ELECTRODIAGNOSIS IN DISEASES OF NERVE AND MUSCLE
This page intentionally left blank
ELECTRODIAGNOSIS IN DISEASES OF NERVE AND MUSCLE Principles and Practice Edition 3 JUN KIMURA, M.D. Professor Emeritus Department of Neurology Kyoto University Kyoto, Japan Professor Division of Clinical Electrophysiology Department of Neurology University of Iowa College of Medicine Iowa City, Iowa
OXFORD UNIVERSITY PRESS
2001
OXFORD UNIVERSITY PRESS
Oxford New York Athens Auckland Bangkok Bogota Buenos Aires Calcutta Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Shanghai Singapore Taipei Tokyo Toronto Warsaw and associated companies in Berlin Ibadan Copyright © 2001 by Oxford University Press, Inc. Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016 Oxford is a registered trademark of Oxford University Press All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Library of Congress Cataloging-in-Publication Data Kimura, Jun. Electrodiagnosis in diseases of nerve and muscle: principles and practice / Jun Kimura.—Ed. 3. p. ; cm. Includes bibliographical references and index. ISBN 0-19-512977-6 (cloth : alk. paper) 1. Neuromuscular diseases—Diagnosis. 2. Electromyography. 3. Electrodiagnosis. I. Title. [DNLM: 1. Neuromuscular Diseases—diagnosis. 2. Electrodiagnosis—methods. 3. Neural Conduction—physiology. 4. Spinal Cord Diseases—diagnosis. 5. Synaptic Transmission—physiology. WE 550 K49e 2000] RC925.7 .K55 2000 616.7'407547—dc21 00-025011 As new scientific information becomes available through basic and clinical research, recommended procedures undergo changes. The author and publisher have done everything possible to make this book accurate, up-to-date, and in accord with accepted standards at the time of publication. Nonetheless, the reader is advised always to check changes and new information regarding the current practice and contraindications before conducting any tests. Caution is especially urged with new or infrequently used equipment.
1 3 5 7 9 8 6 4 2 Printed in the United States of America on acid-free paper
To Junko and our growing family
This page intentionally left blank
PREFACE AND ACKNOWLEDGMENTS
I know it is unwise to open introductory remarks with excuses because they instantaneously weaken the impact and preempt the thrust of the message. Nonetheless, I wish to offer an explanation for the enormous delay of this publication in spite of my good intention to comply with the warm encouragement for a timely revision. In 1989, immediately after the completion of the second edition, unforeseen turns of events prompted my unscheduled return to Kyoto to teach at my alma mater. I was able to maintain close ties with the University of Iowa thanks to Dr. Antonio Damasio, Chair of the Department of Neurology, who offered me a joint appointment with a ten-year leave of absence. Despite this liaison, an unplanned relocation at once doomed any hopes of repeating a six year cycle of revisions from the original date of publication in 1983. Besides, continued involvement with the Muscle & Nerve, as well as the International Federation of Clinical Neurophysiology (IFCN) and the World Federation of Neurology (WFN), in essence, precluded any progress toward a timely finish. Additionally, I found myself in the midst of a re-entry crisis after a 30 year stay abroad which climaxed with a sadly misdirected legal probe into our accounting of research donation from pharmaceutical industry and the fund raising for the 10th International Congress of EMG and Clinical Neurophysiology (X-ICEMGCN Kyoto, 1995). Ironically, the incident provided me with an unexpected opportunity to concentrate on rewriting the manuscripts as prolonged and persistent press surveillance all but forced my unintended seclusion at home. In fact, I found myself in an ideal position to update the text using ample and uninterrupted time with which to analyze bits and pieces of documents assembled over the past several years. The turmoil, which was widely publicized in Japan as one of the biggest scandals of 1997, ended triumphantly (for us) with the resignation of the prosecutor in charge and the disbandment of the special task force, proving yet again that justice may be blindfolded but the truth always sets you free. vii
During this interim, some new fields have emerged in our discipline requiring their inclusion as additional chapters and other existing areas have gained importance in patient care necessitating expanded coverages. These include magnetic stimulation, human reflexes, late responses, motor unit number estimate, quantitative electromyography, and threshold electrotonus to mention only a few. Advanced technology has brought considerable modifications in theory and practice in many other areas, although the basic premises remain the same in electrophysiologic approaches and in clinical problem solving. Parallel advances in other fields of neuroscience have led to equally exciting progress in the exploration of many disease processes in general, and to the understanding of neuropathies, muscular dystrophies, myasthenic syndrome and movement disorders in particular. I have thus rewritten, in their entirety, the chapters on principles and variation of conduction studies; facts, fallacies and fancies of nerve stimulation technique; other techniques to assess nerve function; the F wave and A wave; somatosensory and motor evoked potentials; electrodiagnosis in the pediatric population; and all the clinical sections included in Part VI and Part VII with the addition of ethical consideration in clinical practice as an appendix. The remaining chapters also underwent substantial changes to reflect current understanding. The addition of some 2500 new papers, which I personally reviewed, attest to the incredible advances in what was once considered to be a static, rather than dynamic, field of clinical electrophysiology. To achieve comprehensive coverage, I retained most of the old articles to document earlier contributions. However, for the sake of brevity, the text emphasizes basic principles, summarizing only pertinent points for day to day practice. The inclusion of ample new references should enable interested readers to consult the original papers for further details. My decision to take on this venture affected—directly or indirectly—many innocent bystanders who had to shoulder additional workloads while I devoted myself in writing. In particular, I owe special thanks to our staff in Kyoto guided by Dr. Ichiro Akiguchi who, along with Dr. Shinichi Nakamura, Dr. Nobuyuki Oka and Dr. Shun Shimohama, assumed many administrative chores. Dr. Ryuji Kaji, together with Dr. Nobuo Kohara, supervised the Clinical Neurophysiology Laboratory where our post-doctoral fellows contributed many new research insights useful for this revision. I am most grateful to Professor Hiroshi Shibasaki and his staff, which consisted of Dr. Hidenao Fukuyama, Dr. Takashi Nagamine and Dr. Akio Ikeda of the Department of Clinical Brain Pathophysiology for their support. Our secretarial personnel, including Mari Yamane, Kayoko Morii, Kyoko Maekawa, Tomoko Noboru and Kumiko Imai, processed an enormous amount of English literature without having prior exposure to foreign materials. Last, but not least, Machiko Miyamoto typed and retyped the entire volume single-handedly, as she was the only Japanese assistant proficient in English. It was my good fortune to be able to complete the revision at the Division of Clinical Neurophysiology at the University of Iowa headed by Dr. Thoru Yamada, who directs the EEG section there with the assistance of Drs. Malcom Yeh and Dr. Mark Granner. I enjoyed a most flexible time schedule in the EMG section thanks to Dr. Edward Aul who filled in whenever necessary, along with the help of viii
Dr. Torage Shivapour and Dr. Jon Tippin. Dr. Eric Dyken, in charge of the Sleep Disorder Laboratory, read the entire book and provided useful suggestions. David Walker, M.S.E.E., rewrote the appendix on electronics, which was previously co-authored by Pete Seaba, M.S.E.E., who also gave helpful advice. I am indebted to Sheila Mennen, Shelli Hahn and Leigha Rios for their indispensable technical and clerical help during my renewed part-time work in Iowa. Ms. Mermen had already assisted with the first and second editions, therefore, I appreciated her sentiment when she inquired as to whether there would ever be a fourth edition! I would also like to thank Lauren Enck and Susan Hannan at Oxford University Press for inheriting the 3rd edition from F.A. Davis in the midst of production, and for guiding me with patience and encouragement despite the slow progress of the adopted project. I am indebted to the American Association of Electrodiagnostic Medicine (AAEM) and its Nomenclature Committee, who granted me permission to reprint the AAEE Glossary of Terms in Clinical Electromyography (1987) as Appendix 5. In concluding this acknowledgment, I wish to update a personal note on our household which, in the earlier editions, triggered many kind remarks. We now have an attorney in San Francisco, a resident physician in Madison, and a counselor for handicapped children in our home base in Iowa City. Junko, my wife, often refers to herself as "an international cleaning lady", and periodically visits all five posts including my retreat in Kyoto, which is buried under manuscripts and always ranks bottom in her assessment. Our boys have finally grown old enough to appreciate the magnitude of the work involved in producing a textbook. For my 60th birthday, which in Japan customarily warrants a special celebration in recognition of one's accomplishment (regardless of any achieved), our three sons and a daughter-in-law came to Kyoto to honor my endeavors. I consider it my good fortune to be able to work on this edition in such a warm and conducive environment. This book is again dedicated to Junko in appreciation for her companionship, and to our growing family to acknowledge their compassionate, albeit spiritual, support. I take great comfort in the thought that, at long last, we may endow the royalty from this book as down payment for their first homes rather than tuition: I am thrilled that, for a change, we can actually witness the rewards of our investment! Kyoto, Japan
J. K.
ix
This page intentionally left blank
PREFACE FOR THE SECOND EDITION
The preparation for the second edition began in 1983 with the original volume still in press, literally before the ink had dried. Kind encouragements and constructive criticisms received from different corners of the world added further incentive for early revision. Most suggestions proved helpful in improving the contents and style. A few requests, however, posed problems because they represented mutually exclusive views: for example, inclusion or exclusion of expanded coverage of evoked potential studies. Here, I had to accept the old maxim that, however much one wishes, one cannot please everybody all the time (or even most people much of the time!). Thus, I followed my own bias as to the relative importance of a topic for the principles and practice of electrodiagnosis. This revision, though initially conceived as routine and minor, eventually required major changes, in part reflecting the rapid medical and technologic advances in the field during the past five years. The sections rewritten in their entirety include Facts, Fallacies, and Fancies of Nerve Stimulation Techniques (Chapter 7), Single-Fiber and Macro Electromyography (Chapter 15), Somatosensory and Motor Evoked Potentials (Chapter 19), Polyneuropathies (Chapter 22), Myasthenia Gravis and Other Disorders of Neuromuscular Transmission (Chapter 24), Myopathies (Chapter 25), and Fundamentals of Electronics and Electrical Safety (Appendices 2 and 3). Most other sections also underwent substantial changes to update, clarify, and tighten the contents. The book now cites more than 1200 additional references selected from some 2500 recent publications that I personally reviewed, with the hope that the inclusive bibliography helps encourage further research in the area of electrodiagnostic medicine. The hustle and bustle inherent to the preparation of voluminous manuscripts, by necessity, involve directly or indirectly those who share the work environment with the author. I could not have completed the job without assistance from my colleagues, who endured the fate of "galley" slaves over an extended period of time. Drs. Thoru Yamada and Stokes Dickins ran the busy services of the Division xi
despite my preoccupation with writing. D. David Walker, M.S.E.E., rewrote the appendix on electronics previously coauthored by Pete Seaba, M.S.E.E., who left the ranks for private enterprise. Sheila Mennen, our Chief Technologist, Deborah Gevock, and Cheri Doggett played major roles in maintaining the daily clinical operation and organizing technical as well as secretarial needs. A number of clinical fellows and residents participated in teaching sessions, shedding new insights into the type of coverage essential in an electrodiagnostic text. A total of 35 research fellows from Japan and elsewhere spent one to two years with us during this interim contributing original data in clinical electrophysiology, much of which found its way into the revised text. Dr. Maurice Van Allen, who had provided a kind foreword for the first edition, continued to support my literary endeavor until his untimely death in 1986. I lost a teacher and friend, and a new foreword, which he had promised. He had jokingly, but perhaps with good reason, attributed the success of the first edition to his opening remarks, which are retained in his honor. Dr. A. L. Sahs, who initiated me into neurology, rendering help when I needed it most, also passed away later in the same year. It was my good fortune that the Department prospered under the direction of Dr. Antonio Damasio, who, together with Dr. Robert L. Rodnitzky, provided the kind of environment enticing to academic pursuit. I owe my thanks to Mr. Robert H. Craven, Sr., Mr. Robert H. Craven, Jr., Dr. Sylvia Fields, Ms. Linda Weinerman, Ms. Jessie Raymond, and Mr. Herbert Powell of F.A. Davis for their patience and encouragement. I am indebted to the American Association of Electromyography and Electrodiagnosis and its Nomenclature Committee, who granted permission to reprint the AAEE Glossary of Terms in Clinical Electromyography (1987) as Appendix 4. The work turned into a family project of sorts over the past several years. Our three sons, five years older and perhaps wiser, if not quieter, could now assist in substance by typing the book, cover to cover, into a word processor to facilitate rewriting. I acknowledge the yeomans' service by honoring their request again to dedicate the book to their mother, who, I know, has funded the teenagers' operation from time to time to boost their spirit of devotion. We lost her father and mine during the preparation of the first edition and my mother in this interim. I salute them for their constant support of our venture abroad, with the credit given to whom it is most justifiably due. J. K.
