULATION Another, equally classical thermal pain test is the cold pressor test, in which the subject immerses the hand or foot into water having a temperature between 0 and 5°C. It is, however, likely that it is not the cold itself, but rather local vascular reactions to it that constitute the pain stimulus. This might be the reason why the cold pressor test has repeatedly produced a dichotomy of pain-sensitive and pain-tolerant subjects, which has never been replicated by use of other pain models.27 Although the reliability of the cold pressor test is now looked upon critically,28 it is still a quite attractive tool in experimental pain research due to its low costs, simple administration, and the similarity of the resulting pain to that of certain clinical pain states.7 Computers can sample continuously the pain intensity score and calculate onset, peak latency/score, and area under the visual analog scale (VAS) curve.29 Cold and cold pain thresholds may also be assessed by Peltier devices. There are many parameters that may influence the results of such a thermal testing. The baseline temperature is an important factor that may influence the ability to discriminate a fall in temperature. In many laboratories, the contact probe is applied at a standard temperature of 32°C, thereby reducing the interindividual variability in thresholds. Few attempts have been made to develop rapid cold pain stimulators suitable for eliciting cold pain. Recently, we have been capable of generating temperature fall times of approximately 50°C/sec by applying a gas to an evaporation chamber (unpublished data). The evaporation is facilitated by coadministration of oxygen to the chamber. A thin silver plate separates the chamber from the skin,
and a heat foil is glued to the plate to provide active heating after each cold pain stimulus.
89.2.5 HEAT STIMULATION Human experimental and clinical pain tests of cutaneous pain have frequently employed heat stimulation for physiological, pathological, and pharmacological assessments. Thus, heat pain constitutes a unique quality in nociception and human studies of cutaneous pain. From a neurobiological basis, heat pain may encompass cutaneous receptors and nociceptive systems similar to those in pathological trauma and inflammatory-induced cutaneous pain. 89.2.5.1 Heat-Responding Cutaneous Nociceptors Cutaneous nociceptors for heat pain have been recently identified as part of the vanilloid receptor system, along with capsaicin and proton receptors, a component of proteins in the membrane of nociceptors. Molecular characterization of capsaicin (vanilloid) receptor (VR1) has added to our understanding of pain-producing chemical and thermal stimuli.30 VR1 is part of the larger transient receptor protein (TRP) in the membrane. VR1 is activated by capsaicin and by a temperature increase when it reaches the noxious intensity range, >43°C. A further increase of temperature to 52°C, tissue damage encounters when a variant VRL-1 is activated over the noxious temperature range. 31 It is known that these heat-related nociceptors not only exist at the peripheral nerve endings, but also in the primary sensory neurons as well as in the spinal cord. Thermal and pain sensations employ similar smalldiameter thin fibers in peripheral transmission of temperature and pain sensations. Microneurogaphy in the study of heat pain in monkey and man32 has characterized the afferent properties. A combination of myelinated A- and nonmyelinated C-fibers can result in five classes of distinctive afferents: A-cold afferent, A-high-threshold mechanical afferent, A-mechano-heat afferent, C-warm afferent, and C-mechano-heat afferent. Heat pain is largely carried out by A-mechano-heat and C-mechano-heat afferents. They are largely polymodal, as these fibers can carry pain induced by thermal, mechanical, and chemical stimuli. A-mechano-heat (A-MH) myelinated afferent carries activation by a heat threshold at 43°C, short activation time of less than 50 ms, and a quick adapting to the heat stimuli. Its conduction velocity is estimated at around 15 m/s. 33 In contrast, the slow-acting C-mechano-heat (CMH) afferent is characterized as in the unmyelinated fiber conduction range (0.86 to 1.25 m/s), 34 in response to the high intensity of mechanical and thermal stimulation. They have a mean activation threshold of 43.6°C and no spontaneous activity below 38°C. 35
Assessment of Cutaneous Pain
793
Aδ
B
C
Aδ
C
C
Aδ
First pain (pricking)
0
0.5
Second pain (burning)
1.0 1.5 Sec.
2.0
Pain intensity
C
0
0.5
1.0 1.5 Sec.
2.0
Pain intensity
A
Pain intensity
The major difference between lies not only in the peripheral conduction velocity, but also in their respective receptor field. For the fast-conducting A-MH afferent, its receptor field is confined largely in hairy skin, with a small receptor field of 1 to 4 mm2. It is noted that in the slowconduction C-MH afferent, the receptor field is generally large, 99 ± 21 mm2, 36 with several points of maximum sensitivity. Nevertheless, both have similar heat thresholds, at 43 to 44°C. However, the C-MH afferent receptor has two distinctive classes, slow or fast adapting, while the A-MH afferent receptor shows only fast adapting. The distinction between A-MH and C-MH afferents/receptors also lies in their physiological mode and role. A-MH afferent shows a high firing rate, at 30 Hz, while C-MH can reach 15 to 20 Hz. Both A-MH and C-MH show contrasting perceptual function. A-MH afferent/receptor codes fast-adapting pricking pain (the first pain sensation) and the C-MH codes slow-adapting burning pain (the second pain sensation) (Figure 89.4). Goldscheider3 introduced heat to evoke experimental pain. One of the first attempts to measure pain thresholds with the application of a hot object with graded temperatures was carried out by Elo and Nikula.37 Hardy et al.38,39 developed the first nontouching heat stimulator based on focused white light. The major disadvantages of light stimulation are wavelength-dependent reflections from the skin. This was in a sense the background for using wavelength-specific lasers, which were introduced in the early 1970s.40
0
0.5
1.0 1.5 Sec.
2.0
FIGURE 89.4 An illustration of the two pain qualities that can occur after cutaneous stimulation. Activation of the thin myelinated Aδ-fibers causes a sharp, pricking first pain, whereas activation of the slow-conduction C-fibers causes a burning, diffuse, dull second pain that occurs 1 to 2 s after stimulation.
89.2.5.2 Thermodes For clinical purposes thermodes based on the Peltier principle (one direction of the current causes cooling and the other heating) are widely used. The advantage of this method is that in addition to cold pain and cold thresholds, the warm, heat pain, and heat pain tolerance thresholds can also be assessed by the same device.41 The disadvantages are that the temperature rise times are low (1 to 2°C/s) and that the device touches the skin during examination, which can be a problem when testing patients in allodynic areas. Contact heat can generally evoke both sharp, pricking first pain and slow, burning second pain. The thermode technique is applicable for studying the differentiated effects of drug and anesthetic procedures.11 Recently, a device based on heating foils has been developed. With this device very rapid temperature increases (30°C/s) can be obtained, depending on the electrical current delivered to the foil.42 The foil has a low thermal capacity and is glued (thermal conducting glue) to a Peltier element by which active cooling (15°C/s) rapidly returns the skin temperature to its baseline value. 89.2.5.3 Focused Light Hardy43 developed the first generation of the dolorimeter, which was a radiant heat device based on focused light from a light bulb. The disadvantage using this technique is, as for the argon laser, part of the light is reflected from the skin. We have recently44 designed a focused light bulb system based on a 1000-W xenon flash lamp. The advantage of this system is that high energies can be generated and the focused light can be transmitted to the skin via a liquid light guide. This stimulator can be used both for psychophysics44 and temporal summation studies and to elicit time-locked brain potentials. 89.2.5.4 Lasers In the mid-1970s Mor and Carmon40 introduced the CO2 laser in cutaneous pain research. Different laser emission sources have been developed for pain research, e.g., argon, thulium-YAG, and laser diodes. Normally, the laser pulses used are of short duration (20 to 200 msec) and elicit a distinct pricking pain, which at the highest stimulus intensities is followed by a burning pain. Nociceptors in the skin are divided into three main classes: C-mechano-heat, type I A-mechano-heat, and type II A-mechano-heat. Type I mechano-heat nociceptors exhibit a slowly increasing response with a latency of several seconds to heat stimuli of high intensity and long duration.45 If the skin is activated by rapid heat stimuli from, e.g., a laser, the type II A-mechano-hea nociceptors are activated,46 together with some activation of warmth receptors (C-fibers).
794
Handbook of Non-Invasive Methods and the Skin, Second Edition
When laser stimulation is used, it is very important to monitor skin parameters as thresholds decrease with increasing skin temperature and with decreasing skin thickness.47 In general, heat-evoked pain is modulated by skin constituents, peripheral nociceptors, local transduction, central transmission, and higher cortical regulation. The CO2 laser is the most commonly used laser both for psychophysics and to elicit brain responses, as this laser is technically simple, provides high power (e.g., 100 W), is relatively cheap, and is easy to control and use. As the infrared CO2 laser emission (10,600 nm) is absorbed in the superficial skin layer and reaches the receptors by passive heat conduction, superficial skin lesions may be used due to very high superficial temperatures.48 The advantage of this laser is that very little energy is reflected from the skin surface. The way we control this is to use large-beam diameters (e.g., 1 cm2) and stimuli with relatively long durations (e.g., 100 ms). It is difficult to transmit the infrared heat by flexible fibers, so various articulated arms have been developed that make the laser relatively flexible to use. In all laser studies, the spot of stimulation has to be shifted slightly between consecutive stimuli in order not to cause receptor sensitization/fatigue. The argon47 and copper vapor lasers49 operate after a different principle than CO2, as they emit light with wavelengths specifically targeting absorption in the chromophores (hemoglobin, melanin) located in the superficial vascular plexus and in upper part of the skin. The bluegreen light (488 to 515 nm) from the argon laser and the green-yellow light (510 to 577 nm) from the copper vapor lasers can be transmitted to the skin via a quartz fiber, which makes it flexible to use. The absorbed light is converted to heat and then conducted to the nearby nociceptors located around the superficial vascular plexus in the dermal-epidermal border. As argon lasers have relatively little power (e.g., 5 W), we normally have to use stimuli with relatively long duration (100 to 200 ms) and relatively small spot diameters (e.g., 3 mm diameter).47 Even with stimuli with durations of 200 ms, large time-locked vertex potentials can be elicited.50 The argon and copper vapor lasers can cause coagulations within the superficial plexus, but again it occurs rarely if the duration and spot diameter are properly selected. The disadvantage of these lasers is the reflectance of light from the skin surface, and hence the influence of, e.g., skin pigmentation. In some studies the skin has been blackened in order to minimize this source of variability. The flexible light fiber from, e.g., the argon laser, has the advantage that it can be attached to a step motor and the spot can be moved between consecutive stimuli and temporal summation can be studied.51 The YAG lasers (neodymium-YAG (1064 nm), thuliumYAG (2000 nm)) have also been applied, but may be slightly difficult to control, as they penetrate relatively deep in the
skin. Interestingly, the different lasers can elicit different sensations depending on their penetration depths.49 The more recent developments within the area of semiconductor lasers have made them of interest as pain stimulation for basic and clinical purposes.52–54 The advantages of these lasers are that they are small and therefore applicable for clinical use, and furthermore, flexible fibers can transmit the output. Currently the power and the available wavelengths of lasers diodes are limited, but within years we expect they will be the main laser used in pain research. We have recently used a 970-nm near-infrared semiconductor laser (20 W, GaInAs/GaAs) and were capable of using this laser for psychophysics52 and to elicit brain potentials.53,54 The laser output is transmitted via a flexible 0.42-mm optic fiber to the skin. The beams out of the fiber diverge, and in order to maintain the stimulation spot diameter constant (e.g., 0.6 mm), we have designed a system that moves the fiber tip to different stimulus locations and at the same time maintains the stimulus spot diameter distance by keeping the distance between the fiber tip and skin constant.
89.3 ASSESSMENT OF CUTANEOUS PAIN The assessment methods are based on psychophysical, electrophysiological, and imaging techniques (Table 89.2).
TABLE 89.2 Noninvasive Assessment of Experimentally Induced Cutaneous Pain Psychophysics Visual analog scales Intensity/unpleasantness Stimulus–response function McGill Pain Questionnaire (quality) Pain detection and tolerance thresholds Cross-modality matching Electrophysiological Microneurography Excitatory or inhibitory reflexes Evoked potentials Electroencephalogram Magnetoencephalogram Psychophysiological Cutaneous temperature (thermography) Cutaneous blood flow (laser Doppler flowmetry) Blood pressure Galvanic skin resistance Imaging techniques PET MRI SPECT
Assessment of Cutaneous Pain
89.3.1 PSYCHOPHYSICAL METHODS Psychophysical determinations can roughly be divided into response-dependent and stimulus-dependent methods (see reviews by Chapman et al.55 and Gracely56). The response-dependent methods rely on how the person evaluates the stimulus intensity or unpleasantness on a given scale (e.g., VAS = visual analog scale). The stimulus-dependent methods are based on adjustment of the stimulus intensity until a predefined response, typically a threshold, is reached. The stimulus intensity required to reach the threshold is in physical units, and therefore the use of scales is avoided. Stimulus–response functions are more informative than a threshold determination, as superthreshold response characteristics can be derived from the data. All quantitative sensory tests are psychophysical tests, which require awake and alert patients who fully understand the instructions given. Stimulus–response functions are valuable to assess hyperpathia to various stimulus modalities. The determination of pain thresholds (pain detection and pain tolerance) has most often been used in basic and clinical examinations. Thresholds are of great value because they are reliably, easily, and simply assessed. There are several ways of assessing pain and pain tolerance thresholds: the method of constant stimuli, the method of adjustment, the method of limits, etc. They all look for the least amount of physical energy necessary to elicit pain in the case of pain threshold, or for the most amount of physical energy still tolerable for an individual in the case of pain tolerance threshold.57 Staircase or tracking procedures are designed to assess thresholds not only once, but also repeatedly over time.58,59 A common feature of all of these procedures is the attempt of keeping the stimulation close to the threshold according to the subject’s responses in the preceding trials. This approach is especially useful to assess the time course of analgesic treatments because the analgesic effects can be monitored continuously without making the subjects aware of any change in pain perception because the subjective sensations are kept on the same level.56 Thresholds only demonstrate where the pain range starts (pain threshold) and where it ends (pain tolerance thresholds). There is no information about pain perception between these range delimiters as obtained using stimulus–response functions. The multidimensional scaling using, e.g., MPQ is not often used in experimental studies, as a standardized stimulus is less multidimensional than ongoing clinical pain. Although it seems simple to rate a given pain on a scale or as a threshold, there are many endogenous and exogenous confounding factors — training, instruction/introduction, sex of investigator vs. sex of patient/ volunteer, experimental environment (e.g., is the spouse
795
present or not), educational level of patient/volunteer, if the scores are done by experimenter/clinician/nurse or self-reported, etc. Other situational and individual factors known to influence pain are, for example, diurnal variations (time of day), menstrual cycle, gender, race, and ethnicity.
89.3.2 ELECTROPHYSIOLOGICAL METHODS 89.3.2.1 Evoked Potentials Many studies on electrophysiological assessment of cutaneous pain are based on evoked potentials, and this technique will therefore be discussed in detail, as many pitfalls exist for this method. Spreng and Ichioka60 and Schmidt61 were among the first to show that the amplitude of the human vertex potential, evoked by nociceptive electrical dental stimuli, increased at increasing stimulus intensity, and hence pain intensity. Many methodological studies have shown that potentials evoked by noxious cutaneous laser stimuli correlate with the pain magnitude rating.62–64 Despite the relation between rating and amplitude for the laser-evoked potentials, many factors can disturb this relationship.65 Central habituation65,66 and variation in the interstimulus interval67 change the amplitude of vertex potentials without concurrent changes in the pain intensity. Thus, it seems as if vertex potentials and pain intensity ratings reflect different phenomena, and therefore, both should always be recorded simultaneously. 89.3.2.2 Nociceptive Withdrawal Reflex The nociceptive withdrawal reflex to cutaneous stimulation may be used as an additional supplement to psychophysical methods to elucidate aspects of spinal nociceptive processing. The generation of a withdrawal reflex is initiated by the cutaneous nociceptive input, but an extensive processing takes place within the spinal cord. The reflex can be used for assessment in two ways: either as a reflex threshold or as an amplitude to a fixed suprathreshold stimulus intensity. The nociceptive withdrawal reflex can be elicited by cutaneous heat68,69 or electrical stimulations delivered to a cutaneous area.70,71 The pain and withdrawal reflex thresholds are found to be very similar when radiant heat43 and electrical71 stimulation are used. The reflex has been claimed to be a very robust measure with a causal relationship to pain intensity.71–73 However, a number of studies have recently observed discrepancies and found poor correlation between reflex and pain rating.69,74–76
796
Handbook of Non-Invasive Methods and the Skin, Second Edition
89.4 INDUCTION AND ASSESSMENT OF CUTANEOUS HYPERALGESIA AND ALLODYNIA The main purpose of experimental models to induce cutaneous sensitization is to generate and hence study experimentally conditions in the skin that lead to peripheral or central hyperexcitability. Clinically, such conditions are observed after tissue injury and inflammation and have been studied extensively in animal models. Both a sensitization of nociceptors in the injured tissue and an alteration of the central processing of the sensory input are involved in its pathophysiology.77 Cutaneous hyperalgesia is an increased response to a painful stimulus, and cutaneous allodynia is the induction of pain by an innocuous stimulus. After application of, e.g., capsaicin, the cutaneous inflammation and neurogenic inflammation cause local sensitization of cutaneous receptors (primary hyperalgesia). The otherwise normal skin area surrounding the primary area is termed secondary hyperalgesia and is sensitive to pinprick (static hyperalgesia or pinprick hyperalgesia) and stroking with, e.g., a brush (dynamic hyperalgesia or allodynia). Secondary hyperalgesia results from alterations in the central modulation of the afferent input and is a result of neuronal plasticity in dorsal horn neurones.78 Experimental induction of hyperalgesia has a long history involving the classic work of Lewis.4,79
89.4.1 INDUCTION
OF
CUTANEOUS HYPERALGESIA
Basically, hyperalgesia can be induced by chemical, thermal, and mechanical models. Burn injury, capsaicin, mustard oil, bradykinin, and serotonin can be used to induce hyperalgesia in humans. Primary hyperalgesia (i.e., hyperalgesia at the site of injury) can be studied by determining pain thresholds to heat and mechanical stimulation.80–82 The area of secondary hyperalgesia (i.e., hyperalgesia at tissues outside the injury) can be determined by brush (allodynia) and pinprick (hyperalgesia) stimulation of the skin surrounding the injury. Alloknesis — an itch produced by innocuous mechanical stimulation — shares common features with allodynia. Alloknesis can be obtained in human volunteers following intracutaneous or subcutaneous injections of histamine (0.1, 1, or 10 μg).83 It is not yet clear why the different methods to induce cutaneous hyperalgesia/allodynia result in different manifestations, e.g., hyperalgesia, to different stimulus modalities. 89.4.1.1 Burn Injury For burn injury, the same thermode used for inducing heat is applied. A constant probe (25 × 50 mm) of 47˚C,
applied to the skin for 7 min, does not evoke spontaneous pain after termination of the stimulus, but produces primary and secondary hyperalgesia.84,85 Increased sensitivity of A-fibers33 and C-fibers86 is responsible for primary hyperalgesia after burn injury. Cfiber neuropeptides and excitatory amino acids are involved in spinal cord changes that produce secondary hyperalgesia after heat-induced injury.87 89.4.1.2 Cold Application of noxious cold79 can induce hyperalgesia. Using the freezing trauma model, a copper cylinder of –28˚C is placed on the skin for 10 s. The hyperalgesia is prominent 20 h later.88 89.4.1.3 Capsaicin Intradermal injection or topical application of capsaicin is safe and the most commonly used model. The effect of capsaicin on neuropeptide-containing small fiber afferents and its relation to neurogen inflammation have been extensively studied by Jancsó and coworkers.89 Intradermal injection of 100 μg of capsaicin (10 μl) evokes a shortlasting burning pain at the site of injection followed by secondary hyperalgesia in the surrounding tissue.90 Pain induced by capsaicin is mediated mostly by C-fibers.91–93 Secondary hyperalgesia is the result of changes in the central processing of sensory input from myelinated Afibers that normally transmit nonpainful tactile sensations.92 Primary and secondary hyperalgesia can last up to 24 h.80 Volunteers can, however, better accept 1 h of topical application, which also results in development of hyperalgesia.81 Capsaicin 1% moisturizing cream, applied topically for 30 to 60 min, produces primary and secondary hyperalgesia.75 Repeated topical applications of capsaicin, however, totally anesthetize the application area94 and inhibit the development of capsaicin-evoked flare.95 89.4.1.4 Mustard Oil A compress soaked with mustard oil is applied to the skin for 4 min. This evokes burning pain followed by an inflammatory reaction at the site of application and secondary hyperalgesia in the surrounding tissue.96,97 The burning pain is mediated by C-fibers, while hyperalgesia to light mechanical stimuli is transmitted by Aβ-fibers, which normally encode nonpainful tactile sensations.98 89.4.1.5 Mechanical Trauma Strong, prolonged, or repeated mechanical pressure39,99 can also induce hyperalgesia, but this method is not commonly used. In the pinch model of hyperalgesia, the skin web is pinched by a force of 8 N for 2 min repeatedly.100
Assessment of Cutaneous Pain
797
89.4.1.6 UV Radiation Benrath et al.101 used UVB in doses of 133, 266, and 400 mj/cm2 at a distance of 15 cm from the skin and used irradiation periods of 15, 30, and 45 s. Reddening appeared 3 h after exposure. Skin blood flow was elevated for 96 h and returned to baseline within 216 h. There were two peaks in blood flow, one after 12 h and one after 36 h. Skin blood flow values at 24 h were significantly lower than at 12 and 36 h after irradiation (for all doses).
89.4.2 ASSESSMENT
OF
CUTANEOUS HYPERALGESIA
Generally, the repertoire of assessment techniques (Figure 89.5) to characterize hyperalgesia and allodynia has been predominantly based on psychophysical techniques. In the primary hyperalgesic area, the thermal pain thresholds to radiant heating and thermode stimulation are lowered from approximately 44˚C to 33˚C, and responses to suprathreshold stimulation (slow temperature rise times) are increased.39,79,80,102 The cutaneous temperature in the primary area is normally increased by several degrees. A problem is application of a touching thermode in the primary area, as this by itself can cause pain. It is known that light pressure in capsaicin-induced primary hyperalgesia is very likely to induce pain, which again could contaminate the measurements of the thresholds.103 Von frey Secondary hyperalgesia Brush
Thermo roll Thermode
Primary hyperalgesia Pressure Injury
Electrical
Laser
FIGURE 89.5 Noninvasive techniques to assess hyperalgesia. The figure illustrates a cutaneous injury (e.g., topical application of capsaicin) and how to assess the cutaneous sensitization noninvasively. The nociceptors under the site of capsaicin application are sensitized, and vasodilatation occurs due to local release of peptides. The area surrounding the area of injury is also sensitized to different stimulus modalities (secondary hyperalgesia). In this area, the skin and nociceptors are normal, but sensitization in the spinal cord dorsal root causes, e.g., tactile dynamic allodynia (pain caused by brushing the skin). To characterize primary and secondary hyperalgesia noninvasively, a variety of stimulus modalities are needed for the assessment (tactile, heat, mechanical, electrical). In addition, laser Doppler flowmetry, thermography, and reflectance spectroscopy can be used.
Arendt-Nielsen et al.51 found that the pricking pain threshold to brief argon laser stimuli (high temperature rise times) was not reduced significantly in the primary area, suggesting that the Aδ-nociceptors (probably type II mechano-heat46) are not sensitized. This finding is in accordance with the study by Kilo et al.,88 in which no hyperalgesia to nociceptive impact stimuli (shown to activate Aδ- and C-fibers) could be detected. The secondary area is characterized by hyperalgesia to brush and pinprick, where the pinprick area is the largest.80 The assessment techniques used are mainly based on simple techniques, such as von Frey hair and camel hair brush stimulation. The neurogenic inflammatory response after induction of cutaneous hyperalgesia by, e.g., capsaicin can be assessed by laser Doppler flowmetry, thermography, and reflectance spectroscopy.
89.5 SUMMARY The quantitative assessment of cutaneous pain has become established and can be utilized in basic dermatological and clinical investigations. Activation of cutaneous nociceptors results in pain, but at the same time, the nociceptors release neuropeptides, whereas most are vasoactive and cause neurogenic inflammation. Such fundamental knowledge can be used to explain processes in the skin leading to sensitization/desensitization of the cutaneous receptors. The clinical knowledge, which can be acquired, relates to assessment of the cutaneous sensibility in normal and diseased skin, to follow changes in sensitivity during progression of skin or other diseases, and to monitor changes during and after treatment. It has been demonstrated that the different stimulus modalities (thermal, electrical, mechanical, chemical) activate different cutaneous receptors, and hence different pathways and mechanisms. The selection of the appropriate experimental pain is guided by the scientific or clinical purpose. Furthermore, the various methods for assessing the experimentally evoked pain (e.g., pain and tolerance thresholds, rating scales) reflect different aspects of the complex phenomenon pain perception. Quantitative sensory testing is today mainly used in experimental studies, in clinical neurophysiology and neurology, but also has a role to play in basic and clinical dermatological research.
REFERENCES 1. Levine, J.D., Terrence, J.C., and Basbaum, A.I. The peripheral nervous system and the inflammatory process. In Proceedings of the Vth World Congress on Pain. Elsevier Science Publishers, Amsterdam, 1988, pp. 33–40.
798
Handbook of Non-Invasive Methods and the Skin, Second Edition
2. von Frey, M. Untersuchungen über die menschliche Haut. In Des XXIII Bandes der Abhandlungen der matematich-physischen Classe der Königl, Sächsischen Gesellschaft der Wissenschaften III. Bei S. Hirzel, Leipzig, 1896. 3. Goldscheider, A. Die spezifische Energie der Gefülsnerven der Haut. Monatschr. Prakt. Dermatol., 3, 283–303, 1988. 4. Lewis, T., Ed. Pain. Macmillan, New York, 1942. 5. Hardy, J.D., Wolff, H.G., and Goodell, H. Methods for the study of pain thresholds. In Pain Sensations and Reactions, Hardy, J.D., Wolff, H.G., and Goodell, H., Eds. Williams & Wilkins, Baltimore, 1952, pp. 52–85. 6. Arendt-Nielsen, L. Induction and assessment of experimental pain from human skin, muscle and viscera. Prog. Pain Res. Manage., 8, 393–425, 1997. 7. Handwerker, H.O. and Kobal, G. Psychophysiology of experimentally induced pain. Physiol. Rev., 73, 639–671, 1993. 8. Rollman, G.B. Cognitive effects in pain and pain judgement. In Psychophysical Approaches to Cognition, Algom, D., Ed. Elsevier, Amsterdam, 1992, pp. 515–574. 9. Notermans, S.L. Measurement of the pain threshold determined by electrical stimulation and its clinical application. I. Method and factors possibly influencing the pain threshold. Neurology, 16, 1071–1086, 1966. 10. Della Corte, M., Procacci, P., Bozza, G., and Buzzelli, G.A. Study on the cutaneous pricking pain threshold in normal man. Arch. Fisiol., 64, 141–170, 1965. 11. Brennum, J., Nielsen, P.T., Horn, A., et al. Quantitative sensory examination during epidural anaesthesia and analgesia in man: dose-response effect of bupivacaine. Pain, 56, 315–326, 1994. 12. Eide, P.K., Jørum, E., Stubhaug, A., Bremnes, J., and Breivik, H. Relief of post-herpetic neuralgia with the Nmethyl-D-aspartic acid receptor antagonist ketamine: a double-blind, cross-over comparison with morphine and placebo. Pain, 58, 347–354,1994. 13. Warncke, T., Stubhaug, A., and Jørum, E. Ketamine, an NMDA receptor antagonist, suppresses spatial and temporal properties of burn-induced secondary hyperalgesia in man: a double-blind, cross-over comparison with morphine and placebo. Pain, 72, 99–106, 1997. 14. Stubhaug, A., Breivik, H., Eide, P.K., Kreunen, M., and Foss, A. Mapping of punctate hyperalgesia around a surgical incision demonstrates that ketamine is a powerful suppressor of central sensitisation to pain following surgery. Acta Anaesthesiol. Scand., 41, 1124–1132, 1997. 15. Ziegler, E.A., Magerl, W., Meyer, R.A., and Treede, R.D. Secondary hyperalgesia to punctate mechanical stimuli. Central sensitization to A-fibre nociceptor input. Brain, 122 (Pt. 12), 2245–2257, 1999. 16. Kohllöffel, L.U.E., Koltzenburg, M., and Handwerker, H.O. A novel technique for the evaluation of mechanical pain and hyperalgesia. Pain, 46, 81–87, 1991.
17. Arendt-Nielsen, L., Yamasaki, H., Nielsen, J., Naka, D., and Kakigi, R. Magnetoencephalographic responses to painful impact stimulation. Brain Res., 839, 203–208, 1999. 18. Jensen, K., Andersen, H.Ø., Olesen, J., and Lindblom, U. Pressure pain thresholds in human temporal region. Evaluation of a new pressure algometer. Pain, 25, 313–325, 1986. 19. Atkins, J.C., Zielinski, A., Klinkhoff, A.V., et al. An electronic method for measuring joint tenderness in rheumatoid arthritis. Arthr. Rheum, 35, 407–410,1992. 20. Bendtsen, L., Jensen, R., Jensen, N.K., and Olesen, J. Muscle palpation with controlled finger pressure: new equipment for the study of tender myofascial tissue. Pain, 59, 235–239, 1994. 21. McCarty, D.J., Gatter, R.A., and Phelps, P. A dolorimeter for quantification of articular tenderness. Arthr. Rheum., 8, 551–559, 1965. 22. Fischer, A.A. Pressure algometry over normal muscles. Standard values, validity and reproducibility of pressure threshold. Pain, 30, 115–126, 1987. 23. Forgione, A.G. and Barber, T.X. A strain gauge pain stimulator. Psychophysiology, 8, 102–106, 1971. 24. Brennum, J., Kjeldsen, M., Jensen, K., and Jensen, T.S. Measurements of human pressure-pain thresholds on fingers and toes. Pain, 38, 211–217, 1989. 25. Ohrbach, R. and Gale, E.N. Pressure pain thresholds, clinical assessment, and differential diagnosis: reliability and validity in patients with myogenic pain. Pain 39, 157–169, 1989. 26. Kobal, G. Pain-related electrical potentials of the human nasal mucosa elicited by chemical stimulation. Pain, 22, 151–163, 1985. 27. Lautenbacher, S., Roscher, S., and Strian, F. Tonic pain evoked by pulsating heat: temporal summation mechanisms and perceptual qualities. Somatosens. Mot. Res., 12, 59–70, 1995. 28. Blasco, T. and Bayes, R. Unreliability of the cold pressor test method in pain studies. Methods Find. Exp. Clin. Pharmacol., 10, 767–772, 1988. 29. Poulsen, L., Arendt-Nielsen, L., Brøsen, K., Nielsen, K.K., Gram, L.F., and Sindrup, S.H. The hypoalgesic effect of imipramine in different human experimental pain models. Pain, 60, 287–293, 1995. 30. Caterina, M.J. and Julius, D. The vanilloid receptor: a molecular gateway to the pain pathway. Annu. Rev. Neurosci., 24, 487–517, 2001. 31. Tominaga, M. and Julius, D. Capsaicin receptor in the pain pathway. Jpn. J. Pharmacol., 83, 20–24, 2000. 32. Torebjork, E. Human microneurography and intraneural microstimulation in the study of neuropathic pain. Muscle Nerve, 16, 1063–1065, 1993. 33. Meyer, R.A., Campbell, J.N., and Raja, S.R. Peripheral neural mechanisms of nociception. In Textbook of Pain, Wall, P. and Melzack, R., Eds. Churchill Livingstone, Edinburgh, 1994, pp. 13–44. 34. Gybels, J., Handwerker, H.O., Van Hees, J. A comparison between the discharges of human nociceptive nerve fibres and the subject’s ratings of his sensations. J. Physiol., 292, 193–206, 1979.
Assessment of Cutaneous Pain
35. Hallin, R.G. and Wiesenfeld, Z. A standardized electrode for percutaneous recording of A and C fibre units in conscious man. Acta Physiol. Scand., 113, 561–563, 1981. 36. Schmelz, M., Schmidt, R., Ringkamp, M., Handwerker, H.O., and Torebjork, H.E. Sensitization of insensitive branches of C nociceptors in human skin. J. Physiol., 15, 389–394, 1994. 37. Elo. J. and Nikula, A. Zur Topographie des Wärmesinnes. Skan. Arch. Physiol., 24, 226–246, 1910. 38. Hardy, J.D., Wolff, H.G., and Goodell, H. Studies on pain. A new method for measuring pain threshold: observations on spatial summation of pain. J. Clin. Invest., 19, 649–657, 1940. 39. Hardy, J.D., Wolff, H.G., and Goodell, H. Experimental evidence on the nature of cutaneous hyperalgesia. J. Clin. Invest., 29, 115–140, 1950. 40. Mor, J. and Carmon, A. Laser emitted radiant heat for pain research. Pain, 1, 233–237, 1975. 41. Lindblom, U. Analysis of abnormal touch, pain, and temperature sensation in patients. In Touch, Temperature, and Pain in Health and Disease: Mechanisms and Assessments, Prog. Pain Res. and Manag., Vol. 3, Boivie, J., Hansson, P., and Lindblom, U., Eds. IASP Press, Seattle, 1994, pp. 63–84. 42. Nielsen, J. and Arendt-Nielsen, L. Spatial summation of heat induced pain within and between dermatomes. Somatosens. Mot. Res., 14, 119–125, 1997. 43. Hardy, J.D. Threshold of pain and reflex contraction as related to noxious stimuli. J. Appl. Physiol., 5, 225–239, 1953. 44. Andersen, O.K., Jensen, L.M., Brennum, J., and ArendtNielsen, L. Evidence for central summation of C and A nociceptive activity in man. Pain, 59, 273–280, 1994. 45. Meyer, R.A. and Campbell, J.N. Myelinated nociceptive afferents account for the hyperalgesia that follows a burn to the hand. Science, 213, 1527–1529, 1981. 46. Treede, R.-D., Meyer, R.A., Raja, S.N., and Campbell, J.N. Evidence for two different heat transduction mechanisms in nociceptive primary afferents innervating monkey skin. J. Physiol., 483, 747–758, 1995. 47. Arendt-Nielsen, L. and Bjerring, P. Sensory and pain threshold characteristics to laser stimuli. J. Neurol. Neurosurg. Psychiatry, 51, 35–42, 1988. 48. Bromm, B. and Treede, R.D. CO2 laser radiant pulses activate C nociceptors in man. Pflügers Arch. Physiol., 399, 155–156, 1983. 49. Svensson, P., Bjerring, P., Arendt-Nielsen, L., Nielsen, J., and Kaaber, S. Comparison of four laser types for experimental pain stimulation on oral mucosa and hairy skin. Lasers Surg. Med., 11, 313–324, 1991. 50. Arendt-Nielsen, L. Characteristics, detection and modulation of laser-evoked vertex potentials. Acta Anaesthesiol. Scand., 38 (Suppl. 101), 1–44, 1994. 51. Arendt-Nielsen, L., Andersen, O.K., and Jensen, T.S. Brief, prolonged and repeated stimuli applied to hyperalgesic skin areas: a psychophysical study. Brain Res., 712, 165–167, 1996.
799
52. Greffrath, W., Nemenov, M.I., Schwarz, S., Baumgartner, U., Vogel, H., Arendt-Nielsen, L., and Treede, R.D. Inward currents in primary nociceptive neurons of the rat and pain sensations in humans elicited by infrared diode laser pulses. Pain, 99, 145–155, 2002. 53. Chen, A.C., Niddam, D.M., and Arendt-Nielsen, L. Contact heat evoked potentials as a valid means to study nociceptive pathways in human subjects. Neurosci. Lett., 316, 79–82, 2001. 54. Gulsoy, M., Durak, K., Kurt, A., Karamursel, S., and Cilesiz, I. The 980-nm diode laser as a new stimulant for laser evoked potentials studies. Lasers Surg. Med., 28, 244–247, 2001. 55. Chapman, C.R., Casey, K.L., Dubner, R., et al. Pain measurement: an overview. Pain, 22, 1–31, 1985. 56. Gracely, R.H. Studies of pain in human subjects. In Textbook of Pain, 4th ed., Wall, P.D. and Melzack, R., Eds. Churchill Livingstone, Edinburgh, 1999, pp. 385–407. 57. Price, D.D. and Harkins, S.W. Psychophysical approaches to pain measurement and assessment. In Handbook of Pain Assessment, Turk, D.C. and Melzack, R., Eds. Guilford Press, New York, 1992, pp. 111–134. 58. Gracely, R.H., Lota, L., Walter, D.J., and Dubner, R. A multiple random staircase method of psychophysical pain assessment. Pain, 32, 55–63, 1988. 59. Lautenbacher, S., Galfe, G., Hölzl, R., and Strian, F. Threshold tracking for assessment of long-term adaptation and sensitization in pain perception. Percept. Mot. Skills, 69, 579–589, 1989. 60. Spreng, M. and Ichioka, M. Langsame Rindenpotentiale bei Schmerzreizung am Menschen. Pflügers Arch., 279, 121–132, 1964. 61. Schmidt, J. Die Beeinflussung der langsamen Hirnrindenpotentiale des Menschen nach elektrischer Zahnreizung durch Analgetika. Acta Biol. Med. Germ., 24, 361–368, 1970. 62. Carmon, A., Dotan, Y., and Sarne, Y. Correlation of subjective pain experience with cerebral evoked responses to noxious thermal stimulations. Exp. Brain Res., 33, 445–453, 1978. 63. Arendt-Nielsen, L. First pain event related potentials to argon laser stimuli: recording and quantification. J. Neurol. Neurosurg. Psychiatry, 53, 398–404, 1990. 64. Svensson, P., Arendt-Nielsen, L., Kaaber, S., and Bjerring, P. Vertex potentials evoked by painful argon laser stimulation of human oral mucosa. Relationship to stimulus intensity. Anaesth. Pain Control Dent., 2, 27–33, 1993. 65. Kobal, G. and Raab, W. The effect of analgesics on painrelated somatosensory evoked potentials. Agents Actions, 19, 75–88, 1986. 66. Miltner, W., Larbig, W., and Braun, C. Biofeedback of somatosensory event-related potentials: can individual pain sensations be modified by biofeedback-induced self-control of event-related potentials? Pain, 35, 205–213, 1988.
800
Handbook of Non-Invasive Methods and the Skin, Second Edition
67. Jacobson, R.C., Chapman, R.C., and Gerlach, R. Stimulus intensity and inter-stimulus interval effects on painrelated cerebral potentials. Electroenceph. Clin. Neurophysiol., 62, 352–363, 1985. 68. Willer, J.C., Boureau, F., and Berny, J. Nociceptive flexion reflexes elicited by noxious laser radiant heat in man. Pain, 7, 15–20, 1979. 69. Campbell, I.G., Carstens, E., and Watkins, L.R. Comparison of human pain sensation and flexion withdrawal evoked by noxious radiant heat. Pain, 45, 259–268, 1991. 70. Kugelberg, E., Eklund, K., and Grimby, L. An electromyographic study of the nociceptive reflexes of the lower limb. Mechanism of the plantar responses. Brain, 83, 394–410, 1960. 71. Willer, J.C. Comparative study of perceived pain and nociceptive flexion reflex in man. Pain, 3, 69–80, 1977. 72. Chan, C.W.Y. and Dallaire, M. Subjective pain sensation is linearly correlated with the flexion reflex in man. Brain Res., 479, 145–150, 1989. 73. DeBroucker, T., Willer, J.C., and Bergeret, S. The nociceptive flexion reflex in humans: a specific and objective correlate of experimental pain. In Issues in Pain Measurement, Chapman, C.R. and Loeser, J.D., Eds. Raven Press Ltd., New York, 1989, pp. 337–364. 74. García-Larrea, L., Charles, N., Sindou, M., and Mauguière, F. Flexion reflexes following anterolateral cordotomy in man: dissociation between pain sensation and nociceptive reflex RIII. Pain, 55, 139–149, 1993. 75. Andersen, O.K., Gracely, R.H., and Arendt-Nielsen, L. Facilitation of the human nociceptive reflex by stimulation of A beta-fibres in a secondary hyperalgesic area sustained by nociceptive input from the primary hyperalgesic area. Acta Physiol. Scand., 155, 87–97, 1995. 76. Andersen, O.K., Felsby, S., Nicolaisen, L., Bjerring, P., Jensen, T.S., and Arendt-Nielsen, L. The effect of ketamine on stimulation of primary and secondary hyperalgesic areas induced by capsaicin: a double blind, placebo-controlled, human experimental study. Pain, 66, 51–62, 1996. 77. Woolf, C.J. Evidence for a central component of postinjury pain hypersensitivity. Nature (London), 306, 686–688, 1983. 78. Yaksh, T.L. Behavioral and autonomic correlates of the tactile evoked allodynia produced by spinal glycine inhibition: effects of modulatory receptor systems and excitatory amino acid antagonists. Pain, 37, 111–123, 1989. 79. Lewis, T. Experiments relating to cutaneous hyperalgesia and its spread through somatic nerves. Clin. Sci., 2, 373–421, 1936. 80. LaMotte, R.H., Shain, C.N., Simone, D.A., and Tsai, E.F. Neurogenic hyperalgesia: psychophysical studies of underlying mechanisms. J. Neurophysiol., 66, 190–211, 1991. 81. Baumann, T.K., Simone, D.A., Shain, C.N., and LaMotte, R.H. Neurogenic hyperalgesia: the search for the primary cutaneous afferent fibers that contribute to capsaicin-induced pain and hyperalgesia. J. Neurophysiol., 66, 212–227, 1991.
82. Simone, D.A., Sorkin, L.S., Oh, U., Chung, J.M., Owens, C., LaMotte, R.H., and Willis, W.D. Neurogenic hyperalgesia: central neural correlates in responses of spinothalamic tract neurons. J. Neurophysiol., 66, 228–246, 1991. 83. Simone, D.A., Alreja, M., and LaMotte, R.H. Psychophysical studies of the itch sensation and itchy skin (“alloknesis”) produced by intracutaneous injection of histamine. Somatosens. Mot. Res., 8, 271–279, 1991. 84. Brennum, J., Dahl, J.B., Møiniche, S., and ArendtNielsen, L. Quantitative sensory examination of epidural anaesthesia and analgesia in man: effects of pre- and post-traumatic morphine on hyperalgesia. Pain, 59, 261–271, 1994. 85. Pedersen, J.L., Rung, G.W., and Kehlet, H. Effect of sympathetic nerve block on acute inflammatory pain and hyperalgesia. Anesthesiology, 86, 293–301, 1997. 86. Torebjörk, H.E., LaMotte, R.H., and Robinson, C.J. Peripheral neural correlates of magnitude of cutaneous pain and hyperalgesia: simultaneous recordings in humans of sensory judgments of pain and evoked responses in nociceptors with C-fibers. J. Neurophysiol., 51, 325–339, 1984. 87. Coderre, T.J. and Melzack, R. Central neural mediators of secondary hyperalgesia following heat injury in rats: neuropeptides and excitatory amino acids. Neurosci. Lett., 131, 71–74, 1991. 88. Kilo, S., Schmelz, M., Koltzenburg, M., and Handwerker, H.O. Different patterns of hyperalgesia induced by experimental inflammation in human skin. Brain, 117, 385–396, 1994. 89. Jancsó, N., Jancsó-Gábor, A., and Szolcaányi, J. Direct evidence for neurogenic inflammation and its prevention by denervation and by pre-treatment with capsaicin. Br. J. Pharmacol. Chemother., 31, 138–151, 1967. 90. Eisenach, J.C., Hood, D.D., and Curry, R. Intrathecal, but not intravenous, clonidine reduces experimental thermal or capsaicin-induced pain and hyperalgesia in normal volunteers. Anesth. Analg., 87, 591–596, 1998. 91. Culp, W.J., Ochoa, J., Cline, M., and Dotson, R. Heat and mechanical hyperalgesia induced by capsaicin. Cross modality threshold modulation in human C nociceptors. Brain, 112, 1317–1331, 1989. 92. Torebjörk, H.E., Lundberg, L.E., and LaMotte, R.H. Central changes in processing of mechanoreceptive input in capsaicin-induced secondary hyperalgesia in humans. J. Physiol. (London), 448, 765–780, 1992. 93. LaMotte, R.H., Lundberg, L.E., and Torebjork, H.E. Pain, hyperalgesia and activity in nociceptive C units in humans after intradermal injection of capsaicin. J. Physiol., 448, 749–764, 1992. 94. Bjerring, P. and Arendt-Nielsen, L. Use of a new argon laser technique to evaluate changes in sensory and pain thresholds in human skin following topical capsaicin treatment. Skin Pharmacol., 2, 162–167, 1989. 95. Bjerring, P. and Arendt-Nielsen, L. Inhibition of histamine skin flare reaction following repeated topical applications of capsaicin. Allergy, 45, 121–125, 1990.
Assessment of Cutaneous Pain
96. Schmelz, M., Schmidt, R., Ringkamp, M., Forster, C., Handwerker, H.O., and Torebjork, H.E. Limitation of sensitization to injured parts of receptive fields in human skin C-nociceptors. Exp. Brain Res., 109, 141–147, 1996. 97. Sjolund, K.F., Segerdahl, M., and Sollevi, A. Adenosine reduces secondary hyperalgesia in two human models of cutaneous inflammatory pain. Anesth. Analg., 88, 605–610, 1999. 98. Koltzenburg, M., Torebjörk, H.E., and Wahren, L.K. Nociceptor modulated central sensitization causes mechanical hyperalgesia in acute chemogenic and chronic neuropathic pain. Brain, 117, 579–591, 1994. 99. Magerl, W., Geldner, G., and Handwerker, H.O. Pain and vascular reflexes in man elicited by prolonged noxious mechano-stimulation. Pain, 43, 219–225, 1990.
801
100. Forster, C., Magerl, W., Beck, A., Geisslinger, G., Gall, T., Brune, K., and Handwerker, H.O. Differential effects of dipyrone, ibuprofen, and paracetamol on experimentally induced pain in man. Agents Actions, 35, 112–121, 1992. 101. Benrath, J., Gillardon, F., and Zimmermann, M. Differential time courses of skin blood flow and hyperalgesia in the human sunburn reaction following ultraviolet irradiation of the skin. Eur. J. Pain, 5, 155–167, 2001. 102. Dahl, J.B., Brennum, J., Arendt-Nielsen, L., Jensen, T.S., and Kehlet, H. The effect of pre- versus postinjury infiltration with lidocaine on thermal and mechanical hyperalgesia after heat injury to the skin. Pain, 53, 43–51, 1993. 103. Koltzenburg, M., Lundberg, L.E., and Torebjork, H.E. Dynamic and static components of mechanical hyperalgesia in human hairy skin. Pain, 51, 207–219, 1992.
Sweat Gland Distribution and Function
Techniques for the 90 Classical Localization of Sweat Glands Peter Dykes Cutest Systems Ltd., Cardiff, United Kingdom
CONTENTS 90.1 Introduction............................................................................................................................................................805 90.2 Object.....................................................................................................................................................................806 90.3 Methodological Principle ......................................................................................................................................806 90.3.1 Staining Methods .......................................................................................................................................806 90.3.2 Molding Methods ......................................................................................................................................806 90.4 Sources of Error.....................................................................................................................................................807 90.5 Correlation with Other Methods ...........................................................................................................................808 90.6 Recommendations..................................................................................................................................................808 References .......................................................................................................................................................................808
90.1 INTRODUCTION The eccrine sweat gland is one of the major cutaneous appendages, with approximately 2 to 4 million units distributed over almost the entire body surface. The density varies greatly with body site, from 60 to 100 glands per square centimeter on the trunk and limbs to 600 to 700 glands per square centimeter on the soles and palms. The principal function of the eccrine sweat gland is thermoregulation following physical exercise or exposure to a hot environment. Major illness or even death can occur following sweat gland failure. Another important function of sweat glands is to moisturize the skin of the palms and soles during physical activity, and hence improve grip. For a review of sweat gland structure and function, see Reference 1. The eccrine sweat gland is also an excretory organ, and some of the substances that it delivers to the skin surface may have important physiological functions. For example, the presence of lactate in sweat at approximately five times the plasma concentration may be significant in the stratum corneum function; topical lactate-containing formulations are used extensively for the treatment of hyperkeratosis and dry skin conditions. Similarly, the urea present in sweat may have a role in skin moisturization. There have been reports of various proteins and enzymes in sweat. These include epidermal growth factor, kallikrein, and immunoglobulin A.2,3 However, the possibility of contamination from the epidermis or stratum corneum
has not been rigorously excluded in all these cases, and further studies are needed in order to verify some of these findings. Various drugs may be excreted in sweat, such as phenytoin, sulfadiazine, phenobarbital, and ethanol. Of particular interest to dermatologists are the reports that antifungal drugs such as griseofulvin and ketoconazole may reach the skin surface rapidly via sweat.4 After a single oral dose of griseofulvin, the highest concentrations in the stratum corneum are reached within a few hours.5 This may only be explained by secretion into sweat and rapid transport to the skin surface. As would be predicted, the blocking of sweat secretion by topical antiperspirants prevents the accumulation of griseofulvin in the stratum corneum. It has also been shown that orally administered ketoconazole is delivered to the stratum corneum in a similar manner.6 Clinically, abnormal sweat production is associated with a wide spectrum of diseases, including several of dermatological interest. Impaired sweating or anhydrosis is a common feature of diabetes mellitus, and is related to the polyneuropathy seen in this disorder.7 Subclinical autonomic neuropathy can be demonstrated by a decreased sweat gland responsiveness in patients with chronic bowel dysfunction.8 A reduction in functional eccrine sweat glands is a feature of radiotherapy and may be a useful indicator of the cumulative doses received.9 In alopecia areata a decrease in number and functional activity of sweat glands has been reported, but there did not 805
806
Handbook of Non-Invasive Methods and the Skin, Second Edition
appear to be a correlation between these changes and disease activity.10 The sweating response to moderate thermal stress is reduced in patients with atopic eczema.11 Considering the importance of eccrine sweat gland function in a variety of situations, it is not surprising that there have been several attempts to measure their number and functional status. These include the following: 1. Gravimetric methods based on collection of sweat in bags or pads held in close proximity to the skin.11,12 Change in body weight has also been used as an index of gland output. 2. Direct measurement of water loss from the skin surface using devices to detect changes in relative humidity.7 3. Microcannulation of individual sweat ducts.13 4. Measurement of electrical conductivity/resistance of the skin. 5. Visualization of individual sweat droplets. This can be achieved by a variety of methods and gives information concerning the number of glands per unit skin surface area and, in some circumstances, the rate of sweat production. Visualization methods are by far the most commonly used methods of assessment of sweat gland function. The main methods used are the subject of this review.
90.2 OBJECT We will detect sweat gland activity at different body sites following physical or pharmacological stimulus or in clinical conditions where sweat gland function is considered to be abnormal.
90.3 METHODOLOGICAL PRINCIPLE The methods for visualization of sweat gland function fall into two main categories. In the first category are the staining methods, which use a variety of dyes or stains to detect the sweat itself. In the second category are the molding methods, which detect the physical presence of liquid on the skin surface. These methods can be used after induction of sweating by vigorous exercise, thermal stimulation, or injection of sudorific agents such as phenylephrine14 or methylcholine.15
90.3.1 STAINING METHODS There are several methods for sweat detection that rely on the use of dyes or stains. A method based on the dye bromophenol blue was used extensively in the early 1950s by Herrmann and coworkers.16 The method relies on filter paper impregnated with powdered bromophenol blue, which is applied to the sweating skin surface for a few
seconds. After removal, a series of blue puncta are apparent at the site of sweat droplets. Alternatively, the bromophenol blue may be suspended in a grease and applied as a thin film to the skin surface.17 The sweat droplet pattern is visible within a few moments, and this can be photographed and the number of active glands recorded. Injected dyes have been used to define sweat gland function.18 A range of stains and dyes were injected intradermally at test sites, and after pharmacological stimulation the sweat duct orifices were clearly marked by the stained sweat. Methylene blue was recommended as the dye of choice, but the method has the drawback of permanent tattoo-like markings occurring under certain circumstances. An alternative way of staining expressed sweat was described by Juhlin and Shelley.19 A 5% solution of o-phthalaldialdehyde in xylene is painted directly onto the skin surface and within 2 to 3 minutes a black reaction product is formed at the sweat duct orifice. It is thought that the interaction of o-phthalaldialdehyde with ammonia in the sweat leads to the formation of the black material. By far the most popular staining method is that based on the iodine starch reaction. This approach has the advantage of simplicity and sensitivity without the use of chemicals, which may be irritant or allergenic in nature. At its simplest, a 2% solution of iodine in ethanol is painted onto the skin surface.20 After evaporation of the solvent, sheets of ordinary paper are held against the skin surface for a few seconds. Upon removal, a dark blue imprint is apparent wherever sweating is occurring. An example of the pattern produced by sweating fingers is given in Figure 90.1. A variation on this method is the application of a starch in castor oil solution to the iodine-painted skin. Active sweat glands appear as dark blue spots on the skin surface. Sophisticated image analysis methods may be applied to photographs or camera images21 in order to estimate the number and activity of sweat glands in this situation. The simplest and most versatile iodine starch method appears to be the one-step procedure described by Sato et al.22 In this method a preparation of iodinated starch is sprayed onto the skin surface using an atomizer. Sweat drops are visualized directly on the skin surface as dark blue or purple spots. This method has the advantage of its simplicity, the fact that it can be used at any body site over any area, and that it can be repeated at the same site after removal of the iodinated starch by simply wiping the skin surface. The safety of the operator and the patient should be considered when using this method, as inhalation of iodinated starch is not desirable.
90.3.2 MOLDING METHODS The methods based on silicone rubber impression material or plastic solutions rely on the hydrophobicity of the
Classical Techniques for the Localization of Sweat Glands
807
FIGURE 90.1 Paper imprint taken from fingers previously painted with 2% iodine in ethanol.
applied material. After application to the skin surface, any sweat droplets emerging from the eccrine pores cause a withdrawal of the hydrophobic material, leading to imprints in the surface of the applied material. Historically, the earliest attempts at molding the skin surface used solutions of polyvinyl chloride (PVC) or polyvinyl formal in a suitable solvent, such as ethylene chloride.23 The method, as modified by Harris et al.,17 uses a mixture of 5 g plastic, 10 g di-n-butyl phthalate (as a plasticizer), 40 g colloidal graphite, and 100 g of ethylene chloride. The colloidal graphite provides sharp contrast between the plastic material and the imprints or holes where the sweat droplets appear. After evaporation of the solvent the film of plastic is removed by applying a transparent adhesive tape to the skin surface and gently peeling the tape and plastic imprint away from the skin. If an area of tape is used that is larger than the plastic film, the tape can be used to attach the imprint to a glass slide for projection or microscopic examination. Active sweat glands appear as holes in the imprint, and these can be counted without difficulty. The most widely used method is that based on the use of silicone rubber impression material.7,9,10,17 This has the advantage over the plastic method of materials being readily available, simpler to prepare and use, and safe to apply to any part of the body surface. The method consists of premixing the silicone rubber with the polymerizing agent and applying the mixture directly to the skin surface. Polymerization occurs within a few minutes, usually 2 to 3 min, but this can be adjusted by varying the ratio of silicone rubber to polymerizer. The imprint can then be peeled carefully from the skin surface and prepared for visualization. According to the thickness of the application, two variations of the method can be achieved. If the silicone rubber solution is applied thickly, the sweat droplets appear as spherical depressions in the replica surface. These can be viewed by light microscopy or by scanning electron microscopy. For scanning electron microscopy a
positive image must be prepared using material such as Araldite® epoxy resin. The positive image can then be sputter coated and viewed down the scanning electron microscope. If it is assumed that the sweat droplet is spherical, measurement of the diameter will give an approximate indication of the output of the sweat gland over the time taken for the silicone rubber to polymerize. If a thin application of the silicone rubber impression material is made, the sweat droplets appear as holes in a membrane. An example is given in Figure 90.2, where a replica prepared with Silaplus® silicone dental correction material (Minerva Dental Ltd., Cardiff, U.K.) is illustrated. The replica can be analyzed in a number of ways. The number of glands per unit skin surface area can be determined by direct counting using a low-power microscope. Alternatively, the replica can be mounted and projected as a magnified image onto a grid system to allow counting. Modern image analysis systems can also be used to obtain information rapidly and accurately.10 A variation of the molding technique is the use of a thin layer of petroleum jelly applied to the skin surface.15 Beading of sweat droplets occurs, and these are photographed under standardized conditions, which enhance the contrast between the sweat and petroleum jelly. Photographic slides are produced in this method, and these can be projected onto a calibrated grid system to count the number of glands per unit skin surface area.
90.4 SOURCES OF ERROR The staining methods are those that are prone to the most error. This occurs principally as a result of the merging together of dots in the images generated. This can occur either as a result of sideways movement during the imprint stage or due to excessive sweating causing a running together of the droplets on the skin surface prior to imprinting. The print of the skin surface is also a mirror image of the skin, and so in some circumstances it may
808
Handbook of Non-Invasive Methods and the Skin, Second Edition
replica techniques, with histological data. The number of active glands recorded by the silicone rubber method on a maximally sweating adult human back was reported as agreeing with histological data.17 The number of eccrine glands recorded in the digit pads of mice by the silicone rubber method is also reported to correspond to the number of glands seen on histological sections.24
90.6 RECOMMENDATIONS
(a)
(b)
FIGURE 90.2 Silicone rubber replica taken from the back of an actively sweating adult male.
be difficult to localize accurately the area of skin from which the print was taken. In a direct comparative study17 the images created with bromophenol blue and iodine starch methods were considered to be less sharp than those obtained by plastic or silicone replica techniques. In addition, in the same group of subjects fewer active ducts seemed to be recorded by the staining techniques, and this suggests a lower sensitivity. A possible source of error with the mold technique is the inclusion of air bubbles. This is usually a result of inexperience on the part of the operator when applying the mold material. With a little practice, this can be overcome, and it is the author’s experience that with silicone rubber materials air bubbles are a rare problem.
90.5 CORRELATION WITH OTHER METHODS Very little information is available concerning the correlation of sweat gland activity, as measured by staining or
In a clinical situation where information is required as to the location of areas of hyperhydrosis, the one-step iodine starch method using an atomizer is the method of choice.22 However, in most other clinical and experimental situations the method of choice is the silicone rubber method. It is the simplest procedure and also the most accurate, and gives a permanent record of sweat gland activity, which can be viewed by projection techniques and by both light and scanning electron microscopy. No reagent preparation is required; the silicone rubber material and polymerizer are conveniently stored in the refrigerator. No equipment such as an atomizer is required, and the method is clean to use, as it is nonstaining. Because silicone rubber is nontoxic, the method can be repeatedly used on the same site with no risk of an irritant or allergic reaction developing. In addition, it is safe for use on delicate areas such as the face.
REFERENCES 1. Sato, K., Kang, W.H., Saga, K., and Sato, K.T., Biology of sweat glands and their disorders. I. Normal sweat gland function, J. Am. Acad. Dermatol., 20, 537, 1989. 2. Pesonen, K., Viinikka, L., Koskimies, A., Banks, A.R., Nicolson, M., and Perheentupa, J., Size heterogeneity of epidermal growth factor in human body fluids, Life Sci., 40, 2489, 1987. 3. Okada, T., Konishi, H., Ito, M., Nagura, H., and Asai, J., Identification of immunoglobulin A in human sweat and sweat glands, J. Invest. Dermatol., 90, 648, 1988. 4. Shah, V.P., Epstein, W.L., and Riegelman, S., Role of sweat in accumulation of orally administered griseofulvin in skin, J. Clin. Invest., 53, 1673, 1974. 5. Epstein, W.L., Shah, V.P., and Riegelman, S., Griseofulvin levels in stratum corneum. Study after oral administration in man, Arch. Dermatol., 106, 344, 1972. 6. Harris, R., Jones, H.E., and Artis, W.M., Orally administered ketoconazole: route of delivery to the human stratum corneum, Antimicrob. Agents Chemother., 24, 876, 1983. 7. Kennedy, W.R., Sakuta, M., Sutherland, D., and Goetz, F.C., The sweating deficiency in diabetes mellitus: methods of quantitation and clinical correlation, Neurology, 34, 758, 1984.
Classical Techniques for the Localization of Sweat Glands
8. Altomare, D., Pilot, M.A., Scott, M., Williams, N., Rubino, M., Ilincic, L., and Waldron, D., Detection of subclinical autonomic neuropathy in constipated patients using a sweat test, Gut, 33, 1539, 1992. 9. Morris, W.J., Dische, S., and Mott, G., A pilot study of a method of estimating the number of functional sweat glands in irradiated human skin, Radiother. Oncol., 25, 49, 1992. 10. Elieff, D., Sundby, S., Kennedy, W., and Hordinsky, M., Decreased sweat gland number and function in patients with alopecia areata, Br. J. Dermatol., 125, 130, 1991. 11. Parkinnen, M.U., Kiistala, R., and Kiistala, U., Sweating response to moderate thermal stress in atopic dermatitis, Br. J. Dermatol., 126, 346, 1992. 12. Rees, J.L. and Cox, N.H., Effect of isotretinoin on eccrine gland function, Br. J. Dermatol., 119, 79, 1988. 13. Schultz, I.J., Micropuncture studies of the sweat formation in cystic fibrosis patients, J. Clin. Invest., 48, 1470, 1969. 14. Banjar, W.M.A., Bradshaw, E.M., and Szabadi, E., Seasonal variation in responsiveness of human eccrine sweat glands to phenylephrine, Br. J. Clin. Pharmacol., 27, 276, 1989. 15. Kenney, W.L. and Fowler, S.R., Methylcholine activated eccrine sweat gland density and output as a function of age, J. Appl. Physiol., 65, 1082, 1988. 16. Herrmann, F., Prose, P.F., and Sulzberger, M.B., Studies on sweating. V. Studies on quantity and distribution of thermogenic sweat delivery to the skin, J. Invest. Dermatol., 18, 71, 1952.
809
17. Harris, D.R., Polk, B.F., and Willis, I., Evaluating sweat gland activity with imprint techniques, J. Invest. Dermatol., 58, 78, 1972. 18. Hurley, H.J. and Witkowski, J., Dye clearance and eccrine sweat secretion in human skin, J. Invest. Dermatol., 36, 259, 1961. 19. Juhlin, L. and Shelley, W.B., A stain for sweat pores, Nature, 213, 408, 1967. 20. Muller, S.A. and Kierland, R.R., The use of a modified starch iodine test for investigating local sweating responses to intradermal injection of methacoline, J. Invest. Dermatol., 32, 126, 1959. 21. Sauermann, G., Hoppe, U., and Kligman, A.M., The determination of the antiperspirant activity of aluminium chlorhydrate by digital image analysis, Int. J. Cosmet. Sci., 14, 32, 1992. 22. Sato, K.T., Richardson, A., Timm, D.E., and Sato, K., One step iodine starch method for direct visualisation of sweating, Am. J. Med. Sci., 295, 528, 1988. 23. Thomson, M.L. and Sutarman, A., The identification and enumeration of active sweat glands in man from plastic impressions of the skin, Trans. R. Soc. Trop. Med. Hyg., 47, 412, 1953. 24. Kennedy, W.R., Dakuta, M., Sutherland, D., and Goetz, F., Rodent eccrine sweat glands: a case of multiple efferent innervation, Neuroscience, 11, 741, 1984.
Mapping of Sudoral 91 Micro-Sensor Activity and Skin Surface Hydration Jean-Luc Lévêque Laboratories de Recherche de L’Oreal, Aulnay Sous Bois, France
CONTENTS 91.1 Working Principles ................................................................................................................................................811 91.2 SkinChip Images....................................................................................................................................................811 91.3 Hydration of the Stratum Corneum.......................................................................................................................812 91.4 Measurement of Sweating .....................................................................................................................................813 91.5 Hydration and Sweating ........................................................................................................................................814 91.6 Conclusion .............................................................................................................................................................815 References .......................................................................................................................................................................816
For years, beside some optical methods using infrared light, attempts to quantify the hydration of the skin surface were made through the measurements of the electric and dielectric properties of the skin. During these years, the influences of the frequency, the shape, the number, and the distance between the electrodes were studied, and there is today a general agreement between specialists for considering that Skicon®, Corneometer®, and Nova® are now validated devices for assessing the hydration of the skin. Some others, derived from these latter, are commercialized, but they all suffer from the same type of uncertainties. Two in particular are encountered. The first is the depth of measurement of these devices and, because of the roughness of the skin, which alters the skin–electrode contact, results cannot fully reflect the skin hydration. The second limitation comes from the size of the electrode: too thin, it can give a figure not reflecting the average hydration of the skin, but too large, it could hide the mosaic pattern of some skin sites. Some of these difficulties or limitations can be overcome by the use a new device, SkinChip®, very recently developed, albeit still existing as prototypes in some laboratories.1 Like the devices mentioned above, it is based on the measurement of the capacitance of the skin. This chapter deals with its technical description and some examples of its use in assessing skin hydration and sweating.
91.1 WORKING PRINCIPLES SkinChip utilizes a sensor developed by the STMicroelectronics Company for characterizing the fingerprints of
people for controlling their identity and allowing them (or not) the use of some computer, car, telephone, etc. This sensor is composed of a two-dimensional array of microcapacitors, or sensor cell, located on a surface of 18 × 12.8 cm2 (sensor plate). There are 92,160 measuring units corresponding to the same number of pixels, giving a spatial resolution of 50 microns. The surface of each pixel is composed of two adjacent plates, protected by a hard, thin coating. When skin comes into contact with the sensor plate, the effective feedback capacitance between the electrodes changes. The conditioning electronic, allowing the adjustment of gain and offset before conversion in the analogic form, is integrated on the same board. The device is directly plugged to the USB port of any computer. The optimal settings of the sensor, to yield good images on most of the skin sites, are proposed as default values. When a given skin site is applied onto the sensor plate, the corresponding image is acquired in real time and displayed on the screen of the computer. It corresponds to the mapping of the skin surface capacitance with a twodimensional resolution of 50 microns.
91.2 SKINCHIP IMAGES Some images of different skin sites, obtained by SkinChip, are illustrated in Figure 91.1. The different topographic characteristics of the skin surface are clearly displayed: microrelief lines, hair, pores, and wrinkles. Each pixel of these images is coded under 256 gray levels ranging from very high (black) to very low (white) capacitance. Because of the skin microrelief conditions, the quality of the 811
812
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 91.1 SkinChip images obtained on different sites of the skin. Hair, wrinkles, microrelief lines, and pores are easily visualized. (a) Dorsal hand. (b) Belly.
contact between the skin and sensor plate, and the measured capacitance of the skin surface also vary according to the distance between the skin surface and the plate, enabling a quite detailed image of the skin microrelief wrinkles and pores. Due to the absence of precise information about the relationship between capacitance and the distance from the plate, it remains highly speculative to transform the data in a three-dimensional image. The versatility of the device allows the obtaining of two-dimensional images in a very easy way. These are obtained in a few tenths of a second and can be immediately digitally stored. They illustrate the large variability of the patterns of the human skin.2 Histograms of the gray levels of images can be obtained routinely through usual image analysis software (Photoshop®, Paintshop®, etc.). As an example, in Figure 91.2 the histogram of an image obtained from the ventral side of the forearm is plotted. The maximum peak corresponds to gray levels ranging between 180 and 256. They represent the pixels of the microrelief lines of the image made clear since not in contact with the sensor plate. Inversely, the lower part of the histogram represents the true capacitance of the part of the skin that is in close contact with the device. As explained below, skin capacitance, i.e., its hydration, can be extracted from this histogram.
91.3 HYDRATION OF THE STRATUM CORNEUM As said above, the working principle of the device is the two-dimensional measurement of the skin surface capacitance with a resolution of 50 μm. The corresponding equivalence between some of the parameters characterizing the gray-level histogram of a skin image and the capacitance of this skin site, obtained through a Corneometer on the same skin site, was likely, but needed to be confirmed. Accordingly, we measured the capacitance of the volar forearm skin of 80 women by means of a Corneometer and recorded the SkinChip images of this same site. Attempts were made to correlate the parameters characterizing the histogram given by the Paintshop software (mean gray level (MGL) and number of pixels of the maximum of the histogram (NPMH)) with the Corneometer values, respectively. The plots of these correlations are shown in Figure 91.3a and b. Statistical calculation indicates that both histogram parameters, MGL and NPMH, are statistically correlated with the Corneometer values (R = –0.7 and R = –0.81, respectively, with p < 0.001 in both cases), making it evident that the correlation is of better value for the NPMH parameter than for the MGL. The dynamic
Micro-Sensor Mapping of Sudoral Activity and Skin Surface Hydration
813
FIGURE 91.1 (c) Dry leg. (d) Small wrinkle and pores on the forehead.
of the measurement is also higher when using the first parameter. Because of the nature of coding, black and white for respectively high and low capacitance, correlations are logically negative. The strong correlation obtained with the NPMH parameter, corresponding to the number of pixels at the maximum of the histogram, is difficult to understand. It corresponds to the pixels illustrating the primary lines of the microrelief. As a consequence, the hydrating skin surface has to shift toward the low pixel values, the left part of the histogram, and to decrease the NPMH. Further experiments should precise this question, but SkinChip images clearly show that skin microrelief alters measurements carried out on the skin surface using a flat and solid electrode, as is the case with the classical impedance or conductance devices. It is likely that some other parameters, extracted from the histogram, could be proposed in the future for representing more precisely the true skin surface capacitance by only selecting pixels corresponding to the skin part in contact with the plate. As said above, SkinChip images represent a hydration map of the skin surface. Such a map could in fact be characterized by both MGL and NPMH, which represent the mean hydration of the site, and some other parameters representing the homogeneity of the skin hydration. Indeed, it is worth noting that on some skin sites, hydration is not homogenous, as shown in Figure 91.1c, for example. Studies
aiming at defining such a parameter are currently in progress. This would supply a new characteristic of the skin surface hydration that is not measured today by the classical devices.
91.4 MEASUREMENT OF SWEATING The images afforded by SkinChip make sweat appear as black patches. It corresponds to liquid water droplets of very high conductance and capacitance (Figure 91.4a). The size of these patches increases with both the intensity and the duration of the sweating period. In a first approach, sweating can be quantified via the same parameters previously used for measuring skin hydration. In the first stages of sweating, the histogram of the images is different because the pixels corresponding to the black patches are not included in the histogram of the pixels giving the nonsweating skin area (Figure 91.4b). In such a case, the quantification of sweating can be carried out in a more specific manner by thresholding. If the subjects are maintained in a warm atmosphere for a while, droplets will merge and spread onto the whole skin surface. The histogram becomes monomodal, and the mean gray level (MGL) is likely the simplest way for quantifying sweating. In Figure 91.5 are plotted the variations of the mean MGL for six volunteers before and after staying in a warm room for some time (see legend for the details).
814
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 91.1 (e) Breast. (f) Neck/lateral.
FIGURE 91.2 Image obtained on the volar forearm and its corresponding gray-level histogram.
91.5 HYDRATION AND SWEATING One advantage of SkinChip is that it allows “seeing” the hydration of the skin through the gray levels of the images. As mentioned above, this brings a new look to an old measurement: skin capacitance, as currently measured through rigid electrodes that are not in full contact with the skin surface. This will probably stimulate scientists to
propose other ways for measuring, more specifically, this important parameter. Another advantage concerns the presence of sweating: just by looking at the images, the experimenter can check the presence or absence of sweating on the given sites. Rare today are the publications dealing with the measurement of skin hydration that address this issue. Our own experience with this new technology convinced us that great care must be taken in the selection of people and in the control of the environmental conditions for avoiding the presence of sweat on the measured zones. Three images relative to a stripping experiment are displayed in Figure 91.6. As presented in the figure, sweat appears on the stripped site after 10 strippings and markedly decreases MGL, as it would have increased capacitance measurements. This phenomenon does not appear systematically, but it could have altered numerous results of studies about transepidermal water loss (TEWL) and capacitance measurements vs. the number of strippings. The same remark can be made about the presence or absence of hair, but this is probably less important because in most of the cases the presence of hair is easy to see. However, the presence of lanugo, on the face of women, for example, is very common and hardly visible. Nobody today studies their influence on skin capacitance.
MGL
Micro-Sensor Mapping of Sudoral Activity and Skin Surface Hydration
210 200 190 180 170 160 150 140 130 120
815
y = −0.8982x + 207 R2 = 0.48 N = 80
10
20
30 40 Corneometer
50
60
FIGURE 91.3 (a) Correlation between mean gray level and Corneometer values measured on the volar forearm (see Lévêque and Corcuff2 for details). FIGURE 91.4 Image of sweating (black patches) appearing on the volar forearm and its corresponding histogram.
3500 y = −48.27x + 3350 R2 = 0.6645 N = 80
3000
Npmh
2500 2000 1500 1000 500 0
10
15
20
25
40 30 35 Corneometer
45
50
55
FIGURE 91.3 (b) Correlation between “Max Histo” and Corneometer values measured on the forearm.
91.6 CONCLUSION
Mean gray level
There are now several devices proposed to quantify the skin water content. These devices are useful for studying
160.0 140.0 120.0 100.0 80.0 60.0 40.0 20.0 0.0
1
2
the influence of treatments on the skin or for characterizing some skin diseases, the influence of aging and the environment, etc. These devices suffer various drawbacks, which are known and have been discussed for years, but are still not solved. SkinChip, as mentioned above, could help to solve one of them: the question of the contact electrode. Another question for this type of measurement is the doubtful presence of sweat. This point is solved by using this device. Another interest comes from the fact that it gives images, capacitance maps, allowing us to discuss both mean and homogeneity of the skin surface hydration. Finally, it supplies a very simple and routine way for looking at and quantifying skin surface microrelief on most of the sites. For all these reasons, it is highly probable that SkinChip will have a great future in the research laboratories for either cosmetic or dermatological purposes.
3 Experiments
4
5
FIGURE 91.5 Mean gray level of images of the back before and after different times passed in a warm room (mean and standard deviation on six persons). Experiments1 corresponds to control before sweating, 2 to measurement after 20 minutes in a warm room, and 3 to measurement after 40 minutes. Experiments 3 and 4 correspond to the same conditions (20 and 40 minutes), the day after. (Courtesy of Dr. D. Moyal, L’Oréal Recherche, Clichy, France.)
816
Handbook of Non-Invasive Methods and the Skin, Second Edition
0 Stripping MGL = 175
10 Strippings MGL = 133
20 Strippings MGL = 108
FIGURE 91.6 Aspect of the volar forearm before and after 10 and 20 strippings. Black dots that appear after the strippings correspond to sweat excretion.
REFERENCES 1. JL Lévêque and B Querleux, SkinChip®, a new tool for investigating the skin surface in vivo, Skin Res Technol 9: 343–347, 2003.
2. JL Lévêque and P Corcuff, The surface of the skin, the microrelief, in Noninvasive Methods for the Quantification of Skin Functions, PJ Frosch and AM Kligman, Eds., Springer-Verlag, Berlin, 1993, pp. 3–24.
for the Collection of Eccrine 92 Methods Sweat Julian H. Barth Department of Chemical Pathology, General Infirmary at Leeds, Leeds, United Kingdom
CONTENTS 92.1 Introduction............................................................................................................................................................817 92.2 Methods..................................................................................................................................................................817 92.2.1 Principle .....................................................................................................................................................817 92.2.2 Stimulation of Sweat Production ..............................................................................................................817 92.2.3 Harvesting Sweat .......................................................................................................................................818 92.2.3.1 Filter Paper Technique ...............................................................................................................818 92.2.3.2 Macroduct™ Sweat Collection System .....................................................................................818 92.2.4 Analysis of Sweat......................................................................................................................................818 92.2.5 Problems with Iontophoresis .....................................................................................................................818 92.2.6 Interpretation of the Analysis of Sweat ....................................................................................................819 References .......................................................................................................................................................................819
92.1 INTRODUCTION
92.2 METHODS
Measurements of eccrine sweating have been used to study the physiology of the eccrine gland itself and of the whole-body response to environmental changes, e.g., heat acclimatization. Second, the function of the eccrine gland has been used as a diagnostic test for cystic fibrosis (CF). The concentration of sweat chloride is considered the most reliable single test in the diagnosis of CF. Although gene probes are available for the common forms of CF, there is sufficient genetic heterogeneity for the sweat test to remain the standard diagnostic tool. The study of the physiology of the eccrine gland and its adaptation to changes in climate depend on the ability to measure the volume of fluid lost. Therefore, a spectrum of methods has evolved from sophisticated measurements of total body weight or whole-body chambers to small units adhered to the skin surface with humidity-sensitive flow cells.1,2 Moreover, the total losses of elements in sweat have been studied by complex washing procedures before and after sweat stimulation in saunas.3 However, for studies of the metabolism of the eccrine gland, where sweat composition rather than volume is appropriate, stimulation by iontophoresis of cholinergic agonists is the most convenient method.
92.2.1 PRINCIPLE The sudomotor processes of eccrine sweat glands are predominantly mediated by postganglionic sympathetic cholinergic neurones. The process of iontophoresis is the migration of small ions under the influence of an electrical current. Cholinergic agonists, e.g., pilocarpine, are transferred from the surface to the eccrine glands in the dermis.
92.2.2 STIMULATION
OF
SWEAT PRODUCTION
Wash and dry the skin of the flexor surface of the forearm and place two 5 cm2 gauzes over the washed area. Moisten gauzes well with a freshly prepared solution of 1% pilocarpine nitrate. Place two gauzes saturated with saline on the extensor surface of the arm. The positive electrode of a direct current (DC) power supply designed for iontophoresis is placed over the gauze with pilocarpine, and the negative electrode is placed on the saline-soaked gauze. Ensure good contact and secure the electrodes with rubber strips or similar means. If the arm is too small to secure the electrodes, as in small children, use the thigh or the interscapular area of the back of the patient. Apply a current of 0.16 mA/cm2 for 5 min.4 The current will tend to increase during this time interval and 817
818
Handbook of Non-Invasive Methods and the Skin, Second Edition
should be maintained at the above setting. After 5 min remove the electrodes, clean the skin with distilled water, and dry the area.
92.2.3 HARVESTING SWEAT 92.2.3.1 Filter Paper Technique With the use of forceps, place two prewashed Whatman No: 42 filter papers (diameter, 5.5 cm) into a weighing bottle, stopper, and determine the combined weight accurately, using an analytical balance. Handle the weighing bottle with tissue or gauze to avoid direct contact with the fingers in order to avoid transfer of sweat from the operator’s fingers, which will increase its weight. With the aid of forceps, place the preweighed filter paper over the skin area that was exposed to pilocarpine. Place a plastic sheet over this area and seal airtight with surgical tape. Allow the sweat to accumulate on the gauze or filter paper. This usually takes approximately 20 to 30 min, but the time of sweat collection may be extended as long as necessary. In general, the appearance of droplets on the plastic sheet indicates that enough sweat has accumulated. These droplets must be included in the collection. Remove the filter paper with forceps, place it immediately into the weighing bottle, and stopper. Handle the bottle with tissues as above. Weigh the bottle accurately (within 1 mg) to determine the weight of the filter paper and calculate the amount of sweat by difference. A minimum amount of 100 mg sweat is required for reliable quantitation. Generally, 200 to 500 μl of sweat is obtained (1 g sweat is assumed to be 1 ml).
FIGURE 92.1 Wescor Macroduct™: illustration shows the spiral capillary tube used for the collection of pilocarpine-stimulated sweat. The sweat pools in the concave surface (not shown) which is adjacent to the skin and is drained into the tubing by capillary action.
92.2.3.2 Macroduct™ Sweat Collection System The Macroduct™ (Wescor, Inc.) is the cornerstone of this system (see Figure 92.1 and Figure 92.2). It is a small sweat-collecting unit with a shallow concave surface that is held against the previously stimulated skin. At the apex, a small aperture leads to a spirally configured plastic capillary tube that withdraws the sweat as it pools in the concavity. The duration required for sweat collection is similar to that for the filter paper method. The absolute volume obtained is dependent only on the method required for the sweat analysis.
92.2.4 ANALYSIS
OF
SWEAT
The technique for elution of the sweat from the filter papers will depend on the particular analysis required. If the Macroduct is used, the sweat can be analyzed directly.
92.2.5 PROBLEMS
WITH IONTOPHORESIS
If the patient complains about discomfort, the test should be discontinued since the discomfort does not ease with
FIGURE 92.2 Wescor Macroduct™: illustration shows two infants. The infant on the right has the Wescor iontophoretic system in operation and the infant on the left has the collection system in operation.
time. A tickling sensation at the site of the electrode is a common finding and should be disregarded. After the test, there is usually some redness, which disappears within a few hours. There may also be small grayish papules, or miliaria; these may take 2 or 3 days to fade. Insufficient sweat may be collected using the filter paper technique above. This problem is not usually encountered by experienced operators. It may, however, be obviated by the Macroduct system, which allows accurate analysis on very small quantities of sweat.
Methods for the Collection of Eccrine Sweat
819
TABLE 92.1 Conditions Likely to be Associated with Elevated Sweat Electrolyte Concentrations Systemic Disorders
Cutaneous Disorders
Cystic fibrosis Anhydrotic ectodermal dysplasia Anorexia nervosa6 Untreated adrenal insufficiency7,8 Hereditary nephrogenic diabetes insipidus9 Glucose-6-phosphatase deficiency10 Hypothyroidism11 Mucopolysaccharidoses12 Malnutrition13 Pupillatonia, hyporeflexia, and segmental hypohydrosis with autonomic dysfunction14 Fucosidosis15 Familial cholestasis16 Hypogammaglobnulinemia17,18
92.2.6 INTERPRETATION
OF THE
ANALYSIS
OF
SWEAT
Sodium and chloride elevations in sweat are seen even in the absence of gastrointestinal or respiratory symptoms, or when pancreatic insufficiency cannot be demonstrated by other tests. In most affected infants, the test becomes positive between 3 and 5 weeks of age. Only 1 to 2% of CF patients has sweat chloride values below the value of 60 mmol/l, and only 1 in 1000 has a value of <50 mmol/l.5 The exceptions are observed predominantly in CF patients who do not have pancreatic involvement. Falsely high and low concentrations of sweat electrolytes may occur in a number of metabolic systemic and cutaneous disorders, as outlined in Table 92.1 to Table 92.3. Further variability may be due to physiological factors such as sweat flow rate, salt intake, and heat acclimatization.
TABLE 92.2 Conditions Likely to be Associated with Reduced Sweat Electrolyte Concentrations Systemic Disorders Cystinosis19 Hyperaldosteronism (primary or secondary, e.g., dehydration or severe stress) Hyperthyroidism20
Cutaneous Disorders
TABLE 92.3 Therapies with the Potential to Interfere with Sweat Electrolyte Concentrations Increased Sodium and/or Chloride All sodium salts, e.g., intravenous antibiotics Anticholinergic drugs, e.g., premedications
Decreased Sodium and/or Chloride Corticosteroids Thiazide diuretics9
REFERENCES 1. Kuno, Y., Human Perspiration, Charles C. Thomas, Springfield, IL, 1956. 2. Graichen, H., Rascati, R., and Gonzales, R.R., Automatic dew point temperature sensor, Am. J. Physiol., 52, 1658, 1982. 3. Brune, M., Magnusson, B., Persson, H., and Hallberg, L., Iron losses in sweat, Am. J. Clin. Nutr., 43, 438, 1986. 4. Gibson, L.E., di Sant’Agnes, P.A., and Schwachmann, H., Procedure for the Quantitative Iontophoretic Sweat Test for Cystic Fibrosis, Cystic Fibrosis Foundation, 6000 Executive Blvd., Suite 510, Rockville, MD 20852, 1985. 5. Davis, P.B., Hubbard, V.S., and di Sant’Agnese, P.A., Low sweat electrolytes in a patient with cystic fibrosis, Am. J. Med., 69, 643, 1980. 6. Beck, R., Goldberg, E., Durie, P.R., and Levison, H., Elevated sweat chloride levels in anorexia nervosa, J. Pediatr., 108, 260, 1986. 7. Conn, J.W., Electrolyte composition of sweat: clinical implication as an index of adrenal function, Arch. Intern. Med., 83, 416, 1949. 8. Morse, W.I., Cochrane, W.A., and Landrigan, P.L., Familial hypoparathyroidism with pernicious anaemia, steatorrhoea and adrenal insufficiency: a variant of mucoviscidosis, N. Engl. J. Med., 264, 1021, 1961. 9. Lobeck, C.C., Banta, R.A., and Mangos, J.A., Study of sweat in pitressin-resistant diabetes insipidus, J. Pediatr., 62, 868, 1963. 10. Harris, R.C. and Cohen, H.I., Sweat electrolytes in glycogen storage disease type 1, Pediatrics, 31, 1044, 1963. 11. Strickland, A.L., Sweat electrolytes in thyroid disorders, J. Pediatr., 82, 284, 1973. 12. Durand, P., Bossone, C., Della Cella, G., and Liotta, A., Le mucopolisaccaridosi, Recent Prog. Med., 44, 279, 1968. 13. Mace, J.W. and Scharberger, J.E., Elevated sweat chloride in a child with malnutrition, Clin. Pediatr., 10, 285, 1971. 14. Esterley, N.B., Cantolino, S.J., Alter, B.P., and Brusilow, S.W., Pupillatonia, hyporeflexia and segmental hypohydrosis: autonomic dysfunction in a child, J. Pediatr., 73, 852, 1968.
820
Handbook of Non-Invasive Methods and the Skin, Second Edition
15. Durand, P., Borrone, C., and Della Cella, G., Fucosidosis, J. Pediatr., 75, 665, 1975. 16. Lloyd-Still, J.D., Familial cholestasis with elevated sweat electrolyte concentrations, J. Pediatr., 99, 580, 1981. 17. Corkey, C.W.B. and Gelfand, E.W., Hypogammglobulinaemia and antibody deficiency in patients with elevated sweat chloride concentrations, J. Pediatr., 100, 420, 1982.
18. Rosario, N.A., Neto, R.S., Nitta, A., and Marinoni, L.P., Hypogammglobulinaemia and elevated sweat chloride values, J. Pediatr., 102, 163, 1983. 19. Gahl, W.A., Hubbard, V.S., and Orloff, S., Decreased sweat production in cystinosis, J. Pediatr., 104, 904, 1984. 20. Gibinski, K., Powierza-Kaczynska, C., Zmudzinski, J., Gec, L., and Dosiak, J., Thyroid control of sweat gland function, Metabolism, 21, 843, 1972.
for the Collection of 93 Methods Apocrine Sweat Julian H. Barth Department of Chemical Pathology, General Infirmary at Leeds, Leeds, United Kingdom
CONTENTS 93.1 Introduction............................................................................................................................................................821 93.2 Methods..................................................................................................................................................................821 93.2.1 Visualization of Physiological Apocrine Secretions .................................................................................821 93.2.2 Harvesting Apocrine Secretions ................................................................................................................821 93.2.3 Collection of Volatile Axillary Secretions.................................................................................................822 References .......................................................................................................................................................................822
93.1 INTRODUCTION The secretions of the apocrine gland enter the hair follicle above the point of entry of the sebaceous duct. Several methods have been developed that employ either absorbent material to collect the secretions or their volatile products or direct cannulation of the apocrine duct to collect the secretions; the choice of technique will depend on the question to be solved. However, since the products of the apocrine and sebaceous glands intermingle in the infundibulum of the pilary duct, it is practically impossible to obtain pure apocrine sweat from an in vivo technique. All of the following methods potentially suffer from contamination by either eccrine or sebaceous secretions. At present, it is probable that the only technique suitable for the study of pure apocrine sweat is the use of the isolated gland model.1
93.2 METHODS 93.2.1 VISUALIZATION OF PHYSIOLOGICAL APOCRINE SECRETIONS The study of the physiological secretion of apocrine glands is difficult due to both the intermittent nature of secretion and also the effect of eccrine secretions. These latter secretions effectively dilute the apocrine secretions, spread them over the skin surface, and therefore enhance their evaporation. This use of plaster-of-Paris discs held on the axillary skin by adherent polyethylene holders has been suggested as a method of overcoming this problem.2 Eccrine fluid is watery and is adsorbed by the discs,
whereas the viscid lipid apocrine secretions collect on the disc surface and can be observed after collections are made over a period of at least 3 to 4 hours. The apocrine droplets can be seen by fluorescence under ultraviolet light.
93.2.2 HARVESTING APOCRINE SECRETIONS The pool of apocrine sweat within the collecting duct can be squeezed onto the skin surface by stimulating the smooth muscle cells with adrenergic agents. Shelley and Hurley3 were the first to use this approach, and they collected the apocrine sweat by cannulation of the apo-pilosebaceous duct with a fine capillary tube.3 The axillary skin should be shaved the day before the planned collection, as surface manipulation is sufficient to stimulate muscular contraction of the apocrine apparatus; this results in an emptying of the reservoir of secretions. The skin is cleansed prior to the procedure with a nonionic detergent, e.g., 0.1% Triton X-100, rinsed with water, blotted dry, and finally washed with hexane. Secretion is stimulated with an intradermal injection of adrenaline 1:2000 in physiological saline. Droplets of apocrine secretion should appear almost immediately. They are visible with the naked eye, but are more clearly seen with a magnifying lens. The droplets have been harvested by canulation of the ducts with individually drawn out capillaries, but sufficient fluid can be collected from the surface with a commercially available capillary (10-μl volume). Approximately 1 μl of milky fluid can be obtained from each apocrine-related orifice; however, not all the apo-pilosebaceous ducts within the visibly adrenaline-blanched area release any fluid. It 821
822
Handbook of Non-Invasive Methods and the Skin, Second Edition
should be possible to collect a total volume of 3- to 5-μL ml secretions.
93.2.3 COLLECTION SECRETIONS
OF
VOLATILE AXILLARY
The study of axillary odor and the search for human pheromones demands a different technique from those outlined above. Cotton gauze squares (10 cm2) which that have been thoroughly cleaned and autoclaved are held in place in the axilla for periods of 6 to 9 hours.4 The gauze pads need to be cleaned with the same procedure as they will be subsequently employed to remove the absorbed material. Individual experimental design will demand the need to standardize on such factors as axillary shaving, use of deodorants and perfumes, and frequency of washing and types of soaps employed.
REFERENCES 1. Barth, J.H., Ridden, J., Philpott, M.P., Greenall, M.J., and Kealey, T., Lipogenesis by isolated human apocrine sweat glands: testosterone has no effect during long term organ maintenance, J. Invest. Dermatol., 92, 333, 1989. 2. Fox, R.H., Mullan, B.J., and Thornton, C., A technique for studying apocrine gland function in man, J. Physiol., 239, 75P, 1974. 3. Shelley, W.B. and Hurley, H.J., Jr., The physiology of the human axillary apocrine sweat gland, J. Invest. Dermatol., 20, 285, 1953. 4. Preti, G., Cutler, W.B., Christensen, C.M., Lawley, H., Huggins, G.R., and Garcia, C.-R., Human axillary extracts: analysis of compounds from samples with influence menstrual timing. J. Chem. Ecol., 13, 717, 1987.
Sebaceous Glands and Sebum Excretion
94 The Follicular Biopsy Otto H. Mills, Jr. Hill Top Research, Inc., University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, East Brunswick, New Jersey
CONTENTS 94.1 Introduction ..........................................................................................................................................................825 94.2 Method for Performing the Follicular Biopsy.....................................................................................................825 94.3 Aging Skin ...........................................................................................................................................................826 94.4 Sampling for Demodex Folliculorum ..................................................................................................................826 94.5 In Vitro Assay for Comedolytic Potential............................................................................................................826 94.6 Evaluating Prophylactic Potential of Antiacne Modalities..................................................................................827 94.7 Evaluating Microbial Densities............................................................................................................................827 94.8 Lipid Analysis ......................................................................................................................................................827 94.9 Biochemical Changes...........................................................................................................................................827 94.10 Analyzing Drug Levels ........................................................................................................................................827 94.11 Ultraviolet Examination .......................................................................................................................................827 94.12 Image Analysis .....................................................................................................................................................828 94.13 Comedogenic and Comedolytic Evaluations.......................................................................................................828 94.14 Conclusion............................................................................................................................................................828 References .......................................................................................................................................................................828
94.1 INTRODUCTION The follicular biopsy1 is an extension of a technique reported in the British literature by Marks and Dawber.2 These investigators used a cyanoacrylate polymer to obtain a thin sheet of stratum corneum cells in order to study the skin’s topography under the stereomicroscope. This skin surface biopsy can be stained, and light microscopy reveals bacteria, fungi, and other surface workings. In the course of their report, they noted that “hair follicle and sweat gland openings are seen particularly well.” Also, Holmes and colleagues3 have used cyanoacrylate polymer on the back of acne patients. They noted that they could extract vellus hairs coated with a material that was composed mainly of lipid and keratin. The follicle is central to a number of abnormalities, including acne vulgaris, keratosis pilaris, ichthyotic conditions, and fungal infections. As hyperkeratinization of the sebaceous follicle is key in the pathology of acne, a noninvasive way to sample this impaction, particularly on the face, opens up many study possibilities. Two key features of the follicular biopsy is its noninvasiveness and speed of sampling. Following is a review of some of the past work done using the follicular biopsy, a synopsis of some of the
current research under way with this technique, as well as suggestions and projections for future applications of this noninvasive technique.