xii
FOREWORD FOR THE FIRST EDITION
I found particular pleasure in preparing this foreword to the work of a colleague whose professional development and scientific accomplishments I have followed very closely indeed for some twenty years. Dr. Kimura, very early after his training in neurology, expressed an interest in clinical electrophysiology. His energy and talents led to full-time assignment and responsibility for the development and application of electrodiagnostic techniques in our laboratory of electromyography and then to direction of the Division of Clinical Electrophysiology. From his early assignment, Dr. Kimura has exploited the possibilities for the applications of clinical electrophysiologic techniques to their apparent limits, which, however, seem to continually advance to the benefit of us all. This volume is based on very extensive personal experience with application of all of the now recognized procedures. The beginner will be able to follow this discipline from its historical roots to the latest techniques with the advantage of an explanatory background of the clinical, physiologic, anatomic, and pathologic foundations of the methods and their interpretation. The instrumentation, so essential to any success in application of techniques, is further described and explained. The more experienced diagnostician will both appreciate and profit from this pragmatic, well-organized, and authoritative source with its important bibliographic references; the beginner will find it a bible. There are few areas in electrodiagnosis that Dr. Kimura does not address from his own extensive experience, backed by clinical and pathologic confirmation. The sections on the blink reflex and the F wave reflect his own pioneering work. He has closely followed the application of new techniques to the study of disease of the central nervous system by evoked cerebral potentials from the beginning. These sections reflect a substantial personal experience in establishment of standards and in interpretation of changes in disease. So important are the findings of electrodiagnostic methods that xiii
the clinical neurologist must himself be an expert in their interpretation. Preferably he should perform tests on his own patients or closely supervise such tests. Only in this way can he best derive the data that he needs or direct the examination in progress to secure important information as unexpected findings appear. To acquire the knowledge to guide him either in supervised training or in self-teaching, he needs first an excellent and comprehensive guide such as this text by Dr. Kimura. Dr. Kimura is justifiably regarded as a leader in clinical electrophysiology both nationally and internationally. Those of us who profit from daily contact with him should be pardoned for our pride in this substantial and authoritative work. Maurice W. Van Allen, M.D.
xiv
PREFACE FOR THE FIRST EDITION
This book grew out of my personal experience in working with fellows and residents in our electromyography laboratory. It is intended for clinicians who perform electrodiagnostic procedures as an extension of their clinical examination. As such, it emphasizes the electrical findings in the context of the clinical disorder. Although the choice of material has been oriented toward neurology, I have attempted to present facts useful to practicing electromyographers regardless of their clinical disciplines. I hope that the book will also prove to be of value to neurologists and physiatrists who are interested in neuromuscular disorders and to others who regularly request electrodiagnostic tests as an integral part of their clinical practice. The book is divided into seven parts and three appendices. Part 1 provides an overview of basic anatomy and physiology of the neuromuscular system. Nerve conduction studies, tests of neuromuscular transmission, and conventional and single-fiber electromyography are described in the next three parts. Part 5 covers supplemental methods designed to test less accessible regions of the nervous system. The last two parts are devoted to clinical discussion. The appendices consist of the historical review, electronics and instrumentation, and a glossary of terms. The selection of technique is necessarily influenced by the special interest of the author. Thus, in Part 5, the blink reflex, F wave, H reflex, and somatosensory evoked potential have been given more emphasis than is customary in other texts. I hope that I am not overestimating their practical importance and that these newer techniques will soon find their way into routine clinical practice. This is, of course, not to de-emphasize the conventional methods, which I hope are adequately covered in this text. The ample space allocated for clinical discussion in Parts 6 and 7 reflects my personal conviction that clinical acumen is a prerequisite for meangingful electrophysiologic evaluations. Numerous references are provided to xv
document the statements made in the text. I hope that use of these references will promote interest and research in the field of electrodiagnosis. J. K.
xvi
ACKNOWLEDGMENTS FOR THE FIRST EDITION
I came from the Island of the Rising Sun, where English is not the native language. It was thus with trepidation that I undertook the task of writing an English text. Although its completion gives me personal pride and satisfaction, I hasten to acknowledge that the goal could not have been achieved without help from others. Dr. M. W. Van Allen has provided me with more than a kind foreword. I wish to thank him for his initial encouragement and continued support and advice. He was one of the first to do electromyography in Iowa. During my early years of training I had the pleasure of using his battery-operated amplifier and a homemade loudspeaker (which worked only in his presence). I am indebted to Dr. A. L. Sahs, who initiated me into the field of clinical neurology, and Dr. J. R. Knot, who taught me clinical neurophysiology. I am grateful to Drs. T. Yamada and E. Shivapour for attending the busy service of the Division of Clinical Electrophysiology while I devoted myself to writing. Dr. Yamada also gave me most valuable assistance in preparing the section on central somatosensory evoked potentials, which includes many of his original contributions. Drs. R. L. Rodnitzky, E. P. Bosch, J. T. Wilkinson, A. M. Brugger, F. O. Walker, and H. C. Chui read the manuscript and gave most helpful advice. Peter J. Seaba, M.S.E.E., and D. David Walker, M.S.E.E., our electrical engineers, contributed Appendix 2 and reviewed the text. My special thanks go to the technicians and secretaries of the Division of Clinical Electrophysiology. Sheila R. Mermen, the senior technician of our electromyography laboratory, typed (and retyped time and time again) all the manuscript with devotion and dedication. Deborah A. Gevock, Cheri L. Doggett, Joanne M. Colter, Lauri Longnecker, Jane Austin, Sharon S. Rath, Lori A. Garwood, and Allen L. Frauenholtz have all given me valuable technical or secretarial assistance. Linda C. Godfrey and her staff in the Medical Graphics Department have been most helpful in preparing illustrations. xvii
I owe my gratitude to Mr. Robert H. Craven, Sr., Mr. Robert H. Craven, Jr., Dr. Sylvia K. Fields, Miss Agnes A. Hunt, Ms. Sally Burke, Miss Lenoire Brown, Mrs. Christine H. Young, and two anonymous reviewers of the F. A. Davis Company for their interest and invaluable guidance. A number of previously published figures and tables have been reproduced with permission from the publishers and authors. I wish to express my sincere appreciation for their courtesy. The sources are acknowledged in the legends. The Glossary of Terms Commonly Used in Electromyography of the American Association of Electromyography and Electrodiagnosis is reprinted in its entirety as Appendix 3, with kind permission from the Association and the members of the Nomenclature Committee. My sons asked if the book might be dedicated to them for having kept mostly, though not always, quiet during my long hours of writing at home. However, the honor went to their mother instead, a decision enthusiastically approved by the children, in appreciation for her effort to keep peace at home. In concluding the acknowledgment, my heart goes to the members of my family in Nagoya and those of my wife's in Takayama, who have given us kind and warm support throughout our prolonged stay abroad. The credit is certainly theirs for my venture finally coming to fruition. J. K.
xviii
CONTENTS
Part I BASICS OF ELECTRODIAGNOSIS Chapter 1 ANATOMIC BASIS FOR LOCALIZATION 1. INTRODUCTION 2. CRANIAL NERVES 3. ANTERIOR AND POSTERIOR RAMI 4. CERVICAL AND BRACHIAL PLEXUSES 5. PRINCIPAL NERVES OF THE UPPER LIMB 6. LUMBAR PLEXUS AND ITS PRINCIPAL NERVES 7. SACRAL PLEXUS AND ITS PRINCIPAL NERVES
Chapter 2 ELECTRICAL PROPERTIES OF NERVE AND MUSCLE 1. 2. 3. 4.
INTRODUCTION TRANSMEMBRANE POTENTIAL GENERATION OF ACTION POTENTIAL VOLUME CONDUCTION AND WAVEFORM
Chapter 3 ELECTRONIC SYSTEMS AND DATA ANALYSIS 1. INTRODUCTION 2. ELECTRODES
3 4 5 8 10 13 20 22
27 27 28 30 33
39 40 40 xix
3. 4. 5. 6. 7. 8. 9.
ELECTRODE AMPLIFIERS VISUAL DISPLAYS OTHER RECORDING APPARATUS ARTIFACTS STIMULATORS NORMATIVE DATA AND STATISTICS EXPERT SYSTEMS AND QUALITY DEVELOPMENT
43 45 46 47 52 53 55
Part II NERVE CONDUCTION STUDIES Chapter 4 ANATOMY AND PHYSIOLOGY OF THE PERIPHERAL NERVE 1. INTRODUCTION 2. ANATOMY OF PERIPHERAL NERVES 3. PHYSIOLOGY OF NERVE CONDUCTION 4. TYPES OF NERVE FIBERS AND IN VITRO RECORDING 5. CLASSIFICATION OF NERVE INJURIES 6. INVOLVEMENT OF AXON VERSUS MYELIN IN NEUROPATHIC DISORDERS
Chapter 5 PRINCIPLES AND VARIATIONS OF NERVE CONDUCTION STUDIES 1. 2. 3. 4. 5. 6. 7. 8.