94.2 METHOD FOR PERFORMING THE FOLLICULAR BIOPSY The cyanoacrylate adhesive product we currently use in our laboratories is Loctite® (Prism™ 460 Series, Loctite Canada, Inc). One droplet (approximately 8 mg/cm2) is placed on the surface of the skin. This is spread uniformly by pressing a glass microscope slide to the skin’s surface with light firm pressure. Sixty seconds is allowed for polymerization. This can be extended if the temperature of the subject’s skin or the environment is elevated. The slide with its confluent sheet is then removed slowly, peeling away with attention to preserving the sample with intact contents of sebaceous follicles. During this time we have not seen any instances of contact allergy or other concerns. Avoiding the eye area is, of course, important. The follicular biopsy sample can be viewed immediately under a dissecting scope using light projected at an angle. If present, sebaceous follicle impactions can be clearly seen and an assessment of size and density performed. 825
826
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 94.1 Transverse section of sebaceous follicle.
film, which was able to conform to the nose and cheek areas better than the rigid surface of the glass slide. Ten volunteer patients with dense to very dense involvement of trichostasis follicularis were selected. In general, the lesion distribution centered on the nose extending to the adjacent areas of the cheek. Seven of the 10 subjects were females. No other medical problem was evident at the time of examination. The cyanoacrylate polymer was placed on the area to be treated. Care was taken to protect the eyes. The adhesive was spread to coat the area with a visible film (approximately 8 mg/cm2). This layer of adhesive was allowed to dry (polymerized in 60 seconds). A piece of polyethylene was placed in contact with the adhesive and held in place for 30 seconds. An additional 30 seconds was allowed for polymerization. The polyethylene film was removed from the skin as one would remove tape — peeling away from the skin. Four patients required a second procedure (2 to 4 weeks later) in order to extract a few remaining lesions. By clinical inspection, 8 of the 10 patients so treated remained free of clinically evident trichostasis spinulosa for 1 year following the last procedure. Five of these subjects were followed for 1.5 years post-follicular biopsy extraction and remained 80% free of clinically observable follicular impactions.
94.4 SAMPLING FOR DEMODEX FOLLICULORUM FIGURE 94.2 Transverse section of sebaceous follicle following a follicular biopsy.
Histological evaluation of a punch biopsy done on follicles immediately following the follicular biopsy procedure could not detect horny material in the follicular canal. Also present was a slight perifollicular lymphocytic infiltrate with minor irritation to the follicular epithelium4 (Figure 94.1 and Figure 94.2).
The mite demodex folliculorum has been cited as playing a role in a number of dermatological conditions, including rosacea and a variety of folliculitises.5 Historically, collection of the mite from human skin has involved a number of scraping techniques. The follicular biopsy presents an effective way to retrieve this organism from the follicular canal. In order to visualize the mites, the follicular biopsy sample should be placed under the microscope and a drop of peanut oil added. By using a small probe it is possible to disrupt the keratinous impactions and often see mites in addition to those already visible.1,6
94.3 AGING SKIN As the skin ages, sebaceous follicles appear to develop several abnormalities. This is particularly the case with facial skin and the added effects of solar exposure. One such abnormality that appears to increase with age is trichostasis follicularis. Although this condition is seen in all decades of life, it appears more prominently on the nose, cheeks, and forehead as age increases. The follicular biopsy extracts these bundles of hair and keratin easily. In 1982, Mills, Huber, McGinley, and Kligman (unpublished data) further adapted the follicular biopsy by placing the adhesive on the skin and applying sheets of polyethylene
94.5 IN VITRO ASSAY FOR COMEDOLYTIC POTENTIAL Another possible application for the follicular biopsy is harvesting microcomedones for study in an in vitro comedolytic assay. As there are two topically applied agents currently used in the treatment of acne that have comedolytic activity (all-trans-retinoic acid and salicylic acid), it may be possible to apply these as benchmarks and create a screening assay for potential comedolytic agents. With the use of microscopy and histological stains, it may be possible to detect cellular changes that would suggest
The Follicular Biopsy
potential comedolytic activity. This in vitro approach should be explored.
94.6 EVALUATING PROPHYLACTIC POTENTIAL OF ANTIACNE MODALITIES As the follicular biopsy evacuates the follicular canal, certain numbers of the impactions reform to some degree in approximately 1 to 3 months. Thus, it may be possible to begin to evaluate the potential of agents to reduce or eliminate the reformation of these impactions. In order to do this kind of investigation, it is key to well define the sample area and ensure that subsequent sampling occurs in the exact area. This could be done on the face or back. The former, of course, has many more follicles, while the later offers the opportunity to compare formulations.
94.7 EVALUATING MICROBIAL DENSITIES In addition to the mite, the keratinous impactions in the sebaceous follicle, in particular microcomedones, contain aerobic and anaerobic bacteria as well as yeast-like fungi. Once retrieved by the follicular biopsy, microcomedones can be studied quantitatively for microbiology on an individual or pooled basis. The lesions are cut off the slide using a no. 15 scalpel under the dissecting scope.1 They are then placed singly or pooled into test tubes of appropriate plating fluids, homogenized, and serially plated. One hundred twenty individual microcomedones were homogenized and diluted in Tween 80. These yielded 410,000 ± 0.79 Propionibacterium acnes microorganisms per lesion, whereas closed comedones (75) yielded 100,000 ± 2.09 and open (75) 67,000 ± 2.11. The respective incidences were 98, 92, and 92%.7 When the scrub cup technique8 and the follicular biopsy have been compared in microbiology studies, it becomes clear that one microcomedone can easily yield the equivalent or more organisms than the surface scrub’s sampling.1 The mean in one study for P. acnes was 1.9 × 105 cm2 vs. 6.9 × 105 per microcomedone.1
94.8 LIPID ANALYSIS Of interest to those studying acne is the free fatty acidto-triglyceride ratio found in sebum, as well as the total amount of sebum being excreted. As with microorganisms, the microcomedones extracted by the follicular biopsy can be studied individually or pooled together. Lipids can be extracted using hexane, and incorporating methyl nervonate allows for qualitating and quantitating component lipids via thin-layer chromatography.7 When comparisons were done using the scrub cup technique, the mean free fatty acid-to-triglyceride ratio for the scrub was 0.29 per
827
square centimeter and 1.09 per microcomedone.7 Quantitatively, the means were 165.7 μg/cm2/3 h vs. 10.3 μg/microcomedone. Clearly the hydrolysis of glyceride was greater in the microcomedone than in the surface lipids. This outcome is consistent with the high density of P. acnes found in each lesion.
94.9 BIOCHEMICAL CHANGES The follicular biopsy permits biochemical study of abnormal sebaceous follicles. The presence of neutrophiles in microcomedones has been confirmed using antineutrophile antiserum and chemical assay of lysozyme. It has been possible to estimate IGG, IGM, and the C3 component of complement using immunoprecipitation techniques.9 These confirming findings are important in piecing together the pathology of inflammatory acne vulgaris and suggest further application of this approach.
94.10 ANALYZING DRUG LEVELS Where the abnormal hyperkeratinization of the sebaceous follicle is central to the pathology of disease, it is important to know if orally or topically administered drugs are reaching the target tissue. As the follicular biopsy retrieves the microcomedone in acne, analysis for the presence of administered drug and its metabolites can be done using high-pressure liquid chromatography. A study was conducted using two concentrations of benzoyl peroxide and one of topically applied erythromycin analyzing for their presence in microcomedones and for the degree of reduction of P. acnes.10 The findings were very interesting and indicated that there was a significant correlation between P. acnes reduction and the active drug concentration in the microcomedone. Less anaerobe reduction occurred in subjects who showed low concentrations of benzoyl peroxide and erythromycin and high concentrations of benzoic acid or anhydroerythromycin. These results also suggest the possibility of investigating individual responses of acne patients to active drugs. This approach may also shed light on why some patients are resistant to acne therapies. Drug metabolism or deactivation of the active drug may be occurring in some individuals. This remains to be investigated further.
94.11 ULTRAVIOLET EXAMINATION Some years ago, Martin and coworkers11 showed that by shining ultraviolet light on the surface of the skin, a coral red fluorescence located follicularly could be detected. This was due to P. acnes coporphyrins. Cornelius and Ludwig12 did additional work to define the nature of these porphyrins. This approach has been used on follicular
828
Handbook of Non-Invasive Methods and the Skin, Second Edition
biopsy samples to evaluate the density of P. acnes before and after therapy.13 It has also been used to separate keratinous impactions, which show the presence and location within the follicle of the anaerobe, and therefore allow for the separation of these for further study. Also, by using the follicular biopsy to assess porphyrin folliculitis, indicating the presence of P. acnes, one reduces the possibility of materials blocking this process when done on the surface — topically applied.
94.12 IMAGE ANALYSIS The follicular biopsy sample is ideal for the application of digital image analysis. In the most widely used method to date, follicular biopsies are evaluated under polarized light, which enables rapid measurement of density and size distribution of microcomedones.14 The images are captured through a light microscope, two polarizing filters, and a high-resolution video camera. A system is then used to digitize and analyze the captured video image. In a series of studies,4,14 there has been a high correlation between the traditional stereomicroscope scores for comedogenicity and the readout by digital image analysis (>0.9). This work represents an advance over traditional grading in that these can be central data interpretations of studies conducted at multiple investigative sites, and the method allows for mathematical analysis of results and easy data storage.
94.13 COMEDOGENIC AND COMEDOLYTIC EVALUATIONS Not unlike other disorders of the skin, the course of acne needs to be studied by looking at both intrinsic and extrinsic factors that influence pathology. Acne-prone individuals appear to be susceptible to the effects of certain formulations causing hyperkeratinization of the follicle (comedogenics). In order to assay for comedogenic potential, both animal15 and human16 models have been used. The follicular biopsy plays a key role in the later assay in detecting subclinical microcomedone formation. The original human comedolytic assay17 used an extrinsic agent to induct comedones and then treat these lesions. As microcomedones occur naturally, we are currently developing a human comedolytic assay that will use these lesions.
94.14 CONCLUSION The follicular biopsy offers a noninvasive approach for sampling the contents of the follicle. The biopsy is done quickly, and the extracted material can be saved for further study. The information gained by sampling inside the follicle is important to understanding the follicular canal
environment. Surface sampling techniques intended to gather information about the follicle may also include epidermal contributions. Part of unraveling the key steps in the pathology of follicular abnormalities may well come from further study of the samples retrieved by the follicular biopsy.
REFERENCES 1. Mills, O.H. and Kligman, A.M., The follicular biopsy, Dermatologica, 167, 57, 1983. 2. Marks, R. and Dawber, R.P.R., Skin surface biopsy: an improved technique for the examination of the horny layer, Br. J. Dermatol., 84, 117, 1971. 3. Holmes, R.L., Williams, M., and Cunliffe, W.F., Pilosebaceous duct obstruction and acne, Br. J. Dermatol., 87, 327, 1972. 4. Ayres, J.C., Mills, O.H., Lyssikatos, J.C., Kligman, A.M., and Groh, D.G., Assessment of a new method for determining the acnegenic potential of topically applied materials on human subjects, in Proceedings of the 17th Annual IFSCC International Congress, Vol. 2, Yokohama, Japan, p. 889. 5. Ayres, S. and Ayres, S., Demodectic eruptions (demodicidosis) in the human. 30 years experience with 2 commonly unrecognized entities: pityriasis folliculorum (demodex) and acne rosacea (demodes type), Arch. Dermatol., 83, 816, 1961. 6. Brazeau, C., Mills, O.H., and Kligman, A.M., Does a high population density of demodex folliculorum contribute to “sensitive” skin in adult females? in 18th World Congress of Dermatology: Progress and Perspectives, New York, June 12–18, 1992, p. 28A. 7. Mills, O.H., McGinley, K.J., and Kligman, A.M., The follicular biopsy in the study of acne (program), in 40th Annual Meeting, American Academy of Dermatology, San Francisco, 1981, p. 117. 8. Williamson, P.E. and Kligman, A.M., A new method for the quantitative investigation of cutaneous bacteria, J. Invest. Dermatol., 45, 498, 1965. 9. Webster, G.B. and Kligman, A.M., A new method for the assay of inflammatory mediators in follicular casts, J. Invest. Dermatol., 73, 266, 1979. 10. Wortzman, M., Scott, R., Wong, P., and Mills, O., A quantitative method for analysis of drug levels in the microcomedone, J. Invest. Dermatol., 82, 413, 1984. 11. Martin, R.J., Kahn, G., Gooding, J.W., and Brown, G., Cutaneous porphyrin fluorescence as an indicator of antibiotic absorption and effectiveness, Cutis, 12, 758, 1973. 12. Cornelius, C.E. and Ludwig, G.D., Red fluorescence of comedones: production of porphyrins by Corynebacterium acnes, J. Invest. Dermatol., 49, 368, 1967. 13. Weinstein, M., Mills, O., Berger, R., Dammers, K., and Baker, M., Follicular localization of Propionibacterium acnes by ultraviolet examination, Clin. Res., 36(3), 1988.
The Follicular Biopsy
14. Groh, D.G., Mills, O.H., and Kligman, A.M., The quantitative assessment of cyanoacrylate follicular biopsies by image analysis, J. Soc. Cosmet. Chem., 43, 101, 1992. 15. Kligman, A.M. and Mills, O.H., Acne cosmetica, Arch. Dermatol., 106, 843, 1972.
829
16. Mills, O.H. and Kligman, A.M., A human model for assessing comedogenic substances, Arch. Dermatol., 118, 903, 1982. 17. Mills, O.H. and Kligman, A.M., A human model for assaying comedolytic substances, Br. J. Dermatol., 107, 543, 1982.
of Excreted Sebum 95 Measurement Using Sebum-Absorbent Film and an Optical Reader: The TapeAnalyzer David L. Miller Bionet Incorporated, Dallas, Texas
CONTENTS 95.1 Introduction............................................................................................................................................................831 95.2 Methodological Principle ......................................................................................................................................831 95.2.1 Sebum-Collecting Device ..........................................................................................................................832 95.2.2 Measurement Device .................................................................................................................................832 95.3 Correlation with Other Methods ...........................................................................................................................832 95.4 Applications and Recommendations .....................................................................................................................833 95.4.1 Skin Type Evaluation.................................................................................................................................833 95.4.2 Treatment Evaluation.................................................................................................................................833 95.4.3 Sebum Excretion Rate ...............................................................................................................................833 References .......................................................................................................................................................................834
95.1 INTRODUCTION The use of sebum-absorbent film comprises a simple but elegant method for visualizing the dynamic pattern of sebum flowing out onto the skin surface. At the same time, in connection with an optical reader, this method is capable of providing numerical data on the delivery of sebum to the skin surface. The TapeAnalyzer is a portable, battery-powered device that incorporates detecting and enumerating hardware and electronic circuitry, providing by optical means a numeric readout of the density of absorbed sebum in the sebum-sensitive area of a specially designed DualTape* tester. As described in other chapters in this section of the handbook, the quantification of sebum excretion is the object of a number of techniques. Chapter 98 describes a gravimetric method: weighing the amount of sebum accumulated in various absorbent papers. Chapter 97 addresses a method involving the correlation of the change in light scattered by a frosted plastic film after it has been pressed against the skin surface. This chapter relates to * DualTape is a sebum collector strip manufactured by Cortex Technology.
the use of hydrophobic, microporous films for collecting sebum and visualizing the area distribution of active sebaceous glands as treated in U.S. Patent No. 5,119,828. SEBUTAPE®** and DualTape testers are commercial realizations of the device described in this patent. The DualTape tester is specifically designed for use in the TapeAnalyzer device. This approach is sensitive, specific, and accurate. At the same time, it is well tolerated by test subjects and patients, technically easy to perform, and robust with respect to high ambient temperature and relative humidity, where sweating becomes a problem that interferes with other methods.
95.2 METHODOLOGICAL PRINCIPLE In this method, sebum measurement consists of applying a sebum-collecting device to the skin surface, then measuring the collected sebum with an instrument that provides for objective, relative quantification of the collected amount of sebum. ** SEBUTAPE is the registered trademark for sebum test products manufactured by CuDerm Corporation.
831
832
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 95.1 Comparison of the appearance of DualTape and SEBUTAPE testers. Bar = 2 cm.
95.2.1 SEBUM-COLLECTING DEVICE Both the SEBUTAPE tester and the DualTape sebum collection strip (see Figure 95.1) utilize a layer of microporous polymeric film that specifically absorbs lipophilic liquids, including sebum. The unexposed film is opalescent because of light scattering from the air-filled micropores. Droplets of sebum occupying the sebaceous infundibula are absorbed into the micropore structure upon contact. The filled pores can no longer scatter incident light, but transmit it, while the surrounding air-filled micropores reflect incident light by scattering. This combined transmittance–reflectance principle provides a high degree of optical contrast between the spots of absorbed sebum and the surrounding film. The dark-colored lightabsorbing layer of the tester enhances this effect and provides a proper reference for optical measurement.1
95.2.2 MEASUREMENT DEVICE Figure 95.2 is a photograph of the Cortex TapeAnalyzer, a portable, battery-powered device that incorporates detecting and hardware and quantifying electronic circuitry. The measurement cycle consists of a two-step procedure: Step 1: Illumination of the unexposed and highly reflecting film in order to measure the amount of reflected light. By establishing a baseline for each individual strip of film, measurement variation — e.g., due to batch variation in the film material, light source variation, etc. — is minimized. Step 2: After the sebum-absorbing area has been applied to the skin for a few seconds, the strip is reinserted into the instrument, the film is illuminated, and the amount of reflected light is measured. The amount of reflected light is reduced as a function of sebum absorption.
FIGURE 95.2 The TapeAnalyzer showing end of DualTape inserted in reading slot.
The dynamic range (scaling) is factory set in accordance with expected maximum levels of sebum in order to obtain optimal sensitivity. The electronic circuitry converts the change in light reflection into 1 of 10 graded levels, which is indicated on the face of the instrument.2
95.3 CORRELATION WITH OTHER METHODS No published studies correlating TapeAnalyzer readings with other methods are available at the time of this writing. To provide an idea of the range of sebum levels obtained with the TapeAnalyzer, a number of samplings from various facial sites on several subjects using both the DualTape and SEBUTAPE testers were taken. The samples were taken using the TapeAnalyzer following the manufacturer’s instructions, then immediately subjected to image analysis using a standardized method3 that determines the total area of dark spots on the tester’s sensing area. The image analysis method used is similar to that described in Chapter 96. Figure 95.3 illustrates the correlation between the TapeAnalyzer readings and the appearance of the sample as viewed by the image analysis software. A comparison by eye of the images in Figure 95.3 reveals the SEBUTAPE tester sebum coverage to be slightly less than the DualTape tester at reader output levels 5 and 8. The results of all the testers evaluated in the comparison4 are shown in Figure 95.4. The data for the SEBUTAPE tester generally lie below those of the DualTape tester in the middle range of the instrument output. There is quite a bit of scatter in this graph because sample size is small and the instrument output is categorical rather than continuous. To a rough approximation the reader unit appears to be 1 mm2/cm2 film area (or 1% of film area covered with sebum spots). Because of the categorical output levels, sebum coverage is less than ~1 mm2/cm2 yield an output value of 1 and sebum coverage is greater
833
Sebutape® tester
Dual tape tester
Measurement of Excreted Sebum Using Sebum-Absorbent Film and an Optical Reader: The TapeAnalyzer
Reader result = 1
Reader result = 5
Reader result = 8
Sebum area per sample area film (mm2/cm2)(y)
FIGURE 95.3 Comparison of sebum patterns on DualTape and SEBUTAPE testers yielding three different output levels as measured by the TapeAnalyzer.
12 10 8
y = 0.7133x + 0.2725 R2 = 0.8016
6 4
Dual tape tester Sebutape tester Linear (dual tape tester) Linear (sebutape tester)
95.4.1 SKIN TYPE EVALUATION Skin type evaluation is the simple assessment of an individual’s facial sebum distribution — a snapshot of skin condition. TapeAnalyzer readings are obtained for various facial skin sites (chin, nose, forehead, and cheek). The results can be used to qualify test panelists prior to clinical and efficacy studies of skin care products or to facilitate recommendation of skin care products at the point of sale.
95.4.2 TREATMENT EVALUATION 2 0
0
2
y = 0.8527x − 1.0911 R2 = 0.5668 4 6 8 10 Reader units (x)
FIGURE 95.4 Graph of experimental data comparing TapeAnalyzer readings to area of sebum spots on the tester.
than ~10 mm2/cm2 yield a value of 10. Measurements obtained at the high end of the scale are therefore not analytically useful, as they indicate at least some level of sebum and all greater amounts.
95.4 APPLICATIONS AND RECOMMENDATIONS Measurements obtained using the TapeAnalyzer are most useful for categorizing the state of the skin surface at the time of testing. By varying the protocol of measurement sequences, various features of test products can be evaluated.
Evaluation of specific treatment effects is best accomplished using a paired site model. For example, four 2 × 2 cm sites are marked on the left and right sides of the forehead in a grid fashion. The sebum level at one of the sites is measured pretreatment; then randomized active and placebo treatments are applied to the left and right sides. Follow-up measurements are subsequently made at the remaining sites at specified times after treatment. This type of experiment yields data on efficacy of cleansers and oil control products. Subject qualification is important for meaningful results: pretreatment levels should be in the range of 5 to 9 reader units. For the most reliable results, the test areas should be cleaned with alcohol 2 hours prior to taking the pretreatment measurements.
95.4.3 SEBUM EXCRETION RATE A rank estimate of sebum excretion rate is obtained by preconditioning the measurement area: all makeup is removed and the site is delipidized (e.g., alcohol wash). After a predetermined period (e.g., 1 or 2 hours), the
834
Handbook of Non-Invasive Methods and the Skin, Second Edition
sebum measurement is performed. With a fixed time for comparison, this method provides information about the sebum excretion rate at the given time and should be useful for monitoring the diurnal variation sebum output. This would also be an approach to monitoring sebum output during the course of various drug therapies as well as physical and psychological stressors that have an impact on the activity of the sebaceous gland. Although the TapeAnalyzer can be used for pilot range finding in this application, the preferred instrument is the DermaLab® DualTape Reader,* due to its superior resolution and sensitivity. The DermaLab DualTape Reader employs the same detection hardware as the TapeAnalyzer, but it uses a more sophisticated data readout. * DermaLab DualTape Reader is an application module for the DermaLab modular series of instruments providing a continuous 0 to 100% readout of the film saturation and two sensitivity settings, as well as a built-in tape application timer. DermaLab is a registered trademark of Cortex Technology.
Overall, the TapeAnalyzer is most useful for informal, on-the-spot skin type assessment as a tool in recommending skin care products and treatments. Although handy for survey studies in evaluating product efficacy, as shown by this author’s investigation, the categorical nature of the output data limits its usefulness in formal scientific studies, where the preferred instrument is the DermaLab DualTape Reader module.
REFERENCES 1. Product literature provided by Cuderm Corporation, Box 797686, Dallas, TX 74379-7686. 2. Product literature provided by Cortex Technology, Smedevaenget 10, 9560 Hadsund, Denmark. 3. Bionet Incorporated SEBUTAPE Image Analysis Protocol, Box 797686, Dallas TX 75379-7686. 4. Author’s unpublished data.
of Sebum Output 96 Quantification Using Sebum-Absorbent Tapes (Sebutapes®) Claudia El Gammal,1 Stephan El Gammal,1 Alessandra Pagnoni,2 and Albert M. Kligman3 1
Dermatological Clinic, Hospital Bethesda, Freudenberg, Germany Philadelphia, Pennsylvania 3Department of Dermatology, University of Pennsylvania, Philadelphia, Pennsylvania 2
CONTENTS 96.1 Summary ................................................................................................................................................................835 96.2 Introduction............................................................................................................................................................836 96.3 Material and Methods............................................................................................................................................836 96.3.1 Image Analytical Evaluation .....................................................................................................................836 96.4 Clinical Evaluation of Sebum Production.............................................................................................................837 96.5 Problems and Artifacts ..........................................................................................................................................838 96.5.1 Storage of Sebutapes .................................................................................................................................838 96.5.2 Illumination, Filtering of Images, Interactive Analysis ............................................................................838 96.5.3 The So-Called Reservoir Effect ................................................................................................................839 References .......................................................................................................................................................................840
96.1 SUMMARY With the introduction of Sebutapes®, a method has become available that allows not only the measurement of sebum output as a single global value of the lipid amount present on a given surface, but also the assessment of differences among the activities of individual sebaceous follicles. Sebutapes are microporous, white adherent tapes. When applied on defatted skin, absorbed lipids become visible as transparent spots. Viewing these against a black background in reflection mode results in a black-and-white picture. Image analysis is based upon gray-level thresholding and filtering of the image. Prior to analysis, application of a shading correction algorithm eliminates graylevel gradients due to unequal illumination. The following parameters were evaluated: •
The percent area covered with sebum spots representing the overall amount of sebum produced by the follicles in this area (sebum excretion rate).
•
•
The number of sebum droplets. This correlates with the number of active follicles. However, high sebum production leads to confluence of the spots, resulting in falsly low counts. The mean and maximum size of the sebum spots. Particularly large, confluent spots are typical for patients with seborrhea and acne patients.
For proper interpretation the following factors have to be considered. Due to the reservoir effect of the sebaceous follicle, sebum output after defatting is highest in the beginning, reaching a constant level after a few hours. Sebum spots are subject to changes in their size and transparency depending on storage time. Especially crystallization processes within the droplets significantly affect the resulting values. Sebutapes should be evaluated at a defined time after removal, preferably within 24 hours. When immediate evaluation is not possible, storage in a freezer at –30°C is advisable. Preferentially, Sebutape images should be captured immediately by an analysis program and stored for delayed evaluation. 835
836
Handbook of Non-Invasive Methods and the Skin, Second Edition
96.2 INTRODUCTION For decades there have been various attempts to quantify the excretion of sebaceous glands. They all depend on recovery of lipid from the skin surface. Mostly the forehead is used as a site of high sebum production, where the contribution of epidermal lipids to the overall lipid content is negligible. Already at the turn of the last century absorbance of sebum into paper and consecutive weighing was performed. Based on these approaches Strauss and Pochi (1961) developed the so-called cigarette paper method: the forehead is wiped with gauze and a stack of four cigarette papers is applied in a standard-size rectangle over a 3-hour period. The absorbed lipids are then extracted with diethyl ether and weighed after evaporation of the ether. Another, much more recently introduced material for the collection of sebum is bentonite clay, which is spread on the skin as an aqueous gel containing 0.2% carboxymethylcellulose (Downing et al. 1982). As the clay can absorb large amounts of sebum, it is possible to leave it on the skin for long periods. A standardized protocol includes a 14-hour sebum depletion step followed by a 3-hour measurement step. Since chromatographic techniques have become available, the composition of the collected and extracted lipids can be determined. Skin surface lipids originate from both the stratum corneum and the sebaceous glands; distinguishing the contributions of these two source was a problem at first. Today it is known that epidermal lipids consist of triglycerides, free fatty acids, and cholesterol, while sebum is composed of triglycerides, wax esters, squalene, cholesterol, and cholesterol esters (Stewart et al. 1983). The disadvantage of methods involving lipid collection into certain materials is that they are rather timeconsuming. To quickly obtain information about the lipids present on the skin surface, two similar instruments have been developed: the Lipomètre (Saint-Leger et al. 1979) and the Sebumeter (Thune and Gustavsen 1984; Dikstein et al. 1987; Serup, 1991). Both use photometry to determine the lipid amount, with fat changing the light transmission through opalescent glass or plastic film. With these instruments, usually casual surface lipid levels are measured (amount recovered from skin that has not been recently defatted or protected in any way). More accurate information is provided after standardized defatting and measurement at a defined time point several hours later. All above-mentioned techniques have in common that they yield a single global value of the lipid amount present on a given surface. Differences among the activity of individual sebaceous follicles cannot be evaluated. With the introduction of the Sebutapes® (CuDerm Corp., Dallas, TX) in 1986, however, a reliable morphologic method became available, with which not only the sebum excretion rate in general, but also the output of individual
follicles can be monitored (Kligman et al. 1986; Nordstrom et al. 1986; Pierard 1986; Serup 1991).
96.3 MATERIAL AND METHODS Sebutapes are white, open-celled, microporous, hydrophobic films coated with an adhesive layer that adheres to the skin surface. They come in the size of 2.5 cm2. Usually, they are applied on the forehead; however, sebum output can theoretically be measured on all parts of the body, except in areas covered with terminal hairs, preventing the tape from sticking to the skin. Before application, it is important to defat the skin in order to remove surface lipids, cell debris, or residual cosmetics. Ethanol, hexane, and Pannoclean (Pannoc Chemie, Herentals) are all suitable; it is, however, important to stick to a certain protocol. We used the following procedure: the skin was washed with a detergent solution, 0.1% Triton X 100, rinsed with tap water, and then gently rubbed with a cotton swab soaked in hexane for 15 seconds. Most research groups agree that an application time of 1 hour is sufficient and gives the most accurate results. Figure 96.1 shows application of Sebutapes on different regions of the face.
96.3.1 IMAGE ANALYTICAL EVALUATION Sebum droplets, which are extruded from the follicles, are absorbed by the Sebutape and become visible as transparent spots. Viewing the tape against a black background results in a black-and-white image. For this purpose, the samples are fixed on black filing cards or on glass slides, which are placed on a black board before analysis. Information about several parameters can be obtained: 1. The percent area of the Sebutape covered with sebum spots represents the overall amount of sebum produced by the follicles in this area.
FIGURE 96.1 Sebutape fixed on the face of a 35-year-old woman.
Quantification of Sebum Output Using Sebum-Absorbent Tapes (Sebutapes®)
Many studies have demonstrated the reliability and reproducibility of this value and its good correlation with results gained by other methods (Kligman et al. 1986; Serup 1991). 2. The number of sebum spots provides information about the number of active follicles in this area. Generally each droplet corresponds to one active follicle. However, the more sebum that is produced, the more confluence of the spots that occurs, resulting in falsely low follicle counts. Confluent follicle spots can be recognized by their irregular, for example, triangular, shape, while droplets produced by a single follicle are round or elliptic. The droplet count is especially interesting when comparing the same site in a person at different time points, e.g., during a treatment phase. When, however, different locations and persons with large variations in sebum excretion rates are assessed, this number is of limited value. 3. The sizes of the sebum spots are important as well. Their mean sizes are much higher in the center of the face — on the nose, chin, midforehead — than in its lateral parts, like the cheeks. In seborrhoic patients the spot sizes are very irregular, ranging from tiny spots of 30 μm diameter to big, confluent areas of more than 3 mm diameter. To quantify these parameters, image analysis is an indispensable tool. Evaluation is based on gray-level thresholding of the image, thus telling the computer program what to recognize as foreground for the calculations. Figure 96.2 shows the image of a Sebutape with its histogram of gray-level distribution in the region of interest. The gray levels range from 0, which is black, to 255, which corresponds to white. The histogram of the displayed Sebutape shows two peaks, the left one representing the black
FIGURE 96.2 Sebutape from the forehead of a 27-year-old woman with the region of interest (inner light rectangle).
837
spots caused by the sebum and the right, larger one representing the white or light gray background. According to the histogram curve, a trigger is applied on the image; the gray-level value of its upper border must be lower than the gray levels of the background. Black or dark gray sebum spots are detected. From the marked areas, the computer program calculates the total area covered, the spot number, and the mean and maximum spot sizes.
96.4 CLINICAL EVALUATION OF SEBUM PRODUCTION Sebutapes have largely replaced outdated, older methods of measuring sebum excretion rates, like the solvent extraction and gravimetric methods. Since their introduction, many questions regarding the sebum output in general, as well as the activity of single follicles, have been investigated. Pierard et al. (1987) disclosed four patterns of sebum excretion rates that are relatively specific for certain age groups: In infants, very few, tiny spots are present. The pubertal pattern is characterized by about 100 spots/cm2, which are small and rather uniform in size. The adult pattern shows a greater number of spots (120 to 280/cm2), with greater variation in size. In old age, a decline in spot number, but not size, compared to the adult pattern, can be observed. Additionally, the authors distinguish the socalled acne pattern, which can be recognized by an unusually great number of irregularly shaped and very large spots. We compared sebum secretion with the density of hair follicles in nine different areas of the face in 12 women, aged 20 to 45 years (Pagnoni et al. 1994). The follicular density was counted microscopically in cyanoacrylate imprints. We found that both the follicular density and the sebum production decreased from medial to lateral in the face. While the number of follicles was relatively constant between individuals, sebum secretion varied greatly. A direct correlation between follicular density and sebum production could therefore not be demonstrated. We can conclude that differences in sebum production are due to the varying activity of single follicles and not to their distribution (Pagnoni et al. 1994; El Gammal et al. 1995b). Correlation of Sebutapes with cyanoacrylate surface biopsies (Pierard et al. 1987) revealed that at least 50% of the sebaceous follicles are quiescent, without any detectable sebum excretion, and that there is no relation between the size of a follicular opening and its sebum output. Using Sebutapes, the hypothesis could be confirmed that sebum output underlies seasonal variations with increased delivery of sebum to the skin surface at higher ambient temperatures. The number of active follicles,
838
Handbook of Non-Invasive Methods and the Skin, Second Edition
however, does not change with season (Pierard-Franchimont et al. 1990). In women, the ovarian cycle influences the sebum output with a peak during the week of ovulation; this effect is most pronounced in very oily persons (Pierard-Franchimont et al. 1991). Blume et al. (1991) measured the growth of vellus hair on the forehead, cheek, back, chest, and shoulder of men and compared it to the sebum excretion rate as assessed by Sebutapes. They found no link between the two parameters. Sebutapes are unique with regard to the fact that the sebum production of individual follicles can be assessed. As opposed to other techniques, they provide not only a single value, but a morphological pattern of sebum production. Image analysis makes it possible to exactly quantify these patterns in a time-economic manner.
96.5 PROBLEMS AND ARTIFACTS 96.5.1 STORAGE
OF
SEBUTAPES
When Sebutapes are stored over longer periods of time at room temperature, several changes within the tapes occur that influence the results: •
•
Within the first 24 hours after removal the spots gradually enlarge. This effect is especially pronounced in samples with large sebum spots and a high percent area covered. It is most probably due to spreading of the collected sebum within the tapes. Immediate evaluation may yield considerably lower values than analysis 1 day after removal. Figure 96.3 shows the same Sebutape evaluated under the exact same conditions 5 minutes and 24 hours after removal from the skin. The increase in spot sizes leads to an increase of the total area covered, from 6.77 to 8.12. The number of spots, however, decreased from 165 to 148 due to confluence. After longer periods of time (weeks), many of the spots may become whitish, probably due to crystallization processes within the sebum. Figure 96.4 demonstrates this phenomenon, showing the same Sebutape photographed 15 minutes and 30 days after removal from the skin. Spots that have turned white or light gray are difficult to distinguish from their surroundings and are not detected anymore by gray-level thresholding of the image.
In Figure 96.5 the influence of room temperature storage on the percent area covered is shown. A Sebutape from the forehead was evaluated by image analysis in exactly the same area under the same conditions at different times.
FIGURE 96.3 Sebutape from the forehead of a 16-year-old boy immediately after removal from the skin (upper image) and 24 hours after removal, stored at room temperature (lower image). There is considerable enlargement of the spots from the upper to the lower image.
In the first hours there was a gradual increase in the percent area covered due to the mentioned enlargement of the spots. The values then remained stable during the following week. However, after 4 and 8 weeks, evaluation resulted in a much lower percent area covered because many of the spots had turned white. To obtain the most accurate and comparable results, Sebutapes should be evaluated either immediately or at a certain constant time interval after removal from the skin, for example, at 24 hours. Storing the samples in a freezer at –30°C will prevent, or at least considerably delay, the turning white of the sebum spots. Also, the increase in size of the spots can be kept to a minimum that way. We therefore recommend that all Sebutapes be stored in a freezer when intending to evaluate them at later time points. A convenient way to record Sebutapes is to save the images captured by the analysis program on hard or floppy discs. We use the TIFF picture format, which enables the transfer of the images to other computers and programs. Image analytical evaluation can then be performed at any time, independently from gaining the samples.
96.5.2 ILLUMINATION, FILTERING INTERACTIVE ANALYSIS
OF IMAGES,
Homogeneous illumination of the specimens is often a problem. In our experience, best results are obtained when Sebutapes are placed in a white light box and illuminated from two sides by means of two fiber-optic
Quantification of Sebum Output Using Sebum-Absorbent Tapes (Sebutapes®)
839
12 % Area covered
10 8 6 4
FIGURE 96.4 (a) Sebutape from the forehead of a 28-year-old woman 15 minutes after obtaining the sample.
8 weeks
4 weeks
9 days
12 days
7 days
5 days
1 days
3 days
4 hrs
3 hrs
1 hr
2 hrs
30 min
0
Immediately
2
After removal from the skin
FIGURE 96.5 Percent area covered of a Sebutape from the forehead evaluated by image analysis at different time points after removal.
gray threshold is necessary than without filtering. Once a certain filter is chosen, one has to make sure that all evaluations are performed only after filtering of the image to ensure comparability of the results. Some image analysis programs allow one to interactively manipulate the binary picture before analysis. Spots that have not been recognized can be added, artifacts can be deleted, or such spots can be cut apart, which resulted from confluent sebum spots of several follicles. Although working on the images in that way may improve accuracy, in our opinion such procedures should be avoided as much as possible. They not only reduce the objectivity and reproducibility of the evaluation, but also are very time-consuming. FIGURE 96.4 (b) The same Sebutape 30 days after removal from the skin, stored at room temperature. Many of the sebum spots have turned light gray or white.
light carriers. We use a high-resolution black-and-white charge coupled device (CCD) video camera mounted on a stereomicroscope. To capture the images, the video signal is digitized by an IP8 frame grabber board using 255 gray levels. To further eliminate gray-level gradients due to nonuniform illumination, a shading correction algorithm can be applied on the image frozen by the analysis program. The image is blurred using a median filter with a large matrix, and then subtracted from the original image. In order to enhance contrast, it is very useful to apply an edge detection filter on Sebutape images before analysis. This leads to a sharper image, and especially very small sebum spots can be recognized more readily by application of a trigger level. On filtered images, a lower
96.5.3 THE SO-CALLED RESERVOIR EFFECT When Sebutapes are applied on the skin successively in the same area, each for 1 hour, the sebum output is highest in the beginning, reaching a constant level after a few hours. This phenomenon has been called the reservoir effect of the sebaceous gland and has been ascribed to the fact that after degreasing the delivery of sebum from the follicular reservoir to the stratum corneum, the surface is increased (Downing et al. 1982; Stewart et al. 1983; SaintLeger and Cohen 1985). The pilosebaceous infundibulum contains considerable amounts of sebum, which have been estimated at a few milligrams per square centimeter. Sebum collected in Sebutapes during the first hours of application originates from this reservoir and does not necessarily reflect the production within the sebaceous gland. The latter is more accurately expressed by the plateau level reached after a few hours of sampling. Figure 96.6 illustrates the reservoir effect, showing the percent area covered of six Sebutapes applied consecutively over 1 hour on the nose, chin, and forehead.
840
Handbook of Non-Invasive Methods and the Skin, Second Edition
25
Nose Chin Forehead
% Area covered
20 15 10 5 0
1st
2nd 3rd 4th 5th One hour of application per sebutape®
6th hour
FIGURE 96.6 Percent area covered of consecutively applied Sebutapes on the forehead, nose, and chin, each for 1 hour for a total time of 6 hours. The sebum excretion rate decreases in the beginning, reaching a constant level after 3 hours.
REFERENCES Blume, U., Ferracin, J., Verschoore, M., Czernielewski, J.M., Schaefer, H., Physiology of the vellus hair follicle: hair growth and sebum excretion, Br. J. Dermatol., 124, 21, 1991. Dikstein, S., Zlotogorski A., Avriel, E., Katz, M., Harms, M., Comparison of the Sebumeter and the Lipometer, Bioeng. Skin, 3, 197, 1987. Downing, D.T., Stranieri, A.M., Strauss, J.S., The effect of accumulated lipids on measurements of sebum secretion in human skin, J. Invest. Dermatol., 79, 226, 1982. El Gammal, C., El Gammal, S., Altmeyer, P., Bildanalyse in der Dermatologie: Quantifikation von Schuppung mit Adhaesivfolien (D-Squames®), Akt. Dermatol., 21, 73, 1995a. El Gammal, C., El Gammal, S., Pagnoni, A., Kligman, A.M., Sebum-absorbent tape and image analysis, in In Vivo Examination of the Skin: A Handbook of Non-Invasive Methods, Serup, J., Jemec, G.B.E., Eds., CRC Press, Boca Raton, FL, 1995b, p. 517.
Kligman, A.M., Miller, D.L., McGinley, K.J., Sebutape: a device for visualizing and measuring human sebaceous secretion, J. Soc. Cosmet. Chem., 37, 369, 1986. Nordstrom, K.M., Schmus, H.G., McGinley, K.J., Leyden J.J., Measurement of sebum output using a lipid absorbent tape, J. Invest. Dermatol., 87, 260, 1986. Pagnoni, A., Kligman, A.M., El Gammal, S., Stoudemayer, T., Determination of density of follicles on various regions of the face by cyanoacrylate biopsies: correlation with sebum output, Br. J. Dermatol., 131, 862, 1994. Pierard, G.E., Follicle to follicle heterogeneity of sebum excretion, Dermatologica, 173, 61, 1986. Pierard, G.E., Rate and topography of follicular sebum excretion, Dermatologica, 175, 280, 1987. Pierard, G.E., Pierard-Franchimont, C., Le, T., Lapiere, C., Patterns of follicular sebum excretion rate during lifetime, Arch. Dermatol. Res., 279, 104, 1987. Pierard-Franchimont, C., Pierard, G.E., Kligman, A.M., Rhythm of sebum excretion during the menstrual cycle, Dermatologica, 182, 211, 1991. Pierard-Franchimont, C., Pierard, G.E., Kligman, A.M., Seasonal modulation of sebum excretion, Dermatologica, 181, 21, 1990. Saint-Leger, D., Berrebi, C., Duboz, C., Agache, P., The lipometre: an easy tool for rapid quantitation of skin surface lipids (SSL) in man, Arch. Dermatol. Res., 265, 79, 1979. Saint-Leger, D., Cohen, E., Practical study of qualitative and quantitative sebum excretion on the human forehead, Br. J. Dermatol., 113, 551, 1985. Serup, J., Formation of oiliness and sebum output: comparison of a lipid-absorbant and occlusive tape method with photometry, Clin. Exp. Dermatol., 16, 258, 1991. Stewart, M.E., Downing, D.T., Strauss, J.S., Sebum secretion and sebaceous lipids, Dermatol. Clin., 1, 335, 1983. Strauss, J.S., Pochi, P.E., The quantitative gravimetric determination of sebum production, J. Invest. Dermatol., 36, 293, 1961. Thune, P., Gustavsen, T., Comparison of two techniques for quantitative measurements of skin surface lipids, Acta Dermatol. Venereol. (Stockh.), Suppl. 134, 30, 1984.
Measurement of Sebum 97 Optical Excretion Using Opalescent Film Imprint: The Sebumeter® Ken-ichiro O’goshi Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark
CONTENTS 97.1 97.2 97.3 97.4 97.5
Introduction and Background ................................................................................................................................841 Measuring Principle of the Sebumeter..................................................................................................................842 Measuring Device and Practical Use ....................................................................................................................842 Casual Lipid Level and Sebum Excretion Rate....................................................................................................843 Normal Skin and Physiological Variation .............................................................................................................844 97.5.1 Age, Sex, and Endocrine Status ................................................................................................................844 97.5.2 Recovery of Casual State ..........................................................................................................................844 97.5.3 Skin Temperature.......................................................................................................................................844 97.5.4 Sweat Gland Function ...............................................................................................................................844 97.5.5 Diurnal Variation........................................................................................................................................844 97.5.6 Site Variation..............................................................................................................................................844 97.5.7 Disease .......................................................................................................................................................844 97.6 Study of Normal Skin............................................................................................................................................844 97.7 Study of Skin Diseases..........................................................................................................................................845 97.8 Study of Product Efficacy .....................................................................................................................................845 References .......................................................................................................................................................................845
97.1 INTRODUCTION AND BACKGROUND Skin surface lipid consists of the lipids derived from epidermis and sebaceous glands. Keratinocytes synthesize lipids found in the cell as odd land bodies and between cells as intercellular bilayers important for the barrier function. Sebum produced by the sebaceous gland as a holocrine excrete is an oily mixture of lipids, keratin, and cellular membrane, in the natural form also containing cells in the form of sebocytes. The dominating lipids of sebum are triacylglycerol (triacylglyceride), fatty acid, cholesterol, ceramides, phosphoric acid, and other mixtures of lipids.1 The sebaceous glands form a part of the pilosebaceous unit. Commonly, they are found in connection with a hair follicle. The sebaceous follicles are found on the face, chest, and back, mainly where we can clinically often see acne vulgaris. Acne is a unique disease of
the human because sebaceous follicles with such distribution are not found in any animal. For more than 50 years, several methods for the determination of sebum output have been introduced.2 There have been different methods based on scraping, washing, and extraction. Absorbent paper was introduced as early as 1886 by Krukenberg and in 1899 by Leubuscer.1 Cunliffe measured sebum excretion by a cup method, in which an organic solvent is placed, for various periods, in direct contact with an area of skin such as the forehead, the solvent being contained in a circular glass cup held firmly in contact with the skin.2,3 The solvent is then pipetted from the cup, evaporated, and the quantity of sebum ascertained by weighing. This technique is still in use. For example, a variant of the original method entails cleaning the skin surface by wiping with a gauze pad moistened with a 1% solution of the nonionic detergent Triton X100.4 Using an internal standard, the amount of the lipid can also be analyzed. There are certain reservations about 841
842
Handbook of Non-Invasive Methods and the Skin, Second Edition
this method of collection, the most important of which is that it physically extracts lipids from both the surface and the duct, and it does not measure excretion specifically — albeit on the forehead at least, the epidermal contribution in adolescents is minimal.5 Extraction by washing is strongly dependent on the solubility of various lipids and on the depth of washing in the epidermis. Strauss and Pochi6 used cigarette paper on the forehead for 3 hours and weighed the lipid extracted from the paper. This cigarette paper method was reliable, and it became quite popular despite the fact that it was timeconsuming. The holocrine excretion from the sebaceous gland has a slow process with a duration of days. Cunliffe and Shuster7 suggested that the result of cigarette paper samplings reflected the pool of sebum already excreted within the pilosebaceous duct and gland, which should have a correlation with true glandular activity. The cigarette paper method was later simplified by using a direct gravimetric measurement of the absorbent paper.8 Bentonite clay and wax esters for lipid sampling absorption have been introduced for sampling over a day’s period to aim at a more specific determination for gland product.9 It takes 17 hours to be done. Schäfer and Kuhn-Bussius10 described a more simple photometric method. Using the Lipométre L’Oreal based on opalescent glass becoming more transparent to light upon lipid adhesion, this principle became validated, and sebum excretion rate and the ambient level of skin surface lipids could be measured precisely.11–14 Later, a more convenient equipment with an opalescent plastic film, instead of glass, named the Sebumeter® (Courage + Khazaka, Cologne, Germany), became commercially available.15,16 Now, photometric measurement methods have widely replaced the above-mentioned cigarette paper method. Skin surface lipid and the function and distribution of individual glands can be demonstrated with replicas stained with osmic acid vapor.2,17 The Sebutape® (CuDerm Corp., Dallas, TX) was introduced as a newer method in 1986. It comprises an open-celled, microporous, hydrophobic polymetric film on an adhesive layer that would permit the passage of lipids into the filter as the strip adheres tightly to the skin surface. Lipid absorption with this tape showed a high correlation with hexane extraction of skin surface lipid.18 Active sebaceous glands were seen as prints of holes on the tape. It could be classified by image analysis into small, regular, large, or irregular spots, the last being especially prevalent in acne and in aged skin. The sebum excretion rate determined by image analysis of follicular spots in the tape and by photometry using the Lipométre showed a significant correlation.19 The Sebutape was evaluated by scoring and by densitometric quantification, which had recently been found useful as a simple and accurate tool for quantifying another tape method, D-Squame®, expressing scaliness and dryness.20 Due to
the virtue of simplicity, the tape method is now used in cosmetic salons.21
97.2 MEASURING PRINCIPLE OF THE SEBUMETER The Sebumeter is a modification of the ground of Lipométre with glass technique, but it uses an opaque plastic tape instead of a glass plate. The measuring principle of the Sebumeter is based on the original ground glass plate method.10 In a comparative study, the plastic tape method was found to be equivalent to the ground glass method.22,23 Opacity becomes translucent when the surface is covered by lipids. The translucency is correlated with the amount of skin surface lipids and is measured with a special spectrophotometer.11 The disadvantage of the ground glass plate method was that the glass had to be cleaned between sebum collections, while tape is disposed and easily standardized.
97.3 MEASURING DEVICE AND PRACTICAL USE The Sebumeter consists of a plastic film cassette (Figure 97.1) and the measuring unit with the spectrometer (Figure 97.2). The method of measurement with the Sebumeter is very simple and easy, with the results displayed immediately. The measuring head of the cassette (Figure 97.3) exposes a 64 mm2 measuring section with a mirror mounted in the cassette behind the tape. For a measurement, at first the used tape should be transported forward by a trigger at the side of the cassette to expose a fresh section of the tape. The used tape is maintained inside the cassette. Each measurement starts with a zero calibration. The head of the plastic film cassette is inserted into the measuring opening of the device. A signal indicates measuring of the plastic film to zero. With the zero calibration cycle finished, the plastic film cassette is removed and the application head of the cassette
FIGURE 97.1 The Sebumeter plastic film cassette (side view).
Optical Measurement of Sebum Excretion Using Opalescent Film Imprint: The Sebumeter®
843
Moveable probe part Fixed probe part Mirror Matted plastic tape
FIGURE 97.2 The Sebumeter with the plastic film cassette in the measurement slot. (a)
Moveable probe part Fixed probe part Mirror Matted plastic tape
Photodetector
FIGURE 97.3 The Sebumeter plastic film cassette: close-up of the sebum collection head, showing the matted plastic film over the mirror.
is pressed under a gentle standard pressure onto the skin surface, in order to collect the sebum of the investigative skin area for 30 seconds (Figure 97.4a). A spring in the cassette head provides constant pressure on the skin surface. During sampling, the film changes from opalescent to transparent due to the absorbed sebum. After a beep that is signaled for the end of the skin exposure time, the head of the plastic film cassette is reinserted into the measuring opening of the device once again, and the transparency of the film is measured automatically by the spectrometer housed in the main body (Figure 97.4b). The result is displayed as a numerical value on an LCD display 3 seconds after. Within a range of 50 to 300, the indicated arbitrary value corresponds to the sebum amount on the skin surface in micrograms per square centimeter.
97.4 CASUAL LIPID LEVEL AND SEBUM EXCRETION RATE Sebum excretion varies during the day and from day to day.24 The natural and undistributed skin lipids spread as a nonuniform thin layer on the skin surface. This sponta-
Light source Measuring unit (b)
FIGURE 97.4 (a) Sebum collection step of the Sebumeter. (b) Sebum measurement step of the Sebumeter. (From Elsner, P., in Bioengineering of the Skin: Methods and Instrumentation, Berardesca, E. et al., Eds., CRC Press, Boca Raton, FL, 1995, p. 83.)
neous layer is known as the casual level of skin surface lipids, expressed in micrograms per square centimeter. In oily skin, it may appear as microscope droplets.25 Rode et al.26 described the relationship between casual sebum level and sebum excretion rate. In their studies, there was no statistically significant day-to-day variation using the Sebumeter. The casual level was following decreasing reached after 2 hours in all seborrheic areas, irrespective of casual level of oiliness. The sebum excretion rate (SER) is defined as the sebaceous lipid output in a defined surface area of skin over a defined period, proceeded by a rinsing procedure, bringing the surface level down to a baseline called zero at the inhibition of measurement. Rinsing before measurement of the SER is important and must be standardized and validated. Alcohol swabs
844
Handbook of Non-Invasive Methods and the Skin, Second Edition
may be used. In regions with high casual lipid level, rinsing shall be quite aggressive. It can be checked with the Sebumeter that the baseline is established with values in the range of 5 to 10 μg/cm2. With too aggressive rinsing, too much lipid can be removed. A standard of rinsing using a shampoo was already published.26
97.5 NORMAL SKIN AND PHYSIOLOGICAL VARIATION 97.5.1 AGE, SEX,
AND
ENDOCRINE STATUS
Age, sex, and endocrine status all effect SER because the sebaceous glands are under endocrine control. So, it is not surprising that sebum production varies with age and sex. Cunliffe and Shuster27 investigated 139 control subjects aged between 10 to 70 years. They found that males had a significantly greater SER than females, except in the age group 10 to 15 years. This is probably explained by the earlier development of puberty in the female. The SER continued to increase after puberty, reaching a peak in both sexes between 30 and 40 years; thereafter there was a gradual decline. There was no obvious accelerated rate of decline at menopause, but there is no investigation on the SER at this stage. The SER of infants and children is not well studied, but apparently low. Oral contraceptives decrease SER. The effect of the menstrual cycle is probably of minor importance to SER. Androgens, especially from the testes and adrenals, stimulate the sebaceous gland directly. Estrogens exert an inhibitory effect in pharmacological doses. The action of estrogen is chiefly indirect by inhibiting hormonal release from the pituitary, and the effect of the estrogen is easily overcome by relatively small doses of androgen. Progesterone probably has no major effect on SER in the adult. The pituitary gland, in particular the anterior lobe, has an important role in controlling sebum excretion. The anterior pituitary exerts its effect through its target organs, in particular the gonads and adrenals. Zeller and Huber28 investigated forehead sebum levels as a function of sex and age and found higher Sebumeter readings in young (mean age, 22 years old) females than in young males, while the levels in elderly females were lower than in elderly males.
97.5.2 RECOVERY
OF
CASUAL STATE
Physiological recovery is known from SER studies. Duration of collection affects the apparent SER, and variance of the measurements is most marked in the first and second hours.27 Secretion plateaus after 2 to 3 hours.
97.5.3 SKIN TEMPERATURE A 1˚C change in skin surface temperature produces a 10% change in SER.29
97.5.4 SWEAT GLAND FUNCTION Sweating reduces the amount of lipid collected by absorbent papers, and a thermally neutral environment is necessary during measurement of SER. If the object is sweating visibly during the procedure, the paper technique should not be used.
97.5.5 DIURNAL VARIATION SER is maximal in the middle morning and minimal during the late evening and early morning.24
97.5.6 SITE VARIATION Sebaceous glands are found predominantly on the scalp, face, chest, and back. The SER of the forehead is five times that on the back, but when related to the number of glands at each site, there is no difference.30
97.5.7 DISEASE SER correlates well with acne and past history of the disease. And since sebum excretion correlates with disease activity, sebum excretion is significantly greater in patients with acne, and attempts to pharmacologically reduce the SER are logical in the treatment of acne. Recently, antiandrogen therapy and retinoids have vastly improved the treatment of the more difficult cases of acne.
97.6 STUDY OF NORMAL SKIN Dikstein et al.31 collected a database of Sebumeter measurements in 150 healthy females and compared the quantitative Sebumeter results to a seborrhoea grading as performed by experienced cosmetologists. While a 70 to 80% agreement was present between measurements and gradings for the extremes of dry and oily skin, the correlation in the middle range was uncertain. However, the authors stress the importance of objective and reproducible measurements for the assessment of skin surface lipids because the skin of many consumers is misclassified by subjective evaluation, leading to faulty recommendations about use of cosmetic products. O’goshi et al.32 investigated stratum corneum functions of the lesional scalp skin of patients with alopecia areata and of patients with androgenetic alopecia, i.e., male pattern baldness. The data of the scalp were compared with those of the cheek and the flexor surface of the volar forearm. The values from the androgen-sensitive areas of the scalp and the cheek, which were comparable, were far higher than those from the
Optical Measurement of Sebum Excretion Using Opalescent Film Imprint: The Sebumeter®
volar forearm. Moreover, removal of the skin surface lipids led to a significant decrease in skin surface hydration.32
97.7 STUDY OF SKIN DISEASES Knutson33 has demonstrated ultrastructual changes in the pilosebaceous ducts of patients with and without acne. He examined unaffected skin of patients with acne compared with normal skin of healthy individuals. He noted that the pilosebaceous duct contents were more densely packed. Knutson also showed that the abnormal corneocytes contained lipid droplets not seen in controls, and probably of ductal rather than sebaceous origin. O’goshi et al.34 compared skin surface lipid values on the volar forearm and eight other sites in healthy volunteers with values in patients with atopic dry skin and patients with atopic dermatitis. Casual lipid levels were high in the seborrheic region, such as on the cheek, where they amounted to 117 AU with the Sebumeter, being followed by the front neck and nape. They far exceeded the values on nonseborrheic areas, particularly the forearms. They tended to be lower in the nonlesional and lesional skin of the atopic dermatitis patients than in the skin of the healthy controls.34
97.8 STUDY OF PRODUCT EFFICACY Huschka and Schulewsky35 used the Sebumeter to compare shampoos with regard to their effect on seborrhoea. They studied the scalps and foreheads of 20 volunteers and found they could show a sebostatic effect of a shampoo containing pyrithiondisulfide, in contrast to increased seborrhoea after octopirox treatment. The authors stressed the advantages of the Sebumeter sebum collection head for measuring skin surface lipids in the scalp area without having to clip the hair. Serup36 used the Sebumeter to compare skin surface lipids resulting from some moisturizers’ lipids. He showed that measurement of skin surface lipids closely reflected the lipid content of the moisturizers. There was no clear relationship between lipid content and efficacy, in contrast to that regarding urea content.
REFERENCES 1. Bahmer FA. Morphometry in clinical dermatology. Acta Derm Venereol (Stockh) 72, 52, 1992. 2. Kvorning SA. Investigations into the pharmacology of skin rate and ointments. The collection and quantitative determinations of lipids on the skin. Acta Pharmacol 5, 248, 1949. 3. Emanuel S. Quantitative determinations of the sebaceous glands’ function with particular mention of the method used. Acta Derm Venereol (Stockh) 17, 444, 1936.
845
4. Ruggieri MR, McGinley KJ, Leyden JJ. Reproducibility and precision in the quantitation of skin surface lipid by TLC. In Advances in Thin Layer Chromatography. Touchstone JC, Ed. John Wiley & Sons, New York, 1982, p. 249. 5. Greene RS, Downing DT, Pochi PE. Anatomical variation in the amount of composition of human skin surface lipid. J Invest Dermatol 54, 240, 1970. 6. Strauss JS, Pochi PE. The quantitative gravimetric determination of sebum production. J Invest Dermatol 36, 293, 1961. 7. Cunliffe WJ, Shuster S. The rate of sebum excretion in man. Br J Dermatol 81, 697, 1960. 8. Lookingbill DP, Cunliffe WJ. A direct gravimetric technique for measuring sebum excretion rate. Br J Dermatol 114, 75, 1986. 9. Collison DW, Burns TL, Steward ME, Downing DT, Strauss JS. Evaluation of a method for measuring the sustainable rate of sebaceous was ester secretion. Arch Dermatol Res 278/279, 266, 1987. 10. Schäfer H, Kuhn-Bussius H. Methodik zur quantitativen Bestimminug menshlihen Talgsek. Archiv Klinische Experimantelle Dermstologie 238, 429, 1970. 11. Saint-Léger D, Berrebi C, Duboz C, Agache P. The Lipometer: an easy tool for rapid quantitation of skin surface lipids (SSL) in man. Arch Dermatol Res 265, 79, 1979. 12. Cunliffe WJ, Kearney JN, Simpson NB. A modified photometric technique for measuring sebum excretion rate. J Invest Dermatol 75, 394, 1980. 13. Saint-Léger D, Lévêque JL. A comparative study of refatting kinetics on the scalp and forehead. Br J Dermatol 106, 669, 1982. 14. Saint-Léger D, Cohen E. Practical study of the qualitative and quantitative sebum excretion on the human forehead. Br J Dermatol 113, 551, 1985. 15. Thune P, Gustavsen T. Comparison of skin surface lipids. Acta Dermatol Venereol Stockh, Suppl 134, 30, 1987. 16. Dikstein S, Zlotogorski A, Avriel E, Katz M, Harms M. Comparison of the Sebumeter and the Lipometre. Bioeng Skin 3, 197, 1987. 17. Sarkany I, Gaylarde P. A method for demonstration of the distribution of sebum on the skin surface. Br J Dermatol 80, 744, 1968. 18. Nordstrom KM, Schmus HG, McGinley KJ, Leyden JJ. Measurement of sebum output using a lipid absorbent tape. J Invest Dermatol 87, 206, 1986. 19. Piérard GE. Follicule to follicule heterogeneity of sebum excretion. Dermatologica 173, 61, 1986. 20. Serup J, Winther A, Blichmann C. A simple method for the study of scale pattern and effects of a moisturizer: qualitative and quantitative evaluation by D-Squame® tape compared with parameters of epidermal hydration. Clin Exp Dermatol 14, 277, 1989. 21. Markey AC. Commercially available tape for assessing ‘skin type.’ Br J Dermatol 119, 271, 1988.
846
Handbook of Non-Invasive Methods and the Skin, Second Edition
22. Blume U, Ferracin J, Verschoore M, Czernielewski JM, Schaefer H. Physiology of the vellus hair follicule: hair growth and sebum excretion. Br J Dermatol 124, 21, 1991. 23. Lévêque JL, Pierard-Frachimont C, de Rigal J, SaintLéger D, Piérard A. Effect of topical corticosteroids on human sebum production assessed by two different methods. Arch Dermatol Res 283, 372, 1991. 24. Burton JL, Cunliffe WJ, Shuster S. Circadian rhythm in sebum excretion. Br J Dermatol 83, 650, 1970. 25. Kligman AM. The uses of sebum. In The Sebaceous Glands, Montagna W, Ellis RA, Silver AF, Eds. Pergamon Press, Oxford, 1963, p. 110. 26. Rode B, Ivens U, Serup J. Decreasing method for the seborrheic areas with respect to regaining sebum excretion rate to casual level. Skin Res Technol 6, 92, 2000. 27. Cunliffe WJ, Shuster S. The rate of sebum excretion in man. Br J Dermatol 81, 697, 1969. 28. Zeller K, Huber H. Sebmetrische Messungen des “Casual Level” der Hautoberflächenlipide bei einen studentischen Kollektiv hautgesunder Probanden. Akt Dermatol 9, 101, 1983. 29. Cunliffe WJ, Burton JL, Shuster S. Effect of local temperature variations on the sebum excretion. Br J Dermatol 83, 650, 1970.
30. Cunliffe WJ, Perera WDH, Thackray P. Pilosebaceous physiology. Br J Dermatol 95, 153, 1976. 31. Dikstein S, Harzshtark A, Bercovici R, Courage W. Verteilung von Talgs piegelmessungen bei gesunden erwachsenen Frauen. Arztl Kosmetol 14, 41, 1985. 32. O’goshi K, Okada M, Iguchi M, Tagami H. Functional analysis of the stratum corneum of scalp skin: studies in patients with alopecia areata and androgenetic alopecia. Arch Dermatol Res 292, 605, 2000. 33. Knutson DD. Ultrastructual observations in acne vulgaris. The normal sebaceous follicle and acne. J Invest Dermatol 62, 288, 1974. 34. O’goshi K, Okada M, Iguchi M, Tagami H. The predilection sites for chronic atopic dermatitis do not show any special functional uniqueness of the stratum corneum. Exog Dermatol 1, 195, 2002. 35. Huschka U, Schulewsky A. Hauttalgsekretion und Haarshampoos. Arztl Kosmetol 14, 115, 1984. 36. Serup J. A three-hour test for rapid comparison of effects of moisturizers and active constituents (urea). Arch Derm Venereol (Stockh), Suppl. 177, 29, 1992.
Technique for Measuring 98 Gravimetric Sebum Excretion Rate (SER) W.J. Cunliffe and J.P. Taylor Leeds Foundation for Dermatological Research, Leeds General Infirmary, Leeds, United Kingdom
CONTENTS 98.1 Introduction............................................................................................................................................................847 98.2 Practical Details .....................................................................................................................................................847 98.2.1 Calculation of SER....................................................................................................................................850 98.3 Factors Affecting the Measurement of SER .........................................................................................................850 98.4 Conclusion .............................................................................................................................................................851 References .......................................................................................................................................................................851
98.1 INTRODUCTION
98.2 PRACTICAL DETAILS
The sebaceous glands produce sebum by a holocrine process, in which the cells synthesize lipid as they move toward the center of the gland and eventually disintegrate. The lipid cell contents are then discharged via the sebaceous duct into the pilosebaceous follicle, when they are excreted to the skin surface as sebum. Autoradiographic studies have shown that it takes 2 or 3 weeks for a tritiumlabeled cell to travel from the basal layer of the sebaceous gland to the sebaceous duct, and a further week or so for the sebum to reach the skin surface.1–3 There is, in addition, a large reservoir of sebum in the pilosebaceous orifice, so that the rate of delivery of sebum to the skin surface may not accurately reflect metabolic events in the glands themselves. Nevertheless, changes in sebaceous gland size or mitotic activity of the sebaceous basal layer will eventually affect the sebum excretion rate. Normally this remains reasonably constant in a given subject, and therefore the rates in individual subjects can be compared, and changes, as a result of experimental procedures, can be measured. The most reliable method of assessing sebum excretion is a simple gravimetric method in which sebum is collected onto absorbent papers held in contact with an area of cleaned forehead skin.4 This area is delineated with adhesive tape, and an elastic headband is used to hold the papers in place. After a timed collection period, the lipid is either extracted with ether into a preweighed container, which is again weighed after evaporation of the ether, or simply weighed more directly by pre- and postweighing the collecting papers.
The subsequent sections outline the techniques in detail. 1. Subjects should wash their hair on the evening prior to the test and apply no makeup or other topical application thereafter, but the face should be washed with soap in the normal way on the morning of the test. 2. The samples should, where possible, be collected at the same time of the day, since there may be circadian variation in both rate of sebum excretion and its composition.5,6 3. The room should be ventilated and should not be too hot because sweat reduces the capacity of the papers to absorb lipids. For the same reason, the subjects should not be allowed to take hot drinks during the test. Environmental temperature can have a marked effect on the apparent sebum excretion rate,7 probably by its effect on sebum viscosity,8 but a room temperature of about 20 to 24˚C is satisfactory. 4. The type of absorbent paper used is critical, because there is considerable variation in the amount of sebum absorbed by different papers, and even between different batches of paper from the same maker. The paper originally used by British workers is no longer available, but a suitable alternative is the “special velin tissue, non-fluff”, obtainable from the General Papers & Box Company, Severn Road, Treforest 847
848
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 98.1 Area of collection, delineated by tape.
FIGURE 98.3 Papers covered with gauze.
FIGURE 98.2 Absorbent papers on the forehead.
FIGURE 98.4 Papers and gauze held in position with an elasticated band.
Industrial Estate, Pontypridd, CF37 5SP England (telephone GB 0443–841977).9 5. The papers must be degreased prior to use by immersing them in Analar ether for three 10minute washes using fresh ether each time. A small, but progressive decrease in absorbency is found if the papers are immersed for more than 6 hours.10 Thereafter, the paper must be handled with forceps, and it can be stored conveniently in ether-cleansed tin foil. Some types of tin foil treated with ether readily shed fragments of the shiny surface, which stick to the papers, so it is best to use the dull surface. 6. The adhesive tape used to delineate the collection area on the forehead must be clean. Currently we use Durapore® surgical tape and prepare it as a template, which ensures that the area of collection consists of two perfectly shaped rectangles (Figure 98.1). The template is made such that collections can be made simultaneously from both sides of the forehead.
7. The skin must be prepared, prior to the timed collection, by the serial application of preliminary absorbent papers until the accumulated surface lipid has been removed and only the follicular pinpoint pattern of freshly excreted sebum is visible. Five sheets of absorbent papers, cut to the length of the forehead (Figure 98.2), are applied and covered with gauze (Figure 98.3); then a broad elastic band (50 to 60 cm long) (Figure 98.4) is applied and is attached at the back of the head with nylon mesh (Figure 98.5); the sheet of paper in contact with skin is removed at 10-min intervals (three times) to remove the variable amount of surface lipid collected. Novice sebum collectors usually underestimate the time required for this preliminary procedure, and the use of Sebutape® to view the follicular patterns (Figure 98.6) will often show that the papers are still heavily lipid laden. A correct pattern is
Gravimetric Technique for Measuring Sebum Excretion Rate (SER)
FIGURE 98.5 The band is secured at the back of the head with nylon mesh.
FIGURE 98.6 Pattern of sebum excretion using Sebutape® in a normal subject.
shown in Figure 98.6. The time for the skin preparation varies for each subject, but most will require three or four removals of paper at 10-min intervals. It is now recommended that the papers are held up to the light to simply visualize the follicular pattern, which is seen as small dots, each dot representing sebum emerging. 8. Once prepared as above, the timed sebum collection can begin. This is done by placing ethercleansed papers over the area marked out by the template (Figure 98.7) and replacing the elastic headband after lining with a single sheet of paper and a piece of gauze. The investigator should check the position of the collection papers periodically throughout the test, as many subjects inadvertently move their elastic headbands. 9. The collection time should be standardized to about 3 hours and the time noted precisely. The
849
FIGURE 98.7 Final absorbent papers being placed on the forehead with forceps.
hourly output of collected sebum decreases with time, and it could be argued that the greater rate in the early hours is spurious and should be disregarded (Figure 98.8). The sebum collected during the first and subsequent 3-hour collection periods is linearly related, however, and the first collection appears to produce a valid index of sebaceous activity (Figure 98.9).10 This observation emphasizes that the method does not directly measure the rate of secretion from the sebaceous glands to the follicular reservoir. 10. The routine for extracting and weighing lipid is as follows: (1) The papers are transferred to beakers; sebum lipid is extracted with 3 × 25 ml of analar diethyl ether and transferred into large tared flasks; the ether is removed by rotary evaporation (300 mmHg at 30˚C), leaving a small volume that is transferred to preweighed 5-ml flasks, which are evaporated to dryness and reweighed. (2) Since the weight of lipid is very small, the flasks are handled with ethercleaned forceps and kept in a desiccator for at least 24 hours before being weighed in a microbalance accurate to 0.01 mg. 11. A good weighing technique is vital. The balance must be accurate to 10 mg (e.g., Mettler M5), and it should preferably be kept in a draftfree balance room built with a concrete floor over a main structural crossbeam of the building. Room temperature should be constant, and humidity inside the balance can be controlled by the use of silica gel. The effects of static electricity can be reduced by the use of aluminum cups in place of glass. One further development is the so-called direct gravimetric technique, in which the papers placed on the
850
Handbook of Non-Invasive Methods and the Skin, Second Edition
forehead are preweighed (Figure 98.7). After the collection period the papers are reweighed. This technique is as accurate as the original method of Strauss and Pochi and eliminates the extraction procedure (Figure 98.10). The methodology and attention to detail are therefore similar to the original gravimetric technique, but without the extraction technique.
SER (mg/cm2/min)
1.5
98.2.1 CALCULATION
1.0
SER
The SER is determined by the following formula: weight of sebum ( μg )
(
area cm 2
0.5
1
3
2 Time (hours)
FIGURE 98.8 Decrease of sebum excreted during a 3-hour collection period.
2.0
1st 3 hour collection (mg/cm2/min)
OF
1.5
1.0
)
× time ( min )
Normal values for the sebum excretion rate by this method for subjects of different ages and sex have been given by Pochi and Strauss,12 but obviously each worker must obtain his own normal range. It should be noted that the experimental error of the method is proportionally larger with low sebum excretion rates, and the variance of the method should therefore be determined at each end of the range. Although the sebum excretion rate is usually expressed in terms of the area of skin studied, it can also be related to the number of functioning glands by applying the replica technique,13 in which a hardened plaster-ofParis (dental grade) replica of the skin surface is stained with 0.25% osmium tetroxide. This allows the number of patent follicular orifices to be counted, and it also produces a permanent record of their position in relation to other microscopic features of the skin surface. Cunliffe and Cotterill14 have described an alternative technique in which a red dye (saturated Oil Red O) is used to stain the follicular orifices, followed by direct microphotography of the skin surface.
r = 0.94
98.3 FACTORS AFFECTING THE MEASUREMENT OF SER The SER is influenced by many factors that are relevant to its measurement in practice:15
0.5
0.5 1.0 2nd 3 hour collection (mg/cm2/min)
1.5
FIGURE 98.9 Correlation between sebum collected in two consecutive 3-hour periods.
1. Age, sex, and endocrine status. Age, sex, and endocrine status all affect SER. Some oral contraceptives inflame SER a little, with the exception of anti-androgen-containing pills, which can reduce SER by 30%. The effect of the menstrual cycle is probably of minor importance to clinical studies of SER. 2. Duration of collection. Duration of collection affects the apparent SER, and variance of the
Gravimetric Technique for Measuring Sebum Excretion Rate (SER)
851
12 11 10 9
Direct paper weights (mg)
8 7 6 5 4 3 2 1 0 0
1
2
3 4 5 6 7 Ether extraction weights (mg)
8
9
10
FIGURE 98.10 Sebum weights determined by direct vs. ether extraction techniques •–•, results from 43 normal subjects (r = 0.991, p < 0.001). –, results from 32 premeasured sebum standards (r = 0.997, p < 0.001).
3. 4.
5.
6.
7.
measurements is most marked in the first and second hours. Skin temperature. A 1˚C change in skin temperature produces a 10% change in SER.7 Sweat gland function. Sweating reduces the amount of lipid collected by absorbent papers, and a thermally neutral environment is necessary during measurement of SER. If the patient sweats obviously during the procedure, the paper technique cannot be used. This event is very uncommon in the U.K. Diurnal variation. SER is maximal in the midmorning and minimal during the late evening and early hours of the morning.5,6 Site variation. Sebaceous glands are found predominantly on the scalp, face, back, and chest, and the SER on the forehead is five times that on the back, but when related to the number of glands at each site, there is no difference.16 Disease. SER correlates with acne severity and past history of the disease.17 Changes also occur with endocrine disease, Parkinsonism, and certain drugs.15
98.4 CONCLUSION Thus, in measuring SER, it is desirable to make collections at the same time of the day, in a room of constant ambient temperature, preferably at comparable stages of the menstrual cycle in females, and to stop all relevant therapy 4 to 6 weeks before the measurement. Since the response of the sebaceous gland is slow, such assessments should be after at least 4 weeks’ treatment. As results vary with details of technique (e.g., duration of collection, absorbency of papers), changes in SER are more reliable than absolute rates in comparing results from different laboratories. The method of choice will possibly depend on the requirement and local facilities. The techniques are easy to learn, and a visit to the nearest sebum center will possibly pay dividends.
REFERENCES 1. Epstein, E.H. and Epstein, W.L., New cell formation in the human sebaceous gland, J. Invest. Dermatol., 46, 453, 1966. 2. Weinstein, G.D., Cell kinetics of human sebaceous glands, J. Invest. Dermatol., 62, 144, 1974.
852
Handbook of Non-Invasive Methods and the Skin, Second Edition
3. Plewig, G. and Christopher, E., Renewal rates of human sebaceous glands, Acta Dermato-Venereol., 54, 177, 1974. 4. Strauss, J.S. and Pochi, P.E., The quantitative gravimetric determination of sebum production, J. Invest. Dermatol., 36, 293, 1961. 5. Burton, J.L., Cunliffe, W.J., and Shuster, S., Circadian rhythm in sebum excretion, Br. J. Dermatol., 82, 497, 1970. 6. Cotterill, J.A., Cunliffe, W.J., and Williamson, B., Variations in skin surface lipid composition and sebum excretion rate with time, Acta Dermato-Venereol., 53, 271, 1973. 7. Cunliffe, W.J., Burton, J.L., and Shuster, S., The effect of local temperature variation on the sebum excretion rate, Br. J. Dermatol., 83, 650, 1970. 8. Burton, J.L., The physical properties of sebum in acne vulgaris, Clin. Sci., 39, 757, 1970. 9. Cunliffe, W.J., Williams, S.M., and Tan, S.G., Sebum excretion rate investigations: a new absorbent paper, Br. J. Dermatol., 93, 347, 1975.
10. Cunliffe, W.J. and Shuster, S., The rate of sebum excretion in man, Br. J. Dermatol., 81, 691, 1969. 11. Lookingbill, D.P. and Cunliffe, W.J., A direct gravimetric technique for measuring sebum excretion rate, Br. J. Dermatol., 114, 75, 1986. 12. Pochi, P.E. and Strauss, J.S.J., The effect of ageing on the activity of sebaceous glands in man, in Advances in Biology of Skin, Vol. 6, Montagna, W., Ed., Pergamom Press, New York, 1965, p. 121. 13. Sarkany, I. and Gaylarde, P., A method for demonstration of the distribution of sebum on the skin surface, Br. J. Dermatol., 80, 744, 1968. 14. Cunliffe, W.J. and Cotterill, J.A., in The Acnes: Clinical Features, Pathogenesis and Treatment, W.B. Saunders, London, 1975, p. 293. 15. Cunliffe, W.J., Acne, Martin Dunitz, London, 1989. 16. Cunliffe, W.J., Perera, W.D.H., Thackray, P., et al., Pilosebaceous duct physiology, Br. J. Dermatol., 95, 153, 1976. 17. Cunliffe, W.J. and Shuster, S., The pathogenesis of acne, Lancet, 1, 685, 1969.
Photography of 99 Fluorescence Sebaceous Follicles Andreas Herpens, S. Schagen, and S. Scheede Research Bioengineering — Biophysics, Beiersdorf AG, Hamburg, Germany
CONTENTS 99.1 Introduction............................................................................................................................................................853 99.2 Fluorescence Evaluation of Follicles ....................................................................................................................853 99.3 Origin of Fluorescence in Sebaceous Follicles.....................................................................................................854 99.4 Sample and HPLC Preparations and Results........................................................................................................855 99.5 The SAFIR Fluorescence Imaging System...........................................................................................................856 99.6 Image Analysis and Neural Algorithms ................................................................................................................857 99.7 Results on Comedogenic and Antibacterial Actions.............................................................................................858 99.8 Recommendations..................................................................................................................................................859 Acknowledgments ...........................................................................................................................................................860 References .......................................................................................................................................................................860
99.1 INTRODUCTION Human skin of sebum-rich areas like the forehead or back shows a very dense distribution of vellus hair follicles of open and closed types. Size and reservoir of the open follicular infundibulae are closely connected to their activity in sebum excretion. It is common dermatological knowledge that increased size and excretion activity of the follicles strongly correlates with signs of impure skin and acne. Besides the inflammatory effect, excessive sebum excretion is one of the major cosmetic problems of young adolescents and adults. To proof the efficacy of medications on acne and impure skin, or to ensure the noncomedogenic safety of cosmetic products, a welldefined and sensitive assessment of the change in follicle size and content was necessary. This chapter gives an update on the development of fluorescence photography as well as a description of a new sensitive and precise in vivo method to characterize follicular changes after topical product application.
99.2 FLUORESCENCE EVALUATION OF FOLLICLES The first observations of the sebaceous follicle fluorescence go back to as early as 1927, when the German dermatologist Bommer1 was looking at his patients’ faces irradiated with Woods UVA light peaking at 400 nm, and
visually recognized orange-red dots in the center of the pores, which he could not explain at that time. Forty years later Cornelius and Ludwig2 identified bacterial protoporphyrin as the main source of the red fluorescence in human sebaceous follicles. In 1982 Melö and Johnsson3 published spectra of cultivated propionibacteria, which showed light emissions at wavelengths of 620 and 680 nm, corresponding to orange and red color. Only a small peak at 580 nm contributed to the visual observation of green-yellow color. Porphyrins are powerful photosensitizers, which cause O2-dependent type II photoreactions, while themselves being partially destroyed.4 This is the proposed mechanism of the UVA-mediated deactivation of propionibacteria.5 In 1990 Sauermann et al.6 showed by fluorescence spectroscopy that the follicular fluorescence is mainly caused by porphyrins and, to a lesser extent, by other fluorophors. Furthermore, they showed that almost all conical structures of cyanoacrylate biopsies are fluorescent, some of them at their distal ends only. They also proved that the number and size of fluorescent glands, quantified by image analysis of biopsies, were influenced by irradiation, blood flow, seasons, cleansing, and cosmetic treatment (Figure 99.1). In the mean time, Fulton et al.7 and Mills and Kligman8 measured the so-called comedogenic effects of certain cosmetic ingredients either by employing the rabbit ear model or by image analysis of cyanoacrylate biopsies taken from human skin. In 1993 Sauermann et al.9 were able to detect the comedogenic as 853
854
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 99.2 Fluorescence photography of acne skin. (From Lucchina, L. et al., J. Am. Acad. Dermatol., 35, 58–63, 1996.)
FIGURE 99.1 Wood’s light image of a subject’s nose with UVsensitive film. (From Sauermann, G. et al., J. Toxicol. Cut. Ocular Toxicol., 8, 369–385, 1990.)
well as the comedolytic effects of topical treatments, when they measured the changes of the protoporphyrin fluorescence of sebaceous origin in vivo by employing a blackand-white Hamamatsu sensitive camera system with an intense UVA excitation source, special filter equipment, and computer-assisted image analysis. The whole system did not get into contact with the subjects’ skin and displayed the fluorescence changes in vivo. The camera and filter system was designed to detect the red-orange porphyrin fluorescence, which needed a suitable panel of volunteers with a marked porphyrin fluorescence on the back or forehead. Unlike follicular biopsies, where all follicles could be measured, only one third of the follicles displayed red fluorescence in vivo and would be utilized for further analysis. Lucchina et al.10 was one of the first who published further studies with real color fluorescence photography (Figure 99.2). They used a xenon flashlight combined with a filter and an analog camera system. Lucchina et al. note that the fluorescence is located prominently in open follicles and, to a lesser extent, in closed and inflamed follicles, but that fluorescence photography appears to be a useful tool to evaluate the efficacy of an acne treatment. In 1997 Pagnoni et al.11 introduced digital fluorescence photography for the assessment of the effects of antibacterial agents on acne. They were able to demonstrate a strong decrease of red bacterial porphyrin
FIGURE 99.3 Digital fluorescence photography of the effect of benzoyl peroxide for 7 days and recovery of protoporphyrin fluorescence after 16 days. (From Pagnoni, A. et al., J. Am. Acad. Dermatol., 41, 710–716, 1999.)
fluorescence during treatment with benzoyl peroxide (Figure 99.3), whereas a greenish fluorescence of the follicles remained unaltered.
99.3 ORIGIN OF FLUORESCENCE IN SEBACEOUS FOLLICLES Our objective was to develop a new sensitive in vivo camera system for assessment of the red and yellow-green fluorescence of the follicles and to identify the corresponding emitters. For chemical analysis of fluorescent substances by high-pressure liquid chromatography (HPLC) or gas chromatography (GC), it is useful to have a rough idea of the relevant components and to start with some known substances, which show fluorescence characteristics similar to those found in the follicles. From our own studies and literature, we knew that the excitation maximum for the yellow-green fluorescence is at 385 nm and the maximum of emission is between 500 and 600 nm. Unfortunately,
Fluorescence Photography of Sebaceous Follicles
855
and with its absorption maximum at 435 nm (Figure 99.7) and a fluorescence emission at 500 nm, it potentially contributes to the fluorescence of the sebaceous gland. In order to quantify the source of the orange-red fluorescence, we also established a HPLC procedure for protoporphyrin IX (PPIX) analysis. PPIX (Figure 99.8) is a powerful photosensitizer with a high quantum yield and specific fluorescence characteristics. The main excitation maximum lies between 385 and 415 nm, and it also shows a characteristic double-emission peak at 636 and 705 nm (Figure 99.9).
Molar extinction coefficient
50,000 40,000 30,000 20,000 10,000 0 200
400
600
800
Wavelength (nm)
99.4 SAMPLE AND HPLC PREPARATIONS AND RESULTS
FIGURE 99.4 Riboflavin absorption spectra. (From Oregon Medical Laser Centre.)
Follicles were extracted from the subjects’ faces by the two-finger squeeze method and were visualized under the fluorescence microscope (Figure 99.10). After we had looked at a variety of follicles, we discovered that most of them more or less showed both the green and the red color. So we decided to collect the extracted follicles from about 30 volunteers and dissolved them in tetrahydrofurane (THF) with three drops of sodium hydroxide as a base. The solution was treated with ultrasound for 5 min to dissolve all the fluorophors from the sebum. Then the resulting solution was filtered through a 0.2-μm filter to eliminate insoluble compounds, mostly lipids, before the final analysis. Fluorophors were detected by a fluorescence detector. The eluation protocol was established for PPIX, bilirubin, and riboflavin (Table 99.1). The collected comedonal material of 0.32 mg/ml from about 30 subjects contained about 1 ng/ml PPIX. Bilirubin was detected neither by HPLC nor by the more sensitive liquid chromatography with mass selective detector (LCMSD). Bilirubin is easily oxidized by light or other oxidants. Hence, it has yet to be clarified whether its apparent absence might be due to its decomposition during isolation or analysis. Riboflavin and flavins are essential components of the mitochondrial energy metabolism, and as such, may survive to the process of holocrine secretion of cellular contents, which occurs in the sebaceous gland. According to our results, the follicles of our test panel contained about
Logic of fluorescence emission
7.5 7 6.5 6 5.5 5 400 500 600 700 800 Wavelength (nm) (excitation = 450 nm, quantum yield = 0.30)
FIGURE 99.5 Riboflavin emission fluorescence. (From Oregon Medical Laser Centre.)
there are a lot of natural fluorophors that match our criteria. However, König’s and Schneckenburger’s12 systematic work on human skin’s autofluorescence showed similar absorption spectra of oxidized flavin mononucleotide (FMN), flavin adenine dinucleotide (FAD), and riboflavin (Figure 99.4 to Figure 99.5, respectively). Fluorescence of these compounds peaks at 530 nm, but quantum yield of FAD is less than 1/10, compared to FMN and riboflavin. Taking this into account, riboflavin was the first choice as a reference for our HPLC analysis. We also included bilirubin (Figure 99.6) as a standard in our HPLC search. Bilirubin is a degradation product of heme and porphyrin,
HO
M
V
M
P
P
M
M
V
C
C
C
C
C
C
C
C
C
C N
C H
FIGURE 99.6 Bilirubin molecular structure.
C
C N H
C H2
C H
C N
C H
C
C N
OH
856
Handbook of Non-Invasive Methods and the Skin, Second Edition
60,000
636 nm
Normalized fluorescence
Molar extinction coefficient
1
40,000
20,000
0 200
300
400
500
600
700
Wavelength (nm)
FIGURE 99.7 Bilirubin absorption. (From Oregon Medical Laser Center.)
M
λexc = 407 nm
705 nm
0
600
650 700 Emission wavelength (nm)
750
FIGURE 99.9 PPIX fluorescence spectra. (From Center for Biomedical Optics and New Laser Systems.)
V
N M
M N
N
P
V N
FIGURE 99.10 Sebaceous follicle content under fluorescence microscope, 50×. P
M
FIGURE 99.8 PPIX molecular structure.
10 ng/ml riboflavin in 0.32 mg/ml comedonal material from about 30 subjects. At first glance, this relatively high concentration reflects the intense green fluorescence, which is 5 to 10 times more efficient in its in vivo visibility and might also be influenced by absorption and reabsorption processes and different quantum yields. However, it should be kept in mind that the intense fluorescence could also be produced by high amounts of FAD or FMN. This implies that an in vivo follicular spectroscopy with the capability to assess both green and red fluorescence in parallel would have substantial benefits. The green fluorescence, which we attribute to flavins, would indicate sebaceous gland activity, and the red fluorescence, whose source is bacterial porphyrins, would give an indication of the microbiological status of the sebaceous duct.
99.5 THE SAFIR FLUORESCENCE IMAGING SYSTEM Two very different problems motivated us to develop a new system for the detection of follicles and comedones in human skin. On one hand, the established system13 could only detect the red fluorescence of PPIX. On the other hand, movements of the volunteers during the measurement often caused blurred images. The new Skin Analyzing Fluorescence Imaging Recorder (SAFIR) (Table 99.2, Figure 99.11, and Figure 99.12) is based on the combination of a high-resolution digital camera and a UV flashlight system. Both are connected by a special filter system that enables the combination of the optical pathways of excitation and emission to avoid spectral distortions. The lens focus adapter can be pressed directly on the skin surface to get sharp and
Fluorescence Photography of Sebaceous Follicles
857
TABLE 99.1 HPLC Eluation Protocols Time (min)
Bilirubin/Protoporphyrin (methanol/hydrochloric acid–citric buffer)
Riboflavin (methanol/hydrochloric acid–citric buffer)
0 15 20 30
65/35 100/0 100/0 65/35
20/80 100/0 100/0 20/80
TABLE 99.2 Technical Data of SAFIR Color digital camera UVA flashlight Fluorescence emission Fluorescence excitation Image analysis
Skin area size
Kappa (1200 × 1600 px), 8ms illumination time Rapp optoelectronic flash, UVA < 1 mW/cm2 450 to 750 nm 385 ± 11 nm ZEISS KS400 and neural image matching software — green and red segmentation for mean size and total count of fluorescent follicles 20 × 30 mm
CCD UVA Flash
Objective
Filter
FIGURE 99.12 Scheme and image of the new SAFIR system.
The resulting images are stored on a PC hard disc for further image analysis. After an automatic matching process the corresponding areas of the images are analyzed for intensity and number of spots with red porphyrin and green riboflavin fluorescence, respectively. The resulting data are subject to further statistical evaluation.
99.6 IMAGE ANALYSIS AND NEURAL ALGORITHMS
Skin surface
FIGURE 99.11 Scheme and image of the new SAFIR system.
intense fluorescence images of the sebaceous follicles of the forehead or back.
Image analysis is based on two main steps, the matching of images subsequently taken from a test site to find a common subsite and the data extraction from the subsite. When fluorescence images are taken at the beginning and the end of a 6-week application period, it is very difficult to mark the skin areas of 20 × 30 mm exactly. It is even more difficult to hit these areas exactly for imaging. We experienced rotational and planar shifts within the images due to very small movements of the volunteers and camera system.
858
Handbook of Non-Invasive Methods and the Skin, Second Edition
100%
97%
99%
99%
87%
82%
85%
87%
78%
88%
92%
88%
94%
89%
90%
FIGURE 99.13 Time-based shift of 15 measurements and correction by the neural software.
Together with the university, we designed a program to match the time-linked images based on neural algorithms, employing a multiple-step nearest-neighborhood algorithm. The error of the new matching program was tested by comparison of a series of images taken twice daily over 7 days. The standard deviation of the evaluated parameters for this nontreated area was between 10 and 15%, which is mainly due to aberrations and image processing errors. The mean displacement of the identified areas was about 7%. Figure 99.13 demonstrates how planar and rotational shifts of 15 subsequently taken images of the same skin site are aligned by the software, which then determines the corresponding subsites. The subsite present in all images of this example covers 78% of the original site and includes 130 follicles, which can be taken for evaluation of changes. Data extraction from the common subsite is done by a program based on ZEISS KS400 software. The original image is transformed in gray, and shading is corrected before the final dynamic segmentation is done either for yellow-green or for orange-red fluorescence, which results in six different parameters described in Table 99.3.
TABLE 99.3 Parameters for Statistical Analysis AREA_red
=
COUNT_red AREA_green
= =
COUNT_green
=
AREA_all COUNT_all
= =
Mean area of the orange-red porphyrin follicles Number of orange-red porphyrin follicles Mean area of the yellow-green riboflavin follicles Number of yellow-green riboflavin follicles Mean area of all fluorescent follicles Number of all fluorescent follicles
99.7 RESULTS ON COMEDOGENIC AND ANTIBACTERIAL ACTIONS In a recent study, we investigated the influence of several ingredients known to have an impact on human sebaceous follicles: benzoyl peroxide, salicylic acid, lactic acid, acetylated lanoline, and isopropyl palmitate. Concentrations are summarized in Table 99.4. Benzoyl peroxide and lactic acid were applied in a cream base, whereas salicylic acid was applied as an alcoholic solution. Two milligrams per square centimeter of each product was applied twice daily to the skin of the upper back for 6 weeks by trained staff. Benzoyl peroxide (BPO) strongly acts against skin bacteria11 and oxidizes PPIX. These initial effects of BPO might explain the enhancement of the green fluorescence (Figure 99.15b) after 3 weeks of continuous treatment. However, it drops again to a normal level after another 3 weeks (Figure 99.15c). Moreover, even the shrinking size of the microcomedones indicates a reduction to normal. Interestingly, the whole skin loses its reddish color and changes to a darker green in the end. As expected, bacterial porphyrin fluorescence increased after treatment with acetylated lanolin. However, some of our results, at least in part, contradict previously published data about comedogenicity. As expected, isopropyl palmitate (IPP) and acetylated lanolin showed a comedogenic effect on the size of the follicles, as measured by an increase of the green sebum fluorescence. Salicylic and lactic acid only moderately reduced the red fluorescence of bacterial origin, but on the other hand, they slightly increased the green follicle fluorescence. In summary, in many cases green and red fluorescence change directions during treatment. Therefore, only the test of a complete cosmetic formula and analysis of both kinds of fluorescence will give a
Fluorescence Photography of Sebaceous Follicles
859
TABLE 99.4 Six-Week In Vivo Study: Tested Ingredients and Effect on Follicle Fluorescence Ingredient
% (w/w) In Product
Green Fluorescence (counts)
Red Fluorescence (counts)
Acetylated lanolin Isopropyl palmitate Benzoyl peroxide Salicylic acid Lactic acid
100 100 5 1 2
Strong increase (>50%) Moderate increase (>20%) No change/effect Moderate increase (>20%) Minor increase (<10%)
Moderate increase (>10%) Moderate decrease (<30%) Strong decrease (>80%) Moderate decrease (>30%) Minor decrease (<10%)
FIGURE 99.14 Image segmentation: (a) original, (b) gray transformed, and (c) segmented.