INTRODUCTION ELECTRICAL STIMULATION OF THE NERVE RECORDING OF MUSCLE AND NERVE POTENTIALS MOTOR NERVE CONDUCTION SENSORY NERVE CONDUCTION NERVE CONDUCTION IN THE CLINICAL DOMAIN STUDIES OF THE AUTONOMIC NERVOUS SYSTEM OTHER EVALUATION OF NERVE FUNCTION
Chapter 6 ASSESSMENT OF INDIVIDUAL NERVES 1. INTRODUCTION 2. COMMONLY TESTED NERVES IN THE UPPER LIMB XX
63 63 64 67 69 72 79
91 92 92 94 96 104 108 113 117
130 131 131
3. OTHER NERVES DERIVED FROM THE CERVICAL OR THORACIC NERVE ROOTS 4. COMMONLY TESTED NERVES IN THE LOWER LIMB 5. OTHER NERVES DERIVED FROM THE LUMBOSACRAL NERVE ROOTS 6. CRANIAL NERVES
151 157 167 171
Chapter 7 FACTS, FALLACIES, AND FANCIES OF NERVE STIMULATION TECHNIQUES
178
1. INTRODUCTION 2. COMMON TECHNICAL ERRORS 3. SPREAD OF STIMULATION CURRENT 4. ANOMALIES AS SOURCES OF ERROR 5. PRINCIPLES AND PITFALLS OF WAVEFORM ANALYSIS 6. STUDIES OVER SHORT AND LONG DISTANCES
178 179 180 187 192 205
Chapter 8 OTHER TECHNIQUES TO ASSESS NERVE FUNCTION 1. MOTOR UNIT NUMBER ESTIMATES 2. ASSESSMENT OF NERVE EXCITABILITY 3. THRESHOLD TRACKING
215 215 219 224
Part III ASSESSMENT OF NEUROMUSCULAR TRANSMISSION Chapter 9 ANATOMY AND PHYSIOLOGY OF THE NEUROMUSCULAR JUNCTION 1. INTRODUCTION 2. ANATOMY OF THE NEUROMUSCULAR JUNCTION 3. ELECTRICAL ACTIVITY AT THE END PLATE 4. EXCITATION-CONTRACTION COUPLING 5. ABNORMALITIES OF NEUROMUSCULAR TRANSMISSION 6. TIME COURSE OF NEUROMUSCULAR TRANSMISSION
239 239 240 242 244 245 248
xxi
Chapter 10 TECHNIQUES OF REPETITIVE STIMULATION 1. 2. 3. 4. 5.
INTRODUCTION METHODS AND TECHNICAL FACTORS COMMONLY USED NERVES AND MUSCLES RECOVERY CURVES BY PAIRED STIMULATION DECREMENTAL RESPONSE AT SLOW RATES OF STIMULATION 6. INCREMENTAL RESPONSE AT FAST RATES OF STIMULATION 7. EFFECT OF TETANIC CONTRACTION 8. CHANGES IN MYOGENIC DISORDERS
257 258 258 260 262 263 266 270 273
Chapter 11 ACTIVATION PROCEDURES AND OTHER METHODS 1. 2. 3. 4.
INTRODUCTION PROVOCATIVE TECHNIQUES ELECTROMYOGRAPHY OTHER TECHNIQUES
279 279 280 280 281
Part IV ELECTROMYOGRAPHY Chapter 12 ANATOMY AND PHYSIOLOGY OF THE SKELETAL MUSCLE 1. 2. 3. 4. 5. 6.
INTRODUCTION FUNCTIONAL ANATOMY TYPES OF MUSCLE FIBERS STRETCH-SENSITIVE RECEPTORS ANATOMY OF THE MOTOR UNIT PHYSIOLOGY OF THE MOTOR UNIT
287 287 288 291 294 296 298
Chapter 13 TECHNIQUES TO ASSESS MUSCLE FUNCTION 1. INTRODUCTION
xxii
307 308
2. 3. 4. 5. 6. 7. 8.
PRINCIPLES OF ELECTROMYOGRAPHY INSERTIONAL ACTIVITY END-PLATE ACTIVITIES MOTOR UNIT ACTION POTENTIAL QUANTITATIVE MEASUREMENTS DISCHARGE PATTERN OF MOTOR UNITS OTHER MEASURES OF MUSCLE FUNCTION
Chapter 14 TYPES OF ELECTROMYOGRAPHIC ABNORMALITIES 1. 2. 3. 4. 5. 6.
INTRODUCTION INSERTIONAL ACTIVITY MYOTONIC DISCHARGE SPONTANEOUS ACTIVITY MOTOR UNIT POTENTIALS RECRUITMENT PATTERN
Chapter 15 EXAMINATION OF NONLIMB MUSCLES 1. INTRODUCTION 2. MUSCLES OF THE FACE, LARYNX, AND NECK 3. EXTRAOCULAR MUSCLES 4. TRUNCAL MUSCULATURE 5. ANAL SPHINCTER
Chapter 16 SINGLE-FIBER AND MACRO ELECTROMYOGRAPHY 1. INTRODUCTION 2. RECORDING APPARATUS 3. SINGLE-FIBER POTENTIAL 4. FIBER DENSITY 5. JITTER AND BLOCKING 6. MACRO AND SCANNING ELECTROMYOGRAPHY 7. CLINICAL VALUES AND LIMITATIONS
309 310 312 314 317 320 325
339 339 340 343 346 356 362
370 370 371 373 377 379
384 384 385 386 387 389 394 397
Part V SPECIAL TECHNIQUES AND STUDIES IN CHILDREN xxiii
Chapter 17 THE BLINK REFLEX 1. 2. 3. 4.
INTRODUCTION DIRECT VERSUS REFLEX RESPONSES NORMAL VALUES IN ADULTS AND INFANTS NEUROLOGIC DISORDERS WITH ABNORMAL BLINK REFLEX 5. ANALYSIS OF THE R1 COMPONENT 6. ANALYSIS OF THE R2 COMPONENT
Chapter 18 THE F WAVE AND THE A WAVE 1. 2. 3. 4. 5.
INTRODUCTION PHYSIOLOGY OF THE F WAVE THE A WAVE AND OTHER LATE RESPONSES DETERMINATION OF F-WAVE LATENCY MOTOR CONDUCTION TO AND FROM THE SPINAL CORD 6. THE F WAVE IN HEALTH AND DISEASE
Chapter 19 H, T, MASSETER, AND OTHER REFLEXES 1. INTRODUCTION 2. H REFLEX AND T REFLEX 3. THE MASSETER AND PTERYGOID REFLEX 4. THE TONIC VIBRATION REFLEX 5. THE SILENT PERIOD, LONG-LATENCY REFLEX, AND CORTICAL RESPONSE 6. OTHER REFLEXES
Chapter 20 THE SOMATOSENSORY EVOKED POTENTIAL 1. INTRODUCTION 2. TECHNIQUES AND GENERAL PRINCIPLES 3. FIELD THEORY 4. NEURAL SOURCES OF VARIOUS PEAKS 5. PATHWAYS FOR SOMATOSENSORY POTENTIALS 6. CLINICAL APPLICATIONS
xxiv
409 409 410 416 420 429 430
439 439 440 443 446 448 449
466 466 467 474 477 478 482
495 496 496 499 503 519 525
Chapter 21 MOTOR EVOKED POTENTIALS
553
1. INTRODUCTION 554 2. ELECTRICAL STIMULATION OF THE BRAIN AND SPINAL CORD 554 3. TRANSCRANIAL MAGNETIC STIMULATION 556 4. STUDIES OF THE PERIPHERAL NERVE 562 5. CENTRAL CONDUCTION TIME 565 6. JERK-LOCKED AVERAGING 566 7. CLINICAL APPLICATIONS 567
Chapter 22 ELECTRODIAGNOSIS IN THE PEDIATRIC POPULATION 1. INTRODUCTION 2. PRACTICAL APPROACH 3. MATURATIONAL PROCESS 4. NERVE CONDUCTION STUDIES 5. LATE RESPONSES 6. BLINK REFLEX 7. TESTS OF NEUROMUSCULAR TRANSMISSION 8. ELECTROMYOGRAPHY 9. SOMATOSENSORY EVOKED POTENTIALS 10. THE FLOPPY INFANT
586 586 586 588 589 590 591 591 593 594 594
Part VI DISORDERS OF THE SPINAL CORD AND PERIPHERAL NERVOUS SYSTEM Chapter 23 MOTOR NEURON DISEASES AND MYELOPATHIES 1. 2. 3. 4. 5. 6. 7. 8.
INTRODUCTION MOTOR NEURON DISEASE SPINAL MUSCULAR ATROPHY CREUTZFELDT-JAKOB DISEASE POLIOMYELITIS SYRINGOMYELIA MULTIPLE SCLEROSIS OTHER MYELOPATHIES
599 599 600 606 611 612 613 614 615 XXV
Chapter 24 RADICULOPATHIES AND PLEXOPATHIES 1. 2. 3. 4. 5.
INTRODUCTION CERVICAL AND THORACIC ROOTS BRACHIAL PLEXUS LUMBOSACRAL ROOTS LUMBOSACRAL PLEXUS
Chapter 25 POLYNEUROPATHIES
628 628 629 632 637 641
650
1. INTRODUCTION 651 2. NEUROPATHIES ASSOCIATED WITH GENERAL MEDICAL CONDITIONS 652 3. INFLAMMATORY, INFECTIVE, AND AUTOIMMUNE NEUROPATHIES 661 4. METABOLIC AND TOXIC NEUROPATHIES 668 5. INHERITED NEUROPATHIES 671
Chapter 26 MONONEUROPATHIES AND ENTRAPMENT SYNDROMES 1. INTRODUCTION 2. CRANIAL NERVES 3. PHRENIC NERVE AND NERVES IN THE SHOULDER GIRDLE 4. RADIAL NERVE 5. MEDIAN NERVE 6. ULNAR NERVE 7. NERVES OF THE PELVIC GIRDLE 8. COMMON PERONEAL NERVE 9. TIBIAL NERVE 10. SURAL NERVE 11. OTHER MONONEUROPATHIES
711 712 713 715 717 719 724 727 730 731 732 732
Part VII DISORDERS OF MUSCLE AND THE NEUROMUSCULAR JUNCTION Chapter 27 MYASTHENIA GRAVIS AND OTHER DISORDERS OF NEUROMUSCULAR TRANSMISSION 753 1. INTRODUCTION
xxvi
753
2. 3. 4. 5. 6.
MYASTHENIA GRAVIS LAMBERT-EATON MYASTHENIC SYNDROME MYASTHENIA IN INFANCY BOTULISM OTHER DISORDERS
Chapter 28 MYOPATHIES 1. 2. 3. 4. 5. 6. 7.
INTRODUCTION MUSCULAR DYSTROPHY CONGENITAL MYOPATHY METABOLIC MYOPATHY ENDOCRINE MYOPATHY MYOSITIS OTHER MYOPATHIES
754 758 761 764 765
778 779 779 787 790 796 797 802
Chapter 29 DISEASES CHARACTERIZED BY ABNORMAL MUSCLE ACTIVITY
821
1. INTRODUCTION 2. MYOTONIA 3. PERIODIC PARALYSIS 4. NEUROMYOTONIA 5. SCHWARTZ-JAMPEL SYNDROME 6. MYOKYMIA 7. HEMIFACIAL AND HEMIMASTICATORY SPASM 8. TETANUS 9. TETANY 10. STIFFMAN SYNDROME 11. CRAMPS 12. CONTRACTURE 13. MYOCLONUS 14. TREMOR 15. MIRROR MOVEMENT 16. RESTLESS LEGS SYNDROME 17. DYSTONIA
821 822 827 829 831 831 832 834 834 834 835 836 837 838 839 839 839
Appendix 1 ETHICAL CONSIDERATIONS IN CLINICAL PRACTICE
859 xxvii
Appendix 2 FUNDAMENTALS OF ELECTRONICS 1. 2. 3. 4. 5. 6. 7. 8. 9.