FIGURE 99.16 Effects of cyanoacrylate stripping on follicular fluorescence (a) before stripping and before matching, (b) before stripping and after matching, and (c) after stripping and after matching.
99.8 RECOMMENDATIONS
FIGURE 99.15 Effects of BPO (a) at the beginning (t0; red/green = 60/20), (b) 3 weeks (red/green = 0/96), and (c) 6 weeks (red/green = 0/30).
comprehensive answer about comedolytic, antibacterial, or comedogenic effects on the skin. Because of these contradictions of some of the results, more test results and experiences have to be collected in the future to get a better understanding of the processes in the human follicle during topical treatments. Figure 99.16 demonstrates a single experiment of the effect of cyanoacrylate stripping on follicular fluorescence. Despite this being well known as a powerful tool to extract follicular contents, not much can be seen on the images. However, careful analysis of images taken before and after stripping with different adhesive tapes, or even cyanoacrylate glue, revealed a superiority of nose pore strips, which reduced size and number of yellow-green spots to about 45 and 68% of their initial values, respectively. On the other hand, cyanoacrylate stripping and 10 successive strippings with adhesive tape only reduced the size/number of spots to 60%/81% and 86%/88%, respectively. These effects will be investigated in further studies to get a better understanding of pore deep-cleansing effects.
Fluorescence imaging of sebaceous follicles has become a valuable and widely used tool for the assessment of follicular changes due to topical treatment. Effects of new acne medications as well as the safety of new cosmetic products and ingredients, which should not harm the sensitive balance of the follicular keratinization process, can be investigated at an early stage with high accuracy and good prediction. The use of a sensitive digital camera system with a high optical resolution gives a precise view of the fluorescence phenomena of the sebaceous follicles. However, the limitations of this method lie in the fact that the fluorescence phenomena can be trapped by optical staining, such as the melanization after intense UV tanning or fluorescent markers, or by changing the molecular environment by pH or fluidity. These impacts have to be accounted for by additional controls, and the final result will then only depend on the parameters, follicular size and bacterial growth. Many critical comments in the past focused on the fact that our former system could only detect the bacterial action, and therefore did not characterize the form and size of the whole follicle. On the background of the facts shown before, we have to agree partially, because the porphyrin fluorescence is in some cases only located in a small section of the follicle with a low partial oxygen pressure (Figure 99.10). But with the new, more precise and complete fluorescence detection with SAFIR we succeeded in overcoming the critics.
860
Handbook of Non-Invasive Methods and the Skin, Second Edition
The first results of our new measuring technique already gave some new insights into the mechanism of comedogenicity. Moreover, Kligman14 already concluded that former animal and human models for the assessment of comedogenicity might have produced false positive results, and therefore have to be revised for the future. It is interesting to note that riboflavin seems to be a 5-alpha-reductase inhibitor.15 Consequently, the high levels of riboflavin in the follicular duct might serve as a feedback principle, which regulates the normal status of the sebaceous duct. Additionally, it has been reported that a lack of riboflavin and its coenzymes FMN and FAD may lead to an increased sebum excretion and can worsen the signs of acne. Finally, we summarize that the new SAFIR system is able to detect novel influences on the size and composition of the human microcomedo. It then helps us to identify new ingredients for safe and effective cosmetic products.
ACKNOWLEDGMENTS Special thanks go to the University of Magdeburg, Prof. Kruse, and his coworkers for the neural image tool.
REFERENCES 1. S. Bommer, Hautuntersuchungen im gefilterten Quarzlicht, Klin. Wochenschrift 6, 1142–1144, 1927. 2. E. Cornelius III and G.D. Ludwig, Red fluorescence of comedones: production of porpyrins by Corynebacterium acnes, J. Invest. Dermatol. 49, 368–370, 1967. 3. T.B. Melö and M. Johnsson, In-vivo porphyrin fluorescence from Propionibacterium acnes. A characterization of the fluorescence pigments, Dermatologica 164, 167–174, 1982. 4. S. Sanberg, J. Glette, G. Hopen, C.O. Solberg, and I. Romslo, Porphyrin-induced photo damage to isolated human neutrophiles, Photochem. Photobiol. 34, 471–475, 1981.
5. T.B. Melö, G. Reisaeter, and A. Johnsson, Photo destruction of Propionibacterium acnes porphyrins, Z. Naturforsch. 40, 125–128, 1985. 6. G. Sauermann, B. Ebens, and U. Hoppe, Analysis of facial comedos by porphyrin fluorescence and image analysis, J. Toxicol. Cut. Ocular Toxicol. 8, 369–385, 1990. 7. J.E. Fulton, Jr., Comedogenicity and irritancy of commonly used ingredients in skin care products, J. Soc. Cosmet. Chem. 40, 321–333, 1989. 8. O.H. Mills and A.M. Kligman, A human model for assessing comedogenic substances, Arch. Dermatol. 118, 903–905, 1982. 9. G. Sauermann, A. Herpens, and U. Hoppe, The quantification of comedogenesis and comedolysis by image analysis of fluorescent bacterial porphyrins, paper presented at the Regional Meeting of the International Society for Bioengineering and the Skin, April 23–25, 1993. 10. L. Lucchina et al., Fluorescence photography in the evaluation of acne, J. Am. Acad. Dermatol. 35, 58–63, 1996. 11. A. Pagnoni, A.M. Kligman, N. Kollias, S. Goldberg, and T. Stoudemayer, Digital fluorescence photography can assess the suppressive effect of benzol peroxide on Propionibacterium acnes, J. Am. Acad. Dermatol. 41, 710–716, 1999. 12. K. König and H. Schneckenburger, Laser-induced autofluorescence for medical diagnosis, J. Fluorescence 4, 17–40, 1994. 13. G. Sauermann, A. Herpens, U. Hoppe, and A.M. Kligman, A novel fluorometric method to investigate sebaceous glands in humans, in Noninvasive Methods for the Quantification of Skin Functions, P.J. Frosch and A.M. Kligman, Eds., Springer-Verlag, Heidelberg, 1999, pp. 252–271. 14. A.M. Kligman, Petrolatum is not comedogenic in rabbits or humans: a critical reappraisal of the rabbit ear assay and the concept of “acne cosmetica,” J. Soc. Cosmet. Chem. 47, 41–48 1996. 15. O. Nakayama et al., Riboflavin, a testosterone 5-alphareductase inhibitor, J. Antibiot (Tokyo) 43, 1615–1616, 1990.
for Assessment of 100 Methods Follicular Transport in Ex Vivo and In Vivo Jürgen Lademann Center of Experimental and Applied Cutaneous Physiology (CCP), Department of Dermatology, University Hospital Charité, Humboldt University, Berlin, Germany
CONTENTS 100.1 Introduction ..........................................................................................................................................................861 100.2 In Vivo Method for Assessment of Active and Inactive Follicles .......................................................................862 100.3 In Vivo Method for Assessment of Follicular Transport .....................................................................................864 100.4 Summary...............................................................................................................................................................865 References .......................................................................................................................................................................865
100.1 INTRODUCTION In principle there are three possible penetration pathways of topically applied substances through the skin: the intercellular penetration, the intracellular penetration, and the follicular penetration. In the past, the penetration processes were described as a diffusion through the lipid layers of the stratum corneum. It was presumed that the hair follicles play a dominating role in this process because the amount of hair follicles of the total skin surface was estimated at not more than 0.1%.1 However, several in vitro and in vivo studies on follicular penetration show a surprisingly high influence of this route on the penetration process. Feldman and Maibach2 and Maibach3 found regional variations of percutaneous absorption in different skin areas. They assumed that the density and size of hair follicles might be the reason for their findings. Tenjarla et al.4 investigated the penetration of steroids through human scar skin devoid of follicles and through normal chest skin containing follicles by Franz cell penetration measurements. Significant differences were found. The penetration rate was higher in the case of the skin containing follicles. Hueber et al.5 confirmed these results by comparing the penetration of steroids through follicle containing rat skin and the skin of newborn rats without any follicles. Essa et al.6 performed an in vitro Franz cell experiment for iontophoretic drug delivery. A stratum corneum/epidermis sandwich method was used for blocking the folli-
cular orifice. A five times lower absorption rate was found when the potential shunt routes were blocked. Turner and Guy7 found a significant iontophoretic drug delivery across the skin via follicular structures. Tur et al.8 demonstrated that both the velocity and the intensity of the vasodilatory effect of methyl nicotinate depend on the follicle density of different human skin areas. The evaluation of all these results needs an exact knowledge of hair follicle density, size of follicular orifices, follicular volume, and follicular surface. Follicle density was mainly measured for terminal hair follicles on the scalp. Blume et al.9 determined the hair follicle density on the forehead and on the back by phototrichogram. Seago and Ebling10 measured the hair follicle density on the upper arm and thigh using a classical trichogram. Pagnoni et al.11 used the cyanoacrylate technique for the measurement of follicle density in different regions of the face. Otberg et al.12 determined the follicular reservoir by measuring the follicle density, the size of follicular orifice, the amount of orifice of the skin surface, the hair shaft diameter, the volume, and the surface of the infundibula, in different regions of the body by using noninvasive cyanoacrylate skin surface biopsies and light microscopy. The highest infundibular volume, and therefore a potential follicular reservoir, was found for the forehead and for the sural region, although the sural region showed the lowest hair follicle density. The calculated follicular reservoir 861
862
Handbook of Non-Invasive Methods and the Skin, Second Edition
volume of these two skin areas was as high as the estimated reservoir of the stratum corneum. The study demonstrated that the amount of appendages of the total skin surface could be significantly more than 0.1%, as already proposed. Lademann et al.13 investigated the penetration of coated titanium dioxide microparticles into the stratum corneum of living human skin by tape stripping and biopsies in combination with spectroscopic measurements. Small amounts of microparticles were found in deeper parts of the stratum corneum after long-term application of a sunscreen containing titanium dioxide. These small amounts were clearly located in the follicle orifices, while the surrounding corneocyte aggregates were free of TiO2 in these parts of the horny layer. The analysis of biopsy sections containing hair follicle channels shows that small amounts of TiO2 microparticles penetrated into the acroinfundibulum of follicles without reaching the layer of viable cells. The microparticles were not found in every hair follicle, but only in 1 of 10. It was shown that it has to be distinguished between active and inactive follicles in relation to the penetration characteristics.
100.2 IN VIVO METHOD FOR ASSESSMENT OF ACTIVE AND INACTIVE FOLLICLES Laser scanning microscopy (LMS) in combination with cyanoacrylate skin surface biopsy was used by Lademann et al.14 to observe follicular penetration of topically applied formulations into the hair follicle. The follicle status (active/inactive) was checked by penetration experiments. The fluorescent food dye curcumin (C21H20O6) in an O/W formulation (2% oil-in-water emulsion) was applied onto the investigated skin areas. Curcumin has distinct fluorescent properties and can easily be detected by laser scanning microscopy with argon laser radiation. It shows a strong fluorescence after excitation at 488 nm in the spectral region of 550 nm. The fluorescence signal can be detected at 590 nm, which is distinct from the background fluorescence caused by protein bands and cyanoacrylate. The curcumin emulsion was applied to selected areas on the upper arm of the volunteers. After an application time of half an hour, a cyanoacrylate follicular biopsy was taken from the treated skin area and measured using laser scanning microscopy. Cyanoacrylate follicular biopsy was performed by applying one drop of cyanoacrylate glue (UHU® GmbH, Brühl, Germany) onto the skin and covering it with a glass slide under light pressure. After polymerization, which occurs in 1 minute, the glass slide was gently lifted and removed. The cyanoacrylate follicular biopsy contains the vellus hairs and a cast of the follicle infundibula (Figure 100.1).
FIGURE 100.1 Cyanoacrylate follicular biopsy containing the vellus hairs and a cast of the follicle infundibula. (From Otberg, N., Penetrationsverhalten topisch applizierter Substanzen in den Haarfollikel: Untersuchungen zur Physiologie und der Größenparameter von Körperhaarfollikeln, Doktorarbeit, Humboldt Universität zu Berlin, Medizinische Fakultät Charité, February 2003.)
The penetration behavior of the applied dye into the follicles could be detected analyzing the cross section of the removed follicle contents by fluorescence measurements using laser scanning microscopy.14 In Figure 100.2 the transmission and fluorescence images of two typical cross sections of different follicle channels are compared. The applied dye could be detected by fluorescence measurements in the whole area of the cross section of the follicle channel presented in Figure 100.2a. Identical results were obtained analyzing the cross sections of this follicle channel in different depths. The situation is different in the follicle channel presented in Figure 100.2b. In this case, a fluorescence signal could not be detected in the middle of the cross section in any depths of the follicle channel. This provides evidence that the dye did not penetrate into this follicle channel. Thus, inactive and active follicles could be distinguished.14 The dye was found in most of the hair follicles, although some follicles were obviously closed for penetration. Reasons for this phenomenon are found in the specific follicular properties. A combination of various tape-stripping and staining methods made it possible to measure hair growth and sebum excretion of every single hair follicle in the defined skin area. A correlation between the penetration properties, hair growth activity, and sebum production was found. Telogen hair follicles that were not
Methods for Assessment of Follicular Transport in Ex Vivo and In Vivo
863
FIGURE 100.2 (a) Transmission and fluorescence images of cross sections of an active follicle infundibulum removed by cyanoacrylate tape stripping from a skin area treated with dye-containing sunscreen. (From Lademann, J. et al., Skin Pharmacol. Appl. Skin Physiol., 14, 17, 2001.)
FIGURE 100.2 (b) Transmission and fluorescence images of the cross sections of an inactive follicle infundibulum removed by cyanoacrylate tape stripping from a skin area treated with dye-containing sunscreen. (From Lademann, J. et al., Skin Pharmacol. Appl. Skin Physiol., 14, 17, 2001.)
excreting sebum were closed for the penetration process.14 This means that only in the case of a movement from inside of the hair follicle can a topically applied substance penetrate into the follicle. It was proposed that the inactive follicles should be closed by a cover, which prevents the penetration. This effect was investigated by the in vivo analysis of the follicle infundibula cross section by optical coherence tomography (OCT).16 The employed commercial OCT system (SkinDex 300 from ISIS Optronics, Mannheim, Germany)16 achieves a depth and lateral resolution in tissue of about 5 and 3 μm, respectively. At a center wavelength of 1300 nm, a broadband light source with 110-nm bandwidth was employed. Due to a spatial multiplexing scheme, a two-dimensional image with 512 lateral positions is collected within 2 seconds in its fastest acquisition mode. For enhanced image quality, but at the expense of acquisition time, three consecutive images are collected and averaged. By the OCT measurements it could be demonstrated that the orifices of inactive hair follicles, i.e., follicles that evidence neither hair growth nor sebum excretion,14,17,18 were filled with a plug. This plug demonstrated morphological characteristics similar to those of the stratum
corneum and protruded from the surface of the skin. It seems likely that this structure consists of shedded corneocytes glued together with dry sebum. This plug is well seen in the OCT image of the cross section of the follicle infundibula in Figure 100.3.17 This kind of plug was found on the top of all closed follicles determined in the penetration experiments.
FIGURE 100.3 OCT image of the cross section of an inactive follicle closed by a plug. (From Otberg, N. et al., Las. Phys. Lett., 1, 46, 2004.)
864
Handbook of Non-Invasive Methods and the Skin, Second Edition
Only 74% of the hair follicles on the upper arm were open for the penetration of the curcumin emulsion. After pretreatment with the exfoliant the percentage of open follicles was raised to 100%. The mechanical peeling usually removes the upper layer of the stratum corneum, and the results show that such a pretreatment can also dislodge the follicular plugs and improve follicular penetration.
100.3 IN VIVO METHOD FOR ASSESSMENT OF FOLLICULAR TRANSPORT As described before, there were two different approaches in the past to determine the influence of the hair follicles on the penetration of topically applied substances. The one approach was based on in vitro measurements using diffusion cells. The penetration was compared in different tissue types containing more or less follicles. In all these studies, the tissue samples had not only different amounts of follicles, but also significantly different properties of the stratum corneum. The second approach was based on biopsies, which were taken from tissue after application and penetration of substances. The distribution of the substances in the hair follicles was analyzed in the histological section of these biopsies by autoradiography or fluorescence measurements. These investigations were carried out either in vitro or ex vivo. Penetration kinetics cannot be analyzed by biopsy sampling. The in vivo analysis of the follicle penetration could be realized only by optical and spectroscopic online methods, which utilize the absorption, fluorescence, or scattering properties of the topically applied substances. Additionally, these methods should have a sufficient space resolution to distinguish the follicles from the surrounding tissue. The most promising in vivo methods are optical coherent tomography19,20 and the laser scanning microscopy.21,22 The OCT image is based on differences in the absorption and scattering of different tissue types and body fluids. Absorption and scattering are less sensitive for the detection of the small amounts of topically applied substances, which penetrate into the hair follicles. In contrast, the fluorescence measurements are very selective and sensitive, even for small concentrations, if the topically applied substances have strong fluorescent properties. The online in vivo investigation of the penetration process of topically applied substances into the lipid layers and into the hair follicles on any body site became possible using fiber-based laser scanning microscopes. In this case, the optical imaging and scanning system is incorporated into a handpiece, which can be applied to any skin area. The follicular penetration was investigated using the dermatological laser scanning confocal microscope
FIGURE 100.4 Dermatological laser scanning microscope Stratum (Optiscan Ltd., Melbourne, Australia).
Stratum (Optiscan Ltd., Melbourne, Australia) (Figure 100.4).23 The radiation of an Ar+ laser at 488 nm was used to excite the topically applied fluorescent dye curcumin and sodium fluorescine. The radiation was transferred by an optical fiber to the probe, which contained the scanning system. The investigated field of vision is 200 × 200 μm. The probe scans in real time; scan frequency depends on scan resolution. For maximum resolution, the frame rate is one frame per second. The optical window of the handpiece was set directly onto the skin surface. The fluorescence emission was collected by the objective lens and transferred by the optical fiber to a photodetector. The depth scan was realized by changing the focus position manually, by using the imaging depth control on the handpiece. The investigations were carried out on healthy male and female volunteers.23 The food dyes, curcumin, and sodium fluorescine were applied in ethanol or in emulsion (2 mg/cm2). The concentration of the dyes in the formulations was 1%. The treated skin area was 4 × 5 cm2. A typical distribution of the formulations on the skin surface is presented in Figure 100.4. The dye-containing emulsion is distributed nonhomogeneously in the stratum corneum and partly located in the furrows of the skin (arrow). High fluorescence intensity was always detected in the lipid layers that surround the corneocytes. This demonstrates that the lipid layers of the stratum corneum are an efficient penetration pathway for topically applied substances, as has been discussed in the literature.24–26 The follicles represent another penetration pathway for topically applied substances. Laser scanning confocal microscopy allows the online investigation of the follicular penetration process in vivo.
Methods for Assessment of Follicular Transport in Ex Vivo and In Vivo
865
FIGURE 100.5 Distribution of topically applied substances on the skin surface.
In Figure 100.5, this penetration process of the dye-containing formulation is demonstrated by analyzing the distribution of the fluorescent dye at different time points after application. On the images, it can clearly be seen how the substance has penetrated into the follicles. This method is well suited to analyze the follicular penetration in vivo and to investigate the phenomena of active and inactive follicles, concerning the penetration of topically applied substances. However, these investigations are not limited to hair follicles. The penetration of substances into the sweat glands could also be analyzed in this way.23
100.4 SUMMARY In vivo LSM measurements are well suited to investigate the skin structure and the penetration processes of topically applied substances in real time. These measurements became possible using fiber-based microscopes. In this case, the optical imaging and scanning system is incorporated into a handpiece, which can be applied to any skin area of the human body. The follicular penetration has to be considered an efficient penetration pathway in congruence with the intercellular penetration pathway. The noninvasive LSM can be used to determine active and inactive follicles concerning their penetration properties. This method presents a useful tool for clinical diagnostics and therapy control in dermatology.
REFERENCES 1. Schaefer, H. and Redelmeier, Th.E., In vitro approaches for predicting percutaneous absorption, in Skin Barrier: Principles of Percutaneous Absorption, Karger, Basel, 1996, 146.
2. Feldmann, R.J. and Maibach, H.I., Regional variation in percutaneous penetration of 14c cortisol in man, J. Invest. Dermatol., 48(2), 181, 1967. 3. Maibach, H.I., Feldman, R.J., Milby, T.H., and Serat, W.F., Regional variation in percutaneous penetration in man, Arch. Environ. Health, 23, 208, 1971. 4. Tenjarla, S.N., Kasina, R., Puranajoti, P., Omar, M.S., and Harris, W.T., Synthesis and evaluation of n-acetylprolinate esters: novel skin penetration enhancers, Int. J. Pharm.,192, 147, 1999. 5. Hueber, F., Besnard, M., Schaefer, H., and Wepierre, J., Percutaneous absorption of estradiol and progesterone in normal and appendage-free skin of hairless rat, lack of importance of nutritional blood flow, Skin Pharmacol., 7, 245, 1994. 6. Essa, E.A., Bonner, M.C., and Barry, B.W., Possible role of shunt route during iontophoretic drug penetration, Perspect. Percutan. Penetration, 8, 54, 2002. 7. Turner, N.G. and Guy, R.H., Visualization and quantification of iontophoretic pathways using confocal microscopy, J. Invest. Dermatol. Proc., 3, 136, 1998. 8. Tur, R., Buxaderas, C., Martinez, F., Busquets, A., Coroleu, B., and Barri, P.N., Comparison of the role of cervical and intrauterine insemination techniques on the incidence of multiple pregnancy after artificial insemination with donor sperm, J. Assist. Reprod. Genet., 14(5), 250, 1997. 9. Blume, U., Verschoore, M., Poncet, M., Czernielewski, J., Orfanos, C.E., and Schaefer H., The vellus hair follicle in acne, hair growth and sebum excretion, Br. J. Dermatol., 129, 23, 1993. 10. Seago, S.V. and Ebling, F.J., The hair cycle on the human thigh and upper arm, Br. J. Dermatol., 135, 9, 1995. 11. Pagnoni, A.P., Kligman, A.M., Gammal, S.E.L., and Stoudemayer, T., Determination of density of follicles on various regions of the face by cyanoacrylate biopsy, correlation with sebum output, Br. J. Dermatol., 131, 862, 1994. 12. Otberg, N., Richter, H., Schaefer, H., Blume, U., Sterry, W., and Lademann, J., The vellus hair follicle, a reservoir for topically applied substances, J. Invest. Dermatol., in press.
866
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. Lademann, J., Weigmann, H.J., Rickmeyer, C., Bartelmes, H., Schaefer, H., Müller, G., and Sterry, W., Penetration of titanium dioxide microparticles in a sunscreen formulation into the horny layer and the follicular orifice, Skin Pharmacol. Appl. Skin Physiol., 12, 247, 1999. 14. Lademann, J., Otberg, N., Richter, H., Weigmann, H.J., Lindemann, U., Schaefer, H., and Sterry, W., Investigation of follicular penetration of topically applied substances, Skin Pharmacol. Appl. Skin Physiol., 14(1), 17, 2001. 15. Otberg, N., Penetrationsverhalten topisch applizierter Substanzen in den Haarfollikel: Untersuchungen zur Physiologie und der Größenparameter von Körperhaarfollikeln, Doktorarbeit, Humboldt Universität zu Berlin, Medizinische Fakultät Charité, February 2003. 16. Lademann, J., Knüttel, A., Richter, H., Otberg, N., v. Pelchrzim, R., Audring, H., Meffert, H., Sterry, W., and Hoffmann, K., Application of optical coherent tomography for skin diagnostics, J. Quant. Electr., in press. 17. Otberg, N., Richter, H., Knüttel, A., Schaefer, H., Sterry, W., and Lademann, J., Laser spectroscopic methods for the characterization of open and closed follicles, Las. Phys. Lett., 1, 46, 2004. 18. Schaefer, H. and Lademann, J., The role of follicular penetration: a differential view, Skin Pharm. Appl. Skin. Phys., 14, 23, 2001.
19. Jackle, S., Gladkova, N., Feldchtein, F., Terentieva, A., Brand, B., Gelikonov, G., Gelikonov, V., Sergeev, A., Fritscher, S., Ravens, A., Freund, J., Seitz, U., Soehendra, S., and Schrodern, N., In vivo endoscopic optical coherence tomography of the human gastrointestinal tract: toward optical biopsy, Endoscopy, 32(10), 743, 2000. 20. Puliafito, C.A., Hee, M.R., Schuman, J.S., and Fujimoto, J.G., Optical coherence tomography of diseases, Thorofare NJ, 12, 1995. 21. Meuwissen, M.E., Janssen, J., Cullander, C., Junginger, H.E., and Bouwstra, J.A., A cross-section device to improve visualization of fluorescent probe penetration into the skin by confocal laser scanning microscopy, Pharm. Res., 15(2), 352, 1998. 22. Heise, H.M., Küppel, L., Pittermann, W., and Butvina, L.N., New tool for epidermal and cosmetic formulation studies by attenuated total-reflection spectroscopy, Fres. J. Anal. Chem., 271, 753, 2001. 23. Lademann, J., Richter, H., Otberg, N., Lawrenz, F., Blume, U., and Sterry, W., Application of a dermatological laser scanning microscope for investigation in skin physiology, J. Las. Phys., 13(5), 756, 2003. 24. Pilgram, G.S., Engelsma-van Pelt, A.M., Koerten, H.K., et al., Pharm. Res., 17(7), 796, 2000. 25. Hatziantoniou, S., Rallis, M., Demetzos, C., et al., Pharmacol. Res., 42(1), 55, 2000. 26. Ogiso, T., Yamaguchi, T., Iwaki, M., et al., J. Drug Target., 9(1), 49, 2001.
Hair, Physical Properties, and Growth Rate
101 Measurement of Hair Growth Julian H. Barth1 and D. Hugh Rushton2 1
Department of Chemical Pathology, General Infirmary at Leeds, Leeds, United Kingdom School of Pharmacy and Biomedical Sciences, University of Portsmouth, Portsmouth, United Kingdom 2
CONTENTS 101.1 Introduction ..........................................................................................................................................................869 101.2 Subjective Measurement ......................................................................................................................................869 101.2.1 Scoring Systems for Scalp Hair Growth...............................................................................................869 101.2.2 Scoring Systems for Body Hair Growth ...............................................................................................869 101.3 Objective Measurements ......................................................................................................................................870 101.3.1 Presampling Considerations ..................................................................................................................870 101.3.2 Presampling Recommendations.............................................................................................................871 101.3.3 Sampling Criteria...................................................................................................................................871 101.3.4 The Unit Area Trichogram ....................................................................................................................871 101.3.5 Examining Hair Root Status..................................................................................................................871 101.3.6 Problems with Anagen, Catagen, and Telogen Classifications.............................................................871 101.3.7 Determination of Hair Diameter ...........................................................................................................872 101.3.8 Evaluation of Hair Shaft........................................................................................................................873 101.3.9 Summary of Published Methods of Objective Hair Measurement.......................................................873 References .......................................................................................................................................................................873
101.1 INTRODUCTION Techniques for the measurement of hair growth were initially developed to satisfy the needs of the wool industry, human anthropologists, and the human cosmetics industry. Indeed, major developments have been established for the former two, rather than the latter purpose. The methods described in this chapter derive from our interest in human hair growth and are divided into subjective and objective methods. The former are more suitable for body hair growth since both hair form and the density of hair vary considerably from one site to another. The objective methods are more suitable for the denser growth on the scalp since cosmetic changes are only apparent after there has been a reduction of approximately 20% in hair quantity.
101.2 SUBJECTIVE MEASUREMENT 101.2.1 SCORING SYSTEMS GROWTH
FOR
SCALP HAIR
The patterns produced by the gradual process of scalp hair loss in male pattern balding were first described in men
by Hamilton1 and in women by Ludwig.2 The recognition that an individual’s scalp hair loss fits one of these patterning systems is essential for the diagnosis of scalp hair loss, but the staging is regrettably too insensitive for all but the most crude of measurement. An attempt has been made to scan the appearance of the scalp into a computer imaging system,3 but this is impeded by the presence of hair; more sensitivity is only afforded by shaving an area of the scalp.4
101.2.2 SCORING SYSTEMS GROWTH
FOR
BODY HAIR
Subjective grading systems have been designed by Pedersen, 5 Beek, 6 Garn, 7 Shah, 8 Ferriman and Gallwey,9 Lorenzo,10 and Lunde11 (Table 101.1). All these methods are based on subdividing the body into zones; each is scored and the total is summed. However, it is the study by Ferriman and Gallwey that has been granted approval by most hirsuties investigators, as it is the least complex and easiest to use. Unfortunately, as hirsuties is a subjective condition, the subjective methods of its measurement are subject to considerable observer variation (Figure 101.1). 869
870
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 101.1 Summary of Subjective Methods for the Evaluation of Hair Growth on the Face, Trunk, and Limbs Number of Sites Evaluated
Authors Pedersen Beek Garn
12 19 11
Shah
9
Ferriman and Gallwey Lorenzo Lunde
11 7 subzones 18
Specific Features Evaluated Evaluation only of presence of hair at each site but no quantitative assessment Evaluation only of presence of hair at each site but no quantitative assessment Evaluation of patterns of hair growth at each site and form of hair shaft present, e.g., curly or straight; no quantitative assessment Quantitative assessment of terminal hairs >0.5 cm; (Q)uality (scored 1–3), (D)ensity (0–3) and (P)roportion of zone covered with hair (0–1); total score = Q × D × P Quantitative score based on distribution of hair on each site Quantitative score based upon density and extent of involvement Quantitative assessment of length and density of terminal hairs > 0.5 cm
It could be concluded that the difference is due to variation in the degree of hair growth in different populations of hirsute women, but it is more probably due to a lack of observer conformity. It should be noted that in the myriad of publications that have employed subjective measurement of body hair growth, there are few attempts to evaluate the precision of the method used. In our hands, the Ferriman and Gallwey scoring system has a repeatability coefficient of 3.2 (Ferriman and Gallwey units) at a mean score of 26.12
101.3 OBJECTIVE MEASUREMENTS Quantitative evaluation of scalp hair requires techniques that are sensitive enough to assess fundamental variables such as hair density, fiber diameter, proportion of anagen
hair (i.e., actively growing), and linear growth rate. Such information provides essential details for determining normal morphology as well as understanding changes arising from disease.
101.3.1 PRESAMPLING CONSIDERATIONS Presampling factors are probably the most difficult problems to standardize. These include shampooing, combing, and other cosmetic procedures, all of which must be standardized to provide unbiased measurements. Unfortunately, ideal protocols for hair measurements are impractical since they would include 3-month shampoo-free periods for telogen counts. Therefore, subjects should continue with their normal daily routines, which are used for background data. The shampooing interval prior to sampling should be kept standard.
Ferriman & Gallwey score
40
30
20
10
0
FIGURE 101.1 Comparison of Ferriman and Gallwey scores of hirsute women from 11 studies (mean ± SD). It is assumed that the severity of hirsuties of women attending different centers is similar and therefore this graph illustrates the variation in hirsuties grading perceived by different investigators. (Redrawn from Diseases of the Hair and Scalp, Rook and Dawber, Eds., Blackwell Scientific, Oxford, 1991.)
Measurement of Hair Growth
101.3.2 PRESAMPLING RECOMMENDATIONS Hair should be shampooed daily or on alternate days during the month prior to sampling, and not less than on alternate days throughout the study period. On the day of sampling specimens should be obtained, or visual imaging performed, within 3 hours of shampooing. Such actions produce minimal influences upon the derived anagen estimate, while providing excellent conditions for photographic or image analysis reproductions. It should also be pointed out that an inadequate shampooing action can have a detrimental influence upon the derived anagen estimate. Inadequate and inefficient shampooing can be found in subjects experiencing excessive hair loss, which results in poor cleansing and only partial removal of the hairs due to be shed. As excessive amounts of hair are seen at each wash, the shampooing frequency is decreased in a belief that this will reduce future loses. When this situation is found, sufficient time must elapse to allow the reestablishment of shedding levels representative of the problem under investigation.
101.3.3 SAMPLING CRITERIA The selection of area to be sampled is also of critical importance. Sites must be chosen carefully to represent the changes occurring. Problems arising from patterned presentations within the sampling distribution require careful attention to avoid biased estimates. No real problems should be encountered where the distribution of hair density is uniform, assuming sufficient hairs are obtained. Approximately 100 hairs are required to estimate variables by proportion; consequently, a sufficiently hairy area will need to be sampled. Two sample sites providing the required total number of hairs give better estimates (smaller sampling variation) than a single site.13 However, the surrounding area should be indistinguishable (with respect to hair density) from the sample sites (Figure 101.1). This arrangement allows resampling within ±5 mm of the original sites without the need to find the exact initial spot. Disturbances such as male pattern baldness often present problems because of patterned distributions in affected areas. Hair density within the vertex or frontal hair line can sometimes vary tremendously, and exact resampling is required. This can be achieved by placing a small tattoo between the two original sample sites or within the center of a single site. Where tattooing is not possible, location coordinates are essential for accurate relocation at some future date. Resampling frequencies depend upon the variables being estimated and the method of evaluation being employed. Where plucking techniques are used, hair density cannot be reestimated for 6 months; with noninvasive methods (hair weights or phototrichograms), resampling can be performed within 1 month. However, estimates for the per-
871
centage of hair in the anagen growth phase need to be made 12 months apart due to seasonal variation14 and are a requirement independent of the method of evaluation.
101.3.4 THE UNIT AREA TRICHOGRAM The unit area trichogram evaluates scalp hair variables with excellent reproducibility. It is a semi-invasive (plucking) technique in which all the hairs within a defined area (usually >30 mm2) are counted and measured.15,16 The area to be sampled should first be degreased with an acetone/isopropanol (60:40 v/v) mixture to remove surface lipids, which can cause blurring of the delineating sample line. The area to be sampled should be identified prior to epilation; a roller ball pen produces the sharpest line compared to circles drawn with fiber, felt, or ballpoint pens. The sample area can be quantitatively measured from an enlarged black-and-white photograph containing a scale bar or a computer image. The hairs should be rapidly epilated in a single smooth action in the direction of growth in order to minimize trauma to the roots.
101.3.5 EXAMINING HAIR ROOT STATUS All epilated hairs are placed directly onto double-sided tape (15 mm wide) attached to a 25-mm microscope slide. The collected hairs are subsequently realigned so that their root ends protrude from the tape edge toward the center, thereby allowing easy visualization within the microscope (Figure 101.2). Each hair is classified microscopically (magnification, ×40) according to its growth phase (anagen, catagen, or telogen), diameter, and length. The typical microscopical (dry-mounted) appearance of the various hair growth stages, anagen, catagen, and telogen, are shown in Figure 101.3. Catagen hairs are classified with the telogen population for data analysis since catagen is effectively the first stage of telogen. The catagen bulb structure is fully keratinized and encased in a shriveledup translucent outer root sheath (Figure 101.3), and for these reasons we classify catagen fibers with the telogen population.
101.3.6 PROBLEMS WITH ANAGEN, CATAGEN, TELOGEN CLASSIFICATIONS
AND
A possible area of confusion and contention is the reporting of dysplastic or dystrophic hairs, which are usually observed in plucked samples. Apart from a few rare congenital diseases, dysplastic or dystrophic hair shafts are never seen on the scalps of men, women, or children, nor in scalp biopsy. It is our belief that these terms are in reality descriptions for traumatic features produced by the procedure of hair epilation. Classification of such roots is preferable to their exclusion from data analysis, and a few simple guidelines should ensure that difficult root presentations are assigned to the correct population.
872
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 101.2 Plucked hairs mounted on double-sided tape and ordered by length. The plucked hairs are initially collected onto another strip of adhesive tape and then during transfer into the illustrated format have their length measured.
The principal feature to keep in mind is that actively growing hairs (anagen) have soft prekeratinized tissues up to approximately 2.0 mm above the apex of the dermal matrix.13 This soft tissue bends easily and distorts upon epilation. Catagen and telogen hairs differ since they are fully keratinized, and are therefore rigid and do not easily bend. Moreover, the diameters of catagen and telogen fibers taper toward the bulb, and this zone is devoid of, or has reduced, pigmentation. The following therefore serves as a guide to assigning difficult fibers to the anagen or telogen populations: If the root end exhibits bending with or without pigmentation, we assign it to the anagen population. If not, we assign it to the telogen population. Broken hairs arising from the epilation procedure also cause difficulties with interpretation, but should be assigned to the telogen population if the proximal end exhibits tapering or loss of pigment; otherwise, the fiber should be classified as anagen. Fortunately, the occurrence of such hairs is <5% when hair is epilated singly.
101.3.7 DETERMINATION
FIGURE 101.3 Photomicrograph of three hair shafts. The left shaft has the characteristic drumstick appearance of a telogen bulb. The middle shaft has a tapered straight bulb surrounded by a loose atrophic sack and is a catagen bulb. The shaft on the right has a densely pigmented and distorted bulb with an extravagant collar of external root sheath and is typical of an anagen bulb.
OF
HAIR DIAMETER
The shaft diameters can be measured with an eyepiece micrometer using the shafts mounted as in Figure 101.2. Major and minor axes should be measured and averaged, but single-plane estimates are acceptable if only the anagen or telogen diameter is required. Anagen fibers should be measured 10 to 15 mm from the root end to avoid the influences of weathering. Telogen fibers should be measured 10 mm from the last sign of tapering, which is usually 15 to 20 mm from the bulb. Since telogen hairs of equal length have correspondingly smaller diameters than their anagen counterparts,13 studies involving hair diameter comparisons should be controlled for anagen-totelogen ratios. This will ensure that any changes are true diameter effects and not simply a consequence of changes in the anagen or telogen populations. Recently, studies on linear hair growth in scalp hair have found a correlation between fiber diameter and the rate of daily growth.17
Measurement of Hair Growth
873
TABLE 101.2 Summary of Methods of Hair Evaluation Unit Area Trichogram Sample time (h) Number of hairs measured Accuracy Precision (%) Repeat sampling
Phototrichogram
4–6 50–180 Excellent ±5 6 monthly
4–8 90–200 Good ±15 Monthly
Thicker hairs grow faster than finer hairs; as a result, studies involving linear growth rate must now specify diameter values.
101.3.8 EVALUATION
OF
HAIR SHAFT
A novel method for the measurement of hair growth on the limbs has been described: the vellus index.18 This method is only appropriate for hair growth on the body and limbs. It utilizes hair shaved from the skin, and the ratio of pigmented, medullated terminal to nonpigmented, nonmedullated vellus shafts is simply determined by microscopic examination of the shafts. It has the great advantage of not requiring samples to be shaved from a defined area. This method has been shown to separate men from women, and that hirsute women have an intermediate amount of hair. It has not yet been used for monitoring antiandrogen therapy.
101.3.9 SUMMARY OF PUBLISHED METHODS OBJECTIVE HAIR MEASUREMENT
OF
The methods available for hair evaluation are compared in Table 101.2. Scalp biopsies give a similar degree of quantitative information to the unit area trichogram, but suffer from being unable to resample the original site. Visual hair counting would appear to be of little value in determining the number of hairs present on the scalp, and has now been abandoned. Telogen hair counts offer no obvious information about the rate or severity of hair shedding. There are too many unknown variables that cloud the significance of total telogen counts, and without data for the total numbers of hair present or the duration of anagen, daily counts of telogen hairs are meaningless. Noninvasive methods such as the phototrichogram currently suffer from two-dimensional imaging problems19; however, these may be improved with the introduction of three-dimensional image analysis, and then phototrichograms may become the method of choice. The ability to estimate hair quality demands quantitative techniques capable of measuring fundamental hair components. Interrelating these variables provides additional data giving a sensitive measure of changes
Visual Counting
Scalp Biopsy
Telogen Counts
Sample Weight
1–2 50–450 Poor ±35 Immediate
24–48 5–20 Good ±5 No
1–2 20–500 Poor Unknown Daily
4–6 115–310 Good Unknown Bimonthly
occurring. Although time-consuming and mildly painful, the unit area trichogram offers the trained investigator a valid method for determining fiber thickness, density, and proportion actively growing, and appears the method of choice.
REFERENCES 1. Hamilton, J.B., Patterned loss of hair in man: types and incidence, Ann. N.Y. Acad. Sci., 53, 708, 1951. 2. Ludwig, E., Classification of the types of androgenetic alopecia (common baldness) occurring in the female sex, Br. J. Dermatol., 97, 247, 1977. 3. Gibbons, R.D., Fiedler-Weiss, V.C., West, D.P., and Lapin, G., Quantification of scalp hair: a computer-aided methodology, J. Invest. Dermatol., 86, 782, 1986. 4. Van Neste, D., Dumortier, M., and De Coster, W., Phototrichogram analysis: technical aspects and problems in relation with automated quantitative evaluation of hair growth to computer assisted image analysis, in Trends in Human Hair Growth and Alopecia Research, Van Neste, D., Lachappelle, J.M., and Antoine, J.L., Eds., Kluwer Academic, Dordrecht, The Netherlands, 1989, p. 155. 5. Pedersen, J., Hypertrichosis in women, Acta Dermatol. Venereol., 23, 1, 1943. 6. Beek, C.H., A study on the extension and distribution of the human body hair, Dermatologica, 101, 317, 1950. 7. Garn, S.M., Types and distribution of the hair in man, Ann. N.Y. Acad. Sci., 53, 498, 1951. 8. Shah, P.N., Human body hair: a quantitative study, Am. J. Obstet. Gynecol., 73, 1255, 1957. 9. Ferriman, D. and Gallwey, J.D., Clinical assessment of body hair growth in women, J. Clin. Endocrinol., 21, 1440, 1961. 10. Lorenzo, E.M., Familial study of hirsutism, J. Clin. Endocrinol. Metab., 31, 556, 1970. 11. Lunde, O., A study of body hair density and distribution in normal women, Am. J. Phys. Anthropol., 64, 179, 1984. 12. Barth, J.H., Cherry, C.A., Wojnarowska, F., and Dawber, R.P.R., Cyproterone acetate for severe hirsutism: results of a double-blind dose-ranging study, Clin. Endocrinol., 35, 5, 1991.
874
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. Rushton, D.H., Chemical and Morphological Properties of Scalp Hair in Normal and Abnormal States, Ph.D. thesis, University of Wales, Cardiff, 1988. 14. Randall, V.A. and Ebling, F.J.G., Seasonal changes in human hair growth, Br. J. Dermatol., 124, 146, 1991. 15. Rushton, D.H., Ramsay, I.D., James, K.C., Norris, M.J., and Gilkes, J.J.H., Biochemical and trichological characterisation of diffuse alopecia in women, Br. J. Dermatol., 123, 187, 1990. 16. Rushton, D.H., Ramsay, I.D., Norris, M.J., and Gilkes, J.J.H., Natural progression of male pattern baldness in young men, Clin. Exp. Dermatol., 16, 188, 1991.
17. Van Neste, D.J.J., de Brouwer, B., and Dumortier, M., Reduced linear hair growth rates of vellus and of terminal hairs produced by human balding scalp grafted onto nude mice, Ann. N.Y. Acad. Sci., 642, 480, 1991. 18. Madanes, A.E. and Novotny, M., The vellus index: a new method of assessing hair growth, Fertil. Steril., 48, 1064, 1987. 19. Rushton, D.H., de Brouwer, B., De Coster, W., and Van Neste, D.J.J., Comparative evaluation of scalp hair by phototrichogram and unit area trichogram analysis within the same subjects, Acta Dermatol. Venereol., 73, 150, 1993.
of the Hair: 102 Microscopy The Trichogram Ulrike Blume-Peytavi Department of Dermatology, Hospital Charité, Humboldt University, Berlin, Germany
Constantin E. Orfanos Department of Dermatology, University Medical Center Steglitz, The Free University of Berlin, Berlin, Germany
CONTENTS 102.1 Introduction ..........................................................................................................................................................875 102.2 Object ...................................................................................................................................................................876 102.3 The Trichogram Technique ..................................................................................................................................876 102.3.1 Methodological Principles .....................................................................................................................876 102.3.2 Evaluation ..............................................................................................................................................877 102.3.3 Types of Hair Roots...............................................................................................................................877 102.4 Criteria for Standardizing the Trichogram Procedure and Eliminating Sources of Error..................................878 102.5 Comparison with Other Techniques ....................................................................................................................879 References .......................................................................................................................................................................880
102.1 INTRODUCTION Dermatologists are frequently consulted for focal or diffuse effluvium, hair thinning, and changes in hair structure. In man, hair has no vital biological function but an incalculable psychological one. Hair is a major esthetic display feature of the human body, especially in social and sexual interaction. In the diagnosis of hair disease, clinical history and daily counting of shed hair by the patient are useful, but an objective and reproducible method is needed for exact quantification and characterization of hair loss. In most rodents and other mammals, hair does not grow continuously, and the hair follicle undergoes cyclic changes of growth, regression, and inactivity periods throughout life.1,2 In humans, however, these cyclic changes are not synchronized and seem to be regulated by complex interactions between the individual epithelial and mesenchymal components of the hair follicle.3–6 A large variety of endogenous and exogenous factors may possibly intervene to disturb normal growth regulation processes.4,7 Each individual hair follicle undergoes a cyclic rhythm of:
1. A growing or anagen phase, during which the hair is produced. The hair with medulla, cortex, cuticle, and inner and outer root sheaths is formed by the highly active matrix cell population located in the hair bulb region, probably the most rapidly growing cell population of all normal tissues, reaching a maximum cell turnover rate entering mitosis every 24 hours. Different anagen stages, like proanagen (anagen I to V) and metanagen (anagen VI), can be distinguished by histological examination. The normal duration of anagen in any individual scalp follicle is genetically determined and ranges from 2 to 6 years, with an approximate average anagen duration of 1000 days.8 The growing anagen hair is normally firmly bound to the hair root and can only be plucked with some pain sensation. The anagen phase is followed by a transition or catagen phase. 2. The transition or catagen phase lasts only a few days (2 to 3 weeks). Catagen is characterized by a decreasing and finally stopping activity of the matrix, followed by involution and keratinization of the bulb. Subsequently, the bulb 875
876
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 102.1 Duration of the Anagen and the Telogen Phase in Different Regions of the Body Site3–6,16,20,30,36–38 Body Region
Anagen
Telogen
Scalp Eyebrows Beard Moustache Axilla Pubic region Hand Finger Arms Leg
2–6 years 4–6 weeks 12 months 16 months Months 47 weeks 10 weeks 12 months 13 months 21 months
3–4 months 3 months 3 months 6 months 3 months 2 weeks 7 weeks 9 months 13 months 19 months
moves upward toward the surface of the skin to prepare the following. 3. In the resting or telogen phase the hair is shed; during telogen, the follicle is located just below the orifice of the sebaceous gland. The club hair is held in its envelope by the intercellular junctions and can be pulled without any pain. The subsequent hair cycle starts spontaneously under the influence of an unknown stimulus at the end of telogen or can be induced by plucking the club hair. Normally, the telogen hair either falls out at the beginning of the new anagen phase or is retained in the follicle until metanagen is well established during the next hair cycle. In the human scalp, the telogen phase lasts between 2 and 3 months. The duration of hair cycle phases in each individual is age dependent9 and varies according to body region,5,10 produced hair type (terminal, lanugo, or vellus hair),11 and with seasonal changes12 (Table 102.1).
102.2 OBJECT Studies on the dynamics of the human follicular cycle largely depend on trichogram examinations, a microscopic evaluation of plucked hairs with subsequent quantitative measuring of the number of individual hair roots. Morphological examination of hair roots was first introduced by Van Scott et al.,13 as an indicator of hair growth in toxicologic studies on cytostatic drugs. Initially, the term trichogram was used by Pecoraro and coworkers14,15 to describe several trichometric parameters, such as growth rate, thickness of the hair shaft, and telogen rate. In the following years, the trichogram technique was developed and standardized to serve as a reliable diagnostic measure
in hair diseases. Today, this term is used to describe the examination of a simultaneously epilated tuft of 50 to 80 hairs and the count of various hair root types under the light microscope by a standardized technique. A major target of trichogram measurements is to count and evaluate the status of individual hair roots and to establish the anagen/telogen hair ratio.16–19 Trichogram measurements are based on the hair cycle, and therefore serve as a standard method for quantifying hair follicles in their different growth cycle phases.20 Thus, this simple and repeatable technique is a diagnostic measure to: • • • • • • •
Assess hair growth capacity Obtain an overview on the current state of growth of the hair follicles21 Detect hair growth cycle disturbances22,23 Classify different forms of hair loss and alopecia18,24 Elucidate pathogenetic mechanisms leading to diffuse hair loss Detect exogenous damage or toxic effects on hair structure Assess the effectiveness of hair growth-promoting treatments25
In addition, microscopic examination of hair shaft morphology can also help to detect malformations of the hair shaft as in primary genotrichoses (e.g., monilethrix, trichorrhexis nodosa), secondary genotrichoses (e.g., amino acid disorders, phenyl-ketonuria),26 etc.
102.3 THE TRICHOGRAM TECHNIQUE 102.3.1 METHODOLOGICAL PRINCIPLES Materials necessary for performing a trichogram are trichogram forceps, a pair of scissors, 76 × 26 mm glass slides, coverslips, hair clips, Corbit™ balsam (I. Hecht, Kiel, Germany), preparation needles, and a bioccular microscope or a Reichert projection microscope Visopan® (Reichert & Jung, Germany) with variable objectives (4×, 10×, 40×, 63×) (Figure 102.1). The reproducibility of trichogram measurements depends on the maintenance of high standards in obtaining the hair samples. Trichogram examinations should always be performed 5 days after the last shampooing of the hair. Patients have to interrupt all local treatments, including cosmetic procedures, 2 to 3 weeks earlier. On the 5th day after the last hair shampooing, hairs are taken from at least two specified sites. In diffuse effluvium and in androgenetic alopecia one sample is taken 2 cm behind the frontal hair line and 2 cm from the midline; another sample is obtained from the occipital region, 2 cm lateral from the protuberans occipitalis. In circumscribed alopecia, such
Microscopy of the Hair: The Trichogram
877
may leave a small bald patch that may become slightly erythematous for a short time afterward. Performing a trichogram may cause slight discomfort, caused by the removal of the whole group of hairs. Nevertheless, when correctly performed, trichogram is not painful and the plucked hairs grow back after 3 to 6 weeks. Correctly embedded hair roots can be stored for a long time.
102.3.2 EVALUATION
FIGURE 102.1 Material necessary for trichogram measurements: two trichogram forceps, glass slides, cover slips, a pair of scissors, Corbit balsam.
Evaluation of trichogram is performed with a bioccular microscope or ideally with a projection microscope. The embedded hair roots are individually analyzed by microscopic examination of the mounted glass slides. First, the overall shape, the presence or absence of hair root sheaths, the external contours, and the hair shaft diameters are examined. In addition, the appearance and size of the bulb are also important, as atrophy and shrinkage reflect reduced growth activity.27 At every stage of the hair growth cycle the hair follicle is characterized by a typical morphological structure. Based on microscopic examination of plucked hair roots, it is possible to decide on the hair growth phase of the examined hair as anagen, catagen, or telogen (Figure 102.3a to c), and in addition, also dysplastic anagen, dystrophic, and broken-off hairs (Figure 102.3d and e) can be distinguished. Finally, the percentage of the different hair root types is calculated.
102.3.3 TYPES
FIGURE 102.2 Plucking of a tuft of 50 to 80 hairs, firmly grasped with a trichogram forceps with rubber-protected jaws and pulled in the direction of emergence from the scalp.
as alopecia areata, one sample is taken from the border of the lesion and the control is taken from the contralateral, obviously not affected region. In the examined area, surrounding hair is fixed with clips and a bundle of 60 to 80 hairs is grasped with trichogram forceps, whose jaws are protected with rubber. This facilitates grasping hairs of different thickness and minimizes hair traumatization (Figure 102.2). Subsequently, forceps are completely closed and hairs are then plucked by twisting and lifting the hair shafts in the direction of emergence from the scalp. To prevent dehydration of the hair root shafts, they are immediately placed with their roots on a glass slide in embedding medium (e.g., Corbit balsam). Hair shafts are cut off approximately 1 cm above the root sheath. After they have been arranged one by one with a preparation needle, they are covered with a coverslip and left to dry for 24 hours. Plucking scalp hair
OF
HAIR ROOTS
Anagen hair: The root has a large base with an equally large diameter throughout. A firm inner and frequently outer root sheath can be clearly distinguished. Often, a more than 20˚ angulation of root and shaft can be observed (Figure 102.3a). Catagen hair: It generally has an equal diameter throughout or can become narrower toward the base, frequently with a missing or wrinkled root sheath. Contours are not deformed (Figure 102.3b). Telogen hair: The root is characteristically club shaped with smooth contours, and its sheaths are contracted around the tip (Figure 102.3c). Dysplastic hairs: They are thin but obviously growing anagen hairs, with a diminished matrix diameter and plucked without root sheaths; the lower end of the hair shaft is usually wavy or bent like the handle of a walking cane (Figure 102.3e). Dystrophic hairs: They are mainly described as thin, nongrowing anagen hairs, often with defective keratinization; changes are so severe that the root has broken off at the narrowest level with a tapering, pencil-like, broken end tip, sometimes showing individual bundles of keratin fibrils.
878
Handbook of Non-Invasive Methods and the Skin, Second Edition
(a) (b)
FIGURE 102.3 Different hair root types which can be distinguished by microscopic examination of individual hair roots. (a) anagen, (b) catagen, (c) telogen, (d) broken-off hair, (e) dysplastic anagen hair root.
Broken-off hairs: They are normally growing anagen hair shafts that break off upon plucking (Figure 102.3d). They can be easily recognized since, in contrast to dystrophic hair, the broken ends appear smooth, with a remaining diameter equal to that of the hair shaft. They may amount to up to 5 to 6% of the total number, but their prevalence is higher when hair is fragile or when the plucking technique was not adequate. The existence of variable percentages of such abnormal hair roots may be of interest in various diseases where the hair can serve as a marker for an underlying systemic defect or disease (Table 102.2).
102.4 CRITERIA FOR STANDARDIZING THE TRICHOGRAM PROCEDURE AND ELIMINATING SOURCES OF ERROR The trichogram technique provides reliable results under the condition that hair samples have been obtained under
a standardized procedure; already minor variations may lead to misinterpretation and false results. •
•
•
•
There should be a 5-day period prior to examination to avoid an artificial reduction of the telogen rate.28 The same standardized scalp regions (see above) should always be chosen to permit an optimal comparison of trichogram measurement results, for example, for regular follow-up of individual patients and for clinical studies.20 Correct plucking speed (sharp quick pull) and exact plucking in the direction of the emergence angle of the hairs from the scalp are very important to obtain a reliable hair root pattern.29 Slow or hesitant traction, very rapid plucking, or the wrong pulling direction may induce distorsions and alterations of the hair shafts and cause serious damage to the hair roots. More than 50 hairs have to be evaluated, as the error increases significantly with decreasing size of the hair sample.20,30
Microscopy of the Hair: The Trichogram
879
(d)
(e)
(c)
FIGURE 102.3 Continued.
•
•
Trichograms with >10% broken-off hairs cannot be evaluated correctly and should be repeated.20 Trichogram measurements should always be performed and evaluated by the same experienced investigator to maintain optimal and comparable examination conditions. A short training period with an experienced colleague is recommended.
102.5 COMPARISON WITH OTHER TECHNIQUES The trichogram is a semi-invasive technique with plucking of the entire hair. This makes it unsuitable for the monthly follow-up of patients, for studying the hair growth rate and the duration of the growing stage of individual hairs, and for studying seasonal variations. Noninvasive microscopic techniques such as optical microscopy with image analysis, 31 phototrichogram, 32,33 or the unit area trichogram34 should be chosen for this purpose. The trichogram technique is accurate and reliable for the diagnosis of hair diseases and is highly suitable because of its handiness.35 Although trichogram measurements are only
TABLE 102.2 Normal Distribution Pattern of Hair Roots in the Trichogram of the Scalp20 Hair Root Anagen Dysplastic anagen Catagen Telogen Dystrophic Broken-off hairs
% 60–80 5–20a 1–3 12–15 <2 5–6
a
In children, adolescents, and in persons with thin hair the rate of dysplastic anagen can occasionally reach >50%.
confined to two small scalp regions, earlier studies provided evidence that, except in alopecia areata, the trichogram technique of one site is representative of the neighboring areas.30
880
Handbook of Non-Invasive Methods and the Skin, Second Edition
Trichogram evaluation of hair root pattern is particularly useful at the beginning of abnormal hair loss, when hair density still seems normal at clinical examination. In this case, trichogram results may permit an early diagnosis, early determination of prognosis, and an early start of treatment to prevent further continuation of effluvium. In addition, normal trichogram results may allow a favorable prognosis and help to avoid various expensive and useless treatments.
REFERENCES 1. Kligman, A.M., The human hair cycle, J. Invest. Dermatol., 33, 307, 1959. 2. Uno, H., Biology of hair growth, Sem. Reprod. Endocrinol., 4, 131, 1986. 3. Ebling, F.J.G., The hair, in Textbook of Dermatology, Rook, A., Wilkinson, D.S., Ebling, F.J.G., Champion, R.H., and Burton, J.L., Eds., Blackwell Scientific, Oxford, 1986, p. 25. 4. Ebling, F.J.G., The biology of hair, Dermatol. Clin., 5, 467, 1987. 5. Rook, A. and Dawber, R.P.R., The comparative physiology, embryology and physiology of human hair, in Diseases of the Hair and Scalp, Rook, A. and Dawber, R.P.R., Eds., Blackwell Scientific, Oxford, 1982, p. 1. 6. Sato, Y., The hair cycle and its control mechanism, in Biology and Disease of the Hair, Koboti, T. and Montagna, W., Eds., University Park Press, Baltimore, 1986, p. 3. 7. Braun-Falco, O. and Kint, A., Dynamik des normalen und pathologischen Haarwachstums, Arch. Klin. Exp. Dermatol., 221, 75, 1966. 8. Orentreich, N., Scalp hair replacement in man, in Advances in Biology of Skin, Vol. 9, Montagna, W. and Dobson, R.L., Eds., Pergamon Press, Oxford, 1967, p. 99. 9. Barman, J.M., Astore, J., and Pecoraro, V., The trichogram of people over 50 years but apparently not bald, in Advances in Biology of Skin Hair Growth, Vol. 9, Montagna, W. and Dobson, R.L., Eds., Pergamon Press, Oxford, 1964, p. 211. 10. Trotter, M., The life cycles of hair in selected regions of the body, Am. J. Phys. Anthropol., 7, 427, 1924. 11. Blume, U., Ferracin, J., Verschoore, M., Czernielewski, J.M., and Schaefer, H., Physiology of the vellus hair follicle: hair growth and sebum excretion, Br. J. Dermatol., 124, 21, 1991. 12. Randall, V.A. and Ebling, F.J.G., Seasonal changes in human hair growth, Br. J. Dermatol., 124, 146, 1991. 13. Van Scott, E.J., Reinertson, R.P.A., and Steinmüller, R., The growing hair roots of the human scalp and morphologic changes therein following amethopterin therapy, J. Invest. Dermatol., 29, 197, 1957.
14. Pecoraro, V., Astore, J., Barman, J., and Araujo, C.S., The normal trichogram in the child before the age of puberty, J. Invest. Dermatol., 42, 427, 1964. 15. Pecoraro, V., Astore, J., and Barman, J., The normal trichogram of the pregnant woman, in Advances in Biology of Skin, Vol. 9, Montagna, W. and Dobson, J.M., Eds., Pergamon Press, New York, 1967, p. 203. 16. Barman, J.M., Astore, J., and Pecoraro, V., The normal trichogram of the adult, J. Invest. Dermatol., 44, 233, 1965. 17. Grosshans, E., Che pfer, M.P., and Maleville, J., Le trichogramme. A propos d’une méthode d’étude des cheveux, J. Med. Strasbourg, 378, 1972. 18. Metz, H.G. and Landes, E., Der Haarausfall und seine Untersuchung, Fortschr. Med., 88, 1327, 1970. 19. Zaun, H. and Ludwig, E., Zur Definition ungewöhnlicher Haarwurzeln im Trichogramm, Hautarzt, 27, 606, 1976. 20. Orfanos, C.E., Androgenetic alopecia, in Hair and Hair Diseases, Orfanos, C.E. and Happle, R., Eds., SpringerVerlag, Berlin, 1991, p. 485. 21. Jost, B., Meiers, H.G., Schmidt-Elmendorff, H., and Pfaffenrath, V., Trichometrische Quantifizierung und Verlaufsbeurteilung des Hirsutismus, Dtsch. Med. Wschr., 99, 2395, 1974. 22. James, K.C. and Rushton, D.H., Evaluation techniques for male pattern baldness, J. Am. Acad. Dermatol., 14, 849, 1983. 23. Orfanos, C.E. and Hertel, H., Haarwachstumsstörungen bei Hyperprolaktinämie, Z. Hautkr., 63, 23, 1988. 24. Sterry, W., Konrads, A., and Nase, J., Alopecie bei Schilddrüsenerkrankungen: Charakteristische Trichogramme, Hautarzt, 31, 308, 1980. 25. Orfanos, C.E., Meiers, H.G., Friedrich, H.C., Ludwig, E., Mahrle, G., and Zaun, H., Haarausfall, Trichogramm und hormonelle Haartherapeutika, Dtsch. Arztbl., 3603, 1974. 26. Blume, U., Föhles, J., Gollnick, H., and Orfanos, C.E., Genotrichoses: clinical manifestations and diagnostic techniques, in Proceedings of the 18th World Congress of Dermatology, New York, in press. 27. Barth, J.H., Measurement of hair growth, Clin. Exp. Dermatol., 11, 127, 1986. 28. Braun-Falco, O. and Fischer, C., Über den Einflub des Haarewaschens auf das Haarwurzelmuster, Arch. Klin. Exp. Dermatol., 227, 419, 1966. 29. Bassukas, I.D. and Hornstein, O.P., Effects of plucking on the anatomy of the anagen hair bulb. A light microscopic study, Arch. Dermatol. Res., 281, 188, 1989. 30. Braun-Falco, O. and Heilgemeir, G.P., The trichogram. Structural and functional basis, performance and interpretation, Sem. Dermatol., 1, 40, 1985. 31. Hayashi, S., Miyamoto, I., and Takeda, K., Measurement of human hair growth by optical microscopy and image analysis, Br. J. Dermatol., 25, 123, 1991. 32. Fiquet, C. and Courtois, M., Une technique originale d’appréciation de la croissance et de la chute des cheveux, Cutis, 3, 975, 1979.
Microscopy of the Hair: The Trichogram
33. Van Neste, D., Dumortier, M., and De Coster, W., Phototrichogram analysis: technical aspects and problems in relation to automated quantitative evaluation of hair growth by computer-assisted image analysis, in Trends in Human Hair Growth and Alopecia Research, Van Neste, D., Lachapelle, J.M., and Antoine, J.L., Eds., Kluwer Academic, Dordrecht, The Netherlands, 1989, p. 147. 34. Rushton, H., James K.C., and Mortimer, C.H., The unit area trichogram in the assessment of androgen-dependent alopecia, Br. J. Dermatol., 109, 429, 1983.
881
35. Meiers, H.G., Trichogramm (=Haarwurzelstatus, =Haarbild). Methode und Aussagefähigkeit, Akt. Dermatol., 1, 31, 1975. 36. Astore, J., Pecoraro, V., and Pecoraro, E.G., The normal trichogram in pubic hair, Br. J. Dermatol., 101, 441, 1979. 37. Saitoh, M., Uzuka, M., and Sakamoto, M., Human hair cycle, J. Invest. Dermatol., 54, 65, 1970. 38. Seago, S.V. and Ebling, F.J.G., The hair cycle on the human thigh and upper arm, Br. J. Dermatol., 113, 9, 1985.
and Computerized 103 Photographic Techniques for Quantification of Hair Growth D. Van Neste Skinterface, Tournai, Belgium
CONTENTS 103.1 Introduction ..........................................................................................................................................................883 103.2 Basics about Hair Structure and Function...........................................................................................................883 103.3 Hair Photography .................................................................................................................................................884 103.3.1 Search for Golden Standards.................................................................................................................884 103.3.2 Improving Hair Photography for Computerized Measurements ..........................................................887 103.3.3 Global Vision and Imaging Methods ....................................................................................................888 103.3.3.1 Categorical Classification Systems ......................................................................................888 103.3.3.2 Calibrated Scoring Systems .................................................................................................889 103.3.3.3 Global Photographs ..............................................................................................................889 103.3.4 Analytical Methods................................................................................................................................890 103.3.4.1 Phototrichogram: From Conventional PTGs to Contrast-Enhanced PTGs .........................890 103.3.4.2 Variants of PTG Methods.....................................................................................................891 103.3.4.3 Future Trends in Computerized Methods ............................................................................891 103.3.4.4 Detection of Nonvisible Hair: Less Contrasted and Thinning Hair....................................892 103.4 Conclusion............................................................................................................................................................892 References .......................................................................................................................................................................893
103.1 INTRODUCTION The present review will focus on basic principles involved in human hair evaluation using photographic and computerized methods. Needless to say, visible hair, hair growth, and regrowth are different in nature. Indeed, the visibility of hair will depend not only on the resolution power of the optics and camera, but also on the natural contrast between the object, i.e., hair (with every variation of the fiber components and optical interfaces), and background (including skin as a heterogeneous and variable background and extraneous material). Hair growth is a dynamic process that results in visible hair production at the skin surface reflecting cell proliferation, differentiation, and migration of trichocytes, i.e., epithelial cells that are under the influence of skin-deep dermal papilla. Hair regrowth is more related to the phenomenon of hair cycling (see later). All these biological events happen to occur in the deeper layers of the skin, i.e., in the hair follicle. As far
as hair is concerned, these processes result in the appearance and maintenance of one or more hair fibers at the surface of the skin. Finally, this is what matters as well to the patient as to scientists or observers interested in capturing the hair and analyzing its vital characteristics. Whether or not this biological event makes the person happy is a complex matter that will not be considered here.
103.2 BASICS ABOUT HAIR STRUCTURE AND FUNCTION Here we shall briefly review some basics about hair follicle structure and function necessary for understanding the fine-tuning of technical aspects of hair photography. Those who are acquainted with growth of hair and hair follicle cycling may easily proceed to the next section. Scalp hair is an appropriate example to introduce the concept of global and analytical methods while “zooming” into the field of hair (Figure 103.1). Scalp hair appears as 883
884
Handbook of Non-Invasive Methods and the Skin, Second Edition
1
2
1S
1R
2S
2R
SA n/cm2 RA
200 150 100 50 0
0
50
100
150
global maintenance. The key to the appropriate assessment of hair maintenance lies first in a thorough understanding of the hair cycle (for a review, see Reference 1). This is shown schematically in Figure 103.2 for one single hair follicle. On the scalp there are about 100,000 follicles. Importantly, not all follicles are active at the same time: some produce a hair, while others are resting. After this, hair is shed (exogen hair follicle) and eventually a new hair shows up. After a short time necessary to reach the scalp surface, hair becomes visible. Each follicle appears to function independently from its neighbors; i.e., the process of scalp hair growth is not synchronized. However, follicles in a certain field may express a common phenotype after exposure of some compounds, i.e., sex hormones. Some scalp areas, in genetically predisposed individuals, will show a phenotype of defective hair replacement, i.e., patterned balding (Figure 103.3). In other areas (axilla, face, chest, genital skin, etc.) other patterns will be expressed: the very thin hair follicles will now grow longer and coarser hair. As such, the clinical appearance of hairiness depends as much on regional modulation of the hair follicle activity, and of its cycle in particular, as on the number of hair follicles.
MA n/cm2
FIGURE 103.1 Zoom concept from global view to hair folliculogram. Global view of the top of the head in two male subjects with male pattern baldness.1,2 The top of the head is fragmented into smaller areas (squares) for scalp coverage scoring (for more details on SCS, see text and other figures). Clipping hair allows a more detailed record using close-up (or macro) photographs. Depending on the technique, the area of the target site ranges from 5 to 1 cm2 (white arrows; clipping and close-up photographs not shown). A small tattoo allows relocation of the target area (middle of white ring). Such tattoos can also be removed by a small 4-mm punch biopsy (not shown). We show images taken with a videomicroscope of the surface of the biopsy. This shows a very small number of hairs (1S less than 2S). The same scalp specimens are marked (red dots) and processed for the isolation of the dermal part of the hair fiber (1R and 2R). In the lower left panel we show the results of a validation test of a macrophotographic equipment and analysis system. After taking a macrophotograph for anagen hair counts (MA, n/cm2 46), tattoos were punched out and specimens were tested for anagen hair counts with a stereomicroscope on the surface or on the root view (respectively SA and RA, n/cm2). In most cases (3/4), counts of SA and RA were higher than those obtained with a close-up photograph (MA, n/cm2), showing that the macrophotographic system was less than optimal to detect all growing hair in vivo on the human scalp.4 Finally, the light microscope (lower right panel) gives a very detailed view of the hair as it emerges at the scalp surface (thick arrow) from the deeper parts of the hair infundibulum (three small arrows).
a stable mass of hair. It actually represents the cumulative end result of discrete changes of individual hair follicle dynamics: hair shedding and replacement resulting in
103.3 HAIR PHOTOGRAPHY 103.3.1 SEARCH
FOR
GOLDEN STANDARDS
There is not a single technical modality that will encompass with a sufficient degree of precision the many dimensions of hair.2,3 At the one end of the spectrum, as shown in Figure 103.1 and Figure 103.3, global viewing is very often used for demonstration purposes, i.e., to document health authorities or individual patients. Global viewing, even when highly standardized procedures are being used, does not resolve the question of hair cycling changes. At the other end of the spectrum (Figure 103.1), the microscope has a superb resolution power when a single hair fiber or follicle is concerned. A comparison between stereomicroscopic viewing of scalp samples and photographs of the same sites taken with commercially available macrophotographic equipment showed that the number of growing hair counted on the surface view or from the root view were always higher than the counts generated from printouts.4 However, one may argue that those small numbers generated from scalp biopsies (4 mm diameter) may not be clinically representative. A larger field containing over 100 hair fibers would certainly be more representative,5 but many clinical research programs and investigators do forget that one may not expand the size of the target area without affecting the resolution of the image. A proper balance between sample size and the optical resolution power must be established before launching a hair photographic method.
Photographic and Computerized Techniques for Quantification of Hair Growth
885
A 1 2 3 4 5 6 a
b
c
d
e
f
g
h
i
a
b
c
d
e
f
g
h
i
B 1 2 3 4 5 6
FIGURE 103.2 Hair cycle. The hair growth phase (anagen) during which hair is visible at the skin surface and growing is shown in A, while the apparent resting phase of the hair cycle (telogen) is shown in B. The follicle is represented schematically and the essential components are numbered in the legend: A: pigmented (1) and less pigmented (2) hair shaft produced during the growth phase. Inner root sheath (IRS) (3), club hair (4), stelae (5), dermal papilla (6). B: Inner root sheath (1), tip of new anagen hair (2), release of old club hair (4), new anagen grows out of IRS (5) and is more pigmented (6). A: From growth to rest. The same hair follicle is represented at various times (a to i) at the very end of its growth phase. At the skin surface, there is normal pigmented hair production (time between a–b and b–c is 24 hours). The increased length of visible hair represents the daily hair production from which the linear growth rate can be calculated (μm/24 hours). Growth takes place at the bottom of the hair follicle. The space immediately below the bottom of the hair fiber is where cell proliferation takes place, i.e., hair matrix. This is very close to the dermal papilla. Then the pigmentation of the newly synthesized hair shaft is decreased (c). This early event announces the hair follicle regression and is followed by terminal differentiation of cells in the hair matrix (d) that will turn into the club hair formation (d to g). The shrinking dermal papilla (d) begins an ascending movement together with the hair shaft (time between d and h, 21 days). This characterizes the catagen phase (d to h) with an apparent elongation of the hair fiber. This is not growth, but reflects the outward migration of the hair shaft. What is left over after disappearance of the regressing follicle is usually referred to as streamers or stelae (f to i). The true resting stage begins when catagen is completed, i.e., when the dermal papilla abuts to the bottom of the club hair. As for now, no hair elongation is observed at the surface (g to i). B: From rest to growth. During this stage, one notices absence of hair growth at the skin surface (a to e), but significant changes occur in the deeper parts of the hair follicle. The dermal papilla expands and attracts epithelial cells from the bulge (stem cell zone) in a downward movement (a and b). To create space, previously deposited materials have to be digested (a to c). The epithelial cells then start differentiation in an orderly fashion, starting with the inner root sheath (IRS forms the initial cone in b and then a cuff in c) containing the tip of the newly formed hair fiber. This tip is made of the cuticle and a very thin hair cortex, usually non- or less pigmented (c to e). The hair fiber is released from the IRS cuff when it reaches the isthmus (f to g). The old resting hair remains in the hair follicle for approximately 1 to 3 months (a to e), then a detachment process transforms the old follicle into exogen and, as a consequence, releases an exogen hair (f). Such exogen hair may stick in the follicle for a short time before being shed (f and g). The shiny root end of the shed hair is the club, visible with the naked eye. Before, during, or after hair shedding there may be replacement by a new, gradually thicker and more pigmented hair shaft (e to g). Indeed, under physiological conditions, the follicle may proceed immediately or only after some lag time with new hair production (from b to g; up to 90 days). In conditions like androgenetic alopecia, the shed hair may stick in the follicle for a longer time. Stasis of exogen hair may turn into trichostasis. This reflects an abnormal accumulation of nonadherent or loosely attached elements. Also, there may be a much longer interval before regrowth is recorded at the scalp surface (example, steps b to g may take 6 months). At the earliest visible stages, i.e., when the matrix of the new anagen follicle is again deeply set into the dermis, one notices at the scalp surface a thin, usually nonpigmented hair tip that is seen first (h), soon followed by a thicker, more pigmented and faster-growing hair fiber (i). This sequence, of course, depends on the many regulatory factors controlling the hair follicle activity during its replacement.
886
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 103.3 Patterned hair loss. The present classification shows patterns that affect the scalp of genetically predisposed male subjects. Such patterns develop after puberty, when deficient hair replacement ultimately results in bald appearance. They are subdivided in six categories of increasing severity (I to VI), from mild to severe balding. The anterior pattern (A) indicates a backward progression of hair follicle miniaturization that starts from the frontal hairline. The vertex type (V) indicates isolated regression occurring on the vertex, but this is usually combined with some involvement of the frontal temporal areas. As in Figure 103.1, fragmentation (square areas) help in evaluating the severity — in terms of extension or progression — of the balding process.
One way to evaluate 100% of hair growth potential is to take an exhaustive sample. This means taking a skin biopsy; i.e., there is no way for a single hair follicle to
escape the analysis. After fixation and sectioning the scalp sample from top to bottom, one may examine and analyze all sections with a light microscope. A trained observer
Photographic and Computerized Techniques for Quantification of Hair Growth
FIGURE 103.4 Phototrichogram in a young male with patterned alopecia. The contrast enhanced phototrichogram (CE-PTG) technique (upper panels) matches perfectly histology (lower panels) for hair growth measurement. Evaluation of hair growth with the phototrichogram: At time 0 (day 0), the hairs are clipped close to the scalp surface and photographed. After a given time (48 hours in the author’s experience, day 2), the same scalp site is photographed again. Substantial elongation at day 2 reflects hair growth and indicates anagen in thick (1, 2, 5) and thin (7) hair. Moderate elongation reflects catagen in a thinning hair (3). No elongation reflects resting in thick (4) and thinning (6) hair. After shedding the thick (4) or the typical tiny (9) resting hair, an empty follicle in the exogen stage will remain visible (8 arrows at day 0 and day 2 and upper arrow in histology 8, lower panel). The formation of new follicles may initiate anagen, for a thick (lower arrow) and thin hair (upper right arrowhead) can be seen only with histology (8). Tracking of such empty follicles with CEPTGs will show newly growing hair in some days or weeks (see Figure 103.2B: steps f, g, h, i).
887
will easily stage the hair follicles into anagen, telogen, or catagen. After tracking hair follicles through several hundreds of sections, a complete picture of the hair follicles will be available. Because of its high magnification and the exhaustive sampling, quantitative microscopic analysis of scalp biopsy specimens may serve the purpose of validating other measurement methods such as counting hairs and evaluating growth on scalp hair photographs6 (as shown already in Figure 103.1 and further detailed in Figure 103.4). Taking repeat biopsies is definitely not practical for hair growth monitoring purposes. Hair density or number of hair per unit area is usually reported as number per square centimeter. It reflects the number of functionally active follicular units whether growing (anagen) or not (telogen). Under physiological conditions, i.e., the long duration of anagen phase and the comparatively short duration of telogen, we know that most scalp follicles are engaged in anagen. This will produce long and clearly visible hair at the scalp surface, but this may change dramatically in some hair disorders, leading to hair loss and balding. The percentage of anagen follicles properly reflects the time during which hair follicles are engaged into the growth phase. The anagen/telogen ratio is also often found in the literature. Many clinicians are not aware that this is true only when an exhaustive count of hair or hair follicles is made. This means including those follicles that are growing a fiber that is not yet visible at the scalp surface (anagen III to V; see Figure 103.2B, steps c to f), those that are subject to miniaturization, or those that produce fibers that may not be seen with conventional photography. Therefore, classification of follicles into terminal, intermediate, and miniaturized has been proposed. There is no consensus, however, on the definitions of these categories. Many based their assumptions on the thickness of hair fibers, but we argue that not enough attention has been devoted in the past to variables such as growth rate, natural pigmentation,7 and eventually combined effects such as relative resistance to clipping of thinner fibers influencing the remaining length after clipping, clustering of thin and thick fibers, etc.7–9 From the above sections it becomes clear that the results of an assay largely depend on the technology that was used for measuring hair parameters. Before discussing further the aims of computerizing, we wish to illustrate some tools that are commercially available or systems that can be delivered in the context of the installation of a hair clinic with licensing of technology or on the occasion of a clinical trial project (Figure 103.5).
103.3.2 IMPROVING HAIR PHOTOGRAPHY COMPUTERIZED MEASUREMENTS
FOR
Computers may help in performing routine tasks, but they need to be fed very high quality documents in order to
888
Handbook of Non-Invasive Methods and the Skin, Second Edition
Ga
Gb Ma
Mb
Mc
FIGURE 103.5 Equipment for global and close-up hair photography. Equipment for global scalp viewing (Ga from Skinterface or Gb from Canfield) or for macrophotography (Ma from Canfield or Mb (historical) and Mc from Skinterface). Ga also shows a model used for calibration purposes and the comb that is always necessary to organize the mass of scalp hair properly. Such materials are commercially available at Canfield or can be licensed by Skinterface on the occasion of a clinical trial project. Ga and Mc (developed in 2002) show the recent trend toward miniaturization (compare with Gb and Ma, Mb).
generate acceptable numbers from figures: there is no brain other than the human observer to interpret the clinical value of an image. In the description of noninvasive photographic methods that can be used in vivo in human subjects, we propose to distinguish between global and analytical methods. Global methods apprehend at once various factors responsible for the hair area under examination but cannot resolve the details, as opposed to the analytical methods that shall be described later on. The latter have a unique advantage over the former, because they provide a series of individual measures that reflect the structure/function of the hair follicle as an organ. Such data are subject to critical analysis. By combination of all the analytical data, one can generate a global value, but the reverse is not true. Also, another disadvantage of published global methods is that they are usually not calibrated. As such, it remains difficult to derive clinical relevance of statistically significant findings.
103.3.3 GLOBAL VISION
AND IMAGING
METHODS
103.3.3.1 Categorical Classification Systems Distinctive patterns of defective scalp skin coverage or alopecia have been identified by clinicians as patterns. You could practice in any public space (mall, theater, etc.) and see for yourself what the limits of this method are (Figure 103.1 and Figure 103.3). In the author’s hands, there is a very large variation between sessions when the same views of the top of the head of male subjects with patterned hair loss were presented several times in a randomized sequence (unpublished observations). Also, intervals between successive severity scales
10
20
30
11
21
31
FIGURE 103.6 How do we standardize global scalp hair imaging? In this setting, great care has been taken in the preparation of the background and positioning of the patient (1). Imaging (2) involves adjusting the subject with the headgear to the positioning device. Once the light box is switched on and the camera properly inserted into the light box, one captures an image appearing on the computer screen. Examples show the views of the top of the head (10 to 30), on which a grid can be placed as an overlay (11 to 31).
may not be equal. The consensus among experienced clinicians is that the schemes are of little help in measuring the dynamics of hair changes — whether growth or loss — over time. Based on these patterns, fragmentation of the scalp was proposed on the basis of anatomical criteria.10 However, no evidence was presented to support this new belief. The use of a noncalibrated density scale casts some doubts as to the practical application in real time or “as a bedside” measurement tool.10 In order to circumvent several of the weaker aspects of these methods (lack of standardization, lack of calibration, transposing shaved hair density skinhead type of density patterns onto real heads of patients with a great diversity of hairstyles, etc.), we proposed, as shown in Figure 103.1, Figure 103.3, and Figure 103.6, to fragment the top of the head into a number of fixed and predefined subunits. This helps to focus the attention of the observer in clearly outlined spaces where patterning results in various degrees of defective scalp coverage. We
Photographic and Computerized Techniques for Quantification of Hair Growth
will describe results obtained using this novel approach in greater detail in the next section. 103.3.3.2 Calibrated Scoring Systems A system has been developed where examination of the scalp surface proceeds through fixed external standards. The top of the head is separated into small units (equal size in the projection plane) with no anatomical correspondence of the bones of the skull (e.g., frontal, parietal, etc.) or the immense variation of patterning in human subjects. The relative difficulty to detect scalp skin between the hair is translated into scalp coverage scores (scalp coverage scoring (SCS) score 0 means there is no difficulty and score 5 means that it is barely possible to detect scalp skin through the hair). This is rated — under strict distance and angle control — against objective rulers of density (method and equipment for measuring coverage density of a surface, Skinterface, application patent PCT/EP 01/06970, 2001). The reproducibility (intra- and interobserver) is very high (correlation factors >0.9).11 Several studies using scores generated in the clinic (real-time measurement) and on global photographs have now been performed either as a single center or in the context of multicenter placebo-controlled trials (H.A.I.R. Technology® protocols 02P08, 03P17, and 03P22) using known drugs or new compounds (phases II and III). When all identified sources of variation are kept under control, the variation of scores is less than 5%.11 During calibration studies, the SCS was also correlated with clinically relevant hair parameters such as proportion of anagen hair or density of thinning hair.12 As such, SCS allows quantification of clinically relevant changes (Figure 103.3). The main advantage of the SCS method is its easiness to use in the hair clinic, where it helps in identifying subjects who respond to specific treatment regimens in vivo. In a 1-year period, a clinician is able to disclose hair loss in untreated (randomized placebo-controlled study in vivo) patients.13–15 In the longer run, the clinician is more accurate in detecting clinically relevant changes due to lack of active treatment (e.g., placebo effect), as compared to maintenance or improved scalp coverage under active treatment. Besides the real-time quantitative observations at the hair clinic or at bedside, SCS can also be applied to standardized global scalp photographs, as will be discussed in the next section. 103.3.3.3 Global Photographs Global photography has been a significant step forward in scalp hair documentation by creating a permanent record.
889
Global photography apprehends all factors involved in hairiness at once and can be used for drug efficacy evaluation providing that adequate scalp preparation and hairstyle are maintained throughout the study. This is the most patient-friendly photographic method. This method is used in the clinic under standardized conditions of exposure.16 Processing and rating have to be performed under controlled (e.g., blinded as to treatment or time) conditions. Trained experts could generate reproducible data. It appears that paired comparison of global photographs is more realistic in its appreciation of hair growth after drug treatment than subjective evaluations of investigators and patients.17 As sets of photographs can also be scored individually with the SCS system — one photograph at a time and in circumstances one usually encounters in day-to-day practice (see earlier) — the SCS could provide real-time scalp coverage values. Such a quality control before inclusion of subjects into clinical trials may be of great value. Indeed, quantifying disease severity and precise evaluation of distribution patterns are immediately available, as opposed to global photographs that need expert evaluation after completion of the study. Indeed, in practice, the quality control happens only after processing in a special laboratory. The quality of photographs is evaluated in terms of optics (focus, contrast, etc.), but not in terms of hair patterning or alopecia. The techniques of global picturing and density documentation appear to be important issues for future development, and work is in progress in our laboratory. Some computer approaches have already been published,18,19 but can hardly be used on real documents. In general, paired rating of global photographs and SCS are two techniques that are significantly correlated inasmuch as all factors of variation (hair length, room and scalp light, standardized combing, etc.) are well controlled (for some examples of technological problems, see Figure 103.7). SCS has the advantage of helping the clinician in his real-time evaluation of efficacy in vivo before analyzing comparatively scalp views on global photographs. Incidentally, the accuracy of the image setting used for the SCS three-dimensional control method (as demonstrated in Figure 103.6) allowed us to detect subtle changes in distance or angle of vision between pairs of before and after photographs. Such deviations remain usually unnoticed even to a panel of internationally recognized independent experts (as a practical exercise, trace a grid on a transparency and apply the grid on the photographs appearing in panels 1 and 2 of Figure 103.7; personal unpublished data).
890
Handbook of Non-Invasive Methods and the Skin, Second Edition
1
2
3
4
FIGURE 103.7 Coping with technical and styling variability of global photographs. These two sets of before and after pictures illustrate difficulties encountered in evaluation of time or treatment-related changes of scalp coverage. The commercially available Canfield system (1 and 2) and clinical trial set developed at Skinterface (3 and 4) illustrate some differences between the systems (backscattering of light, background, inclusion of color codes during imaging, etc.). The vertex view (1) turned into a top-of-the-head view (2) at the follow-up visit. The technician did not properly position the head of the patient on the two sessions (estimated angle error, ± 45˚). The two top-of-the-head views (3 and 4) illustrate the difficulty of managing properly the hairstyle. Although the technical quality is almost perfect (see overlay of targets, distance, angle, color codes, etc.), any possible change of hair growth is masked by the change in hairstyle.
103.3.4 ANALYTICAL METHODS 103.3.4.1 Phototrichogram: From Conventional PTGs to Contrast-Enhanced PTGs 103.3.4.1.1 What Did We Learn with the Conventional PTG? The basic principle of the phototrichogram (PTG) consists of taking a close-up photograph of a certain area of the scalp. The hair is cut very close in preparation for the first photograph, followed by repeat photographic documentation after a certain period. This period of time should be long enough to allow the growth of a hair segment (time window that is usually between 24 and 72 hours), but not too long, in order to prevent outgrowth or too much overlapping of growing hair. The growth is then evaluated by comparing the two pictures. Hairs that have grown are in
the anagen phase, and those that have not are in the telogen phase (Figure 103.4). Analytical methods that document major aspects of the hair cycling process have been developed over the years20–26 and are subject to continuous reevaluation and improvement.27–29 Some PTG data have been computed for mathematical modeling so as to develop virtual patterns of defective hair replacement, mimicking those observed in the hair clinic.30 Nevertheless, neither all hair fibers nor all productive hair follicles are taken into account during the conventional PTG procedures. In some assays, we noticed that the length of hair significantly affected its visibility. As growing hair will become more visible on the second photograph, a bias was detected (unpublished data). In 1989, we devised scalp immersion photography,31 or proxigraphy, 22 well before epiluminescence microscopy became the accepted term. After comparison with another, more invasive method,28 some weaknesses have been considered with great care and considered contrast enhanced (CE) as a further improvement.6,27 This brought the phototrichogram technique (CE-PTG) to a resolution almost equal to that of transverse microscopy of scalp biopsies, which are usually considered the golden standard.32–37 So far, the CE-PTG remains the only method that has documented all transitions of thick and thinning follicles, from anagen through catagen into telogen phases, on a follicular basis.6 In early stages of androgenetic alopecia (AGA) in man, this sensitive method12 was able to detect a subclinical phase of AGA with obvious shortening of the anagen phase in the absence of hair miniaturization. This preclinical stage evolves into patterning, i.e., the fullblown phenotype associated with a further shortening of the growth phase along with reduction in hair diameter.22–29,38–42 The follicular regression process finally results in production of clinically nonvisible hair.43 The CE-PTG method, as a refined, noninvasive, and validated technique, could be used for calibration purposes for any new method that would be developed for use in the skin and hair clinic. Hence, we identified a new global measurement method that integrates cumulative hair growth and reflects clinically relevant scalp skin coverage, as described earlier (patent application PCT/EP 01/06970, June 2001). 103.3.4.1.2 What Can We Learn to Measure with the CE-PTG? The assessment is made on one or a number of predefined scalp sites considered representative of the condition. The data that can be generated from a PTG are total number of hair present in a certain area, i.e., hair density (n/cm2), the percentage of hair in the growth phase (anagen %), the linear hair growth rate (LHGR; mm/day), and the hair thickness.
Photographic and Computerized Techniques for Quantification of Hair Growth
Thickness can be measured on hair clippings, on scalp biopsies, and on scalp hair photographs. It may seem trivial to state that the hair diameter evaluation is more precise with the microscope. These measures reflect diameter whenever the fiber section is circular, and in all races inasmuch as the hair remains thinner than 40 μm.44 In Caucasians, thicker hair loses this property and the section becomes elliptical. In Africans, the hair flattens even more, and we are not aware of any published data on the value of hair thickness evaluation on scalp photographs of those subjects. The same holds true for other body sites where curly and flattened hair are the rule rather than the exception. Nevertheless, we found a significant correlation between paired microscopic measurements and thickness evaluation on Caucasian scalp hair photographs in the range of <20- to >100-μm-thick hair fibers (less than 1% variation in a test using multiple technicians; personal unpublished data). By this approach measurements relate to sample population on a hair-to-hair basis and temper the apparent statistical superiority of microscopic measures. Indeed, we noticed that the sample collection and processing of clipped hair for microscopic display (and probably any other type of measurements) may be difficult to standardize and become sources of error.45 The method was applied with success to various genetic conditions, including male pattern baldness in vivo.3 This technique was able to document the earliest changes of hair growth in AGA in man. From shortening of anagen duration of thick hair, regression appears to evolve in sequence into a phase of further shortening associated with slower growth rates and more severe thinning, turning finally into a stage of reversible miniaturization without production of any visible hair. At this stage, one speculates that drug response is still possible before the follicle drops into total irreversible atrophy, though without scarring. 103.3.4.2 Variants of PTG Methods Subtle modifications in the preparation of the target site can help identification of the hair in the growing from resting phases, especially when less than optimal magnification lenses are used (e.g., less than 3×24). Indeed, after clipping the hair short (first step of the PTG as control for density or hair counts), a close shave will further reduce the visible length of the hair fiber. Then, usually 3 rather than 2 days later, the second photograph is taken. A new hair count of the long hair fibers reflects anagen hair follicles. This procedure has been used to monitor changes occurring after finasteride in man with AGA, demonstrating a significant induction of growth compared with placebo.46 When the photographic camera is replaced by a video or charge coupled device (CCD) camera equipped with
891
specific lenses, other variants of PTG recording are obtained. In fact, reports in Orientals and Caucasians were published. In the latter subjects, the contrast between hair and scalp seems favorable for the application of this method, and the low figures of hair density could possibly be racial in origin.21 As shown already,27,45 the use of CE is advisable — especially for Caucasian hair — even with the use of computer applications for textile fiber analysis47 or hair recognition software already developed for other applications.48 Such a system has been proposed49 and the promoters of the method recognize that the explored area is small (less than 0.25 cm2). This is yet another source of variability because the number of hairs so analyzed is definitely less than 100 and drops below statistically acceptable population sample size.5 Furthermore, all of these automated systems generate numbers without securing that all hairs in the field were detected. Various commercial brands have been proposed, such as Trichoscan49 or Capillicare (data in file), and customers reported lack of satisfaction with such methods because the dream of an easy and fast method vanishes as it remains time-consuming. The method analyzes only a limited number of clinical situations, and even when all clinical conditions are suitable for the computer, it does not generate all data that may be of interest to the observer in terms of diagnosis, prognosis, or any other clinically relevant information. More importantly, independent clinical observers50 also state that “the potential of computerassisted technology in this field is yet to be maximized and the currently available tools are less than ideal.” 103.3.4.3 Future Trends in Computerized Methods More than a decade ago, we listed the different problems arising when automated computer-assisted image analysis (ACAIA) comes into the scope of hair growth measurement.51 Some problems have been solved. Accurate analysis still requires expert human intervention during the sampling and processing of images. Today we are easily coping with the three dimensions involved in hair; we mastered the requirements in terms of enlargement factor or pixel size, immersion, and polarized light. We also wiped out the problem of backscattering of light and other problems linked to presence of sebum droplets, sweat, and scales. More recent developments helped us to clear the scene from any other loosely attached elements, such as exogen hair (application patent process, method and apparatus for removing nonadherent elements from the skin of living beings and measuring the hair loss of living beings, Skinterface, patent PCT/EP 02/06434, June 12, 2002), and add information on trichostasis, a largely unsuspected phenomenon that significantly influences hair counts. Increasing contrast between scalp and hair allows us to express data in a single case observation or to perform accurate
892
Handbook of Non-Invasive Methods and the Skin, Second Edition
time-course studies of hair growth changes using highly standardized operating and monitoring procedures. As an example of the technological resolving power, we recently discussed hair pigmentation during aging and how it might affect the outcome of a clinical trial.7,52 Initially, nonvisible hair under conventional PTG may become visible for drug effects such as increased pigmentation, initiation of shortlived but thick anagen hair, etc. The spectrum of follicular bioresponses is actually underestimated and requires sophisticated investigations.
1
3
5
2 4 a
b
c
d
e
6 78
103.3.4.4 Detection of Nonvisible Hair: Less Contrasted and Thinning Hair Clinicians occasionally express their fears about new information generated by these up-to-date technologies, such as taking into account that thin hair is becoming visible.10 This clearly illustrates that the message that contrast enhancement also makes thick hair more visible did not come through. There is only one way to know whether miniaturized hair has any potential meaning or is just useless as a potential site for drug action: use a method that is able to measure their presence and monitor their potential modification over time. Anything else should be considered as speculative issues contributing to maintaining hair science and technology in the marshlands of “tricho-quackery.” As a practical example of a critical comparative study, we refer to the illustrations and comments in Figure 103.8. Therefore, we conclude that fully automated analysis systems,31 notwithstanding the apparent easiness of more recent systems,49 remain generally unsatisfactory when a detailed description of hair variables, such as hair counts, diameter — especially for thinning hair — and growth rate, is required. Biologically significant data in terms of follicle distribution, productivity, cycling, regression, or progression are not yet published — but will be in the near future — with our updated technologies. We do hope that the work done during the last decade will definitely bring such systems to the status of medically acceptable as diagnostic, prognostic, and therapeutic monitoring tools.
103.4 CONCLUSION A bold statement would be to say that the assessment of hair loss requires some experience and a lot of technological effort in order to grasp all the parameters involved in hair measurement. There was a time when some colleagues argued that measurement methods were not necessary, as the patient could tell when hair was growing or not. It is obvious that this is untrue as soon as one enters the continuum of the hair replacement process, especially when some hair is still present. Such a statement also looks outdated when one acknowledges the
FIGURE 103.8 Quality process evaluation comparing manual vs. automated hair analysis software. Here we show a typical example of quality process evaluation — as it is usually practiced in our laboratory — comparing results of our routine hair processing method vs. those generated by the automated hair analysis software. Image processing results: An image (a, source image) was downloaded in July 2003 from a website (trichoscan.com) displaying a full automatic system for hair analysis. Accordingly, after the automated hair selection (b, yellow outline of selected elements by software) and analysis (c, binary image of selected elements), the same source document (a) was also submitted to a technician for routine processing in our laboratory. The technician had no information about the origin of the document or results of automated analysis. This modality is part of our routine processing quality control (d). Our technician identified 43 hair segments in the circular outline of that image. Fourteen hair fibers had a thickness estimated to be less than 40 μm and 29 were considered thicker hair (40 μm). Quality analysis of process: At the final revision by the author, discrepancies were noted between the two processing modalities. They are outlined visually (e) against a pink background. One very thin hair partly hidden within a trio (white rectangle with arrowhead) was missed by both measurement systems. A display of hair results from automated analysis of image (a) is shown in (f): 29 hairs were counted in the 0.227 cm2 area (parts of circular outline), generating a density of 127.8 hair/cm2. The automated system also missed seven other hairs that were identified by our technician (empty numbered spaces in (e)). Automated counts also included three hairs crossing the edges (white rectangles, top (one hair) and bottom (two hairs) of (e)). Hairs that stuck together were not individualized by the automated system. Results: While the automated analysis counted 29 hairs, we found 42 in the same area. The technician’s error (1/43) is much lower than the automation error (12, or 13/43), i.e., 2.3% instead of 30.2% of the total sample. Conclusion: Hair evaluation as performed with automated systems does not match the quality standards prevailing in our laboratory. Many problems identified in 198931 still cause problems for automatic evaluation in 2004.
Photographic and Computerized Techniques for Quantification of Hair Growth
importance of the placebo effect. Indeed, subjective evaluation may reach 60% or more satisfaction, while significantly decreased hair counts clearly document the natural worsening of the condition.17 Our experience points to the fact that a combination of a highly sensitive and precise analytical approach with a global calibrated method seems advisable in the context of kinetic monitoring of hair growth and hair loss in the hair clinic in general, and this is warmly recommended in the context of efficacy analysis of new (and recognized) compounds in future clinical trials.
REFERENCES 1. Stenn KS, Paus R. Controls of hair follicle cycling. Physiol Rev 81:449–494, 2001. 2. Sinclair R, Jolley D, Mallari R, Magee J, Tosti A, Piracinni BM, Vincenzi C, Happle R, Ferrando J, Grimalt R, Leroy T, Van Neste D, Zlotogorski A, Christiano AM, Whiting D. Morphological approach to hair disorders. J Invest Dermatol Symp Proc 8:56–64, 2003. 3. Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A. Hair Science and Technology. Skinterface, Tournai, 2003. 4. Van Neste D. Folliculogram demonstrates more anagen hair roots in male androgenetic alopecia after one year treatment with finasteride 1mg/d. In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 311–316. 5. Van Neste DJJ. Hair growth evaluation in clinical dermatology. Dermatology 187:233–234, 1993. 6. Van Neste DJJ. Contrast enhanced phototrichogram (CE-PTG): an improved non-invasive technique for measurement of scalp hair dynamics in androgenetic alopecia — validation study with histology after transverse sectioning of scalp biopsies. Eur J Dermatol 4:326–331, 2001. 7. Van Neste D. Thickness, medullation and growth rate of female scalp hair are subject to significant variation according to pigmentation and scalp location during ageing. Eur J Dermatol 1:1–2, 2004. 8. Van Neste D, Hughes TC, Herd H, Gibson WT. Thickness, medullation and growth rate of human scalp hair are subject to significant variation according to pigmentation and scalp location. Australas J Dermatol 38 (Suppl. 2):19, 1997 (abstract). 9. Van Neste D, Demortier Y. Detailed monitoring of hair cycle transitions in vivo using contrast enhanced phototrichogram (CE-PTG). In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 211–222. 10. Olsen EA. Current and novel methods for assessing efficacy of hair growth promoters in pattern hair loss. J Am Acad Dermatol 48:253–262, 2003.
893
11. Van Neste D, Leroy T, Sandraps E. Validation and clinical relevance of a novel scalp coverage scoring method. Skin Res Technol 9:64–72, 2003. 12. Sandraps E, Leroy T, Demortier Y, Van Neste D. Validation and clinical relevance of a novel scalp coverage scoring (SCS) method. In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 255–270. 13. Van Neste D. My management plan of the male patient with androgenetic alopecia. In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 301–310. 14. Van Neste D. Scalp coverage scoring (SCS) documents natural worsening within less than 12 months in a majority of men with androgenetic alopecia (AGA). In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 271–276. 15. Van Neste D, Leroy T, de Ramecourt A. Hair removal and hair follicle targeting. In Hair Science and Technology, Van Neste D, Blume-Peytavi U, Grimalt R, Messenger A, Eds. Skinterface, Tournai, 2003, pp. 401–412. 16. Canfield D. Photographic documentation of hair growth in androgenetic alopecia. Dermatol Clin 14:713–721, 1996. 17. Kaufman KD. Long-term (5-year) multinational experience with finasteride 1 mg in the treatment of men with androgenetic alopecia. Eur J Dermatol 12:38–49, 2002. 18. Gibbons RD. Computer-aided quantification of scalp hair. Dermatologic Clinics 4, 627–640, 1986. 19. Gibbons RD, Fiedler-Weiss VC, West DP, Lapin G. Quantification of scalp hair: a computer-aided methodology. J Invest Dermatol 86:78–82, 1986. 20. Guarrera M, Ciula MP. A quantitative evaluation of hair loss: the phototrichogram. J Appl Cosmetol 4:61–66, 1986. 21. Hayashi S, Miyamoto I, Takeda K. Measurement of human hair growth by optical microscopy and image analysis. Br J Dermatol 125:123–129, 1991. 22. Van Neste DJJ, Dumortier M, De Brouwer B, De Coster W. Scalp immersion proxigraphy (SIP): an improved imaging technique for phototrichogram analysis. J Eur Acad Dermatol Venerol 1:187–191, 1992. 23. Courtois M, Loussouarn G, Hourseau C, Grollier JF. Hair cycle and alopecia. Skin Pharmacol 7:84–89, 1994. 24. Van Neste D, De Brouwer B, De Coster W. The phototrichogram: analysis of some technical factors of variation. Skin Pharmacol 7:67–72, 1994. 25. Guarrera M, Rebora A. Anagen hairs may fail to replace telogen hairs in early androgenic female alopecia. Dermatology 192:28–31, 1996. 26. Guarrera M, Semino MT, Rebora A. Quantitating hair loss in women: a critical approach. Dermatology 194:12–16, 1997. 27. Blume U, Ferracin I, Verschoore M, Czernielewski JM, Schaefer H. Physiology of the vellus hair follicle: hair growth and sebum excretion. Br J Dermatol 124:21–28, 1991.
894
Handbook of Non-Invasive Methods and the Skin, Second Edition
28. Rushton DH, De Brouwer B, De Coster W, Van Neste D. Comparative evaluation of scalp hair by phototrichogram and unit area trichogram analysis within the same subjects. Acta Derm Venereol (Stockh) 73:150–153, 1993. 29. Van Neste DJJ, Rushton DH. Hair problems in women. Clin Dermatol 15:113–125, 1997. 30. Halloy J, Bernard BA, Loussouarn G, Goldbeter A. Modeling the dynamics of human hair cycles by a follicular automaton. Proc Natl Acad Sci USA 97:8328–8333, 2001. 31. Van Neste D, Dumortier M, De Coster W. Phototrichogram analysis: technical aspects and problems in relation to automated quantitative evaluation of hair growth by computer-assisted image analysis. In Trends in Human Hair Growth and Alopecia Research, Van Neste D, Lachapelle JM, Antoine JL, Eds. Kluwer Academic Publishers, Lancaster, U.K., 1989, pp. 155–165. 32. Headington JT. Histological findings in androgenic alopecia treated with topical minoxidil. Br J Dermatol 107(Suppl. 22):20–21, 1982. 33. Headington JT, Novak E. Clinical and histologic studies of male pattern baldness treated with topical minoxidil. Curr Ther Res 36:1098–1106, 1984. 34. Whiting DA. The value of horizontal sections of scalp biopsies. J Cutan Aging Cosmet Dermatol 1:165–173, 1990. 35. Headington JT. Telogen effluvium. New concepts and review. Arch Dermatol 129:356–363, 1993. 36. Whiting DA. Diagnostic and predictive value of horizontal sections of scalp biopsy specimens in male pattern androgenetic alopecia. J Am Acad Dermatol 28:755–763, 1993. 37. Whiting DA. Scalp biopsy as a diagnostic and prognostic tool in androgenetic alopecia. Dermatol Ther 8:24–33, 1998. 38. Tsuji Y, Ishino A, Hanzawa N, Uzaka M, Okazaki K, Adachi K, Imamura S. Quantitative evaluations of male pattern baldness. J Dermatol Sci 7:136–141, 1994. 39. Courtois M, Loussouarn G, Hourseau C, Grollier JF. Ageing and hair cycles. Br J Dermatol 132:86–93, 1995. 40. Courtois M, Loussouarn G, Hourseau S, Grollier JF. Periodicity in the growth and shedding of hair. Br J Dermatol 134:47–54, 1996.
41. Ishino A, Uzuka M, Tsuji Y, Nakanishi J, Hanzawa N, Imamura S. Progressive decrease in hair diameter in Japanese with male pattern baldness. J Dermatol 24:758–764, 1997. 42. Rushton DH. Androgenetic alopecia in men: the scale of the problem and prospects for treatment. Int J Clin Pract 53:50–53, 1999. 43. Dawber R, Van Neste D. Hair and scalp disorders . In Common Presenting Signs, Differential Diagnosis and Treatment. Martin Dunitz, London, 1995. (Note: New edition planned in 2004.) 44. Rushton DH. Chemical and Morphological Properties of Scalp Hair in Normal and Abnormal States. University of Wales, Cardiff, 1988. 45. Leroy T, Van Neste D. Contrast enhanced phototrichogram pinpoints scalp hair changes in androgen sensitive areas of male androgenetic alopecia. Skin Res Technol 8:106–111, 2002. 46. Van Neste D, Fuh V, Sanchez-Pedreno P, Lopez-Bran E, Wolff H, Whiting D, Roberts J, Kopera D, Stene JJ, Calvieri S, Tosti A, Prens E, Guarrera M, Kanojia P, He W, Kaufman K. Finasteride increases anagen hair in men with androgenetic alopecia. Br J Dermatol 143:804–810, 2000. 47. Van Neste D, De Coster W. Phototrichogram: technical problems in relation with automated quantitative evaluation of hair growth by computer assisted image analysis. Nouv Dermatol 7(Suppl. 1):56, 1988. 48. Lee T, Vincent NG, Gallagher R, Coldman A, McLean D. Dullrazor®: a software approach to hair removal from images. Comput Biol Med 27:533–543, 1997. 49. Hoffmann R. TrichoScan: combining epiluminescence microscopy with digital image analysis for the measurement of hair growth in vivo. Eur J Dermatol 11:362–368, 2001. 50. Chamberlain AJ, Dawber RP. Methods of evaluating hair growth. Australas J Dermatol 44:10–18, 2003. 51. Van Neste D. Dynamic exploration of hair growth: critical review of methods available and their usefulness in the clinical trial protocol. In Trends in Human Hair Growth and Alopecia Research, Van Neste D, Lachapelle JM, Antoine JL, Eds. Kluwer Academic Publishers, Lancaster, U.K., 1989, pp. 143–154. 52. Van Neste D, Tobin DJ. Hair cycle and hair pigmentation: dynamic interactions and changes associated with aging. Micron 1:1–2, 2004.
of the Mechanical 104 Measurement Strength of Hair R. Randall Wickett College of Pharmacy, University of Cincinnati, Cincinnati, Ohio
CONTENTS 104.1 Introduction ..........................................................................................................................................................895 104.1.1 Stress–Strain Curves ..............................................................................................................................895 104.1.2 The Two-Phase Model of the Hookean Region....................................................................................896 104.1.3 α-β Transformation in the Yield Region...............................................................................................896 104.1.4 The Series Zone Model and the Postyield Region ...............................................................................896 104.1.5 Variations among Fiber Types ...............................................................................................................897 104.2 Object ...................................................................................................................................................................897 104.2.1 Tensile Measurements of Hair Damage ................................................................................................897 104.2.2 Bending and Torsional Measurements ..................................................................................................897 104.2.3 Chemical Relaxation Methods...............................................................................................................897 104.3 Methodological Principle .....................................................................................................................................898 104.3.1 Overview ................................................................................................................................................898 104.3.2 Modulus Calculation..............................................................................................................................898 104.3.3 Tensile Testers........................................................................................................................................898 104.4 Sources of Error ...................................................................................................................................................898 104.4.1 Variability in Hair ..................................................................................................................................898 104.4.2 Relative Humidity ..................................................................................................................................899 104.4.3 Other Sources of Error ..........................................................................................................................899 104.5 Recommendations ................................................................................................................................................899 104.5.1 Reducing Variability ..............................................................................................................................899 104.5.2 Determination of Hair Diameters ..........................................................................................................899 104.5.3 Gripping the Hair...................................................................................................................................899 104.5.4 Breaking Strength from an Inexpensive Device ...................................................................................900 References .......................................................................................................................................................................900
104.1 INTRODUCTION 104.1.1 STRESS–STRAIN CURVES The mechanical behavior of hair and other keratin fibers is most conveniently, and thus most frequently, measured in extension. Figure 104.1 illustrates typical stress–strain curves for adjacent sections of the same hair obtained either in air at 40% relative humidity (RH) or immersed in water. These curves can be characterized by three different regions. In the first region the curve is approximately linear and a slope can be determined. This is called the Hookean region, and it extends to about 102% of the equilibrium length of the fiber (2% strain). Between 2 and
4% strain the curve “turns over” into the yield region. In the yield region, very little increase in force is required to increase extension. In the postyield region, which begins at about 25% strain in the dry fiber and 28% strain in the wet fiber, the force again increases markedly with strain. For this particular hair, under the conditions tested, the slope in the postyield region was about one fifth of that in the Hookean region of the dry fiber. There is little difference in postyield slopes between the wet and dry sections of the fiber. Published reports on the mechanical properties of keratin fibers date back to the work of Speakman1–3 in the 1920s. Since that time, extensive research on hair and wool has led to an interpretation of each region of the 895
896
Handbook of Non-Invasive Methods and the Skin, Second Edition
is close to that of ice, as expected for a hydrogen-bonded network.17 The matrix contributes viscous forces that decay with time, causing stress relaxation. The viscosity of the matrix decreases greatly as the water content of the fiber increases.4,11 The two-phase model of keratin fibers accounts well for the effects of water on the mechanical properties,2,11,18 the effect strain rate on Young’s modulus,4,19 the stress relaxation behavior in the Hookean region,13,20 and the behavior of wet, dry, and permanently set fibers in torsion.2,9,21
50 Hookean slopes
45 40
40% RH
Grams force
35
Water
30 25 Post yield region
20 15
Yield region
10 5 0
0
10
20
30 % Strain
40
50
60
FIGURE 104.1 Stress/strain curves for different sections of the same hair at 40% RH and in water.
stress–strain curves of keratin fibers in terms of changes occurring in the molecular structure. Feughelman4 has reviewed this work very thoroughly and his review is highly recommended to anyone with a serious interest in the mechanical properties of hair. The discussion below is confined to a brief overview.
104.1.2 THE TWO-PHASE MODEL OF THE HOOKEAN REGION In the region of the force extension curve below 1.5% extension keratin fibers are generally considered to behave as Hookean springs. Bendit5,6 has pointed out that the curve is not truly Hookean in this region. However, the curve is approximately linear and Young’s modulus of elasticity can be calculated.4,7 Since this region has been referred to as Hookean for at least 50 years, it is likely that the terminology will persist for the foreseeable future.4 The mechanical properties of hair or wool in the Hookean region are well explained by the two-phase model of Feughelman.4,8–14 This model considers the mechanical properties of the fiber to be determined by a water-impenetrable phase, C, the microfibrils, and a waterpermeable phase, M, the matrix. The microfibrils are primarily composed of a-helical proteins aligned parallel to the fiber axis,15 and the matrix is composed of water and high-sulfur proteins that may be globular.16 The composite may be modeled mechanically as a fixed Hookean spring in parallel with a spring and viscous dashpot in series.4,10 The spring contributes about 1.4 × 109 Pa to the Young’s modulus and is contained in the water-impenetrable microfibrils. The main resistance to extension of the microfibrils probably comes from the hydrogen bond network in the a-helical proteins. When a keratin fiber is immersed in liquid nitrogen to prevent segment mobility, the modulus
104.1.3 α-β TRANSFORMATION REGION
IN THE
YIELD
Above extensions of about 2 to 3% the stress–strain curve turns over into the yield region. The stress does not increase markedly until about 25% extension. The mechanical properties of a fiber extended into this region can be recovered by relaxing the fiber in water overnight if the fiber is not held too long in extension3 and the extension is carefully confined to the yield region. As we shall see, this fact is of great practical importance in designing protocols to measure hair strength. High-angle x-ray diffraction results have demonstrated that there is a progressive loss of α-helical content and a concomitant increase in β-sheet as a fiber is extended through the yield region.24 By the end of the yield region about 30% of the original α-helix has been unfolded. It has also been shown24,25 that the mechanical behavior of keratin in the yield region can be completely accounted for by application of a Burte–Halsey26 model. The fiber is considered to contain a continuum of units that can exist in a short state, A (α-helix), or an extended state, B (βsheet), with an energy barrier between the states. The yield region corresponds to a phase transition between state A and state B at constant stress. This first-order phase transition, producing a large length change at constant stress and temperature, is analogous to the transformation of water to steam, producing a large-volume change at constant temperature and pressure.
104.1.4 THE SERIES ZONE MODEL POSTYIELD REGION
AND THE
Speakman3 found the postyield slope to be independent of the water content of the fiber. The increase in stiffness in the postyield region was shown to result from a covalently bonded network involving cystine. The postyield slope has been shown to be dependant on the disulfide content of the fibers.27,28 Extension to about 50% strain leads to loss of all a-helical structure in the fiber24 as judged by x-ray diffraction. The behavior of keratin fibers in the yield and postyield regions has been interpreted in terms of a series
Measurement of the Mechanical Strength of Hair
zone model.29,30 The model postulates two alternating zones, X and Y, along the microfibrils. The X zones contain the 30% of the α-helices that unfold reversibly in the yield region. The Y zones contain regions of α-helix that cannot be unfolded without the breakdown of disulfide bonds. Thus, unfolding of the Y zones in the postyield region is irreversible.
897
treatments42 on hair. These studies have all shown tensile measurements to be very useful for the study of damaging treatments that break disulfide bonds in the cortex. However, Robbins and Crawford43 have shown that treatments that cause severe damage to hair cuticle may have little or no effect on tensile properties.
104.2.2 BENDING 104.1.5 VARIATIONS
AMONG
FIBER TYPES
Much of the work on the mechanical properties of keratin fibers has been carried out with wool. In the discussion above, the term keratin fiber has been used to refer to either hair or wool. The mechanical behavior of the two fibers is very similar. A comparative study by Menkart et al.31 found the elastic modulus and stress at 20% extension of hair to be somewhat higher than that of wool, while Chaiken and Chemberlain32 found the dynamic elastic modulus of hair to be nearly equivalent to that of wool. Tolgyesi et al.33 found beard hair to behave similarly to head hair in extension, but to have a slightly lower elastic modulus and stress at 20% extension. In general, head hair, beard hair, and wool may be considered similar enough in behavior that all of the conclusions about structure and mechanical properties discussed above can be considered to be equally valid for each structure.
104.2 OBJECT 104.2.1 TENSILE MEASUREMENTS
OF
HAIR DAMAGE
Measurements of hair tensile properties are most frequently made to assess the effects of chemical treatments on hair strength.7 The mechanical properties of hair are greatly affected when the number of disulfide cross-links is reduced. This is especially true of wet hair in the yield and postyield regions of the stress–strain curve.4,7,27,28,34–37 A typical protocol is to strain an untreated hair into the yield region and measure either the force or the work of extension. The work of extension is the area under the force vs. extension curve. Beyak et al.38 assessed the effects of bleaching and permanent waving on the relative stress to extend a hair to 15% strain before and after treatment, the 15% index (I15), in water. They found that a 30-min bleach treatment reduced I15 by an average of 10% and a 5-min permanent wave treatment reduced I15 by about 13%. Wolfram et al.39 reported that a 30-min bleaching treatment reduced the yield stress by about 12%, and Robbins7 presented data showing that a commercial permanent wave caused an 18% reduction in the work to extend a hair to 20% in the wet state compared to an 11% decrease in the work to extend hair to 20% in the dry state. Tensile measurements have also been used to characterize the effect of ultraviolet radiation,40 surfactant binding,41 and chlorine
AND
TORSIONAL MEASUREMENTS
A further objective of mechanical measurements on hair is to understand the processes involved in setting and permanent waving. Measurements of extensional properties are not necessarily the best way to achieve this goal. Bogaty44 pointed out that the behavior of hair under torsional and bending strains is very important to the permanent waving process because forming a curl from straight hair involves a combination of twisting and bending deformations. He found that permanent waving decreased the torsional rigidity of hair in the wet state, but actually increased it slightly at 65% RH. Wolfram and Albrecht45 made torsional measurements on hair and concluded that the cuticle is very stiff in the dry state and may make a significant contribution to the torsional rigidity, especially for fine hairs. However, in the wet state the cuticle was found to be so plasticized as to make no contribution to mechanical behavior. Scott and Robbins46,47 described a practical, balanced fiber method for measuring the bending stiffness of hair. A long hair is draped over a small wire with small weights attached to each end. The bending stiffness can be calculated from the distance between the two ends. It is also possible to measure bending strength by a three-point beam deflection method, and this method has been applied to measuring the stiffness of beard hairs.48 The balanced fiber method has the disadvantage of requiring a relatively long fiber, but in the author’s experience it is far easier to use than three-point bending methods.
104.2.3 CHEMICAL RELAXATION METHODS The dramatic effect that breaking disulfide bonds has on the tensile properties of hair can be used to study the kinetics of the reduction reaction. If a hair is stress relaxed to a constant level of force at a constant extension in water or buffer, and then the solution is switched to reducing agent, the force decays with time, due to the breaking of disulfide bonds,49–52 and kinetic parameters can be determined. An alternate method is to repeatedly stretch the hair in the linear region, measuring the reduction in elastic modulus as the reaction proceeds. 53 Wortman and coworkers54,55 have used a combination of these methods to attempt to predict permanent set based on relaxation parameters. I have recently reviewed these methods along with other work on the effects of permanent waving on the physical properties of hair.56
898
Handbook of Non-Invasive Methods and the Skin, Second Edition
104.3 METHODOLOGICAL PRINCIPLE 104.3.1 OVERVIEW This discussion of methodology will focus on use of extensional measurements to evaluate hair strength in tension, including determination of elastic modulus, yield stress, and breaking strength. These are the most widely used means to evaluate treatment effects on hair. Robbins7 has provided a thorough discussion of other methods for the evaluation of hair physical properties.
104.3.2 MODULUS CALCULATION Young’s modulus of elasticity, E (Y in some older papers), is calculated from the slope of the Hookean region of the stress–strain curve (Figure 104.1). E is defined as follows: E = ΔF*L/ΔL*A
(104.1)
where DF is the change in force induced by a change in length, DL, L is the equilibrium length of the fiber, and A is the cross-sectional area. For example, assume that a cylindrical hair of 80 mm (8 × 10–5 M) diameter, 5 × 10–9 M2 cross-sectional area, requires 10.2 grams-force (0.1 Newtons) to extend 1%: E = 100*0.1N/5 × 10–9 M2 = 2 × 109 N/M2 (104.2) A Newton/M2 is 1 Pa, so the modulus is 2 × 109 Pa. Older papers report E or Y in dynes/cm2. A pascal is 10 dynes/cm2, so the modulus of this fiber is 2 × 1010 dynes/cm2. When reporting E, the rate of extension must be specified because the slope of the Hookean region is known to vary with extension rate7,19 due to stress relaxation in the matrix.4 Young’s modulus is typically about 1.5 to 2.0 × 109 Pa for wet hair and 3.5 to 4.5 × 109 Pa for dry hair, depending on strain rate and hair source used.
104.3.3 TENSILE TESTERS Stress–strain measurements on hair are usually made with commercially available tensile testers. Most measurements reported in the literature over the years have been made using one of the several varieties manufactured by the Instron® Corporation, usually a table model such as the 4201. The Instron® uses a large, very robust screw drive system to move either a crosshead containing a load cell or a large mechanical stage at a constant rate of extension. Modern Instrons® are computer controlled, and a wide variety of stress–strain protocols can be programmed by using “canned” software provided with the instrument. By using a programming language such as PASCAL and drivers provided with the instrument, soft-
ware can be created to collect data using virtually any protocol that the programmer imagines. Advantages of the Instron® are its high precision and flexibility. Forces from less than 0.1 g to several kilograms can be measured, depending on the load cell selected, and extensions from less than a millimeter to more than a meter can be accurately performed. Disadvantages are its cost, $50,000 and up, depending on options, and its rather large size. Even a table model Instron® weighs a few hundred pounds and takes up a considerable amount of lab space. It is definitely not a portable instrument. A recently available option for measuring the mechanical properties of hair is the Dia-stron Rheometer (DR). The DR is portable. The measuring jig weighs just 3 kg, and the control unit is about the size of a typical personal computer box. The cost of the DR is about one fourth of that of an Instron®. The curves in Figure 104.1 were obtained using the DR. While completely adequate for obtaining stress–strain curves from hair, the DR has some limitations. The light weight of the DR measuring jig that makes it portable also makes it susceptible to vibrational noise, such as might arise from general activity in the lab. Stress–strain and stress relaxation protocols are entered into one of two methods available to the user at any one time. Each protocol allows one extension, and one compression phase with data collection after each, and each extension–compression cycle can be repeated several times. This gives considerable flexibility to the operator, but not the extreme flexibility provided by the Instron®. I should also point out that the software we obtained with the DR does not calculate Young’s modulus correctly. Whenever using a software package provided with an instrument, it is wise to check all calculations by hand.
104.4 SOURCES OF ERROR 104.4.1 VARIABILITY
IN
HAIR
A major source of error in measurements on hair is the inherent variability in the thickness and shape of hair. This can be compensated for to some extent by determining the dimensions of each hair and normalizing the result to cross-sectional area. Measurements of breaking strength are particularly prone to variability, and even different sections of the same hair will often break at different strengths. Another source of error when evaluating treatment effects is the inherent difference in treatment response between individuals. For example, reaction rates to reducing agent may vary by a factor of 10 between individuals who do not have a history of chemical treatment, and may vary more if chemically treated hair is used.50,52 Some possible solutions to the variability problem are discussed below.
Measurement of the Mechanical Strength of Hair
899
104.4.2 RELATIVE HUMIDITY When making dry measurements, exact control of relative humidity is very important. This is well illustrated by the fact that simple room relative humidity gauges often use a horsehair to move the dial as the room RH changes. Thus, variations in RH can be a significant source of error when making dry measurements of elastic modulus or yield stress. As can be seen in Figure 104.1, postyield slope and breaking strength are much less sensitive to RH.
104.4.3 OTHER SOURCES
OF
ERROR
Other sources of error are slipping of the hair gripping system, damage to the hair by the gripping system, and slack in the hair resulting in an inaccurate value of equilibrium length. Slack can be easily accounted for by modern computerized methods, which can recalculate the equilibrium length when significant force is first sensed. Gripping problems are discussed in detail below.
TABLE 104.1 Dia-stron® Hair Breaking Compared to Lever Breaking Hair # 3 5 1 10 9 2 8 7 6 4 Average
Dia-stron® 183.0 164.0 126.0 106.0 105.0 102.0 101.5 93.5 79.5 70.0 113.05
Lever
Difference Diast – Lever
193.0 159.0 150.0 111.0 97.0 89.0 95.0 95.0 65.0 82.0 113.6
+10 –5 +24 +5 –8 –13 –6.5 +1.5 –14.5 +12 0.55
104.5 RECOMMENDATIONS
variations in the hair, as described below. The only other alternative is to run literally hundreds of samples to get statistical significance.
104.5.1 REDUCING VARIABILITY
104.5.2 DETERMINATION
The best way to account for variability between hairs when evaluating treatment effects is to use each hair as its own control. As discussed above, the mechanical properties of a keratin fiber can be recovered if the hair is not strained into the postyield region.1,4,7,22 A typical protocol is that used by Beyak et al.38 Hairs were strained in water to 15% extension on an Instron® tensile tester and then soaked in water for 16 hours, treated, and rerun. The change in grams-force required to reach 15% extension was evaluated. Force values at 15% extension from 25 hairs ranged from 11.4 to 35.5 g in the first run. A second run without a treatment between showed an average change in force for each hair of only 0.33% of the original force, with a standard deviation of 2.68%. The changes observed ranged from +5.9% to –5.7%. Bleach treatments caused an average reduction in force at 15% strain of 10 to 20%, depending on treatment time. This study and many others like it7,39–42 clearly show the value of using each hair as its own control. When doing breaking or chemical relaxation studies, it is not possible to use each hair segment as its own control. The next best approach is to cut the hairs into different sections and compare results between sections, as was done to obtain the data in Table 104.1, discussed below. Breaking values on different sections are not as reproducible as reruns of stress–strain curves to 20% extension on the same section of hair, but still provide a large improvement over simply comparing different hairs. When it is not possible to either use each hair as its own control or compare different sections of the same hair, one must attempt to at least normalize for the dimensional
Determining the cross-sectional dimensions of a hair is a difficult problem. Not only is hair a fine fiber, but it is also not necessarily uniform in cross section. While Caucasian hair is generally considered elliptical in shape, significant variations from ellipticity can occur. Robbins7 has reviewed methods for diameter determination and recommends the linear density method as the method of choice. I concur with his recommendation. To use the linear density method, a hair is cut to a given length and weighed on a microbalance. The fiber density is assumed to be 1.32 g/cm3 (Reference 7) and the cross-sectional area, A, in square centimeters, is then given by
OF
HAIR DIAMETERS
A = W/(1.32*L)
(104.3)
where W is the weight of the hair in grams and L is the length of the segment measured in centimeters. If necessary, the effective diameter of the hair can then be determined from simple geometry if the hair is assumed to be circular in cross section.
104.5.3 GRIPPING
THE
HAIR
An annoying difficulty when making tensile measurements on hair is the problem of obtaining a good grip on the hair without causing damage at the gripping point. This is especially vexing when one wants to stress the hair all the way to breaking. The plastic fiber grips provided by Instron® are totally inadequate, as most hairs will slip from the grips well before they break.
900
Handbook of Non-Invasive Methods and the Skin, Second Edition
I have previously crimped each end of the hair into short sections of fine-diameter aluminum tubing, which can then be gripped by any good spring-loaded clamp, but the sections must be cut and filed carefully to avoid having any sharp edges that could cut the hair. What seems to be a reasonable solution to the gripping problem has recently been provided by Dia-stron®. It sells 14-mm sections of 2-mm-diameter metal tubing with plastic tubing inside. These sections can be crimped onto the hair to provide good handles for gripping. Dia-stron® also sells a crimping tool for this purpose that is convenient for crimping onto each end of a standard 3-cm hair section. We have found that an adequate crimp can also be obtained by careful use of an electrician’s crimping tool, available in nearly any hardware store. Even using the Dia-stron® crimping system one must be careful to get a good crimp if making a hair-breaking measurement. We usually move the sample block slightly and recrimp at least twice to ensure a good crimp. Otherwise, the hair may slip before breaking.
104.5.4 BREAKING STRENGTH FROM AN INEXPENSIVE DEVICE If you want to screen a treatment for its effect on hair strength and do not have access to a tensile tester, I recommend trying a simple and inexpensive device we have developed to measure the dry-breaking strength of hair.57 The device is based on a lever principle and is shown schematically in Figure 104.2. A meter stick pivots on a bolt through a hole in its center. The hair is fastened to one end and a weight on a sliding hanger is moved slowly and carefully along the other side, away from the center hole. The breaking strength is Wt* (L2/L1) – Wc, where Wt is the total weight of the sliding hanger and attached weight, L2 is the distance from the center hole to the slide at break, L1 is the distance from the center hole to the hair, and Wc is the weight of the clip used to hold the hair. I have made sliding weight hangers from either a L1 L2 Sliding mount Pivot rod Meter stick Upper grip Hair Lower grip Weight
FIGURE 104.2 Schematic diagram of a simple device to measure hair breaking strength.
mirror mount bracket or a closet door mount. No doubt other objects could be used for this purpose. Our current hanger weighs 30 g, and we use an additional 150 g of balance weights attached using wire. If the hairs are mounted into the Dia-stron® crimps discussed above, the grips can be made from alligator clips. Ten hairs were cut into two sections each, and the breaking strength of one section was determined on the DR and the other on the lever device. The results are shown in Table 104.1. The agreement between the average values for breaking strength obtained from each device is remarkable. Of course, the lever device does not allow one to produce a stress–strain curve, but it seems to work well as a potential screening tool considering that it took about a half hour and $12 worth of materials purchased from a local hardware store to construct.
REFERENCES 1. Speakman, J.B., The gel structure of the wool fibre, J. Text. Inst., 17, T457, 1926. 2. Speakman, J.B., The rigidity of wool and its changes with adsorption of water-vapor, Trans. Faraday Soc., 25, 92, 1929. 3. Speakman, J.B., The intracellular structure of the wool fibre, J. Text. Inst., 18, T431, 1927. 4. Feughelman, M., The physical properties of alpha keratin fibers, J. Soc. Cosmet. Chem., 33, 385, 1982. 5. Bendit, E.G., Properties of the matrix in keratins. II. The “Hookean” region in the stress-strain curves of keratins, Text. Res. J., 48, 717, 1978. 6. Bendit, E.G., There is no Hookean region in the stressstrain curve of keratin or other viscoelastic fibers, J. Macromol. Sci. Phys., B17(1), 129, 1980. 7. Robbins, C.R., Chemical and Physical Behavior of Human Hair, 2nd ed., Springer-Verlag, New York, chap. 8. 8. Feughelman, M., A two phase structure for keratin fibers, Text. Res. J., 29, 223, 1959. 9. Mitchell, T.W. and Feughelman, M., The torsional properties of single wool fibers. I. Torque-twist relationships and torsional relaxation in wet and dry fibers, Text. Res. J., 30, 662, 1960. 10. Feughelman, M., The relation between structure and the mechanical properties of keratin fibers, Appl. Polym. Symp., 18, 757, 1971. 11. Feughelman, M. and Robinson, M.S., Some mechanical properties of wool fibers in the “Hookean” region from zero to 100% relative humidity, Text. Res. J., 41, 469, 1971. 12. Feughelman, M., Keratin, in Encyclopedia of Polymer Science and Engineering, Vol. 8, 2nd ed., John Wiley & Sons, New York, 1987. 13. Feughelman, M. and Robinson, M.S., Stress relaxation of wool fibers in water at extensions in the Hookean region over the temperature range 0˚–90˚C, Text. Res. J., 39, 196, 1969.
Measurement of the Mechanical Strength of Hair
14. Feughelman, M., A note on the water impenetrable component of a-keratin fibers, Text. Res. J., 59, 739, 1989. 15. Fraser, R.D.B., MacRea, T.P., and Suzuki, E., Structure of the a-keratin microfibril, J. Mol. Biol., 108, 435, 1976. 16. Fraser, R.D.B., Macrae, T.P., and Rogers, G.E., Molecular organization in alpha keratin, Nature, 193, 1052, 1962. 17. Feughelman, M. and Robinson, M.S., The tensile behavior of wool fibers in liquid nitrogen, Text. Res. J., 37, 705, 1967. 18. Breuer, M.M., The binding of small molecules to hair. I. The hydration of hair and the effect of water on the mechanical properties of hair, J. Soc. Cosmet. Chem., 23, 447, 1972. 19. Sikorski, J. and Woods, H.J., the effect of rate of extension on Young’s modulus of keratin fibers, Leeds Phil. Soc., 5, 313, 1950. 20. Wortmann, F.J. and De Jong, S., Analysis of the humidity-time superposition for wool fibers, Text. Res. J., 55, 750, 1985. 21. Feughelman, M., Microfibril/matrix relationships in the mechanical properties of keratin fibers. I. The torsional properties of “melted” and permanently set keratin fibers, Text. Res. J., 48, 518, 1978. 22. Feughelman, M., A note on the recoverability of mechanical properties of wool, J. Text. Inst., 59, T548, 1968. 23. Bendit, The α-β transformation in keratin, Nature, 179, 535, 1957. 24. Feughelman, M., Creep of wool fibres in water, J. Text. Inst., 45, T630, 1954. 25. Feughelman, M. and Rigby, B.J., A two energy state model for the stress relaxation and creep of wool fibres in water, in Proceedings of the International Wool Textile Conference, Australia, 1955, p. D-62. 26. Burte, H. and Halsey, G., A new theory of non-linear viscoelasticity, Text. Res. J., 17, 465, 1947. 27. Feughelman, M., The mechanical properties of permanently set and cystine reduced wool fibers at various relative humidities and the structure of wool, Text. Res. J., 33, 1013, 1963. 28. Cannell, D.W. and Carothers, L.E., Permanent waving: utilization of the post-yield slope as a formulation parameter, J. Soc. Cosmet. Chem., 29, 685, 1978. 29. Feughelman, M. and Haly, A.R., Structural features of keratin suggested by its mechanical properties, Biochim. Biophys. Acta, 32, 596, 1959. 30. Feughelman, M., The post-yield region and the structure of keratin, Text. Res. J., 34, 539, 1964. 31. Menkart, J., Wolfram, L.J., and Mao, I., Caucasian hair, Negro hair and wool: similarities and differences, J. Soc. Cosmet. Chem., 17, 769, 1966. 32. Chaiken, M. and Chemberlain, W.H., The propagation of longitudinal stress pulses in textile fibers, J. Text. Inst., 46, T44, 1955. 33. Tolgyesi, E., Coble, D.W., Fang, F.S., and Kairinen, E.O., A comparative study of beard and scalp hair, J. Soc. Cosmet. Chem., 34, 361, 1983.
901
34. Weigman, H.D., Rebenfield, L., and Danziger, C., The Role of Sulfhydryl Groups in the Mechanism of Permanent Setting of Wool, III Cirtel, Paris, 1965, Sec. 2, p. 319. 35. Hermans, K.W., Hair keratin, reaction, penetration and swelling in mercaptan solutions, Trans. Faraday Soc., 59, 1633, 1963. 36. Weigman, H.D. and Danziger, C.J., Effects of crosslinks on the mechanical properties of keratin fibers, Appl. Polym. Symp., 18, 795, 1971. 37. Robinson, M.S. and Rigby, B.J., Thiol differences along keratin fibers: stress/strain and stress-relaxation behavior as a function of temperature and extension, Text. Res. J., 55, 597, 1985. 38. Beyak, R., Meyer, C.F., and Kass, G.S., Elasticity and tensile properties of human hair. I. Single fiber test method, J. Soc. Cosmet. Chem., 20, 615, 1969. 39. Wolfram, L.J., Hall, K., and Hui, I., The mechanism of hair bleaching, J. Soc. Cosmet. Chem., 21, 875, 1970. 40. Beyak, R., Kass, G.S., and Meyer, C.F., Elasticity and tensile properties of human hair. II. Light radiation effects, J. Soc. Cosmet. Chem., 22, 667, 1971. 41. Breuer, M.M., The interaction between surfactants and keratinous tissue, J. Soc. Cosmet. Chem., 30, 41, 1979. 42. Fair, N.B. and Gupta, B.S., The chlorine-hair interaction. II. Effect of chlorination at varied pH levels on hair properties, J. Soc. Cosmet. Chem., 38, 371, 1987. 43. Robbins, C.R. and Crawford, R.J., Cuticle damage and tensile properties of human hair, J. Soc. Cosmet. Chem., 42, 59, 1991. 44. Bogaty, H., Torsional properties of hair in relation to permanent waving and setting, J. Soc. Cosmet. Chem., 18, 575, 1967. 45. Wolfram, L.J. and Albrecht, L., Torsional behavior of human hair, J. Soc. Cosmet. Chem., 36, 87, 1985. 46. Scott, G.V. and Robbins, C.R., A convenient method for measuring fiber stiffness, Text. Res. J., 39, 975, 1969. 47. Scott, G.V. and Robbins, C.R., Stiffness of human hair fibers, J. Soc. Cosmet. Chem., 29, 469, 1978. 48. Savenije, E.P.W. and De Vos, R., Mechanical properties of human beard hair, Bioeng. Skin, 2, 215, 1986. 49. Reese, C. and Eyring, H., Mechanical properties and the structure of hair, Text. Res. J., 20, 743, 1950. 50. Wickett, R.R., Kinetic studies of hair reduction using a single fiber technique, J. Soc. Cosmet. Chem., 34, 301, 1983. 51. Wickett, R.R. and Barman, B.G., Factors affecting the kinetics of disulfide bond reduction in hair, J. Soc. Cosmet. Chem., 36, 75, 1985. 52. Wickett, R.R. and Mermelstein, R., Single fiber stress decay studies of hair reduction and depilation, J. Soc. Cosmet. Chem., 37, 461, 1986. 53. Szadurski, J.S. and Erlman, G., The hair loop test: a new method of evaluating perm lotions, Cosmet. Toilet., 41(12), 41, 1984. 54. Wortman, F.J. and Souren, I., Extensional properties of hair and permanent waving, J. Soc. Cosmet. Chem., 38, 125, 1987.
902
Handbook of Non-Invasive Methods and the Skin, Second Edition
55. Wortman, F.J. and Kure, N., Bending relaxation properties of human hair and permanent waving performance, J. Soc. Cosmet. Chem., 41, 123, 1990. 56. Wickett, R.R., Disulfide bond reduction in permanent waving, Cosmet. Toilet., 106(7), 37, 1991.
57. Wickett, R.R., An inexpensive device to measure hair breaking strength, manuscript in preparation.
the Strength of Human 105 Evaluating Hair Sidney B. Hornby Neutrogena Corporation, Los Angeles, California
CONTENTS 105.1 Introduction ..........................................................................................................................................................903 105.2 Methodology.........................................................................................................................................................903 105.2.1 Preparation of the Sample .....................................................................................................................903 105.2.2 Importance of Hair Fiber Dimensions ..................................................................................................904 105.3 Instruments ...........................................................................................................................................................904 105.3.1 Cyclic Tester ..........................................................................................................................................904 105.3.2 Impact Loading ......................................................................................................................................905 105.3.3 Flexabrasion Method .............................................................................................................................905 105.4 Analysis of Hair Failure Data..............................................................................................................................905 105.4.1 Characteristic Life or Alpha ..................................................................................................................905 105.4.2 Shape Parameter or Beta .......................................................................................................................905 105.4.3 Calculating Characteristic Life Values from Experimental Data..........................................................905 105.4.4 Survival Probability Curves...................................................................................................................906 105.5 Concluding Remarks ............................................................................................................................................907 References .......................................................................................................................................................................907
105.1 INTRODUCTION Strong, healthy-looking hair is almost as important as the condition of the skin to many people. However, most people engage in styling routines that lead to progressive damage to the hair fibers, which frequently manifests itself in undesirable hair splitting and breakage. In Chapter 104 the measurement of the mechanical break strength of hair by tensile testing was described. Reduction in the break strength can be observed in hair that has been severely damaged, especially by bleaching and permanent waving.1 However, damage in the surface regions of the hair that does not involve significant changes to the structure of the hair cortex cannot easily be distinguished by tensile testing. Nor can the benefit of conditioners and certain treatments always be demonstrated.2 Accelerated wear tests such as cyclic or fatigue testing are routinely employed to analyze the propensity of everything from aerospace composites to mechanical motors to fail in ordinary use. Items are repeatedly stressed until they break or fail. The failure data are then analyzed to determine whether the item can withstand the rigors of regular usage over time. In this chapter, the applications
of three accelerated wear methodologies to human hair fibers are described.
105.2 METHODOLOGY 105.2.1 PREPARATION
OF THE
SAMPLE
Human hair can be obtained from volunteers or purchased from a supplier. The effects of applied conditioning treatments can be best distinguished if the substrate hair is predamaged. A commercial bleach treatment is a good method of predamaging hair, since this is a very common procedure. Single fibers should be randomly selected from the hair bundle. Then a suitable length of each fiber (depending on the instrument used) should be affixed at each end. An easy mounting system consists of polyvinyl chloride (PVC)-lined brass ferrules that are crimped over each end of the single hair filaments. The PVC lining provides a cushion so that the hair fiber is not cut or crushed. Typically, 20 single hair fibers are tested in each sample, and the experiment should always include a control sample. Since the viscoelastic properties of hair are greatly 903
904
Handbook of Non-Invasive Methods and the Skin, Second Edition
Failure cycles
2000 1500 1000 500 0 0.01
0.015 0.02 0.025 Stress applied (grams/square micron)
FIGURE 105.1 The relationship between stress applied to hair fibers under a constant cyclic load of 60 g and the corresponding cycles to break.
affected by moisture in the environment,3 the relative humidity in the test area should be carefully controlled.
105.2.2 IMPORTANCE
OF
HAIR FIBER DIMENSIONS
It is known that the thickness of individual hair fibers can vary widely even if they are obtained from the head of one volunteer.3 Therefore, since the load applied to each fiber is constant, the applied stress on each fiber varies with different fiber dimensions. The relationship between the failure cycles of hair fibers and the stress applied during cyclic testing has been explored for chemically unaltered (virgin) hair, bleached hair, and bleached hair treated with two different conditioning systems.2 It was found that, except at the very low stress values (high-fiber cross-sectional area) and very high stress values (low-fiber cross-sectional area), there was no clear systematic relationship between the applied stress and cycles required to break each hair (Figure 105.1). One explanation for this result is that at the high stress values, preexisting fiber damage in the form of cracks or flaws propagates easily and rapidly through the hair fiber, resulting in breakage at a relatively low number of stress cycles. Therefore, for small-diameter hairs, the number of cycles to break each hair fiber is more strongly dependent upon the number of significant faults in the hair than the number of applied cycles. Conversely, the applied stress in fibers with larger than usual diameters may not be high enough to cause preexisting flaws to rapidly progress right through the fiber. As a result, much higher numbers of cycles will be required before the hair fails. Therefore, at low applied stress levels, the numbers of cycles at break becomes more strongly dependent on the duration of cyclic fatiguing endured by the fiber, and less dependent on the presence of surface damage. Therefore, the cross-sectional areas of each hair fiber in the sample should be determined microscopically or by laser micrometer before testing. Fibers with cross-
FIGURE 105.2 Cyclic tester for evaluating hair strength by resistance to break.
sectional areas between 2500 and 5000 μ2 can be included in the sample, and fibers outside of this range should be discarded.
105.3 INSTRUMENTS 105.3.1 CYCLIC TESTER The cyclic tester (Dia-Stron, Ltd., Andover, U.K.) is an apparatus that is designed to stretch single fibers to a selected load at a given speed and then relax them, with these cycles repeating until the hair breaks.2,4 Each hair fiber is placed in turn into the machined sample holder of the cyclic tester either automatically or by hand. The sample holder is designed so that one end of the hair fiber is attached to a sensitive load cell and the other to a movable sophisticated drive system that employs a closed-loop controlled linear actuator (Figure 105.2). This drive system can be programmed to achieve a wide range of movement distances, elongation speeds, and accelerations, and is capable of positional accuracy of better than 10 μ. Each stress cycle applied to the hair filament consists of stretching at the selected speed (10 to 20 mm/s) until the load cell registers the predetermined load. The load should be chosen to avoid stretching hair fibers well into their yield region. Typically, the load can be set to between 40 and 60 g at an elongation speed of 20 mm/s.4 Once the appropriate load is detected, the direction of the moving arm reverses, removing the load from the fiber. These oscillations are repeated until the fiber breaks. The computer then records the number of elongation cycles elapsed at break. The data obtained are in the form of number of cycles at break of each hair fiber, which can be sorted in order of breaking. In addition, the computer can record the actual load elongation curves at selected intervals. Such data allow us to observe the changes in the viscoelastic behavior of the hair fiber as it is stressed.
Evaluating the Strength of Human Hair
105.3.2 IMPACT LOADING This method (TRI Fatigue Tester, TRI/Princeton, Princeton, NJ)5,6 can stress up to 40 hair fibers simultaneously, rather than sequentially. The fibers are suspended from individual holders in a canopy with a weight on each one. At the beginning of the experiment each weight rests in a pocket equipped with a switch on a movable platform underneath the fibers. Then the platform oscillates up and down with amplitude of movement sufficient to relax and then load the hairs in a cyclic fashion. As each hair fiber breaks, the corresponding switch is triggered and the computer records the cycles elapsed. At the end of the experiment the data obtained are in an array of number of cycles at break of each hair fiber, but no load elongation data are recorded.
905
on a normally distributed data cannot be used to compare different samples of hair. Instead, the experimental data are usually exponentially distributed and are best analyzed using the twoparameter form of the Weibull distribution,10–15 shown in Equation 105.1:
F ( x) = 1 − e
⎛ x ⎞ −⎜ ⎝ alpha ⎟⎠
beta
, for x > 0
(105.1)
F(x) is the probability that a given hair fiber will break at x cycles, alpha is the Weibull characteristic life, and beta is the Weibull shape parameter. To compare the relative breakage resistance of the hair fiber samples, we first need to understand alpha and beta.
105.3.3 FLEXABRASION METHOD
105.4.1 CHARACTERISTIC LIFE
This technique is different from the cyclic or impact fatigue testing described above. In addition to the stress imparted to each hair fiber by loading it, this apparatus simulates the abrasive effect of a comb or brush rubbing against individual hair fibers during grooming.7,8 The Flexabrasion apparatus (Croda, Inc., Edison, NJ) can test 20 fibers at once. Typically, 14-mm lengths are cut from adjacent sections of a single hair fiber to minimize the effect of fiber-to-fiber variations in cross-sectional area. Each specimen is glued onto mounts. Each fiber is then repeatedly run over a drawn tungsten wire under a weight at a 90˚ angle, as diagrammed in Figure 105.3.9 Typically a 2- to 4-mm section of the fiber is abraded at a frequency between 0.25 and 7 Hz in a temperature- and humiditycontrolled cabinet. At the end of the experiment the data obtained are in the form of an array number of cycles at break of each hair fiber.
The characteristic life of failure data is analogous to the mean of normally distributed data and is rigorously defined as the number of stress cycles elapsed when 63.2% of the hair fibers in the sample have broken. Therefore, a larger characteristic life of a sample suggests that it has a greater resistance to breaking during everyday brushing, combing, and styling than a sample of hair with a lower characteristic life value.
105.4 ANALYSIS OF HAIR FAILURE DATA The data obtained from each of the techniques described above are in the form of an array of the cycles required to break each hair fiber, from first fiber to break to the last. Failure data are usually not normally distributed, so calculating the average failure cycles of the sample based Reciprocating motion
105.4.2 SHAPE PARAMETER
OR
OR
ALPHA
BETA
The shape parameter of the Weibull distribution is related to the failure rate of the hair fibers in the sample. Generally, shape factors greater than 1 indicate that the failure rate is increasing with the time that the fibers are subjected to fatigue testing. This behavior is typical of items that wear out during use, such as tires or clothing.11 Shape parameters less than 1 suggest that items in the sample are breaking early during the stress cycling. Very often, as the cyclic testing continues, the shape parameter may become equal to 1, indicating the onset of constant failure rate. This behavior would be typical in populations where flawed specimens are eliminated early in the testing. An example would be light bulb filaments, which experience a burn-in period where a number will burn out quickly during normal use, after which, the surviving filaments will burn for extended periods. A shape parameter of less than 1 is usually observed in hair samples, indicating that previously flawed hairs break quickly, leaving hair fibers that can withstand stressing for extended times.
Wire W
FIGURE 105.3 Schematic drawing of a hair fiber specimen being subjected to flex fatigue on the Flexabrasion apparatus.
105.4.3 CALCULATING CHARACTERISTIC LIFE VALUES FROM EXPERIMENTAL DATA Obviously, the form of Equation 105.1 does not conveniently lend itself to the calculation of characteristic life
906
Handbook of Non-Invasive Methods and the Skin, Second Edition
from the array of failure data generated during one of the three techniques. Therefore, Equation 105.1 should be transformed into a linear relationship of the form Y = mx + b by taking the natural log of each side twice, yielding Equation 105.2:
TABLE 105.1 Characteristic Life Values of Treated Hair
Treatment
⎡ ⎛ 1 ln ⎢ ln ⎜ ⎜ ⎢ ⎝ 1− F x ⎣
()
⎞⎤ ⎟ ⎥ = beta ln x − beta ln alpha ⎟⎠ ⎥ ⎦ (105.2)
( )
( (
))
It is recommended to estimate the probability of failure, F(x), by using the method of median ranks.11–15 First the data are ordered from first failure to last. Then F(x) is derived using Equation 105.3: F ( x) ≈
failure order number − 0.3 N + 0.4
(105.3)
N is the total number of broken specimens. Once F(x) is calculated for each fiber in the ordered array, ⎡ ⎛ ⎞⎤ 1 ln ⎢ ln ⎜ ⎟ ⎥ can be calculated and plotted vs. the ⎢ ⎜⎝ 1 − F x ⎟⎠ ⎥ ⎣ ⎦ natural log of the failure cycles, ln(x), obtained from the experiment. Figure 105.4 shows a typical plot of the linearly transformed Weibull distribution data obtained from a sample of bleached hair. The points lie close to the bestfit regression line, indicating that the two-parameter form of the Weibull distribution is a suitable model for the experimental data.
()
Unaltered (virgin) Bleached Bleached and leave-on conditioner Bleached and rinse-off conditioner
Characteristic Life, alpha (cycles)
Shape Parameter, beta
34,187 1,213 10,979
0.51 0.31 0.41
5,021
0.33
A poor fit to a linear regression line suggests that an alternative form of the Weibull equation that utilizes more than two parameters is required. This is rarely observed for hair fibers. The slope of the regression line is the Weibull shape parameter, beta, and its Y-axis intercept is –beta (1n (alpha)) from inspection of Equation 105.2. Once these two values are identified, it is easy to calculate alpha, the characteristic life. Examples of characteristic life values and shape parameters of unaltered hair and that of hair after bleaching are given in Table 105.1.2 The characteristic life of bleached hair is substantially less than that of virgin hair. Bleaching weakens the fiber structure,1 and it becomes less resistant to breakage during styling routines. The increase in characteristic life values after applying conditioner to the bleached hair suggests that the treatment helps it resist breaking. The comparison of alpha values also can be used to rate the effects of a leave-on or a rinse-off conditioner.
105.4.4 SURVIVAL PROBABILITY CURVES 1.5 1 0.5
1
ln ln
1 − F(x)
0 0
5
10
15
−0.5 −1 −1.5 −2 −2.5 −3 ln(x)
FIGURE 105.4 A typical plot of the linearly transformed Weibull distribution data obtained from a sample of bleached hair. The solid line is the best-fit regression line of the calculated data points.
A powerful feature of the Weibull analysis of failure data is the ability to predict the probability of hair fibers to break under stress cycling, which simulates everyday grooming and styling stresses. The procedure is to rewrite Equation 105.1 using the characteristic life value and shape parameter derived from the experimental data. Then the equation is used to calculate F(x), the probability of survival as a function of the number of applied stress cycles (ln[x]). The output can be plotted as demonstrated in Figure 105.5 for the data of Table 105.1. From the curves, it is apparent that the bleached hair fibers that were tested have about a 49% chance of surviving 400 cycles of stress under the conditions of the experiment. Unaltered hair has approximately a 90% probability of surviving the same number of cycles. The application of the leave-on conditioner on the bleached hair raises its survival probability to 77%. Therefore, plots of survival curves are a powerful tool for assessing the beneficial effects of treatments on damaged hair.
Evaluating the Strength of Human Hair
907
1 0.9
Survival probability
0.8 0.7
Unaltered hair
0.6
Bleached hair
0.5
Bleached + leave in conditioner
0.4
Bleached + rinse off conditioner
0.3 0.2 0.1 0 0
400
800 1200 Cycles to failure
1600
2000
FIGURE 105.5 Survival probability curves calculated using the alpha and beta values shown in Table 105.1.
105.5 CONCLUDING REMARKS Direct measurement of the mechanical strength of hair fibers is usually insufficient to elucidate the benefit of conditioning treatments on damaged hair. Hair strength is not just the amount of direct force to break the fiber; it is also reflected by the ability of the hair fiber to resist breakage under relatively low stress levels encountered during combing or brushing. Therefore, cyclic fatigue testing is an invaluable tool in assessing hair damage and evaluating the beneficial effect of formulations.
REFERENCES 1. Tate, M.L., Kamath, Y.K., Ruetsch, S.B., and Weigmann, H.-D., Quantification and prevention of hair damage, J. Soc. Cosmet. Chem., 44, 155, 1993. 2. Hornby, S.B., Cyclic testing: demonstrating conditioner benefits on damaged hair, Cosmet. Toilet., 116, 35, 2001. 3. Robbins, C.R., Chemical and Physical Behavior of Human Hair, 4th ed., Springer-Verlag, New York, 2002, p. 395. 4. Hornby, S.B., Wiunsey, N.J.P., and Bucknell, S., New technique to capture viscoelastic changes in hair induced by mechanical stress, IFSCC Magazine, 5, 93, 2002.
5. Kamath, Y.K., Hornby, S.B., and Weigmann, H.-D., Mechanical and fractographic behavior of Negroid hair, J. Soc. Cosmet. Chem., 35, 21, 1984 6. Kamath, Y.K., Hornby, S.B., and Weigmann, H.-D., Effect of chemical and humectant treatments on the mechanical and fractographic behavior of Negroid hair, J. Soc. Cosmet. Chem., 36, 39, 1985. 7. Swift, J.A., Chahal, S.P., and Challoner, N., Flexabrasion: a method for evaluating hair strength, Cosmet. Toilet., 116, 53, 2001. 8. Leroy, F. et al., Flexabrasion: A New Test for Predicting Human Hair Resistance, poster at the 1st Tricontinental Meeting of Hair Research Societies, Brussels, Belgium, October 8–10, 1995. 9. Personal communication. 10. Weibull, W., A statistical distribution function of wide applicability, J. Appl. Mech., 9, 292, 1951. 11. Kececioglu, D., Reliability and Life Testing Handbook, Vol. 1, Prentice Hall, Englewood Cliffs, NJ, 1993. 12. Collins, J., Failure of Materials in Mechanical Design, 2nd ed., John Wiley & Sons, New York, 1993. 13. Griffith, A.A., The phenomena of rupture and flow in solids, Phil. Trans. R. Soc. London, 163–197, 1921. 14. Epstein, B., Statistical aspects of fracture, J. Appl. Physics, 19, 140, 1948. 15. Evans, G. and Jones, R.L., Evaluation of a fundamental approach for the statistical analysis of fracture, J. Am. Ceram. Soc., 61, 156, 1978.
Nail Structure and Growth
for Nail Assessment: 106 Methods An Overview David de Berker Bristol Dermatology Centre, Bristol Royal Infirmary, Bristol, United Kingdom
CONTENTS 106.1 Methods for Nail Assessment: An Overview ...................................................................................................911 106.2 Photography.......................................................................................................................................................911 106.3 Photodermatoscopy and Dermatoscopy............................................................................................................912 106.4 Capillaroscopy ...................................................................................................................................................913 106.5 Profilometry .......................................................................................................................................................913 106.6 Surface Replicas ................................................................................................................................................913 106.7 Magnetic Resonance Imaging ...........................................................................................................................913 106.8 Microscopy: Light, Scanning, and Transmission Microscopy .........................................................................914 106.9 Other Techniques...............................................................................................................................................914 106.10 Scoring Systems ................................................................................................................................................914 106.11 Strength..............................................................................................................................................................915 106.12 Measuring Nail Strength ...................................................................................................................................915 106.13 Permeability.......................................................................................................................................................916 References .......................................................................................................................................................................916
106.1 METHODS FOR NAIL ASSESSMENT: AN OVERVIEW Much science applied to nails is a microcosm of dermatology in general. Nails are a specialized appendage that call for adaptation and experimentation of standard techniques. Only in some areas do they command techniques special to themselves, such as in the measure of longitudinal growth or their strength. In this chapter we cover how the nail can be assessed with the use of techniques described in detail in other parts of the book. This will lead to many brief sections, best read in conjunction with the core chapters. Many of these techniques are very simple, but small tips can make a big difference to the success of your methods. We also cover the borderline topic of constituent analysis. Nails and hair afford themselves to providing “tissue” without invading the body. In legal terms, it is not clear whether obtaining a nail clipping or a cut hair sample is an invasive process. However, in scientific terms and for medicine in general, most people would be happy to provide such samples.
106.2 PHOTOGRAPHY There are three points that have great importance when undertaking photography of the nail. First is focus. Although this might sound banal, it is important to realize that the nail is a limited field of focus usually within a larger picture. It is also curved in two axes. The depth of focus of digital cameras is generally poor and there is little tolerance. The autofocus should be set on central spot focus or manual to minimize the chance of having a beautifully focused background and a blurred digit. The second point is illumination (Figure 106.1). Most digital cameras have their own flash. This introduces problems with macrophotography, as you will want to be as close as possible. The angle between the origin of the flash and the point of focus means that the light is slanted and introduces a shadow down one side of the digit. This is avoided if you use ambient light, but that leads to slow exposures and blur. The problem can be overcome to some extent by taking high-resolution images at some distance and then blowing up the area of interest as part of the digital editing (Figure 106.2). The ideal situation is to work with the small number of cameras with high 911
912
Handbook of Non-Invasive Methods and the Skin, Second Edition
A
B
C
D
FIGURE 106.1 Toe photographed with digital camera (A) in bright ambient light in the longitudinal axis of the toe. (B) In the same light as (A), but with flash. (C) In poor ambient light and in the same light with flash. The skin tones are superior in (A) and (D). The proximity of the camera in (B) and (D) means that there is low exposure adjusted for flash, but the light does not reach the subject. The zoom can accommodate this to some degree, but the figure illustrates a point.
F E
A
G
FIGURE 106.2 Where there is difficulty obtaining a good close-up image due to depth of focus, it is possible to take a more distant image (E), crop it (F), and blow it up 4× (G), which compares favorably with the original detailed shot (A) from Figure 106.1.
tolerance of low light or a separate light source. Separate light sources can be proper studio lights or a handheld flash. The first is preferable, as it will allow you to gauge the distortion introduced by reflection from the nail surface. However, while it may be okay for standardized laboratory photographs, it is not so good for recording clinical scenarios. It is also not so flexible for controlling the angle of any shadow you may wish to introduce.
FIGURE 106.3 A patient with evolving pitting has mascara rubbed into the nail plate prior to photography to assist in demonstration of the pattern of pits.
Remember that if you do not want shadow, you can use a ring flash or orient the flash longitudinally up the digit. Thirdly is the artifact of reflection from the nail surface. This makes it difficult to visualize detail of the surface or color characteristics of the subungual tissues. This problem is resolved with the use of ambient or studio lighting. All flash techniques are vulnerable to production of surface glare from the nail plate. Subtle surface irregularities are often difficult to discern with photography, even with the above techniques. These features can be enhanced by using mascara rubbed into the nail surface and then partially removed with a soft cloth. This leaves mascara within the surface impressions and outlines them for photography (Figure 106.3). Video is less commonly used as a nail research tool. Goyal and Griffiths1 employed images grabbed from video to make serial measurements for nail dimensions for morphometric analysis.
106.3 PHOTODERMATOSCOPY AND DERMATOSCOPY The dermatoscope is a useful tool for close inspection and can be used with or without intervening liquid medium. Nikon makes attachments that provide compatibility with standard dermatoscopes and allow digital photography. When not using a medium, the dermatoscope can be used to assess fine surface detail with the benefits of the 10× magnification, most effectively delivered using the lightemitting diode (LED) dermatoscope head. The tool allows analysis of nail surface characteristics in vivo and fine changes in periungual epidermal morphology, as can be seen with defining the margins for excision of Bowen’s disease of the nail unit (Figure 106.4). The most convenient medium for nail assessment is aqueous clear gel as used for skin ultrasound or
Methods for Nail Assessment: An Overview
FIGURE 106.4 The patient is undergoing Mohs’ micrographic surgery. Intraoperative examination with the dermatoscope assists in defining the clinical margin that was previously obscured by the nail plate. The subtle border of change of skin markings can be seen.
lubrication. This comes in a tube and can be easily wiped off. The alternative of mineral oil is not so good, as it creates only a thin film and it is difficult for the flat lens of the dermatoscope to form a good seal. With this technique, many nail bed and matrix features can be more closely assessed. This is of particular value in pigmentary changes. The dermatoscope allows a high level of clinical confidence in the differentiation of blood, melanin of matrix origin, and pigment of fungal origin. In the nail fold, blood vessels are easily qualitatively assessed with this technique.2
106.4 CAPILLAROSCOPY Modified forms of the dermatoscope can be used for quantitative capillaroscopy in normal and disease (Figure 106.5). Some of these have a video element,3 which in turn requires computer analysis. Diseases include connective tissue disorders4 and psoriasis.5 Where basic morphology is being measured, descriptive terms are used according to taxonomies created within each study,6 as well as some more accepted terms.7
106.5 PROFILOMETRY Profilometry is the technique of measuring the profile of a surface. It can be used on the nail surface to assess pitting, grooves, and trachyonychia using the measures of roughness, mean depth of roughness, and number of peaks or crests.8 Changes in these characteristics can be equated with disease activity and used as measures of response to therapy, such as in psoriatic trachyonychia during lowdose cyclosporin9 and the rate of nail growth during itraconazole treatment.10
913
FIGURE 106.5 Nail-fold capillaries viewed with dermatoscope, ultrasound gel, and using a Nikon attachment.
106.6 SURFACE REPLICAS It is possible to make surface replicas of the nail with a range of materials. The two that are easiest to obtain are cyanoacrylate11 and silicone molding material. The first provides a replica that is translucent and allows examination by normal light microscopy as well as scanning electron microscopy. However, it is a difficult balance between obtaining a dry replica on the surface of a glass slide apposed to the nail and finding that the slide is irrevocably stuck to the nail. Given that the technique requires a glass slide to allow subsequent light examination, it can be a problem if significant force is needed to remove the slide from the nail surface. The rate of drying is a function of the temperature, force applied, and characteristics of the glue. It takes some practice and experimentation to master the technique — something that I have not managed. Silicone is more flexible and forgiving, but also introduces small artifacts not seen with cyanoacrylate. It is a good alternative to photography where a three-dimensional record needs to be kept of a morphological characteristic of the nail or periungual tissues (Figure 106.6). Where I have used this, I have glued the mold to a glass slide and used it to attach identification. The replica method has been evaluated alongside scanning electron microscopy, indicating a high degree of agreement between techniques.12
106.7 MAGNETIC RESONANCE IMAGING Magnetic resonance imaging (MRI) produces very clear images of the nail, phalanx, and periungual tissues. It has been used for anatomical delineation of structures in normal digits.13 Where there is unexplained pain or dystrophy of an isolated digit, MRI may reveal a soft tissue tumor where traditional radiography is normal.14 It is most useful when the tumor contrasts with surrounding tissues with respect to density, fluid, or fat content. The most marked example of this is with myxoid pseudocysts.15 It is of
914
Handbook of Non-Invasive Methods and the Skin, Second Edition
A
B
C
FIGURE 106.6 (A) Silicone rubber can be used as a molding material to make an imprint of the nail surface. (B) Detail of the imprint reveals the pattern of nail pitting. (C) The mold is glued to a slide, which in turn is labeled on the reverse.
particular value when there is pain after an initial operation for an entity such as a glomus tumor, characterized by pain. In this setting it is not certain whether the pain is due to recurrence and warrants further surgery or whether it is related to the surgery itself, which makes further surgery a bad idea.16
106.8 MICROSCOPY: LIGHT, SCANNING, AND TRANSMISSION MICROSCOPY Nail plate is amenable to examination by all forms of light and electron microscopy. Light microscopy is sometimes of higher quality if thinner sections (6 μ) are cut after embedding in epoxy, but the improvement over standard techniques is not great.17 Standard techniques entail thicker sections, cut from wax-embedded nail. However, one thing that does make a difference is how long the nail is left in formalin. There is a case for avoiding this stage altogether, as formalin hardens an already hard structure and makes it even more difficult to cut high-quality sections. It can be helpful to practice whatever staining procedure you anticipate on both fixed and unfixed specimens. Cutting sections typically causes problems with the microtome, leaves a jagged profile on sections and makes the blade blunt. This makes nail histology unpopular with laboratory technicians. A further problem is getting the sections to stick to the glass slides. Where there is no soft tissue as part of the specimen, there is little if any natural adhesiveness to keep the section on the slide. Some pretreated slides will minimize this problem, but it is still advisable to produce multiple sections in order that a percentage of failures can be tolerated.
Once ready for staining, sections can be stained to demonstrate different things. A periodic acid Schiff is useful for delineating the outline of the nail plate cells throughout the nail and also highlights fungus. Histological evaluation of nail clippings is one of the main supplements to routine mycology that can clarify uncertain diagnoses.18 Scanning electron microscopy can be used on nail fragments — on the dorsal or ventral aspect or on sections. The main use of this technique has been to explore the natural tendency of the lamellar fabric of the nail to separate at the free edge in onychoschizia.19 At the magnification used, there is not really any advantage over using light microscopy, which is much less expensive. However, scanning electron microscopy is superior at examining the nail plate surface and defining the features of individual corneocytes. Transmission electron microscopy has been used in the past for basic ultrastructural analysis of nail plate and its soft tissue attachments.20 It can be used on nail fragments and demonstrates the details of intercellular and intracellular structures within the nail plate.
106.9 OTHER TECHNIQUES Laser Doppler can be used to assess the blood flow in the nail unit and has been applied alongside capillaroscopy in diabetics21 and in Raynaud’s disease22 and normals.23 This combined approach provides an opportunity to try to correlate anatomical and functional aspects of blood flow. Dynamic aspects of flow are accentuated by cooling and rewarming to broaden the range for evaluation. Optical coherence tomography produces a series of cross-sectional images down to a depth of 1 mm, separated by 15 mm. It has some potential for examining periungual tissues, but has been little explored as a technique.24 Confocal microscopy is in a similar category, where light penetrates the unsectioned tissue to give three-dimensional information and has been used to examine the normal nail plate25 and to evaluate onychomycosis.26 Older techniques for the assessment of shape, and clubbing in particular, include brass templates,27 shadow graphs,28 plaster casts and planimetry,29 and plethysmography.30
106.10 SCORING SYSTEMS A range of scoring systems have been developed to measure disease or suffering in nail disease. Objective measures of disease mainly pertain to psoriasis and onychomycosis. Both of these are the subject of therapeutic trials, and it is important to have some objective measure of the disease activity. de Jong et al.31 has used a nail area severity (NAS) score, consisting of the separate parameters nail
Methods for Nail Assessment: An Overview
A
B
C
D
E
FIGURE 106.7 Five nails demonstrating the manner of dividing the nail into eight zones, representing 12.5% surface area per zone. This can lead to problems with interpretation — for instance, where the partial nail loss of (A) leads to uncertainty concerning where to place the transverse division. In (E), where there is superficial involvement, it is apparent why this scoring system might not work — although it is a score of disease as detected at the surface. In (D), part of every upper section is clear and part abnormal. If the nail was clipped vigorously, the fractions would alter.
TABLE 106.1 Scoring System for Evaluation of a Pigmented Streak in a Nail and Diagnosis of Subungual Melanoma A B C D E F
Age of patient (fifth to seventh decade) Breadth of streak > 3 mm Color change from brown to black High-risk digit, such as thumb or big toe Extension of the pigment onto periungual tissues (Hutchinson’s sign) Family history of melanoma or dysplastic nevi
From Levit, E.K. et al., J. Am. Acad. Dermatol., 42, 269–274, 2000.
pitting area, number of nail pits, subungual keratosis, onycholysis, oil spots, and a score for overall improvement. Baran32 has used a similar system, rating features on a scale of 1 to 3. A further scoring system for the same disease evaluates some of the subjective aspects of the disease, including pain and disability.33 In onychomycosis, the main scoring variable is percentage of involved nail surface. A common standard for judging this is to bisect the nail transversely, then longitudinally, giving quadrants. Each lateral segment is then bisected again with a longitudinal line, resulting in
915
division of the nail into eight segments, representing 12.5% of surface area each (Figure 106.7). There can be problems of interpretation with this method, as illustrated in the figure. When the nail is naturally small or partially disintegrated, this will mean that there is less diseased tissue than in someone who may have bigger nails but a smaller percentage score. Attempts have also been made to develop an onychomycosis disease-specific questionnaire that assesses functional impact, social stigma, and psychological distress associated with the disease.34 As with most quality of life indicators, the group is preselected by having sought treatment. This will tend to increase measures of subjective suffering, especially if working within a health care system where it is in the interests of the clinician that the patient pursues treatment. Where there are pigmented streaks of the nail, Levit et al.35 has proposed an ABC scoring system for clinical detection of streaks likely to be subungual melanoma.
106.11 STRENGTH Nail strength is partly due to its composition and partly its anatomical relationships with the underlying bone and surrounding soft tissues. The main protein constituent is keratin, the most abundant of intracellular proteins in epithelial cells. The nail plate is a modified epithelium and has a different complement of keratins from those in skin. In addition to those often termed soft or epithelial keratins, there are hard or hair/nail keratins. These contain a higher fraction of sulfur-rich amino acids than their soft counterparts. These enhance disulfide bridging between the threedimensional convolutions of the keratin molecule. This increases the rigidity and strength of the protein and creates a molecular environment that excludes water. This in turn makes the nail strong and resistant to chemical breakdown. One measure of this strength is the in vitro preparation needed for nail specimens when undergoing amino acid analysis.
106.12 MEASURING NAIL STRENGTH Several techniques have been developed to study the physical properties of nails.36–38 Techniques have been described to test the tearing, flexural, and tensile strength of nails.37 Finlay et al.38 devised a nail flexometer able to repeatedly flex longitudinal nail sections through 90˚, recording the number it took to fracture the nail. In this way, the strength could be quantified. Soaking the nails in water increased their flexibility.38 Wessel et al.’s work39 suggests that this may be due to the loosening of the alpha helix conformation seen with Raman spectroscopy in vivo after soaking in water for 10 min, where water occupies the interstices.
916
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 106.2 Different Methods of Nail Constituent Analysis Method
Element
Reference
I. Structural and Mineral Constituents Raman spectroscopy Immunohistochemistry Near infrared spectrometer Polymerase chain reaction Electron microscopy X-ray diffraction Colorimetry
Water, proteins, and lipid Keratin Water Deoxyribonucleic acid Cystine Mg, Cl, Na, Ca, S, Cu Fe
46 47 48 49, 50 51 52, 53 54
II. Exogenous Materials Atomic absorption spectometry Mass fragmentography Gas chromatography–mass spectroscopy High-performance liquid chromatography Flow injection hydride generation atomic Inductively coupled plasma–mass spectrometry Absorption spectrometry
Cd, Pb, Zn, Ca, Cr, Fe, Cu, Mn, Ni, Co, Na, K Methamphetamine Amphetamine, cocaine, cannabinoids Nicotine, terbinafine, testosterone, pregnenolone Arsenic Arsenic
53, 55, 56 57 58–61 62–64 65 66
DNA Furosine (glycosylated keratin) Lipid: triglyceride Ni Steroid sulfatase Zinc, selenium
67 68 51 69 70 71–73
III. Biological Markers Polymerase chain reaction High-performance liquid chromatography Microscopy Adsorption differential pulse voltametry Enzymic assay Neutron activation analysis
Zaun40 has used a method of assessment of brittleness that relies on the swelling properties of nail, employing the technique before and after therapy for brittle nails. Splitting can be partially overcome by applications of emollient after soaking the nails in water.
106.13 PERMEABILITY Nail permeability is relevant to topical drugs on the dorsal surface and systemic drugs from the ventral surface. Transonychial water loss can be measured in vivo41,42 and in vitro,43 where the nail appears to be 1000 times more permeable to water than is skin. This means that watersoluble drugs are likely to penetrate the nail as long as the molecular size is not large.44 However, most drug delivery processes across nail need more than fragments for their assessment and are not covered here. Nevertheless, it is possible to assess movement of systemic drugs into nail by obtaining material taken either from bore holes in the nail plate or from the free edge. This is particularly relevant for systemic antifungals taken for onychomycosis.45
REFERENCES 1. Goyal S, Griffiths WAD. An improved method of studying fingernail morphometry: application to the early detection of fingernail clubbing. J Am Acad Dermatol 39:640–642, 1998. 2. Bergman R, Sharony L, Schapira D, Nahir MA, BalbirGurman A. The handheld dermatoscope as a nail-fold capillaroscopic instrument. Arch Dermatol 139:1027–1030, 2003. 3. Cutolo M, Pizzorni C, Tuccio M, Burroni A, Craviotto C, Basso M, Seriolo B, Sulli A. Nailfold videocapillaroscopic patterns and serum autoantibodies in systemic sclerosis. Rheumatology (Oxford) 43:719–726, 2004. 4. Nagy Z, Czirjak L. Nailfold digital capillaroscopy in 447 patients with connective tissue disease and Raynaud’s disease. J Eur Acad Dermatol Venereol 18:62–68, 2004. 5. Bhushan M, Moore T, Herrick AL, Griffiths CE. Nailfold video capillaroscopy in psoriasis. Br J Dermatol 142:1171–1176, 2000. 6. Jones BF, Oral M, Morris CW, Ring EF. A proposed taxonomy for nailfold capillaries based on their morphology. IEEE Trans Med Imaging 20:333–341, 2001.
Methods for Nail Assessment: An Overview
7. Hu Q, Mahler F. New system for image analysis in nailfold capillaroscopy. Microcirculation 6:227–235, 1999. 8. Nikkels-Tassoudji N, Piérard-Franchimont C, de Doncker P, Piérard GE. Optical profilometry of nail dystrophies. Dermatology 190:301–304, 1995. 9. Piérard GE, Piérard-Franchimont C. Dynamics of psoriatic trachyonychia during low dose cyclosporin A treatment: a pilot study on onychochronobiology using optical profilometry. Dermatology 192:116–119, 1996. 10. de Doncker P, Piérard GE. Acquired nail beading in patients receiving itraconazoleaan indicator of faster nail growth? A study using optical profilometry. Clin Exp Dermatol 19:404 –406, 1994. 11. Marks R. Histochemical applications of skin surface biopsy. Br J Dermatol 86:20–26, 1972. 12. Hashimoto K. New methods for surface ultrastructure: comparative studies of scanning electron microscopy, transmission electron microscopy and replica method. Int J Dermatol 13:357–381, 1974. 13. Drapé JL, Wolfram-Gabel R, Idy-Peretti I, et al. The lunula: a magnetic resonance imaging approach to the subnail matrix area. J Invest Dermatol 106:1081–1085, 1996. 14. Drape JL. Imaging of the tumors of the perionychium. Hand Clin 18:655–670, 2002. 15. Drapé JL, Idy-Peretti I, Goettmann S, et al. MR imaging of digital mucoid cysts. Radiology 200:531–536, 1996. 16. Theumann NH, Goettmann S, Le Viet D, Resnick D, Chung CB, Bittoun J, Chevrot A, Drape JL. Recurrent glomus tumors of fingertips: MR imaging evaluation. Radiology 223:143–151, 2002. 17. de Berker D, Mawhinney B, Sviland L. Quantification of regional matrix nail production. Br J Dermatol 134:1083–1086, 1996. 18. Reisberger EM, Abels C, Landthaler M, Szeimies RM. Histopathological diagnosis of onychomycosis by periodic acid-Schiff-stained nail clippings. Br J Dermatol 148:749–754, 2003. 19. Wallis MS, Bowen WR, Guin JD. Pathogenesis of onychoschizia (lamellar dystrophy). J Am Acad Dermatol 24:44 –48, 1991. 20. Parent D, Achten G, Stouffs-Vamhoof F. Ultrastructure of the normal human nail. Am J Dermatopathol 7:529–535, 1985. 21. Meyer MF, Pfohl M, Schatz H. Assessment of diabetic alterations of microcirculation by means of capillaroscopy and laser-Doppler anemometry. Med Klin (Munich) 96:71–77, 2001. 22. Creutzig A, Hiller S, Appiah R, Thum J, Caspary L. Nailfold capillaroscopy and laser Doppler fluxmetry for evaluation of Raynaud’s phenomenon: how valid is the local cooling test? Vasa 26:205–209, 1997. 23. Lutolf O, Chen D, Zehnder T, Mahler F. Influence of local finger cooling on laser Doppler flux and nailfold capillary blood flow velocity in normal subjects and in patients with Raynaud’s phenomenon. Microvasc Res 46:374–382, 1993.
917
24. Welzel J, Lankenau E, Birngruber R, Engelhardt R. Optical coherence tomography of the human skin. J Am Acad Dermatol 37:958–963, 1997. 25. Kaufman SC, Beuerman RW, Greer DL. Confocal microscopy: a new tool for the study of the nail unit. J Am Acad Dermatol 32:668–670, 1995. 26. Hongcharu W, Dwyer P, Gonzalez S, Anderson RR. Confirmation of onychomycosis by in vivo confocal microscopy. J Am Acad Dermatol 42:214–216, 2000. 27. Stavem P. Instrument for estimation of clubbing. Lancet 2:7–8, 1959. 28. Bentley D, Moore A, Schwachman H. Finger clubbing: a quantitative survey by analysis of the shadowgraph. Lancet 2:164–167, 1976. 29. Regan GM, Tagg B, Thomson ML. Subjective measurement and objective measurement of finger clubbing. Lancet 1:530–532, 1967. 30. Cudowicz P, Wraith DG. An evaluation of the clinical significance of clubbing in common lung disorders. Br J Tuberculous Dis Chest 51:14–31, 1957. 31. de Jong EM, Menke HE, van Praag MC, van De Kerkhof PC. Dystrophic psoriatic fingernails treated with 1% 5fluorouracil in a nail penetration-enhancing vehicle: a double-blind study. Dermatology 199:313–318, 1999. 32. Baran RL. A nail psoriasis severity index. Br J Dermatol 150:568–569, 2004. 33. de Jong EM, Seegers BA, Gulinck MK, Boezeman JB, van de Kerkhof PC. Psoriasis of the nails associated with disability in a large number of patients: results of a recent interview with 1,728 patients. Dermatology 193:300–303, 1996. 34. Turner RR, Testa MA. Measuring the impact of onychomycosis on patient quality of life. Qual Life Res 9:39–53, 2000. 35. Levit EK, Kagen MH, Scher RK, Grossman M, Altman E. The ABC rule for clinical detection of subungual melanoma. J Am Acad Dermatol 42:269–274, 2000. 36. Baden HP. The physical properties of nail. J Invest Dermatol 55:115, 1970. 37. Maloney MJ, Paquette EG. The physical properties of fingernails. I. Apparatus for physical measurements. J Soc Comp Chem 28:415, 1977. 38. Finlay AF, Frost P, Keith AC, Snipes W. An assessment of factors influencing flexibility of human fingernails. Br J Dermatol 103:357–365, 1980. 39. Wessel S, Gniadecka M, Jemec GB, Wulf HC. Hydration of human nails investigated by NIR-FT-Raman spectroscopy. Biochim Biophys Acta 17:210–216, 1999. 40. Zaun H. Brittle nails. Objective assessment and therapy follow-up. Hautarzt 48:455–461, 1997. 41. Jemec GBE, Agner T, Serup J. Transonychial water loss: relation to sex, age and nail plate thickness. Br J Dermatol 121:443–446, 1989. 42. Spruit D. Measurement of water vapor loss through human nail in vivo. J Invest Dermatol 56:359–361, 1971. 43. Walters KA, Flynn GL, Marvel JR. Physicochemical characterization of the human nail. 1. Pressure sealed apparatus for measuring nail plate permeability. J Invest Dermatol 76:76–79, 1981.
918
Handbook of Non-Invasive Methods and the Skin, Second Edition
44. Mertin D, Lippold BC. In vitro permeability of the human nail and of a keratin membrane from bovine hooves: influence of the partition coefficient octanol/water and the water solubility of drugs on their permeability and maximum flux. J Pharm Pharmacol 49:30–34, 1997. 45. Munro CS, Shuster S. The route of rapid access of drugs to the distal nail plate. Acta Dermato-Venereol 72:387–388, 1992. 46. Gniadecka M, Nielsen O, Christensen D, et al. Structure of water, proteins and lipids in intact human skin, hair and nail. J Invest Dermatol 110:393–-398, 1998. 47. Heid HW, Moll I, Franke WW. Patterns of expression of trichocytic and epithelial cytokeratins in mammalian tissues. Differentiation 37:215–230, 1988. 48. Egawa M, Fukuhara T, Takahashi M, Ozaki Y. Determining water content in human nails with a portable near-infrared spectrometer. Appl Spectrosc 57:473–478, 2003. 49. Kaneshige T, Takagi K, Nakamura S, et al. Genetic analysis using fingernail DNA. Nucleic Acid Res 20:5489–5490, 1992. 50. Tahir M, Watson N. Typing of DNA HLA-DQa alleles extracted from human nail material using polymerase chain reaction. J Forens Sci 40:634–636, 1995. 51. Salamon T, Lazovic-Tepavac O, Nikulin A, et al. Sudan IV positive material of the nail plate related to plasma triglycerides. Dermatologica 176:52–54, 1988. 52. Sirota L, Straussberg R, Fishman P, Dulitzky F, Djaldetti M. X-ray microanalysis of the fingernails in term and preterm infants. Pediatr Dermatol 5:184–186, 1988. 53. Forslind B. Biophysical studies of the normal nail. Acta Dermato-Venereol 50:161–168, 1970. 54. Jacobs A, Jenkins DJ. The iron content of finger nails. Br J Dermatol 72:145–148, 1960. 55. Wilhelm M, Hafner D, Lombeck I, Ohnesorge FK. Monitoring of cadmium, copper, lead and zinc status in young children using toenails: comparison with scalp hair. Sci Total Environ 103:199–207, 1991. 56. Nowak B. Occurrence of heavy metals, sodium, calcium and potassium in human hair, teeth and nails. Biol Trace Elem Res 52:11–22, 1996. 57. Suzuki S, Inoue T, Hori H, Inayama S. Analysis of methamphetamine in hair, nail, sweat and saliva by mass fragmentography. J Anal Toxicol 13:176–178, 1989. 58. Suzuki O, Hattori H, Asano M. Nails as useful materials for detection of methamphetamine or amphetamine abuse. Forens Sci Int 24:9–16, 1984. 59. Cirimele V, Kintz P, Mangin P. Detection of amphetamines in fingernails: an alternative to hair analysis. Arch Toxicol 70:68–69, 1995.
60. Miller M, Martz R, Donnelly B. Drugs in keratin samples from hair, fingernails and toenails. In Second International Meeting on Clinical and Forensic Aspect of Hair Analysis, Genoa, Italy, June 6–8, 1994, p. 39 (abstract). 61. Raharjo TJ, Verpoorte R. Methods for the analysis of cannabinoids in biological materials: a review. Phytochem Anal. 15:79–94, 2004. 62. Al-Delaimy WK, Mahoney GN, Speizer FE, Willett WC. Toenail nicotine levels as a biomarker of tobacco smoke exposure. Cancer Epidemiol Biomarkers Prev 11:1400–1404, 2002. 66. Dykes PJ, Thomas R, Finlay AY. Determination of terbinafine in nail samples during treatment for onychomycoses. Br J Dermatol 123:481–486, 1990. 67. Choi MH, Yoo YS, Chung BC. Measurement of testosterone and pregnenolone in nails using gaschromatography-mass spectrometry. J Chromatogr B Biomed Sci Appl 25:495–501, 2001. 68. Das D, Chatterjee A, Badal K, et al. Arsenic in ground water in six districts of West Bengal, India: the biggest arsenic calamity in the world. Analyst 120:917–924, 1995. 69. Chen KL, Amarasiriwardena CJ, Christiani DC. Determination of total arsenic concentrations in nails by inductively coupled plasma mass spectrometry. Biol Trace Elem Res 67:109–125, 1999. 70. Oz C, Zamir A. An evaluation of the relevance of routine DNA typing of fingernail clippings for forensic casework. J Forens Sci 45:158–160, 2000. 71. Sueki H, Nozaki S, Fujisawa R, et al. Glycosylated proteins of skin, nail and hair: application as an index for long-term control of diabetes mellitus. J Dermatol 16:103–110, 1989. 72. Gamelgaard B, Anderson JR. Determination of nickel in human nails by adsorption differential-pulse voltametry. Analyst 110:1197–1199, 1985. 73. Matsumoto T, Sakura N, Ueda K. Steroid sulphatase activity in nails: screening for X-linked ichthyosis. Pediatr Dermatol 7:266–269, 1990. 74. Rogers M, Thomas DB, Davis S, et al. A case control study of oral cancer and pre-diagnostic concentrations of selenium and zinc in nail tissue. Int J Cancer 48:182–188, 1991. 75. Van Noord PAH, Collette HJA, Maas MJ, de Waard F. Selenium levels in nails of premenopausal breast cancer patients assessed prediagnostically in a cohort-nested case-referent study among women screened in the DOM project. Int J Epidemiol 16 (Suppl.):318–322, 1987. 76. Yoshizawa K, Willett WC, Morris SJ, et al. Study of the prediagnostic selenium levels in toenails and the risk of advanced prostate cancer. J Natl Cancer Inst 90:1219–1224, 1998.
of Longitudinal 107 Measurement Nail Growth Jeffrey S. Roth and Richard K. Scher Department of Dermatology, College of Physicians and Surgeons, Columbia University, New York, New York
CONTENTS 107.1 Introduction ..........................................................................................................................................................919 107.2 Historical Overview .............................................................................................................................................919 107.3 Measurement of Nail Growth ..............................................................................................................................920 107.3.1 Fixed Landmarks ...................................................................................................................................920 107.3.2 Distal Landmarks...................................................................................................................................921 107.3.3 Technique of Measurement ...................................................................................................................921 107.4 Summary...............................................................................................................................................................921 References .......................................................................................................................................................................921
107.1 INTRODUCTION Initial thoughts on the longitudinal measurement of nail growth turn to the simple and intuitive application of a ruler to the nail and recording of differences in length over time. While this is sound in principle, precision in nail growth measurement demands a more rigorous technique and has been the subject of numerous articles over many decades. Achieving consistency, reproducibility, and accuracy of results requires that there be consensus about the definition of stable landmarks, avoiding sources of variability such as nail plate wear. In addition, nail growth has been reported to vary with season, time of day, digit, sex, and state of health,1 making it important to establish a biologically meaningful interval over which nail growth should be measured. Though a technique may claim to allow measurement of nail growth over a 15-min period,2 this may not be meaningful biologically. Who might be interested in measuring nail growth? Research applications would include the response of the onychomycotic nail to novel antifungals under investigation or to quantify the response of psoriatic nails to treatment. Such applications would require relative precision to allow meaningful statistical evaluation of study data. Clinical applications might be the measurement of Beau’s lines to time recent trauma to the nail matrix from disease or chemotherapy, the quantification of nail growth in a patient who complains of nails that grow poorly, or the establishing of efficacy of specific therapy for nail disease
in the clinic setting. Clinical and research applications may impact differently on the choice of technique. Research techniques should be precise, may involve specialized equipment, need not be quick, and need not (though should) be inexpensive. Clinical methods should be rapid, inexpensive, need no specialized equipment, and can be relatively less precise. Both research and clinical applications should do no harm to the patient (an intuitive “given” that is not always adhered to). The following will take into account the history of the treatment of this issue and a review of the major techniques that have been described.