INTRODUCTION ELECTRICAL CONCEPTS AND MEASURES ELECTRIC CIRCUITS AND CIRCUIT LAWS CAPACITANCE INDUCTANCE AC CIRCUITS FILTERS SOLID-STATE DEVICES DIGITAL ELECTRONICS
Appendix 3 ELECTRICAL SAFETY
861 862 862 863 865 868 871 872 875 876
879
1. INTRODUCTION 879 2. THE ELECTRICAL HAZARD SITUATION 879 3. THE SAFETY PROBLEM—LEAKAGE CURRENT AND LOSS OF GROUND 882 4. ADDITIONAL SAFETY CONCERNS 882 5. SAFETY REGULATION DOCUMENTS 883 6. PROTOCOL FOR LABORATORY SAFETY 883 7. SPECIAL SAFETY DEVICES AND CIRCUITS 885
Appendix 4 HISTORICAL REVIEW 1. 2. 3. 4.
INTRODUCTION EARLY DEVELOPMENTS CLASSICAL ELECTRODIAGNOSIS ELECTROMYOGRAPHY AND NERVE STIMULATION TECHNIQUES 5. RECENT DEVELOPMENTS
Appendix 5 AAEE GLOSSARY OF TERMS IN CLINICAL ELECTROMYOGRAPHY INTRODUCTION ALPHABETICAL LIST OF TERMS WITH DEFINITIONS ILLUSTRATIONS OF SELECTED WAVEFORMS TERMS GROUPED BY SUBJECT WITHOUT DEFINITION
INDEX xxviii
887 887 888 889 890 892
897 898 898 919 943
951
Part I BASICS OF ELECTRODIAGNOSIS
This page intentionally left blank
Chapter 1 ANATOMIC BASIS FOR LOCALIZATION
1. INTRODUCTION 2. CRANIAL NERVES Facial Nerve Trigeminal Nerve Accessory Nerve 3. ANTERIOR AND POSTERIOR RAMI 4. CERVICAL AND BRACHIAL PLEXUSES Phrenic Nerve Dorsal Scapular Nerve Suprascapular Nerve Musculocutaneous Nerve Axillary Nerve 5. PRINCIPAL NERVES OF THE UPPER LIMB Radial Nerve Median Nerve Ulnar Nerve General Rules and Anomalies 6. LUMBAR PLEXUS AND ITS PRINCIPAL NERVES Iliohypogastric Nerve Ilioinguinal Nerve Genitofemoral Nerve Lateral Femoral Cutaneous Nerve Femoral Nerve Saphenous Nerve Obturator Nerve 7. SACRAL PLEXUS AND ITS PRINCIPAL NERVES Superior and Inferior Gluteal Nerves Sciatic Nerve Tibial Nerve Common Peroneal Nerve Sural Nerve
3
Basics of Electrodiagnosis
4
1 INTRODUCTION Electrodiagnosis, as an extension of the neurologic examination, employs the same anatomic principles of localization, searching for evidence of motor and sensory compromise (Fig. 1-1). Neurophysiologic studies supplement the clinical examination by providing additional precision, detail, and objectivity. They delineate a variety of pathological changes that may otherwise escape detection, and they help examine atrophic, deeply situated, or paretic muscles, which tend to defy clinical evaluation. Specialized techniques provide means to test the neuromuscular junction per se, even though it is an integral part of the motor system. Electrical studies also allow quantitative assess-
DTRWith Sensory Change
ment of reflex amplitude and latencies as well as complex motor phenomena. Individual methods can be used to explore different groups of overlapping neural circuits. Meaningful analysis of electrophysiologic findings, therefore, demands an adequate knowledge of neuroanatomy. In addition, an electromyographer must learn superficial anatomy of skeletal muscles and peripheral nerves as a prerequisite for accurate placement of the recording and stimulating electrodes. The first part of this chapter contains a review of peripheral neuroanatomy important for the performance and interpretation of electrodiagnostic studies. A concise summary of clinically useful information serves as a framework for the rest of the text. Despite the recognized importance of understanding muscle and nerve anatomy,
Upper Motor Neuron
Cerebrum
Stroke
Brainstem
M.S.
Spinal Cord
Tumor
Neuropathy
,DTR
Lower Motor Neuron
Peripheral Nerve
DTR
Upper Motor Neuron
ALS
Weakness
Without Sensory Change
Anterior Horn Cells DTR
Poliomyelitis
Neuromuscular Junction
Myasthenia Gravis Myasthenic Syndrome Myotonia
Anatomical Dx Examples
Muscle
Without Muscle Tenderness
Muscular Dystrophy
With Muscle Tenderness
Polymyositis
Metabolic Myopathy
Figure 1-1. Simplified diagram illustrating the differential diagnosis of weakness, with major divisions into those with or without sensory abnormalities. Patients having sensory symptoms must show involvement of the nervous system rather than of muscle or neuromuscular junction. Disease of upper motor neurons characteristically shows increased stretch reflexes, while disease of the lower motor neurons is characterized by decreased stretch reflexes. Patients without sensory disturbances may still have nervous system disease, especially if the weakness is associated with hyperreflexia, as in amyotrophic lateral sclerosis. Most patients, however, exhibit hyporeflexia, as seen in anterior horn cell lesions, diseases of the neuromuscular junction, or primary muscle disorders (see Chapters 23 through 29).
Anatomic Basis for Localization written descriptions render the subject complicated and rather dry. The use of schematic illustrations in this chapter simplifies the discussion to compensate for this inherent problem. Existing texts provide more detailed accounts in regard to the superficial anatomy of skeletal muscles6,7,10,14 or to the general peripheral neuromuscular anatomy.1-3,9,13
5
2
CRANIAL NERVES
Of the 12 cranial nerves, nine innervate voluntary muscles, as summarized in Table 1-1. The oculomotor, trochlear, and abducens nerves control the movement of the eyes. The trigeminal nerve innervates the muscles of mastication; the facial
Table 1-1 Muscles Innervated by the Cranial Nerves and Cervical Plexus Medulla C-4 Pons C-2 C-3 Mesencephalon Nerve Levator Oculomotor palpebrae Superior rectus Medial rectus
Trochlear Trigeminal Abducens Facial
Inferior rectus Inferior oblique _ Superior oblique Masseter Temporalis Pterygoid _ Lateral rectus Frontalis Orbicularis oculi
Orbicularis oris
Platysma Digastric & stylohyoid muscles Glossopharyngeal Vagus Accessory (cranial root) Hypoglossal Accessory (spinal root) Cervical plexus Phrenic
Laryngeal muscles Laryngeal muscles Laryngeal muscles Tongue Sternocleidomastoid Trapezius Upper Middle Trapezius Lower Diaphragm
Basics of Electrodiagnosis
6
nerve, the muscles of facial expression. The laryngeal muscles receive innervation from the glossopharyngeal and vagal nerves and the cranial root of the accessory nerve. The hypoglossal nerve supplies the tongue. The spinal root of the accessory nerve innervates the sternocleidomastoid and upper portion of the trapezius. Of these, the nerves most commonly tested in an electromyographic laboratory include the facial, trigeminal, and accessory nerves.
Facial Nerve The course of the facial nerve, from the nucleus to the distal trunk, consists of four arbitrarily subdivided segments (Fig. 1-2). The central component, referred to as the intrapontine portion, initially courses posteriorly to loop around the sixth nerve nucleus. Its elongated course makes it vulnerable to various pontine lesions, which
cause a peripheral, rather than central, type of facial palsy. The facial nerve complex exits the brainstem ventrolaterally at the caudal pons. Acoustic neuromas or other cerebellopontine angle masses may compress the nerve in this area. After traversing the subarachnoid space, the facial nerve enters the internal auditory meatus. Here it begins the longest and most complex intraosseous course of any nerve in the body. Within this segment lies the presumed site of lesion in Bell's palsy. Upon exiting from the skull through the stylomastoid foramen, the facial nerve penetrates the superficial and deep lobes of the parotid gland. It then branches with some variation into five distal segments (Fig. 1-3).
Trigeminal Nerve The trigeminal nerve subserves all superficial sensation to the face and buccal and
Figure 1-2. Functional components of the facial nerve and the three major divisions of the trigeminal nerve. The facial nerve (N, VII), consists of the portion at the stylomastoid foramen (A), middle segment distal to the geniculate ganglion (B), and a more proximal segment that includes extrapontine and intrapontine pathways (C). [From Carpenter,3 with permission.]
Anatomic Basis for Localization
7
Figure 1-3. Major types of facial nerve branching and intercommunication with percentage occurrence of each pattern in 350 recordings. [From Anson,1 with permission.)
nasal mucosa. It also supplies the muscles of mastication, which consist of the masseters, temporalis, and pterygoids. The ophthalmic and maxillary divisions of the trigeminal nerve supply sensation to the upper and middle parts of the face, whereas the mandibular division carries the sensory fibers to the lower portion of the face as well as the motor fibers (see Fig. 1-2). The first-order neurons, concerned primarily with tactile sensation, have their cell bodies in the gasserian ganglion. Their proximal branches enter the lateral portion of the pons and ascend to reach the main sensory nucleus. Those fibers subserving pain and temperature sensation also have cell bodies in the gasserian ganglion. Upon entering the pons, their fibers descend to reach the spinal nucleus of the trigeminal nerve. The first-order afferent fibers, subserving proprioception from the muscles of mastication, have their cell bodies in the mesencephalic nucleus. They make monosynaptic connection with the motor nu-
cleus of the trigeminal nerve located in the midpons, medial to the main sensory nucleus. This pathway provides the anatomic substrate for the masseter reflex. The first component of the blink reflex probably follows a disynaptic connection from the main sensory nucleus to the ipsilateral facial nucleus. The pathway for the second component, relayed through polysynaptic connections, include the ipsilateral spinal nucleus and the facial nuclei on both sides (see Fig. 17-1). Accessory Nerve The cranial accessory nerve has the cell bodies in the nucleus ambiguus. The fibers join the vagus nerve and together distribute to the striated muscles of the pharynx and larynx. Thus, despite the traditional name, the cranial portion of the accessory nerve functionally constitutes a part of the vagus nerve. The spinal accessory nerve has its cells of origin in the
Basics of Electrodiagnosi
8
Figure 1-4. Communication between the last four cranial nerves on the right side viewed from the dorsolateral aspect. Note the division of the accessory nerve into the cranial accessory nerve, which joins the vagal nerve, and the spinal accessory nerve, which supplies the trapezius and sternocleidomastoid muscles. [From Williams and Warwick,15 with permission.]
spinal nucleus located in the first five or six cervical segments of the spinal cord (Figs. 1-4 and 1-5). The fibers ascend in the spinal canal to enter the cranial cavity through the foramen magnum and then leave by the jugular foramen to end in the trapezius and the sternocleidomastoid muscles. These two muscles receive additional nerve supply directly from C2 through C4 roots although their motor contribution is minimal.8 The spinal accessory nerve provides the sole motor function, whereas the cervical roots subserve purely proprioceptive sensation (Fig. 1-6). The accessory nucleus consists of several separate portions. Thus, a lesion in the spinal cord may affect it only in part, causing partial paralysis of the muscle groups innervated by this nerve. This central dissociation could mimic a peripheral lesion affecting individual branches.
3
ANTERIOR AND POSTERIOR RAMI
The anterior and posterior roots, each composed of several rootlets, emerge from the spinal cord carrying motor and sensory fibers, respectively (Fig. 1-7). They join to form the spinal nerve that exits from the spinal canal through the respective intervertebral foramina. A small ganglion, containing the cell bodies of sensory fibers, lies on each posterior root in the intervertebral foramina just proximal to its union with the anterior root but distal to the cessation of the dural sleeve. There are 31 spinal nerves on each side: 8 cervical, 12 thoracic, 5 lumbar, 5 sacral, and 1 coccygeal nerve. After passing through the foramina, the spinal nerve branches into two divisions, the anterior and posterior primary rami.
Figure 1-5. The sternocleidomastoid divides the field bounded by the trapezius, mandible, midline of neck, and clavicle into anterior and posterior triangles. The obliquely coursing omohyoid further subdivides the posterior triangle into occipital and subclavian triangles. The contents of the occipital and subclavian triangles include the cervical plexus, spinal accessory nerve, and brachial plexus. The spinal accessory nerve becomes relatively superficial in the middle portion of the stemocleidomastoid along its posterior margin, thus making it accessible to percutaneous stimulation. [From Anson,1 with permission.]
CERVICAL TRIANGLES
Figure 1-6. Anterior rami of the cervical spinal nerves, forming the cervical plexus. Note the phrenic nerve supplying the diaphragm, and the branches from C2, C3, and C4 roots and the accesory nerve, both innervating1 the trapezius. [From Anson, with permission.]
9
10
Basics of Electrodiagnosis cles branch off the spinal nerve just distal to the intervertebral foramen. Hence, denervation found at this level differentiates radiculopathy from more distal lesions of the plexus or peripheral nerve. The reverse does not necessarily hold, especially in early stages of the disease, when the compressing lesions may only irritate the root without causing structural damage. Futhermore, spontaneous discharges appear in the denervated muscles only two to three weeks after nerve injury. Similar to the innervation of the intercostal muscles by the anterior rami, the paraspinal muscles receive supplies from multiple posterior rami with substantial overlap. Therefore, the distribution of abnormalities in the limb muscles rather than the paraspinal muscles determines the level of a radicular lesion.
4 CERVICAL AND BRACHIAL PLEXUSES Figure 1-7. Ventral and dorsal roots forming the spinal nerve, which divides into the anterior and posterior rami. The sensory ganglion of the dorsal root lies within the respective12 intervertebral foramen. [From Ranson and Clark, with permission.]
The posterior rami supply the posterior part of the skin and the paraspinal muscles, which include the rectus capitis posterior, oblique capitis superior and inferior, semispinalis capitis, splenius capitis, longus capitis, and sacrospinalis. These muscles extend the head, neck, trunk, and pelvis, respectively. The anterior rami supply the skin of the anterolateral portion of the trunk and the limbs. They also form the brachial and lumbosacral plexuses, which, in turn, give rise to peripheral nerves in the arms and legs. The anterior rami of the thoracic spinal nerves become 12 pairs of intercostal nerves supplying the intercostal and abdominal muscles. At least two adjoining intercostal nerves supply each segmental level in both the thoracic and abdominal regions. The diagnosis of a root lesion depends on identifying abnormalities confined to a single spinal nerve without affecting adjacent higher or lower levels. The posterior rami that supply the paraspinal mus-
The anterior rami of the upper four cervical nerves, C1 through C4, form the cervical plexus (Fig. 1-6). It innervates the lateral and anterior flexors of the head, which consist of the rectus capitis lateralls, anterior longus capitis, and anterior longus colli. The brachial plexus, formed by the anterior rami of C5 through T1 spinal nerves, supply the muscles of the upper limb. Occasional variations of innervation include the prefixed brachial plexus with main contributions from C4 through C8, and the postfixed brachial plexus derived primarily from C6 through T2. Tables 1-1 and 1-2 present a summary of the anatomic relationship between the nerves derived from cervical and brachial plexuses and the muscles of the shoulder, arm, and hand. Topographic divisions of the brachial plexus include the root, trunk, cord, and peripheral nerve (Fig. 1-8). Two nerves originate directly from the roots before the formation of the trunks: dorsal scapular nerve from C5, innervating levator scapulae and rhomboid, and long thoracic nerve from C5, C6, and C7, supplying serratus anterior. The roots then combine to give rise to three trunks. The union of C5 and
Anatomic Basis for Localization
11
Figure 1-8. Anatomy of the brachial plexus with eventual destination of all root components. The brachial plexus gives rise to (A) dorsal scapular, (B) suprascapular, (C) lateral pectoral, (D) musculocutaneous and its sensory branch, (E) lateral antebrachial cutaneous, (F) median, (G) axillary, (H) radial, (/) ulnar, (J) medial antebrachial cutaneous, (K) thoraccorsal, (I) subscapular, (M) medial pectoral, and (N) long thoracic nerves. In addition, the radial nerve gives off the posterior antebrachial cutaneous nerve (not shown) at the level of the spiral groove. [Modified from Patten,9 with permission.]
C6 forms the upper trunk, and that of C8 and T1, the lower trunk, whereas the C7 root alone continues as the middle trunk. Each of the three trunks gives off anterior and posterior divisions. The posterior cord, formed by the union of all three posterior divisions, gives off the subscapular nerve innervating teres major, thoracodorsal nerve supplying latissimus dorsi and axillary nerve subserving deltoid and teres minor, and continues as the radial nerve. The anterior divisions of the upper and middle trunks form the lateral cord,
which gives rise to the musculocutaneous nerve and its sensory branch, lateral antebrachial cutaneous nerve, and the outer branch of the median nerve. The anterior division of the lower trunk, forming the medial cord, gives off the ulnar nerve, medial antebrachial cutaneous nerve, and the inner branch of the median nerve. The trunks pass through the supraclavicular fossa under the cervical and scalenus muscles, forming the cords just above the clavicle at the level of the first rib. Accompanied by the subclavian artery, the cords
Basics of Electrodiagnosis
12
traverse the space known as the thoracic outlet between the first rib and the clavicle. Consequently, injuries above the clavicle affect the trunks; those below, the cords. A more distal lesion involves the peripheral nerves that emerge from the cords between the clavicle and axilla. Phrenic Nerve The phrenic nerve, one of the most important branches of the cervical plexus, arises from C3 and C4 roots and innervates the ipsilateral hemidiaphragm (Table 1-1). Dorsal Scapular Nerve The dorsal scapular nerve, derived from C4 and C5 roots through the most prox-
imal portion of the upper trunk of the brachial plexus, supplies the rhomboid major and minor and a portion of the levator scapulae, which keeps the scapula attached to the posterior chest wall during arm motion. The rhomboid receives innervation only from a single root (C5) in contrast to the other shoulder girdle muscles supplied by multiple roots. Suprascapular Nerve The suprascapular nerve arises from C5 and C6 roots through the upper trunk of the brachial plexus. It reaches the upper border of the scapula behind the brachial plexus to enter the suprascapular notch, a possible site of entrapment. The nerve supplies the supraspinatus and infraspinatus (Fig. 1-8).
Figure 1-9. Musculocutaneous nerve (A) and ulnar nerve (B), and the muscles they supply. The common sites of lesion include ulnar groove and cubital tunnel (1), Guyon's canal (2), and midpalm (3). [Modified from The Guarantors of Brain: Aids to the Examination of 7the Peripheral Nervous System. ]
Anatomic Basis for Localization
13
Figure 1-10. Axillary nerve (A) and radial nerve (B) with its main terminal branch, the posterior interosseous nerve (C), and the muscles they supply. The nerve injury may occur at the axilla (1), spiral groove (2), or elbow (3), as in the posterior interosseous nerve syndrome. [Modified from The Guarantors of Brain: Aids to the Examination of7 the Peripheral Nervous System. ]
Musculocutaneous Nerve The musculocutaneous nerve originates from the lateral cord of the brachial plexus near the lower border of the pectoralis minor (Fig. 1-9). Its axons, chiefly derived from C5 and C6 roots, reach the biceps, brachialis, and coracobrachialis, with some variations of innervation for the last two muscles. Its terminal sensory branch, called the lateral antebrachial cutaneous nerve, supplies the skin over the lateral aspect of the forearm. Axillary Nerve The axillary nerve, originating from C5 and C6 roots, arises from the posterior cord as the last branch of the brachial plexus. It supplies the deltoid and teres
minor, and a small area of the skin over the lateral aspect of the arm (Fig. 1-10).
5
PRINCIPAL NERVES OF THE UPPER LIMB Radial Nerve
The radial nerve, as a continuation of the posterior cord, derives its axons from C5 through C8, or all the spinal roots contributing to the brachial plexus (Fig. 1-8). The nerve gives off its supply to the three heads of the triceps and the anconeus, which originates from the lateral epicondyle of the humerus as an extension of the medial head . The radial nerve then enters the spiral groove winding around the humerus posteriorly from the medial
14
to the lateral side (Fig. 1-10) giving off a sensory branch, posterior antebrachial cutaneous nerve, which innervates the skin of the lateral arm and the dorsal forearm. As the nerve emerges from the spiral groove, it supplies the brachioradialis, the only flexor innervated by the radial nerve, and, slightly more distally, the extensor carpi radialis longus. Located lateral to the biceps at the level of the lateral epicondyle, it enters the forearm between the brachialis and brachioradialis. At this point, it divides into a muscle branch, the posterior interosseous nerve, and a sensory branch, which surfaces in the distal third of the forearm. The muscle branch innervates the supinator, the abductor pollicis longus, and all the extensor muscles in the forearm: extensor carpi radialis longus and brevis, extensor carpi ulnaris, extensor digitorum communis, extensor digiti minimi, extensor pollicis longus and brevis, and extensor indicis. The sensory fibers, originating from the C6 and C7 roots, pass through the upper and middle trunks and the pos-
Basics of Electrodiagnosis terior cord, branching off the main truck about 10 cm above the wrist as the superficial radial nerve, which supplies the skin over the lateral aspect of the dorsum of the hand. Median Nerve The median nerve arises from the lateral and medial cords of the brachial plexus as a mixed nerve derived from the C6 and T1 roots (Fig. 1-8). It supplies most forearm flexors and the muscles of the thenar eminence. It also subserves sensation to the skin over the lateral aspect of the palm and the dorsal surfaces of the terminal phalanges, along with the volar surfaces of the thumb, the index and middle fingers, and half of the ring finger. The sensory fibers of the index and middle fingers enter the C7 root through the lateral cord and middle trunk, whereas the skin of the thumb receives fibers mainly from C6, with some contribution from the C7 root, through the lateral cord and upper or mid-
Figure 1-11. Median nerve (A) with its branch, the anterior interosseous nerve (B), and the muscles they supply. The nerve may undergo compression at the elbow between the two heads of pronator teres (1), or slightly distally (2), as in the anterior interosseous syndrome, or at the palm (3), as in the carpal tunnel syndrome. [Modified from The Guarantors of Brain: Aids to the Examination of the Peripheral Nervous System.7]
Anatomic Basis for Localization dle trunk. The median nerve innervates no muscles in the upper arm (Fig. 1-11). It enters the forearm between the two heads of the pronator teres, which it supplies along with the flexor carpi radialis, palmaris longus, and flexor digitorum superficialis. A pure muscle branch, called the anterior interosseous nerve, innervates the flexor pollicis longus, pronator quadratus, and flexor digitorum profundus I and II. The main median nerve descends the forearm and, after giving off the palmar sensory branch, which innervate the skin over the thenar eminence, passes through the carpal tunnel between the wrist and palm. It supplies lumbricals I and II after giving rise to the recurrent thenar nerve at the distal edge of the carpal ligaments. This muscle branch to the thenar eminence innervates the abductor pollicis brevis, the lateral half of the flexor pollicis brevis, and the opponens pollicis.