107.2 HISTORICAL OVERVIEW The development of methods for measuring nail growth spans many decades. Interest in the growth of nails initially stemmed from the notion that the health of the nail and the vigor of its growth closely parallel the general health. While this notion remains generally sound, it is now appreciated that although the nail is not a direct reflection of the general health, it may offer clues to underlying systemic illness. Early reports of methods of measurement and of patterns of nail growth were flawed by poor scientific rigor and anecdotal experience. Thus, Berthold3 measured only the growth of his own fingernails (using the lunula as a fixed landmark). Similarly, Sharpley-Schafer4 measured only his own hand. Bloch5 criticizes the 919
920
Handbook of Non-Invasive Methods and the Skin, Second Edition
previous pioneering work of Dufour,6 published in 1872, which measures the distance between the cuticle and a permanent mark made in the nail plate using silver nitrate, stating that the cuticle is a constant landmark only in the well groomed. (He offers in place of Dufour’s methodology two techniques, one in which a permanent mark in the nail plate is measured against a fixed india ink tattoo on the dorsal finger and another in which the fixed point on the nail plate is measured against the knuckle of the finger bent at a fixed angle.) LeGros Clark and Buxton7 criticize the measurement of nail growth using the proximal nail fold as a landmark. Their methodology in turn is derided as too vague in subsequent studies. Internal inconsistencies abound as well. For example, they use as their fixed landmark a point on the nail “about 2 mm from the margin of the lunula,” yet they go on to give measurement of change in “micromillimeters” (microns). William Bennett Bean,8,9 perhaps the most indefatiguable nail watcher of all time, measured his own fingernails over a 35-year period both longitudinally and by weighing his nail clippings. His perseverance allowed him to observe the deceleration of nail growth with advancing years, from a rate of 0.123 mm/day at the age of 32 to 0.095 mm/day at age 67. Intuitively, measuring nail growth requires no sophisticated instrumentation; nevertheless, it has been subject to sometimes faddish application of technology as it became available. The modern age brought with it a flurry of new techniques: photography,10,11 magnifying devices such as the “biomicrometer” of Basler18 and the tool of LeGros Clark and Buxton,7 the “split image rangefinder adapted to a trinocular microscope” described by Orentreich et al.,2 and time-lapse photography.12 Authorities have opined on the suitability of anatomic structures for use as landmarks from which to measure the growing nail. Some favor the proximal nail fold.1 Others prefer the lunula, though it is conceded that on some fingers the lunular margin is in fact blurred under magnification, and therefore unsuitable for submillimeter measurements.7 Furthermore, many individuals lack a visible lunula. Some favor the use of bony landmarks by xray or physical examination, relying on the intimate relationship between the distal phalanx and the nail plate.10,12,14 The results of animal studies may be difficult to extrapolate to humans, owing to fundamental differences in the shape of the nail in other species, such as the rat.17 In summary, such a simple issue as how to measure the growing nail is beset with good-natured controversy. We intend to endorse no single technique, but to advise a common-sense approach: simplicity, reliability, ease, inexpensiveness, and, above all, avoidance of harm to the patient.
107.3 MEASUREMENT OF NAIL GROWTH We will approach the measurement of nail growth by presenting techniques and points of view in each of several aspects of the problem.
107.3.1 FIXED LANDMARKS The issue of finding a stable landmark as a point of reference from which to measure nail growth is important since if this point varies between determinations, the measurements will be rendered meaningless. Several proximal (fixed) landmarks have been proposed: 1. The cuticle. This landmark appears to be relatively stable in patients who do not manipulate or cut back the cuticle, especially if submillimeter measurements are not needed. The cuticle is obviously invalid as a point of reference if pushed back or cut. 2. The proximal nail fold. This is similar to the cuticle in its ability to be pushed back, though not as easily or permanently as the cuticle. Whether the edge of the proximal nail fold can be precisely defined for very fine measures is unclear. This has been shown to have a small interobserver error rate13 and may be the most suitable landmark. 3. The distal interphalangeal (DIP) joint. This is among the more precise landmarks. It can be used in two ways: one in which a mold or rigid brace is constructed so that the DIP joint is flexed at an identical angle at each determination and the other as a radiographic landmark (used with a radiopaque marker cemented to the nail plate). The first technique requires a simple tool and may be unsuitable for submillimeter measurements. The second technique, though perhaps the most precise, involves exposure to ionizing radiation and is therefore suitable only in research settings with informed consent, if ever. 4. The lunula. Using this structure has several disadvantages. First, not everyone has a visible lunula, especially on the index, third, ring, and small fingers. Second, while the lunula appears well defined, under magnification its edge is somewhat vague and would therefore introduce unnecessary error into measurement. 5. A structure cemented to the skin of the dorsal distal phalanx. This has the advantage of being relatively stable and easy to see, but may come loose during a meaningful interval of nail
Measurement of Longitudinal Nail Growth
growth. This technique was cited more than 50 years ago12 and later echoed.2
921
physician to follow the dictum primum non nocere, “first, do no harm.”
107.3.2 DISTAL LANDMARKS Once the issue of proximal landmarks is settled, it becomes relatively easy to choose a way of marking the growing portion of the nail. It makes no difference if the nail is etched, drilled, or if a radiopaque band is cemented to the nail plate. The only specific requirement is that the marking be permanent over the course of measurement.
107.3.3 TECHNIQUE
OF
MEASUREMENT
Several options exist in this regard: 1. A straight rule with at least millimeter (and preferably submillimeter) increments indicated. Such a device should be used with a loupe of 8× magnification or greater. 2. A caliper similar to those used in electrocardiography,16 which is then applied to the straight rule. This is preferable, as the direct application of the straight rule to the curved nail plate may introduce inaccuracy. 3. Devices such as the “split image rangefinder”2 have only very specialized applications, as does time-lapse photography.12
107.4 SUMMARY The technique employed to measure the longitudinal growth of nails should satisfy several criteria: it should be simple, disfigure as little as possible, be safe, and be biologically relevant. The extremes of precision that many methods aim to achieve are only of value in research settings. Except in specialized situations when very short term growth rates are measured, the rate of nail plate growth is so variable between day and night and season15 in an individual as to make overly precise or overly frequent measurements meaningless. No dermatologic diagnosis rests on accurate measurement of longitudinal growth of the nail. In pursuing his interest in this parameter, therefore, it behooves the
REFERENCES 1. Dawber, R. and Baran, R., Nail growth, Cutis, 39, 99, 1987. 2. Orentreich, N., Markofsky, J., and Vogelman, J.H., The effect of aging on the rate of linear nail growth, J. Invest. Dermatol., 73, 126, 1979. 3. Berthold, Beobachtungen uber das quantitative Verhaltniss der Nagel: und Haarbildung beim Menschen, Mullers Arch., 156, 1850. 4. Sharpley-Schafer, E., Relative growth of nails on right hand and left hand respectively: on seasonal variations in rate; and on influences of nerve section upon it, Proc. R. Soc. Edinburgh, 51(1), 8. 5. Bloch, A.M., Etude de la croissance des ongles, C. R. Soc. Biol., 58, 253, 1905. 6. Dufour, 1872. 7. LeGros Clark, W.E. and Buxton, L.H.D., Studies in nail growth, Br. J. Dermatol. Syph., 50, 221, 1938. 8. Bean, W.B., A note on fingernail growth, J. Invest. Dermatol., 20, 2, 1953. 9. Bean, W.B., Nail growth. Thirty five years of observation, Arch. Int. Med., 140, 73, 1980. 10. Babcock, M.J., Methods of measuring fingernail growth rates in nutritional studies, J. Nutr., 55, 323, 1955. 11. Sibinga, M.S., Observations on growth of fingernails in health and disease, Pediatrics, 24, 225, 1959. 12. Morton, R., Visual assessment of nail growth, J. Audiovis. Media Med., 14(1), 31, 1991. 13. Dawber, R., Fingernail growth in normal and psoriatic subjects, Br. J. Dermatol., 82, 454, 1970. 14. Kandil, E., Accurate measurement of nail growth, Int. J. Dermatol., 11(1), 54, 1972. 15. Scher, R.K. and Daniel, R.C., Nails: Therapy, Diagnosis, Surgery, W.B. Saunders, Philadelphia, 1991. 16. Hillman, R.W., Fingernail growth in the human subject. Rates and variations in 300 individuals, Hum. Biol., 27, 255, 1955. 17. Godwin, K.O., An experimental study of nail growth, J. Nutr., 69, 121, 1959. 18. Basler, A., Growth processes in fully developed organisms, Med. Klin., 33, 1664, 1937.
108 Measurement of Nail Thickness Gregor B.E. Jemec Department of Dermatology, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark
CONTENTS 108.1 Variable.................................................................................................................................................................923 108.2 Methods ................................................................................................................................................................923 108.3 Correlation between the Methods........................................................................................................................924 108.4 Practical Recommendations .................................................................................................................................924 References .......................................................................................................................................................................924
The nail is a well-defined keratin structure. The larger part of it is clearly visible and immediately accessible for studies. Yet comparatively few studies have been made of the nail compared with, e.g., hair. In these, mostly nail longitudinal growth has been measured, although a few studies have also paid attention to the structure of the nail plate and volume of nail growth. Studies of nail thickness may be relevant to quantification of nail matrix output, to penetration studies, and to studies of the nails as markers of nondermatological disease.1
108.1 VARIABLE The variable studied is simply the thickness of the nail, defined as the distance between the superficial and profound surfaces of the nail. The nail plate is of uneven thickness, being thinner at the nail matrix where it is formed and gradually thickening toward the free edge.2 All measurements of the thickness should therefore be made at a defined point along the nail plate, most often at the free edge or immediately proximal to this. The thickness is most appropriately expressed in millimeters. Normal values range between 0.3 and 0.9 mm, with the first finger having a thicker nail than the fifth finger. Toenails are usually thicker than fingernails, and men appear to have thicker nails than women.
108.2 METHODS Several noninvasive methods are available for the measurement of nail thickness: simple callipers or micrometers, high-frequency ultrasound, and optical coherence tomography. However, only the first two are discussed, as the necessary apparatus is more readily available.
Simple callipers or micrometers are readily available and have sufficient precision to measure nails. The nails can be measured in vivo provided that a sufficient free length of nail is available for the instrument to grip at the distal end of the nail plate. The instrument should also preferably exert a standardized pressure on the gripped tissue, although this is less important in hard tissues such as nail than in, e.g., skin. The positioning should always be perpendicular to the nail plate surface, and the shortest measurements are the most correct.3 In one published study the coefficient for such measurements was 5.3% (SD = 2.4%). The main disadvantages of the method are that it can be difficult to get a grip on the free edge of the nail if it is cut short, and that it is only the thickness at the free edge that is measured. The nail progressively increases in thickness toward the free edge, and the calliper measurements therefore represent a maximum value of nail thickness rather than, e.g., an average or a minimum value. High-frequency ultrasound offers the advantage of being able to measure the thickness of the nail plate anywhere along the nail, but also here repeated measurements are necessary to identify the correct (smallest) thickness of the plate. It should be specified as well as possible where along the nail plate the measurements have been made.4 Both A- and B-scans give the thickness, but the B-scan provides additional information (Figure 108.1). Optical coherence tomography gives images similar to Bscans. The speed of sound within the nail has been found to be 2459 m/sec. This can be used for exact calculations of distance. Nail hydration may lead to slowing of the speed of sound within the nail, and hence to an overestimation of 923
924
Handbook of Non-Invasive Methods and the Skin, Second Edition
108.4 PRACTICAL RECOMMENDATIONS B
A D C
FIGURE 108.1 Ultrasound image with A-scan superimposed on B-scan showing the thickness of the normal human fingernail, marked between the arrows. A: Ultrasound membrane (machine). B: Proximal nail fold. C: Dorsal nail plate. D: Ventral nail plate.
actual distances. The coefficient of variation for ultrasound measurements was 4% (SD = 1.3%) in one study.
108.3 CORRELATION BETWEEN THE METHODS The two methods correlate well (r = 0.79, p < 0.001).3
Position your instrument of choice so as to be perpendicular to the surface of the nail. The smallest thickness is the correct value. If you are using ultrasound, mark where you measure, and do not soak the nail.
REFERENCES 1. Wollina, U., Berger, M., and Karte, K., Calculation of nail plate and nail matrix parameters by 20 MHz ultrasound in healthy volunteers and patients with skin disease, Skin Res. Technol. 7, 60, 2001. 2. Johnson, M. and Shuster, S., Continuous formation of nail along the bed, Br. J. Dermatol. 128, 277, 1993. 3. Finlay, A.Y., Western, B., and Edwards, C., Ultrasound velocity in human fingernail and effects of hydration: validation of in vivo thickness measurement techniques, Br. J. Dermatol. 123, 365, 1990. 4. Jemec, G.B.E. and Serup, J., Ultrasound structure of the human nail plate, Arch. Dermatol. 125, 643, 1989.
109 Image Analysis of the Nail Surface Claudine Piérard-Franchimont and Gérald E. Piérard Department of Dermatopathology, University Hospital Sart Tilman, Liège, Belgium
CONTENTS 109.1 Introduction ..........................................................................................................................................................925 109.2 Nail Surface Modifications ..................................................................................................................................925 109.2.1 Longitudinal Striations ..........................................................................................................................925 109.2.2 Herringbone Nail ...................................................................................................................................925 109.2.3 Beau’s Lines...........................................................................................................................................926 109.2.4 Pitting and Rippling...............................................................................................................................926 109.2.5 Trachyonychia........................................................................................................................................926 109.2.6 Ridging of the Nail Underface ..............................................................................................................926 109.3 Instrumental Assessments ....................................................................................................................................926 109.3.1 Static Microtopography of the Nail Surface .........................................................................................926 109.3.2 Dynamic Microtopography of the Nail Surface....................................................................................926 109.3.3 Nail Microindentation............................................................................................................................926 109.3.4 Nail Sclerometry ....................................................................................................................................927 109.4 Conclusion............................................................................................................................................................927 References .......................................................................................................................................................................926
109.1 INTRODUCTION Nail is a hard but flexible structure growing continuously and potentially submitted to many types of microtraumatisms. It may grossly look smooth, but closer examination shows it is not. The structure of the nail surface has attracted so far little interest from researchers. Descriptive reports are rarely supported by quantification of the nail plate microrelief. However, several patterns of nail surface abnormalities are well identified by clinical inspection. They result from endogenous dermatosis or from trauma and weathering.
109.2 NAIL SURFACE MODIFICATIONS 109.2.1 LONGITUDINAL STRIATIONS Longitudinal striations at the nail surface present as indented grooves separated by projecting ridges. This condition may be considered a physiological feature when presenting as shallow depressions, usually parallel, and separated by low projecting ridges. They become more prominent with age and in some particular conditions described hereafter.
Onychorrhexis consists of a series of narrow, longitudinal, parallel superficial striations with the appearance of having been scratched by an awl. Sometimes dust particles are ingrained into the nail surface. Splitting of the free edge of the nail is common. The small rectilinear projections extend from the proximal nail fold to the free edge of the nail. They may also stop short or be interrupted at regular intervals, giving rise to a beaded appearance. In some patients a wide, longitudinal median ridge has the appearance, in cross section, of a circumflex accent. Median nail dystrophy is an uncommon condition consisting of a longitudinal groove in the mid-line or just off center of the thumbnails, starting at the cuticle and growing out of the free edge. Tumors nearby the nail matrix may exert pressure and produce a single wide, deep, longitudinal groove or canal. This aspect disappears when the cause is removed.
109.2.2 HERRINGBONE NAIL Nail ridging, with oblique lines pointing centrally to meet in the mid-line, is an uncommon pattern occurring in childhood. 925
926
Handbook of Non-Invasive Methods and the Skin, Second Edition
109.2.3 BEAU’S LINES Beau’s lines are transverse grooves extending from one lateral edge of the nail to the other. They affect all nails at corresponding levels. The width of the transverse groove relates to the duration of the process that has altered the nail matrix activity. An abrupt distal limit of the groove indicates a sudden outbreak of disease. A sloping aspect suggests a more progressive onset. The proximal limit of the depression may also be abrupt or sloped.
109.2.4 PITTING
AND
RIPPLING
The eponym Rosenau’s depressions refers to nail pitting and rippling. Pits develop as a result of defective nail formation in punctate areas located in the proximal portion of the nail matrix. The surface of the nail plate may be studded in a buckshot pattern with small punctate depressions that vary in number, size, depth, and shape. It is usual to accept five pits as an arbitrary number for the physiological condition. The depth and width of the pits relate to the extent of the matrix involved. Their length is determined by the duration of the damage. Deep and irregularly shaped pits often suggest psoriasis, but they are not pathognomonic for any disease. Pits may be randomly distributed or evenly patterned in series along one or several longitudinal lines. They are sometimes arranged in a crisscross pattern, and may thus resemble the external surface of a thimble. Regular pitting may convert to rippling or ridging, and these conditions may correspond to variants of uniform pitting.
109.2.5 TRACHYONYCHIA Trachyonychia refers to a spectrum of alterations resulting in severe nail roughness, as if the surface had been rubbed with sandpaper.
109.2.6 RIDGING
OF THE
NAIL UNDERFACE
The undersurface of the nail disclosed after avulsion exhibits a topographical aspect unrelated to the outer surface of the same nail. Deep longitudinal striations are present. They deepen with age.
109.3
INSTRUMENTAL ASSESSMENTS
The distinctive features of the nail microrelief can be studied using various procedures. Clinical examination allows qualitative assessment of the gross microtopography. Low-magnification photographs under carefully controlled and repeatable conditions can be used to document the nail microrelief. More precise quantitative examinations can be made on the outer portion of the nail in vivo, or after its avulsion, or after making a copy of negative
replicas. The rigorous use of optical profilometric methods or any other microtopographic assessment1 on this material brings quantitative information. The diverse typical alterations are clearly evidenced.2–5 In addition, due to the continuous nail growth, onychochronobiology can be studied by the same means.6,7 The same methods are suitable for assessing the mechanical properties of nail in combination with the microindentation and the sclerometry methods.
109.3.1 STATIC MICROTOPOGRAPHY SURFACE
OF THE
NAIL
Quantitative assessments of the nail microtopography are usually performed by longitudinal scans. Transversal scans are less easy to interpret due to the natural curvature of the nail. When information must be obtained in this direction, it is recommended to examine sections 5 mm in length to minimize this pitfall. It should be kept in mind that native alterations are better revealed at the proximal part of the nail. Weathering and natural microabrasions may add their effects in a cumulative way when moving toward the distal part of the nail. Sources of variability such as nail plate wear should be discarded. Controlled positioning of the nail is of the upmost importance.
109.3.2 DYNAMIC MICROTOPOGRAPHY OF THE NAIL SURFACE Repeated controlled assessments over time give insight in onychochronobiology. The effects of therapies can thus be assessed. The speed of growth of the nail can be assessed simultaneously when a mark has been engraved at the initial examination. Thus, it is possible to evaluate the rate of improvement or degradation of the nail condition. A biologically meaningful interval should be respected between successive measurements. In this consideration the speed of nail growth must be taken into consideration. Indeed, there may be interdependence between the disclosed microtopography changes and variations in the nail growth rate. An example is given by Beau’s lines and beaded nail.1 Seasonal variations in the nail surface microtopography may vary from insignificant to quite obvious.6
109.3.3 NAIL MICROINDENTATION Experimental microindentation allows the study of some mechanical properties of the nail. A load is applied under controlled conditions on a small surface. The indentation is usually measured during the test procedure. If any residual plastic deformation persists after releasing the force, the imprint of the device can be observed by profilometry.
Image Analysis of the Nail Surface
109.3.4 NAIL SCLEROMETRY Sclerometry deals with the dynamic assessment of the response of an object during microabrasion. In addition to the classical abrasion parameters, profilometry may describe in another way the groove traced in the nail. The effects of nail-hardening products and nail protectors can be conveniently assessed by that way. Similarly, nail softening by xenobiotics or altered states of health can also be quantified.
109.4 CONCLUSION The nail microrelief is subjected to variability due to physiological parameters, weathering, external trauma, and pathological features altering the nail matrix. Some microtopographic alterations are linked to changes in the nail growth rate and in the nail hardness. Objective assessments of the nail surface topography have been seldom addressed in the literature. Presumably the methods developed for the skin surface microtopography can be applied to the nail apparatus. These methods could give insight into onychochronobiology, providing unique information about the physiopathological processes having affected the nail over the past weeks and months.
927
REFERENCES 1. Lévêque, J.L., EEMCO guidance for the assessment of skin topography, J. Eur. Acad. Venereol., 12, 103, 1999. 2. De Doncker, P. and Piérard, G.E., Acquired nail beading in patients receiving itraconazole: an indicator of faster nail growth? A study using optical profilometry, Clin. Exp. Dermatol., 19, 404, 1994. 3. Nikkels-Tassoudji, N., Piérard-Franchimont, C., De Doncker, P., and Piérard, G.E., Optical profilometry of nail dystrophies, Dermatology, 190, 301, 1995. 4. Piérard, G.E. and Piérard-Franchimont, C., Fractal microrelief of the skin and nail, Giorn. Int. Dermatol. Ped., 8, 75, 1996. 5. Piérard-Franchimont, C. and Piérard, G.E., Surface image analysis of nail alterations in juvenile pityriasis rubra pilaris, Skin Res. Technol., 4, 34, 1998. 6. Piérard, G.E. and Piérard-Franchimont, C., Dynamics of psoriatic trachyonychia during low dose cyclosporin A treatment. A pilot study on onychochronobiology using optical profilometry, Dermatology, 192, 116, 1996. 7. Piérard-Franchimont, C., Jebali, A., Ezzine, N., Letawe, C., and Piérard, G.E., Seasonal variations in polymorphic nail surface changes associated with diabetes mellitus, J. Eur. Acad. Dermatol. Venereol., 7, 182, 1996.
Section III Clinical Experimentation, Evaluation, and Quantification
Guidelines for Assessment 110 General of Skin Diseases Elisabeth A. Holm and Gregor B.E. Jemec Department of Dermatology, Roskilde Hospital, University of Copenhagen, Roskilde, Denmark
CONTENTS 110.1 110.2 110.3 110.4 110.5
Introduction .......................................................................................................................................................931 Validity...............................................................................................................................................................932 Reliability ..........................................................................................................................................................933 Sensitivity and Responsiveness.........................................................................................................................933 Generic Instruments ..........................................................................................................................................934 110.5.1 Sickness Impact Profile (SIP).............................................................................................................934 110.5.2 Nottingham Health Profile (NHP) ......................................................................................................934 110.5.3 Medical Outcomes Study (MOS) 36-Item Short Form (SF-36)........................................................934 110.6 Dermatology Quality of Life Instruments ........................................................................................................935 110.7 Generic Quality of Life Instruments for Skin Disorders .................................................................................935 110.7.1 Dermatology Life Quality Index (DLQI)...........................................................................................935 110.7.2 Children’s Dermatology Life Index (CDLQI) ...................................................................................935 110.7.3 Skindex................................................................................................................................................935 110.8 Disease-Specific Dermatology Quality of Life Instruments ............................................................................935 110.8.1 Atopic Dermatitis................................................................................................................................935 110.8.2 Psoriasis Index of Quality of Life (PSORIQoL) ...............................................................................936 110.9 Disease-Specific Assessment.............................................................................................................................936 110.9.1 Atopic Eczema ....................................................................................................................................936 110.9.2 Psoriasis...............................................................................................................................................938 110.9.3 Acne ....................................................................................................................................................938 110.10 Other Skin Disorders.........................................................................................................................................938 110.11 Conclusion .........................................................................................................................................................938 References .......................................................................................................................................................................940
110.1 INTRODUCTION The skin can be assessed by a diversity of methods as demonstrated by the breadth of topics covered in this book. The accurate and appropriate measurement of health outcome is an important aspect of clinical work and research, and forms the basis of good evidence-based practice. This is of particular importance in chronic recurrent diseases such as skin diseases, where the absolute endpoints in treatment are not death or survival but relative improvement. Assessing disease severity in dermatological disorders presents practical problems because laboratory tests of
disease severity often do not exist. Biopsy and blood test provide valuable information regarding diagnosis, but are often only supplementary to the assessment of disease activity. Measurement as quality of life (QoL) has grown to be an important — and sometimes the decisive — endpoint of disease severity considerations. In a busy dermatological department or practice, the assessment of the patient’s skin disease is often the quick clinical look. The outcome depends upon the doctor’s knowledge, experience, and how detailed the severity is described in the patient’s records. Comments such as “much better” or “fine” convey little to a doctor who did not see the patient previously, and probably also very little to the original author when reviewing the notes months later. A more 931
932
Handbook of Non-Invasive Methods and the Skin, Second Edition
objective and reproducible assessment would therefore contribute significantly to the quality of even routine clinical work. All disease severity assessments should satisfy basic requirements if they are to be clinically useful. These requirements are primarily validity, reliability, sensitivity, and responsiveness. These four properties are described in the following using examples from the assessment of atopic eczema.
X X X X X X X X X X X X X X X X X X X X
X X X X X X X X X X X X X X X X X X X X (a)
110.2 VALIDITY
X X X X X X X X X X
Validation of methods is the process of determining whether the method or instrument measures what it is intended to measure, and if it is useful for the intended purpose. For example, to what extent is it reasonable to claim that a clinical scoring system like SCORAD for atopic eczema (AE) really is assessing the severity of AE? Since we are attempting to measure an ill-defined variable (severity of AE), we can only infer that the instrument is valid insofar as it correlates with other observable behavior in AE. This validation process consists of a number of stages, and it is traditionally subdivided into three main aspects: (1) content, (2) criterion, and (3) construct validity. Content validity concerns the extent to which the symptoms we include in our new scoring system are sensible and reflect the intended domain (the specific disease) of interest. Are all relevant areas of interest covered in our new construct? For assessing the symptomatology of atopic eczema, the method should include items relating to all major symptoms of the disease. As an example, we design a scoring system that only includes symptoms like erythema, papulation, and edema, which are dominant features in the acute phase of the disease (Figure 110.1a). Such a scorings systems would have no content validity for the general population of patients with AE, in whom chronic lesions are prominent. We must therefore include symptoms like lichenification and skin dryness to ensure that it also covers patients in a stable phase, if we want a valid system for both the acute and chronic phases (Figure 110.1b). Methods of content validation are not amenable to formal statistical testing. To ensure that the instrument covers all the relevant issues, it is important in the construction phase to:
X X X X X X X X X X
• • •
Review literature, including published results from clinical trials Include input from specialists Collect information from relevant patient associations
After the scoring system has been constructed and before it is tested in a pilot study, it is useful to check
No content validity, since only one small area of interest is covered by our focus.
X X X X X X X X X X
Content validity, since items cover a wide aspect of the phenomena that we want to measure.
X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X (b)
FIGURE 110.1 See text for details.
whether the instrument covers the intended area by once more contacting experts in the field of interest and relevant patient groups. This critical review of an instrument after it has been constructed is called by some authors face validity. Criterion validity describes how well an instrument agrees with the true value, ideally a golden standard. For atopic eczema no golden standard exists; therefore, criterion validities for new instruments are often compared against one or more well-established instruments, e.g., SCORAD. This method may be problematic, if the motivation for creating a new tool is to compensate for inadequacies of the existing instrument. In this case, the comparison of new instruments against established instruments is of limited value. Comparing a new scoring system with an established one is, however, a good method if the motive is to develop either a shorter or simpler system. The statistical methods used for comparison of new and old are correlation, which is the method of analysis of the possible association between two continuous variables, and regression, if we want to describe more precisely the degree of association between the two methods.1 Construct validity is one of the most important properties of a measurement instrument. It is an assessment of the degree to which an instrument measures the construct that it was designed to measure. For example, does SCORAD really measure severity of atopic eczema? The process of analyzing construct validity involves several steps, and it is a lengthy and ongoing process. A hypothetical model is constructed and data are collected and analyzed. If the relationship between data and other well-known factors of the same construct relates well, the instrument appears to be valid. The greater the supporting evidence, the more confident we can be that the model is
General Guidelines for Assessment of Skin Diseases
an adequate representation of the construct that we want to measure, e.g., atopic eczema severity. Every study in which the new instrument is seen to correlate with other aspects of the disease supports the validity of the tool. In AE such supportive evidence could be, e.g., steroid hormone consumption, doctors visits, and time spent on treatment. In contrast, a single negative finding may lead to reconsideration of the whole theoretical background for the new instrument. Construct validity is the most applicable to numerical analysis of the three validation processes. For greater detail, see Fayers and Machin.2 Cases where our methods or instruments correlate positively with other aspects of our construct display convergent validity; e.g., an increase in objective SCORAD is normally followed by an increase of pruritus. If dimensions of the severity appear to be negatively correlated, this is called divergent validity. Correlation and factor analysis plays an important role in construct validation.2 One of the simpler forms of construct validation is known-groups validity. This is based on the principle that specified groups of patients are anticipated to score differently from each other, and the instrument should be able to distinguish this difference. Known-groups comparison is therefore a combination of test for validity and a form of sensitivity or responsiveness. An example of known-group validity in dermatology is Jemec and Wulf’s study,3 where two disease-specific scoring systems (PASI for psoriasis and SCORAD for AD) were compared for patients suffering from psoriasis and AD.
110.3 RELIABILITY Assessment of reliability is to determine if, e.g., a scoring system produces reproducible and consistent results. A reliable scoring system will give reproducible or similar values if it is used repeatedly on the same patient, when the patient’s condition is stable. This repeatability reliability is based upon analysis of correlations between repeated measurements. The measurements can be repeated over time (test–retest reliability), by different observers (interobserver reliability), or by the same observer on two or more occasions (intraobserver reliability). In many cases information about correlations gives insufficient data to assess the methods. If one is more interested in prediction or estimation, regression analysis should be used. The simplest method of assessing repeatability for binary data is the proportion of agreement, when the same instrument is applied on two occasions. For example, the same patients are assessed twice and the results are presented in a 2 × 2 table:
933
First Assessment Second Assessment Positive Negative Total
Positive A11 C21 AC
Negative B12 D22 BD
Total AB CD N
The number of agreements, that is, the number of patients who respond in the same way in both assessments, is A11 + D22, and the proportion of agreement is PAgree = (A11 + D22)/N Another widely used method for assessing repeatability for binary data is the kappa coefficient, κ. The kappa coefficient provides a better method than the above concept, because the kappa coefficient also reflects the agreement, which exists purely by chance. If there is perfect agreement, κ = 1. If the agreement is no better than by chance, κ = 0. If the agreement is less than would be expected by chance, κ = negative value. Pearson’s correlation is often used as a measure of reliability. This is not recommended since repeated measures may be highly correlated even though they can be systematically different. For example, if all patients in a second assessment score 20 points higher than the first assessment, and then data for both measures are plotted in a X–Y diagram, the slope for the two graphs will be exactly the same. This gives a correlation value of 1, indicating perfect association, while in reality the agreement between the two measurements is poor, as they are parallel displaced by 20. A method recommended for assessing reliability is analysis of variance (ANOVA). For details, see the Altman.1 The word reliability is rather confusingly used for another property of scale validation: internal reliability. Internal reliability assesses if the scores from different items correlate with each other and with the total scale score. Are all items related to the same latent variable? Cronbach’s coefficient α is one of the most widely used methods of assessing internal consistency. If items are uncorrelated, αCronbach = 0, and if all items are identical and have perfect correlation, αCronbach = 1. Cronbach’s coefficient α will increase with the number of items used in the scale.
110.4 SENSITIVITY AND RESPONSIVENESS Sensitivity and responsiveness are two closely related properties to repeatability reliability. Sensitivity is the instrument’s ability to detect differences between groups. This can, e.g., be the difference between patients with various degrees of disease severity or different treatment groups. Sensitivity is one of the most important properties of an instrument, since the usefulness
934
Handbook of Non-Invasive Methods and the Skin, Second Edition
of a measure is dependent upon its ability to detect clinically relevant differences. There are situations where an established sensitive method fails to detect differences. This occurs if the data only take up a small spectrum of the scale. For example, we have selected patients with either very mild (SCORAD 1 to 10) or very severe (SCORAD 93 to 103) atopic eczema. If a new instrument is only sensitive to differences, which correspond to a change in SCORAD of 10, it will not be able to detect any differences within the groups of mild or severe eczema. This phenomenon is called the floor-and-ceiling effect. An instrument that is valid and reliable in one study can therefore lose its properties in another study, depending upon the selection of sample size, type of study populations, types of interventions, etc. Sensitivity is usually assessed by cross-sectional comparisons of groups of patients in which differences are expected. In practice, sensitivity is closely related to known-groups validity. A highly sensitive scale will usually also be reliable and highly responsive. Responsiveness is closely related to sensitivity, but relates to within patients, where sensitivity relates to between patients. If a method is responsive, it is able to detect changes in a patient’s disease status over time. Responsiveness can also be regarded as providing additional evidence of our new instrument’s validity, since it confirms that the anticipated responses occur when the patient’s status changes. The construct of our hypothetical model is confirmed. Finally, besides being valid, reliable, and sensitive to changes, our instrument also needs to be acceptable for both patient and investigator.
110.5 GENERIC INSTRUMENTS Generic instruments have been constructed for general use, irrespective of the illness or condition of the patients, to compare data between different diseases, including skin disorders. The generic questionnaires may often also be applicable to healthy people, and are constructed to cover a spectrum from healthy people to very sick patients. Because of the width of the spectrum, individual sensitivity and responsiveness suffer. The first generic instruments were developed primarily with population surveys in mind, although they were later applied in clinical trials. They are commonly described as quality of life (QoL) questionnaires, but many of these only measure physical symptoms, and are therefore more appropriately termed health status instruments than QoL instruments. Many different generic assessment instruments exist. In this chapter, three different instruments will be described: Sickness Impact Profile (SIP), Nottingham Health Profile (NHP), and Medical Outcomes Study 36Item Short Form (SF-36).
110.5.1 SICKNESS IMPACT PROFILE (SIP)* Bergner et al.4 developed SIP; the final version was finished in 1981. The SIP is a questionnaire designed to measure sickness-related behavioral dysfunction and was developed for use as an outcome measure in the evaluation of health care. The full questionnaire (16 pages) consists of 136 items and takes about 30 minutes to complete. The items describe everyday activity and cover the impact upon health of activities, behavior, social functioning, and emotional well-being. All items are negatively worded, representing dysfunction. Twelve main areas of dysfunction are covered. A standard scoring method is used for each of these 12 dysfunctions. These 12 represent two higher-order dimensions, (1) physical and (2) psychosocial, which can be scored in a similar manner as the 12. SIP has been tested in respect to validity, reliability, responsiveness, and sensitivity. Test–retest reliability (r = 0.92) and internal consistency (r = 0.94) were high, and SIP is sensitive even to minor changes in morbidity.
110.5.2 NOTTINGHAM HEALTH PROFILE (NHP)** Hunt et al.5 developed the NHP in 1981. It is often used in population studies of general health assessment and in clinical trials. Although it is less sensitive to minor changes, it was mainly developed to assess whether there are any health problems. It measures health-related quality of life within the sections of energy, sleep, emotions, pain, mobility, and social isolation, as well as the frequency of health-related problems pertaining to paid employment, housework, hobbies, family life, social life, sex life, and holidays. It consists of 38 items, where the respondents are given a list of statements that they answer yes or no. As for SIP, all items are negatively worded, representing dysfunction. Compared to SIP, the NHP is short and easy to complete. The NHP is well documented with regard to reliability and validity, and it is useful in describing the impact of chronic disease.6
110.5.3 MEDICAL OUTCOMES STUDY (MOS) 36-ITEM SHORT FORM (SF-36)*** The most widely used of the general health status questionnaires is SF-36.7 Ware and Sherbourne7 developed SF36 nearly 10 years after both the SIP and NHP were constructed. SF-36 was constructed to survey health status * For permission to use contact: Health Services Research & Development Center, John Hopkins School of Hygiene and Public Health, 624 North Broadway, Baltimore, MD 21205-1901. ** For permission to use contact: Dr. Stephen McKenna, Galen Research, Enterprise House, Manchester Science Park, Lloyd Street North, Manchester M15 6SU, U.K. *** For permission to use contact: Dr. John Ware, Medical Outcomes Trust, 20 Park Plaza, Suite 1014, Boston, MA 02116-4313.
General Guidelines for Assessment of Skin Diseases
in the Medical Outcomes Study. The SF-36 was also designed for use in clinical practice and research, health policy evaluations, and general population surveys. As the name implies, there are 36 questions, where most refer to the past 4 weeks. The SF-36 includes one multi-item scale that assesses eight health concepts: (1) limitations in physical activities because of health problems, (2) limitations in social activities because of physical or emotional problems, (3) limitations in usual role activities because of physical health problems, (4) bodily pain, (5) general mental health (psychological distress and well-being), (6) limitations in usual role activities because of emotional problems, (7) vitality (energy and fatigue), and (8) general health perceptions. These eight can be scored separately, but also summarized into two measures: (1) physical health and (2) mental health. SF-36 is used in several dermatological studies8–10 and is tested for validity, reliability, and sensitiveness/responsiveness.11–13
110.6 DERMATOLOGY QUALITY OF LIFE INSTRUMENTS Generic instruments are constructed to cover a wide range of conditions and diseases. Therefore, they often lack the sensitivity and responsiveness to detect differences that arise as a consequence of treatments compared in, e.g., clinical trials. As a consequence, disease-specific questionnaires have been developed.
110.7 GENERIC QUALITY OF LIFE INSTRUMENTS FOR SKIN DISORDERS
935
110.7.2 CHILDREN’S DERMATOLOGY LIFE INDEX (CDLQI)** CDLQI is designed for use in children, i.e., patients from age 5 to age 16. It was developed by Lewis-Jones and Finlay in 199316 and is based upon the DLQI. Only a few questions have been changed, so the construct is more relevant for children. It is widely used, self-explanatory, and self-administered. It is usually completed in 1 to 2 min, and like the DLQI, it is tested for validity, reliability, and sensitivity to changes and is translated into several languages.
110.7.3 SKINDEX*** The Skindex is constructed to measure the effect of skin disease on patients’ quality of life. Initially a 61-item selfadministered survey instrument was developed, the socalled Skindex. The original Skindex had eight scales, each of which addressed a construct, or an abstract component, in a comprehensive conceptual framework: cognitive effects, social effects, depression, fear, embarrassment, anger, physical discomfort, and physical limitations. Item responses were standardized from 0 (no effect) to 100 (maximal effect).17,18 The 61-item instrument was tested for validity and reliability, but the acceptability by the respondents was weak, because it was time-consuming. As a consequence, the Skindex-2919 and later the Skindex-1620 were developed. These new and shorter versions have improved discriminative and evaluative capability and have decreased respondent burden compared to Skindex-61. The Skindex is commonly used and translated into several languages.
110.7.1 DERMATOLOGY LIFE QUALITY INDEX (DLQI)*
110.8 DISEASE-SPECIFIC DERMATOLOGY QUALITY OF LIFE INSTRUMENTS
DLQI is a simple practical questionnaire technique for routine clinical use for any skin disease. It was developed by Drs. Finlay and Khan in 199214,15 and is today the most widely used general life quality questionnaire in dermatology. It covers different aspects of life impairment, e.g., symptoms, feelings, daily activities, work or school, personal relationships, and treatment. It consists of 10 questions, where each question has four alternative responses and scores from 0 to 3. The DLQI is calculated by summing the score of each question, resulting in a score range from 0 to 30. The higher the score, the greater the impairment of quality of life. The DLQI is easy and quick to use. It is tested according to validity, reliability, and sensitivity to changes and is translated into many different languages.
110.8.1 ATOPIC DERMATITIS****
* For permission to use contact: A.Y. Finlay, Head of Department of Dermatology, University of Wales College of Medicine, Heath Park, Cardiff, CF 14 4XN, Wales, U.K.
The Dermatitis Family Impact Questionnaire (DFI)21 and the Infants’ Dermatitis Quality of Life Index (IDQOL)22 are two simple questionnaires that are designed to assess the impact on quality of life of infants with atopic dermatitis and their families. The authors of CDLQI developed both questionnaires, and the two instruments therefore have the same basic construction and scoring system as ** For permission to use contact: Ms. Lewis-Jones or A.Y. Finlay, Head of Department of Dermatology, University of Wales College of Medicine, Heath Park, Cardiff, CF 14 4XN, Wales, U.K. *** For permission to use contact: M.M. Chren, R.J. Lasek, S.A. Flocke, and S.J. Zyzanski, Dermatology Service, Cleveland Veterans Affairs Medical Center, Cleveland, Ohio. **** For permission to use contact: Ms. Lewis-Jones or A.Y. Finlay, Head of Department of Dermatology, University of Wales College of Medicine, Heath Park, Cardiff, CF 14 4XN, Wales, U.K.
936
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 110.1 Disease-Specific Quality of Life Instruments for Dermatological Diseases Type of Disease
Name of Instrument
Reference
Acne Atopic dermatitis
The Acne-Specific Quality of Life Questionnaire (Acne-QoL) The Family Dermatitis Impact Questionnaire (FDI) The Infants’ Dermatitis Quality of Life Index (IDQOL) Hairdex Melasma Quality of Life Scale (MELASQOL) Psoriasis Disability Index Psoriasis Life Stress Inventory Psoriasis Index of Quality of Life (PSORIQoL) Scalpdex Hornheide questionnaire for psychosocial support
24 21 22 25 26 27 28 23 29 30
Hair disorder Melasma Psoriasis
Scalp, dermatitis Skin cancer
the CDLQI and DLQI. In addition to the 10 items, the IDQOL also has a global severity question. Both the DFI and the IDQOL are tested for validity, reliability, and sensitivity to changes and are translated into several languages.
110.8.2 PSORIASIS INDEX (PSORIQOL)
OF
QUALITY
OF
LIFE
The PSORIQoL is a new (2003) psoriasis-specific measure of QoL. It consists of 25 dichotomous items, making it short and practical for use in clinical studies. It differs from existing patient-reported outcome measures used in dermatology in the way that items do not directly assess impairment or disability, but rather the impact of these and other influences on the QoL of the patient.23 For other disease-specific QoL instruments, see Table 110.1.
110.9 DISEASE-SPECIFIC ASSESSMENT In lieu of objective measures in dermatology, many scoring systems have been developed. These methods generally describe a few clinical parameters: area involved, assessment of severity by scoring of three to six relevant symptoms on a scale from 0 = no involvement to 3 = severe, in the inflamed area. Some are designed for overall assessment, e.g., PASI and SCORAD, and some for target lesions, such as ADSI for AE. Measurements of extent have been claimed to be an “impossible task” for many dermatological diseases, in particular AE.31 Therefore, many different assessment methods such as the rule of hand, color coding,32 a system of tick boxes,33 computer-assisted body surface area34 image analysis system,35 and the rule of nines in, e.g., SCORAD,36 have been developed.
9%
Chest 18% 9%
Back 18%
9%
1% 18%
18%
FIGURE 110.2 Body surface area by Rules of Nines.
110.9.1 ATOPIC ECZEMA For atopic eczema at least 15 different disease-specific objective skin examination scales exist.37 In this section some of the most commonly used, like SCORAD and EASI, are presented, while other scoring instruments for AD are listed in Table 110.2. SCORAD (Scoring Atopic Dermatitis) was developed by the European Task Force on Atopic Dermatitis in 199336 and is the most extensively tested of all existing AE severity indices. It is an overall assessment and includes scoring of extent, intensity, and subjective symptoms (pruritus and sleep loss). The subjective symptoms can be left out in what is then termed an objective SCORAD. It consists of a combination of: 1. Assessment of extent (the rule of nines). Be aware of the differences in body area size
General Guidelines for Assessment of Skin Diseases
937
TABLE 110.2 Disease-Specific Instruments for Dermatological Diseases Type of Disease Acne Atopic dermatitis
Dry skin
Dyshidrotic eczema Hand dermatitis Inflammatory disease Lipodystrophy Mastocytosis Melasma Morphea Pemphigus Psoriasis
Rosacea Scleroderma
Name of Instrument
Reference
Leeds Acne Grading Scale SCORAD Eczema Area and Severity Index (EASI) Self-administered EASI (SA-EASI) Assessment Measure for Atopic Dermatitis (ADAM) Atopic Dermatitis Area and Severity Index (ADASI) Atopic Dermatits Severity Index (ADSI) Basic Clinical Scoring System (BCSS) Costa’s Simple Scoring System (Costa’s SSS) Leicester score Nottingham Eczema Severity Score Rajka and Langeland Six-Area, Six-Sign Atopic Dermatitis Severity Index (SASSAD) Skin Intensity Score (SIS) Six-Area, Total Body Severity Assessment (TBSA) Objective Severity Assessment of Atopic Dermatitis Score (OSAAD) Overall Dry Skin Score (ODS) Dry Skin Area and Severity Index (DASI) Specified Symptom Sum Score (SRRC) Dyshidrotic Eczema Area and Severity Index Work Productivity and Activity Impairment–Chronic Hand Dermatitis Questionnaire in chronic hand dermatitis Dermatology Index of Disease Severity (DIDS) Objective lipodystrophy severity grading scale Scoring Index of Mastocytosis (SCORMA) Melasma Area and Severity Index (MASI) Two methods to assess morphea: skin scoring and the use of a durometer Pemphigus Area and Activity Score (PAAS) A new grading system for oral pemphigus PASI Self-administered Psoriasis Area and Severity Index (SAPASI) National Psoriasis Foundation Psoriasis Score (NPF-PS) Rosacea staging Standard classification of rosacea European Scleroderma Study Group (EscSG) activity indices for systemic sclerosis Scleroderma visual analog scales U.K. Scleroderma Functional Score (UKFS) Arthritis Hand Function Test in adults with systemic sclerosis (scleroderma) Self-administered Systemic Sclerosis Questionnaire (SySQ)
42 36 38 39 43 32 44 45 46 47 33 48 49 50 51 34 52
between infants of <2 years and adults. Max score of 100. 2. Assessment of six clinical features of intensity: erythema/darkening, edema/papulation, oozing/crust, excoriation, lichenification/prurigo, and dryness. The first five are assessed on a single average and representative site and dryness is evaluated on uninvolved skin. Each feature is assessed on a scale of 0 to 3: 0 = absence, 1 = mild, 2 = moderate, and 3 = severe. Maximum score of 18.
53 54 55 56 57 26 58 59 60 40 41 61 62 63 64 65 66 67 68
3. Two visual analog scales (VASs), where the patient has to evaluate the average degree of pruritus and sleep loss for the last 3 days or nights. Items 1 and 2 can be measured alone and presented as objective SCORAD, giving a maximum sum score of 83, or if item 3 is added, then the maximum sum score is 103, which is the most widely used method and gives the SCORAD.
938
Handbook of Non-Invasive Methods and the Skin, Second Edition
SCORAD is quick and easy to use, but it has problems with interobserver variation in the assessment of both extent and intensity. Eczema Area and Severity Index (EASI) is designed to evaluate extent and severity, but also therapeutic response.38 It combines: 1. Assessment of disease extent on a scale from 0 to 6 in four body regions: Extent: 0 = no eruption; 1 < 10%, 2 = 10 to 29%, 3 = 30 to 49%, 4 = 50 to 69%, 5 = 70 to 89%, 6 = 90 – 100% Body proportions: head = 10%, trunk = 30%, upper extremities = 20%, lower extremities = 40% 2. Assessment of four clinical features of intensity: erythema, infiltration or papulation, excoriation, and lichenification. Each feature is assessed on a scale from 0 to 3: 0 = absence, 1 = mild, 2 = moderate, and 3 = severe. Half steps are allowed. A formula is then used to calculate the score for each body region; the scores are then added to give the total score. The EASI has a score range from 0 to 72. EASI has also been constructed in a patient selfadministered version, SA-EASI.39
110.9.2 PSORIASIS Psoriasis Area and Severity Index (PASI) was designed in 1978 to evaluate extent and severity, but also therapeutic response.40 It is the counterpart to and has the same construct as EASI in AE. It combines: 1. Assessment of disease extent on a scale from 0 to 6 in four body regions: Extent: 0 = no eruption, 1 < 10%, 2 = 10 to 29%, 3 = 30 to 49%, 4 = 50 to 69%, 5 = 70 to 89%, 6 = 90 to 100% Body proportions: head = 10%, trunk = 30%, upper extremities = 20%, lower extremities = 40% 2. Assessment of three clinical features of intensity: erythema, infiltration, and desquamation. Each feature is assessed on a scale from 0 to 4: 0 = absence, 1 = mild, 2 = moderate, 3 = severe, and 4 = very severe. A formula is then used to calculate the score for each body region; the scores are then added to give the total score. The PASI has a score range from 0 to 72. The PASI is validated and has the same problems with interobserver variation as other dermatological scoring systems that assess disease extent.
A self-administered PASI, the SAPASI, has also been developed.41 It assesses the extent of the same four body regions as PASI, but severity of erythema, induration, and scaling are assessed on an average psoriatic lesion with three modified visual analog scales (VASs). The SAPASI has been examined for validity, reliability, and responsiveness to change in severity over time.
110.9.3 ACNE For assessment of acne, the Leeds Acne Grading Scale is the most widely used scoring system.42 The Leeds Acne Grading Scale (Figue 110.3) is originally based on a large study of 435 patients with acne of varying severity. The technique requires examination of the patient in good light. Lesions are inspected visually and palpated to detect nodules or cysts. The acne severity can then be assigned a grade ranging from 0 to 10 (severe nodulocystic acne). In daily practice, the range is from 0.25 to 0.75 (physiological acne) to 7 (predominantly nodular cystic acne). The patient’s acne should be graded separately for the (1) face, including the neck, (2) chest, and (3) back, because at these sites acne frequently varies in severity and treatment responses. The grading system is only supplementary to a clinical description of the types of lesions present, e.g., predominantly comedones or mixed inflamed lesions, as this will have a bearing on the type of treatment chosen. Be aware that skin irritation caused by some topical therapies can make acne look worse than it is, and this must be considered when acne severity is assessed.
110.10 OTHER SKIN DISORDERS In dermatology there is a tendency that for every new randomized trial, a new assessment instrument is developed. Therefore, it is not possible to give a complete list of all the available clinical scoring systems for skin diseases in this book. Examples are given in Table 110.2 of different assessment instruments of various skin disorders.
110.11 CONCLUSION In this chapter general guidelines for scorings systems are given, and many examples of available assessment instruments for various skin diseases are listed. The time from the development of a clinical scoring system to its clinical use and its validation is a lengthy and ongoing process. If new clinical instruments continuously are being constructed, earlier instruments will never be fully validated, and we will miss important information about their usefulness. Most importantly, we will miss the chance to find a disease-specific scoring system that satisfies the basic requirements of validity, reliability, sensitivity, and responsiveness. Without a standard measure, we will not be able to compare study results in the
General Guidelines for Assessment of Skin Diseases
939
FIGURE 110.3 The Leeds Acne Grading Scale. (Reproduced from Retinoids Today and Tomorrow, 15, 1989. With permission.)
940
Handbook of Non-Invasive Methods and the Skin, Second Edition
future, and this will in the end slow down the progress in dermatology research. In other disease groups, e.g., rheumatoid arthritis, a standard measure has been accepted. Rheumatoid arthritis is similar to many dermatological diseases as being chronic, complex in both its symptoms and signs, and without a cure. If such standardization is possible in the field of rheumatology, it should also be possible in the area of dermatology.
REFERENCES 1. Altman D. Practical Statistics for Medical Research. 2004, pp. 1–611. 2. Fayers PM, Machin M. Quality of Life: Assessment, Analysis and Interpretation. 2004, pp. 3–404. 3. Jemec GB, Wulf HC. The applicability of clinical scoring systems: SCORAD and PASI in psoriasis and atopic dermatitis. Acta Derm Venereol 77(5):392–393, 1997. 4. Bergner M, Bobbitt RA, Pollard WE, Martin DP, Gilson BS. The sickness impact profile: validation of a health status measure. Med Care 14(1):57–67, 1976. 5. Hunt SM, McKenna SP, Williams J. Reliability of a population survey tool for measuring perceived health problems: a study of patients with osteoarthrosis. J Epidemiol Commun Health 35:297–300, 1981. 6. Wiklund I. The Nottingham Health Profile: a measure of health-related quality of life. Scand J Prim Health Care Suppl 1:15–18, 1990. 7. Ware JE, Jr., Sherbourne CD. The MOS 36-item shortform health survey (SF-36). I. Conceptual framework and item selection. Med Care 30(6):473–483, 1992. 8. Bingefors K, Lindberg M, Isacson D. Self-reported dermatological problems and use of prescribed topical drugs correlate with decreased quality of life: an epidemiological survey. Br J Dermatol 147(2):285–290, 2002. 9. Schiffner R, Brunnberg S, Hohenleutner U, Stolz W, Landthaler M. Willingness to pay and time trade-off: useful utility indicators for the assessment of quality of life and patient satisfaction in patients with port wine stains. Br J Dermatol 146(3):440–447, 2002. 10. Iliev D, Furrer L, Elsner P. [Assessment of the quality of life of patients in dermatology]. Hautarzt 49(6):453–456, 1998. 11. Brazier JE, Harper R, Jones NM, O’Cathain A, Thomas KJ, Usherwood T, et al. Validating the SF-36 health survey questionnaire: new outcome measure for primary care. BMJ 305(6846):160–164, 1992. 12. McHorney CA, Ware JE, Jr., Raczek AE. The MOS 36Item Short-Form Health Survey (SF-36). II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care 31(3):247–263, 1993.
13. McHorney CA, Ware JE, Jr., Lu JF, Sherbourne CD. The MOS 36-Item Short-Form Health Survey (SF-36). III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups. Med Care 32(1):40–66, 1994. 14. Finlay AY. Quality of life assessments in dermatology. Semin Cutan Med Surg 17(4):291–296, 1998. 15. Finlay AY, Khan GK. Dermatology Life Quality Index (DLQI): a simple practical measure for routine clinical use. Clin Exp Dermatol 19(3):210–216, 1994. 16. Lewis-Jones MS, Finlay AY. The Children’s Dermatology Life Quality Index (CDLQI): initial validation and practical use. Br J Dermatol 132(6):942–949, 1995. 17. Chren MM, Lasek RJ, Quinn LM, Mostow EN, Zyzanski SJ. Skindex, a quality-of-life measure for patients with skin disease: reliability, validity, and responsiveness. J Invest Dermatol 107(5):707–713, 1996. 18. Chren MM, Lasek RJ, Quinn LM, Covinsky KE. Convergent and discriminant validity of a generic and a disease-specific instrument to measure quality of life in patients with skin disease. J Invest Dermatol 108(1):103–107, 1997. 19. Chren MM, Lasek RJ, Flocke SA, Zyzanski SJ. Improved discriminative and evaluative capability of a refined version of Skindex, a quality-of-life instrument for patients with skin diseases. Arch Dermatol 133(11):1433–1440, 1997. 20. Chren MM, Lasek RJ, Sahay AP, Sands LP. Measurement properties of Skindex-16: a brief quality-of-life measure for patients with skin diseases. J Cutan Med Surg 5(2):105–110, 2001. 21. Lawson V, Lewis-Jones MS, Finlay AY, Reid P, Owens RG. The family impact of childhood atopic dermatitis: the Dermatitis Family Impact Questionnaire. Br J Dermatol 138(1):107–113, 1998. 22. Lewis-Jones MS, Finlay AY, Dykes PJ. The Infants’ Dermatitis Quality of Life Index. Br J Dermatol 144(1):104–110, 2001. 23. McKenna SP, Cook SA, Whalley D, Doward LC, Richards HL, Griffiths CE, et al. Development of the PSORIQoL, a psoriasis-specific measure of quality of life designed for use in clinical practice and trials. Br J Dermatol 149(2):323–331, 2003. 24. Fehnel SE, McLeod LD, Brandman J, Arbit DI, McLaughlin-Miley CJ, Coombs JH, et al. Responsiveness of the Acne-Specific Quality of Life Questionnaire (Acne-QoL) to treatment for acne vulgaris in placebocontrolled clinical trials. Qual Life Res 11(8):809–816, 2002. 25. Fischer TW, Schmidt S, Strauss B, Elsner P. [Hairdex: a tool for evaluation of disease-specific quality of life in patients with hair diseases]. Hautarzt 52(3):219–227, 2001. 26. Balkrishnan R, McMichael AJ, Camacho FT, Saltzberg F, Housman TS, Grummer S, et al. Development and validation of a health-related quality of life instrument for women with melasma. Br J Dermatol 149(3):572–577, 2003. 27. Finlay AY, Kelly SE. Psoriasis: an index of disability. Clin Exp Dermatol 12(1):8–11, 1987.
General Guidelines for Assessment of Skin Diseases
28. Gupta MA, Gupta AK. The Psoriasis Life Stress Inventory: a preliminary index of psoriasis-related stress. Acta Derm Venereol 75(3):240–243, 1995. 29. Chen SC, Yeung J, Chren MM. Scalpdex: a quality-oflife instrument for scalp dermatitis. Arch Dermatol 138(6):803–807, 2002. 30. Rumpold G, Augustin M, Zschocke I, Strittmatter G, Sollner W. [The validity of the Hornheide questionnaire for psychosocial support in skin tumor patients: a survey in an Austrian and German outpatient population with melanoma]. Psychother Psychosom Med Psychol 51(1):25–33, 2001. 31. Charman CR, Venn AJ, Williams HC. Measurement of body surface area involvement in atopic eczema: an impossible task? Br J Dermatol 140(1):109–111, 1999. 32. Bahmer FA. ADASI score: atopic dermatitis area and severity index. Acta Derm Venereol Suppl (Stockh) 176:32–33, 1992. 33. Emerson RM, Charman CR, Williams HC. The Nottingham Eczema Severity Score: preliminary refinement of the Rajka and Langeland grading. Br J Dermatol 142(2):288–297, 2000. 34. Sugarman JL, Fluhr JW, Fowler AJ, Bruckner T, Diepgen TL, Williams ML. The objective severity assessment of atopic dermatitis score: an objective measure using permeability barrier function and stratum corneum hydration with computer-assisted estimates for extent of disease. Arch Dermatol 139(11):1417–1422, 2003. 35. Yune YM, Park SY, Oh HS, Kim DJ, Yoo DS, Kim IH, et al. Objective assessment of involved surface area in patients with psoriasis. Skin Res Technol 9(4):339–342, 2003. 36. Severity scoring of atopic dermatitis: the SCORAD index. Consensus Report of the European Task Force on Atopic Dermatitis. Dermatology 186(1):23–31, 1993. 37. Charman C, Williams H. Outcome measures of disease severity in atopic eczema. Arch Dermatol 136(6):763–769, 2000. 38. Hanifin JM, Thurston M, Omoto M, Cherill R, Tofte SJ, Graeber M. The Eczema Area and Severity Index (EASI): assessment of reliability in atopic dermatitis. EASI Evaluator Group. Exp Dermatol 10(1):11–18, 2001. 39. Housman TS, Patel MJ, Camacho F, Feldman SR, Fleischer AB, Jr., Balkrishnan R. Use of the self-administered Eczema Area and Severity Index by parent caregivers: results of a validation study. Br J Dermatol 147(6):1192–1198, 2002. 40. Freddriksson T, Pettersson U. Severe psoriasis: oral therapy with a new retinoid. Dermatologica 157:238–244, 1978. 41. Sampogna F, Sera F, Mazzotti E, Pasquini P, Picardi A, Abeni D. Performance of the self-administered Psoriasis Area and Severity Index in evaluating clinical and sociodemographic subgroups of patients with psoriasis. Arch Dermatol 139(3):353–358, 2003. 42. Burke BM, Cunliffe WJ. The assessment of acne vulgaris: the Leeds technique. Br J Dermatol 111(1):83–92, 1984.
941
43. Charman D, Varigos G, Horne DJ, Oberklaid F. The development of a practical and reliable assessment measure for atopic dermatitis (ADAM). J Outcome Meas 3(1):21–34, 1999. 44. Van Leent EJ, Graber M, Thurston M, Wagenaar A, Spuls PI, Bos JD. Effectiveness of the ascomycin macrolactam SDZ ASM 981 in the topical treatment of atopic dermatitis. Arch Dermatol 134(7):805–809, 1998. 45. Sprikkelman AB, Tupker RA, Burgerhof H, Schouten JP, Brand PL, Heymans HS, et al. Severity scoring of atopic dermatitis: a comparison of three scoring systems. Allergy 52(9):944–949, 1997. 46. Costa C, Rilliet A, Nicolet M, Saurat JH. Scoring atopic dermatitis: the simpler the better? Acta Derm Venereol 69(1):41–45, 1989. 47. Berth-Jones J, Thompson J, Graham-Brown RA. Evening primrose oil and atopic eczema. Lancet 345(8948):520, 1995. 48. Rajka G, Langeland T. Grading of the severity of atopic dermatitis. Acta Derm Venereol Suppl (Stockh) 144:13–14, 1989. 49. Berth-Jones J. Six area, six sign atopic dermatitis (SASSAD) severity score: a simple system for monitoring disease activity in atopic dermatitis. Br J Dermatol 135 (Suppl. 48):25–30, 1996. 50. Wuthrich B, Kagi MK, Joller-Jemelka H. Soluble CD14 but not interleukin-6 is a new marker for clinical activity in atopic dermatitis. Arch Dermatol Res 284(6):339–342, 1992. 51. van Joost T, Heule F, Korstanje M, van den Broek MJ, Stenveld HJ, van Vloten WA. Cyclosporin in atopic dermatitis: a multicentre placebo-controlled study. Br J Dermatol 130(5):634–640, 1994. 52. Serup J. EEMCO guidance for the assessment of dry skin (xerosis) and ichthyosis: clinical scoring systems. Skin Res Technol 1:109–114, 1995. 53. Vocks E, Plotz SG, Ring J. The dyshidrotic eczema area and severity in. Dermatology 198(3):265–269, 1999. 54. Reilly MC, Lavin PT, Kahler KH, Pariser DM. Validation of the Dermatology Life Quality Index and the Work Productivity and Activity Impairment-Chronic Hand Dermatitis Questionnaire in chronic hand dermatitis. J Am Acad Dermatol 48(1):128–130, 2003. 55. Faust HB, Gonin R, Chuang TY, Lewis CW, Melfi CA, Farmer ER. Reliability testing of the Dermatology Index of Disease Severity (DIDS). An index for staging the severity of cutaneous inflammatory disease. Arch Dermatol 133(11):1443–1448, 1997. 56. Carr A, Law M. An objective lipodystrophy severity grading scale derived from the lipodystrophy case definition score. J Acquir Immune Defic Syndr 33(5):571–576, 2003. 57. Heide R, Middelkamp Hup MA, Mulder PG, Oranje AP. Clinical scoring of cutaneous mastocytosis. Acta Derm Venereol 81(4):273–276, 2001. 58. Seyger MM, van den Hoogen FH, de Boo T, de Jong EM. Reliability of two methods to assess morphea: skin scoring and the use of a durometer. J Am Acad Dermatol 37(5 Pt. 1):793–796, 1997.
942
Handbook of Non-Invasive Methods and the Skin, Second Edition
59. Agarwal M, Walia R, Kochhar AM, Chander R. Pemphigus Area and Activity Score (PAAS): a novel clinical scoring method for monitoring of pemphigus vulgaris patients. Int J Dermatol 37(2):158–160, 1998. 60. Saraswat A, Bhushan K, India C. A new grading system for oral pemphigus. Int J Dermatol 42(5):413–414, 2003. 61. Gottlieb AB, Chaudhari U, Baker DG, Perate M, Dooley LT. The National Psoriasis Foundation Psoriasis Score (NPF-PS) system versus the Psoriasis Area Severity Index (PASI) and Physician’s Global Assessment (PGA): a comparison. J Drugs Dermatol 2(3):260–266, 2003. 62. Plewig G, Kligman AM. Acne and Rosacea, 3rd ed., Springer-Verlag, Berlin, 2004. 63. Wilkin J, Dahl M, Detmar M, Drake L, Feinstein A, Odom R, et al. Standard classification of rosacea: report of the National Rosacea Society Expert Committee on the Classification and Staging of Rosacea. J Am Acad Dermatol 46(4):584–587, 2002.
64. Valentini G, Bencivelli W, Bombardieri S, D’Angelo S, Della RA, Silman AJ, et al. European Scleroderma Study Group to define disease activity criteria for systemic sclerosis. III. Assessment of the construct validity of the preliminary activity criteria. Ann Rheum Dis 62(9):901–903, 2003. 65. Steen VD, Medsger TA, Jr. The value of the Health Assessment Questionnaire and special patient-generated scales to demonstrate change in systemic sclerosis patients over time. Arthr Rheum 40(11):1984–1991, 1997. 66. Smyth AE, MacGregor AJ, Mukerjee D, Brough GM, Black CM, Denton CP. A cross-sectional comparison of three self-reported functional indices in scleroderma. Rheumatology (Oxford) 42(6):732–738, 2003. 67. Poole JL, Gallegos M, O’Linc S. Reliability and validity of the arthritis hand function test in adults with systemic sclerosis (scleroderma). Arthr Care Res 13(2):69–73, 2000. 68. Ruof J, Bruhlmann P, Michel BA, Stucki G. Development and validation of a self-administered systemic sclerosis questionnaire (SySQ). Rheumatology (Oxford) 38(6):535–542, 1999.
Lauryl Sulfate (SLS) Testing: 111 Sodium ESCD Application and Reading Standards R.A. Tupker Department of Dermatology, St. Antonius Hospital, Nieuwegein, The Netherlands
CONTENTS 111.1 Introduction ..........................................................................................................................................................944 111.2 Effects of SLS ......................................................................................................................................................944 111.2.1 Characteristics of SLS ...........................................................................................................................944 111.2.2 Histopathological and Immunological Effects ......................................................................................944 111.2.3 Clinical Effects.......................................................................................................................................944 111.3 SLS Exposure Methods........................................................................................................................................945 111.3.1 One-Time Occlusive Tests (Closed Patch Tests)...................................................................................945 111.3.1.1 Type of Test Chamber ..........................................................................................................945 111.3.1.2 Quantity of Test Solution .....................................................................................................945 111.3.1.3 Evaporation and Temperature of the Solution .....................................................................945 111.3.1.4 Concentration of Test Solution and Duration of Exposure .................................................946 111.3.1.5 Time of Evaluation ...............................................................................................................946 111.3.2 Repeated Occlusive Tests ......................................................................................................................946 111.3.3 Repeated Open Tests..............................................................................................................................946 111.3.4 Immersion Tests .....................................................................................................................................946 111.3.5 Wash Tests..............................................................................................................................................947 111.3.6 Correlation between Occlusive and Open Tests....................................................................................947 111.4 Factors Influencing SLS Reactivity .....................................................................................................................947 111.4.1 Constitutional Factors ............................................................................................................................947 111.4.1.1 Age........................................................................................................................................947 111.4.1.2 Race ......................................................................................................................................948 111.4.1.3 Sex ........................................................................................................................................948 111.4.1.4 Anatomical Site ....................................................................................................................948 111.4.1.5 Baseline TEWL and Predictive Testing ...............................................................................948 111.4.1.6 Genetic Factors .....................................................................................................................948 111.4.1.7 Prior Exposure to Irritants....................................................................................................948 111.4.2 SLS Testing in Abnormal Skin..............................................................................................................949 111.4.2.1 Dryness/Ichthyosis................................................................................................................949 111.4.2.2 Dermatitis .............................................................................................................................949 111.4.3 Environment-Related Variables .............................................................................................................949 111.5 Recommendations ................................................................................................................................................949 111.5.1 Test Individuals, Reactivity, and Location ............................................................................................949 111.5.2 Test Substance and Applications ...........................................................................................................950 111.5.3 Specific Test Procedures ........................................................................................................................950 111.5.4 Individual Pilot Study ............................................................................................................................951 111.5.5 Interpretation of SLS Exposure/Results from Noninvasive Evaluation Methods ................................952 References .......................................................................................................................................................................953 943
944
Handbook of Non-Invasive Methods and the Skin, Second Edition
111.1 INTRODUCTION Sodium lauryl sulfate (SLS) has been used extensively as a model irritant for skin irritancy testing. Irritancy testing has been performed using various exposure methods, and evaluated by several (clinical) scoring systems. It is not surprising, therefore, that there is nowadays a multitude of SLS testing methods, which hampers the interpretation of results from different laboratories. This chapter is based on the guidelines on SLS exposure tests, a report from the standardization group of the European Society of Contact Dermatitis (ESCD).1 A short description of the characteristics of SLS and its histopathological and clinical effects is followed by a review of the various exposure methods, with the ultimate goal of introducing a uniform approach to SLS testing in humans. Since different study aims warrant different testing conditions, we propose several testing methods tailored to the purpose of study. With respect to aims, studies are divided into two categories. The first category, provocative testing, concerns studies in which SLS is used to induce a definite skin reaction in all individuals. Examples of aims of the first category are (1) elucidate the mechanisms of skin irritation, (2) predict the irritant potency of different detergents, (3) study the time course after irritation, (4) compare the sensitivity of different noninvasive methods, and (5) compare the efficacy of different moisturizers in preventing or healing skin irritation. The second category, susceptibility evaluation, concerns studies aimed to predict the irritant susceptibility of individuals, and investigate individual and environmental factors determining this susceptibility. Initially, evaluation of irritancy testing was based on visual scoring only. Examples are classic studies, such as those by Bettley2 and Frosch and Kligman.3 Later, the irritant reaction was evaluated by various noninvasive techniques, such as transepidermal water loss (TEWL) measurement, laser Doppler flowmetry, colorimetry, ultrasound measurement, and electrical capacitancy measurement. For a description of these techniques, refer to the appropriate sections in this book.
111.2 EFFECTS OF SLS 111.2.1 CHARACTERISTICS
OF
SLS
SLS (synonym: sodium dodecyl sulfate) is an anionic detergent with the molecular weight of 288.38 g/mol. The molecule is a 12-carbon chain with the composition CH3–(CH2)10–CH2–O–SO2–O–Na. Pure SLS is white and the crystals, flakes, or powder are soluble in water (1 g/10 ml = 10%) and somewhat soluble in ethanol. In aqueous solution SLS reduces the surface tension and forms oilin-water emulsions. The critical micelle concentration for
SLS in aqueous solution is 8.2 × 10–3 mol/l (0.23% w/v) at 25oC. Irritancy patch test studies with SLS generally have been performed using aqueous solutions, although petrolatum was also used as a vehicle.4 SLS is available in different purities with a variable content of carbon chain lengths other than 12. C12 is known to elicit a maximum of irritant response clinically.5 Analysis of different grades of SLS by high-performance liquid chromatography revealed that in some grades of SLS part of the highly irritating C12 substance had been substituted by longer and less irritating carbon chains.6 Thus, high-purity SLS (99%) must be used in any study, preferably in water solution.
111.2.2 HISTOPATHOLOGICAL AND IMMUNOLOGICAL EFFECTS Ultrastructurally, SLS at low concentration does not alter the existing lipid structure of the stratum corneum, but rather influences the synthesis of new lipids.7 These changes might explain the protracted TEWL enhancement following SLS. Single application of SLS, with sampling between 24 and 72 h, induces spongiosis, intracellular vacuolation, and lipid accumulation in those reactions that are mild to moderate, and necrosis in those that are more severe.8 Parakeratosis is also a common feature, most probably arising from the stimulatory effects of SLS on keratinocytes. SLS reactions are characterized by the influx of leukocytes into the epidermis and dermis, accompanied in some reactions by edema and collagen disruption/degeneration. T lymphocytes are the predominant infiltrating cells, with CD4+ cells almost always outnumbering CD8+ cells.9 Following exposure to SLS, keratinocytes release proinflammatory cytokines, such as interleukine-1 (IL-1) and tumor necrosis factor-α (TNF-α), induce expression of adhesion molecules, and upregulate MHC class I and II and integrin receptors.10,11 Furthermore, antioxidative enzymes and heat shock proteins are formed.10
111.2.3 CLINICAL EFFECTS Björnberg in his classical thesis on primary irritancy described the morphology of SLS reactions.12 Erythema, sometimes associated with infiltration and sometimes with superficial erosion of the epithelium, is the main feature of acute reactions (Figure 111.1). With higher concentrations, vesicular and even pustular reactions may be seen. During healing of acute reactions, scaling and fissuring will take over. The same appearance with erythema, scaling, and fissuring is seen during repeated application of SLS (Figure 111.2).13,14 The so-called soap effect or effet de savon consists of a fine wrinkled surface contour (Figure 111.3) associated or followed by chapping, the latter
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
945
representing a characteristic roughening of the skin. This characteristic soap effect may be observed when SLS at low concentration is used.
111.3 SLS EXPOSURE METHODS 111.3.1 ONE-TIME OCCLUSIVE TESTS (CLOSED PATCH TESTS)
FIGURE 111.1 Acute reaction with moderate erythema. According to Table 111.2A (simple scoring system), score 2. According to Table 111.2B (elaborate scoring system), erythema score 2, roughness 0, scaling 0, edema 0, fissures 0.
For many years, closed patch testing with SLS has been the favored test method, both for practical and traditional reasons. It is undoubtedly easy and rapid to perform. The disadvantage of this one-time occlusive test is the fact that it mimics only an acute irritant reaction. The more common situation in real life is the development of chronic irritant contact dermatitis, resulting from multiple repeated exposures.15 Regrettably, numerous studies have used a multitude of exposure techniques. Variables include type of test chamber, quantity of test solution, concentration of detergent solution, evaporation of the solution, temperature of the solution, duration of exposure, and time of evaluation. 111.3.1.1 Type of Test Chamber
FIGURE 111.2 Cumulative reaction with erythema, scaling, and fissuring. According to Table 111.3 (simple scoring system), score 2. According to Table 111.2B (elaborate scoring system), erythema score 2, roughness 0, scaling 2, edema 1, fissures 1.
An important variable is the internal diameter of the chamber. The most frequently used chamber is the Finn device, constructed by Pirilä.16 Originally this chamber had an inner diameter of 8 mm, with a small capacity (20 μl), which was found to be too small to yield positive reactions to certain irritants.17 Nowadays, Finn chambers with 12and 18-mm inner diameters, having capacities of 75 and 300 μl, respectively, are available. Frosch and Kligman introduced the Duhring chamber and found that this chamber had better test properties than the small Finn chamber.17 This was corroborated in a more recent study, showing that the 18-mm Finn chamber yielded the strongest reactions, followed by the 12-mm chamber, in its turn followed by the 8-mm chamber.18 111.3.1.2 Quantity of Test Solution In closed patch testing with SLS, as with other irritants, the quantity of test solution per square millimeter of skin is of importance for the skin response.17 Even when applied at the same concentration, larger quantities of SLS solution tend to induce more intense skin reactions. Furthermore, small test areas may be too limited to elicit an irritant skin reaction.17
FIGURE 111.3 Cumulative reaction with soap effect. According to Table 111.3 (simple scoring system), score 1. According to Table 111.2B (elaborate scoring system), erythema score 0, roughness 0, scaling 1, edema 0, fissures 0.
111.3.1.3 Evaporation and Temperature of the Solution Evaporation from the patch test before application on the skin may inhibit the inflammatory response, even though the relative concentration of test substance per square
946
Handbook of Non-Invasive Methods and the Skin, Second Edition
millimeter is the same.19 Also, the temperature of the test substance may be of significant importance. Berardesca et al. have reported significantly different skin responses to 4, 20, and 40˚C SLS test solution.20 111.3.1.4 Concentration of Test Solution and Duration of Exposure Several studies have demonstrated a dose dependence for SLS reactions in one-time occlusive tests.21,22 For closed testing with SLS, application times of 24 and 48 h are commonly used. In a study comparing these two periods, Staberg and Serup4 have found that for various SLS concentrations, higher mean blood flow values are present after 48 h than after 24 h application, as evaluated by laser Doppler flowmetry. Recently, the interrelationship between the concentration of SLS and exposure time (3, 6, 12, 24, and 48 h) was determined.23 There was a clear distinction between the concentrations used only at 12-, 24-, and 48-h exposures. The factor of concentration had a greater impact on the outcome of the irritant response than the factor of exposure duration. To achieve the same irritant response as found with a doubled SLS dose, the application time must be tripled. Recently, a simple 4-h patch test was designed to examine crude differences in the effects of SLS on the skin, evaluated by a simple visual scoring system, using extremely high SLS concentrations (20 and 100%).24 Hannuksela and Hannuksela, however, did not notice responses after a 4-h patch test, whereas the same solution caused proper reactions after 24 and 48 h.18 111.3.1.5 Time of Evaluation For closed patch testing, the time between removal and reading of the patch should be carefully considered. The time course for TEWL after occlusive patch testing with SLS demonstrated a significant decrease at 0 to 30 min and 30 to 60 min after removal of the patch, whereas the values at 60 to 180 min were stable.25 In the same study, the time course for TEWL after patch testing with an empty chamber was studied. A statistically significantly increased TEWL, as compared to normal skin, was reported 30 min after removal of the empty patch, but not after 60 min. In contrast, other investigators found that longer times between removal and reading were necessary.17,26 The TEWL value of a 24-h water patch had returned to a level slightly elevated above its own baseline value only after 2 h in most individuals.26 The reactions during the days after removal of the SLS patch were studied by several investigators. Different patterns in individual time courses could be recognized, dividing the group into subjects with early peak values (day of removal, day 2), late peaks (day 3), or very late
peaks (day 4).21,27–29 The time necessary to restore barrier function exceeds 10 days.14,28
111.3.2 REPEATED OCCLUSIVE TESTS The concept of repeated occlusive testing is not new. Kligman and Wooding5 developed a test method in which substances are brought into contact with the skin under occlusion 24 h a day for 20 consecutive days.5 This method was designed for substances with a very low expected irritant potency. Later, Frosch and Kligman3 used an exposure model in which a substance is applied repeatedly during five consecutive days: the first day for 24 h and the following days for 6 h a day.3 In order to reflect better the conditions in daily practice, models were developed in which detergents were applied in a multiple repeated short-time way, namely, 2 times daily for 45 min during the working days of 1 week28,30–32 or 3 consecutive weeks.13,33 Using this test method, it was possible to rank detergents according to irritant potency, SLS being by far the most irritant detergent.13 The advantages of this multiple repeated short-time occlusive test method are that it offers a reasonable simulation of the daily practice, and has a relatively high reproducibility. However, being occlusive rather than open, it does not entirely mimic the situation in most reallife conditions.
111.3.3 REPEATED OPEN TESTS Solutions of 2, 5, and 7.5% SLS were pipetted onto areas on the back and allowed to air-dry in the studies by Lammintausta et al.34,35 Applications were done once daily for 6 days total. Concentration-dependent TEWL increases were observed, accompanied with no34 or only minor35 visible reactions. Other repeated open test models were based on once daily applications on the forearm skin using a glass ring during 20 to 45 min for 1 to 3 weeks.36–40 Increasing dose-dependent values for TEWL and erythema36–38 and clear differences in the irritancy of various detergents39,40 have been demonstrated by this method. By means of a repetitive open irritation test it was possible to test the efficacy of different barrier creams.38 The repeated open test can offer a good simulation of conditions in daily practice, as it mimics repetitive and cumulative toxic insults, while avoiding occlusive influences. The disadvantage is the fact that it is time-consuming for the subjects.
111.3.4 IMMERSION TESTS Twice daily 30-min forearm immersions of various detergents of 0.1% concentration were shown to induce different degrees of erythema and scaling, with SLS proving to be the most irritant.41 SLS at 0.5% appeared to elicit
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
erythema in most subjects after three immersions of 10 min duration.42 The skin response was markedly increased in arms immersed in solutions of higher temperature (40˚C), as compared with low temperature (20˚C).43 When the immersion procedure was applied to the hands only, no distinctions could be found visually between two different detergent solutions.44 Instead, the dominant hand had clinically more severe reactions, irrespective of the type of solutions applied. The dominant hands had a lower baseline hydration value, as measured with the Skicon® device. This was explained by a more intense exposure of the dominant hand before the experiment. The advantage of the immersion model is that it reproduces the cumulative insults by detergents in a more realistic way, but it is cumbersome compared to patch testing. Another disadvantage is the fact that only two agents can be compared in the same volunteer.
111.3.5 WASH TESTS The wash test (or use test) represents another open test model, in which the act of washing is mimicked in a repetitive way.45,46 The washing was performed in the elbow fold two times a day for 1 min.46 Repetitive washing with grit-containing cleansers four times daily for 2 min for 1 week on the forearm led to differences in TEWL values, skin redness, and hydration.47 In another wash test, minor erythema responses were observed, as measured with a Minolta Chroma Meter®, after 7 days of two times daily 1-min washings of the upper arm with 3 ml of a 10% dishwashing liquid, moving the hand up and down one time per second.18 TEWL values were moderately elevated compared to the baseline values. Gehring et al.48 let the volunteers wash their upper and lower arms with three types of solutions, five times daily during 1 min for 12 consecutive days.48 Fifty percent of the atopic panelists had to discontinue washing with 0.1 molar (= 4%) SLS solution halfway through the study procedure because of severe inflammatory reactions. In the nonatopic group, most subjects did not react at all, and others showed signs of hardening, as analyzed with various noninvasive methods. In the atopic subgroup, exposure of a soap-free washing emulsion and water did not cause hardening. The washing test mimics perfectly the act of washing, and thus offers the most realistic view for this real-life event. Moreover, it is the only way in which the effects of grits can be tested. However, this test procedure has the lowest standardization potential, because of the multitude of factors influencing the test outcome.
111.3.6 CORRELATION OPEN TESTS
BETWEEN
OCCLUSIVE
AND
A few studies have compared the patch test with the wash test with respect to the irritancy potential ranking order
947
of a series of detergents.18,45,46 In one study,45 discrepancies were found in the ranking order obtained by the two techniques. Another study found similar ranking orders; however, the wash test appeared to have a greater power of discrimination.46 Hannuksela and Hannuksela did not notice any correlation between the wash test, repeated open test, and patch test with respect to degree of responses.18 Studying the irritant potential of a limited number of detergents, Smeenk41 showed that there was a relatively good correlation between the results from the one-time occlusive patch test and those from the immersion test.41 The exception was soap, which was irritant only in the patch test, but not in the immersion test.41 This discrepancy may be explained by the earlier observation that the pH of a soap solution decreased after contact with human skin,49 and that soap at low pH (7.5) induced erythema and pruritus, which did not occur with soap solution at higher pH (9.5).49 More recently, the same finding has been observed in our study investigating four different detergents using one-time occlusive, repeated occlusive, and repeated open tests.40 Only in the one-time occlusive test did soap induce more erythema than the other detergents, whereas in the other models SLS and sodium cocoyl isethionate had higher visual scores than soap. In another study on the effects of occlusion, it was found that postexposure occlusion by a plastic wrap caused more severe irritation than unoccluded exposures to SLS.37 The degree of augmentation of the irritant response by occlusion differs between varying types of substances. The irritant effect of polar compounds is less influenced by occlusion than that of nonpolar compounds.50 Hence, it appears that the type of exposure may influence the outcome of the ranking order in irritancy testing. The central question is: Which exposure method offers the best prediction of real-life exposure? This poses another question: What are the real-life conditions encountered? In most in-use situations the uncovered skin is exposed to irritants several times daily. In other situations, alreadyexposed skin is covered by protective gloves or other impermeable materials. In the first situation mentioned the open tests seem the best way to simulate this. In the second situation, however, the one-time or repeated occlusive test may be preferred.
111.4 FACTORS INFLUENCING SLS REACTIVITY 111.4.1 CONSTITUTIONAL FACTORS 111.4.1.1 Age More pronounced irritative responses were noted in young subjects than in elderly individuals.51 However, once increased, the raised TEWL levels are more persistent in
948
Handbook of Non-Invasive Methods and the Skin, Second Edition
elderly subjects, implying that healing capacity decreases with age.52 111.4.1.2 Race In one study, TEWL responses after SLS exposure tended to be higher in blacks than in caucasians.53 Among Asians, the Chinese showed significantly higher values than the Malays.54 Fair skin within the Caucasian population has been shown to correlate with increased SLS sensitivity.35,55 111.4.1.3 Sex Experimental studies do not find a sex-related difference in skin susceptibility to irritants.30,34,55 Variation in skin susceptibility during the menstrual cycle has been demonstrated.56 In this study of 29 females with regular menstrual cycles, an increased skin reactivity at day 1, as compared to day 10, was confirmed by measurement of TEWL and edema formation (skin thickness).
to SLS, a correlation between baseline TEWL and postexposure TEWL was found only by Wilhelm et al.,36 and not by Lammintausta et al.35 Field studies on workers in high-risk occupations have shown that prework barrier function is not a valid predictor of the risk of hand dermatitis,62-64 in contrast to preexposure barrier function in experimental irritancy models. In these laboratory models, irritants are exposed for a relatively short period on the skin. In daily practice, however, the result of repeated exposures to damaging influences of various kinds is a complex interrelationship of different (partly unknown) factors, such as severity of chemical and mechanical insults, recovery time between exposures, seasonal influences, barrier function, history of atopic dermatitis, ability to develop adaptation, and other factors that are probably under genetic control. 111.4.1.6 Genetic Factors
Regional variations in baseline TEWL exist, with higher levels on the palms and forehead than on the arm and back.57 However, there are also significant differences within the same area of the body. On the forearm, for instance, significantly higher TEWL values were found next to the wrist than at the other sites of the forearm.58 It has been suggested, therefore, that this area should be excluded when testing.58 There is also a site variation in susceptibility to SLS on the forearm, as evaluated by TEWL and visual scoring.58 Postexposure TEWL next to the elbow was significantly higher than the other sites on the forearms, the area next to the wrist having the lowest value.58 Surprisingly, the baseline TEWL values had an inverse pattern.58 This points to the fact that susceptibility to SLS is not solely dependent on barrier function of the stratum corneum. Other factors, such as reactivity of the layers beneath the stratum corneum, are also important.
Recently the influence of a genetic marker on irritant susceptibility was investigated.65 Visual irritant thresholds were determined using graded concentrations of SLS and benzalkonium chloride in a large group of nonatopics. Transition polymorphism has been identified in the TNF-α gene. Individuals carrying a haplotype that includes the A allele are high secretors of TNF-α, which is a key mediator in the pathogenesis of irritant contact dermatitis. In the low irritant threshold groups of both SLS and benzalkonium chloride, a significantly increased number of persons having the A allele has been found.65 This study offers the first description of a non-barrier-related marker of susceptibility in nonatopics. The strength of response to irritants is not only influenced by the levels of TNF-α and IL-1, but also by the individual’s ability to quench free radicals, levels of antioxidant enzymes, and the ability to form protective heat shock proteins.10 It has been hypothesized that the above-mentioned mechanisms may all be under genetic control, which thus determine the variability in responsiveness to irritants.10
111.4.1.5 Baseline TEWL and Predictive Testing
111.4.1.7 Prior Exposure to Irritants
In 1986 Murahata et al. found that subjects with increased basal TEWL reacted more severely to detergent irritation, as evaluated visually, indicating an association between high basal TEWL and sensitive skin.59 Subsequently, our group demonstrated that the preexposure baseline TEWL value was strongly correlated with the TEWL value after 1 week of repeated exposures,26,30 and only moderately correlated with TEWL after 3 weeks of exposures.33 A positive correlation between basal TEWL and TEWL after skin exposure to SLS was also found by Agner.55,60 In other studies using one-time occlusive tests, these results could not be confirmed; no correlation 27,61 or low correlation53 was found. Using repeated open exposures
Hyporeactive skin may be induced by preceding long-time irritant exposure, as was demonstrated by Lammintausta et al.66 Hyporeactivity was also noted by Widmer et al. 6 and 9 weeks after short-time repeated 3-week occlusive exposures, but not after 3 weeks.67 The same phenomenon has been identified by our group, in multiple repeated exposures on humans.13 Clinical changes were accompanied or preceded by a downgrade curve in the TEWL time course, after an initial TEWL increase. Similar downgrade TEWL time courses were observed on skin sites showing no or only mild clinical signs.13,48 The hyporeactivity found in these studies signifies a functional adaptation, or hardening, of the skin. The
111.4.1.4 Anatomical Site
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
specificity of hardening is not yet exactly known. If the specificity of hardening is high, one must be cautious of this phenomenon in predictive irritancy tests comparing different types of chemicals.
111.4.2 SLS TESTING
IN
ABNORMAL SKIN
111.4.2.1 Dryness/Ichthyosis Skin dryness is a symptom with a heterogenous background. In atopic dermatitis patients, the barrier function (baseline TEWL) was found to correlate well with dryness,68 whereas in individuals with ordinary winter xerosis dryness was poorly correlated with barrier function.69 In subjects with dry skin (of which the majority had atopic dermatitis), a higher susceptibility to SLS was observed than in subjects having a nondry skin.33 In contrast, Lammintausta et al. did not observe a difference in susceptibility between normal skin and skin that looked and felt rough, in a group of atopics and nonatopics.35 111.4.2.2 Dermatitis A higher TEWL was noted in the involved skin in various types of dermatitis than in uninvolved sites.70,71 Shahidullah et al. have observed an increased TEWL in the involved and uninvolved skin in several types of dermatitis, TEWL of uninvolved skin being related to the severity of the dermatitis.70 In a more recent study, it was noted that susceptibility to SLS was related to the severity of manifest atopic dermatitis.31 In other studies on several types of dermatitis, it was observed that existing dermatitis anywhere on the body may enhance susceptibility to irritants on another body location.12,60,72 This has been termed the excited skin syndrome.73 Since the type of dermatitis present did not appear to influence the reactivity, the enhanced susceptibility noted was regarded as secondary to dermatitis activity per se.72 However, van der Valk et al. found a significantly elevated TEWL before and after SLS exposure only in the subgroup of patients with manifest atopic dermatitis.74 In patients with manifest irritant contact dermatitis there was no difference in the postexposure TEWL values compared with control subjects.74 In a study comparing patients with a history of atopic dermatitis (but without manifest dermatitis) and patients with a history of contact dermatitis, higher preexposure and postexposure TEWL values could be noted only in patients with a history of atopic dermatitis.33 In another study, patients with active and those with inactive atopic dermatitis reacted more strongly to SLS than nonatopic controls.22 Also, in the open test model, subjects with a history of atopic dermatitis were proven to be more susceptible.35 Therefore, atopic dermatitis skin may be considered unique with respect to irritant susceptibility.
949
111.4.3 ENVIRONMENT-RELATED VARIABLES Seasonal factors have a major influence on skin reactivity.3 A significantly higher skin susceptibility to SLS was found in winter than in summer in healthy volunteers.32,75 Differences between winter and summer were even more pronounced in patients with atopic dermatitis.32 A significantly lower preexposure barrier function and hydration state of the stratum corneum was reported in winter in healthy volunteers and atopic dermatitis patients.32 The pathophysiological background for this may be that the low ambient water vapor pressure in the winter impairs the ability of stratum corneum to retain water. Water vapor pressure is a better estimate of quantity of water vapor in the air than percent relative humidity, since the latter is dependent on the ambient temperature. Air of low temperature can contain a lower quantity of water vapor than air of high temperature. Water vapor pressure is proportional to dew point, which is defined as the temperature of the air at which the gaseous moisture begins to condense in the air.76 It is likely that differences in atmospheric conditions influence skin susceptibility in various geographical regions, determined by water vapor pressure, temperature, wind, and number of sunshine hours.
111.5 RECOMMENDATIONS In general, research in the field of irritancy testing has two opposite tendencies. One may wish to obtain enough signal to be able to discriminate between different detergents or different influencing factors, on the one hand. On the other hand, however, this signal must not be too strong to prevent suffering by the subjects involved in the testing. A nice compromise would be to discontinue further exposure in case of slight clinically apparent skin changes by one of the products. When in that case exposure of all other products is stopped, there may be not enought signal from these remaining products. Therefore, it may be considered to discontinue only the product that induces the clinical reaction, continuing exposure of the other products and stopping exposure of the one that gives the next visually discernable reaction. This should be done within a predetermined period, in repeated exposure schemes. An alternative compromise may be to stop exposure of all products when one of these reaches a certain level of (moderate) clinical irritation, to be determined beforehand. This latter procedure offers a better relative comparison. Authors should mention exactly which procedure they followed.
111.5.1 TEST INDIVIDUALS, REACTIVITY, LOCATION
AND
The possible influences of age, race, menstrual cycle, anatomical site, and season and climate on SLS reactivity
950
Handbook of Non-Invasive Methods and the Skin, Second Edition
should be considered. Also, previous conditions, such as atopic dermatitis or other skin diseases and exposures, including topical and oral medications, need consideration. Flexor side forearm skin, with cubital fossa and the wrist excluded, is a preferred study site. When different detergents or treatments are to be compared, it is recommended to change the position of these treatments cyclically from one subject to the other. Pre- and postexposure values should be given. The use of differences between pre- and postexposure values is optional. For repeated testing, measurements should be performed at least twice weekly, namely, on the first day of exposure and on the last day, before the applications.
111.5.2 TEST SUBSTANCE
AND
APPLICATIONS
High-purity (99%) SLS must be used in any study, dissolved in distilled water (w/v%) in occlusive and open testing, while tap water may be acceptable in immersion testing. Standard sized occlusion chambers with filter paper discs corresponding to large (12 mm Ø; maximal capacity, 70 μl; recommended volume, 60 μl) and extra large (18 mm Ø; maximal capacity, 300 μl; recommended volume, 200 μl) should be used. Finn® chambers are recommended, the extra large chambers for repeated applications. Test sites on the skin must be properly marked and remarked during repeated exposure studies. For repeated exposure, the use of chambers on a strip is preferred (Figure 111.4). Manufacturer’s standard filter discs should be used. Chambers should be applied to the skin immediately, i.e., within 1 min of preparation of the test solution. There must be good contact between the application system and the skin. For open exposures, a 20-mm-diameter plastic ring is advised. This ring must be fixed to the volar forearm skin, preferably by means of a strip and nonadhesive bandages (Figure 111.4). The volume of the solutions must be such that the total exposure area is covered (about 800 μl). For immersion testing, the forearms should be used, since these are usually not influenced by previous exposures to irritants, as opposed to the hands. In studying the hands, effects of more intense exposure of the dominant hand may interfere with the effects to be studied. The temperature of the immersion solution should be given, since this has been proven an important factor. For wash tests, also the forearms (or upper arms) are the preferred sites for the reasons mentioned above. The elbow folds are not advised in case noninvasive methods other than visual scoring are used, since elbow folds exhibit a higher level af sweat gland activity, which represents higher hydration and temperature. The temperature of the washing solution should be stated. Washing two times daily for 1 min is advisable, making one wash move-
FIGURE 111.4 Examples of a strip with 20-mm-diameter holes (left) for repeated open testing on the forearm, and a strip with extra large (18-mm) Finn chambers (right) for repeated occlusive testing on the forearm. (From Tupker RA, Willis C, Berardesca E, Lee CH, Fartasch M, Agner T, Serup J. Guidelines on sodium lauryl sulphate (SLS) exposure tests. A report from the standardization group of the European Society of Contact Dermatitis. Contact Derm 37:53–69, 1997. With permission.)
ment per second. This should be performed by the volunteer under supervision of a technician, or even better, by the technician himself or herself, in order to achieve compliance and some kind of standardization in this difficult field. TEWL measurement should be performed a minimum of 1 h after removal of test chambers, with the skin uncovered under laboratory room conditions in the meantime.
111.5.3 SPECIFIC TEST PROCEDURES The ESCD standards for one-time occlusion, repeated occlusion, open, and immersion tests are presented in Table 111.1. Procedures fall into the following categories: Provocative testing aims to induce an irritant reaction in all test individuals, without creating a caustic burn, in order to study irritant reactions, their modification by therapeutic or prophylactic measures, or for comparison with other irritants, particularly detergents. Susceptibility testing aims to study the individual by determining the susceptibility to SLS, either by a simple procedure with one standard concentration or, more detailed, using a dilution series and assessment of the irritation dose–response curve. Dilution series application may also be used in provocative testing in order to provoke reactions of a defined strength. Acute reaction testing aims at experimental study of the initial irritative damage to the skin. This may be desired for studying the time course after
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
951
TABLE 111.1 ESCD Guidelines on SLS Exposure Tests with TEWL Measurement Susceptibility Evaluation Acute Reaction One-time occlusion test Application time Mode of application SLS w/v% Repeated occlusion test Application time Application period Mode of application SLS w/v% Repeated open test Application time Application period Mode of application SLS w/v% Immersion testd Immersion time Application period Mode of application SLS w/v% Wash test Wash time Wash period Mode of washing SLS w/v%
Provocative Testing
Cumulative Reaction
Acute Reaction
Cumulative Reaction
24 h Chamber, 12 mm 0.5%e
Not applicable
24 h Chamber, 12 mm 2%e
Not applicable
Not applicable
2 h once daily 3 weeksf Chamber, 18 mm 0.25%
Not applicable
2 h once daily 3 weeksf Chamber, 18 mm 1%
60 min twice daily 1 day 20-mm guard ring 10%
30 min twice dailya 4 daysa,b 20-mm guard ring 2%a
Not possiblec
30 min twice dailya 4 daysa,b 20-mm guard ring 10%a
30 min twice daily 1 day Forearm immersion 0.5%
10 min twice dailya 4 daysa Forearm immersion 0.1%a
30 min twice daily 1 day Forearm immersion 2%
10 min twice dailya 4 daysa Forearm immersion 0.5%a
10 min 5 times daily 1 day Movement by hand, each s 5%
1 min twice daily 4 days Movement by hand, each s 5%
10 min 5 times daily 1 day Movement by hand, each s 10%
1 min twice daily 4 days Movement by hand, each s 10%
a
Deviation from the earlier guidelines.1 A longer application period is optional when a more delicate response over a more prolonged period is preferred. In that case, a 3-week model can be used, with 10-min once daily applications of 1% solution for susceptibility evaluation, and 2% solution for provocative testing. c In temperate zones it is not possible to elicit an irritation response in all subjects using 10% SLS for 60 min twice daily, and longer exposure times are not feasible. d Water temperature, 35˚C. e May also be performed as a one-time application of a dilution series (SLS 4, 2, 1, 0.5, 0.25, 0.125, 0.0625%, 0% w/v). f One week is five application days.
b
this irritation, or to compare the efficacy of various moisturizers in healing this reaction. Cumulative reaction testing aims at clinically realistic study of chronic irritant contact dermatitis taking ongoing repair, with increase or lowering of the threshold, into consideration. A time course of clinical signs may be erythema in the first week, changing in scaling and roughness in the third week, although in other circumstances or other subjects erythema (together with the other signs) may be observed in the third week only. The clinical appearance of acute and cumulative reactions differs. The guideline therefore includes separate
standard clinical scoring schemes for the two types of reactions (Table 111.2A and B and Table 111.3). The use of a scoring system for subjective irritation is optional (Table 111.4). When large areas are involved, such as in immersion testing, the clinical signs may have an uneven spread. In that case, it is advised to score a representative area.
111.5.4 INDIVIDUAL PILOT STUDY In recognition of the range of influencing variables and their complex interaction, it is recommended that a pilot study be performed in order to ensure that any deviation from the recommended standard concentrations given in the guidelines is justifiable.
952
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 111.2A ESCD Guideline on Clinical Scoring of Acute SLS Irritant Reactions, Simple Scoring Systema Score
Qualification
Description
0 1/2 1 2
Negative Doubtful Weak Moderate
3
Strong
4
Very strong/caustic
No reaction Very weak erythema or minute scaling Weak erythema, slight edema, slight scaling, or slight roughnessb Moderate degree of erythema, edema, scaling, or roughness, or minor degree of erosions, vesicles, crusting, or fissuring Marked degree of erythema, edema, scaling, roughness, erosions, vesicles, bullae, crusting, or fissuring As 3, with necrotic areas
Note: Reading 25 to 96 h after one-time exposure. a
The ESCD simple scoring system is based on the scoring system according to Frosch and Kligman 3 and may be used when no subdivision into the different qualities of irritation (erythema, scaling, roughness, edema, fissures) is wanted. b The term roughness is used for reactions that can be felt as rough or dry, sometimes preceeded or followed by visible changes of the surface’s contour, in contrast to scaling, which is accompanied by visible small flakes.
TABLE 111.2B ESCD Guideline on Clinical Scoring of Acute SLS Irritant Reactions, Elaborate Scoring Systema Quality
1
/2
1
Each of the irritation qualities is graded according to the following Erythema Very weak Weak, diffuse, or spotty Roughness/contour Shiny surface Slight roughnessb or wrinkled surface Scaling — Minute flakes Edema — Slight Fissures — Fine fissures
2
3
system: Moderate Moderate roughness
Marked Marked roughness
Moderate flakes Moderate Broad fissures
Large flakes Marked Wide fissures with hemorrhage or exudation
a
The elaborate scoring system is based on the scoring system according to Frosch and Kligman3 and may be used when a subdivision into the different qualities of irritation (erythema, scaling, roughness, edema, fissures) is necessary, or when a better balanced estimation of reaction intensity is warranted. b The term roughness is used for reactions that can be felt as rough or dry, sometimes preceeded or followed by visible changes of the surface’s contour, in contrast to scaling, which is accompanied by visible small flakes.
111.5.5 INTERPRETATION OF SLS EXPOSURE/RESULTS FROM NONINVASIVE EVALUATION METHODS SLS is a model irritant, suited for precise testing and the highly sensitive measurement of response by means of TEWL. However, irritant substances in general differ greatly in skin penetration capabilities and mode of responses. Results of testing with SLS cannot therefore
be uncritically projected on to other types of irritation. TEWL is not a clinical sign or symptom per se, but it is an indirect measure of the skin response. The same holds true for the other noninvasive evaluation methods, all of which measure only one particular aspect of skin irritation. Results must be compared with clinical observations, epidemiological data, or comparative data obtained using other noninvasive methods.
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
953
TABLE 111.3 ESCD Guideline on Clinical Scoring of Subacute/Cumulative SLS Irritant Reactions, Simple Scoring Systema Score
Qualification
0 1/2 1
Negative Doubtful Weak
2
Moderate
3
Strong
4
Very strong/caustic
Description No reaction Very weak erythema or shiny surfaceb Weak erythema, diffuse or spotty, slight scaling, or slight roughnessc Moderate degree of erythema, scaling, roughness or weak edema or fine fissures Marked degree of erythema, scaling, roughness, edema, fissures or presence of papules or erosions or vesicles As 3, with necrotic areas
Note: The elaborate scoring system may be used when a subdivision into the different qualities of irritation (erythema, scaling, roughness, edema, fissures) is necessary, or when a better balanced estimation of reaction intensity is warranted. Each of the irritation qualities is graded according to the scoring system described in Table 111.2B. a
The ESCD simple scoring system may be used when no subdivision into the different qualities of irritation (erythema, scaling, roughness, edema, fissures) is wanted. b The term shiny surface is used for those minimal reactions that can only be discerned when evaluated in skimming light as a shiny area. It may be the precursor of the so-called soap effect (see Figure 111.3). c The term roughness is used for reactions that can be felt as rough or dry, sometimes preceeded or followed by visible changes of the surface’s contour, in contrast to scaling, which is accompanied by visible small flakes.
TABLE 111.4 ESCD Guideline on Subjective Scoring of the SLS Irritant Reactions during or after Exposure (Optional) Qualification Negative Weak Moderate Strong
Score
Description
0 1 2 3
No burning/stinging sensation Weak burning/stinging Moderate burning/stinging Strong burning/stinging
REFERENCES 1. Tupker RA, Willis C, Berardesca E, Lee CH, Fartasch M, Agner T, Serup J. Guidelines on sodium lauryl sulphate (SLS) exposure tests. A report from the standardization group of the European Society of Contact Dermatitis. Contact Derm 37: 53–69, 1997. 2. Bettley FR. The irritant effect of detergents. Trans St Johns Hosp Derm Soc 58: 65–74, 1972. 3. Frosch PJ, Kligman AM. The soap chamber test. J Am Acad Dermatol 1: 35–41, 1979. 4. Staberg B, Serup J. Allergic and irritant skin reactions evaluated by laser Doppler flowmetry. Contact Derm 18: 40–45, 1988.
5. Kligman AM, Wooding WM. A method for the measurement and evaluation of irritants on human skin. J Invest Dermatol 49: 78–94, 1967. 6. Agner T, Serup J, Handlos V, Batsberg W. Different skin irritation abilities of different qualities of sodium lauryl sulphate. Contact Derm 21: 184–188, 1989. 7. Fartasch M. Ultrastructure of the epidermal barrier after irritation. Microsc Res Tech 37: 193–199, 1997. 8. Willis CM, Stephens CJM, Wilkinson JD. Epidermal damage induced by irritants in man: a light and electron microscopic study. J Invest Dermatol 93: 695–699, 1989. 9. Willis CM, Stephens CJM, Wilkinson JD. Differential patterns of epidermal leukocyte infiltration in patch test reactions to structurally unrelated chemical irritants. J Invest Dermatol 101:364–370, 1993. 10. Willis CM. Variability in responsiveness to irritants: thoughts on possible underlying mechanisms. Contact Derm 47:267–271, 2002. 11. Driesch von den P, Fartasch M, Hüner A, Ponec M. Expression of integrin receptors and ICAM-1 on keratinocytes in vivo and in an in vitro reconstructed epidermis: effect of sodium lauryl sulphate. Arch Dermatol Res 287:249–253, 1995. 12. Björnberg A. Skin Reactions to Primary Irritants in Patients with Hand Eczema. Thesis, Oscar Isacsons Tryckeri AB, Gothenburg, 1968.
954
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. Tupker RA, Pinnagoda J, Coenraads PJ, Nater JP. The influence of repeated exposure to surfactants on the human skin as determined by transepidermal water loss and visual scoring. Contact Derm 20:108–114, 1989. 14. Wilhelm K-P, Freitag G, Wolff HH. Surfactant-induced skin irritation and skin repair. J Am Acad Dermatol 30:944–949, 1994. 15. Malten KE. Thoughts on irritant contact dermatitis. Contact Derm 7:238–247, 1981. 16. Pirilä V. Chamber test versus patch test for epicutaneous testing. Contact Derm 1:48–52, 1975. 17. Frosch PJ, Kligman AM. The Duhring chamber. Contact Derm 5:73–81, 1979. 18. Hannuksela A, Hannuksela M. Irritant effects of a detergent in wash, chamber and repeated open application tests. Contact Derm 34:134–137, 1996. 19. Dahl MV, Roering MJ. Sodium lauryl sulphate irritant patch tests. III. Evaporation of aquous vehicle influences inflammatory response. J Am Acad Dermatol 11:477–479, 1984. 20. Berardesca E, Vignoli GP, Distante F, Brizzi P, Rabbiosi G. Effects of water temperature on surfactant-induced skin irritation. Contact Derm 32:83–87, 1995. 21. Agner T, Serup J. Sodium lauryl sulphate for irritant patch testing: a dose-response study using bioengineering methods for determination of skin irritation. J Invest Dermatol 95:543–547, 1990. 22. Nassif A, Chan SC, Storrs FJ, Hanifin JM. Abnormal skin irritancy in atopic dermatitis and atopy without dermatitis. Arch Dermatol 130:1402–1407, 1994. 23. Aramaki J, Loffler C, Kawana S, Effendy I, Happle R, Loffler H. Irritant patch testing with sodium lauryl sulfate: interrelation between concentration and exposure time. Br J Dermatol 145:704–708, 2001. 24. Basketter DA, Griffiths HA, Wang XM, Wilhelm KP, McFadden J. Individual, ethnic and seasonal variability in irritant susceptibility of skin: the implications for a predictive human patch test. Contact Derm 35:208–213, 1996. 25. Agner T, Serup J. Time course of occlusive effects on skin evaluated by measurement of transepidermal water loss (TEWL). Contact Derm 28:6–9, 1993. 26. Pinnagoda J, Tupker RA, Coenraads PJ, Nater JP. Prediction of susceptibility to an irritant response by transepidermal water loss. Contact Derm 20:341–346, 1989. 27. Freeman S, Maibach HI. Study of irritant contact dermatitis produced by repeat patch test with sodium lauryl sulphate and assessed by visual methods, transepidermal water loss, and laser Doppler velocimetry. J Am Acad Dermatol 19:496–502, 1988. 28. Tupker RA, Pinnagoda J, Nater JP. The transient and cumulative effect of sodium lauryl sulphate on the epidermal barrier assessed by transepidermal water loss: inter-individual variation. Acta Derm Venereol (Stockh) 70:1–5, 1990. 29. Serup J, Staberg B. Differentiation of allergic and irritant reactions by transepidermal water loss. Contact Derm 16:129–132, 1987.
30. Tupker RA, Coenraads PJ, Pinnagoda J, Nater JP. Baseline transepidermal water loss (TEWL) as a prediction of susceptibility to sodium lauryl sulphate. Contact Derm 20:265–269, 1989. 31. Tupker RA, Coenraads PJ, Fidler V, de Jong MCJM, van der Meer JB, De Monchy JGR. Irritant susceptibility and wheal and flare reactions to bio-active agents in atopic dermatitis. I. Influence of disease severity. Br J Dermatol 133:358–364, 1995. 32. Tupker RA, Coenraads PJ, Fidler V, de Jong MCJM, van der Meer JB, De Monchy JGR. Irritant susceptibility and wheal and flare reactions to bio-active agents in atopic dermatitis. II. Influence of season. Br J Dermatol 133:365–370, 1995. 33. Tupker RA, Pinnagoda J, Coenraads PJ, Nater PJ. Susceptibility to irritants: role of barrier function, skin dryness and history of atopic dermatitis. Br J Dermatol 123:199–205, 1990. 34. Lammintausta K, Maibach HI, Wilson D. Irritant reactivity in males and females. Contact Derm 17:276–280, 1987. 35. Lammintausta K, Maibach HI, Wilson D. Susceptibility to cumulative and acute irritant dermatitis. Contact Derm 19:84–90, 1988. 36. Wilhelm K-P, Saunders JC, Maibach HI. Increased stratum corneum turnover induced by subclinical irritant contact dermatitis. Br J Dermatol 122:793–798, 1990. 37. Van der Valk PGM, Maibach HI. Post-application occlusion substantially increases the irritant response to the skin to repeated short-term sodium lauryl sulfate (SLS) exposure. Contact Derm 21:335–338, 1989. 38. Frosch PJ, Kurte A, Pilz B. Efficacy of skin barrier creams. III. The repetitive irritation test (RIT) in humans. Contact Derm 29:113–118, 1993. 39. Tupker RA, Vermeulen K, Fidler V, Coenraads PJ. Irritancy of sodium laurate and other anionic detergents using an open exposure model. Skin Res Technol 3:133–136, 1997. 40. Tupker RA, Bunte EE, Fidler V, Wiechers JW, Coenraads PJ. Irritancy ranking of anionic detergents using one-time occlusive, repeated occlusive and repeated open tests. Contact Derm 40:316–322, 1999. 41. Smeenk G. The influence of detergents on the skin (a clinical and biochemical study). Arch Klin Exp Dermatol 235:180–191, 1969. 42. Allenby CF, Basketter DA, Dickens A, Barnes EG, Brough HC. An arm immersion model of compromised skin. I. Influence on irritation reactions. Contact Derm 28:84–88, 1993. 43. Øhlenslæger J, Friberg J, Ramsing D, Agner T. Temperature dependency of skin susceptibility to water and detergents. Acta Derm Venereol (Stockh) 76:274–276, 1996. 44. Paye M, Gomes G, Zerweck CR, Piérard GE, Grove GL. A hand immersion test under laboratory-controlled usage conditions: the need for sensitive and controlled assessment methods. Contact Derm 40:133–138, 1999. 45. Imokawa G, Mishima Y. Cumulative effect of surfactants on cutaneous horny layers: adsorption onto human keratin layers in vivo. Contact Derm 5:357–366, 1979.
Sodium Lauryl Sulfate (SLS) Testing: ESCD Application and Reading Standards
46. Frosch PJ. Irritancy of soaps and detergent bars. In Principles of Cosmetics for the Dermatologist, Frost P, Horwitz SN, Eds. CV Mosby Company, St. Louis, 1982, pp. 5–12. 47. Wigger-Alberti W, Fischer T, Greif C, Maddern P, Elsner P. Effects of various grit-containing cleansers on skin barrier function. Contact Derm 41:136–140, 1999. 48. Gehring W, Gloor M, Kleesz P. Predictive washing test for evaluation of individual eczema risk. Contact Derm 39:8–13, 1998. 49. Blank JH, Gould J. Penetration of anionic surfactants into skin. III. Penetration from buffered sodium laurate solutions. J Invest Dermatol 37:485–488, 1961. 50. Treffel P, Muret P, Muret-D’Aniello P, Coumes-Marquet S, Agache P. Effect of occlusion on in vitro percutaneous absorption of two compounds with different physicochemical properties. Skin Pharmacol 5:108–113, 1992. 51. Cua AB, Wilhelm KP, Maibach HI. Cutaneus sodium lauryl sulphate irritation potential: age and regional variability. Br J Dermatol 123:607–613, 1990. 52. Elsner P, Wilhelm D, Maibach H. Sodium lauryl sulphate-induced irritant contact dermatitis in vulvar and forearm skin of premenopausal and postmenopausal women. J Am Acad Dermatol 23:648–652, 1990. 53. Berardesca E, Maibach HI. Racial differences in sodium lauryl sulphate induced cutaneus irritation: black and white. Contact Derm 18:65–70, 1988. 54. Goh CL, Chia SE. Skin irritability to sodium lauryl sulphate — as measured by skin water vapour loss — by sex and race. Clin Exp Dermatol 13:16–19, 1988. 55. Agner T. Basal transepidermal water loss, skin thickness, skin blood flow and skin colour in relation to sodium-lauryl-sulphate-induced irritation in normal skin. Contact Derm 25:108–114, 1991. 56. Agner T, Damm P, Skouby SO. Menstrual cycle and skin reactivity. J Am Acad Dermatol 24:566–570, 1991. 57. Pinnagoda J, Tupker RA, Agner T, Serup J. Guidelines for transepidermal water loss (TEWL) measurement. A report from the standardization group of the European Environmental and Contact Dermatitis Society. Contact Derm 22:164–178, 1990. 58. Van der Valk PGM, Maibach HI. Potential for irritation increases from the wrist to the cubital fossa. Br J Dermatol 121:709–712, 1989. 59. Murahata R, Crove DM, Roheim JR. The use of transepidermal water loss to measure and predict the irritation response to surfactants. Int J Cosmet Sci 8:225–231, 1986. 60. Agner T. Skin susceptibility in uninvolved skin of hand eczema patients and healthy controls. Br J Dermatol 125:140–146, 1991. 61. Wilhelm K-P, Maibach HI. Susceptibility to irritant dermatitis induced by sodium lauryl sulphate. J Am Acad Dermatol 23:122–124, 1990.
955
62. John SM, Uter W, Schwanitz HJ. Relevance of multiparametric skin bioengineering in a prospectively-followed cohort of junior hairdressers. Contact Derm 43:161–168, 2000. 63. Smit HA, van Rijssen A, Vandenbroucke JP, Coenraads PJ. Individual susceptibility and the incidence of hand dermatitis in a cohort of hairdressers and nurses. Scan J Work Environ Health 20:113–121, 1994. 64. Goh CL, Gan SL. Efficacies of barrier creams and afterwork emollient cream against cutting fluid dermatitis in metal workers: a prospective study. Contact Derm 31:176–181, 1994. 65. Allen MH, Wakelin SH, Holloway D, Lisby S, Baadsgaard O, Barker JN, McFadden JP. Association of TNFA gene polymorphism at position –308 with susceptibility to irritant contact dermatitis. Immunogenetics 51:201–205, 2000. 66. Lammintausta K, Maibach HI, Wilson D. Human cutaneous irritation: induced hyporeactivity. Contact Derm 17:193–198, 1987. 67. Widmer J, Elsner P, Burg G. Skin irritant reactivity following experimental cumulative irritant contact dermatitis. Contact Derm 30: 35–39, 1994. 68. Werner Y, Lindberg M. Transepidermal water loss in dry and clinically normal skin in patients with atopic dermatitis. Acta Derm Venereol (Stockh) 65:102–105, 1985. 69. Lévêque JL, Grove G, de Rigal J, Corcuff P, Kligman AM, Saint Leger D. Biophysical characterization of dry facial skin. J Soc Cosmet Chem 82:171–177, 1987. 70. Shahidullah M, Raffle EJ, Rimmer AR, Frain-Bell W. Transepidermal water loss in patients with dermatitis. Br J Dermatol 81:722–730, 1969. 71. Blichman C, Serup J. Hydration studies on scaly hand eczemas. Contact Derm 16:155–159, 1987. 72. Gloor M, Völlm P, Gehse M, Ringelmann R. Irritationseffekt von Tensiden bei Patienten mit Gewerbeekzemen im Friseur-und Krankenpflegeberuf. Dermatosen 33:86–89, 1985. 73. Maibach HI. The E.S.S.: excited skin syndrome (alias the “angry back”). In New Trends in Allergy, Ring J, Burg G, Eds. Springer-Verlag, Berlin, 1981, pp. 208–221. 74. Van der Valk PGM, Nater JP, Bleumink E. Vulnerability of the skin to surfactants in different groups of eczema patients and controls as measured by water vapour loss. Clin Exp Dermatol 10:98–103, 1985. 75. Agner T, Serup J. Seasonal variation of skin resistance to irritants. Br J Dermatol 121:323–328, 1989. 76. Gaul LE, Underwood GB. Relation of dew point and barometric pressure to chapping of normal skin. J Invest Dermatol 19:9–19, 1952.
and Computer-Based 112 Instrumental Methods for Measurement of Surface Area Afflicted with Disease Chil Hwan Oh Department of Dermatology, School of Medicine, Korea University, Seoul, Korea
CONTENTS 112.1 112.2 112.3 112.4 112.5
Introduction ..........................................................................................................................................................957 Computer Image Analysis Using a Digital Camera (CIAD) ..............................................................................958 Computer Image Analysis of Direct Tracings (CIAT) ........................................................................................959 Computer Image Analysis with Analog Camera .................................................................................................960 Clinical and Experimental Use ............................................................................................................................959 112.5.1 Validity of the Rule of Nines ................................................................................................................960 112.5.2 Assessment of Involved Area in Atopic Dermatitis (Computer Image Analysis with CIAD and CIAT) ...................................................................................................................................961 112.5.3 Assessment of Involved Area in Psoriasis (Computer Image Analysis with Analog Camera).....................................................................................................................................963 112.6 Future, Potential Limitations, and Pitfalls...........................................................................................................964 References .......................................................................................................................................................................965
112.1 INTRODUCTION The objective quantification of the severity of cutaneous diseases is very important, because it is a fundamental aspect of the clinical evaluation of the disease. This type of quantification is required not only for the etiological study of cutaneous disorders, but also for deciding upon the most effective treatment modality.1,2 To measure the effects of health care intervention accurately, reliable methods of assessment should be used. There are many tools used as clinical scoring systems to measure skin disease activity in various skin diseases, e.g., the psoriasis area severity index (PASI), the severity scoring of atopic dermatitis (SCORAD), the Leeds acne grading scale, the melasma area severity index (MASI), the severity weighted assessment tool (SWAT) for mycosis fungoides, the systemic lupus activity measure (SLAM), and so on.3–7 The estimation of the involved surface area has been the weakest aspect of clinical scoring systems.8 There are marked interobserver variations between the estimates of different clinicians, and this may have a significant impact on the resulting scoring system.3
Visual scoring methods, which essentially constitute a subjective scoring system whose reliability cannot be demonstrated, may produce misleading results that should therefore be interpreted with caution.8–10 The most widely used visual scoring method for measuring surface areas afflicted with diseases is the rule of nines, in which the body surface area is divided into nine anatomical regions, consisting of the combined head and neck area, each arm, the front and back of each leg, and the trunk, with each of these regions representing 9% of the total surface area, plus 1% for the genitalia. This method was originally developed for estimating the involved surface area of burns.9,11 In one study involving visual scoring with clinical photographs, the mean surface area, as assessed by a nonexpert investigator, ranged from 20 to 100%, compared with the estimations made by the different experts, which ranged from 40 to 60%. Ideally, all observers involved in using body surface area measurement should be trained in this technique. It is generally recommended that the extent of individual lesions be drawn on the printed figure of the evaluation sheet, in order to take into account skin lesions. The currently used rule-of-nines method for assessing the total area of involvement 957
958
Handbook of Non-Invasive Methods and the Skin, Second Edition
represents a critical component of the SCORAD index4 in atopic dermatitis, and is also critical in the PASI system3 in the case of psoriasis. However, previous studies have indicated the existence of wide variations in both interobserver and intraobserver estimations.8,13 The other way to simply and rapidly assess the approximate area of involvement is referred to as the rule of hand, with the hand being used as an area measure. The area of the flat of one adult hand has often been taken to be approximately 1% of the body surface area. More accurate measurement, however, has shown that on average it covers 130 cm2, that is, approximately 0.75% of the adult body surface area.12 The PASI scoring system has been most commonly used to objectively evaluate the level of clinical symptoms in psoriasis until recently. The PASI scores include the involved surface areas of the body and the severity of individual psoriatic lesions, such as erythema, scaling, and infiltration. The involved area of the body is divided into four parts, the head, trunk, upper limbs, and lower limbs, which are taken to occupy 10, 30, 20, and 40% of the total body surface area, respectively.3 The involved areas are calculated as percentages of the individual parts of the body, and these percentages are then converted into numerical values ranging from 0 to 6. The severity of each psoriatic lesion is evaluated on a scale of 0 to 4, in order to assess the extent of erythema, infiltration, and scaling. As a result, the final PASI scores are expressed in the form of numerical values ranging from 0 to 72.3 Although this PASI scoring system is economic, speedy, and convenient, individual psoriatic lesions may give rise to different interpretations, depending on the physician’s experience or knowledge.8 New techniques have been developed to assess psoriasis disease severity, including computer-assisted planimetry from color photographs, as well as a computer analysis system based on whole-body black-and-white photographs.9,13 According to a report by Ramsay and Lawrence,9 in which surface area assessments were made from whole-body black-and-white photographs and were compared with estimates obtained from the image analysis and those of traced outlines, untrained observers were found to overestimate the extent of psoriasis.13 Marks et al.8 suggested that even experienced clinicians may differ in their estimates of the body area involved in psoriasis. This may have an important effect on the PASI score. Since it involves the calculation of the involved area based on the observer’s subjective estimation, the conventional PASI scoring system is considered not to provide an accurate objective measurement.3 The other objective method of determining the area of skin involvement is by mapping the disease onto a body diagram, and then evaluating the extent of involvement by grid point counting of area fractions drawn on the diagram. This evaluation is performed by calculating the number
of point intersections divided by the total number of points that fall on the whole-body diagram.10 By using a color image and comparing the ratio of green and red light, areas of similarly colored normal skin can be identified, and the color difference due to psoriatic plaque will have a different green/red ratio, and thus be able to be distinguished from the surrounding normal skin.14 There are many clinical severity scales that are used in atopic dermatitis, such as baseline grading, the eczema area and severity index (EASI), the six area and six sign atopic dermatitis (SASSAD) severity index, the Nottingham eczema severity score, the atopic dermatitis area and severity index (ADASI), the severity scoring of atopic dermatitis (SCORAD) index, and so on.4,15–20 This plethora of techniques suggests that there is no absolute right or wrong way to perform this evaluation. In order to be able to compare results from different studies, it would be advantageous if a more accurate and uniform system were used to evaluate the severity of skin diseases. Of the severity scoring systems currently available for atopic dermatitis, the SCORAD index has been most extensively tested. However, the rule-of-nines method currently used to assess the total surface area of patients with atopic dermatitis, which is included in the SCORAD index, is neither an accurate nor a reproducible method. This may have an important effect on the PASI score, which is similar to the SCORAD index in atopic dermatitis. In addition, this approach is fraught with difficulties because of the diffuse nature of atopic dermatitis.20 Even in the case of psoriasis, a relatively well demarcated disease, difficulties are recognized. Atopic dermatitis has illdefined skin lesions, so it is difficult to make measurement accurately. According to many investigators, the reliability of the methods currently being used is poor.8–10 There exist marked interobserver variations between the estimates of different clinicians, and these may have a significant impact on the resulting scoring system.3 In this chapter, the surface area afflicted with the skin disease is measured by computer image analysis with a digital camera (CIAD), computer image analysis of direct tracing (CIAT), and computer image analysis with analog camera.21,22
112.2 COMPUTER IMAGE ANALYSIS USING A DIGITAL CAMERA (CIAD) Digital images of eight parts of the body (both anterior and posterior arms, anterior and posterior legs, and anterior and posterior trunk) were taken by one dermatologist using the same digital camera (DSC-707®, SONY Co., Tokyo, Japan), lighting, and patient posture. A rectangular 10 × 1.9 cm2 sized calibration bar was applied to the skin surface (Figure 112.1 and Figure 112.2).
Instrumental and Computer-Based Methods for Measurement of Surface Area Afflicted with Disease
959
Digital camera
Frame grabber
Analogue camera
PC Mouse Image analysis program
Composite video
FIGURE 112.1 Schematic diagram of the computerized image analysis system with digital and analog cameras.
FIGURE 112.2 Computer image analysis system using a digital camera (CIAD).
Based on these images, the total skin surface area and the involved area of each part of the body were measured using an image analysis program (Image-Pro Plus 4.1®, Media Cybernetics, Inc., GA). The digital images were composed of 2048 × 1536 pixels. By comparing the number of pixels in the image to the number of pixels in the known area of the calibration bar, the computer image analysis system produced the total surface area and the involved area of each part of the body in square centimeters (Figure 112.2). The eight parts of the body were analyzed separately, taking 60 to 90 min/patient, and the area estimates of each part were added to calculate the total involved area.
112.3 COMPUTER IMAGE ANALYSIS OF DIRECT TRACINGS (CIAT) A transparent vinyl chloride sheet containing six parts that fitted the body area (both arms and legs, and the anterior and posterior trunk) was used for these measurements. The skin lesions were then traced taking 60 minutes/patient. Once the same calibration bar that had been used in the computer image analysis using a digital camera (CIAD) had been traced on these sheets, six photographs
FIGURE 112.3 Computer image analysis system of direct tracings of the lesions (CIAT).
of the sheets were taken using the digital camera. From the images of the six traced sheets, the total skin surface area and the involved area of each part of the body were determined in the same way as that described for the CIAD method (Figure 112.3). The computer image analysis of direct tracing (CIAT) method was assumed to provide the most accurate measurement of the surface area. In the tracing method, the involved surface area had to be drawn, photographed by the digital camera, and measured using the computer image analysis system. This took 2 to 2.5 hours per patient to complete, indicating that this method is unsuitable for routine use in large studies.4 According to the rule of nines9 and Long et al.’s study,12 the total area measured was reduced by 19% to exclude the areas of the head, neck, groin, hands, and feet, which might produce a high measurement error and lead to low compliance for the tracing method. The involved surface areas obtained from the CIAD and CIAT methods were expressed as a percentage of this corrected body surface area.
112.4 COMPUTER IMAGE ANALYSIS WITH ANALOG CAMERA Eight photographs were taken for each patient (head, anterior and posterior trunk, both anterior and posterior upper arms, both anterior and posterior lower legs) using an analog camera (F80®, Nikon Co., Japan) under the same conditions of lighting and posture for each patient. Then the involved area of each part of the body was converted to a percentage using the image analysis system (MI-PCP Version 2.32, Belvoir Consulting & AIC), following the conversion of the photographs taken into digital images by means of a frame grabber (PC vision plus, Image Technology, Inc.) (Figure 112.1 and Figure 112.4A and B). Also, the total involved area was calculated for each patient by adding up the percentages of the involved areas for each part of the body, according to the following
960
Handbook of Non-Invasive Methods and the Skin, Second Edition
A
B
FIGURE 112.4 (A) The lesions of a patient with psoriasis. (B) The application of image analysis to the psoriatic lesions (computer image analysis with analog camera).
repartition: 10% (head), 30% (trunk), 20% (upper limbs), and 40% (lower limbs).
112.5 CLINICAL AND EXPERIMENTAL USE 112.5.1 VALIDITY
OF THE
RULE
OF
However, no significant difference was observed for the arms (Table 112.2). The rule of nines is therefore more accurate in patients aged 3 years or younger than in those aged 4 years or older.
NINES
In order to confirm the validity of the rule of nines, the total percentage area of each body region defined by the rule of nines was compared with CIAT determined areas. The 23 atopic patients were divided into two groups. Group 1 consisted of 8 patients aged 3 years or younger, and group 2 consisted of 15 patients aged 4 years or older. The visual scoring method indicated the area of the skin lesions as a percentage of the whole area of the respective part of the body by the visual assessment of four dermatologists, and the total involved area was calculated. According to the rule of nines, the percentage areas of the trunk, arms, and legs are 36, 18, and 36% of the total body area, respectively. Four dermatologists were asked to simultaneously assess the involved area using the photograph captured by the camera in the form of two separate assessments on two consecutive days using the rule of nines. The dermatologists had no record of their previous estimations when performing the new assessment on the second day of observation, and the order of the assessments was randomized to minimize memory recall bias. The percentage area of each total body region based on the rule of nines was compared to that obtained from CIAT. In group 1, no significant difference in the percentage areas of any part of the body was observed between the rule of nines and the CIAT. On the other hand, group 2 showed significant differences in this respect. In the case of the trunk, the estimate from CIAT was lower than that obtained using the rule of nines, and for the legs, the CIAT estimate was greater than that of the rule of nines.
TABLE 112.1 Comparisons of the Total Area of Each Body Region in Groups 1 and 2a
Trunk Arm Leg a b c d
Rule of Nines
Group 1b (n = 8)
Group 2c (n = 15)
Total
36% 18% 36%
34.62 ± 3.41 18.18 ± 2.00 37.21 ± 3.38
31.59 ± 3.76 d 18.48 ± 1.88 39.93 ± 3.17 d
32.64 ± 3.85d 18.38 ± 1.88 38.98 ± 3.43
Wilcoxon signed rank test was done. Patients aged 3 years or younger. Patients aged 4 years or older. p < 0.05.
TABLE 112.2 Comparisons of Estimates of the Involved Surface Area in Each Anatomic Region by CIAT and CIA
Trunk Arm Leg a b
CIAT (%)a
CIAD (%)a
p Valueb
9.96 ± 12.35 7.21 ± 8.24 7.23 ± 9.40
9.45 ± 10.74 7.84 ± 9.39 8.40 ± 10.31
>0.05 >0.05 >0.05
Mean ± SD. Compared by Student’s paired t-test.
Instrumental and Computer-Based Methods for Measurement of Surface Area Afflicted with Disease
112.5.2 ASSESSMENT OF INVOLVED AREA IN ATOPIC DERMATITIS (COMPUTER IMAGE ANALYSIS WITH CIAD AND CIAT) In comparison to psoriatic lesions, atopic dermatitis gives rise to ill-defined, diffuse-natured individual lesions. So, it is preferable to use the digital camera system to improve the quality of the photographs in the case of clear cutaneous lesions. The skin lesions of atopic dermatitis patients were estimated by using a visual scoring method, computer image analysis using a digital camera (CIAD), and computer image analysis of direct tracings of the lesions (CIAT). The comparison between the visual scoring method and the CIAT showed that, on the whole, the four dermatologists overestimated the involved area, and that, for each dermatologist, the differences between the results obtained using the visual scoring method and those obtained using the CIAT on each individual patient were statistically significant. Moreover, the mean overestimate was 3.5-fold the results of the CIAT (Figure 112.5). There were highly significant differences between the assessments of the individual dermatologists. On the other hand, three of the four dermatologists showed little day-to-day variation; only one dermatologist’s estimation differed significantly between days 1 and 2. This lack of difference is attributed to the interval between estimations being limited to 24 hours, and the fact that the observers were relatively well-trained dermatologists (Figure 112.6). However, there was some correlation between the visual scoring method and CIAT in that those patients having high scores on the CIAT system also had high scores on
961
the visual scoring method. In general, the visual scoring system overestimated the surface area involved (Figure 112.7). This may have a significant effect on the SCORAD index in atopic dermatitis, which is similar to the PASI score. The reliability of CIAD is tested with CIAT, which was presumed to precisely measure the involved area, although admittedly no method used in this study was absolutely accurate. On comparing CIAT with CIAD, no significant difference was found between the two methods. In terms of the area measurements, the agreement was high, especially on the trunk and arm areas. However, the leg areas of CIAD slightly exceeded the corresponding CIAT areas (Figure 112.8). It should be noted that in a photograph, a three-dimensional object is transformed into a two-dimensional photograph. As a consequence, there is some loss of lateral surface area, especially for cylindrical objects.9,10 In addition, the lesions of patients with atopic dermatitis are located mainly on the anterior or posterior areas of the limbs, rather than the lateral areas. Therefore, the percentage areas (the involved area/total surface area) as assessed by CIAD were overestimated compared to the corresponding values obtained using CIAT. The CIAT and CIAD methods might seem to show more variation in their scores between individual patients than the visual scoring method; however, this is not because these systems are more sensitive, but because of the wide variation in the estimates of the involved area. In 3 of the 23 patients these estimates were greater than 34% of the total body surface area, compared with the remaining 20 patients, in which the involved areas were lower than 13%.
Observer 1 80.00 P < 0.01 70.00
Area(%)
60.00
Visual grading method Computer image analysis of direct tracings
50.00 40.00 30.00 20.00 10.00 0.00
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 Patients
FIGURE 112.5 Comparison of estimates between the visual scoring made using the rule of nines and the computer image analysis of direct tracings (CIAT) of the lesions. The results show that, on the whole, the four dermatologists overestimated the involved area compared to the values obtained using CIAT, and that, in each case, the differences between the values obtained by the visual scoring and those obtained using CIAT were statistically significant (p < 0.01, Student’s paired t-test).
962
Handbook of Non-Invasive Methods and the Skin, Second Edition
Observer 2 80.00 P < 0.01 70.00
Area(%)
60.00
Visual grading method Computer image analysis of direct tracings
50.00 40.00 30.00 20.00 10.00 0.00
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 Patients
Observer 3 80.00 P < 0.01 70.00
Area(%)
60.00
Visual grading method Computer image analysis of direct tracings
50.00 40.00 30.00 20.00 10.00 0.00
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 Patients
Observer 4 80.00 P < 0.01 70.00
Area(%)
60.00
Computer image analysis of direct tracings
50.00 40.00 30.00 20.00 10.00 0.00
Visual grading method
1
2
3
FIGURE 112.5 (Continued.)
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 19 20 21 22 23 Patients
Instrumental and Computer-Based Methods for Measurement of Surface Area Afflicted with Disease
963
80 70
Area (%)
60 50 40 30 20 10 0
Day 1 Day 2
Day 1 Day 2
Day 1* Day 2*
Day 1 Day 2
Dermatologist 1 Dermatologist 2 Dermatologist 3** Dermatologist 4 * P < 0.05 comparison between consecutive days by 3rd dermatologist (Student’s t-test) ** P < 0.05 comparison between four dermatologists (ANOVA)
FIGURE 112.6 Comparison of the results obtained by the four dermatologists on consecutive days using the visual scoring method. 80 n = 23
n = 23
n = 23
Area (%)
60
40
20
0
CIAD
Visual scoring method
CIAT
Visual scoring method
CIAT
CIAD
FIGURE 112.7 Comparisons of results between the computer image analysis of tracing (CIAT), the computer image analysis with the digital camera (CIAD), and the visual scoring method.
The visual scoring method cannot accurately estimate the total involved surface area in patients with skin diseases, whereas the CIAD method shows accuracy similar to that obtained by the CIAT. In addition, the CIAD is much less time-consuming than CIAT. Therefore, the CIAD can be recommended for use as an objective and accurate method for measurement of surface area afflicted with skin disease.
112.5.3 ASSESSMENT OF INVOLVED AREA IN PSORIASIS (COMPUTER IMAGE ANALYSIS WITH ANALOG CAMERA) The areas of psoriatic lesions were measured by the visual scoring method and computerized image analysis with
analog camera, by dividing the surface area of the body into eight parts: the head and anterior and posterior parts of the trunk, upper limbs, and lower limbs. The percentages of the involved areas measured by the visual scoring method were significantly higher than those obtained from the computerized image analysis with analog camera on the trunk, upper and lower limbs, and total involved body area (Table 112.3). The measurement of the involved area in the case of psoriasis showed the necessity for an objective assessment that could overcome the difference between the observers, which was statistically significant in this study. On the other hand, the measurements of the head area were not statistically different, showing that it was difficult to observe the involved area present on the head and to accurately estimate these
964
Handbook of Non-Invasive Methods and the Skin, Second Edition
25 CIAT 20 CIAD
%
15
10
5
0
Trunk Arm ∗Compared by student’s paired t-test
Leg∗
FIGURE 112.8 Comparisons of estimates of involved area in each anatomic region between the computer image analysis of tracing (CIAT) and the computer image analysis with digital camera (CIAD).
TABLE 112.3 The Average Area Percentages of Psoriatic Lesions in Four Body Regions Visual Scoring (%) Head and neck Upper extremity Trunk Lower extremity Total corrected %b
1.00 8.25 11.33 11.58 9.78
± ± ± ± ±
0.89 3.99 5.95 4.78 3.39
Image Analysis with Analog Camera (%) 0.28 2.79 4.45 5.52 4.13
± ± ± ± ±
0.27 1.78 3.01 3.95 2.50
p Valuea >0.05 <0.05 <0.05 <0.05 <0.05
a
Compared by Wilcoxon signed rank test. Total corrected %: 0.1AH + 0.3AT + 0.2AU + 0.4AL (A = area, H = head, T = trunk, U = upper extremity, L = lower extremity). b
lesions, because of the smaller size of the lesions.5 In addition, we could measure the involved surface areas of the patients with psoriasis whose lesional areas were less than 4% of the total surface area using the image analysis system, but not with the visual scoring method. These results suggest that more accurate and objective data can be obtained from the computerized image analysis with the analog camera system than with the visual scoring method.
112.6 FUTURE, POTENTIAL LIMITATIONS, AND PITFALLS There are some limitations in the computer-based methods for measured surface area, which need to be improved in the near future. First, the image analysis system was not developed specifically for this type of work; therefore, many operations are excessively time-consuming,
including the measurement of the area of the calibration bar, the area of each body region, and the involved area in square centimeters, and the conversion of the area in square centimeters into a percentage of the total body area. In the future, the method should be faster because the photographs will be digitized faster and the computer image analysis system. Second, the three-dimensional shape of the body makes this assessment difficult to perform with the two-dimensional schematic figure outlines. Third, assessing the extent of the disease in patients with dark skin may be difficult, as erythema is less apparent in darker-skinned persons than in white-skinned persons. Fourth, it is important to assess functional limitations, e.g., the involvement of the hands and feet, which can severely limit activity. However, the computer-based methods for measured surface area are not able to quantitatively assess symptoms such as locations and disabilities because their measurement has not been standardized. Lastly, objective
Instrumental and Computer-Based Methods for Measurement of Surface Area Afflicted with Disease
methods of recording disease activity in the case of cutaneous disorders include the measurement of skin function and physiological properties. The ability to assess diseases by evaluating the area of involvement is one of the most important aspects of these measurement techniques. However, the severity of the skin lesions should also be taken into consideration by including a measure of the skin color change (erythema, pigment) and assessment of skin surface contours.23 Recently, digital cameras have been improved to the extent that they now have the potential to produce highquality pictures. Compared to an analog camera, the digital camera has the advantage of allowing the image to be immediately checked after taking the photograph. In addition, it is easy to obtain a magnified image, thus allowing the extent of the involved area to be assessed precisely. The focal length of the lens has become much shorter in digital cameras, so that they have a larger depth of focus than analog cameras. In addition, they have improved color-balancing capabilities when faced with various illuminating conditions. Some cameras even have an antiblurring function, which allows sharp pictures to be taken even at slower shutter speeds. Such functionalities are precisely what are needed to take clear clinical photographs. Therefore, the use of digital cameras may be the ideal and most practical solution.
REFERENCES 1. Buccheri, L., Katchen, B.R., Karter, A.J., et al., Acitretin therapy is effective for psoriasis associated with human immunodeficiency virus infection, Arch. Dermatol., 133, 711, 1997. 2. Townsend, B.L., Cohen, P.R., Duvic, M., Zidovudine for the treatment of HIV-negative patients with psoriasis: a pilot study, J. Am. Acad. Dermatol., 32, 994, 1995. 3. Fredriksson, T., Pettersson, U., Severe psoriasis: oral therapy with a new retinoid, Dermatologica, 157, 238, 1978. 4. European Task Force on Atopic Dermatitis, Severity scoring of atopic dermatitis: the SCORAD index, Dermatology, 186, 23, 1993. 5. Barke, B.M., Cunliffe, W.J., The assessment of acne vulgaris, Br. J. Dermatol., 111, 83, 1984. 6. Steven, S.R., Ke, M.S., Parry, E.J., Mark, J., Cooper, K.D., Quantifying skin disease burden in mycosis fungoides type cutaneous T-cell lymphoma, Arch. Dermatol., 138, 42, 2002. 7. Goudfield, M., Measuring the activity of disease in cutaneous LE, Br. J. Dermatol., 142, 399, 2003.
965
8. Marks, R., Barton, S.P., Shuttleworth, D., et al., Assessment of disease progress in psoriasis, Arch. Dermatol., 125, 235, 1989. 9. Ramsay, B., Lawrence, C.M., Measurement of involved surface area in patients with psoriasis, Br. J. Dermatol., 124, 564, 1991. 10. Tiling-Grosse, S., Rees, J., Assessment of area of involvement in skin disease: a study using schematic figure outlines, Br. J. Dermatol., 128, 69, 1993. 11. Wallace, A.B., The exposure treatment of burns, Lancet, 1, 501, 1951. 12. Long, C.C., Averill, R.W., Finlay, A.Y., The rule of hand: 4 hand measures = 2FTU = 1g, Arch. Dermatol., 128, 1130, 1992. 13. Yune, Y.M., Park, S.Y., Oh, C.H., Objective assessment of involved surface area in patients with psoriasis, Skin Res. Technol., 9, 339, 2003. 14. Savolainen, L., Kontinen, J., Roning, J., Oikarinen, A., Application of machine vision to assess involved surface in patients with psoriasis, Br. J. Dermatol., 137, 395, 1997. 15. Finlay, A.Y., Measurement of disease activity and outcome in atopic dermatitis, Br. J. Dermatol., 135, 509, 1996. 16. Hanifin, J.M., Rajka, G., Diagnostic features of atopic dermatitis, Acta. Derm. Venereol. (Stockh.), Suppl. 92, 44, 1980. 17. Costa, C., Rilliet, A., Nicolet, M., et al., Scoring atopic dermatitis: the simpler the better, Acta. Derm. Venereol. (Stockh.), 69, 41, 1989. 18. Berth-Jones, J., Six area, six sign atopic dermatitis (SASSAD) severity score, Br. J. Dermatol., 135 (Suppl. 48), 25, 1996. 19. Bahmer, F.A., ADASI score: atopic dermatitis area and severity index, Acta Derm. Venereol., 176, 32, 1992. 20. Emerson, R.M., Charman, C.R., William, M.C., The Nottingham eczema severity score, Br. J. Dermatol., 142, 288, 2000. 21. Charman, C.R., Venn, A.J., William, H.C., Measurement of body surface area involvement in atopic dermatitis: an impossible task? Br. J. Dermatol., 141, 109, 1999. 22. Moon, J.S., Yi, G.J., Oh, C.H., Lee, M.H., Lee, Y.S., Kim, M.G., A new technique for three-dimensional measurements of skin surface contours: evaluation of skin surface contours according to the ageing process using a stereo image optical topometer, Physiol. Meas., 23, 247, 2003. 23. Park, S.Y., Oh, H.S., Kim, D.J., Kim, I.L., Moon, J.S., Kim, M.K., Oh, C.H., Quantitative evaluation of severity in psoriatic lesions using 3-dimesional morphometry, Exp. Dermatol., in press.
Evaluation of Wheal113 Instrumental and-Flare Reactions D. Van Neste Skinterface, Tournai, Belgium
CONTENTS 113.1 Introduction ..........................................................................................................................................................967 113.2 Clinical Approach and Skin Color.......................................................................................................................968 113.3 Optical Laser Doppler Methods...........................................................................................................................969 113.4 Ultrasound Echographic Methods........................................................................................................................970 113.5 Conclusion............................................................................................................................................................971 Acknowledgments ...........................................................................................................................................................971 References .......................................................................................................................................................................971
113.1 INTRODUCTION Everyone has probably experienced a vanishing redness after stroking the skin. Nettle stings are probably among everyone’s most itchy and unpleasant experiences; they develop along with vanishing erythema (flare) around the central swelling (wheal) within the first minutes after contact with this plant. The cutaneous response of wheal and flare is known as a physiological triple vascular response, described earlier last century by Sir Thomas Lewis.1 The response depends on the presence of two functionally normal superficial dermal networks: nervous and vascular. After stimulation, nerve endings are activated or depolarized. While sending information to the brain, i.e., transmitting the depolarizing signal to the more central parts of the nervous system,2 there is also a depolarization in the more distal nerve endings, i.e., those that belong to the same axon and that are located more distally from the stimulated skin site. This phenomenon is known as the axon reflex and is due to antidromic activation. These associated nerve endings release mediators at the periphery.3 This will in turn influence blood vessel function in the immediate surroundings of the initial stimulus. The total area so involved is a function of the intensity of the triggering stimulus or the reactivity of the skin site of a given subject. If the nerve circuitry is interrupted, this spreading vascular response is not present. In case of gradual exhaustion from mediator content at the nerve endings — for example, by repeated stimulation — the intensity of the skin responses will
gradually decrease.4 The emotional status of the individual may also play a modulator role.5 In very sensitive subjects, these nerve–vessel interactions may be abnormally amplified beyond the usual redness. The activation of the microscopic blood vessels (capillaries are as thin as a hair; same Latin origin) in those subjects results in leakage. Fluid extravasation results in a wheal formation in the middle of the erythematous reaction. This type of wheal-and-flare reaction occurs under various clinical disorders known as dermographism, urticaria, and allergic (type 1 or immediate) reactions. Whealand-flare reactions are also part of physiological skin responses to insect bites, such as skin reactions to the saliva injected by mosquitoes.6 Histamine is one of the mediators intervening in wheal-and-flare skin reactions, especially at the site of the stimulus-induced wheal.7 The total amount of agonist — e.g., histamine or histamine-releasing compounds — delivered in a certain time in a certain area is a critical factor for pharmacodynamic studies. When histamine is administered to the skin by various methods, including pricking the skin, various intradermal administration modalities, or iontophoresis, the skin response is characterized by the development of a vanishing wheal-and-flare reaction. The field of skin disorders and skin pharmacodynamics is beyond the scope of this chapter. Here, we will describe and illustrate the approaches with simple and practical examples. Images and figures are results that we obtained with various noninvasive methods while investigating wheal-and-flare reactions. The basic principles of 967
968
Handbook of Non-Invasive Methods and the Skin, Second Edition
those methods will be briefly mentioned here, and we refer the reader to other chapters in this volume for a more extensive and in-depth review. Those who are more interested in tracing our work in this field may read the following papers: intradermal administration of histamine,8–14 specificity of histamine-induced laser Doppler flowmetry (LDF) signals,15 histamine iontophoresis,9,16 and agonist–antagonist interactions.17–20 We and others have investigated alternative modalities for agonist administration and dynamic investigation of skin responses with instrumental methods: the least invasive appears to be histamine iontophoresis,21 as will be illustrated later in this chapter. These skin challenging methods were most often developed for investigative purposes and have been used for testing comparatively the efficacy and skin pharmacodynamic effects of antihistamines. One preliminary remark for any clinical trial: please keep in mind during the planning of clinical investigations that the skin is a reactive organ. The administration modality of the stimulus (a chemical compound or a physical agent) can modify the skin vascular response, and a lot of energy should always be spent in control experiments (effects of procedures, administration modalities of the stimulus, agonist concentration, pH and control solutions or any other controls for site, gender, etc.). A second preliminary remark comes from the technological perspective: one should also be cautious in correlation studies between methods because some methods may be — while others are not — sensitive to apparently minimal technical errors (such as a small, almost unnoticed microbleeding and crust formation after skin pricking, the effect of a slight angle in capturing images or data, modifications of skin surface geometry in the presence of the wheal, etc.).
113.2 CLINICAL APPROACH AND SKIN COLOR Visual inspection of the skin before, during, and after application of the stimulus is a nonsophisticated observation method. A slight pressure on the skin will be sufficient to clear the miniature blood vessels from their content. As red blood cell content is reduced, a relative skin whitening is observed, more so in the wheal than in the flare area. Outlining the papule and the erythematous area at certain times after application of the stimulus is a means for measuring the dynamics of wheal-and-flare reactions. An alternative is to observe the skin through transparencies, on which the outlines of the wheal and the flare areas are drawn (Figure 113.1). Measuring the area of the wheal and the area of the flare (including the wheal, i.e., wheal-and-flare area) on those transparencies is a convenient method for dynamic studies. Simple as it is, the method requires a proper selection of materials and methods
#1 Histamine control
#2 Histamine control
FIGURE 113.1 Clinical record of wheal-and-flare reactions. Two subjects participated in an assay where histamine and a control solution were administered by skin pricking. In this figure, outlines on the source documents were partly reprocessed for better illustration, e.g., blue outline of edges of erythematous flare and red filling for contrasting wheal amidst the flare area. The outlines of the flare in subject 1 are more corrugated than those in subject 2 as a result of a between-observer difference. The wheal appears fragmented in subject 2 for the same reasons. Subject 1 reacts more readily to the skin administration (see size of control) than subject 2.
(type of transparency, skin markers, definition of outlines, etc.) and standardized procedures for recording. The area where erythema is continuous or homogeneous may be of great significance (see between-subject and betweenobserver differences for skin erythema outlining, Figure 113.1). Indeed, as the distance increases from the triggering insult, less and less axons are recruited and involved for neuromediator release at the periphery. As a consequence of the microanatomical distribution of skin nerve endings, a mottled erythema is observed at the periphery. Chromametry appears as a rigorous and standardized method where light is delivered from a specific source with a known spectrum to a well-defined area. A white control reference surface serves to calibrate the instrument. The amount of light that is reflected by the skin is described as a point in absolute values in the three-dimensional color space (e.g., La*b*). This point constitutes a record of the color of the skin site at a given time. After skin stimulation, the color distribution will principally move toward red along the green-red axis (higher a* values) in the flare area.22 In the wheal there is a slightly more complex event. Whitening of the skin is due to swelling of the skin. This modifies the spatial position of the skin, and the surface of the wheal becomes closer to the measuring device. A typical example of immediate erythematous reactions after stinging by two species of mosquitoes is shown at the early and late recording times along the a* axis in Figure 113.2. In this case the wheal-and-flare reaction extends
Instrumental Evaluation of Wheal-and-Flare Reactions
969
6 5
Aedes Anopheles
∂a∗
4 3 2 1 0 t0
t5
t15
t30
t45
t60
t90
t120
t180
t1440
FIGURE 113.2 Chromametry of skin reactions induced by mosquitoes. Two species of mosquitoes were tested in one session in a group of healthy volunteers. The increased (average of differences from baseline t0 value) a* values reflect increased redness as time elapses after stinging (t5 to t1440 minutes). The skin reaction toward Aedes was significantly more intense, as well as early as at later times, compared with the response to Anopheles.
into a delayed response and reflects intervention of other mechanisms — due to the chemical composition of the saliva (intermosquito differences, with Aedes being more severe than Anopheles) and possibly immunologic (human initial exposure or reexposure and sensitization) in nature, i.e., much more than just wheal and flare. The weight of the instrument and the need for physical contact between the instrument and the skin may be regarded as minor drawbacks in very contrasted situations with intense redness. These disadvantages gain in importance when weak skin reactions are being observed because the slightly variable pressure may alter the skin redness. The use of clinical photography is also possible for recording purposes.22 Categories of mild, moderate, and severe reactions can be defined, but great care must be taken when numbers (surfaces, color density, etc.) are being generated from clinical photographs (see other chapters in this book). Similarly, skin temperature changes — as evidenced by contact or telethermography — have been used23,24 for documentation purposes rather than detailed measurements. Clinical, photographic, and chromametric differences in skin color (between individuals, sites, and at various times in the same individual) are easily established on a pale background. As the presence of natural pigment, such as skin in black African subjects (Figure 113.3), may interfere with the readings, other instrumental methods may appear more appropriate, as will be described in the next sections.
113.3 OPTICAL LASER DOPPLER METHODS Laser Doppler measurements have been used to characterize skin vascular responses. Laser light is emitted at a certain frequency at a low energetic level for safety reasons. It is delivered into the skin with optical fibers. As
FIGURE 113.3 Experimental setup for histamine iontophoresis. Electric current delivery and time are as important variables as the histamine concentration. In this particular experiment microvascular changes were recorded with laser Doppler flowmetry at the site of agonist penetration and during delivery of current.
the light enters the skin it bounces on various interfaces, some of which are in motion, some not. The frequency of the reflected laser light is shifted only as it bounces back on mobile elements (e.g., blood cells in the most superficial blood vessels) contained in a very small volume in the superficial dermis.25 With the conventional method (LDF), one or a combination of several probe holders are fixed to the skin at selected distances from the stimulus. Some probe holders are redesigned as a function of the study protocol (Figure 113.3) in order to monitor changes in skin blood motion. One measures activity changes inside the blood vessels at the site of agonist delivery and during the process of agonist administration or at distance from it in order to measure the expansion of the erythematous wave (see also Figure 113.4). Laser Doppler imaging (LDI)26 has several advantages over the conventional method. LDI covers a bigger area
970
Handbook of Non-Invasive Methods and the Skin, Second Edition
1–00:00:00
2–00:01:04
3–00:01:35
4–00:02:07
5–00:02:39
6–00:03:10
7–00:03:42
8–00:04:14
9–00:04:45
10–00:05:17
11–00:05:48
12–00:06:20
13–00:06:52
14–00:07:23
15–00:07:55
16–00:08:27
17–00:08:58
18–00:09:30
19–00:10:02
20–00:10:33
1–00:00:00
2–00:00:53
3–00:01:25
4–00:01:57
5–00:02:28
6–00:03:00
7–00:03:32
8–00:04:03
9–00:04:35
10–00:05:06
11–00:05:38
12–00:06:10
13–00:06:41
14–00:07:13
15–00:07:45
16–00:08:16
17–00:08:48
18–00:09:19
19–00:09:51
20–00:10:23
A: Control
B: Histamine
FIGURE 113.4 Laser Doppler imaging and histamine iontophoresis. Each image is a record of the skin before agonist application (left of first row) and after control solution or histamine iontophoresis. One record is made almost every 30 seconds (upper panel A, control; lower panel B: histamine). Using laser Doppler imaging in an experiment similar to the one shown in Figure 113.3, one records a rapid increase in blood flow at the site of electrical current delivery (A, control), and response already fades away after the fifth minute. This response also occurs after histamine delivery, but stays on for a longer time. In the very center of the images, as from the fifth minute, one notices a decrease of the skin perfusion (darker colors), indicating wheal formation. The wheal expands until the eighth minute and afterwards is again gradually masked by the increased skin perfusion signal.
than the laser Doppler conventional method. As there is no direct contact with the skin, one records changes over time during agonist delivery (same method as in Figure 113.3), as well as the wheal and the flare areas. Such an experiment clearly supports our initial observations (performed with the single LDF probe) of an acute increase of blood flow followed by a relative inhibition,
as reflected by a reduction at the site at the wheal formation, as compared with the higher LDF values recorded in the flare. This observation was considered nonspecific by many reviewers, and the publication of our papers was delayed or eventually refused. The expansion of the dermal tissues initiated by the extravasation of fluid from the blood vessels will increase intratissular pressure, counteract vessel dilation, and eventually increase the space between vessels so that the echo returning to the measuring probe will result in a decreased laser Doppler signal (Figure 113.4). Noninvasive dermal administration modalities, e.g., iontophoresis, have also been calibrated by dosing the amount of histamine that could be collected in the dermis by microdialysis (M. Church and D. Van Neste, Skinterface, unpublished data). Also, histamine is found in significant amounts in the immediate vicinity of the wheal, and almost none is found in the flare.7 Hence, with LDI there was a clear indication that the dumping of skin perfusion was reflecting the wheal and was directly related to the effect of the agonist. Accordingly, our observations — initially formulated in the late 1980s with dermal pharmacodynamic experiments, e.g., histamine–antihistamine assays in vivo in humans — were finally confirmed. Those who thought that events recorded at the site of histamine administration were only artifacts were wrong. Hence, advising measurements at a distance from the histamine administration site is easier and may appear to be less subject to other influences, such as the microtrauma associated with the administration modalities (injection, pricking, or electrical current), but this is only indirect evidence of histamine-related skin reactions and, as a consequence, indirect evidence of antihistamine activity. As such, we and others pursued to track specific measurements at the site of histamine administration using other approaches, such as ultrasound.
113.4 ULTRASOUND ECHOGRAPHIC METHODS Ultrasound two-dimensional27,28 (Figure 113.5) and threedimensional (personal files, Figure 113.6) imaging systems have both been used for measuring wheal-and-flare reactions. The use of three-dimensional probes is more convenient because the scanned area is larger and twodimensional together with the three-dimensional images are recorded in one session (Figure 113.6). The exact location of the administration site and skin measurement is of primary importance for the characterization of these skin responses. The complex expansion of the wheal depends on the mode of delivery of the agonist and also the viscoelastic properties of the skin test area. Gender of the test person and body site must be taken into account as factors influencing the responses to the skin challenge.
Instrumental Evaluation of Wheal-and-Flare Reactions
971
ACKNOWLEDGMENTS The author acknowledges the technical and scientific support of Th. Leroy (Skinterface, Tournai, Belgium) and S. Conil (Spiroon, Tournai, Belgium) during the preparation of this manuscript.
REFERENCES 0 min.
5 min.
7 min.
9 min.
13 min.
FIGURE 113.5 Ultrasound time course study of whealing. Histamine release is followed by decreased density and echo in the superficial dermis (between 5 and 13 minutes). The surface of the skin is lifted up by the fluid extravasation, and the dermal structures appear pushed aside and downward: tissue is clearly subject to a three-dimensional redistribution according to local tissue viscoelastic properties.
z
z
1
z
2
x
3
4
FIGURE 113.6 Three-dimensional imaging of the wheal. Recombination of echogenic slices (one section every 100 μm) shows a virtual wheal in a three-dimensional space. Measurement of the expansion, i.e., dermal volume, is another way of exploring wheal-and-flare reactions and reflects subtle regional variations of skin reactivity after endogenous histamine release.
113.5 CONCLUSION In conclusion, the range from clinical approach to the most sophisticated instrumental approach is very wide. The interconnection between clinical observation and instrumental methods must always remain, as each approach has its advantages and disadvantages or limitations. If well understood and integrated into the study plan, a combined clinical-instrumental approach can clearly help the investigator in measuring objectively the wheal-and-flare response. When the test is performed in the context of a clinicalpharmacological study, dose range effects, appropriate controls, and time range effect studies must be performed. In comparative efficacy trials, the agonist–antagonist equilibrium in the skin and the affinity of the agonists to their receptors must be studied in great detail before concluding that one agent is superior to another. We advocated for some time already that skin pharmacodynamics may be regarded as a more relevant indicator of clinical efficacy than drug concentration profiles in the peripheral blood, i.e., classical pharmacokinetics.
1. Lewis T, Grant RT. Vascular reactions to injury. Heart 2:209–265, 1924. 2. Bernstein JE. Neuropeptides and the skin. In Biochemistry and Physiology of the Skin, Goldsmith LA, Ed. Oxford University Press, New York, 1983, pp. 1217–1233. 3. Hagermark O, Hokfelt T, Pernow B. Flare and itch induced by substance P in human skin. J Invest Dermatol 71:233–225, 1978. 4. Drummond PD. Attenuation of axon reflexes to compound 48/80 after repeated iontophoresis of compound 48/80 in skin of the human forearm. Skin Pharmacol Appl Skin Physiol 16:263–270, 2003. 5. Zachariae R, Jorgensen MM, Egekvist H, Bjerring P. Skin reactions to histamine of healthy subjects after hypnotically induced emotions of sadness, anger, and happiness. Allergy 56:734–740, 2001. 6. Geveaux C, Leroy T, Van Neste D. A comparative study of skin reaction after mosquito bite with either Aedes aegypti or Anopheles stephensi species. Nouv Dermatol 18:79–82, 1999. 7. Petersen LJ, Church MK, Skov PS. Histamine is released in the wheal but not the flare following challenge of human skin in vivo: a microdialysis study. Clin Exp Allergy 27:284–295, 1997. 8. Van Neste D. Skin response to histamine: reproducibility study of the dry skin prick test method and of the evaluation of microvascular changes with laser Doppler flowmetry. Acta Derm Venereol (Stockh) 71:25–28, 1991. 9. Van Neste D, Leroy T, De Brouwer B, Rihoux J-P. Histamine-induced skin reactions using iontophoresis and H1-blockade. Inflamm Res 45:S48–S49, 1996. 10. Thysman S, Jadoul A, Leroy T, Van Neste D, Preat V. Laser Doppler evaluation of skin reaction in volunteers after histamine iontophoresis. J Control Release 36:215–219, 1995. 11. Van Neste D, De Brouwer B, Tasset C, Valentin B, Coulie P. Laser Doppler flowmetry on histamineinduced skin response. Eur J Dermatol 5:537–541, 1995. 12. Van Neste D, Rihoux J-P. Dynamics of the skin blood flow response to histamine. Dermatology 186:281–283, 1993. 13. Van Neste D. Skin response to histamine dry skin prick test: influence of duration of the skin prick on clinical parameters and on skin blood flow monitoring. J Dermatol Sci 1:435–440, 1990.
972
Handbook of Non-Invasive Methods and the Skin, Second Edition
14. Van Neste D. Evaluation of histamine induced itch with the histamine dry skin prick test (PhazetTM). J Head Neck Pathol 9:232–234, 1990. 15. Leroy T, Jadoul A, Thysman S, Preat V, Van Neste D. In vivo laser Doppler evaluation of specific histamineagonist effects in the skin. J Eur Acad Dermatol Venerol 1995 (abstract). 16. Leroy T, Van Neste D. Reproducibility of skin responses induced by histamine iontophoresis. J Eur Acad Dermatol 1996 (abstract). 17. Rihoux J-P, Van Neste D. Quantitative time course study of the skin response to histamine and the effect of H1 blockers. Dermatologica 179:129–134, 1989. 18. Van Neste D, Coussement C, Ghys L, Rihoux J-P. Agonist-antagonist interactions in the skin: comparison of effects of loratidine and cetirizine on skin vascular responses to prick tests with histamine and substances P. J Dermatol Sci 4:172–179, 1992. 19. Van Neste D, Ghys L, Antoine JL, Rihoux J-P. Pharmacological modulation by cetirizine and atropine of the histamine- and methacholine-induced wheals and flares in human skin. Skin Pharmacol 2:93–102, 1989. 20. Van Neste D, Rihoux JP. Inhibition of the cutaneous response to histamine by H1-blocking agents. Skin Pharmacol 1:192–199, 1988. 21. Magerl W, Westerman RA, Möhner B, Handwerker HO. Properties of transdermal histamine iontophoresis: differential effects of season, gender, and body region. J Invest Dermatol 94:347–352, 1990.
22. Boysen L, Sorensen P, Larsen M, Serup J, Kristensen F. Evaluation of skin erythema by use of chromametry and image analysis of digital photographs after intradermal administration of histamine in dogs. Am J Vet Res 63:565–569, 2002. 23. Stüttgen G, Eilers J. Reflex heating of the skin and telethermography. Arch Dermatol Res 272:301–310, 1982. 24. Van Neste D. Etude comparative de deux méthodes d'évaluation non invasives de la réponse cutanée à l'histamine. Téléthermographie versus laser doppler velocimétrie. Nouv Dermatol 9:32–35, 1990. 25. Serup J, Staberg B. Quantification of weal reactions with laser Doppler flowmetry. Comparative blood flow measurements of the oedematous centre and the perilesional flare of skin-prick histamine weals. Allergy 40:233–227, 1985. 26. Mattsson U, Krogstad A-L, Pegenius G, Elam M, Jontell M. Assessment of erythema and skin perfusion by digital image analysis and scanner laser Doppler. Skin Res Technol 3:53–59, 1997. 27. Seidenari S, Giusti G, Bertoni L, Magnoni C, Pellacani G. Thickness and echogenicity of the skin in children as assessed by 20-MHz ultrasound. Dermatology 201:218–222, 2000. 28. Serup J. Diameter, thickness, area, and volume of skinprick histamine weals. Measurement of skin thickness by 15 MHz A-mode ultrasound. Allergy 39:359–364, 1984.
Evaluation of 114 Instrumental Occluded Patch Test Reactions Stefania Seidenari, Francesca Giusti, and Giovanni Pellacani Department of Dermatology, University of Modena and Reggio Emilia, Modena, Italy
CONTENTS 114.1 Introduction ..........................................................................................................................................................973 114.2 Transepidermal Water Loss..................................................................................................................................973 114.3 Infrared Thermography ........................................................................................................................................974 114.4 Colorimetric Methods ..........................................................................................................................................974 114.5 Cutaneous Blood Flow Measurement..................................................................................................................975 114.6 Ultrasound ............................................................................................................................................................976 References .......................................................................................................................................................................977
114.1 INTRODUCTION Patch testing is a widely used procedure in both clinical and experimental settings. Patch tests are routinely employed in order to diagnose allergic contact dermatitis (ACD), and are evaluated according to internationally agreed rules using a clinical rating scale based on the degree of erythema, induration, and the presence or absence of vesicles in daily practice. However, visual assessment and palpation of patch test responses lack in objectivity and reproducibility and are unsuitable in experimental studies investigating the sensitizing properties of chemicals or the efficacy of drugs on ACD, when a comparison is needed. Patch testing has also been widely used in the study of skin irritation induced by numerous chemicals. Repetitive patch tests or single 24-h patch test with sodium lauryl sulfate (SLS) are validated tools in experimental studies on irritant reactions. Since skin damage caused by repeated irritant insults may not be apparent to the eye and may not cause clinical symptoms for a prolonged period, there is a need for sensitive and reliable methods to study irritant responses, especially when they are weak. A number of instrumental techniques have been applied to patch test assessment in order to quantify skin responses to allergens and irritants, producing objective and reproducible data. Because of their noninvasive approach, these methods allow follow-up examinations. Moreover, they provide continuous data for grading disease intensity that are suitable for statistical analysis and dose–responses studies.
Bioengineering methods cannot replace the clinician’s eye and finger in routine patch test assessment; however, their introduction into the evaluation of irritant and allergic responses has opened up a new field of investigation in experimental dermatology. When evaluating the usefulness of each technique, its application area together with variables and limitations must be considered. In fact, each method highlights only one aspect of the inflammatory reaction: the evaporimeter evaluates barrier disruption, the laser Doppler flowmeter and the colorimeter quantify the component of the allergic response due to increase in blood flow, and ultrasound measures edema, which is secondary to vasodilation, increase in blood vessel permeability, and extravasation of water within the tissue. An improved characterization of skin reactions and understanding of their kinetics can be achieved by the combined use of a number of noninvasive techniques.
114.2 TRANSEPIDERMAL WATER LOSS Transepidermal water loss (TEWL) refers to the rate at which water migrates from the dermis through the layers of the stratum corneum to the external environment. The Evaporimeter records the total evaporation from the skin, and since sweating should be suppressed during measurements, the term TEWL is generally accepted for the recording of this instrument. When evaluating the irritant properties of substances by patch testing, TEWL measurement has been demonstrated to enable an objective scoring. A good correlation between physical or chemical damage of 973
974
Handbook of Non-Invasive Methods and the Skin, Second Edition
the skin barrier and increase in TEWL values has been demonstrated; however, besides a great interindividual variability, the elicited responses, as evaluated by TEWL, vary from irritant to irritant.1–10 The best correlation between TEWL values and the irritant response was found for SLS: a positive dose–response relationship to SLS has been demonstrated in many studies.4,5,11–14 So far, the Standardization Group of the European Society of Contact Dermatitis has recommended the measurement of TEWL as the best method for evaluating the effect of SLS patch testing.15 Comparing 24-h patch test reactions with different irritant substances, we found that modifications of TEWL are less pronounced for 4% hydrochloric and 40% nonanoic acid in comparison with 3% SLS.9 Whereas SLS patch test sites showed an increase in TEWL as the primary event, topical D vitamin and all-trans-retinoic acid application for 48 h did not affect TEWL at day 3 assessment.10 A delayed increase in TEWL after application of all-trans-retinoid acid 24-h patch tests was described.16 It is likely that in this case the deficient skin barrier is secondary, resulting from the underlying inflammatory process and concomitant repair mechanism. Vitamin D, vitamin D analogs, and all-transretinoic acid belong to the noncorrosive type of irritants, defined as low degree irritants, but with no increase in TEWL, while SLS, inducing impairment of skin barrier function and an increase in TEWL values, is a corrosivetype irritant.17 TEWL assessment is particularly useful when evaluating subjects in those cases where a chemical can induce slight damage, which cannot be appreciated clinically. In a study performed on 63 patients affected by eczema undergoing a 5% SLS patch test on the volar aspect of the forearm we observed significant modifications in TEWL values, even at clinically negative patch test sites.6 By investigating the irritant effects of three different metalworking fluids in 10 healthy subjects, TEWL measurement provided reliable information about skin barrier function, before barrier disruption became clinically evident. 7 Furthermore, TEWL variation in patch test responses to the three metalworking fluids significantly differed from one other, in spite of their similar alkalinity. Whereas barrier impairment in irritant contact dermatitis has been thoroughly studied, data about TEWL at allergic patch test sites are scarce. Van der Valk et al.18 assessed TEWL in nickel-allergic patients undergoing 48h patch testing with nickel sulfate and irritants. An increased TEWL was observed only in strong reactions to nickel. Investigating patch test responses to nickel and SLS in 12 nickel-sensitized patients by means of evaporimetry, Serup and Staberg11 found that after 48 h only strong reactions resulted in increased TEWL. These data were confirmed by us, evaluating skin responses (+ or ++) to nickel sulfate at different concentrations in 12 nickelsensitive women by means of evaporimetry and other
instrumental methods: no concentration- and time-dependent modifications of TEWL were detectable.19 On the contrary, both colorimetry and 20-MHz ultrasound provide numerical data that increase proportionally to the intensity of the patch test response. In a study performed on 18 patients with alopecia areata aiming at evaluating their sensitization status before and 4 months after diphenylcyclopropenone (DPCP) contact immunotherapy, TEWL values on DPCP patch test areas proved significantly lower after 4 months with respect to the beginning of immunotherapy, but they correlated to visual scores worse than laser Doppler flowmetry ones.20 In conclusion, evaporimetry does not represent a suitable method for measuring the intensity of the patch test response to allergens. The impairment of the barrier in this case is probably secondary to the inflammatory process, and it is evident only for strong reactions. In the studies reported above, measurements were performed up to a maximum of 72 h after the application of the contact sensitizers and no mention was made of barrier status during recovery.
114.3 INFRARED THERMOGRAPHY Infrared thermography is a noninvasive technique that provides information on the radiative heat loss from the skin surface. It has been employed to assess patch test responses in patients undergoing routine patch testing with a wide range of substances for contact eczema.21 The sites of allergic reactions appear as hot areas when compared to the surrounding normal skin as a result of local vasodilation or increased local metabolic activity. A good correlation between clinical assessment and thermographic parameters, temperature and area of involvement, has been demonstrated, being dependent on the severity of the response. However, the ability of thermography to discriminate between weak allergic responses and equivocal reactions is poor. As regards the thermographic assessment of irritant reactions, no thermal response to SLS could be detected over a range of concentrations and application times.21,22 This may be due to increased TEWL associated with irritation, providing sufficient evaporating cooling to counteract any heat production arising from metabolic activity or vasodilation. Since irritant reactions appear cold by thermography, this method can be used to discriminate between allergic and irritant responses.
114.4 COLORIMETRIC METHODS Increased perfusion of the test area represents the primary response in the positive patch test, although the increase in blood amount in the upper papillary dermis,
Instrumental Evaluation of Occluded Patch Test Reactions
a∗ 12
975
0–30 pixels 15000
10
12000
8
9000
6 6000
4
3000
2 0
+
++
+++
0
+
++
+++
FIGURE 114.1 Colorimetric (a*) and echographic (0 to 30 pixels) values at allergic patch test sites according to the positivity degree (+, ++, and +++ reactions). a* values correspond to erythema, whereas 0 to 30 pixel values express inflammation-induced dermal hyporeflectivity.
corresponding to the appearance of erythema, represents only one substantial aspect of inflammation. Measurement of the erythema index by the Cortex Dermaspectrometer, which is proportional to erythema, has been employed for quantification of eczematous patch test reactions.23–25 This index did not prove to be more sensitive than the visual evaluation of skin responses in 10 allergic patients patch tested with serial dilutions of the appropriate antigen in a study by Quinn et al.24 When comparing the erythema index of 56 allergic reactions to regional controls and 189 patch test negative responses, Jemec and Johansen25 observed that the method correlates well with clinical patch test assessment. However, since factors other than erythema also play a role in the clinical assessment of eczema, its ability to distinguish between the different grades of eczematous reactions was unsatisfactory, as no significant differences were found between ++ and +++ responses and between ?+ and + responses. We confirmed these results by measuring the color parameter a* by the Minolta Chromameter, describing the color range from green to red, in 120 nickel-sensitive patients challenged with 5% nickel sulfate on the flexor site of forearms.26 An increasing trend in a* values according to the increase in clinical scores was observed, but a maximum response was already visible for ++ reactions, reflecting the pathogenetic sequence of the allergic inflammation, showing an initial stage dominated by vasodilation, and a more advanced stage dominated by the formation of edema compressing the vessels and preventing erythema from appearing on the skin surface (Figure 114.1). Therefore, going on these data, colorimetry does not represent a suitable method for the grading of allergic patch test responses of strong intensity, whereas it has been found to provide a suitable supplement to clinical scoring in irritant reactions.27 As regards the evaluation of irritant responses to patch tests, chromametry did not clearly distinguish between irritant and nonirritant substances and was unable to detect an early irritant response.28 In 10 healthy volunteers patch tested with sodium salts of nalkyl sulfates, no direct correlation was observed between
skin irritation and color measurements.29 Unlike TEWL and hydration values, colorimetry did not enable the ranking of n-alkyl sulfates based on irritation potential. A poor correlation between color changes and the index of irritation was found in a study investigating surfactant irritation.30 Colorimetric assessment of irritant reactions induced by dithranol patch tests showed a good correlation between clinical scores for erythema and colorimetric values, whereas for SLS the concordance between clinical judgment and colorimetry was low.27
114.5 CUTANEOUS BLOOD FLOW MEASUREMENT Laser Doppler flowmeter (LDF) is suitable to measure the blood flow in skin capillaries as well in arteriovenous anastomoses. By measuring the cutaneous blood flow at the sites of negative and positive standard patch tests in 11 eczematous patients, Staberg et al.31 observed a fivefold increase in doubtful reactions and a tenfold one in both weak and strong reactions. However, blood flow in + and ++ responses did not differ from each other. This method therefore proved useful to separate negative, doubtful, and positive allergic reactions, whereas it failed in discriminating between weak and strong positive responses. LDF values correlated to visual score at DPCP patch test sites in 18 patients with alopecia areata treated with DPCP contact immunotherapy for 4 months.20 In a study performed on 20 healthy volunteers patch tested on the flexor side of the upper arm with three different irritant substances (SLS, nonanoic acid, and hydrochloric acid), a positive correlation was observed between LDF values and visual assessment of irritant responses.2 No significant differences were found in blood flow between the three irritants. Thus, the method was helpful to objectively quantify and monitor irritant responses, but it was unsuitable for classification of individual irritants, in contrast to TEWL measurement. Laser Doppler imaging (LDI) is a noninvasive technique, which was recently introduced for the mapping of
976
Handbook of Non-Invasive Methods and the Skin, Second Edition
cutaneous blood flow, enabling the visualization of the spatial variation of the perfusion. Being a nontouch technique, it has no mechanical influence on skin perfusion and provides more objective and reproducible data with respect to the unidimensional method. It has been widely used to scan superficial blood perfusion in allergic and irritant patch test responses.24,32–36 Useful data over a wide range of allergen concentrations were best obtained by measurement of erythema or area of reaction using LDI in comparison to other instrumental methods, like TEWL, erythema index, and conventional laser Doppler flowmetry.24 In subjects patch tested with Kathon CG and nickel sulfate, Bjarnason et al.34,35 observed large interindividual variations in the perfusion profiles of subjects tested in identical fashion. Initial reactions were detected earlier with laser Doppler scanning than with the visual assessment technique in both studies.34,35 When investigating skin irritation after application of two different substances, SLS and all-trans-retinoic acid, in a hairless guinea pig model, LDI was able to quantify both the intensity and the extension of irritant skin reactions, with a very high concordance with the subjective clinical scoring.32 Moreover, LDI can be used for the assessment of patch test responses induced by irritants not affecting the cutaneous barrier function, in contrast to TEWL measurement.
114.6 ULTRASOUND Edema and inflammatory infiltration at a positive patch test site, increasing when the reaction becomes stronger, induce skin thickening, which is proportional to the intensity of the skin response and can be measured by unidimensional (A-scanning) ultrasound methods.37 B-scanning echography enables real-time cross-sectional imaging of the skin. The echographic appearance of a positive patch test response is characteristic (Figure 114.2). Increased skin thickness and homogeneity of the tissue, with expansion of the hypoechogenic dermal areas and attenuation of hyperreflecting areas in the lower dermis, have been described in allergic reactions.26,37–43 As the reaction grows more intense, the extension of hypoechogenic dermal areas increases, as evaluated by image analysis procedures. In 120 nickel-sensitive patients patch tested with 5% nickel sulfate on the flexor site of forearms, we observed an increase in 0 to 30 pixel area values, corresponding to inflammation-induced dermal hyporeflectivity, correlating with the clinically assigned positivity degree.26 Hypoechogenicity values were able to distinguish between +, ++, and +++ reactions, and values referring to +++ reactions were almost double those of ++ reactions (Figure 114.1). Regional variations in skin reactivity to allergic patch tests on the forearm were demonstrated by means of 20-MHz ultrasound.38 By applying 24-h patch tests with nickel at different skin sites on the
FIGURE 114.2 Twenty-megahertz ultrasound image of positive allergic patch test reaction: skin thickening and expansion of the hypoechogenic dermal areas.
forearm in 17 sensitive patients, a greater increase both in hyporeflecting dermal areas and in thickness values was demonstrated in skin areas near the wrist crease. Moreover, the echographic method has been used to measure the intensity of a wide range of allergic responses, including those induced by dilutions of allergens producing a subthreshold response.40 In a study performed on 70 nickel-allergic patients challenged with 0.05% nickel sulfate for 24 h on the volar aspect of forearms and evaluated at 24 and 72 h, we observed a significant increase in the 0 to 30 pixel area with respect to baseline values for both doubtful reactions at 72 h and negative reactions at 24 h developing as positive at 72 h (Table 114.1).
TABLE 114.1 Echographic Evaluation of Doubtful and Negative Allergic Reactions to 0.05% Nickel Sulfate in 70 Nickel-Sensitive Patients 0–30 Pixel Values
Control sites (empty chambers) Negative reactions at 24 h, developing positive at 72 h Doubtful allergic reactions at 72 h
Δ 24ha
Δ 72hb
110 890
80 —
—
520
Note: At negative or doubtful patch test sites the increase in extension of the 0 to 30 band area in respect to baseline values is significant (p < 0.05). a b
Difference between 24-h evaluation and baseline. Difference between 72-h evaluation and baseline.
Instrumental Evaluation of Occluded Patch Test Reactions
In conclusion, echography is a suitable method for patch test assessment and for grading of allergic patch test reactions, since correlation with clinical scoring covers the whole range of positivities, including subclinical responses. As regards irritant responses, processing of echographic images enables the assessment of both the inflammatory and epidermal components. Intervals of interest are the 0 to 30 amplitude band, marking the hyporeflecting parts of the dermis, corresponding to edema and inflammatory infiltration, and the 201 to 255 band, evaluating the superficial hyperreflecting part of the skin, corresponding to the epidermis.9,37 Dermal edema and inflammatory infiltration at SLS patch tests appear echographically with a hypoechogenic area, which is subepidermal in the first phase, and spreads to the underlying dermis as the reaction grows in intensity. In comparison to allergic responses, even in strong reactions, the hyperreflecting part of the lower dermis at volar forearm skin does not disappear completely; i.e., the inflammatory process appears more superficial. The decrease of the superficial hyperreflecting band corresponding to the epidermis is characteristic of patch test responses to SLS. Moreover, patch testing with 0.5 to 5% SLS showed that skin thickness and 0 to 30 pixel values increase according to SLS concentration and the intensity of the inflammatory component, while a decrease of 201 to 255 pixel values corresponds to attenuation of the epidermal reflectivity.44 Whereas data calculated by the 0 to 30 band elaboration showed a fair correlation to clinical scoring, 201 to 255 values were invertly related to TEWL. Ultrasound assessment of irritation is particularly useful when a chemical induces a slight damage, which cannot be appreciated clinically. To evaluate subclinical irritant reactions, 63 eczema patients were patch tested with 5% SLS on the volar aspect of the forearm for 30 min.6 Processing of echographic images showed a significant decrease of the superficial hyperreflecting band in 15 subjects, where no visible reactions were present at 24 h. Ultrasound assessment of skin damage induced by SLS patch tests proved helpful in studying skin reactivity variations in different patient groups.42 By patch testing 34 nickel-sensitive patients, 14 of whom were affected by atopic dermatitis, with 0.05% nickel sulfate at untreated and 5% SLS-pretreated sites, we observed that at SLS patch test sites the intensity of the dermal inflammatory response was greater in atopic dermatitis patients than in allergic contact dermatitis ones at 24 h. Skin barrier damage, too, as assessed by evaluation of the superficial reflectivity, was higher in atopics than in nonatopics. These data indicate a higher reactivity to SLS in the atopic dermatitis group than in the allergic contact dermatitis group, and a specific susceptibility of atopic skin to surfactants. Moreover, subsequent to a slight irritant stimulus, such as a 30-
977
min patch test to SLS, an earlier inflammatory response and a more pronounced skin damage are induced in atopics, followed by a more marked allergic reaction. While the inflammatory component of irritant reactions has a homogeneous echographic appearance, epidermal damage caused by diverse irritants shows different patterns, characterized by thickening or thinning of the superficial hyperreflecting band corresponding to the epidermis at different time points.27,45 Thus, 20-MHz sonography associated with image analysis contributes to characterizing patch test responses to different irritant substances, which are distinguished by a variable combination of inflammatory and epidermal aspects.
REFERENCES 1. Berardesca, E. and Maibach, H.I., Bioengineering and the patch test, Contact Derm., 18, 3, 1988. 2. Agner, T. and Serup, J., Skin reactions to irritants assessed by non-invasive bioengineering methods, Contact Derm., 20, 352, 1989. 3. Wilhelm, K.P., Pasche, F., Surber, C., and Maibach, H.I., Sodium hydroxide-induced subclinical irritation. A test for evaluating stratum corneum barrier function, Acta Derm. Venereol. (Stockh.), 70, 463, 1990. 4. Wilhelm, K.P. and Maibach, H.I., Susceptibility to irritant dermatitis induced by sodium lauryl sulfate, J. Am. Acad. Dermatol., 23, 122, 1990. 5. Agner, T. and Serup, J., Sodium lauryl sulphate for irritant patch testing: a dose-response study using bioengineering methods for determination of skin irritation, J. Invest. Dermatol., 95, 543, 1990. 6. Seidenari, S. and Belletti, B., Instrumental evaluation of subclinical irritation induced by sodium lauryl sulfate, Contact Derm., 30, 175, 1994. 7. Hüner, A., Fartasch, M., Hornstein, O.P., and Diepgen, T.L., The irritant effect of different metalworking fluids, Contact Derm., 31, 220, 1994. 8. Rogiers, V., Transepidermal water loss measurements in patch test assessment: the need for standardisation, Curr. Probl. Dermatol., 23, 152, 1995. 9. Seidenari, S., Image processing of 20 MHz B-scan recordings of irritant reactions, Curr. Probl. Dermatol., 23, 169, 1995. 10. Fullerton, A. and Serup, J., Characterization of irritant patch test reactions to topical D vitamins and all-trans retinoic acid in comparison with sodium lauryl sulphate. Evaluation by clinical scoring and multiparametric noninvasive measuring techniques, Br. J. Dermatol., 137, 234, 1997. 11. Serup, J. and Staberg, B., Differentiation of allergic and irritant reactions by transepidermal water loss, Contact Derm., 16, 129, 1987. 12. Van Neste, D. and de Brouwer, B., Monitoring of skin response to sodium lauryl sulphate: clinical scores versus bioengineering methods, Contact Derm., 27, 151, 1992.