15
hamate to reach the lateral aspect of the hand, where it reaches the adductor pollicis and medial half of the flexor pollicis brevis. Along its course from hypothenar to thenar eminence, the deep branch also innervates the three volar and four dorsal interossei, and lumbricals III and IV. General Rules and Anomalies
Table 1-2 summarizes the pattern of nerve supply in the upper limbs. One cannot memorize the exact innervation for all the individual muscles, but learning certain rules helps broadly categorize muscles. The radial nerve innervates the brachioradialis, triceps, and with its main terminal branch, the posterior interosseous nerve, all the extensors in the forearm, but none of the intrinsic hand muscles. The radial nerve innervates only the extensors, with the exception of brachioradialis, an elbow flexor in the neutral or half-pronated position. The nerve subserves all the extensors of the upper limb Ulnar Nerve except for the four lumbricals, which, The ulnar nerve, as a continuation of the supplied by median and ulnar nerves, exmedial cord of the brachial plexus, derives tend the digits at the interpharangeal its fibers from the C8 and T1 roots (Fig. joints. The median nerve supplies most 1-8). It lies in close proximity to the me- flexors in the forearm, in addition to the dian nerve and brachial artery at the ax- intrinsic hand muscles of the thenar emilla. In this position, the ulnar nerve inence and lumbricals I and II. The antepasses between the biceps and triceps, rior interosseus nerve branches off the and then deviates posteriorly at the mid- median nerve trunk in the forearm to inportion of the upper arm and becomes su- nervate the flexor digitorum profundus I perficial behind the medial epicondyle and II, flexor pollicis longus, and prona(Fig. 1-9). After entering the forearm, it tor quadratus. With the exception of the supplies the flexor carpi ulnaris and flexor flexor carpi ulnaris and the flexor digitodigitorum profundus III and IV, and gives rum profundus III and IV, the ulnar nerve rise to the dorsal cutaneous branch of the supplies only intrinsic hand muscles, inulnar nerve, which innervates the skin cluding all the interossei. over the medial aspect of the dorsum of The most common anomaly of innervathe hand. It then passes along the medial tion in the upper limb results from the aspect of the wrist to enter the hand, presence of a communicating branch from where it gives off two branches (see Chap- the median to the ulnar nerve in the foreter 26-6). The superficial branch supplies arm. The fibers involved in this crossover, the palmaris brevis and the skin distally called the Martin-Gruber anastomosis, from the wrist over the medial aspect of usually supply ordinarily ulnar-innervated the hand, including the hypothenar emi- intrinsic hand muscles. Thus, the anomnence, the fifth digit, and half of the fourth alous fibers form a portion of the ulnar digit. The deep muscle branch first in- nerve that, instead of branching off from nervates the hypothenar muscles, that is, the medial cord of the brachial plexus, abductor, opponens, and flexor digiti min- takes an aberrant route distally along with imi. It then deviates laterally around the the median nerve and then reunites with
Table 1-2 Innervation of Shoulder Girdle and Upper Limb Muscles Nerve
C-4
C-5
C-6
Dorsal scapular
Levator scapulae Rhomboideus major & minor
Supra scapular
Supraspinatus Infraspinatus
Axillary
Teres minor . Deltoid Anterior Middle Posterior
Subscapular
C-7
Teres major .
Musculocutaneous
Brachialis Biceps brachii Coracobrachialis Serratus.
Long thoracic
anterior
Pectoralis major . (clavicular part)
Anterior thoracic
Pectoralis major . (sternocostal part) Pectoralis minor . Thoracodorsal
Latissimus dorsi
Radial nerve
Brachioradialis Extensor carpi radialis longus & brevis Triceps—long, lateral & medial heads Anconeus Posterior interosseous nerve
Supinator .
Extensor carpi Extensor digitorum Extensor digiti minimi Abductor pollicis longus. Extensor pollicis longus Extensor pollicis brevis . Extensor indicis
C-8
T-l
Pronator teres Flexor carpi radialis
Median nerve
Palmaris longus. Flexor digitorum sublimis Abductor pollicis brevis Flexor pollicis brevis (superficial head) Lumbrlcals I&II Opponens pollicis Anterior Interosseous nerve
Ulnar nerve
Flexor digitorum profundus I&II Flexor pollicis longus — Pronator quadratus Flexor dlgltorum profundus III & IV _ Flexor carpi ulnarls . Adductor pollicls . Flexor pollicls brevls (deep head) . Abductor digitl minimi — Opponens dlgltl minimi — . Flexor digitl minimi — Volar interossel . Dorsal interossel Lumbrlcals III & IV
Table 1-3 Innervation of Pelvic Girdle and Lower Limb Muscles Nerve
L-2
Iliopsoas _ Pectineus Sartorius Vastus
Femoral nerve
Obturator nerve
L-3
L-5
8-1
8-2
Sartorius Intermedius Rectus femoris Vastus lateralis . Vastus medialis
Gracilis Adductor longus & brevis Adductormagnus Obturator
Superior gluteal nerve
L-4
externus medius GluteusGluteus minimus Tensor fasciae latae
Inferior gluteal nerve
Gluteus maximus
Sacral plexus
Obturator internus Superior & inferior gemelli Quadratus femoris Piriformis
Sciatic nerve trunk Peroneal division Tibial division
Biceps femoris short head Semi Semi
tendinosus membranosus
Biceps femoris long head. Common peroneal nerve Deep peroneal nerve
. Tibialis anterior Extensor digitorum longus Extensor digitorum brevis. Peroneus tertius Extensor hallucis longus .
Superficial peroneal nerve Tibial nerve
Peroneus longus Peroneus brevis Tibialis posterior Popliteus Flexor digitorum longus Flexor longus hallucis Gastrocnemius Medial head. Lateral head Soleus
Medial plantar nerve
Flexor digitorum brevis Flexor hallucis brevis Abductor hallucis Lumbrical I
Lateral plantar nerve
Abductor digiti minimi Adductor hallucis Flexor digiti minimi Interossei Quadratus plantae Lumbricals II, III, IV
Basics of Electrodiagnosis
20
the ulnar nerve proper in the distal forearm. Other anomalies reported in the literature include communication from the ulnar to the median nerve in the forearm, and all median or all ulnar hands, in which one or the other nerve supplies all the intrinsic hand muscles. These extremely rare patterns stand in contrast to the high incidence of the median-to-ulnar anastomosis. Failure to recognize an anomaly leads to misinterpretation in clinical electrophysiology as a common source of error (see Chapter 7-4).
6
LUMBAR PLEXUS AND ITS PRINCIPAL NERVES
The spinal cord ends at the level of the L1 to L2 intervertebral space as the preconus, which consists of the L5 and S1 cord segments, and the conus medullaris, which contains the S2 through S5 levels. The fibers of the cauda equina, formed by the lumbar and sacral roots, assume a downward direction from the conus toward their respective exit foramina. The fibrous filum terminale interna extends from the lowermost end of the spinal cord to the bottom of the dural sac at the level of the S2 vertebra. Table 1-3 summarizes the nerves derived from the lumbar plexus and the muscles they innervate. The anterior rami of the first three lumbar spinal nerves, originating from the L1, L2, and L3, and part of L4 roots, unite to form the lumbar plexus within the psoas major muscle (Figs. 1-12 through 1-14). The iliohypogastric and ilioinguinal nerves arise from the L1 root and supply the skin of the hypogastric region and medial thigh, respectively. The genitofemoral nerve, derived from the L1 and L2 roots, innervates the cremasteric muscle and the skin of the scrotum or labia major. The lateral femoral cutaneous nerve originates from the L2 and L3 roots. It leaves the psoas muscle laterally to supply the lateral and anterior thigh. The anterior divisions of the L2 through L4 anterior rami join to form the obturator nerve, which exits the psoas muscle medially to innervate the adductor muscles of the thigh. The posterior divisions of the same rami give rise to the femoral
Figure 1-12. Anterior rami of the lumbar spinal nerve forming the lumbar plexus, with the major nerves derived from this plexus. The shaded portion indicates the dorsal divisions. [From Anson,1 with permission.]
nerve, which leaves the psoas muscle laterally. It then descends under the iliacus fascia to reach the femoral triangle beneath the inguinal ligament. Though primarily a muscle nerve, it also gives off sensory branches—the intermediate and medial cutaneous nerves, and the saphenous nerve. Iliohypogastric Nerve The iliohypogastric nerve originates from the L1 root and supplies the skin of the upper buttock and hypogastric region. Ilioinguinal Nerve The ilioinguinal nerve, arising from the L1 and L2 roots, supplies the skin over the upper and medial part of the thigh, the root of the penis, and the upper part of the scrotum or labia major. It also innervates the transversalis and internal oblique muscles. The nerve follows the basic pattern of an intercostal nerve, winding around
Anatomic Basis for Localization
21
Figure 1-13. Lumbosacral plexus and the courses of the femoral, obturator, and sciatic nerves. [From Anson,1 with permission.]
the inner side of the trunk to the medial anterior iliac spine. Genitofemoral Nerve The genitofemoral nerve, arising from the L1 and L2 roots, branches into lumboinguinal and external spermatic nerves. The lumboinguinal nerve supplies the skin over the femoral triangle. The external spermatic nerve innervates the cremasteric muscle and the skin of the inner aspect of the upper thigh, scrotum, or labium.