978
Handbook of Non-Invasive Methods and the Skin, Second Edition
13. Hinnen, U., Elsner, P., and Burg, G., Assessment of skin irritancy by 2 short tests compared to acute irritation induced by sodium lauryl sulfate, Contact Derm., 33, 236, 1995. 14. Aramaki, J., Effendy, I., Happle, R., Kawana, S., Löffler, C., and Löffler, H., Which bioengineering assay is appropriate for irritant patch testing with sodium lauryl sulfate?, Contact Derm., 45, 286, 2001. 15. Tupker, R.A., Willis, C., Berardesca, E., Lee, S.H., Fartash, M., Agner, T., and Serup, J., Guidelines on sodium lauryl sulfate (SLS) exposure testing, Contact Derm., 37, 53, 1997. 16. Effendy, I., Weltfriend, S., Patil, S., and Maibach, H.I., Differential irritant skin responses to topical retinoic acid and sodium lauryl sulphate: alone and in crossover design, Br. J. Dermatol., 134, 424, 1996. 17. Serup, J., The spectrum of irritancy and application of bioengineering techniques, Curr. Probl. Dermatol., 23, 131, 1995. 18. Van der Valk, P.G.M., Kruis de Vries, M.H., Nater, J.P., Bleumink, E., and de Jong, M.C.J.M., Eczematous (irritant and allergic) reactions of the skin and barrier functions as determined by water vapour loss, Clin. Exp. Dermatol., 10, 185, 1985. 19. Seidenari, S., Di Nardo, A., and Giannetti, A., Valutazione di reazioni a patch test eseguiti con solfato di nichel e laurilsolfato di Na (SLS) mediante sonografia in B scan, colorimetria, evaporimetria e laser Doppler flussimetria, G. Ital. Dermatol. Venereol., 127, 15, 1997. 20. Seo, K.I. and Eun, H.C., Loss of contact sensitization evaluated by laser Doppler blood flowmetry and transepidermal water loss measurement, Contact Derm., 34, 233, 1996. 21. Baillie, A.J., Biagioni, P.A., Forsyth, A., Garioch, J., and McPherson, D., Thermographic assessment of patch-test responses, Br. J. Dermatol. 122, 351, 1990. 22. Agner, T. and Serup, J., Contact thermography for assessment of skin damage due to experimental irritants, Acta Derm. Venereol., 68, 192, 1988. 23. Gawkrodger, D.J., McDonagh, A.J.G., and Wright, A.L., Quantification of allergic and irritant patch test reactions using laser-Doppler flowmetry and erythema index, Contact Derm., 24, 172, 1991. 24. Quinn, A.G., McLelland, J., Essex, T., and Farr, P.M., Quantification of contact allergic inflammation: a comparison of existing methods with a scanning laser Doppler velocimeter, Acta Derm. Venereol. (Stockh.), 73, 21, 1993. 25. Jemec, G.B.E. and Johansen, J.D., Erythema index of clinical patch test reactions, Skin Res. Technol., 1, 26, 1995. 26. Seidenari, S. and Belletti, B., The quantification of patch test responses: a comparison between echographic and colorimetric methods, Acta. Derm. Venereol. (Stockh.), 78, 364, 1998. 27. Schiavi, M.E., Belletti, B., and Seidenari, S., Ultrasound description and quantification of irritant reactions induced by dithranol at different concentrations. A comparison with visual assessment and colorimetric measurements, Contact Derm., 34, 272, 1996.
28. Zuang, V., Rona, C., Archer, G., and Berardesca, E., Detection of skin irritation potential of cosmetics by non-invasive measurements, Skin Pharmacol. Appl. Skin Physiol., 13, 358, 2000. 29. Di Nardo, A., Conti, A., Martini, M., and Seidenari, S., In vivo assessment of n-alkyl-sulfate-induced skin irritation by means of non-invasive methods, Skin Res. Technol., 4, 192, 1998. 30. Gabard, B., Chatelain, E., Bieli, E., and Haas, S., Surfactant irritation: in vivo corneosurfametry and in vivo bioengineering, Skin Res. Technol., 7, 49, 2001. 31. Staberg, B., Klemp, P., and Serup, J., Patch test responses evaluated by cutaneous blood flow measurements, Arch. Dermatol., 120, 741, 1984. 32. Fullerton, A. and Serup, J., Laser Doppler image scanning for assessment of skin irritation, Curr. Probl. Dermatol., 23, 159, 1995. 33. Wårdell, K., Andersson, T., and Anderson, C., Analysis of laser Doppler perfusion images of experimental irritant skin reactions, Skin Res. Technol., 2, 149, 1996. 34. Bjarnason, B. and Fischer, T., Visual and laser Doppler perfusion scanning assessments of Kathon CG patch test reactions, Am. J. Contact Derm., 9, 224, 1998. 35. Bjarnason, B. and Fischer, T., Objective assessment of nickel sulphate patch test reactions with laser Doppler perfusion imaging, Contact Derm., 38, 112, 1998. 36. Bjarnason, B., Flosadóttir, E., and Fischer, T., Objective non-invasive assessment of patch tests with the laser Doppler perfusion scanning technique, Contact Derm., 40, 251, 1999. 37. Serup, J. and Staberg, B. Ultrasound for assessment of allergic and irritant patch test reactions, Contact Derm., 17, 80, 1987. 38. Seidenari, S., Di Nardo, A., Pepe, P., and Giannetti, A., Ultrasound B-scanning with image analysis for assessment of allergic patch test reactions, Contact Derm., 24, 216, 1991. 39. Seidenari, S. and Di Nardo, A., Cutaneous reactivity to allergens at 24-h increases from the antecubital fossa to the wrist: an echographic evaluation by means of a new image analysis system, Contact Derm., 26, 171, 1992. 40. Seidenari, S. and Di Nardo, A., B-scanning evaluation of allergic reactions with binary transformation and image analysis, Acta Derm. Venereol. (Stockh.), 175, 3, 1992. 41. Seidenari, S., Echographic evaluation of subclinical allergic patch test reactions, Contact Derm., 29, 156, 1993. 42. Seidenari, S., Reactivity to nickel sulfate at sodium lauryl sulfate pre-treated skin sites is higher in atopics: AD echographic evaluation by means of image analysis performed on 20 MHz B-scan recordings, Acta Derm. Venereol. (Stockh.), 74, 245, 1994. 43. Seidenari, S., Di Nardo, A., and Giannetti, A., Assessment of topical corticosteroid activity on experimentally induced contact dermatitis: echographic evaluation with binary transformation and image analysis, Skin Pharmacol., 6, 85, 1993.
Instrumental Evaluation of Occluded Patch Test Reactions
44. Seidenari, S. and Di Nardo, A., B scanning evaluation of irritant reactions with binary transformation and image analysis, Acta Derm. Venereol. (Stockh.), Suppl. 175, 9, 1992.
979
45. Seidenari, S., Echographic evaluation with image analysis of irritant reactions induced by nonanoic acid and hydrochloric acid, Contact Derm., 31, 146, 1992.
Sources, Sunlight, and 115 Light Radiation Dosimetry Hans Christian Wulf Department of Dermatology, Bispebjerg Hospital, University of Copenhagen, Copenhagen, Denmark
CONTENTS 115.1 Introduction ..........................................................................................................................................................981 115.2 UV Sources ..........................................................................................................................................................981 115.3 Sunlight.................................................................................................................................................................982 115.4 Dosimetry .............................................................................................................................................................984 115.5 Prediction of UV Dose for Treatment .................................................................................................................985 115.6 Test Equipment for Photosensitivity....................................................................................................................985 115.7 Photodynamic Therapy ........................................................................................................................................985 References .......................................................................................................................................................................988
115.1 INTRODUCTION Ultraviolet radiation is one of the most common treatment modalities in dermatology and is used by many patients as sun exposure to cure or minimize their skin disease. The more reliable artificial UV sources can be used all year, but sunlight is still commonly used not only in the summertime, but also as heliotherapy at the Dead Sea or in Southern Europe (Abels and Kattan-Byron, 1985). Ultraviolet radiation came into medicine around the year 1900 as a treatment for skin tuberculosis (lupus vulgaris), a treatment that was invented by Niels Finsen (Finsen, 1896). He and his staff invented the carbon arc lamp, and shortly after the mercury vapor lamp was invented by Kromayer (1906). These lamp types were used for many decades until the development of the modern UV light tubes (Anderson, 1993). UVB tubes became common from the mid-1960s. From the mid-1970s the UVA part of the spectrum was used together with psoralens in the PUVA treatment, whereas the UVB was used without medication as a direct healing radiation (Parrish et al., 1974). The emission from the tubes depends on the type of fluorescing material coated on the inside of the glass tube, and it is possible to have broadband UVB, narrowband UVB, and broadband UVA or UVA1 sources that have all been shown to be beneficial in different skin diseases. It has been very common to measure the effect of the different UV sources by measurement of the physical
irradiance (W/m2) or as a dose (J/m2). However, this information has no bearing on the effect on the skin, and therefore the doses are now given as biologically weighted doses according to the erythema action spectrum (McKinlay and Diffey, 1987). This has been chosen so that the dose is independent of light source and the erythema reaction in the skin will be identical for identical units. This dose is called the standard erythema dose (SED) (Wulf and Lock-Andersen, 1996; Diffey et al., 1997). It is, however, also important to realize that a person’s skin color will be of importance for how many SEDs that can be tolerated without getting burnt. It is today possible to measure the optimal treatment dose for patients by skin reflectance measurement (Bech-Thomsen et al., 1994).
115.2 UV SOURCES Emission of ultraviolet radiation can be obtained in different ways. For full-body treatment it is most common to use fluorescent tubes, but radiation with bulb-type sources equipped with filters to achieve the desired wavelengths may also be used. Nowadays the most common sources are emitting broadband UVB, narrowband UVB, and broadband UVA or UVA1. The emission spectra are shown in Figure 115.1. Broadband UVB was the first to be used in the treatment of psoriasis and eczema and replaced carbon arc and mercury vapor lamps. The mercury vapor spectrum consists of spikes with very high 981
982
Handbook of Non-Invasive Methods and the Skin, Second Edition
1.0
0.8 Cleo 0.6
TL10 TL01 UV6
0.4
TL12
0.2
0.0 260
280
300
320
340
360
380
400
FIGURE 115.1 Emission spectra of light sources used in the treatment of skin diseases. The dashed vertical line represents the border between UVA and UVB.
output at 254, 297, 303, 313, and 365 nm. The carbon arc spectrum is sun spectrum like, but may also have shorter wavelengths. The emission spectrum will depend on the impurities (additives) put into the carbon electrodes. Broad-spectrum UVB is mainly giving radiation between 280 and 350 nm (Philips TL12) or from about 290 to 370 nm (Waldmann UV6). Both lamp types have shown a convincing efficacy in the treatment of light to moderate psoriasis and eczema. In the last 15 years the narrowband UVB (Philips TL01) has been spread because of a higher efficiency in the treatment of the mentioned diseases. Since most radiation is emitted at about 311 nm, absorbing molecules may be hit very hard compared to radiation from the more broad-spectrum light sources that have only smaller UV amounts at the different wavelengths. It also turns out that the narrowband UVB is more erythrogenic than should be expected from the CIE erythema action spectrum, compared to the broadband UVB sources, and it should therefore also be expected to be more carcinogenic (Hansen et al., 1994). Unpublished studies in mice comparing broad-spectrum and narrowband UVB indicate that narrowband UVB is more carcinogenic than broadband UVB. Equal effective doses seem to have more or less the same carcinogenic potential, and narrowband UVB should be reserved for severe psoriasis (Dawe et al., 2003). PUVA is a combination of psoralen drug and broadspectrum UVA (Parrish et al., 1974). Psoralen may be given systemically or topically. The phototoxic reaction from the treatment is diminished by about a factor of 3 if UVA1 is used as a radiation source (Bech-Thomsen and Wulf, 1994). In fact, psoralen and UVA result in crosslinks in DNA, and it has long been known that PUVA is more (about a factor of 20) carcinogenic than UVB, and the treatment should be limited to severe skin diseases, where UVB is not efficient (Stern and Vakeva, 1999).
Besides the effect on DNA, UV treatment also affects the immune system of importance for the control of skin diseases. UVA1 has very limited effect on DNA, and only makes DNA changes that are easily reparable for the cells (Petersen et al., 2000). Besides, it induces collagenase and has a suppressing effect on immune cells; UVA1 has accordingly been shown to be effective in some immunological disorders like localized scleroderma and possibly lupus erythematosus, and may be used instead of UVB in atopic dermatitis (Krutmann, 1995; El-Mofty et al., 2004; Scharffetter et al., 1991).
115.3 SUNLIGHT Sunlight has been used for thousands of years in the treatment of skin diseases and is still used at the Dead Sea and other Mediterranean areas, where patients with chronic skin diseases like psoriasis are exposed to sunlight (Abels and Kattan-Byron, 1985) (Figure 115.2). In Northern Europe the sun spectrum starts at about 295 nm in summer, and in the wintertime around 305 nm (Figure 115.3). Going closer to the equator will give shorter wavelengths, but no substantial radiation below 290 nm. It is often claimed that only UVA wavelengths penetrate down to the sea level at the Dead Sea; however, that is not the case when measured spectroradiometrically, and the sun is thus not at all gentle at that site, as often claimed. Figure 115.3 shows the UV spectrum of the sun, and Figure 115.4 shows the biological erythema efficacy of the sunlight over the year and over the day (Figure 115.5), both indicating that the sun altitude is most important for the biological efficacy of sunlight. Figure 115.6 shows the erythemaefficient wavelengths in sunlight, and it is therefore most important to be protected from those wavelengths when the sun is used for cosmetic purposes.
Light Sources, Sunlight, and Radiation Dosimetry
983
FIGURE 115.2 Psoriasis patients having heliotherapy at the Dead Sea in Israel.
Erythema potential
Summer
1
−6 Winter
0.1
−7
0.01
−8
0.001
−9
0.0001
280
300
320
340
360
380
400
Sun intensity (log)
−5
10
−10
Wavelength (nm)
FIGURE 115.3 Emission spectra of the winter and summer sun in Denmark seen in relation to the erythema action spectrum for human skin (full line). 40 35
SED per day
30 25 20 15 10 5 0 Jan Feb Mar Apr May Jun
Jul
Aug Sep Oct Nov Dec
FIGURE 115.4 Ambient biological active UV dose per day during 1 year. The average Dane tolerates about 5 SED before erythema is provoked.
984
Handbook of Non-Invasive Methods and the Skin, Second Edition
5. July
SED/hour
6
4 21. April 2 15. December 0 5
7
9
15 13 Time of day
11
17
19
21
23
FIGURE 115.5 Ambient biological active UV dose on one sunny day during summer, spring, and winter.
100
FIGURE 115.7 Instrument (UV optimize 558) consisting of a pen for measuring ultraviolet radiation in biological units (SED) and a pen that measures the UV sensitivity of a patient referred to phototherapy. The central box indicates the optimal treatment time or dose in a specific treatment cabin. Measurements take less than 10 seconds.
22. July
CIE × SUN
80 14. April
60 40 20
15. Dec.
0 280
300
320
340
360
380
3
400
Wavelength (nm)
1
2
FIGURE 115.6 The erythema action spectrum multiplied by the sun’s spectrum on days in summer, spring, and winter to illustrate which wavelengths in sunlight contribute the most to erythema of skin. The relative erythema effectiveness at different times of the year is shown.
115.4 DOSIMETRY The sensitivity of the skin to ultraviolet radiation is very wavelength dependent, and the CIE has published an erythema action spectrum, seen in Figure 115.3 (McKinlay and Diffey, 1987). Together with sunlight spectra, it can roughly be seen that wavelengths around 300 nm and shorter have the highest erythema effectiveness, and that at 328 nm the effectiveness has fallen into a 1000th, and at 400 nm 1/10,000th, of the wavelength effectiveness at 300 nm. This also illustrates that a shift in sun emission wavelengths between summer and winter will highly affect the number of SEDs reaching the earth per day. On a sunny summer day in Denmark the number of SEDs may be up to 38, while on a sunny winter day it may reach 1 SED. In connection with treatment of skin diseases, erythema becomes very important, as the limitation in the UV dose that can be given to the patients is in fact erythema. The wavelength must therefore be taken into
FIGURE 115.8 Spectroradiometer for measurement of the emission spectrum of a light source. The light source (1) is exchangeable with any light source for direct measurement or via a light guide. The light passes a monochrometer (2) and goes into a detector unit (3). A computer calculates the emission spectrum.
the equation, when determining the time to 1 SED from a light source. Therefore, the specter of the light source must be measured spectroradiometrically (Taylor et al., 2002). The intensity at each wavelength is determined and used to compute the SED. SED can also be measured by a detector, which has the same sensitivity as the erythema action spectrum. The time to 1 SED is then given directly without computerization and is therefore considerably cheaper and very easy to handle. Both such systems are available and can be seen in Figure 115.7 and Figure 115.8.
Light Sources, Sunlight, and Radiation Dosimetry
115.5 PREDICTION OF UV DOSE FOR TREATMENT In the treatment of psoriasis it seems optimal to give a just perceptible erythema dose at each treatment, whereas in atopic dermatitis it is advantageous to give nonerythrogenic doses. Treatment doses will of course also depend on skin type at the start of treatment, and later depend on the development of pigmentation (Bech-Thomsen et al., 1994). The sensitivity of the skin can be measured by skin reflectance measurements, and such equipment can also predict the desirable dose on a personnel level (BechThomsen et al., 1994; Wulf, 1986). It has been shown that the UVB doses to erythema of different grades are directly correlated to measured pigment protection factor (PPF) of skin and degree of pigmentation (Lock-Andersen et al., 1997). Both parameters can be measured by skin reflectance, and the number of SEDs that should be given to a person to optimize psoriasis treatment or to optimize the treatment of other diseases is computerized (Bech-Thomsen and Wulf, 1995). The calculated SED is not only correlated to the PPF, but also to the number of weekly treatments and the total length of the treatment. Since persons are generally most UV sensitive on the body, a good measuring place will be between the shoulder blades, normally the least irritated area. It has to be remembered that such a measurement is a compromise for all body sizes and may not hold if there are large differences in pigmentation between body parts. The lower extremities normally tolerate more UV than the other topographic body parts. The measuring system is normally used at the beginning of the treatment and thereafter once a week to compensate for changes in pigmentation during treatment. The treatment dose to a very sensitive person for atopic dermatitis and treatment for psoriasis in a very dark skinned person may differ by about a factor of 20. This illustrates that the doctor’s guess of treatment time is not very valid, and dose measurement has also been shown to shorten treatment time or reduce the total number of SED by at least 50 to 60% (Bech-Thomsen and Wulf, 1995; Selvaag et al., 2000).
115.6 TEST EQUIPMENT FOR PHOTOSENSITIVITY Photosensitivity diseases are quite common, and differentiation between the different diagnoses can be quite difficult (Stratigos et al., 2002). Test systems for diagnostic use (Lehmann, 1993) shall determine the minimal erythema dose (MED) from different wavelengths in the UVB, UVA, and blue part of the visible spectrum. MED can be determined by irradiating with small bits of the spectrum through a monochromator. Typically, six different doses are given in 25 to 40% increments. In one of the test fields minimal erythema is located 24 hours after
985
exposure, and the UV dose given in that field represents the MED (Lock-Andersen and Wulf, 1996). MED is most often determined by a solar simulator emitting light from a xenon arc bulb filtered to give UV like in sunlight. Exchanging filters also allows testing of UVA separately. In some of the solar simulators the six different doses are given in one and the same session (Figure 115.12). Each irradiated field covers about 1.4 cm2. A few solar simulators also allow testing with visible light (e.g., blue). An even simpler test procedure may be performed by means of sets of UVB tubes or UVA tubes. For UVB testing, Philips TL 12 may be used. These also emit some UVA, but because of very short irradiation times (few minutes), this may be neglected. The UVA source, however, must be practically free of UVB since very long irradiation times (hours) must be used, and even very small UVB intensities will add up to clinically relevant doses. Philips TL10 (UVA1) can be used, preferably with a window glass filter. Blue light testing may likewise be performed with OSRAM deluxe L71 tubes. With these lamp sources the patients are irradiated through a frame with holes, which is closed at different time intervals to give different doses. A better alternative is to make the MED test by means of a disposable neutral density filter sheet automatically giving six different doses (MED test patch, Chromo-Light, Denmark). Having determined the MED, one must decide whether it is abnormal. In some instances this is performed according to skin type (Fitzpatrick, 1988) and experience with earlier test doses in the local population. It can, however, be evaluated on an individual basis by reflectance measurement (UV-optimize 558, Chromo-Light, Denmark) (Lock-Andersen et al., 1999). The test fields are measured before irradiation, and the instrument calculates the number of SEDs to erythema assuming normal sensitivity. The day after UV exposure the MED spot is found and the number of SEDs to give erythema is compared to that expected. From these figures the increase in sensitivity may be found. The highest increase in sensitivity we have ever seen was 95 times normal (the patient tolerated 95 times less than normal). Test equipment is seen in Figure 115.12 and Figure 115.7, and interpretation of the test results in relation to ideopatic photodermatoses is shown in Table 115.1. Pratically all phototoxicity and photoallergy reactions are associated with exposure to long-wave ultraviolet irradiation, and the UVA1 (Philips TL10) is thus suitable for this purpose. Most often 5 J/cm2 is given on patch tests applied the day before (British Photodermatology Group, 1997; Lehmann, 1993). Some, however, prefer to use higher doses in order not to miss weak reactions.
115.7 PHOTODYNAMIC THERAPY Another treatment modality that has come into dermatology within the last 10 years is photodynamic therapy
986
Handbook of Non-Invasive Methods and the Skin, Second Edition
Glycine + Succinyl-CoA Pyridoxal PO4 ALA-Synthetase HOOC
CH2
CH2
CO
CH2
NH2 (ALA)
ALA Ac
Pr (PBG)
H2N
N
H2C
PBG Deaminase PBG Isomerase Pr A
Pr
H H
H2
N
N
H
N
H
H2 Ac
Ac B
N
Ac
H2
Ac
C
D
H2 Pr Pr Urogen III Urogen decarboxylase
H N
N H
H
H2 Pr Pr Coprogen III
H2
N
CH3
Oxidase
H3C
Vi
N
H2
CH3 B
H
H
N H3C
A
Coprogen
N
H
H2
H 3C
Pr
H2
H
H3C
Vi
H2
N
CH3
N
H2
H
Pr
CH3
C
D
H2 Pr Pr Protoporphyrinogen Protogen oxidase CH3
Vi
Vi
N
N
H
H
N
N
H 3C
H3C
CH3 Pr
Pr PP IX Fe+ +
Ferrochelatase
HEME
FIGURE 115.9 Heme cascade. Adding ALA or methyl-ALA (MAL) to the skin will lead to accumulation of protoporphyrin IX (PPIX). This makes the cells light sensitive, and irradiation with red or blue light leads to cell death.
(PDT) with 5-aminolevulenic acid (5-ALA) or its esters (Metvix®, MAL) (Braathen, 2002). The substances are applied locally on the skin after preparing the skin by superficial curettage and pricking holes in tumors. An occlusive dressing is put on the top of the cream and left there for 3 hours. In that time the MAL is going into the
cells, and via the heme cascade protoporphyrin IX (PPIX) is accumulated (Figure 115.9 and Figure 115.10). PPIX is very phototoxic and irradiation with light, especially red light, may lead to the cytotoxic effect. Broad-spectrum red light or narrowband red light (Figure 115.11) is used to obtain the phototoxic reaction and formation of free
Light Sources, Sunlight, and Radiation Dosimetry
987
TABLE 115.1 Diagnostic Guidance to Photodermatoses
Disease
MED-Test Times Increased Sensitivity
PLE
Normal >1.5
Actinic prurigo Chronic actinic dermatitis Solar urticaria Photo-allergy (on drug) Photo-toxicity (on drug)
5 Days Photoprovocation by UVB and UVA
Most Common Time of Year
1.5–2 2–100
Abnormal (most often papules) Abnormal papules Abnormal (dermatitis)
Spring Summer All year All year
Normal (wheal)
Abnormal (urticaria)
>1.5 (UVA)
Eczema
>1.5 (UVA)
Eczema
Summer All year Spring Summer Spring Summer
Who Young adults Children Middle-aged men Middle-aged women All (photo patch test) All (photo patch test)
100 MAL Normalized intensity
ALA
AK
FIGURE 115.10 Fluorescence photo of sun-damaged skin with actinic keratosis (AK) treated with 5-ALA (left) or MAL (Metvix) (right). In both cases there is a high uptake in AK (white spots), but more unspecific PPIX in normal skin in the 5-ALAtreated area than in the MAL-treated area.
oxygen radicals, which the effect depends on (Huang et al., 2003). Since the light dose is depending on how well the light spectrum is absorbed in PPIX, the doses will vary; 37 J/cm2 is normally used for narrowband irradiation from light-emitting red diodes and about 70 J/cm2 for broadband red light sources (Figure 115.11). The light penetration into the skin is very wavelength dependent; that is why only red light is used for deeper lesions. Superficial actinic keratoses may also be treated with 5-ALA (Levulan®) and
80 60 40 20 0 500
525
550
575
600
625
650
675
700
725
750
Wavelength (nm)
FIGURE 115.11 Protoporphyrin IX has a peak absorption around 630 nm. The light source used for light activation must therefore contain this wavelength range. The broad spectrum is produced by a Waldman PDT 1200 (GE) and the narrow spectrum by an ahtilite 128 (PhotoCure, Norway). Only about half the light dose is needed when the narrowband light-emitting diode source is used.
blue light at 400 to 430 nm, where the penetration into the skin is low but the absorption of PPIX is very high. This treatment is especially used for actinic keratosis and superficial basal cell carcinoma and has been proven to be very efficient, with a cure rate of about 90% (Szeimies et al., 2002). It has recently been shown that just one session will cure about 80% of actinic keratosis and that 60% of the rest will be cured by one more treatment after 3 months (Tarstedt et al., 2003). In basal cell carcinoma, two treatments, one week apart, are normally given after careful preparation, especially if nodular tumors are treated (Horn et al., 2003; Rhodes et al., 2004). Light and UV treatment has developed tremendously through the last decades, and there is no reason to believe
988
Handbook of Non-Invasive Methods and the Skin, Second Edition
FIGURE 115.12 Computerized solar simulator (ChromoLight, Denmark). The instrument can reach the back of a person lying on a bed. It can be lowered to be in contact with the skin. For MED testing it will give six different UV doses of either solar radiation, UVA, or UVB. Together with the optimizer (Figure 115.7), it can calculate the degree of deviation from the normal MED. For provocation testing it irradiates a 5 × 5 cm area and may give full-spectrum, UVA, UVB, solar radiation, or visible light.
that it is at its end. Laser also emits light, but it was not the scope of this chapter.
REFERENCES Abels DJ, Kattan-Byron J. Psoriasis treatment at the Dead Sea: a natural selective ultraviolet phototherapy. J Am Acad Dermatol 12: 639–643, 1985. Anderson TF. Light sources in photomedicine. In Clinical Photomedicine, Lim HW, Soter NA, Eds. Marcel Dekker, New York, 1993, pp. 37–58. Bech-Thomsen N, Angelo HR, Wulf HC. Skin pigmentation as a predictor of minimal phototoxic dose after oral Methoxsalen. Arch Dermatol 130: 464–468, 1994.
Bech-Thomsen N, Wulf HC. A polychromatic action spectrum for photosensitivity to orally administered 8-methhoxypsoralen in humans. Clin Exp Dermatol 19: 12–15, 1994. Bech-Thomsen N, Wulf HC. Skin reflectance-guided UVB phototherapy of psoriasis reduces the cumulative UV dose significantly. J Dermatol Treat 6: 207–210, 1995. Braathen LR, Ed. Photodynamic Therapy Basic Principles and Clinical Experience with Metvix PDT. PhotoCure ASA, Oslo, 2002, pp. 1–66. British Photodermatology Group. Photopatch testing: methods and indications. Br J Dermatol 136: 371–376, 1997. Dawe RS, Cameron H, Yule S, Man I, Wainwright NJ, Ibbotson SH, Ferguson J. A randomized controlled trial of narrowband ultraviolet B vs. bath-psoralen plus ultraviolet A photochemotherapy. Br J Dermatol 148: 1194–1204, 2003. Diffey BL, Jansén CT, Urbach F, Wulf HC. Standard erythema dose, a review. CIE J 125: 1–5, 1997. El-Mofty M, Mostafa W, E-Darouty M, Bosseila M, Nada H, Yousef R, Esmat S, El-Lawindy, Assaf M, El-Enani G. Different low doses of broad-band UVA in the treatment of morphea and systemic sclerosis, a clinico-pathologic study. Photodermatol Photoimmunol Photomed 20: 148–156, 2004. Finsen NR. Om anvendelse i medicinen af koncentrerede kemiske lysstraaler. Gyldendalske Boghandels Forlag 1–54, 1896. Fitzpatrick TB. The validity and practicality of sun-reactive skin types I through IV. Arch Dermatol 124: 869–870, 1988. Hansen AB, Bech-Thomsen N, Wulf HC. Erythema after irradiation with ultraviolet B from Philips TL12 and TL01 tubes. Photodermatol Photoimmunol Photomed 10: 22–25, 1994. Horn M, Wolf O, Wulf HC, et al. Topical methyl aminolevulinate photodynamic therapy in patients with basal cell carcinoma prone to complications and poor cosmetic outcome with conventional treatment. Br J Dermatol 149: 1242–1249, 2003. Huang Z, Chen Q, Shakil A, Chen H, Beckers J, Shapiro H, Hetzel FW. Hyperoxygenation enhances the tumor cell killing of Photofrin-mediated photodynamic therapy. Photochem Photobiol 78: 496–502, 2003. Kromayer E. Quecksilberwassenlampen zur Behandlung von Haut und Schleimhaut. Dtsch Med Wochenschr 48–51, 1906. Krutmann J. UVA-1 radiation induced immunomodulation, UVA-1 for atopic dermatitis. In Photoimmunology, Krutmann J, Elmets C, Eds. Blackwell Scientific, Oxford, 1995, pp. 246–256. Lehmann P. Principles of photo testing and photopatch testing: a European perspective. Retinoids Today Tomorrow 31: 36–41, 1993. Lock-Andersen J, de Fine Olivarius F, Hædersdal M, Poulsen T, Wulf HC. Minimal erythema dose in UV-shielded and UV-exposed skin predicted by reflectance measured pigmentation. Skin Res Technol 5: 88–95, 1999.
Light Sources, Sunlight, and Radiation Dosimetry
Lock-Andersen J, Therkildsen P, de Fine Olivarius F, Gniadecka M, Dahlstrøm K, Poulsen T, Wulf HC. Epidermal thickness, skin pigmentation and constitutive photosensitivity. Photodermatol Photoimmunol Photomed 13: 153–158, 1997. Lock-Andersen J, Wulf HC. Threshold level for measurement of UV sensitivity: reproducibility of phototest. Photodermatol Photoimmunol Photomed 12: 154–161, 1996. McKinlay AF, Diffey BL. A reference action spectrum for ultraviolet induced erythema in human skin. CIE J 6: 17–22, 1987. Parrish JA, Fitzpatrick TB, Tanenbaum L, Pathak MA. Photochemotherapy of psoriasis with oral methoxsalen and longwave ultraviolet light. N Engl J Med 291: 1207–1211, 1974. Petersen AB, Gniadecki R, Vicanova J, Thorn T, Wulf HC. Hydrogen peroxide is responsible for UVA-induced DNA damage measured by alkaline comet assay in HaCaT keratinocytes. J Photochem Photobiol B Biol 59: 123–131, 2000. Rhodes LE, de Rie M, Enström Y, Groves R, Morken T, Goulden V, Wong GAE, Grob J-J, Varma S, Wolf P. Photodynamic therapy using topical methyl aminolevulinate vs surgery for nodular basal cell carcinoma. Arch Dermatol 140: 17–23, 2004. Scharffetter K, Wlaschek M, Hogg A, Bolsen K, Schothorst A, Goerz G, Krieg T, Plewig G. UVA irradiation induces collagenase in human dermal fibroblasts in vivo and in vitro. Arch Dermatol Res 283: 506–511, 1991. Selvaag E, Caspersen L, Bech-Thomsen N, de Fine Olivarius F, Wulf HC. Optimized UVB treatment of psoriasis: a controlled, left-right comparison trial. JEADV 14: 19–21, 2000.
989
Stern R, Vakeva L. Noncutaneous malignant tumors in the PUVA follow-up study: 1975–1996. J Invest Dermatol 108: 897–900, 1997. Stratigos AJ, Antoniou C, Katsambas AD. Polymorphous light eruption. JEADV 16: 193–206, 2002. Szeimies RM, Karrer S, Radakovic-Fijan S, Tanew A, CalzavaraPinton PG, Zane C, Sidoroff A, Hempel M, Ulrich J, Proebstle T, Meffert H, Mulder M, Salomon D, Dittmar HC, Bauer JW, Kernland K, Braathen L. Photodynamic therapy using topical methyl 5-aminolevulinate compared with cryotherapy for actinic keratosis: a prospective, randomized study. J Am Acad Dermatol 47: 258–262, 2002. Tarstedt M, Wennberg A-M, Rosdahl I, Persson BE, Berne B, Bojs G, Nyberg F, Andersson K, Bendsöe N. A comparison of two treatment regimes using photodynamic therapy with Metvix® in actinic keratosis. Paper presented at the Finsen 100 Years Anniversary Symposium, Copenhagen, 2003 (abstract). Taylor DK, Anstey AV, Coleman AJ, Diffey BL, Farr PM, Ferguson J, Ibbotson S, Langmack K, Lloyd JJ, McCann P, Martin CJ, Menagé du P, Moseley H, Murphy G, Pye SD, Rhodes LE, Rogers S. Guidelines for dosimetry and calibration in ultraviolet radiation therapy: a report of a British photodermatology group work-shop. Br J Dermatol 146: 755–763, 2002. Wulf HC, Lock-Andersen J. Standard erythema dose. Skin Res Technol 3: 192, 1996. Wulf HC. Method and Apparatus for Determining an Individual’s Ability to Stand Exposure to UV. U.S. Patent 14: 882, 598: 1–32, 1986.
Phototoxicity and 116 Phototesting: Photoallergy Takeshi Horio Department of Dermatology, Kansai Medical University, Osaka, Japan
CONTENTS 116.1 116.2 116.3 116.4
Introduction ..........................................................................................................................................................991 Light Source .........................................................................................................................................................991 Phototoxic and Photoallergic Reactions ..............................................................................................................992 Phototesting without Chemicals ..........................................................................................................................992 116.4.1 UVA Irradiation .....................................................................................................................................992 116.4.2 UVB Irradiation .....................................................................................................................................993 116.4.3 Visible Light Irradiation ........................................................................................................................993 116.5 Photopatch Testing ...............................................................................................................................................994 116.5.1 Principle and Indication.........................................................................................................................994 116.5.2 Method ...................................................................................................................................................994 116.5.3 Test Site..................................................................................................................................................994 116.5.4 Materials.................................................................................................................................................994 116.5.5 Light Shielding ......................................................................................................................................994 116.5.6 Irradiation...............................................................................................................................................995 116.5.7 Interpretation..........................................................................................................................................995 116.6 Drug Phototesting.................................................................................................................................................995 116.7 Serum Phototesting ..............................................................................................................................................996 116.8 Conclusion............................................................................................................................................................996 References .......................................................................................................................................................................996
116.1 INTRODUCTION
116.2 LIGHT SOURCE
In some patients it is possible to make a confident diagnosis of photosensitivity based on the distribution pattern and history of cutaneous changes. However, subsequent phototesting is required to make an accurate decision. Phototesting is usually employed to make a miniature of skin lesions in patients with suspected photosensitivity disease by means of irradiation with artificial light sources. The existence of photosensitivity, precise diagnosis, action spectrum, or etiologic factors can be confirmed when the characteristic skin changes are reproduced in phototesting. This type of in vivo testing is useful for several photosensitivity diseases, listed in Table 116.1. Light is exposed to the patient’s skin with or without an application of chemicals that are suspected to cause the photosensitivity state.
The selection of a light source is the most essential step in phototesting. The sun is the reasonable light source since clinical photosensitivity reaction in patients is induced by natural sunlight. However, light energy from the sun is unpredictable and hard to control. It is impossible to obtain the irradiance at any intensity and at any time desired. A variety of artificial light sources are commercially available.1 However, it is imperative that the light source for phototesting has adequate irradiance in the action spectrum of the disease being examined. Otherwise, phototesting may yield a false negative result. The emission spectrum should be in the long-wave ultraviolet light (320 < UVA < 400 nm), mid-wave ultraviolet light (290 < UVB < 320 nm), or visible light (400 to 800 nm) range. Infrared radiation seems to be irresponsible to photosensitivity 991
992
Handbook of Non-Invasive Methods and the Skin, Second Edition
TABLE 116.1 Indication of Phototesting Disease
Action Spectrum
Polymorphous light eruption Photocontact dermatitis Drug-induced photosensitivity Chronic actinic dermatitis Solar urticaria Hydroa vacciniforme Lupus erythematosus Xeroderma pigmentosus Cockayne’s syndrome
UVA or UVB UVA UVA UVA, UVB (visible light) UVA, UVB, or visible light UVA UVB UVB (UVA) UVB
reaction. The light source for in vivo testing should not emit short-wave ultraviolet light (UVC < 290nm), which is not included in the natural sunlight and may induce a false positive reaction. Furthermore, it is desirable that the light source be easily and economically available for practice use. Fluorescent tubes are useful and convenient light sources for UVA and UVB irradiation. The so-called black light and sunlamp produce irradiation primarily in the UVA and UVB ranges, respectively. These fluorescent tubes have a large enough field size to test a number of materials at once in photopatch testing. A slide projector can be used as a visible light source to especially examine solar urticaria. The solar simulator equipped with a xenon lamp emits a broad waveband with a mix of UVB, UVA, and visible light similar to that occurring in natural sunlight. However, this light source is expensive and has a small field size.
to the causative drugs. Clinically, however, there exists an apparent refractory period, because the photosensitizing chemicals and the required wavelengths are only rarely present at the start of drug administration in the proper amounts for a reaction to occur. Drug-induced photoallergy usually involves a delayed hypersensitivity response. It is especially well established that photoallergic contact dermatitis develops through a T cell-mediated type IV immunologic reaction.2,3 Experimental photoallergic contact dermatitis can be induced in guinea pigs and mice using a modification of an animal model of contact hypersensitivity. Therefore, the cutaneous change is a clinically eczematous eruption that is identical to allergic contact dermatitis. However, the difference in clinical features is not always clear-cut between a phototoxic and photoallergic reaction. Action spectrum studies cannot differentiate two reactions. A single drug may cause phototoxic as well as photoallergic reactions. A photoallergic reaction is usually induced with less doses of the drug and light than a phototoxic reaction. Like an allergic drug eruption or contact dermatitis, there is an incubation period of at least 7 days in man. Some patients with drug-induced photoallergic dermatitis may retain a persistent reactivity to sunlight that continues long after the exposure to the causative photosensitizing compound has ceased. These patients are called persistent light reactors.4
116.4 PHOTOTESTING WITHOUT CHEMICALS 116.4.1 UVA IRRADIATION
116.3 PHOTOTOXIC AND PHOTOALLERGIC REACTIONS The most common cause of photosensitivity diseases is drug or chemical substance. Drug photosensitivity reactions are cutaneous responses to the combined action of a chemical and a physical agent. The activating light must be absorbed by the drug to initiate the photosensitizing reaction, since photobiologic responses are predicated upon photochemical reactions. Thus, drugs that do not absorb light energy do not induce a photosensitivity reaction. Druginduced photosensitivity can be divided into phototoxic and photoallergic reactions based on the mechanisms involved, whether nonimmunological or immunological. The mechanism of action of drug-induced phototoxic reaction is not uniform, depending on the responsible chemicals. Phototoxic reactions can occur theoretically in 100% of the population if sufficient doses of the drug are administered and appropriate wavelengths (action spectrum) of light are irradiated. Experimentally, reactions can develop without an incubation period after first exposure
The UVA-sensitive photodermatoses include druginduced photosensitivity (phototoxic and photoallergic), photocontact dermatitis, chronic actinic dermatitis, certain cases of polymorphous light eruption, hydroa vacciniforme, and some cases of solar urticaria (Table 116.1). There is no standardized procedure for UVA phototesting. The irradiation dose to reproduce the clinical lesion depends on the severity of photosensitivity in the patients, and therefore it varies not only among diseases, but also among patients even of an identical disease. In our photodermatology unit, UVA at doses of 1.5, 3.0, 6.0, and 9.0 J/cm2 is exposed to four areas (2 × 2 cm each) on normal-appearing skin (usually on the back) using fluorescent black light tubes. Exposed areas are read 24 and 48 hours after irradiation. Only the patients with possible solar urticaria are evaluated for wheal formation until 30 min after exposure. In normal subjects, the UVA exposure under this condition does not produce any reaction, except immediate pigment darkening, which disappears within a few hours of irradiation. Therefore, an
Phototesting: Phototoxicity and Photoallergy
FIGURE 116.1 Phototesting in hydroa vacciniforme. An edematous erythema with vesicles was induced with UVA radiation.
erythematous reaction can be estimated as an abnormal photosensitive state. Patients with chronic actinic dermatitis are most reactive to UVA irradiation,6 while most cases with polymorphous light eruption are least susceptible. When no reaction is produced with this phototesting in patients who are strongly suspected to have photosensitivity disease, the UVA exposure is repeated two or three times at 24- or 48hour intervals at the same area. Eczematous or dermatitic changes appear in the UVAinduced photosensitivity, chronic actinic dermatitis, and polymorphous light eruption. In hydroa vacciniforme, edematous erythema appears with vesicle or bullae, which may develop into necrosis, when a sufficient dose of UVA is exposed (Figure 116.1).
116.4.2 UVB IRRADIATION The UVB-sensitive photodermatoses are less common than the UVA-reactive diseases. The UVB radiation easily produces sunburn reaction even in normal persons. Therefore, it is not always easy to interpret the reaction induced by UVB irradiation because the photosensitive patients can develop pathological changes mixed with a physiological reaction. The reactions should be evaluated quantitatively and qualitatively. There are two procedures of UVB phototesting: the minimal erythema dose (MED) and the delayed erythema dose (DED) tests.15 The MED test is described in detail elsewhere in this text. In patients with xeroderma pigmentosum or Cockayne’s syndrome, the MED to induce sunburn reaction is lowered. The cutaneous reaction is indistinguishable from that of normal subjects macroscopically and microscopically. However, the time course of UVB erythema is characteristic in these photosensitive genodermatoses in comparison with normal sunburn reaction. The peak of reaction is delayed, being reached at 48 to 72 hours after irradiation, and the erythema persists longer, extending to 5 to 7 days.
993
Patients with other photodermatoses, such as chronic actinic dermatitis, often react to much lower UVB dose than MED in normal subjects. However, the UVB-induced reaction is not a simple sunburn, but an eczematous change in these patients. It may be necessary to deliver larger amounts of UVB (DED) for reproduction of skin eruption in certain photosensitivity diseases, in which MED is normal, such as polymorphous light eruption, in some patients with druginduced photosensitivity, and in the systemic as well as discoid lupus erythematosus. There is no established procedure for this test. The cutaneous changes of other photoaggravated dermatoses may also be induced with high doses of UVB radiation. A single or divided exposure of 6 to 8 MEDs on nonlesional skin is often used for DED test. Sunburn erythema appears at 10 to 24 hours after the last irradiation. As it is subsiding, a second, erythematous, or papular reaction may become visible at 3 to 10 days after the irradiation. It is evaluated as positive when the reaction is identical with the clinical eruption (Figure 116.2). The positive responses may persist for 4 to 14 days. In most instances, repeated exposures of divided dose at 24- to 48hour intervals are more useful than single exposure of the same UVB dose. Using this technique, Epstein7 demonstrated a positive response in 90% of the patients with polymorphous light eruption. It may be necessary to irradiate previously diseased skin in some patients to demonstrate the positive phototest results.
116.4.3 VISIBLE LIGHT IRRADIATION The phototesting with visible light source is of diagnostic value, especially in solar urticaria, whose action spectrum lies in this range. A slide projector is an easily available and valuable light source for provocative phototesting.
FIGURE 116.2 Phototesting in polymorphous light eruption. 2 MEDs of UVB were exposed three times at a 48-hour interval to the same area. Papular reaction appeared 3 days after the last irradiation.
994
Handbook of Non-Invasive Methods and the Skin, Second Edition
1. Application of chemical substances in duplicate patches 2. Covering of patches with light opaque material 3. Evaluation of uncovered patches 48 hours later 4. Irradiation of one set of patches with UVA 5. Covering of both sets 6. Evaluation after an additional 24, 48, and 72 hours
116.5.3 TEST SITE FIGURE 116.3 Urticarial wheal induced in a patient with solar urticaria by exposure to a slide projector.
The exposure dose varies depending on the severity of the disease and the light source used. We routinely use a slide projector equipped with a 500-W bulb, since a larger bulb emits too much infrared radiation and may heat the exposed skin. A wheal formation can be produced in the patients with solar urticaria immediately after the exposure for 1 to 10 min at a distance of 20 cm from the projector lens (Figure 116.3). Precise action spectra of solar urticaria can be obtained by using cutoff glass filters in combination with a slide projector. In selected patients, not only action spectra but also inhibition and augmentation spectra can be found.8,9 In these cases, pre- or postirradiation with wavelengths longer than action spectrum may inhibit or augment the urticarial response. In a few patients with chronic actinic dermatitis, visible light radiation may evoke an eczematous change at 24 to 48 hours after exposure. These patients are also photosensitive to the UVA range.
116.5 PHOTOPATCH TESTING 116.5.1 PRINCIPLE
AND INDICATION
The principle of photopatch testing is exactly the same as that of patch testing in allergic contact dermatitis. It is performed in order to reproduce the photoallergic contact dermatitis in miniature by topical application of the offending photosensitizer and subsequent exposure by activating ultraviolet light. Photopatch testing may be valuable in some patients who are photoallergically sensitized by systematically administered drugs. It cannot be applied for diagnosis or to prove the cause of phototoxic dermatitis, because phototoxic reaction is not specific to the photosensitive patients, but can occur in all normal persons.
116.5.2 METHOD Photopatch testing is performed with the following procedure, although certain modifications may be made:
Any site of the whole-body skin can reveal a positive photopatch test result. However, the test area must be large enough, since the chemicals to be examined are applied in duplicate. The patches are placed symmetrically on nonlesional skin of the interscapular or lumbar regions of the back.
116.5.4 MATERIALS Suspected photoallergenic chemicals must be selected to be examined. The chemical substances listed in Table 116.2 are often used as the photopatch test series. Selection should be modified by information from the patient’s history. Most substances are usually incorporated into petrolatum in 1% concentration. However, photosensitizing chemicals have the ability to induce a phototoxic as well as photoallergic reaction. Therefore, test materials must be used in nonirritating and nonphototoxic concentrations. For example, chlorpromazine is a strongly phototoxic substance, and also frequently induces a photoallergic reaction. It is commonly used in 0.1 to 0.01% concentration. In contrast, the phototoxicity of halogenated salicylanilides is low in in vivo tests. However, higher concentrations can inadvertently photosensitize patients after repeated photopatch testing. Furthermore, Harber and Bickers1 observed a “broadening of the base” phenomenon in several patients with 1% tetrachlorosalicylanilide (TCSA). These patients were initially photosensitive only to TCSA but developed photosensitivity to bithionol on the fourth photopatch testing.
116.5.5 LIGHT SHIELDING In the severely photosensitive patients, even trace amounts of light exposure to patched sites may yield false positive test results. Epstein10 designated such a reaction as “masked photopatch test,” since it is a positive photopatch test due to unintentional exposure to light. Willis and Kligman11 were able to elicit a positive photopatch test reaction to halogenated salicylanilide with light exposure through cotton and woolen cloth, adhesive tape, and bond paper. To avoid false positive reaction, the patches must be completely covered with light opaque material. For this purpose, a fin chamber or ALTest is recommended. During
Phototesting: Phototoxicity and Photoallergy
995
TABLE 116.2 Compounds for Routine Photopatch Testing Substance
%
Vehicle
Tetrachlorosalicylanilide
0.1–1.0
Tribromosalicylanilide
1.0
Dibromosalicylanilide
1.0
Bithionol Hexachlorophene Trichlorocarbanilide Irgasam Fentichlor Chlorpromazine
1.0 1.0 1.0 1.0 1.0 0.01–0.1
Perphenazine
0.01–0.1
Promethazine Levomepromazine Thioridazine Diphenhydramine Musk ambrette 6-Methylcoumarin Para-aminobenzoic acid Oxybenzone Methoxycinnamate Butyl methoxy dibenzoylmethane Ketoprofen Suprofen Piroxicam Dibucaine Psoralens
1.0 1.0 1.0 1.0 1.0–5.0 1.0–5.0 5.0 10.0 10.0 10.0 1.0–3.0 1.0 1.0 0.1 0.001
Petrolatum Ethanol Petrolatum Ethanol Petrolatum Ethanol Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Ethanol Petrolatum Ethanol Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Ethanol Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Petrolatum Water Ethanol
the irradiation of one set of patches, another set for patch testing must be carefully shielded.
116.5.6 IRRADIATION The action spectrum of photoallergic contact dermatitis is primarily in the UVA range. Therefore, the most important instrument for photopatch testing is an artificial light source with sufficient spectral irradiance in the UVA range. A bank of multiple black light lamps is convenient equipment for this purpose because of their large field size, nonnecessity of filters, and easy availability. Test sites are irradiated with 1 to 5 J/cm2 or with 50% of the threshold dose of UVA. Willis and Kligman11 irradiated test areas immediately after application of photosensitizers. This came from their thesis that the role of light energy is simply to transform the photosensitizer into a more potent contact sensitizer, and therefore that it is not necessary to allow time for penetration to occur prior to irradiation.12 Usually, however, test sites are irradiated 24 to 48 hours after
FIGURE 116.4 Positive photopatch testing to ketoprofen. Patch testing yielded a negative result.
application, since the immediate irradiation may yield a false negative result.
116.5.7 INTERPRETATION Results of photopatch testing are evaluated by comparing them with another set of patch testing. When the irradiated patch shows a positive reaction and the nonirradiated patch shows a negative one to the identical test material, the result is interpreted as the presence of photoallergic contact dermatitis (Figure 116.4). If both sites show equally positive reactions, plain contact dermatitis is present because the light exposure does not play any role in the reaction. In a few cases, both irradiated and nonirradiated sites show positive results, but the former reaction is more pronounced than the latter. Such a result is interpreted as a coexistence of allergic and photoallergic contact dermatitis to the same chemical. If there is no reaction at either site, there is no contact or photocontact sensitization to the substances tested.
116.6 DRUG PHOTOTESTING Photopatch testing may be of no value in the diagnosis of photosensitivity reaction induced by systematically administered drugs. In principle, the causative agent should be administered through the same route as in the clinical use to reproduce the reaction. Unlike the provocative test in ordinary drug eruption, drug phototesting can be safely performed because the hypersensitivity reaction appears only on the skin localized in the exposed area. Drug phototesting is performed according to the following steps in our photodermatology clinic: 1. 2. 3. 4. 5.
Discontinuation of drug intake for 1 week Determination of MED (especially to UVA) Readministration of drug for 1 day Reexamination of MED Comparison of MED before and after drug readministration
996
Handbook of Non-Invasive Methods and the Skin, Second Edition
When the MED is significantly reduced after readministration, the drug can be estimated as the responsible photosensitizer. The above-mentioned procedure must be modified depending on the sensitizing drugs, degree of patient’s photosensitivity, and type of eruption.
116.7 SERUM PHOTOTESTING Solar urticaria can be passively transferred to normal persons by means of an intradermal injection of a patient’s serum and subsequent exposure to the activation wavelengths.13 This is a modification of the Prausnitz–Küstner technique used in the immediate hypersensitivity reaction. Therefore, the passive transfer test is a useful procedure to examine whether the disease process is allergic in nature. However, this test is not performed at the present time, since it can transfer not only allergic reaction, but also viral diseases. Some patients with solar urticaria develop a wheal at the site of injection of their own serum, which was previously exposed to light in vitro.14,15 This indicates that the wheal-forming factor is a substance produced by light energy in the patient’s serum (Figure 116.5).16
FIGURE 116.5 A wheal formation in a patient with solar urticaria at an injection site of in vitro-irradiated patient’s serum.
116.8 CONCLUSION Photosensitivity diseases include a variety of disorders with diverse pathomechanisms, etiologies, action spectra of light, and clinical manifestations. The disorders can be classified into endogenous and exogenous photosensitivity diseases, based on the origin of responsible photosensitizers or chromophores. Identification of the individual disease is important to treat and prevent the photosensitivity of patients. For this purpose, in vivo phototestings can be performed easily and safely.
REFERENCES 1. Harber, L.C. and Bickers, D.R., Photosensitivity Diseases, B.C. Decker, Philadelphia, 1989. 2. Takigawa, M. and Miyachi, Y., Mechanisms of contact photosensitivity in mice. I. T cell regulation of contact photosensitivity to tetrachlorosalicylanilide under the genetic restrictions of the major histocompatibility comples, J. Invest. Dermatol., 79, 108, 1982. 3. Maguire, H.C. and Kaidbey, K., Experimental photoallergic contact dermatitis: a mouse mode, J. Invest. Dermatol., 79, 147, 1982. 4. Wilkinson, D.S., Patch test reactions to certain halogenated salicylanilides, Br. J. Dermatol., 74, 302, 1962. 5. Jillson, O.F. and Curwen, W.L., Phototoxicity, photoallergy and photoskin tests, Arch. Dermatol., 80, 678, 1959. 6. Norris, P.G. and Hawk, J.L.M., Chronic actinic dermatitis: a unifying concept, Arch. Dermatol., 126, 376, 1990. 7. Epstein, J.H., Polymorphous light eruption, Ann. Allergy, 24, 397, 1966. 8. Hasei, K. and Ichihashi, M., Solar urticaria: determinations of action and inhibition spectra, Arch. Dermatol., 118, 346, 1982. 9. Horio, T. and Fujigaki, K., Augmentation spectrum in solar urticaria, J. Am. Acad. Dermatol., 18, 1189, 1988. 10. Epstein, S., “Masked” photopatch tests, J. Invest. Dermatol., 41, 369, 1963. 11. Willis, I. and Kligman, A.M., Photocontact allergic reactions: elicitation by low doses of long ultraviolet rays, Arch. Dermatol., 100, 535, 1969. 12. Willis, I. and Kligman, A.M., The mechanism of photoallergic contact dermatitis, J. Invest. Dermatol., 51, 378, 1968. 13. Horio, T., Solar urticaria: sun, skin and serum, Photodermatology, 4, 115, 1987. 14. Horio, T. and Minami, K., Solar urticaria: photoallergen in a patient’s serum, Arch. Dermatol., 113, 157, 1977. 15. Horio, T., Photoallergic urticaria induced by visible light: additional cases and further studies, Arch. Dermatol., 114, 1761, 1978. 16. Horio, T., Solar urticaria: idiopathic? Photodermatol. Photoimmunol. Photomed., 19, 147, 2003.
Index A Abnormal skin, 949, see also Normal anatomy and skin Abscesses, ultrasound imaging, 524 Absolute recovery, 444 ACAIA, see Automated computer-assisted image analysis (ACAIA) Acanthotic epidermis, 251, 252 Accuracy capacitance measurement, 339 high-frequency electrical conductance measurement, 331 measurements, semiopen systems TEWL, 389 statistical analysis example, 54 stratum corneum hydration state, 355 Accurate estimates, 54 Accurate positioning, images, 261 Acid mantle, see pH, skin surface Acid stratum corneum buffer system, 412 Acne fiber-optic microscopy, 131–132 Hi-scope system, 129–130 skin disease assessment guidelines, 938, 939 ultrasound imaging, 482 Acne estivalis, 35 Acoustic output, 516 Acquisition, images basics, 97 confocal scanning laser microscopy, 278–279 SIAscopy, 320 Actinic aging, 607–608 Active contact thermometry, 773–774 Active follicles, 862–864 Acute contact dermatitis, 270 Acute reaction testing, 950 Adenosine triphosphate (ATP), 531–532 Adhesive methods, 366–367, 372 Administration, fluorescent markers, 287–290 Agache, Pierre, 12 Agarose, 429–430 Aging and age differences, see also Children; Infants B-mode imaging, 498 capacitance measurement, 341 capillaroscopy and videocapillaroscopy, 683 characterizing appearance differences, 96, 97 conductance levels, 332 Cutometer, 582 DermFlex, 574 echogenicity, 480–481, 512–513 follicular biopsy, 826 indentometry, 618–620 levarometry, 614–615 pH, 414 sebum excretion, 844 skin integument, 27–29 skin thickness, 512–513 sodium laurel sulfate, 948 tribological studies, 222–223 twistometry, 605–608 ultrasound imaging, 476–477, 511–513
Air-bearing table, 171 Air convections, 386, 390 Air temperature and humidity, see Humidity; Temperature Algorithm, laser profilometry, 174–175 Allergic patch test reaction, 482, 500, 501 Allergies contact dermatitis comparison, 6 plants, 35 preservatives, 35 Allergology, 449–450 Allodynia, 796–797 Alloknesis, 796 Alopecia areata, 877 Ambient air humidity and temperature, see Humidity; Temperature American Academy of Dermatology, 20 A-mode scanning, 474 Amonton’s law, 215–216 AM 400 WB NMR unit, 539 Anagen classification, 871–872, 875, 877 Analog camera, 959–960 AnalySIS, 248 Analysis, see also Image analysis; Statistical analysis drug levels, follicular biopsy, 827 furrows and wrinkles, image analysis, 157, 160 hair growth measurements, 890–892 sweat, 818–819 Analysis, image desquamation rate, 365 dry skin and scaling, 375–378 fluorescence photography, sebaceous follicles, 857–858 follicular biopsy, 828 nail surface, 925–927 Sebutapes, 836–837 thermal imaging, skin temperature, 778 ultrasound imaging, 477–478 Analytical Morphological of the Arbeitsgemeinschaft Dermatologische Foschung, 119 Analyzer instrumentation, 156–157 Anatomical region sodium laurel sulfate, 948 tribological studies, 222–223 Androgenetic alopecia, 876 Anechoic structures, 516 Angiomatosis, 526 Angular errors, 720–721 Animal studies fluorescein sodium, 287, 290 hair, 875 lymph flow evaluation, 744 magnetic resonance spectroscopy, 531 melanocytes, 299 microdialysis sampling, 446 sodium lauryl sulfate, 502 trauma effect, skin, 447 ultrasound imaging, 487 Anisotropic spectral filtering, 198–200 Anisotropy, 516 Antibacterial actions, 858–859
997
998
Handbook of Non-Invasive Methods and the Skin, Second Edition
Antigravity suit, 575 Antimicrobial function, pH, 415 Apocrine sweat collection, 821–822 Apparatus improvements, 180–182 Applications colorimetry, 641–645 confocal microscopy, product research and development, 299–302 digital Cutometer, 589 diseased surface areas, 960–964 dynamic capillaroscopy, 674–676 erythema and melanin indices, 669–670 fluorescent markers, 286 Hi-scope system, 129–130 laser profilometry, 177 nuclear magnetic resonance, 308–312 optical coherence tomography, 262–264 SIAscopy, 320–321 sodium laurel sulfate, 950 TapeAnalyzer, 833–834 thermal imaging, skin temperature, 778–782 transcutaneous oxygen tension measurements, 400–402 twistometry, 609–610 ultrasound imaging, 481–487, 496–497 AquaFlux system, 393–395 Archiving, dermatoscopy, 122 Area under curve (AUC), 57–58 Argon lasers, 794 Arterial flow, ultrasound imaging, 522 Arterial insufficiency, 703, 704 Arterial ischemia, 486 Arteries, 520 Arteriosclerosis, 675–676 Arteriovenous hemangiomas, 526 Artificial dichotomy, 56 Aspartate aminotransferase (AST) test, 53–59 Assessment color calibration, 656–657 follicular transport, 861–865 microcirculation, 684 nail surface image analysis, 926 Assessment, cutaneous pain allodynia, 796–797 basics, 787–788, 794–795, 797 chemical stimulation, 792 cold stimulation, 792 electrical stimulation, 790–791 electrophysiological methods, 795 heat stimulation, 792–794 hyperalgesia, 796–797 ideal cutaneous stimulator, 788–790 induction, cutaneous pain, 789, 790–794 mechanical stimulation, 791 psychophysical methods, 795 Assessment, guidelines acne, 938, 939 atopic dermatitis, 935–936 atopic eczema, 936–938 basics, 931–932, 938, 940 Children’s Dermatology Life Index, 935 Dermatology Life Quality Index, 935 dermatology quality, instruments, 935 disease-specific assessment, 936–938 disease-specific dermatology quality, 935–936 generic quality, instruments, 935 instruments, 934–935 36-Item Short Form, 934–935
Medical Outcomes Study, 934–935 Nottingham Health Profile, 934 psoriasis, 938 Psoriasis Index of Quality of Life, 936 reliability, 933 responsiveness, 933–934 sensitivity, 933–934 Sickness Impact Profile, 934 Skindex, 935 validity, 932–933 AST, see Aspartate aminotransferase (AST) test Atopic dermatitis, see also Contact dermatitis basics, 42 bioengineering techniques, 43 capacitance, 42–43 conductance, 42–43 dermatoscope, 116 diseased surface areas, 961–964 irritants reactivity, 43 skin disease assessment guidelines, 935–936 transepidermal water loss, 42 Atopic eczema, 936–938 Atopic skin, dermatoscope, 116 ATP, see Adenosine triphosphate (ATP) Atraumatic local labeling, 734, 734 Attenuation, ultrasound imaging, 516 AUC, see Area under curve (AUC) Auspitz’s sign, 105 Authoritative references, method selection, 11–12 Authorship, web pages, 25 Automated assessment, melanoma diagnosis, 138–140 Automated computer-assisted image analysis (ACAIA), 891 Automation, furrows and wrinkles, 158–159 Autonomy, 74 Autoregressive modeling, 702 Axial positioning, 479
B Background, photography, 83, 85 Bacteria, sampling available methods, 458–463 basics, 457 correlations, 460, 463 follicular sampling methods, 462–463 impression methods, 459 method selection factors, 457–458 recommendations, 460, 463–465 replica methods, 459 swabbing methods, 460–461 washing methods, 461–462 Bacterial infections, environmental influences, 35 Ballistometry, 627–632 Bandwidth, ultrasound imaging, 516 Basal cell carcinoma high-resolution sonography, 248 magnetic resonance spectroscopy, 541–542 ultrasound imaging, 485 Basal cell layer, 280 Bayes’ theorem, 55, 59 Beau’s lines, 926 Bending measurements, 897 Beneficence, 74 Benign conditions dermatoscopy, 114–116
Index
nevi, 485 tumoral pathology, 525–527 Biased prevalence, 54 Biochemical background, 531–532 Biochemical changes, 827 Bioengineering atopic dermatitis assessment, 43 basics, 3–4, 6–7 contact dermatitis, 6 current issues, 4–5 future directions, 5–6 photoaging, 6 predisease diagnosis, 6 quality management system, 69–72 systemic sclerosis, 41 Bioequivalence potential, 451 Bioethics, 73–74 Bioimpedance, 345–349 Biological variables, 477, 480–482 Biological zero, 710–711 Biomechanics, skin Langer’s lines, 566 racial differences, 29–30 twistometry, 608–609 Biphasic flow, ultrasound imaging, 520 Blackbody, 756 Blackheads, 462 Blanching effect, 642–643 Bleaching effect, 645 Blinding, investigator responsibilities, 50 Blisters, 434–436 Blood clearance, lymph flow evaluation, 747 Blood flow, 38–39, 738 Blood flux, laser Doppler measurement, 691–695 Blood studies, skin blisters, 436 B-mode scanning, 474 Body hair growth, 869–870 Body region, skin integument, 27–29 Boltzmann’s constant, 756 Border effects correction, 655–656 Bowen’s disease, 912 Brevibacterium spp., 458 Broken-off hairs, 878 Bullous disorders, 524 Bürger model, 604 Burn injury cutaneous pain assessment, 796 semiopen systems, transepidermal water loss, 384 ultrasound imaging, 484 Burte-Halsey model, 896
C Cable, fiber-optic microscopy, 127 Café au lait spots, 113 Calcinosis, 524 Calcipotriol, 40, 364 Calculations, microdialysis sampling, 448–449 Calibrated scoring systems, 886, 889 Calibration closed-chamber systems, TEWL, 394–395 electrodes, transcutaneous PCO2, 409 laser Doppler flowmetry, 710–711 laser profilometry, 171–172 measurements, semiopen systems TEWL, 389
999
Calibration, dermascopic images assessment of γ, 656–657 basics, 661 border effects correction, 655–656 illumination, 655–656 importance, 653–654 instruments characteristics, 654–659 multi-instrument calibration, 659–661 RGB/XYZ conversion, 657–658 video camera’s physical properties, 654–655 XYZ conversions, 657–659 Cameras, see also Compact digital cameras; Computerized image analysis analog, 959–960 digital, 958–959 photography, 87 Candida spp., 467–468 Capacitance atopic dermatitis assessment, 42–43 basics, 337–338, 343 corneometer, 338–341 dermato-cosmetic applications, 341 diseases, 341 lesions, 341 measurements, 339–341 MoistureMeter, 341 moisturizing products, 342–343 objectives, 338 oscillating method, 225 probe, 338, 339 Capilab Toolbox, 684–685 Capillaries detection, microcirculation, 684 hemangiomas, 526 morphology, 681–682 network analysis, 684–685 Capillaroscopy, 673–676, 679–680 Capillary blood cell velocity (CBV), 673–674 Capsaicin, 796 Capture software, 128 Carbon replication, 148 Carcinoma, basal cell, 248 Carpal tunnel syndrome, 525 Case studies, Raman spectroscopy, 556, 558–559 Casual state, sebum excretion, 844 Catagen classification, 871–872, 875–877 Cavernous hemangiomas, 526 CBF, see Cutaneous blood flow (CBF) CDLQI, see Children’s Dermatology Life Index (CDLQI) CDSS, see Coefficient of developed skin surface (CDSS) Chamber technique, desquamation rate, 362 Change dynamics, 402 Characteristic life (alpha) value, 905–906 Chemical indicators, 429 Chemical relaxation methods, 897 Chemical stimulation, 792 Children, see also Aging and age differences; Infants atopic dermatitis, 42–43 conductance levels, 332 indentometry, 618–620 Pityrosporum ovale, 469 positioning, 86–87 skin surface pH, 414 Children’s Dermatology Life Index (CDLQI), 935 Chi-square test, dichotomous data, 65 Cholesterol metabolism, 424
1000
Handbook of Non-Invasive Methods and the Skin, Second Edition
Chromameters, 71, 968, see also Minolta Chromameter Chronic ulcer, dermatoscope, 118 CIE, see Commission Internationale de l’Eclairage (CIE) color system CIELAB, 636, 643 CIE-L*a*b* system, 651, 669 Circadian rhythm, 414 Clarity, 11 Classifications dysplastic/dystrophic hairs, 871–872 FDA rules, 12 global vision and imaging systems, 884, 886, 888 Clemson University Physics Online Laboratory (CUPOL), 20–21 Closed-chamber systems, transepidermal water loss, 393–396 Closed patch tests, 945–946 Clothing, photography, 86 C-mode scanning, 474 Coefficient of developed skin surface (CDSS), 158 Cohesion, desquamation rate, 367–368 CO2 lasers, 794 Cold, cutaneous pain assessment, 796 Cold provocation, 674 Cold stimulation, 792 Cole equations, 345, 347 COLIPA, see European Cosmetic Toiletry and Perfumery Association (COLIPA) Collection techniques desquamation rate, 362–363 sebum collection, 832 Color, skin facial averaging methods, 96 SIAscopy, 318–319 wheals-and-flare reactions, 968–969 Color calibration, dermascopic images assessment of γ, 656–657 basics, 661 border effects correction, 655–656 illumination, 655–656 importance, 653–654 instruments characteristics, 654–659 multi-instrument calibration, 659–661 RGB/XYZ conversion, 657–658 video camera’s physical properties, 654–655 XYZ conversions, 657–659 ColorChecker, 657, 659–660 Colorimeter technical details, 637–638 Colorimetric methods, 974–975 Colorimetry applications, 641–645 basics, 635–636, 645 blanching effect, 642–643 bleaching effect, 645 CIE color system, 636–637 colorimeter technical details, 637–638 contact allergens, 642 correlations, 639–640 corticosteroids, 642–643 depigmenting agents, 645 dose-response curves, 644 error sources, 639 erythema, 641–642, 644 instruments, 639 irritants, 641–642 methods, 636–638 objective, 636 PASI, 39 pigmentation, 643–644
reproducibility, 639 skin color, 643 ultraviolet radiation, 637, 641–642, 643–644 Colors assessment, melanoma, 135–142 Color temperature, 83 Comedogenic evaluations, 828 Comedogenicity, 858–859 Comedolytic evaluations and potential, 826–828 Comedone extractors and extraction, 460, 462, 464 Commercial electrical measurement instruments accuracy, 355 basics, 351, 356 corneometers, 352–353 correlations, 355 DermaLab moisture unit, 353 measurement depth, 356 MoistureMeters, 353 multifrequency impedance spectrometer, 354 Nova Dermal Phase Meters, 353–354 principle, 351–352 sensitivity range, 355–356 Skicon 200/200EX, 354 SkinChip, 354 variation coefficient, 356 Commission Internationale de l’Eclairage (CIE) color system, 636–637, 984 Compact digital cameras, 89–93 Competent consent, 76 Compression, 522 Computation, graphs, 316, 319, 322 Computer, 165 Computer assessment, melanoma, 136–142 Computer image analysis, 959–964 Computerized and photographic hair growth measurements analytical methods, 890–892 basics, 883, 892–893 function, hair, 884–886 global vision and imaging methods, 888–889 improvements, 887–888 photography, 884–892 standards, 884, 886–887 structure, hair, 883–884 Computerized image analysis, 95–99 Computerized laser capillary microscopy, 673–674 Computerized suction cup, DermaLab, 596 Computer systems, surface skin temperature, 764 Conditional independence, 59 Conditional probabilities, 54 Conductance, 42–43, 329–334 Conduction, 772–775 Confocal microscopy, fluorescent markers administration, 287–290 applications, 286 basics, 285–286 fluorophore selection, 286–287 functional imaging, 291–292 future directions, 292–293 green fluorescent protein expression, 292 immunofluorescence labeling, 291 intradermal administration, 287, 288, 289–290 intravenous administration, 290 skin barrier function, 291, 292 thermal burn-induced autofluorescence, 292 topical application, 287 transdermal drug delivery, 291 Confocal microscopy, product research and development
Index
basics, 297 epidermal thickness measurement, 298–300 images application, 299–302 melanin distribution visualization, 299, 300 methodological principles, 298–299 objectives, 297–298 reflectance confocal microscopy, 297 skin research suitability, 302–303, 305 sunscreen agent evaluation, 300, 302, 303–304 ultraviolet radiation protection, 300, 302, 303–304 Confocal microscopy, reflectance acute contact dermatitis, 270 basics, 268, 274–275 cutaneous infections, 270–271 difficulties, 274 inflammatory conditions, 270–271 issues, 274 margin assessment, 274 neoplastic skin lesions, 271–274 nonpigmented lesions, 271–272 normal skin, 269–270 pigmented lesions, 272–274 psoriasis, 271 reflectance principles, 268–269 solutions, potential, 274 therapy adjunct, 274 treatment response evaluation, 274 Confocal scanning laser microscopy (CSLM), 41, 277–282 Connective tissue diseases dynamic capillaroscopy, 675 twistometry, 610 ultrasound imaging, 483–485 Consent, 50–51, 74–76, 82 Consistency, 86, 88 Constant velocity, 216 Constitutional factors, SLS, 947–949 Construct validity, 932 Consumer evaluation, 230 Contact allergens, colorimetry, 642 Contact dermatitis, 6, 949, see also Atopic dermatitis Contact Dermatitis, 12 Contact plates, bacteria sampling, 459–460 Contact pressure, semiopen systems TEWL, 388 Contact temperature measurement, 754 Content validity, 932 Continuous data, t-test, 64–65 Continuous variables, 932 Contours and wrinkles, methods comparison advantages, 211–212 basics, 205, 212 contours, 209 FOITS8, 208 image analysis, 206–207, 209 instruments, 206–209 limitations, 211–212 measurement, 206–207 optical profilometry, 209 parameters, 208–209 PRIMOS7, 208 replicas, 206–208, 211 shadowing method, 207 three-dimensional analysis, 208–209 in vitro measurement, 208, 211 wrinkles, 207, 209, 210 Control, laser profilometry, 170–171 Control unit, Hi-scope system, 128
1001
Convection, thermal imaging, 772 Convergent validity, 933 Copper vapor lasers, 794 Corneocytes, CSSS, 240 Corneometers capacitance measurement, 338–341 changing instruments, 71 epidermal capacitance, 338–341 sensitivity, 355–356 skin friction measurement, 227 stratum corneum hydration state, 332, 352–353 Corpora aliena, 116 Correction, verification bias, 54, 57, 58–60 Correlation and factor analysis, 933 Correlations, other methods bacteria sampling, 460, 463, 463 colorimetry, 639–640 confocal scanning laser microscopy, 278–279 dry skin and scaling evaluation, D-Squames and image analysis, 377–378 erythema and melanin indices, 668–669 furrows and wrinkles, image analysis, 160–161 high-frequency electrical conductance measurement, 334 Langer’s lines, 566–568 laser Doppler methods, 694–695, 721 Levarometry, 615 lymph flow evaluation, 749 magnification, 106 nail thickness, 924 nuclear magnetic resonance, 313 quasi-L*a*b* color measurement, 650–651 skin chamber techniques, 439 skin replication, 148–149 sodium laurel sulfate, 947 stratum corneum hydration state, 355 stratum corneum material, harvesting, 372 stylus method, contour measurement, 166–167 sweat gland localization, 808 TapeAnalyzer, 832–833 transcutaneous oxygen tension measurements, 399–400 transcutaneous PCO2, 409 trichograms, 879–880 ultrasound imaging, 475–476 133Xenon wash-out technique, 739 Cortex DermaSpectrometer, 666–668, 975 Corticosteroids atrophy, 483 colorimetry, 642–643 ultrasound imaging, 501 Corynebacterium spp., 458 Cosmetics magnetic resonance spectroscopy, 533 microcirculation, 685 regulation, 5 Costs, test evaluations, 61 Couperose assessment, 685, 686 Courage-Khazaka software, 584–585, 588 Creams, measurements, 385 Creeping, Cutometer, 581, 583 Criterion validity, 932 Critical value, statistical analysis, 56 Cronbach’s coefficient, 933 Cross-sectional view, 518 Crow’s feet, 114–115 CSSS, see Cyanoacrylate skin surface stripping (CSSS) Cultural habits, 35
1002
Handbook of Non-Invasive Methods and the Skin, Second Edition
Cumulative reaction testing, 951 CUPOL, see Clemson University Physics Online Laboratory (CUPOL) Current issues, bioengineering, 4–5 Curved arrays, ultrasound imaging, 516 Cutaneous and subcutaneous blood flow rates, 133Xenon wash-out technique atraumatic local labeling, 734 basics, 733 blood flow rate calculation, 738 correlations, 739 data management, 734–735 error sources, 739 loss, 738 methodological principle, 733–738 objectives, 733 physical properties, 733–734 recommendations, 739–740 registration, 734–735 washout model, 735–737 Cutaneous blood flow (CBF) heat wash-in, heat wash-out, 723–730 laser Doppler measurement, 692–694 occluded patch tests reactions, 975–976 periodic fluctuations, 697–704 Cutaneous infections, 270–271 Cutaneous lymphoma, 486 Cutaneous melanoma, 528 Cutaneous neoplasms, 241, 485–486 Cutaneous pain assessment allodynia, 796–797 basics, 787–788, 794–795, 797 chemical stimulation, 792 cold stimulation, 792 electrical stimulation, 790–791 electrophysiological methods, 795 heat stimulation, 792–794 hyperalgesia, 796–797 ideal cutaneous stimulator, 788–790 induction, cutaneous pain, 789, 790–794 mechanical stimulation, 791 psychophysical methods, 795 Cutaneous pigmented lesions, 121 Cutaneous vessel alterations, 116–118 Cutometers basics, 579–582 DermFlex, 571–572 skin friction, 227 Cyanoacrylate biopsy, cytologic evaluation, 239–241 Cyanoacrylate cement and glue, 372, 460, 462–463 Cyanoacrylate skin surface stripping (CSSS), 239–241 cyberDerm software, 593, 596 Cyberware scanning, 98 Cyclic tester, hair strength evaluation, 904 Cystic hygroma, 526 Cystic lesions, 525–526 Cystic lymphangiomas, 526 Cytologic evaluation, cyanoacrylate biopsy, 239–241
D Dage-MTI CCD72, 376 Dark areas, melanoma, 136–137 Data ballistometry, 629, 632 collection, laser profilometry, 173
error, 11 hair strength evaluation, 905–906 interpretation, Raman spectroscopy, 554–556 laser Doppler imaging, 718–719 lymph flow evaluation, 743–744 management, 133Xenon wash-out technique, 734–735 reduction, gas-bearing electrodynamometer, 624–625 sample size calculation, 64–65 twistometry, 605 ultrasound imaging, 518 verification, monitors, 49 Dawson’s erythema index, 668 Declaration of Helsinki, 12, 47, 74–75 Decreased skin temperature, 780–782 Deeper epidermis, 102 Deformation reproducibility, 587 Delaunay triangulation, 684 Delegation, duties, 50 Delta 10 dermatoscope, 111–112 Deming cycle, 69, 70 Demodex folliculorum, 826 Depigmenting agents, 645 Depot clearance, 743 Depth of field, 82–83, 127 Dermaflex, 571–576 DermaLab, see also Semiopen systems, transepidermal water loss epidermal capacitance, 338 sensitivity, 355–356 stratum corneum hydration state, 353 Dermal alterations, 102 Dermal disease, ultrasound imaging, 522, 524 Dermal echoes, 476 Dermal-epidermal junction, 280, 280 Dermal layer, ultrasound imaging, 520 Dermal papillae, 280–281 Dermal Torque Meter, 610 Dermal water and edema, ultrasound assessment, 507–509 Dermascan C scanner B-mode imaging, 494–495 image analysis, 478 variation, 479 Dermascopic correlations, 278–279 Dermascopic images, color calibration assessment of γ, 656–657 basics, 661 border effects correction, 655–656 illumination, 655–656 importance, 653–654 instruments characteristics, 654–659 multi-instrument calibration, 659–661 RGB/XYZ conversion, 657–658 video camera’s physical properties, 654–655 XYZ conversions, 657–659 DermaSpectrometer, 666–668, 975 Dermatitis, sodium laurel sulfate, 949, see also Contact dermatitis Dermatitis Family Impact Questionnaire (DFI), 935–936 Dermato-cosmetic applications, 341 Dermatoglyphics, 115 Dermatologic surgery, 532 Dermatology Life Quality Index (DLQI), 935 Dermatology quality, instruments, 935 Dermatopharmacology, 533 Dermatophyte infections, 271 Dermatoscopy archiving, 122 basics, 109
Index
benign conditions, 114–116 cutaneous vessel alterations, 116–118 follow-up, 122 infrared photography, 112–113 instrumentation, 110–112 malignant conditions, 118–121 objective, 109–110 surface microscopy, 114 ultraviolet photography, 113 validation, 121 Dermavision 2D, 511 Dermis, 310–312, 518 DermNet web page, 16, 19 Dermoid cysts, 525 Desquamation rate adhesive methods, 366–367 basics, 361–362, 367 cohesion, 367–368 collection techniques, 362–363 image analysis technique, 365 measurement techniques, 362–363 microscopic techniques, 365–366 photometric techniques, 363–364 staining techniques, 363–366 tape methods, 366–367 visual techniques, 363–364 Detectors, 760–764 Detergents, 423 Detergent scrub technique, 460, 461, 462 DFI, see Dermatitis Family Impact Questionnaire (DFI) Diabetes type II dynamic capillaroscopy, 675 laser Doppler nail assessment, 914 skin surface pH, 417 Dialysate analysis, 445, 446 Diapers, 414, 416–417 Dia-stron, 900, 904 Dichotomous data, 65 Difference method, 448 Difference to detect, 63, 65 Differentiation, ultrasound imaging, 522 Difficulties, reflectance confocal microscopy, 274 Diffuse effuvium, 876 Digital cameras, 89–93, 958–959, see also Cameras; Computerized image analysis Digital Cutometer, 583–589 Diode lasers, 551 Direct gravimetric technique, 849–850 Directional extraction, 199–200 Direct light, measurements, 390 Direct tracings, 959 Disc attachment, skin, 605 Disease bioimpedance, 347–348 capacitance measurement, 341 confocal scanning laser microscopy, 277–278 dermatology quality, 935–936 extent, 38, 241 sebum excretion, 844–845 ultrasound imaging, 521–522, 524 vascular, 522–523 Disease assessment acne, 938, 939 atopic dermatitis, 935–936 atopic eczema, 936–938 basics, 931–932, 938, 940
1003
Children’s Dermatology Life Index, 935 Dermatology Life Quality Index, 935 dermatology quality, instruments, 935 disease-specific assessment, 936–938 disease-specific dermatology quality, 935–936 generic quality, instruments, 935 instruments, 934–935 36-Item Short Form, 934–935 Medical Outcomes Study, 934–935 Nottingham Health Profile, 934 psoriasis, 938 Psoriasis Index of Quality of Life, 936 reliability, 933 responsiveness, 933–934 sensitivity, 933–934 Sickness Impact Profile, 934 Skindex, 935 skin disease assessment guidelines, 936–938 validity, 932–933 Diseased surface areas, instrument and computer-based methods comparison analog camera, 959–960 applications, 960–964 atopic dermatitis, 961–964 basics, 957–958 digital camera, 958–959 direct tracings, 959 future directions, 964–965 limitations, 964–965 pitfalls, 964–965 rule of nines, 959, 960 Disproportional superficial strain, 571 Distal landmarks, nails, 920–921 Distance errors, laser Doppler imaging, 720–721 Distensibility, DermFlex, 575 Distributed multimedia hypermedia, 15 Distribution volume, 748 Dithranol, 40 Diurnal variation, 480, 844 Divergent validity, 933 DLQI, see Dermatology Life Quality Index (DLQI) Doctors, prescribing habits, 35 Documents, essential, 49 Doppler methods, see also Laser Doppler methods measurement, skin blood flux, 691–695 optical coherence tomography, 264 ultrasound imaging, 517 Dose prediction, see Light sources, sunlight, and radiation dosimetry Dose-response curves, 644 Dosimetry, 984, see also Light sources, sunlight, and radiation dosimetry Dr. Lange Micro Color instrument, 638–640 Drug level analysis, 827 Drug phototesting, 995–996 Dry, scaly lesions, 341 Dryness, sodium laurel sulfate, 949 Dry skin and scaling evaluation, D-Squames and image analysis, 375–378 DSC-707 camera, 958 D-Squames desquamation rate, 365–366 image analysis, dry skin and scaling, 375–378 method, 366 Sebutape, 842 stratum corneum material, harvesting, 372–373 tape striping, 425 DSR, see Dynamic spring rate (DSR)
1004
Handbook of Non-Invasive Methods and the Skin, Second Edition
DualTape testers, 831–832, 834 Dupuytren contracture, 527 Dyes, 422, 742 Dynamic capillaroscopy, 673–676 Dynamic microtopography, 926 Dynamic movements, 519 Dynamic range, 517 Dynamic spring rate (DSR), 624 Dynamometer, 621 Dysplastic hairs, 877 Dysplastic nevi, 485 Dystrophic hairs, 877
E EASI, see Eczema Area and Severity Index (EASI) Eccrine sweat collection, 817–819 Echogenicity age differences, 512–513 measurements, 41 Echographic characterization, 499, see also Ultrasound B-mode imaging, in vivo structure analysis Echographic methods, 970, 971 Echo poor/rich band, 248–251 Ectatic vessels, dermatoscope, 117–118 Ectodermal dysplasia, 115, 116 Eczema, dermatoscope, 117 Eczema Area and Severity Index (EASI), 938 Eczematous diseases, pH, 416 Edema, 507–509, 519–520, 523–524 Educational courses, Internet, 20–21 EEMCO, see European Group for Efficacy Measurements on Cosmetics and Other Topical Products (EEMCO) Ehlers-Danlos syndrome, 576, 610 EIT, see Electrical impedance tomography (EIT) Elasticity, skin, 40, 571, 575 Elasticity, Twistometry actinic aging, 607–608 aging, 605–608 applications, 609–610 basics, 601–602, 610 connective tissue diseases, 610 equipment, 602 interpretation, 605 intrinsic aging, 605–607 mechanical testing, 602–603 parameters, mechanical, 603–605 retinoic acid, topical, 609 scleroderma, 609–610 skin biomechanics, 608–609 stratum corneum, 608–609 technical considerations, 602–605 testing, mechanical, 602–603 theoretical considerations, 602–605 topical retinoic acid, 609 validation, 605 Elasticity parameters, Twistometry, 603–604 Elastosis, photoaging, 6 Electrical conductance measurement, 329–334 Electrical impedance tomography (EIT), 349 Electrical measurement instruments accuracy, 355 basics, 351, 356 corneometers, 352–353 correlations, 355
DermaLab moisture unit, 353 measurement depth, 356 MoistureMeters, 353 multifrequency impedance spectrometer, 354 Nova Dermal Phase Meters, 353–354 principle, 351–352 sensitivity range, 355–356 Skicon 200/200EX, 354 SkinChip, 354 variation coefficient, 356 Electrical stimulation, 790–791 Electrochemical electrode, 408–409, see also Probes Electrodes, see also Probes calibration, 409 transcutaneous PCO2, 408–409 Electrodynamometer, gas-bearing (GBE), 621–625 Electronic flash, see Lighting (photography) Electrophysiological methods, 795 Elevations, Levarometry, 615 Emerging technology, 764 Emollients high-frequency electrical conductance measurement, 334 tribological studies, 217, 220–221 Emulsion application results, 228–229 Endocrine status, 574, 844 Endogenous factors influence, 413–414 Environmental influences, 33–35, 340 Environment-related variables measurements, semiopen systems TEWL, 387–390 sodium laurel sulfate, 949 Enzymes, pH evaluation, 423–425 Epidermal capacitance, measurement basics, 337–338, 343 corneometer, 338–341 dermato-cosmetic applications, 341 diseases, 341 lesions, 341 measurements, 339–341 MoistureMeter, 341 moisturizing products, 342–343 objectives, 338 probe, 338, 339 Epidermal disease, ultrasound imaging, 521–522, 524 Epidermal echoes, 476 Epidermal hydration, 329–334 Epidermal thickness measurement, 298–300 Epidermis, ultrasound imaging, 522 Epidermis examination, nuclear magnetic resonance basics, 307 characterization, 310–311 correlations, 313 error sources, 312–313 methodological principle, 308–312 objectives, 307–308 recommendations, 313 skin layer parameters, 310–312 in vivo high-resolution imaging, 308–310 in vivo measurements, 310–312 Epiluminescence microscopy (ELM), 119–121 Equipment contour measurement, 164–165 Dermaflex, 572–573 handling, stylus method, 166 nuclear magnetic resonance, 308, 309 suction chamber method, 572–573 twistometry, 602
Index
ultrasound imaging, 478–480, 494–495 Error sources basics, 11 colorimetry, 639 erythema and melanin indices, 667–668 furrows and wrinkles, image analysis, 159–160 hair strength measurement, 898–899 high-frequency electrical conductance measurement, 331–332 Langer’s lines, 568 laser Doppler imaging, 720–721 lymph flow evaluation, 747–748 magnetic resonance spectroscopy, 534 measurements, semiopen systems TEWL, 387–391 nuclear magnetic resonance, 312–313 oil immersion examination, 105–106 pH, 413 quasi-L*a*b* color measurement, 650 skin chamber techniques, 436–439 skin replication, 148–150 stratum corneum material, harvesting, 372 stylus method, contour measurement, 166 sweat gland localization, 807–808 transcutaneous oxygen tension measurements, 399 transcutaneous PCO2, 409 trichograms, 878–879 133Xenon wash-out technique, 739 Erythema colorimetry, 641–642, 644 index, 665–670 psoriasis assessment, 38–39 Erythrosis assessment, 685 Essential documents, 49 Ethical considerations, 73–76, 82 Ethics committees, 50 European Contact Dermatitis Society, 693 European Cosmetic Toiletry and Perfumery Association (COLIPA), 12 European Group for Efficacy Measurements on Cosmetics and Other Topical Products (EEMCO), 12, 395 European Society of Contact Dermatitis (ESCD) abnormal skin, 949 applications, 950 basics, 944 characteristics, 944 clinical effects, 944–945 closed patch tests, 945–946 constitutional factors, 947–949 correlations, 947 effects, 944–945 environment-related variables, 949 exposure methods, 945–947 histopathological effects, 944 immersion tests, 946–947 immunological effects, 944 individual pilot study, 951 individuals, 949–950 location, 949–950 occlusive tests, 945–946 one-time tests, 945–946 open tests, 946 pilot study, 951 reactivity, 947–950 recommendations, 949–952 repeated tests, 946 results interpretation, 952 testing, 949–951 transepidermal water loss, 395, 974