Lateral Femoral Cutaneous Nerve The lateral femoral cutaneous nerve, the first sensory branch of the lumbar plexus, receives fibers from the L2 and L3 roots. It emerges from the lateral border of the psoas major muscle and runs forward, coursing along the brim of the pelvis to the lateral end of the inguinal ligament. The nerve reaches the upper thigh after passing through a tunnel formed by the lateral attachment of the inguinal ligament and the anterior superior iliac spine. About 12 cm below its exit from the tun-
Basics of Electrodiagnosis
22
femoris, vastus lateralis, vastus intermedius, and vastus medialis. Of the muscles innervated by this nerve, the iliopsoas flexes the hip at the thigh, the quadriceps femoris extends the leg at the knee, the sartorius flexes the leg and the thigh, and the pectineus flexes the thigh. Saphenous Nerve
Figure 1-14. Anterior rami of the lumbosacral spinal nerve forming the sacral plexus with the major nerves derived from this plexus. The shaded portion indicates the dorsal divisions. [From Anson,1 with permission.]
nel, the nerve gives off an anterior branch, which supplies the skin over the lateral and anterior surface of the thigh, and a posterior branch, which innervates the lateral and posterior portion of the thigh. Femoral Nerve The femoral nerve, formed near the vertebral canal, arises from the anterior rami of the L2 through L4 roots (Fig. 1-15). The nerve reaches the front of the leg passing along the lateral edge of the psoas muscle, which it supplies together with the iliacus. It then exits the pelvis under the inguinal ligament just lateral to the femoral artery and vein. Its sensory branches supply the skin of the anterior thigh and medial aspect of the calf. The muscle branch innervates the pectineus and the sartorius, as well as the quadriceps femoris, which consists of the rectus
The saphenous nerve, the largest and longest sensory branch of the femoral nerve, receives maximum 11innervation through the L3 and L4 roots and supplies the skin over the medial aspect of the thigh, leg, and foot. It accompanies the femoral artery in the femoral triangle, then descends medially under the sartorius muscle. The nerve gives off the infrapatellar branch at the lower thigh, which supplies the medial aspect of the knee. The main terminal branch descends along the medial aspect of the leg, accompanied by the long saphenous vein. It passes just anterior to the medial malleolus, supplying the medial side of the foot. Obturator Nerve The obturator nerve arises from the anterior divisions of the L2 through L4 roots (Fig. 1-15). Formed within the psoas muscle, it enters the pelvis immediately anterior to the sacroiliac joint. As it passes through the obturator canal, the obturator nerve gives off an anterior branch, which supplies the adductor longus and brevis and the gracilis, and a posterior branch, which innervates the obturator externus and half of the adductor magnus muscle. The sensory fibers supply the skin of the upper thigh over the medial aspect and send anastomoses to the saphenous nerve.
7 SACRAL PLEXUS AND ITS PRINCIPAL NERVES
The sacral plexus arises from the L5, S1, and S2 roots in front of the sacroiliac joint (Figs. 1-13 and 1-14). Designation as the
Anatomic Basis for Localization
23
Figure 1-15. Femoral nerve (A), obturator nerve (B), and common peroneal nerve (C) branching into superficial (D) and deep peroneal nerve (E) and the muscles they supply. The compression of the peroneal nerve commonly occurs at the flbular head (1). [Modified from The Guarantors of Brain: Aids to the Examination of the Peripheral Nervous System.7]
lumbosacral plexus implies an interconnection between the sacral and lumbar plexi. Common anomalous derivations include a prefixed pattern with a major contribution of the L4 root to the sacral plexus or a postfixed form with the L5 root supplying mainly the lumbar plexus. The sacral plexus gives rise to the superior gluteal nerve, derived from the L4, L5, and S1 roots, and the inferior gluteal nerve, which arises from the L5, S1, and S2 roots. The sciatic nerve, the largest nerve in the body, arises from the L4 through S2 roots. After giving off branches to the hamstring muscles, it divides into the tibial and common peroneal nerves.
Table 1-3 summarizes the nerves derived from the sacral plexus, and the muscles that they innervate. Superior and Inferior Gluteal Nerves The superior gluteal nerve, derived from the L4 through S1 roots, innervates the gluteus medius and minimus, and the tensor fascia lata, which together abduct and rotate the thigh internally. The inferior gluteal nerve, arising from the L5 through S2 roots, innervates the gluteus maximus, which extends, abducts, and externally rotates the thigh.
Basics of Electrodiagnosis
24
Sciatic Nerve The union of all of the L4 to S2 roots gives rise to the sciatic nerve, which leaves the pelvis through the greater sciatic foramen (Fig. 1-16). The nerve consists of a peroneal portion derived from the posterior division of the anterior rami, and a tibial portion composed of the anterior divisions. The peroneal and tibial components eventually separate in the lower third of the thigh to form the common peroneal and tibial nerves or, in the older terminology, anterior and posterior tibial nerves. In the posterior aspect of the thigh, the tibial component of the sciatic trunk gives off a series of short branches to innervate the bulk of the hamstring muscles, which consist of the long head of the biceps femoris, semitendinosus, and semimembranosus. The peroneal component supplies the short head of the biceps femoris, which, if
affected in patients with foot drop, implies a lesion above the knee. The adductor magnus, primarily supplied by the obturator nerve, also receives partial innervation from the sciatic trunk. Tibial Nerve The tibial nerve arises as an extension of the medial popliteal nerve that bifurcates from the sciatic nerve in the popliteal fossa (Fig. 1-16). After giving off branches to the medial and lateral heads of the gastrocnemius and soleus, it supplies the tibialis posterior, flexor digitorum longus, and flexor hallucis longus in the leg. The nerve enters the foot, passing through the space between the medial malleolus and the flexor retinaculum. Here it splits into medial and lateral plantar nerves after giving off a small calcaneal nerve. This bi-
Figure 1-16. Superior gluteal nerve (A), inferior gluteal nerve (B), and sciatic nerve trunk (C), and the muscles they supply. The sciatic nerve bifurcates to form the common peroneal nerve (D) and the tibial nerve (E). The tibial nerve in turn gives rise to the medial (F) and lateral plantar nerve (G). The compression of the tibial nerve may occur at the medial malleolus in the tarsal tunnel (1). [Modified from The Guarantors of Brain: Aids to the Examination of the7 Peripheral Nervous System. ]
Anatomic Basis for Localization furcation occurs within one centimeter of the malleolar-calcaneal axis in 90 percent of feet.4 The medial plantar artery, which accompanies the medial plantar nerve, serves as the landmark to locate the nerve just below the medial malleolus. The muscle branches innervate the abductor hallucis, flexor digitorum brevis, and flexor hallucis brevis. The sensory fibers of the medial plantar nerve supply the medial anterior two thirds of the sole and the plantar skin of the first three toes and part of the fourth toe. The lateral plantar nerve winds around the heel to the lateral side of the sole to innervate the abductor digiti minimi, flexor digiti minimi, abductor hallucis, and interossei. It supplies the skin over the fifth toe, the lateral half of the fourth toe, and the lateral aspect of the sole.
Common Peroneal Nerve The common peroneal nerve arises as an extension of the lateral popliteal nerve, which branches off laterally from the sciatic trunk in the politeal fossa (Fig. 1-15). It consists of fibers derived from the L4, L5, and S1 roots. Immediately after its origin, the nerve becomes superficial as it winds around the head of the fibula laterally. After entering the leg at this position, it gives off a small recurrent nerve that supplies sensation to the patella and then bifurcates into the superficial and deep peroneal nerves. The superficial peroneal nerve, also known as the musculocutaneous nerve, supplies the peroneus longus and brevis, which plantar-flex and evert the foot. After descending between the peroneal muscles, it divides into medial and intermediate dorsal cutaneous nerves. These sensory branches pass anterior to the extensor retinaculum and supply the anterolateral aspect of the lower half of the leg and dorsum of the foot and toes. The deep peroneal nerve innervates the muscles that dorsiflex and evert the foot. These muscles include the tibialis anterior, extensor digitorum longus, extensor hallucis longus, peroneus tertius, and extensor digitorum brevis. An anomalous
25
communicating branch called the accessory deep peroneal nerve may arise from the superficial peroneal nerve at the knee to innervate the lateral portion of the extensor digitorum brevis (see Chapter 7-4). The deep peroneal nerve also supplies the skin over a small, wedge-shaped area between the first and second toes.
Sural Nerve The sural nerve originates from the union of the medial sural cutaneous branch of the tibial nerve and the sural communicating branch of the common peroneal nerve. It arises below the popliteal space, descends between the two bellies of the gastrocnemius, winds behind the lateral malleolus, and reaches the dorsum of the fifth toe. It receives maximum innervation from the S1 root, with the remainder coming from the L5 or S2 root,11 and supplies the skin over the posterolateral aspect of the distal leg and lateral aspect of the foot. As one of the few readily accessible sensory nerves in the lower limbs, the sural nerve offers an ideal site for biopsy, especially because its removal induces only minimal sensory changes. A fascicular biopsy of the sural nerve allows in vitro recording of nerve action potentials (see Chapter 4-4). Therefore, in vivo studies of the sural nerve before such a procedure provide an interesting opportunity to correlate the data directly with in vitro conduction characteristics and the histologic findings of the biopsy speciman.5
REFERENCES 1. Anson BJ: An Atlas of Human Anatomy, ed 2. WB Saunders, Philadelphia, 1963. 2. Berry MM, Standing SM, Banister LH: Nervous System. In Williams (ed): Gray's Anatomy, ed 38. Churchill Livingstone, New York, 1995. 3. Carpenter MB: Human Neuroanatomy, ed 7. Williams & Wilkins, Baltimore, 1976. 4. Dellong AL, Mackinnon SE: Tibial nerve branching in the tarsal tunnel. Arch Neurol 41:645-646, 1984. 5. Dyck PJ, Lambert EH, Nichols PC: Quantitative measurement of sensation related to compound action potential and number and size of myelinated fibers of sural nerve in health, Friedreich's
Basics of Electrodiagnosis
26
6. 7. 8.
9.
ataxia, hereditary sensory neuropathy and tabes dorsalis. In Remond A (ed): Handbook of Electroencephalography and Clinical Neurophysiology, Vol 9. Elsevier, Amsterdam, 1971, pp 83-118. Goodgold J: Anatomical Correlates of Clinical Electromyography. Williams & Wilkins, Baltimore, 1974. The Guarantors of Brain: Aids to the Examination of the Peripheral Nervous System. WB Saunders, Philadelphia, 1987. Nori S, Soo KC, Green RF, Strong EW, Mee SM: Utilization of intraoperative electroneurography to understand the innervation of the trapezius muscle. Muscle Nerve 20:279-285, 1997. Patten J: Neurological Differential Diagnosis, ed 2. Springer-Verlag, New York, 1995, p. 297.
10. Perotto A: Anatomical Guide for the Electromyographer: The Limbs and Trunk, ed 3, Charles C Thomas, Springfield, IL, 1996. 11. Phillips II LH, Park TS: Electrophysiological mapping of the segmental innervation of the saphenous and sural nerves. Muscle Nerve 16: 827-831, 1993. 12. Ranson SW, Clark SL: The Anatomy of the Nervous System: Its Development and Function, ed 10. WB Saunders, Philadelphia, 1959. 13. Sunderland S: Nerves and Nerve Injuries, ed 2. Churchill Livingstone, New York, 1978. 14. Warfel JH: The Head, Neck and Trunk, ed 6. Lea & Febiger, Philadelphia, 1993. 15. Williams PL, Warwick R: Gray's Anatomy, ed 36 (British). Churchill Livingstone, Edinburgh, 1980.