1005
variables, 949 wash tests, 947 Evaporimeter, see Semiopen systems, transepidermal water loss Evidence-based medicine, 9 Evoked potentials, 795 Exaggeration of benefits, 81 Examination room, 777–778 Examiner, ultrasound imaging, 478–480 Excretion, sebum measurement, 831–834 Sebumeter, 843–844 TapeAnalyzer, 833–834 Exogenous factors influence, 413–414 Experimental designs, tribological studies, 216–217 Experimental setup, microdialysis sampling, 446–448 Expert evaluation, 229–230 Expert systems, 764 Explanatory mistake, 11 Exposure methods, sodium laurel sulfate, 945–947 External environmental factors influence, 340, 413–414 Extrinsic forces, lymph flow evaluation, 748 Ex vivo/in vivo follicular transport assessment, 861–865
F Face validity, 932 Facial averaging methods, 96 Facial sagging, measuring improvement, 99 False-color renditions, 113 False positives and negatives, 54 Fast Fourier transform, 700–702, see also Fourier transform Fast start-up, 91 Fauna, 35 FDA, see Food and Drug Administration (FDA) Female differences, see Sex differences Fiber dimensions, 904 Fiber-optic microscopy acne, 129–130 application, 129–130 basics, 125–126 cable, 127 capture software, 128 control unit, 128 depth of field, 127 future directions, 133 good practices, 133 hairs, 130, 133 hi-scope system, 128–130 horizontal illumination, 128–129 image grabber board, 128 light-gathering power, 126–127 light source, 127 optical head, 128 optics, 126–127 photodamage, 129 polarized light imaging, 129 rosacea, 129, 131 scaliness, 130, 132 surface and subsurface imaging theory, 127 vertical illumination, 128–129 video imaging, 126 working distance, 126 wound healing, 130, 132 Fiber type variations, 897 Fibromatous tumors, 527
1006
Handbook of Non-Invasive Methods and the Skin, Second Edition
Fick’s diffusion law, 385 File format, photography common formats, 92 projection images, 88 sebutape storage, 838 web sites, 89 Filenames, 15–16 Film, photography, 87 Filtering eccrine sweat collection, 818 Sebutapes, 838–839 stylus method equipment, 165 Final reports, 51 Finn chamber, 247 FITC-Dextran, 292, 742 5 M’s (man, machine, methods, materials, and marginal conditions), 72 Fixed landmarks, nails, 920–921 Fixed monitoring systems, 757 Flash, see Lighting (photography) Flashpoint 128, 128 Flat pH glass electrode, 422 Flexabrasion method, 905 Fluid accumulation Raman spectroscopy, 558–559 ultrasound imaging, 519–520, 523–524 Fluid movement, measurement, 508 Fluorescein sodium, 287, 289–290 Fluorescence confocal microscopy administration, 287–290 applications, 286 basics, 285–286 fluorophore selection, 286–287 functional imaging, 291–292 future directions, 292–293 green fluorescent protein expression, 292 immunofluorescence labeling, 291 intradermal administration, 287, 288, 289–290 intravenous administration, 290 skin barrier function, 291, 292 thermal burn-induced autofluorescence, 292 topical application, 287 transdermal drug delivery, 291 Fluorescence microlymphangiography, 742 Fluorescence origin, 854–855 Fluorescence photography, sebaceous follicles antibacterial actions, 858–859 basics, 853 comedogenicity, 858–859 fluorescence origin, 854–855 follicle evaluation, 853–854 high-pressure liquid chromatography preparation, 855–856 image analysis, 857–858 neural algorithms, 857–858 recommendations, 856, 859–860 SAFIR system, 856–857 Fluorescent markers, confocal microscopy administration, 287–290 applications, 286 basics, 285–286 fluorophore selection, 286–287 functional imaging, 291–292 future directions, 292–293 green fluorescent protein expression, 292 immunofluorescence labeling, 291 intradermal administration, 287, 288, 289–290 intravenous administration, 290
skin barrier function, 291, 292 thermal burn-induced autofluorescence, 292 topical application, 287 transdermal drug delivery, 291 Fluorophore selection, 286–287 F80 Nikon camera, 959 Focal pane arrays (FPAs), 757–760 Focused light, cutaneous pain assessment, 793 Focusing system, three-dimensional evaluation, 180–181 FOITS8, 208 Follicle evaluation, 853–854 Follicular biopsy, 825–828 Follicular sampling methods, 462–463 Follicular transport assessment, ex vivo and in vivo, 861–865 Folliculitis, 271 Follow-up, dermatoscopy, 122 Food and Drug Administration (FDA) bioequivalence, 451 classification rules, 12 good clinical practice, 47 medicotechnical instruments, 12 Forced desquamation, 362–363 Forearm skin, ultrasound imaging, 499 Foreign bodies, ultrasound imaging, 521, 524 Formaldehyde allergy, 35 Forward predictive model, 318–319 FotoFinder, 654, 656, 659 Fourier transform, see also Fast Fourier transform morphological tree, cutaneous network of lines, 196–198 optical coherence tomography, 258 skin surface textural analysis, 185 FPAs, see Focal pane arrays (FPAs) Fractional removal rate, 743 Fraud, 51 Free fatty acid pathway, 412 Frequency range, 196–198, 520 Frequency response analysis, 172 Friction coefficient measurement studies age, 222–223 anatomic region, 222–223 basics, 215–216, 222–223 emollients, 217, 220–221 experimental designs, 216–217 friction coefficient values, 218–222 gender differences, 222–223 hydration, 217–219 lubricants, 217, 220 moisturizers, 217, 220–221 normal load, 217–218 probes, 217, 218–219, 221–222 racial differences, 222–223 talcum powder, 220 Friction coefficient values, 218–222 Friction evaluation, unidirectional stress basics, 225 consumer evaluation, 230 emulsion application results, 228–229 expert evaluation, 229–230 methodological principle, 225–226 physiological parameters relationship, 227–228 sensory evaluation, 228–231 skin friction, 226–231 in vivo measurement, 226–227 Full-thickness tears, 525 Function hair, growth measurements, 884–886
Index
stylus method equipment, 165–166 Functional imaging, 291–292 Fungal infections, 35 Fungi, mapping, 467–469 Furrow directional quantification, 185–187 Furrows and wrinkles, replica image analysis analysis, 157, 160 analyzer instrumentation, 156–157 automation, 158–159 basics, 155 correlations, 160–161 error sources, 159–160 image analysis, 156–159 lighting angle, 159–160 measurement parameters, 157–158 methodology, 156–159 objectives, 155–156 recommendations, 161 replica artifacts, 159, 160 shadowing method, 156 Future directions bioengineering, 5–6 diseased surface areas, 964–965 fiber-optic microscopy, 133 fluorescent markers, 292–293 magnetic resonance spectroscopy, 548 microdialysis sampling, 451 phototrichograms, 891–892 Raman spectroscopy, 559–560 SIAscopy, 321 surface skin temperature sensors and handheld devices, 764–765
G Gain and gain setting, 479, 516 Ganglions, 525 Garments, photography, 86 Gas-bearing electrodynamometer (GBE), 621–625 GCP, see Good clinical practice (GCP) standards Gel layer thickness, 479 Gel pad, ultrasound imaging, 518 Gender differences, see also Sex differences levarometry, 614 pH, 414 tribological studies, 222–223 Generic quality, instruments, 935 Generic 2 × 2 table, 54 Genetic factors, sodium laurel sulfate, 948 Glabrous skin, normal, 247, 249–250, 252 Global photographs, 888, 889, 890 Global vision and imaging methods, 888–889 Glomus tumors, 525 GMP, see Good manufacturing practice (GMP) Gold sputtering, 152 Gold standard, 54, 55–56 Good clinical practice (GCP) standards basics, 11–12, 47–48 Declaration of Helsinki, 47 essential documents, 49 ICH-GCP principles, 48–49 investigator responsibilities, 50–52 monitor responsibilities, 49 regulatory requirements, 48 responsibilities and roles, 49–52 sponsor responsibilities, 49
1007
Good laboratory practice (GLP) standards, 11 Good manufacturing practice (GMP), 49 Good practices, fiber-optic microscopy, 133 Google search engine, 21–24 Gorlin’s syndrome, 152 Graphs, SIAscopy, 316, 319, 322 Green fluorescent protein expression, 292 GretagMacbeth ColorChecker Color Rendition Chart, 657, 659–660 Gripping, hair strength measurement, 899–900 Guard ring, Levarometry, 615 Guidelines measurements, 573–574 photography publication, 88 Guidelines, skin disease assessment acne, 938, 939 atopic dermatitis, 935–936 atopic eczema, 936–938 basics, 931–932, 938, 940 Children’s Dermatology Life Index, 935 Dermatology Life Quality Index, 935 dermatology quality, instruments, 935 disease-specific assessment, 936–938 disease-specific dermatology quality, 935–936 generic quality, instruments, 935 instruments, 934–935 36-Item Short Form, 934–935 Medical Outcomes Study, 934–935 Nottingham Health Profile, 934 psoriasis, 938 Psoriasis Index of Quality of Life, 936 reliability, 933 responsiveness, 933–934 sensitivity, 933–934 Sickness Impact Profile, 934 Skindex, 935 validity, 932–933
H Habits, seasonal variations effects, 35 Hair and hair follicles breaking strength, 900 damage, 897 diameter, 872–873, 899 evaluation, strength, 903–907 fiber-optic microscopy, 133 high-resolution sonography, 250 Hi-scope system, 130 mechanical strength, 895–900 nonvisible hair, 892 roots, 871–872, 877–878 scoring systems, 869–870 shaft, 873 strength, 895–900, 903–907 trichogram, 871, 875–880 ultrasound imaging, 521 Hair failure data, 905–906 Hair growth measurements basics, 869 body hair growth, 869–870 classifications, 871–872 diameter, hair, 872–873 hair root status, 871, 872 objective measurements, 870–873 presampling, 870–871
1008
Handbook of Non-Invasive Methods and the Skin, Second Edition
published methods, 873 sampling, 870, 871 scalp hair growth, 869 scoring systems, 869–870, 883–893 shaft, 873 trichograms, 871 unit area trichogram, 871 Hair growth measurements, photographic and computerized techniques analytical methods, 890–892 basics, 883, 892–893 function, hair, 884–886 global vision and imaging methods, 888–889 improvements, 887–888 photography, 884–892 standards, 884, 886–887 structure, hair, 883–884 Handling equipment, stylus method, 166 Haptic finger, 233–236 Hardware, DermaLab, 593–595 Harvesting sweat, 821 H4300 device, 393 Healing theory and evolution, 191 Healthy skin, see Normal anatomy and skin Heat-responding cutaneous nociceptors, 792–793 Heat stimulation, 792–794 Heat wash-in/heat wash-out, cutaneous blood flow applications, 728–730 basics, 723 blood flow rate calculation, 725 correlations, 727 data management, 724, 725 error sources, 726–727, 728 heat loss, 725–726 methodological principles, 724–726 model, 724–725 objectives, 723 physical principles, 724 probes, 724 recommendations, 730 registration, 724, 725 Helsinki Declaration, see Declaration of Helsinki Hemangiomas, 526 Hematomas, 524 Hemodialysis, 417 Hereditary hemorrhagic telangiectasia, 117 Hering’s opponent color theory, 637 Herpes infections, 271 Herringbone nail, 925 High-frequency electrical conductance measurement, 329–334 High-frequency ultrasound, skin examination animal studies, 487 applications, 481–487 basics, 473–474 biological variables, 477, 480–481 connective tissue diseases, 483–485 correlations, 475–476 cutaneous neoplasms, 485–486 equipment, 478–480 examiner, 478–480 histology, 475–476 image analysis, 477–478 inflammatory conditions, 477, 481–483 laboratory facility, 478–480 leg ulcers, 486 nails, 487 normal skin, 474, 476–477
physical principles and techniques, 475 practical guidance, 478–480 radiography, 476 skinfold caliper, 476 ultrasonography, 475–476 variables, 477, 478–481 vascular system, 486 velocity, 475 High-perfusion microangiopathy, 712 High-pressure liquid chromatography (HPLC), 854–856 High-resolution sonography analysis, 247 basics, 246, 251–253 glabrous skin, normal, 247, 249–250 methods, 246–247 palmar skin, normal, 247–250 patients, 247–248 processing, 246–247 results, 248–251 skin diseases, 247–248, 250–251 statistics, 248 three-dimensional sonography, 247 Hippocratic Oath, 74 Hi-scope system, 128–130, 133 Histamine, wheal-and-flare reactions, 967 Histidine-to-urocanic acid pathway, 412–413 Histogram intersection, 660 Histology, ultrasound imaging, 475–476 Histopathological correlations, 278–279 Histopathological effects, 944 Historical background lymph flow evaluation, 742 nail longitudinal growth, 919 thermal imaging, skin temperature, 769–770 HMCC, see Home Made ColorChecker (HMCC) HM-2200 MD, 299 Home Made ColorChecker (HMCC), 659–660 Homeostasis, 450 Hookean region, 896 Hooke’s law, 594, 596–597 Horizontal illumination, 128–129 Horny cells, 376 Horny layer, 248–251 Hough transform, 185 HPLC, see High-pressure liquid chromatography (HPLC) Human body temperature, 770–772 Human studies DermaLab, 598–599 experiment preparation, 446–447 lymph flow evaluation, 744–745 Humidity, 35, 388, 899 Hydration bioimpedance, 346–347 capacitance measurement, 340 high-frequency electrical conductance measurement, 329–334 micro-sensor mapping, 812–813, 815 nuclear magnetic resonance imaging, 312 stratum corneum, 812–813, 815 sweating, 814, 816 tribological studies, 217–219 Hydration state, stratum corneum accuracy, 355 basics, 351, 356 corneometers, 352–353 correlations, 355 DermaLab moisture unit, 353
Index
measurement depth, 356 MoistureMeters, 353 multifrequency impedance spectrometer, 354 Nova Dermal Phase Meters, 353–354 principle, 351–352 sensitivity range, 355–356 Skicon 200/200EX, 354 SkinChip, 354 variation coefficient, 356 Hyperalgesia, 796–797 Hyperechoic structures, 516 Hypertension assessment, 682 Hypoechoic structures, 516 Hyporeactive skin, 948 Hysteresis, 581, 583
I Ichthyosis dermatoscope, 117 pH, 416–417 sodium laurel sulfate, 949 Ideal cutaneous stimulator, 788–790 Identification, Langer’s lines basics, 565–566 correlations, 566–568 error sources, 568 methodological principle, 567–568 multiple simple tension tests, 567 objective, 566–567 qualitative correlates, 566–567 recommendations, 568 skin biomechanics, 566 suction chamber methods, 567–568 IDQOL, see Infants’ Dermatitis Quality of Life Index (IDQOL) Illumination color calibration, 655–656 dry skin and scaling evaluation, D-Squames and image analysis, 377 Sebutapes, 838–839 Image acquisition basics, 97 confocal scanning laser microscopy, 278–279 SIAscopy, 320 Image analysis, see also Analysis; Statistical analysis compact digital camera photography, 89–90 desquamation rate, 365 dry skin and scaling, 375–378 fluorescence photography, sebaceous follicles, 857–858 follicular biopsy, 828 furrows and wrinkles, image analysis, 156–159 ionic gradation, 430–431 nail surface, 925–927 Sebutapes, 836–837 stylus method, contour measurement, 167 surface contours and wrinkles, methods comparison, 206–207, 209 thermal imaging, skin temperature, 778 ultrasound imaging, 477–478 Image analysis, furrows and wrinkles analysis, 157, 160 analyzer instrumentation, 156–157 automation, 158–159 basics, 155 correlations, 160–161 error sources, 159–160 image analysis, 156–159
1009
lighting angle, 159–160 measurement parameters, 157–158 methodology, 156–159 objectives, 155–156 recommendations, 161 replica artifacts, 159, 160 shadowing method, 156 Image-Pro Plus 4.1, 959 Images analysis software, 495–496 application, 299–302 bank storage, 89, 91, 92 blocks, 141–142 capturing, 777–778 display, 778 filtering, 838–839 frame, 517 grabber board, 128 processing, 93, 684 software, 93 Imaging impedance, 348–349 nuclear magnetic resonance, 308–309 optical coherence tomography, 261, 263–264 Immediacy, compact digital cameras, 89 Immersion tests, 946–947, 950 Immunofluorescence labeling, 291 Immunological effects, 944 IMP, see Investigational medicinal product (IMP) Impact loading, 905 Impact stimulation, 791 Impedance, skin, 330 Implementation, Raman spectroscopy, 552–556 Impression methods, 459 Improvements, hair growth measurements, 887–888 IMP Spectrometer, 338 Inactive follicles, 862–864 Inadequate measuring conditions, 11 Increased skin temperature, 778–780 Indentometry, 617–620 Independent data, 64–65 Indications, ultrasound imaging, 529 Indirect lymphography, 742 Individual-related variables, 390–391 Individuals, 581, 949–950 Induction cutaneous pain, 789, 790–794 skin blisters, 434–435 Induration, psoriasis assessment, 39 Industries, prevailing types, 35 Infants, see also Aging and age differences; Children positioning, 86 transcutaneous PCO2, 409 ultrasound imaging, 476–477 Infants’ Dermatitis Quality of Life Index (IDQOL), 935–936 Infections, skin, 35 Inflammation and inflammatory conditions CSSS, 240–241 high-resolution sonography, 251, 252 microdialysis sampling, 449–450 optical coherence tomography, 264 reflectance confocal microscopy, 270–271 ultrasound imaging, 477, 481–483 Informed consent, 50–51, 74–76 Infrared photography, 112–113 Infrared radiometers, 757–758
1010
Handbook of Non-Invasive Methods and the Skin, Second Edition
Infrared theory, 756 Infrared thermography, 974 Infrared thermometers, 756–757 Injection depth and trauma, 747–748 Injex needless injectors, 289 Insertion, microdialysis sampling, 447 Instron, 898–899 Instrumentation changing, 71 dermatoscopy, 110–112 gas-bearing electrodynamometer, 623–624 hair strength evaluation, 904–905 laser Doppler imaging, 718 microdialysis sampling, 445–446 ultrasound imaging, 516–517 Instruments color calibration, 654–659 colorimetry, 639 erythema and melanin indices, 667 method selection, 9–11 pH, 413 skin disease assessment guidelines, 934–935 surface contours and wrinkles, methods comparison, 206–209 ultrasound imaging, 494–495 variables, 387–390 Integument, see Skin integument Interactive analysis, Sebutapes, 838–839 Interinstrumental variability, 389 Internal reliability, 933 International Conference on Harmonization (ICH), 48–49 International Society for Bioengineering and the Skin (ISBS), 12, 16 International Society for Digital Imaging of Skin (ISDIS), 12 International Society for Skin Imaging (ISSI), 12, 16 International Standards Organization (ISO), 67–69, 756 Internet basics, 15, 24–25 educational courses, 20–21 literature searches, 19, 22–23 search engines, 21–24, 23, 25 subject directories, 16–17, 19 tutorials, 20–21 uniform resource locators, 15–16, 17–18 Interobserver reliability, 933 Interpretation microdialysis sampling, 448–449 phototesting, 995 sweat analysis, 819 twistometry, 605 Intradermal administration, 287, 288, 289–290 Intrainstrumental variability, 389 Intraobserver reliability, 933 Intrasubject standard deviation, 64 Intravenous administration, 290 Intravital dyes, 742 Intravital microscopy, 698–699 Intrinsic aging, twistometry, 605–607 Invasiveness, microdialysis sampling, 447 Inversion, graphs, 316, 319, 319, 322 Inversion-recovery sequence, 539–540 Investigational medicinal product (IMP), 50 Investigator responsibilities, 50–52 In vitro measurements, surface contours and wrinkles, 208, 211 In vitro microdialysis sampling, 445, 446 In vivo confocal microscopy, fluorescent markers administration, 287–290 applications, 286
In
In In In In
In
In
In In
basics, 285–286 fluorophore selection, 286–287 functional imaging, 291–292 future directions, 292–293 green fluorescent protein expression, 292 immunofluorescence labeling, 291 intradermal administration, 287, 288, 289–290 intravenous administration, 290 skin barrier function, 291, 292 thermal burn-induced autofluorescence, 292 topical application, 287 transdermal drug delivery, 291 vivo confocal microscopy, product research and development basics, 297 epidermal thickness measurement, 298–300 images application, 299–302 melanin distribution visualization, 299, 300 methodological principles, 298–299 objectives, 297–298 reflectance confocal microscopy, 297 skin research suitability, 302–303, 305 sunscreen agent evaluation, 300, 302, 303–304 ultraviolet radiation protection, 300, 302, 303–304 vivo dermal water and edema ultrasound assessment, 507–509 vivo experiments, haptic finger, 234–235 vivo/ex vivo follicular transport assessment, 861–865 vivo high-resolution sonography analysis, 247 basics, 246, 251–253 glabrous skin, normal, 247, 249–250 methods, 246–247 palmar skin, normal, 247–250 patients, 247–248 processing, 246–247 results, 248–251 skin diseases, 247–248, 250–251 statistics, 248 three-dimensional sonography, 247 vivo measurements friction evaluation, 226–227 nuclear magnetic resonance, 310–312 vivo nuclear magnetic resonance (NMR) basics, 307 correlations, 313 error sources, 312–313 methodological principle, 308–312 objectives, 307–308 recommendations, 313 skin layer parameters, 310–312 in vivo high-resolution imaging, 308–310 in vivo measurements, 310–312 vivo reflectance confocal laser microscopy, 277 vivo reflectance confocal microscopy acute contact dermatitis, 270 basics, 268, 274–275 cutaneous infections, 270–271 difficulties, 274 inflammatory conditions, 270–271 issues, 274 margin assessment, 274 neoplastic skin lesions, 271–274 nonpigmented lesions, 271–272 normal skin, 269–270 pigmented lesions, 272–274 psoriasis, 271 reflectance principles, 268–269
Index
solutions, potential, 274 therapy adjunct, 274 treatment response evaluation, 274 In vivo structure analysis, ultrasound B-mode imaging allergic patch test reaction, 500, 501 applications, 496–497 basics, 493–494, 503–504 corticosteroids, topical, 501 equipment, 494–495 image analysis software, 495–496 instruments, 494–495 irritant reactions, 501–503 normal skin, 497–499 psoriatic skin, 503–504 technical aspects, 495 tissue characterization, 495–503 wheals, 503 In vivo water sorption-desorption test, 333–334 Ionic gradation, visualization techniques, 429–431 Iontophoresis, 674, 818 Ion visualization process, 430 Irradiation, phototesting, 992–993, 995, see also Ultraviolet light and radiation Irritants atopic dermatitis assessment, 43 colorimetry, 641–642 measurements, semiopen systems TEWL, 384–385 racial differences, 29–30 sodium laurel sulfate, 482, 948 ultrasound imaging, 501–503 Irritations, capacitance measurement, 341, 342 ISBS, see International Society for Bioengineering and the Skin (ISBS) ISDIS, see International Society for Digital Imaging of Skin (ISDIS) ISO, see International Standards Organization (ISO) ISSI, see International Society for Skin Imaging (ISSI) Issues, reflectance confocal microscopy, 274 36-Item Short Form (SF-36), 934–935
J Java, 377 Joints, ultrasound imaging, 523–524, 525 Justice, 74
K Kaposi sarcoma, 486 Kappa coefficient, 933 Keratin fiber, 897 KES-SE Friction Tester, 225–226 Kinetics, 742 Kirchoff’s law, 756 Known-groups validity, 933 Kodak Gray Scale, 656–657 Kubelka-Munk theory, 318, 320
L Laboratory facility, ultrasound imaging, 478–480 Labscan 6000 instrument, 637–638 Lamé’s coefficient, 603 Landmarks, nails, 920–921 Langerhans cells, 34
1011
Langer’s lines basics, 195, 565–566 correlations, 566–568 error sources, 568 methodological principle, 567–568 multiple simple tension tests, 567 objective, 566–567 qualitative correlates, 566–567 recommendations, 568 skin biomechanics, 566 suction chamber methods, 567–568 twistometry, 601 Laser capillary microscopy, 673–674 Laser Doppler methods, see also Doppler methods erythema and melanin indices, 669 flowmetry, 639–640, 709–713 imaging, 717–721 periodic fluctuations, CBF, 699–700 skin blood flux measurement, 691–695 transcutaneous oxygen tension, 403 wheals-and-flare reactions, 969–970 Laser profilometry air-bearing table, 171 algorithm, 174–175 applications, 177 basics, 169–170 calibration methods, 171–172 control, 170–171 data collection, 173 frequency response analysis, 172 measuring, 173 methodological principle, 170–177 objective, 170 optical sensor, 171 parameter calculations, 173–174 recommendations, 177 replicas, 170 roughness specimens, 172 software, 172 stratified structure, characterization, 173 stylus instrument comparison, 172–173 stylus method, contour measurement, 166–167 three-dimensional parameters, 175–177 Lasers, 551, 793–794 LCMSD, see Liquid chromatography with mass selective detector (LCMSD) Ledderhose disease, 527 Leeds Acne Grading Scale, 938 Legal references, method selection, 11–12 Leg ulcers, 486 Lesions capacitance measurement, 341 cystic, ultrasound imaging, 525–526 epiluminescence microscopy, 121 high-frequency electrical conductance measurement, 332–333 Levarometry, 613–615 Lévêque and de Rigal device, 330, 603, 605 Lichen planus, 247–248 Lichen ruber planus, 116–117 Ligaments, ultrasound imaging, 519, 521 Light, direct, 390 Light-gathering power, 126–127 Lighting angle, 159–160 Lighting (photography), 83–85 Light shielding, phototesting, 994 Light sources, sunlight, and radiation dosimetry
1012
Handbook of Non-Invasive Methods and the Skin, Second Edition
basics, 981 dose prediction, 985 dosimetry, 984 fiber-optic microscopy, 127 photodynamic therapy, 985–988 photosensitivity, 984, 985, 987–988 phototesting, 991–992 sunlight, 982, 983–984 UV sources, 981–982 Likelihood, 55 Likelihood ratio, 55 Limitations diseased surface areas, 964–965 microdialysis sampling, 450–451 statistical analysis, 55–56 surface contours and wrinkles, methods comparison, 211–212 Linear array transducers, 516 Linear transducers, 523 Linear variable differential transformer (LVDT), 613, 623 Lines family, directional extraction, 200 Lines identification, 198–200 Lipid analysis, follicular biopsy, 827 Lipid level, Sebumeter, 843–844 Lipodermatosclerosis, 524 Lipomas, 527 Lipomatous tumors, 526, 527 Lipophilic yeasts, 467–469 Liposarcoma, 528 Liquid chromatography with mass selective detector (LCMSD), 855 Liquid crystal contact temperature measurement, 774–775 Literature searches, online, 19, 22–23 Load influence, digital Cutometer, 588 Localization, sweat gland, 805–808 Localized scleroderma, 40, 524 Location, sodium laurel sulfate, 949–950 Loctite, 835 Longitudinal growth, nails, 919–921 Longitudinal relaxation time (T1), 540–544, 545–546 Longitudinal scans, 518 Longitudinal striations, 925 Long-term effects, moisturizing products, 343 Long-wave photon detectors, 761, 762 Loss, 133Xenon wash-out technique, 738 Low-perfusion microangiopathy, 712 Lubricants, 217, 220 Lucid, Inc., 297 Lupus erythematosus, 117 LVDT, see Linear variable differential transformer (LVDT) Lymphadenopathy, 520–521, 524, 528 Lymphangiography, 742 Lymphangiomas, 526 Lymphatic hemangiomas, 526 Lymphedema, 486 Lymph flow evaluation basics, 741 blood clearance, 747 correlations, 749 data analysis, 743–744 distribution volume, 748 error sources, 747–748 extrinsic forces, 748 injection depth and trauma, 747–748 lymphangiography, 742 lymphoscintigraphy, 742–743 objectives, 743–746 procedure, 743
radiolabeled tracers, 743 recommendations, 749 reliability, 744 reproducibility, 744 skin lymph flow, 743 tracer migration, 747 Lymph nodes, 524 Lymph node uptake, 742 Lymphoscintigraphy, 742–743 Lymph transport kinetics, 742–743
M 5 M’s (man, machine, methods, materials, and marginal conditions), 72 Macroduct system, 818 Macrorelief, three-dimensional evaluation, 187–191 Magnetic resonance spectroscopy (MRS), 531–534 Magnetic resonance spectroscopy (MRS), normal skin and pigmented tumors basal cell carcinoma, 541–542 basics, 537–538, 544–547 future outlook, 548 longitudinal relaxation time, 540–541, 544, 545–546 malignant melanoma, 543–544 materials, 539–541 methods, 539–541 microscopy, 539 nevocellular nevus, 542–543 non-invasive imaging methods, 538 principles, 538–539 results, 541–544 spin-echo sequence, 539–540 statistics, 541 transversal relaxation time, 541, 544, 545–546 voxel reconstruction, 541 Magnification, 101–106 Malassezia furfur, 465 Malconclusion, 11 Male differences, see Sex differences Malignant conditions, dermatoscopy, 118–121 Malignant melanoma (MM), see also Melanoma automated assessment, 138–140 basics, 135, 142 colors assessment, 135–142 computer assessment, 136–142 dark areas, 136–137, image blocks, 141–142 magnetic resonance spectroscopy, 543–544 nevi, 138–141 SIAscopy, 316, 319, 322 Malignant tumoral pathology, 528 Mann-Whitney-Wilcoxon test, 248, 540 Manufacturer’s manual, 12 Mapping fungi, 467–469 Marfan’s syndrome, 610 Marginal probability, 54 Margin assessment, 274 MAT, see Moisture accumulation test (MAT) Materials magnetic resonance spectroscopy, 539–541 morphological tree, cutaneous network of lines, 196 photopatch testing, 994 Mean skin temperature, 771–772 Measurements capacitance measurement, 339–341
Index
closed-chamber systems, TEWL, 393–396 Cutometer, 579–580 depth, stratum corneum hydration state, 356 dermal edema, 508–509 desquamation rate, 362–363 digital Cutometer, 584–587, 588–589 excreted sebum, 831–834 friction evaluation, 226–227 furrows and wrinkles, image analysis, 157–158 hair growth, 869–873, 883–893 indentometry, 618–619 laser Doppler flowmetry, 711–712 laser profilometry, 173 location, surface skin temperature, 754–755 nails, 920–921, 923–924 physiological fluid movement, 507–508 procedure design, laser Doppler imaging, 719–720 sebum excretion, 831–839, 841–845, 847–851 serial tape stripping pH evaluation, 422 stylus method, contour measurement, 164 surface contours and wrinkles, methods comparison, 206–208, 211 Measurements, commercial instruments accuracy, 355 basics, 351, 356 corneometers, 352–353 correlations, 355 DermaLab moisture unit, 353 measurement depth, 356 MoistureMeters, 353 multifrequency impedance spectrometer, 354 Nova Dermal Phase Meters, 353–354 principle, 351–352 sensitivity range, 355–356 Skicon 200/200EX, 354 SkinChip, 354 variation coefficient, 356 Measurements, epidermal capacitance basics, 337–338, 343 corneometer, 338–341 dermato-cosmetic applications, 341 diseases, 341 lesions, 341 measurements, 339–341 MoistureMeter, 341 moisturizing products, 342–343 objectives, 338 probe, 338, 339 Measurements, semiopen systems TEWL barrier composition, 385 barrier function evaluation, 384 basics, 384, 391 creams, 385 environment-related variables, 387–390 error sources, 387–391 individual-related variables, 390–391 instrument-related variables, 387–390 irritants, 384–385 methodological principle, 385–387 moisturizers, 385 objective, 384–385 predictive irritancy testing, 384–385 protective creams, 385 sweat gland activity, 384 theory, 385 validity, method and variables, 387 Measurements, transcutaneous oxygen tension (TcPO2)
1013
applications, 400–402 basics, 397 change dynamics, 402 correlations, 399–400 error sources, 399 methodological principle, 397–399 recommendations, 398, 403 Mechanical properties, sclerotic skin, 41 Mechanical scanning, 758–760 Mechanical stimulation, 791 Mechanical strength, hair, 895–900 Mechanical testing, twistometry, 602–603 Mechanical trauma, 796 Mechanics, ballistometry, 628–630 Median nerve, ultrasound imaging, 525 Medical care, investigator responsibilities, 50 Medical ethics, 73–74 Medical Outcomes Study (MOS), 934–935 Medicaments, topical, 35 Medicotechnical instruments, 12 Medscape web page, 17, 20–21 Megapixels, 91–92 Melanin CSSS, 240–241 distribution visualization, 299, 300 index, 665–670 Melanocytic lesions, 279–282 Melanocytic nest features, 280–281 Melanoma, see also Malignant melanoma (MM) automated assessment, 138–140 basics, 135, 142 colors assessment, 135–142 computer assessment, 136–142 cutaneous, 528 dark areas, 136–137 image blocks, 141–142 nevi, 138–141 reflectance confocal microscopy, 272, 273, 274 SIAscopy, 316, 319, 322 ultrasound imaging, 485 Membrane antiporters, 412 Memory cards, 91, see also Storage Men, see Sex differences Metabolism, microdialysis sampling, 450 Metal/carbon replication, 148 Method, bacteria sampling, selection, 457–458 Methodological principles apocrine sweat collection, 821–822 confocal microscopy, product research and development, 298–299 digital Cutometer, 584–585 dry skin and scaling evaluation, D-Squames and image analysis, 376–378 dynamic capillaroscopy, 673–674 eccrine sweat collection, 817–819 erythema and melanin indices, 665–667 follicular biopsy, 825–826 friction evaluation, 225–226 furrows and wrinkles, image analysis, 156–159 gas-bearing electrodynamometer, 622–623 gravimetric measurement technique, 847–850 hair strength, 898, 903–904 high-frequency electrical conductance measurement, 330–331 Langer’s lines, 567–568 laser Doppler methods, 692–693, 718–720 laser profilometry, 170–177 Levarometry, 613–614
1014
magnetic resonance spectroscopy, 532–534 measurements, semiopen systems TEWL, 385–387 nail thickness, 923–924 nuclear magnetic resonance, 308–312 pH, 413 photopatch testing, 994 quasi-L*a*b* color measurement, 649–650 Sebumeter, 842 Sebutapes, 836–837 SIAscopy, 317–320 skin chamber techniques, 434–436 stratum corneum hydration state, 351–352 stratum corneum material, harvesting, 371–372 stylus method, contour measurement, 164–166 sweat gland localization, 806–807 TapeAnalyzer, 831–832 transcutaneous oxygen tension measurements, 397–399 transcutaneous PCO2, 407–409 trichograms, 876–877 133Xenon wash-out technique, 733–738 Methods bacteria sampling, 458–463 colorimetry, 636–638 high-resolution sonography, 246–247 magnetic resonance spectroscopy, 539–541 morphological tree, cutaneous network of lines, 196 non-invasive technique selection, 9–12 Methotrexate treatment, 53 Mexameter, 666–668 Microbial densities, 827 Microcirculation, 679–685 Microcomedomes, 6 Microdialysis, skin sampling advantages, 450–451 allergology, 449–450 animal experiment preparation, 446 basics, 443–444 bioequivalence potential, 451 calculations, 448–449 challenges, 450–451 dialysate analysis, 445, 446 experimental setup, 446–448 future directions, 451 homeostasis, 450 human experiment preparation, 446–447 inflammation, 449–450 insertion, 447 instrumentation, 445–446 interpretation, 448–449 invasiveness, 447 limitations, 450–451 metabolism, 450 persusate, 446 physiology, 449–450 principles, 444–445 probes, 445–448 pumps, 446 recovery, 444–445 regional drug delivery, 449 regulatory aspects, 451 systemic drug delivery, 450 target organ measurements, 450 test areas, 449–450 thickness, skin, 447–448 topical drug delivery, 449 transdermal drug delivery, 449
Handbook of Non-Invasive Methods and the Skin, Second Edition
trauma, insertion, 447 troubleshooting, 445, 451 ultrasound imaging, 447–448 validation, 448–449 in vitro setup, 445, 446 Microelectromechanical systems (MEMS), 348–349 Microindentation, 926 Microrelief, three-dimensional evaluation, 182–187 Microscan System (Fort UK Microscan MS 2500), 112, 115 Microscopic techniques desquamation rate, 365–366 trichograms, 871, 875–880 Microscopy, 539, see also specific type Micro-sensor mapping basics, 811, 815 hydration, 814, 816 principles, 811 skinchips images, 811–812 stratum corneum, hydration, 812–813 sweating, 813–814 Microtopography, static, 926 Miliaria, 35 Minolta Chromameter, 638–642, 650, 669, 947, 975, see also Chromameters MI-PCP Vers. 2.32 system, 959 MM, see Malignant melanoma (MM) M-mode scanning, 474 Mobility, compact digital cameras, 92 Models autoregressive, 702 Bürger, twistometry, 604 Burte-Halsey, 896 forward prediction, SIAscopy, 318–319 heat wash-in/heat wash-out, 724–725 rheological, twistometry, 604–605 series zone model, 896–897 skin coloration, 317 133Xenon wash-out technique, 735–737 Modesty garments, 86 Modified Rodnan total skin thickness score, 40–41 Modulus of Young digital Cutometer, 587–588 hair strength measurement, 896, 898 Hookean region, 896 suction chamber method, 595, 597 twistometry, 603–604, 606, 608, 610 Moisture accumulation test (MAT), 42 MoistureMeter capacitance measurement, 341 epidermal capacitance, 338, 341 sensitivity, 355–356 stratum corneum hydration state, 353 Moisturizers capacitance measurement, 342–343 high-frequency electrical conductance measurement, 334 measurements, semiopen systems TEWL, 385 tribological studies, 217, 220–221 Molding methods, 806–807, 808 Monitoring, investigator responsibilities, 50 Monitor responsibilities, 49 Monophasic flow, 520 Morality, 73 Morphea, dermatoscope, 115 Morphological tree, cutaneous network of lines anisotropic spectral filtering, 198–200 basics, 195–196, 201–203
Index
directional extraction, 199–200 Fourier transform, 196–198 frequency range, 196–198 lines family, directional extraction, 200 lines identification, 198–200 material, 196 method, 196 orientation, 198 spatial frequency range, 197 statistical analysis, 200–201 three-dimensional tree skin spectrum, 200–202 wrinkles, directional extraction, 199 MOS, see Medical Outcomes Study (MOS) Movement, 87, 720 MRS, see Magnetic resonance spectroscopy (MRS) MT-8C probe, 331 Mucoid cysts, 525 Multifrequency impedance spectrometer, 354 Multi-instrument calibration, 659–661 Multiple simple tension tests, 567 Munsell color-order system, 635, 637 Muscle fascia, 477 Mustard oil, 796
N Nail area severity (NAS) score, 914–915 Nails longitudinal growth measurement, 919–921 surface image analysis, 925–927 thickness measurement, 923–924 ultrasound imaging, 487, 521 Nails, assessment basics, 911 capillaroscopy, 913 dermascopy, 912–913 light microscopy, 914 magnetic resonance imaging, 913–914 measurement, 915–916 microscopy, 914 permeability, 916 photodermatoscopy, 912–913 photography, 911–912 profilometry, 913 scanning microscopy, 914 scoring systems, 914–915 strength, 915 surface replicas, 913, 914 transmission microscopy, 914 Nail surface image analysis, 925–927 NAS, see Nail area severity (NAS) score National Institutes of Health (NIH), 4, 89 Nd/YAG lasers, 551, 794 Negative mold, 150 Negative replica, 149–152 Neoplastic skin lesions, 271–274 Nerves, ultrasound imaging, 519, 521 NETD, see Noise-equivalent temperature difference (NETD) Neural algorithms, 857–858 Neural disease, ultrasound imaging, 524–525 Neurofibromas, 527 Neurogenic tumors, 527 Neurological dysfunction, 781 Nevi melanoma diagnosis, 138–141
1015
ultrasound imaging, 485 Nevocellular nevus, 542–543 Nevus flammeus, 118 NHP, see Nottingham Health Profile (NHP) NIH, see National Institutes of Health (NIH) NMR, see Nuclear magnetic resonance (NMR) Nociceptive withdrawal reflex, 795 Nodular prurigo, 483 Nodular tenosynovitis, 525 Noise-equivalent temperature difference (NETD), 760 Noncomputerized, stand-alone suction cup, 595–596 Noncontact temperature measurement, 756–760 Noninfectious erythemato-squamous disorders, 240–241 Noninvasive assessment methods atopic dermatitis, 42–43 barrier assessment, 39 basics, 37–38 bioengineering techniques, 43 blood flow, 38–39 capacitance, 42–43 conductance, 42–43 desquamation, 39 erythema, 38–39 ethical considerations, 76 extent of disease, 38 induration, 39 irritant reactivity, 43 localized scleroderma, 40 oil immersion examination, 101–106 psoriasis, 38–40 recent techniques, 39 sclerotic skin, 40–41 systemic scleroderma, 40–41 topical drugs side effects, 40 transepidermal water loss, 42 Non-invasive imaging methods, 538 Nonmaleficence, 74 Nonpigmented lesions, 271–272 Nonvisible hair, 892 Normal anatomy and skin confocal scanning laser microscopy, 277–278 CSSS, 239–240 Cutometer, 582 Dermaflex, 574–575 high-frequency electrical conductance measurement, 332 oil immersion examination, 103 optical coherence tomography, 262–263, 264 reflectance confocal microscopy, 269–270 sebum excretion, 844–845 SIAscopy, 317–318 sonograms, 246 suction chamber method, 574–575 ultrasound imaging, 476–477, 497–499, 520–523 Normal glabrous skin, 247, 249–250, 252 Normal load, tribological studies, 217–218 Normal palmar skin, 247–250 Normal skin, magnetic resonance spectroscopy (MRS) basal cell carcinoma, 541–542 basics, 537–538, 542–547 future outlook, 548 longitudinal relaxation time, 540–541, 544, 545–546 malignant melanoma, 543–544 materials, 539–541 methods, 539–541 microscopy, 539 nevocellular nevus, 542–543
1016
Handbook of Non-Invasive Methods and the Skin, Second Edition
non-invasive imaging methods, 538 principles, 538–539 results, 541–544 spin-echo sequence, 539–540 statistics, 541 transversal relaxation time, 541, 544, 545–546 voxel reconstruction, 541 Nottingham Health Profile (NHP), 934 Nova Dermal Phase Meter, 338, 353–356 Nuclear magnetic resonance (NMR) basics, 307 correlations, 313 error sources, 312–313 methodological principle, 308–312 objectives, 307–308 recommendations, 313 skin layer parameters, 310–312 in vivo high-resolution imaging, 308–310 in vivo measurements, 310–312 Nuremberg Code, 74 Nyquist plot, 345, 347
O Objectiveness, dry skin and scaling evaluation, 378 Objective ratings, facial scans, 98–99 Objectives capacitance measurement, 338 colorimetry, 636 confocal microscopy, product research and development, 297–298 dermatoscopy, 109–110 digital Cutometer, 584 erythema and melanin indices, 665–667 furrows and wrinkles, image analysis, 155–156 hair strength measurement, 897 high-frequency electrical conductance measurement, 329–330 Langer’s lines, 566–567 laser Doppler imaging, 717–718 laser profilometry, 170 lymph flow evaluation, 743–746 magnetic resonance spectroscopy, 532–534 measurements, semiopen systems TEWL, 384–385 nuclear magnetic resonance, 307–308 skin chamber techniques, 434 stratum corneum material, harvesting, 371 sweat gland localization, 806 transcutaneous PCO2, 407 trichograms, 876 133Xenon wash-out technique, 733 Object-related error, 11 Observed prevalence, sensitivity, and specificity, 54 Occlusion chambers, 950 Occlusive tests, 945–946, 973–977 OCDR, see Optical coherence-domain reflectometry (OCDR) OCT, see Optical coherence tomography (OCT) Odds ratios, 55 Oil immersion examination, 101–106 Oil-in-water emulsions, 342 One-time tests, sodium laurel sulfate, 945–946 Online information, see Internet Onychomycosis, 915 Opalescent film imprint, sebum excretion, 841–845 Open exposures, 950 Open GL mode, 211 Open tests, sodium laurel sulfate, 946
Operation, compact digital cameras, 91 Optical coherence-domain reflectometry (OCDR), 257 Optical coherence tomography (OCT) accurate positioning, images, 261 applications, 262–264 basics, 257–258, 264 clinical considerations, 260–262 Doppler imaging, 264 follicular assessment, 863 imaging, 261, 263–264 inflammatory conditions, 264 normal skin, 262–263, 264 patient interface, 261–262 penetration depth, 258, 259 probe, 260–261 psoriasis, 39 spatial resolution, 258 speed, imaging, 261 technical considerations, 258–260 time-domain, 258 tumors, 264 Optical head, Hi-scope system, 128 Optical profilometry, 205, 208–209, see also Profilometry, laser Optical sensor, laser profilometry, 171 Optics, fiber-optic microscopy, 126–127 Optimal exposure, photography, 87 Ordinal values, 378 Orientation, morphological tree, 198 Output, photography, 88
P P. ovale, see Pityrosporum spp. Pads, bacteria sampling, 459 Paired data, 65 Palmar fibromatosis, 527 Palmar skin, normal, 247–250 Palpation, ultrasound imaging, 518 Panjet needless injectors, 289 Papillae, dermal, 280–281 Parameter calculations, 173–174 Parameters, surface contours and wrinkles, 208–209 Parameters, twistometry, 603–605 Parametric maps, 317 Parapsoriasis en plaque, 115 Parasitic disorders, CSSS, 240 Partial-thickness tears, 525 PASI, see Psoriasis area and severity index (PASI) Passive technique, desquamation rate, 362 Patch tests allergic reaction, ultrasound imaging, 500, 501 closed, sodium laurel sulfate, 945–946 occluded, evaluation, 973–977 Pathological epidermis, 308–310 Pathological skin, studies, 745–746 Pathology Dermaflex, 575–576 suction chamber method, 575–576 ultrasound imaging, 523–524 Patients examination technique, 518–520 high-resolution sonography, 247–248 interface, OCT, 261–262 positions, 86–87 preconditioning, Cutometer, 581
Index
rights, 74 PCO2, see Transcutaneous PCO2 PDT, see Photodynamic therapy (PDT) Pearson’s correlation, 933 Penetration depth, optical coherence tomography, 258, 259 Percutaneous absorption, 29 Performance error, 11 measurements, semiopen systems TEWL, 389–390 quality management system, 72 three-dimensional evaluation, micro- and macrorelief, 189–191 Perfusion, temporal changes, 720 Periodic fluctuations, cutaneous blood flow, 697–704 Periungual capillaroscopy, 680 Perspective, photography, 82 Persusate, 446 Petrolatum treatment, 251 pH, serial tape stripping evaluation basics, 421–422 detergent effects, 423 dyes, 422 enzymes, 423–425 flat pH glass electrode, 422 measurement methods, 422 recommendations, 425 regulation, 421–422 skin pH, 422–425 soap effects, 423 tape stripping, 425–426 pH, skin surface age differences, 414 antimicrobial function, 415 basics, 411–412, 417 circadian rhythm, 414 eczematous diseases, 416 endogenous and exogenous factors influence, 413–414 error sources, 413 external environmental factors influence, 413–414 gender differences, 414 histidine-to-urocanic acid pathway, 412–413 ichthyosis, 416–417 instruments, 413 membrane antiporters, 412 methodological principle, 413 phospholipid-to-free fatty acid pathway, 412 racial differences, 414 SC functions, 412, 415–416 skin barrier, 415–416 skin disorders, 416–417 tape stripping, 414 test areas, 414, 415 Pharmacological inhibition, 683 Phased arrays, ultrasound imaging, 516 Philips TL01/TL12 lamp, 982 Phospholipid-to-free fatty acid pathway, 412 Photoaging, bioengineering, 6 Photoallergic reactions, 992 Photoconductive detectors, 762 Photodamage, Hi-scope system, 129 Photodynamic therapy, 986–987, see also Light sources, sunlight, and radiation dosimetry Photodynamic therapy (PDT), 292 Photographic and computerized hair growth measurements analytical methods, 890–892 basics, 883, 892–893 function, hair, 884–886
1017
global vision and imaging methods, 888–889 improvements, 887–888 photography, 884–892 standards, 884, 886–887 structure, hair, 883–884 Photography compact digital cameras, 89–93 computerized image analysis, 95–99 hair growth measurements, 884–892 infrared, 112–113 ultraviolet, 113, 114 Photography, medical/clinical aspects background, 83, 85 basics, 81–82, 88 cameras, 87 consent, 82 depth of field, 82–83 ethical guidelines, 82 film, 87 guidelines, publication, 88 lighting, 83–85 movement, 87 optimal exposure, 87 output, 88 patient positions, 86–87 perspective, 82 processing, 87 publication guidelines, 88 ring flash, 85, 87 room space, 85 scaling, 82 specialist photography, 87 standardization, 88 tracking photos, 82 viewing, 88 Photometric techniques, 363–364 Photon detectors, 760–761 Photopatch testing, 994–995 Photoplethysmography, 695 Photosensitivity, 985, 987–988, see also Light sources, sunlight, and radiation dosimetry Phototesting basics, 991, 992, 996 drug phototesting, 995–996 interpretation, 995 irradiation, 995 light shielding, 994 light source, 991–992 photoallergic reactions, 992 photopatch testing, 994–995 phototoxic reactions, 992 serum phototesting, 996 test area, 994 UV irradiation, 992–993 visible light radiation, 993–994 without chemicals, 992–994 Phototoxic reactions, 992 Phototrichogram (PTG) technique, 890–891, see also Trichograms Photovoltaic detectors, 761–762 Physical principles and techniques, 475 Physical properties, 133Xenon wash-out technique, 733–734 Physicians, prescribing habits, 35 Physiology apocrine secretions, 821 fluid movement measurement, 508 friction evaluation, 227–228
1018
Handbook of Non-Invasive Methods and the Skin, Second Edition
microdialysis sampling, 449–450 racial differences, 29 variations, sebum excretion, 844 Pick-up, 164, 165 Picture quality, 91 Pigmentation, colorimetry, 643–644 Pigment distribution assessment method automated assessment, 138–140 basics, 135, 142 colors assessment, 135–142 computer assessment, 136–142 dark areas, 136–137 image blocks, 141–142 nevi, 138–141 Pigmented lesions, 272–274 Pigmented neoplasms, 241 Pigmented skin, magnetic resonance spectroscopy (MRS) basal cell carcinoma, 541–542 basics, 537–538, 542–547 future outlook, 548 longitudinal relaxation time, 540–541, 544, 545–546 malignant melanoma, 543–544 materials, 539–541 methods, 539–541 microscopy, 539 nevocellular nevus, 542–543 non-invasive imaging methods, 538 principles, 538–539 results, 541–544 spin-echo sequence, 539–540 statistics, 541 transversal relaxation time, 541, 544, 545–546 voxel reconstruction, 541 Pilonidal cysts, 526 Pilot study, sodium laurel sulfate, 951 Pitfalls, 528, 964–965 Pitting, nail surface, 926 Pityrosporum spp., 465, 467–469 Pixels, 91–92 Planck’s formula, 756 Plan-do-check-act cycle, 69, 70 Plantar fibromatosis, 527 Plants, seasonal variations effects, 35 Plastic impression technique, 149 Plastic surgery, 532 Playback, compact digital cameras, 91 Point of no-net-flux method, 448 Poiseuille’s law, 697 Poisson’s ratio, 603 Polarization-sensitive optical coherence tomography (PS-OCT) system, 260 Polarized light imaging, 129 Polyvinylidene fluoride (PVDF) film, 233–234 Port wine stains, 118, 119 Positions, patients, 86–87 Positive predictive value, 55 Positive replica, 149–150, 152 Posterior enhancement and shadowing, 516 Posterior probability, 55 Postocclusive reactive hyperemia, 698 Postthrombotic syndrome, 702–703 Postyield region, 896–897 Power, 63–64, 516 Power angio, see Doppler methods Power Doppler, see Doppler methods PPIX, see Protoporphyrin IX (PPIX) analysis
Practical guidance, ultrasound imaging, 478–480 Preconditioning, 581, 589 Prediction rule, 60 Predictive irritancy testing, 384–385 Predictive testing, 948 Predictive transforms, 97 Predictive value, 55–56 Predisease diagnosis, 6 Premature termination, clinical trials, 51 Prerequisites, 573–574 Presampling, hair growth, 870–871 Prescribing habits, 35 Present status, skin replication, 152–153 Preservatives allergies, 35 Pressure cutaneous pain assessment, 791 digital Cutometer, 589 dynamic capillaroscopy, 674 twistometry, 605 Prestudy checklist, 10–11 PRF, see Pulse repetition frequency (PRF) PRIMOS7, 208, 211 Principalism, 74 Principles magnetic resonance spectroscopy, 538–539 microdialysis sampling, 444–445 stylus method equipment, 165 Prints, photography, 88 Prior odds, 55 Prior probability, 55 Prism 460 Series, 835 Probes, see also Electrochemical electrode capacitance measurement, 338, 339 Cutometer, 580 digital Cutometer, 585, 588–589 Levarometry, 615 microdialysis sampling, 445–448 optical coherence tomography, 260–261 semiopen systems TEWL measurement, 389 ultrasound imaging, 495, 516 Processing high-resolution sonography, 246–247 microcirculation, 684 photography, 87 Product efficacy, 582, 845 Product-induced changes, 96–97 Product research and development, confocal microscopy basics, 297 epidermal thickness measurement, 298–300 images application, 299–302 melanin distribution visualization, 299, 300 methodological principles, 298–299 objectives, 297–298 reflectance confocal microscopy, 297 skin research suitability, 302–303, 305 sunscreen agent evaluation, 300, 302, 303–304 ultraviolet radiation protection, 300, 302, 303–304 Profile characterization, 165 Profilometry, laser, see also Optical profilometry air-bearing table, 171 algorithm, 174–175 applications, 177 basics, 169–170 calibration methods, 171–172 control, 170–171 data collection, 173
Index
frequency response analysis, 172 measuring, 173 methodological principle, 170–177 objective, 170 optical sensor, 171 parameter calculations, 173–174 recommendations, 177 replicas, 170 roughness specimens, 172 software, 172 stratified structure, characterization, 173 stylus instrument comparison, 172–173 stylus method, contour measurement, 166–167 three-dimensional parameters, 175–177 Prony spectral line estimation (PSLE) technique, 702 Propagation velocity, 516 Prophylactic potential, 827 Propionibacterium spp., 413–415, 458, 462 Proportional full-thickness strain, 571 Proportion of agreement, 933 Protection covers, probes, 389 Protective creams, 385 Protocol compliance, 50 Protoporphyrin IX (PPIX) analysis, 855, 858, 986–987 Provocative testing, 944, 950 PSLE, see Prony spectral line estimation (PSLE) technique PS-OCT, see Polarization-sensitive optical coherence tomography (PSOCT) system Psoriasis barrier, 39 basics, 38 dermatoscope, 116–117 DermFlex, 575 erythema, 38–39 extent of disease, 38 induration, 39 recent techniques, 39 reflectance confocal microscopy, 271 semiopen systems, transepidermal water loss, 384 skin blood flow, 38–39 skin disease assessment guidelines, 938 ultrasound imaging, 482, 503–504, 524 Psoriasis area and severity index (PASI), 38–39, 64, 936, 938, 958 Psoriasis Index of Quality of Life (PSORIQoL), 936 Psoriasis pustulosa, 117 Psoriasis vulgaris, 247–248, 251 PSORIQoL, see Psoriasis Index of Quality of Life (PSORIQoL) Psychophysical cutaneous pain assessment, 795 PTG, see Phototrichogram (PTG) technique Publication guidelines, 88 PubMed web page, 19, 22–23 Pull, Levarometry, 615 Pulse repetition frequency (PRF), 516 Pumps, microdialysis sampling, 446
Q QMH, see Quality management handbook (QMH) Qualification, investigator responsibilities, 50 Qualitative correlates, Langer’s lines, 566–567 Qualitative cultures, 468 Qualitative research methods, 9 Quality basics, 67 compact digital camera photography, 89
1019
good clinical practice, 52 Quality management handbook (QMH), 70 Quality management system, 67–72 Quantimet 720/970, 156 Quantitative cultures, 468 Quantitative research methods, 9 Quantum well infrared detectors (QWIP), 762–763 Quasi-L*a*b* color measurement, 649–651, 654
R Racial differences pH, 414 skin integument, 29–30 sodium laurel sulfate, 948 tribological studies, 222–223 Radiation, dosimetry, 981, 984, see also Light sources, sunlight, and radiation dosimetry Radiation, thermal imaging, 772 Radiative heat transfer, 777 Radiography, ultrasound imaging, 476 Radiolabeled tracers, 743 Radiometers, infrared, 757–758 Raman spectroscopy basics, 551 case studies, 556, 558–559 data interpretation, 554–556 future directions, 559–560 implementation, 552–556 skin characterization, 551–556 water in skin, 558–560 Randomization, investigator responsibilities, 50 Rapid scanning optical delay (RSOD), 260 Rayleigh scattering, 651 Raynaud’s phenomenon, 780–781 Raynaud’s syndrome, 675, 914 RCM, see Reflectance confocal microscopy (RCM) Reactivity, sodium laurel sulfate, 947–950 Real-time scanning, 474 Receiver operating characteristic (ROC) curves, 54, 55–60 Recent techniques, psoriasis assessment, 39 Recommendations bacteria sampling, 460, 463–465 dry skin and scaling evaluation, D-Squames and image analysis, 378 erythema and melanin indices, 670 fluorescence photography, sebaceous follicles, 856, 859–860 furrows and wrinkles, image analysis, 161 hair strength measurement, 899 high-frequency electrical conductance measurement, 334 indentometry, 619–620 Langer’s lines, 568 laser profilometry, 177 lymph flow evaluation, 749 magnetic resonance spectroscopy, 534 nail thickness, 924 nuclear magnetic resonance, 313 quasi-L*a*b* color measurement, 651 serial tape stripping pH evaluation, 425 skin chamber techniques, 433, 439 skin replication, 150–152 sodium laurel sulfate, 949–952 statistical analysis, 61 stratum corneum material, harvesting, 372 stylus method, contour measurement, 167–168 sweat gland localization, 808
1020
Handbook of Non-Invasive Methods and the Skin, Second Edition
transcutaneous oxygen tension measurements, 398, 403 transcutaneous PCO2, 410 ultrasound imaging, 517 133Xenon wash-out technique, 739–740 Record retention requirements, 51 Records, investigator responsibilities, 51 Recovery microdialysis sampling, 444–445 sebum excretion, 844 Recovery by loss, 448 Reference method, 448 References, infrared images, 778 References, legal or authoritative, 11–12 Reflectance confocal microscopy (RCM) acute contact dermatitis, 270 basics, 268, 274–275 confocal microscopy, product research and development, 297 cutaneous infections, 270–271 difficulties, 274 inflammatory conditions, 270–271 issues, 274 margin assessment, 274 neoplastic skin lesions, 271–274 nonpigmented lesions, 271–272 normal skin, 269–270 pigmented lesions, 272–274 psoriasis, 271 reflectance principles, 268–269 solutions, potential, 274 therapy adjunct, 274 treatment response evaluation, 274 Reflectance principles, 268–269 Reflection errors, laser Doppler imaging, 720–721 Region, body, 27–29 Regional drug delivery, 449 Registration, 133Xenon wash-out technique, 734–735 Regression, skin disease assessment, 932 Regression methods, statistical analysis, 60 Regulation good clinical practice standards, 48 microdialysis sampling, 451 serial tape stripping pH evaluation, 421–422 Relative recovery, 444 Reliability, 744, 932–933 Repeatability reliability, 933 Repeated tests, sodium laurel sulfate, 946 Repetitive cycles, DermaLab, 597–598 Repetitive stress-time curves, 587, 588 Replica artifacts, 159, 160 Replica image analysis, furrows and wrinkles analysis, 157, 160 analyzer instrumentation, 156–157 automation, 158–159 basics, 155 correlations, 160–161 error sources, 159–160 image analysis, 156–159 lighting angle, 159–160 measurement parameters, 157–158 methodology, 156–159 objectives, 155–156 recommendations, 161 replica artifacts, 159, 160 shadowing method, 156 Replicas bacteria sampling, 459
laser profilometry, 170 stylus method, contour measurement, 164, 166 surface contours and wrinkles, methods comparison, 206–208, 211 Replication, skin basics, 147–148 carbon replication, 148 correlations, 148–149 error sources, 148–150, 151 metal/carbon replication, 148 plastic impression technique, 149 present status, techniques, 152–153 recommendations, 150–152 silicone elastomer method, 149–152 Reports, investigator responsibilities, 51 Representative population, 54 Reproducibility capacitance measurement, 339–340 colorimetry, 639 digital Cutometer, 587 dry skin and scaling evaluation, D-Squames and image analysis, 378 Levarometry, 614 lymph flow evaluation, 744 Research, confocal microscopy, 302–303 Research, ethical guidelines, 74 Reservoir effect, Sebutapes, 839, 840 Resolution, photography, 91–92, 516 Resources, investigator responsibilities, 50 Responsibilities, good clinical practice standards, 49–52 Responsiveness, skin disease assessment, 932–934 Results DermaLab, 598–599 digital Cutometer, 587–588 high-resolution sonography, 248–251 magnetic resonance spectroscopy, 541–544 sodium laurel sulfate, 952 three-dimensional evaluation, micro- and macrorelief, 189–191 Retinoic acid, topical, 609 Retinoids, topical microcomedones, 6 photoaging, 6 Retrodialysis by calibrator, 448 Retrodialysis principle, 448 RGB/XYZ conversion, 657–658 Rheological model, Twistometry, 604–605 Rhodotorula spp., 467 Ridging, nails, 926 Ring flash, photography, 85, 87 Rippling, nails, 926 Risks, consent, 76 Rodnan total skin thickness score, 40–41 Roles, good clinical practice standards, 49–52 Room space, photography, 85 Rosacea assessment, 685 fiber-optic microscopy, 131 Hi-scope system, 129 Rosenau’s depression, 926 Roughness, 172, 182–183 RSOD, see Rapid scanning optical delay (RSOD) Rule of nines, 959, 960
S Safety classification rules, 12 Safety reporting, investigator responsibilities, 51
Index
SAFIR, see Skin Analyzing Fluorescence Imaging Recorder (SAFIR) system Sagging, measuring improvement, 99 Sampling hair strength evaluation, 903 ionic gradation, 430 methodology, 63 size calculation, 63–66 Sampling, bacteria available methods, 458–463 basics, 457 correlations, 460, 463 follicular sampling methods, 462–463 impression methods, 459 method selection factors, 457–458 recommendations, 460, 463–465 replica methods, 459 swabbing methods, 460–461 washing methods, 461–462 Sampling, microdialysis advantages, 450–451 allergology, 449–450 animal experiment preparation, 446 basics, 443–444 bioequivalence potential, 451 calculations, 448–449 challenges, 450–451 dialysate analysis, 445, 446 experimental setup, 446–448 future directions, 451 homeostasis, 450 human experiment preparation, 446–447 inflammation, 449–450 insertion, 447 instrumentation, 445–446 interpretation, 448–449 invasiveness, 447 limitations, 450–451 metabolism, 450 persusate, 446 physiology, 449–450 principles, 444–445 probe depth, 447–448 probes, 445–446 pumps, 446 recovery, 444–445 regional drug delivery, 449 regulatory aspects, 451 systemic drug delivery, 450 target organ measurements, 450 test areas, 449–450 thickness, skin, 447–448 topical drug delivery, 449 transdermal drug delivery, 449 trauma, insertion, 447 troubleshooting, 445, 451 ultrasound imaging, 447–448 validation, 448–449 in vitro setup, 445, 446 Saprophitic microorganisms, 240 Saturation-recovery technique, 540 Saving digital images, 92–93 Scabies, CSSS, 240 Scaliness, 130, 132 Scaling, photography, 82 Scaling evaluation, D-Squames and image analysis, 375–378
1021
Scalp coverage scoring (SCS), 889 Scalp hair growth, 869 Scaly lesions, 341 Scan angle and direction, 479 Scanning, mechanical, 758–760 Scanning, three-dimensional image analysis, 97 Scanning electron microscopy (SEM), skin replication basics, 147–148 carbon replication, 148 correlations, 148–149 error sources, 148–150 metal/carbon replication, 148 plastic impression technique, 149 present status, techniques, 152–153 recommendations, 150–152 silicone elastomer method, 149–152 Scars, ultrasound imaging, 484 Schade’s measuring system, 617 Schwanomas, ultrasound imaging, 527 Scleroderma DermFlex, 575 twistometry, 609–610 ultrasound imaging, 524 Sclerometry, nails, 927 Sclerotic skin, 40–41 SCORAD, see Scoring Atopic Dermatitis (SCORAD) Scoring Atopic Dermatitis (SCORAD), 932–933, 936–937 Scoring systems, see also specific type global vision and imaging systems, 886, 889 hair growth, 869–870 irritant reactions, 951, 952–953 SCS, see Scalp coverage scoring (SCS) SDT, see Sorption-desorption test (SDT) Search engines, 21–24 Seasonal variation effects, 33–35, 340, 390 Sebaceous cysts, ultrasound imaging, 525 Seborrheic eczema, dermatoscope, 115 Sebumeter, 841–845 Sebum excretion gravimetric measurement technique, 847–851 Sebumeter, 841–845 Sebutapes, 835–839 TapeAnalyzer, 831–834 Sebum production, 837–838 Sebutapes, 831–832, 834–839 Selection bias, 54 Selection of methods, 9–12 Sellotape stripping, 459 SEM, see Scanning electron microscopy (SEM), skin replication Semiopen systems, transepidermal water loss barrier composition, 385 barrier function evaluation, 384 basics, 384, 391 creams, 385 environment-related variables, 387–390 error sources, 387–391 individual-related variables, 390–391 instrument-related variables, 387–390 irritants, 384–385 methodological principle, 385–387 moisturizers, 385 objective, 384–385 predictive irritancy testing, 384–385 protective creams, 385 sweat gland activity, 384 theory, 385
1022
validity, method and variables, 387 Senile skin, dermatoscope, 117–118 Sensitivity capacitance measurement, 339 Levarometry, 614 skin disease assessment guidelines, 932–934 statistical analysis, 55–56 stratum corneum hydration state, 355–356 Sensors and handheld devices, surface skin temperature basics, 753–754 computer systems, 764 contact temperature measurement, 754 detector arrays, 760 detectors, 760–764 emerging technology, 764 expert systems, 764 fixed monitoring systems, 757 focal pane arrays, 758–760 future directions, 764–765 infrared radiometers, 757–758 infrared theory, 756 infrared thermometers, 756–757 long-wave photon detectors, 761, 762 measurement location, 754–755 noncontact temperature measurement, 756–760 photoconductive detectors, 762 photon detectors, 760–761 photovoltaic detectors, 761–762 quantum well infrared detectors, 762–763 scanning, mechanical, 758–760 short-wave photon detectors, 761, 762 temperature measurement, 754 terminology, 765–767 thermal detectors, 763–764 thermocouples, 754 thermometers, 754–755 Sensory evaluation, friction, 228–231 Serial tape stripping pH evaluation basics, 421–422 detergent effects, 423 dyes, 422 enzymes, 423–425 flat pH glass electrode, 422 measurement methods, 422 recommendations, 425 regulation, 421–422 skin pH, 422–425 soap effects, 423 tape stripping, 425–426 Series zone model, 896–897 Serine protease, 424 Seromas, ultrasound imaging, 524 Serum phototesting, 996 Sex differences, see also Gender differences capacitance measurement, 341 conductance levels, 332 Cutometer, 582 sebum excretion, 844 skin integument, 27 sodium laurel sulfate, 948 SF-36, see 36-Item Short Form (SF-36) Shadowing method, 156, 207 Shape, facial averaging methods, 96 Shape parameter (beta), hair, 905 Shooting, compact digital cameras, 91 Short-term effects, moisturizing products, 342–343
Handbook of Non-Invasive Methods and the Skin, Second Edition
Short-wave photon detectors, 761, 762 SIAscopy, see Spectrophotometric intracutaneous imaging (SIAscopy) Sickle cell disease, 703–704 Sickness Impact Profile (SIP), 934 Sigma imaging system, 546 Signal processing, haptic finger, 234 Significance level, sample size calculation, 63 SILFLO, 156, 189 Silicone elastomer method, 149–152 Simple tension tests, 567 Simplicity, 11 Single-lens reflex (SLR) cameras, 87 Single stress-time curves, 586, 587–588 SIP, see Sickness Impact Profile (SIP) Site variation, sebum excretion, 844 Skicon devices epidermal capacitance, 338–339 sensitivity, 355–356 stratum corneum hydration state, 330–331, 354 Skin aging, ultrasound assessment, 511–513 Skin Analyzing Fluorescence Imaging Recorder (SAFIR) system, 856–857, 859–860 Skin barrier bioimpedance, 347–348 composition, 385 fluorescent markers, 291, 292 function evaluation, 384 pH, 415–416 psoriasis assessment, 39 Skin biomechanics, 566, 608–609 Skin blood flow, psoriasis assessment, 38–39 Skin chamber techniques basics, 433–434 blisters, 434–436 correlations, 439 error sources, 436–439 methodologic principle, 434–436 objectives, 434 recommendations, 433, 439 Skin characterization, Raman spectroscopy, 551–552 SkinChip, 354, 811–815 Skin color, 643, 968–969 Skin coloration model, 317 Skin deformation reproducibility, 587 SkinDex, 261, 863, 935 Skin disease assessment guidelines acne, 938, 939 atopic dermatitis, 935–936 atopic eczema, 936–938 basics, 931–932, 938, 940 Children’s Dermatology Life Index, 935 Dermatology Life Quality Index, 935 dermatology quality, instruments, 935 disease-specific assessment, 936–938 disease-specific dermatology quality, 935–936 generic quality, instruments, 935 instruments, 934–935 36-Item Short Form, 934–935 Medical Outcomes Study, 934–935 Nottingham Health Profile, 934 psoriasis, 938 Psoriasis Index of Quality of Life, 936 reliability, 933 responsiveness, 933–934 sensitivity, 933–934 Sickness Impact Profile, 934
Index
Skindex, 935 validity, 932–933 Skin diseases Cutometer, 582 high-resolution sonography, 247–251 Skin disorders, pH, 416–417 Skin elasticity, sclerotic skin, 40 Skin elasticity, Twistometry actinic aging, 607–608 aging, 605–608 applications, 609–610 basics, 601–602, 610 connective tissue diseases, 610 equipment, 602 interpretation, 605 intrinsic aging, 605–607 mechanical testing, 602–603 parameters, mechanical, 603–605 retinoic acid, topical, 609 scleroderma, 609–610 skin biomechanics, 608–609 stratum corneum, 608–609 technical considerations, 602–605 testing, mechanical, 602–603 theoretical considerations, 602–605 topical retinoic acid, 609 validation, 605 Skin emissivity, thermal imaging, 775–776 Skin entry echo, sonograms, 246, 252 Skin examination, high-frequency ultrasound animal studies, 487 applications, 481–487 basics, 473–474 biological variables, 477, 480–481 connective tissue diseases, 483–485 correlations, 475–476 cutaneous neoplasms, 485–486 equipment, 478–480 examiner, 478–480 histology, 475–476 image analysis, 477–478 inflammatory conditions, 477, 481–483 laboratory facility, 478–480 leg ulcers, 486 nails, 487 normal skin, 474, 476–477 physical principles and techniques, 475 practical guidance, 478–480 radiography, 476 skinfold caliper, 476 ultrasonography, 475–476 variables, 477, 478–481 vascular system, 486 velocity, 475 Skinfold caliper, 476 Skin friction evaluation, 226–231 Skin impedance, 330 Skin infections, environmental influences, 35 Skin integument, 27–30 Skin layer parameters, NMR, 310–312 Skin lymph flow, 743 Skin mechanical properties, sclerotic skin, 41 Skin microvasculature, 692 Skin pH, 423, see also pH, serial tape stripping evaluation Skin research suitability, 302–303 Skin rheological model, Twistometry, 604–605
1023
Skin surface contours and wrinkles, methods comparison advantages, 211–212 basics, 205, 212 contours, 209 FOITS8, 208 image analysis, 206–207, 209 instruments, 206–209 limitations, 211–212 measurement, 206–207 optical profilometry, 209 parameters, 208–209 PRIMOS7, 208 replicas, 206–208, 211 shadowing method, 207 three-dimensional analysis, 208–209 in vitro measurement, 208, 211 wrinkles, 207, 209, 210 Skin surface pH, 413 Skin temperature, 391, 772–775, 844 Skin temperature, thermal imaging applications, 778–782 basics, 782 conduction, 772–775 convection, 772 decreased skin temperature, 780–782 examination room, 777–778 historical background, 769–770 human body temperature, 770–772 increased skin temperature, 778–780 mean skin temperature, 771–772 principles, 772 radiation, 772 radiative heat transfer, 777 skin emissivity, 775–776 skin temperature measurement, 772–775 thermal imaging systems, 777 thermal radiation, 775–778 thermal skin patterns, 778, 779 thermal symmetry, 780 Skin texture, dermatoscope, 114–115 Skin thickness atopic dermatitis, 43 microdialysis sampling, 447–448 sclerotic skin, 41 Skin type evaluation, TapeAnalyzer, 833 Slackness, see Levarometry Slides, photography, 88 Soaps, 423, 944 Sodium lauryl sulfate (SLS) atopic dermatitis, 43 irritancy, 364, 385, 482 racial differences, 29 side effects, 40 skin surface pH, 413 ultrasound imaging, 501–502 Sodium lauryl sulfate (SLS), ESCD standards abnormal skin, 949 applications, 950 basics, 944 characteristics, 944 clinical effects, 944–945 closed patch tests, 945–946 constitutional factors, 947–949 correlations, 947 effects, 944–945 environment-related variables, 949
1024
exposure methods, 945–947 histopathological effects, 944 immersion tests, 946–947 immunological effects, 944 individual pilot study, 951 individuals, 949–950 location, 949–950 occlusive tests, 945–946 one-time tests, 945–946 open tests, 946 pilot study, 951 reactivity, 947–950 recommendations, 949–952 repeated tests, 946 results interpretation, 952 testing, 949–951 test substance, 950 variables, 949 wash tests, 947 Software Capilab Toolbox, 684–685 capture, Hi-scope system, 128 Courage-Khazaka, 584–585, 588 cyberDERM, 593 Fpg23, 128, 133 GmbH, TeachScreen, 654 image processing, 93 laser profilometry, 172 ultrasound imaging, 495–496 Solar radiation, 33–34 Solution, sodium laurel sulfate, 945–946 Solutions, potential, 274 Sonography, high-resolution analysis, 247 basics, 246, 251–253 glabrous skin, normal, 247, 249–250 methods, 246–247 palmar skin, normal, 247–250 patients, 247–248 processing, 246–247 results, 248–251 skin diseases, 247–248, 250–251 statistics, 248 three-dimensional sonography, 247 Sonography, PASI, 39 SOPs, see Standard operating procedures (SOPs) Sorption-desorption test (SDT), 42 Soundness of research, 11 Sources of errors, see Error sources Spatial frequency range, 197 Spatial resolution, OCT, 258 Specialist photography, 87 Specificity, statistical analysis, 55–56 Spectrophotometric intracutaneous imaging (SIAscopy) applications, 320–321 basics, 315–317, 321–324 coloration, skin, 318–319 computation, graphs, 316, 319, 319, 322 forward predictive model, 318–319 future potential, 321 graphs, 316, 319, 319, 322 image acquisition, 320 inversion, graphs, 316, 319, 319, 322 malignant melanoma, 316, 319, 319, 322 methodological principles, 317–320 normal skin, 317–318
Handbook of Non-Invasive Methods and the Skin, Second Edition
validation, 320 Speed, imaging, 261 Spin-echo sequence, 539–540 Sponsor responsibilities, 49 Squamous cell carcinoma, 485 Staining methods, 363–366, 806, 807 Staircase procedures, 795 Standardization, photography, 88 Standardization Group of the European Society for Contact Dermatitis, 12 Standard operating procedures (SOPs), 70 Standards, hair growth measurements, 884, 886–887 Standoff pad, 518 Staphylococcus aureus, 415, 459 Staring array, 758 Start-up, fast, 91 Start-up and use, 387 Static microtopography, 926 Statistical analysis, see also Analysis; Image analysis artificial dichotomy, 56 basics, 53–55 clinical context, 56 gold standard, 54, 55–56 limitations, 55–56 morphological tree, cutaneous network of lines, 200–201 predictive value, 55–56 receiver operating characteristic curves, 56–60 recommendations, 61 regression methods, 60 sensitivity, 55–56 specificity, 55–56 stylus method equipment, 165 three-dimensional evaluation, micro- and macrorelief, 183–184 validity, 61 Statistics, 248, 541 Stefan-Boltzmann’s constant, 775 Sterile bag technique, 462 Sticky slides, 372 Stiffness, 571–572 Stiffness index, 597 Stimulus-dependent methods, 795 Storage image banks, 89, 91 saving images, 93 Sebutapes, 838–839 specialist photography, 89 Strain-time curves, 585–586 Strategic errors, 11 Stratified structure, characterization, 173 Stratum corneum dansyl chloride fluorescence, 376 epidermal hydration, 329–334 harvesting material, 371–373 high-resolution sonography, 248–249 micro-sensor mapping, 812–813, 815 racial differences, 29 twistometry, 608–609 Stratum corneum, hydration state accuracy, 355 basics, 351, 356 corneometers, 352–353 correlations, 355 DermaLab moisture unit, 353 measurement depth, 356 MoistureMeters, 353 multifrequency impedance spectrometer, 354
Index
Nova Dermal Phase Meters, 353–354 principle, 351–352 sensitivity range, 355–356 Skicon 200/200EX, 354 SkinChip, 354 variation coefficient, 356 Stratum corneum chymotryptic enzyme, 424 Stratum malphighii, 249–250 Stratum microscope, 864 Stress-strain curves, 585, 586–588, 895–896 Stress-time curves, 587–588 Stripping, 341, 459 Stripping, pH evaluation basics, 421–422 detergent effects, 423 dyes, 422 enzymes, 423–425 flat pH glass electrode, 422 measurement methods, 422 recommendations, 425 regulation, 421–422 skin pH, 422–425 soap effects, 423 tape stripping, 425–426 Structure, hair, 883–886 Stylus instrument comparison, 172–173 Stylus method, contour measurement, 163–168 Subcutaneous space, 477 Subcutaneous tissue and adjacent structure, ultrasound basics, 515 benign tumoral pathology, 525–527 cutaneous melanoma, 528 cystic lesions, 525–526 dermal disease, 522, 524 edema, 519–520, 523–524 epidermal disease, 521–522, 524 fibromatous tumors, 527 fluid accumulation, 519–520, 523–524 foreign bodies, 521, 524 indications, 529 instrumentation, 516–517 joints, 523–524, 525 lipomatous tumors, 526, 527 liposarcoma, 528 lymphadenopathy, 520–521, 524, 528 lymphatic hemangiomas, 526 malignant tumoral pathology, 528 neural disease, 522–523, 524–525 neurogenic tumors, 527 normal anatomy, 518–519, 520–523 pathology, 523–524 patients examination technique, 518–520 pitfall potential, 528 recommendations, 517 tendons, 523–524, 525 vascular disease, 522–523, 524–525 vascular hemangiomas, 526 Subepidermal band, 482, 522 Subject directories, Internet, 16–17, 19–21 Suction chamber methods Cutometer, 579–582 Dermaflex, 571–576 DermaLab, 593–598 digital Cutometer, 583–589 Langer’s lines, 567–568 Suction cup, DermaLab, 593–597
1025
Suction device diameter, 588 Suitability, skin research, 302–303 Sunlight, 574, 983–984, see also Light sources, sunlight, and radiation dosimetry Sunscreen agent evaluation, 300, 302, 303–304 Superficial epidermal layers, 279 Surface and subsurface imaging theory, 127 Surface contours and wrinkles, methods comparison advantages, 211–212 basics, 205, 212 contours, 209 FOITS8, 208 image analysis, 206–207, 209 instruments, 206–209 limitations, 211–212 measurement, 206–207 optical profilometry, 209 parameters, 208–209 PRIMOS7, 208 replicas, 206–208, 211 shadowing method, 207 three-dimensional analysis, 208–209 in vitro measurement, 208, 211 wrinkles, 207, 209, 210 Surface microscopy, dermatoscopy, 114 Surface plane measurements, 388 Surface skin temperature, sensors and handheld devices basics, 753–754 computer systems, 764 contact temperature measurement, 754 detector arrays, 760 detectors, 760–764 emerging technology, 764 expert systems, 764 fixed monitoring systems, 757 focal pane arrays, 758–760 future directions, 764–765 infrared radiometers, 757–758 infrared theory, 756 infrared thermometers, 756–757 long-wave photon detectors, 761, 762 measurement location, 754–755 noncontact temperature measurement, 756–760 photoconductive detectors, 762 photon detectors, 760–761 photovoltaic detectors, 761–762 quantum well infrared detectors, 762–763 scanning, mechanical, 758–760 short-wave photon detectors, 761, 762 temperature measurement, 754 terminology, 765–767 thermal detectors, 763–764 thermocouples, 754 thermometers, 754–755 Surface temperature, skin, 391 Survival probability curves, 906–907 Susceptibility testing, 944, 950 Suspension, clinical trials, 51 Swabbing methods, 460–461 Sweat and sweating analysis, 818–819 eccrine, 817–819 harvesting, 818, 821 measurements, semiopen systems TEWL, 391 micro-sensor mapping, 813–814, 815–816 production, stimulation, 817–818
1026
Handbook of Non-Invasive Methods and the Skin, Second Edition
skin temperature effect, 34–35 Sweat glands high-resolution sonography, 249, 250 localization, 805–808 measurements, semiopen systems TEWL, 384 sebum excretion, 844 Swept-gain curve, 495 Synovial cysts, 525 Systemic drug delivery, 450 Systemic scleroderma, 40–41
T T1, see Longitudinal relaxation time (T1) T2, see Transversal relaxation time (T2) Tactile stimulation, 791 Talcum powder, 220 TapeAnalyzer, 831–834 Tape methods, 366–367, 372 Tape stripping, pH evaluation basics, 414, 421–422, 425–426 detergent effects, 423 dyes, 422 enzymes, 423–425 flat pH glass electrode, 422 measurement methods, 422 recommendations, 425 regulation, 421–422 skin pH, 422–425 soap effects, 423 Target organ measurements, 450 TcPO2, see Transcutaneous oxygen tension (TcPO2) Technical equipment, stylus method, 164–165 Technological error, 11 Technologies, new, 713–714 Teledermatology, 93 Telogen classification, 871–872, 876–877 Temperature color, 83 measurements, semiopen systems TEWL, 388, 390–391 seasonal variations effects, 34–35 sebum excretion, 844 surface skin temperature sensors and handheld devices, 754 Tendinitis, 525 Tendinosis, 525 Tendons, 519, 521, 523–524, 525 Tenosynovitis, 525 Tensile measurements, 897–898 Tension tests, 567 Terminal hair shaft alterations, 102 Termination, clinical trials, 51 Test areas capacitance measurement, 340–341 microdialysis sampling, 449–450 pH, 414, 415 phototesting, 994 stylus method, contour measurement, 166 Test chamber, 945 Test individuals, sodium laurel sulfate, 949–950 Testing, twistometry, 602–603 Test procedures, sodium laurel sulfate, 950–953 Test results, inaccurate, 61, see also False positives and negatives; True positives and negatives Test-retest reliability, 933 Test solution, sodium laurel sulfate, 945–946
Test substance, sodium laurel sulfate, 950 Tetralogy, fallot, 674–675 Tewameter, see Semiopen systems, transepidermal water loss TEWL, see Transepidermal water loss (TEWL) Textural analysis, three-dimensional evaluation, 184–187 Texture, facial averaging methods, 96 Texture improvement, measuring, 98, 99 Theories erythema and melanin indices, 665–667 laser Doppler flowmetry, 709–710 measurements, semiopen systems TEWL, 385 twistometry, 602–605 Therapeutic activity, 241 Therapy adjunct, 274 Thermal burn-induced autofluorescence, 292 Thermal detectors, 763–764 Thermal imaging, skin temperature applications, 778–782 basics, 782 conduction, 772–775 convection, 772 decreased skin temperature, 780–782 examination room, 777–778 historical background, 769–770 human body temperature, 770–772 increased skin temperature, 778–780 mean skin temperature, 771–772 principles, 772 radiation, 772 radiative heat transfer, 777 skin emissivity, 775–776 skin temperature measurement, 772–775 thermal imaging systems, 777 thermal radiation, 775–778 thermal skin patterns, 778, 779 thermal symmetry, 780 Thermal imaging systems, 777 Thermal injury, 484 Thermal radiation, 775–778 Thermal skin patterns, 778, 779 Thermal symmetry, 780 Thermistors, 773 Thermocouples, 754, 754, 773 Thermodes, 793 Thermographic methods, 695 Thermometers, 754–755 The World of Skin Care web page, 20 Thickness, nails, 923–924 Thickness, skin, 41, 447–448, 512–513 Three-dimensional analysis, 208–209 Three-dimensional evaluation, micro- and macrorelief apparatus improvements, 180–182 basics, 179–180, 191–192 focusing system, 180–181 furrow directional quantification, 185–187 healing theory and evolution, 191 macrorelief, 187–191 microrelief, 182–187 performance and results, 189–191 roughness parameters, 182–183 statistical analysis, 183–184 study purpose, 180 textural analysis, 184–187 triangulation system, 181–182 wound quantification, 188–191 wrinkle quantification, 187–188, 189
Index
Three-dimensional image analysis, 97–99 Three-dimensional parameters, laser profilometry, 175–177 Three-dimensional scanning, 474 Three-dimensional sonography, 247 Three-dimensional tree skin spectrum, 200–202 Thrichosporum spp., 467 Thromboflebitis, 524 Time-domain, OCT, 258 Tissue characterization, 495–503 Topical application, 287 Topical corticosteroids, 501 Topical drug delivery, 449 Topical medicaments, 35 Topical retinoic acid, 609 Topical retinoids, 6 Torque application rate, 605 Torsional measurements, 897 Touchless acoustic stylus, 165–166 Tour de France, sun exposure study, 607 Tracer migration, 747 Trachyonychia, 926 Tracking photos, 82 Tracking (staircase) procedures, 795 Traction forces, 573 Traditional habits, 35 Transcutaneous oxygen tension (TcPO2) applications, 400–402 basics, 397 change dynamics, 402 correlations, 399–400 error sources, 399 methodological principle, 397–399 recommendations, 398, 403 Transcutaneous PCO2, 407–410 Transdermal drug delivery, 291, 449 Transducers, 164, 516, 523 Transepidermal water loss (TEWL) age and body region, 28 atopic dermatitis assessment, 42 closed-chamber systems, 393–396 occluded patch tests reactions, 973–974 oscillating method, 225 psoriasis, 39 racial differences, 29–30 sodium laurel sulfate, 948 Transepidermal water loss (TEWL), semiopen systems barrier composition, 385 barrier function evaluation, 384 basics, 384, 391 creams, 385 environment-related variables, 387–390 error sources, 387–391 individual-related variables, 390–391 instrument-related variables, 387–390 irritants, 384–385 methodological principle, 385–387 moisturizers, 385 objective, 384–385 predictive irritancy testing, 384–385 protective creams, 385 sweat gland activity, 384 theory, 385 validity, method and variables, 387 α-β transformation, 896 Transit times, lymph flow evaluation, 742 Transversal relaxation time (T2), 541, 544, 545–546
1027
Transverse unit, 164, 165 Trauma, insertion, 447 Traumatic scars, 484 Treatment evaluation, 274, 833 Triangulation system, 181–182 Tribological studies, skin age, 222–223 anatomic region, 222–223 basics, 215–216, 222–223 emollients, 217, 220–221 experimental designs, 216–217 friction coefficient values, 218–222 gender differences, 222–223 hydration, 217–219 lubricants, 217, 220 moisturizers, 217, 220–221 normal load, 217–218 probes, 217, 218–219, 221–222 racial differences, 222–223 talcum powder, 220 Trichograms, 871, 875–880, see also Phototrichogram (PTG) technique TRI Fatigue Tester, 905 Triphasic flow, 520 Troubleshooting, microdialysis sampling, 445, 451 True positives and negatives, 54 t-test, continuous data, 64–65 Tumoral pathology, benign, 525–527 Tumoral pathology, malignant, 528 Tumor parenchyma and stroma, 251, 252 Tumors, 264, 526–527 Turgor, 581 Tutorials, Internet, 20–21 Twistometry, skin elasticity actinic aging, 607–608 aging, 605–608 applications, 609–610 basics, 601–602, 610 connective tissue diseases, 610 equipment, 602 interpretation, 605 intrinsic aging, 604–607 mechanical testing, 602–603 parameters, mechanical, 603–605 retinoic acid, topical, 609 scleroderma, 609–610 skin biomechanics, 608–609 stratum corneum, 608–609 technical considerations, 602–605 testing, mechanical, 602–603 theoretical considerations, 602–605 topical retinoic acid, 609 validation, 605 Two-dimensional facial prototyping methods, 96–97 Two-finger squeeze method, 855
U Ulcers, leg, 486 Ultrasonography, 475–476 Ultrasound dermal water and edema, 507–509 microdialysis sampling, 447–448 occluded patch tests reactions, 975–977 PASI, 39 sclerotic skin, 40, 41
1028
skin aging, 511–513 systemic sclerosis, 41 wheals-and-flare reactions, 971 Ultrasound, subcutaneous tissue and adjacent structure basics, 515 benign tumoral pathology, 525–527 cutaneous melanoma, 528 cystic lesions, 525–526 dermal disease, 522, 524 edema, 519–520, 523–524 epidermal disease, 521–522, 524 fibromatous tumors, 527 fluid accumulation, 519–520, 523–524 foreign bodies, 521, 524 indications, 529 instrumentation, 516–517 joints, 523–524, 525 lipomatous tumors, 526, 527 liposarcoma, 528 lymphadenopathy, 520–521, 524, 528 lymphatic hemangiomas, 526 malignant tumoral pathology, 528 neural disease, 522–523, 524–525 neurogenic tumors, 527 normal anatomy, 518–519, 520–523 pathology, 523–524 patients examination technique, 518–520 pitfall potential, 528 recommendations, 517 tendons, 523–524, 525 vascular disease, 522–523, 524–525 vascular hemangiomas, 526 Ultrasound B-mode imaging, in vivo structure analysis allergic patch test reaction, 500, 501 applications, 496–497 basics, 493–494, 503–504 corticosteroids, topical, 501 equipment, 494–495 image analysis software, 495–496 instruments, 494–495 irritant reactions, 501–503 normal skin, 497–499 psoriatic skin, 503–504 technical aspects, 495 tissue characterization, 495–503 wheals, 503 Ultraviolet light and radiation colorimetry, 637, 641–644 cutaneous pain assessment, 797 follicular biopsy, 827–828 immediate effects, 34 long-term effects, 34 phototesting, 992–993 protection, 300, 302, 303–304 Ultraviolet photography, 113, 114 Understandable consent, 75 Unidirectional stress, friction evaluation basics, 225 consumer evaluation, 230 emulsion application results, 228–229 expert evaluation, 229–230 methodological principle, 225–226 physiological parameters relationship, 227–228 sensory evaluation, 228–231 skin friction, 226–231 in vivo measurement, 226–227
Handbook of Non-Invasive Methods and the Skin, Second Edition
Uniform resource locators (URLs), 15–18 Unit area trichogram, 871 Units, patch tests, 502 Upper dermis features, 282 Upper epidermis alterations, 102 Urocanic acid pathway, 412–413 Urticaria, wheals, 482 U-test, 248, 540 UV sources, 982, see also Light sources, sunlight, and radiation dosimetry
V Validation DermaLab, 596–597 dermatoscopy, 121 Levarometry, 614–615 microdialysis sampling, 448–449 SIAscopy, 320 twistometry, 605 Valid estimates, 54 Validity measurements, semiopen systems TEWL, 387 skin disease assessment guidelines, 932–933 statistical analysis, 61 Valsalva maneuvers, 520 Valvular incompetence, 524 VapoMeter, 394–395 Variables cutaneous blood flow, 693–694 Cutometer, 582 Dermaflex, 573–574 hair, 898–899 nails, 923 skin disease assessment guidelines, 932 sodium laurel sulfate, 949 suction chamber method, 573–574 ultrasound imaging, 477, 478–481, 520 Variation coefficient, 356 Vascular disease, 519, 522–523, 524–525 Vascular hemangiomas, 526 Vascular structures, 521, 523 Vascular system, 486 Vasodilatation, 29 Vasomotion laser Doppler flowmetry, 712–713 periodic fluctuations, CBF, 697–704 Velocity, 475 Velus hairs, CSSS, 240 Venous flow, 522 Venous insufficiency capillaroscopy and videocapillaroscopy, 682–683 periodic fluctuations, CBF, 702–703 ultrasound imaging, 486 Venous leg ulcer, 486 Venous occlusion plethysmography, 694–695 Venous pressure, 702–703 Verification bias, correction, 54, 57, 58–60 Vertical illumination, 128–129 Video camera’s physical properties, 654–655 VideoCap 100, 278 Videocapillaroscopy, 679–685 Video imaging, 126, 376–377 Videomicroscopy acne, 129–130, 131–132
Index
application, 129–130 basics, 125–126 cable, 127 capture software, 128 confocal microscopy image applications, 299 confocal scanning laser microscopy, 278 control unit, 128 depth of field, 127 future directions, 133 good practices, 133 hairs, 130, 133 hi-scope system, 128–130 horizontal illumination, 128–129 image grabber board, 128 light-gathering power, 126–127 light source, 127 optical head, 128 optics, 126–127 photodamage, 129 polarized light imaging, 129 rosacea, 129, 131 scaliness, 130, 132 skin lines network, 203 surface and subsurface imaging theory, 127 vertical illumination, 128–129 video imaging, 126 working distance, 126 wound healing, 130, 132 Viewing photography, 88 Viscosity, 621 Viscosity parameters, Twistometry, 603 Visible light radiation, 993–994 Visualization, apocrine sweat collection, 821 Visual techniques, desquamation rate, 363–364 Vitamin A metabolism, 423 Vitiligo lesions, 113 Vivascope 1000, 278 Volatile axillary secretions, 822 Volume, skin, 711–712 Voluntary consent, 75–76 Voronoï diagram, 684 Voxel reconstruction, 541
W Waldmann UV6 lamp, 982 Warts, 271 Washing methods, 461–462 Washout model, 735–737, 738 Wash tests, 947, 950 Water, in skin, 558–559 Water behavior, epidermis and dermis, 311–312 Water-in-oil emulsions, 342 Water sorption-desorption test, 333–334 Waveform patterns, 700–702 Wavelet analysis, 96 Wein’s law, 756 Wheals, ultrasound imaging, 482, 503 Wheals-and-flare reactions, instrumental evaluations, 967–971 Whole-body Sigma imaging system, 546 Wickham’s striae, 104 Wilcoxon-Mann-Whitney statistic, 58
1029
Wild M650 Epiluminscence microscopy, 101, 111 Williamson-Kligman scrub technique, 469 Windowing, 701 WMA, see World Medical Association (WMA) Women, see Sex differences Working distance, fiber-optic microscopy, 126 Working instructions, 71 World Medical Association (WMA), 12, 74 World Wide Web, see Internet Wounds fiber-optic microscopy, 132 Hi-scope system, 130 three-dimensional evaluation, 188–191 Wrinkles, see also Furrows and wrinkles, replica image analysis directional extraction, 199 methods comparison, 207, 209, 210 three-dimensional evaluation, micro- and macrorelief, 187–188, 189
X 133
Xenon clearance, 694 Xenon wash-out technique, cutaneous and subcutaneous blood flow rates atraumatic local labeling, 734 basics, 733 blood flow rate calculation, 738 correlations, 739 data management, 734–737 error sources, 739 loss, 738 methodological principle, 733–738 objectives, 733 physical properties, 733–734 recommendations, 739–740 registration, 734–735 washout model, 735–737, 738 XYZ conversions, 657–659
133
Y YAG lasers, 551, 794 Yield region, 896 Young’s modulus digital Cutometer, 587–588 hair strength measurement, 896 Hookean region, 896 suction chamber method, 595, 597 twistometry, 603–604, 606, 608, 610
Z Zeiss instrument, 111, 376, 641, 858 Zemtsov-Declercq-Cowie approach, 534 Zero calibration, 710–711 Zero drift, 388 Zeroing, measurements, 387 Zero situation, 574, 581 Zoom lenses, compact digital cameras, 91, 92
180 170 160 150
MeanR MeanG MeanB
140 130 120 110 100 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.1 Mean RGB values of patch 22 in successive acquisitions (5 min distance). 180 170 160 150
MeanR MeanG MeanB
140 130 120 110 100 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.2 Mean RGB values of patch 22 in successive acquisitions (1 min distance). 180 160 140 120 MeanR MeanG MeanB
100 80 60 40 20 0 0
10
20
30
40
50
60
70
80
90
100
110
120
130
Minutes
FIGURE 76.3 Mean RGB values of patch 22 in successive acquisitions (1 min distance with pauses every six acquisitions). 100 90 80 70 60 50 40
Colorimeter Y
30 20 10 0 0
10
20
30
40
50
60
70
80
90
Declared reflectance
FIGURE 76.4 Comparison of declared and measured reflectance of the KODAK Gray Scale.
100
1 0,9 0,8 0,7 0,6
dr dg db
0,5 0,4 0,3 0,2 0,1 0 0
0,05
0,1
0,15
0,2
0,25
0,3
0,35
0,4
0,45
0,5
Colorimeter measured ratio Y/Yn
FIGURE 76.5 Normalized RGB values of the video camera for gamma estimation.
FIGURE 76.6 Declared sRGB values of the ColorChecker and corresponding values as measured by the instrument.
FIGURE 76.7 Comparison of the visual aspect of the same color patches acquired with two different equipments.
FIGURE 76.8 Worst calibration result: first line shows the original images, second line shows the results of lighting correction, third line shows the calibrated images.
FIGURE 76.9 Best calibration result: first line shows the original images, second line shows the results of lighting correction, third line shows the calibrated images.