Chapter 2 ELECTRICAL PROPERTIES OF NERVE AND MUSCLE
1. INTRODUCTION 2. TRANSMEMBRANE POTENTIAL Ionic Concentration of Cells Nernst Equation Sodium-Potassium Pump Goldman-Hodgkin-Katz Equation 3. GENERATION OF ACTION POTENTIAL All-or-None Response Local Current Afterpotentials 4. VOLUME CONDUCTION AND WAVEFORM Diphasic Recording of Action Potential Effect of Volume Conduction Analysis of Triphasic Waveform Near-Field and Far-Field Potentials
1
INTRODUCTION
The nervous system conveys information by means of action potentials, which, under physiological conditions, originate in the cell body or axon terminal and propagate along the nerve fibers. An electrophysiologic study takes advantage of such neural impulses activated artificially by electrical stimuli applied at certain points of the nerve. Motor conduction studies depend on recording a muscle action potential elicited by stimulation of the mixed nerve, whereas sensory studies use either mixed or sensory nerve action potentials. Electromyography permits analysis of electrical properties in the skeletal muscle at rest and during voluntary contraction. Thus, proper interpretations of electrodi-
agnostic data in the clinical domain requires an understanding of the electrical properties of nerve and muscle. Despite different anatomic substrates subserving electrical impulses, the same basic membrane physiology applies to both nerve and muscle. Excitability of the tissues reflects the magnitude of the transmembrane potential in a steady state. When stimulated electrically or by other means, the cell membrane undergoes an intensity-dependent depolarization. If the change reaches a critical level, called threshold, it generates an action potential, which then propagates across the membrane. In contrast to intracellular recording in animal experiments, clinical electrodiagnostic procedures analyze extracellular potentials by surface or needle electrodes. Here the interstitial tissues 27
Basics of Electrodiagnosis
28
act as a volume conductor, where the position of the recording electrode relative to the generator source dictates the waveform of the recorded potentials.
2
TRANSMEMBRANE POTENTIAL
Understanding membrane physiology at the cellular level forms the basis for electrophysiologic examination in the clinical domain. This section deals with the ionic concentration of cell plasma and its role in maintaining transmembrane potentials. The next sections summarize the basic physiology of the propagating action potential recorded through volume conductors. The following comments, intended merely as a background for forthcoming discussion, covers only the fundamental principles relevant to clinical electrophysiology. Subsequent sections, such as Chapters 7, 8, and 20 further elaborate on these points. Interested readers can find a more detailed account of basic cell physiology in established texts.4,7,26,27,29,33,34,44,47,62,63
Ionic Concentration of Cells The muscle membrane constitutes the boundary between intracellular fluid in cell cytoplasm and extracellular interstitial fluids. Both contain approximately equal numbers of ions dissolved in water but differ in two major aspects. First, an electrical potential exists across the cell memTable 2-1 Compositions of Extracellular and Intracellular Fluids of Mammalian Muscle ExtraIntraEquilibrium cellular cellular Potential (mmol/1) (mmol/1) (mV) Cations + Na+ K
145 4
cl
120
Others Anions HCO 3 Others Potential
5
27 7 0 mV
12 155
—
4 8 155 -90 mV
From Patton,48 with permission.
66 -97
brane, with a relative negativity inside the cell as compared to outside. This steady transmembrane potential measures approximately 52 -90 mV in human skeletal muscle cells, but it varies from one tissue to another, ranging from -20 mV to -100 mV. Second, intracellular fluid has a much higher concentration of potassium (K+) and lower concentration of sodium (Na+) and chloride (Cl~) ions relative to the extracellular fluid (Table 2-1).
Nernst Equation In the steady state, the influx of an ion precisely counters the efflux, maintaining an equilibrium. Thus, various factors that determine the direction and the rate of the ionic flow together must exert a balanced force. Measuring the ionic concentration, therefore, provides a calculation of the equilibrium potential—that is, the transmembrane potential theoretically required to establish such a balance (Fig. 2-1). In the case of potassium, for example, the ionic difference tends to push potassium from inside to outside the cell, reflecting the higher concentration inside. This force per mole of potassium, or its chemical work (Wc), increases in proportion to the logarithm of the ratio between internal and external concentration of the cation, (K+)i and (K+)0, according to the equation where R represents the universal gas constant, T, the absolute temperature, i, inside, o, outside, and log(e), natural logarithm. The energy required to counter this force must come from the negative equilibrium potential (Ek) pulling the positively charged potassium from outside to inside the cell. This force per mole of potassium, or the electrical work (We), increases in proportion to the transmembrane voltage Ek, according to the equation
—
-90 -32 —
where F represents the number of coulombs per mole of charge and Zk the valence of the ion. In the steady state, the sum of these two energies, Wc and We, must equal zero, as
Electrical Properties of Nerve and Muscle
29
and Table 2-1 shows the values of Ek (-97 mV), Ena (+66 mV), and Ecl (-90 mV) determined on the basis of their ionic concentrations. These compare with the actual transmembrane potential (-90 mV) in the example under consideration. Thus, ionic concentration and transmembrane potential alone can maintain chloride ions in perfect balance. To keep potassium and sodium in equilibrium at transmembrane potentials of -90 mV, therefore, other factors must exert a substantial influence on ionic movements. These include selective permeability of the cell membrane to certain ions and the energydependent sodium-potassium pump.
Sodium-Potassium Pump
Figure 2-1. Simplified scheme of active and passive fluxes of potassium (K+), sodium (Na+), and chloride (Cl~) in the steady state with driving force on each ion shown by vectors. For potassium, the efflux along the concentration gradient equals the influx caused by the electrical force plus the active influx by the sodiumpotassium pump. For sodium, the electrical and chemical gradient produces only a small influx because of membrane resistence. The sum of the two equals the active efflux by the sodium-potassium pump. For chloride, the concentration gradient almost exactly counters the electrical force. The ratio of sodium and potassium exchange by a common electrogenic pump averages 3:2, although this diagram illustrates a neutral pump with a ratio of 1:1.
they represent forces with opposite vectors. Therefore, Thus, the Nernst equation provides the theoretical potassium equilibrium potential Ek as follows The same equation applies to calculate the sodium and chloride equilibrium potentials, Ena and Ecl, as follows:
In the case of potassium, an additional factor, the active transport of potassium by an energy-dependent pump, explains the small discrepancy between Ek(-97 mV) and Em(-90 mV). Here, the forces pulling potassium from outside to inside the cell consist of the potential difference (-90 mV) and the active potassium transport (approximately equivalent to —7 mV). Together they counter almost exactly the concentration gradient pushing potassium from inside to outside the cell. In the case of sodium, both the concentration gradient and potential difference (-90 mV) pull the ion from outside to inside the cell. Nonetheless, this cation remains in equilibrium because of its impermeability through a mechanical barrier imposed by the structure of the cell membrane. Active transport of sodium from inside to outside counters the small amount of sodium that does leak inwards. This energy-dependent process, known as the potassium-sodium pump, transports sodium outward in exchange for the inward movement of potassium. Although Figure 2-1 depicts a neutral pump that exchanges one sodium ion for every potassium ion actively transported inward, the actual ratio of sodium and potassium ex-
Basics of Electrodiagnosis
30
change averages 3 to 2 in most tissues.47 Such an unbalanced arrangement, called an electrogenic potassium-sodium pump, directly contributes to the membrane potential, but only minimally compared with changes in membrane permeability. Goldman-Hodgkin-Katz Equation The Nernst equation closely predicts membrane potential for highly diffusible chloride and potassium ions. It does not fit well with much less permeable sodium ions, because it ignores relative membrane permeability. The addition of this factor leads to the more comprehensive Goldman-Hodgkin-Katz formula, which incorporates the concentration gradients and membrane permeabilities of all ions. Em = (RT/F) loge PNa (Na + 0 ) + PK(K+0)
+ P cl (Cl - i ) PNa (Na+i) + PK(K+i) + Pcl(Cl-0) where PNa, Pk,, and PCl represent permeabilities of the respective ions. According to this equation, the concentration gradient of the most permeable ions dictates the transmembrane potentials. In the resting membrane with very high PK relative to negligible PNa, the GoldmanHodgkin-Katz equation would approximate the Nernst equation using the potassium concentration gradients. The transmembrane potentials calculated using either equation range from -80 to -90 mV. Conversely, the Goldman-Hodgkin-Katz potential would nearly equal the Nernst potential for sodium, with negligible Pk relative to high PNa. In this situation, the calculated membrane potentials range from +50 to +70 mV. This reversal of polarity in fact characterizes the generation of an action potential as outlined below.
3
GENERATION OF ACTION POTENTIAL
Generation of an action potential consists of two phases: subthreshold and threshold. Subthreshold activation produces a
graded response or a self-limiting local potential in transmembrane potential that diminishes with distance. If, on the other hand, the membrane potential reaches a critical level with about 15-25 mV of depolarization, from -90 mV to -65 to -75 mV in the case of human muscle cell,52 the action potential develops in an allor-none fashion; that is, the same maximal response occurs through a complex energydependent process regardless of the kind or magnitude of the stimulus, as described below (Fig. 2-2). All-or-None Response In the living cell, a voltage-sensitive molecular structure regulates the conductance of sodium and potassium ions across the membrane. One set of channels controls the movement of sodium ions and another set controls potassium ions, depending on the transmembrane potential. When open, they provide adequate pathways for that specific ion to cross the membrane. In the resting stage, potassium ions move freely, through potassium channels kept open at this transmembrane potential, whereas sodium ions remain static. Depolarization to a critical level opens the sodium channels, giving rise to a 500-fold increase in sodium permeability. An externally applied current for nerve stimulation, for example, will depolarize the nerve under the cathode, or negative pole, inducing negativity outside the axon and thus making the inside relatively more positive. When this positivity, or depolarization, reaches a critical level, voltagedependent sodium channels open, initiating the sequence of events leading to nerve excitation. In short, nerve stimulation accomplishes its objective by opening sodium channels. This intrinsic property of nerve and muscle underlies the all-or-none response: regardless of the nature of the stimulus, the same action potential occurs as long as depolarization reaches the critical level. The increased conductance or permeability allows sodium ions to enter the cell seeking a new steady state. Sodium entry further depolarizes the cell, which in turn accelerates inward movement of this ion. Because
Electrical Properties of Nerve and Muscle
31
Figure 2-2. Schematic diagram of graded responses after subthreshold stimuli and generation of action potentials after suprathreshold stimuli. The experimental arrangement shows intracellular stimulation (I) and recording electrodes (E) on top (A) and polarity, strength, and duration of a constant current on bottom (B): Hyperpolarizing (1) and subthreshold depolarizing current (2) induces a nonpropagating local response. Current of just threshold strength will produce either local change (3a) or an action potential (3b). Suprathreshold stimulation (4) also generates an action potential, but with a more rapid time course of depolarization. [From Woodbury,62 with permission.]
of this regenerative sequence, an action potential develops explosively to its full size. The dramatic change in sodium permeability during the course of the action potential results in a reversal of membrane potential from -80 or -90 mV to +20 or +30 mV. In other words, a switch from the potassium to the sodium equilibrium constitutes generation of an action potential. This shift of intracellularly recorded membrane potential from negative to positive gives rise to negative spike when recorded extracellularly according to the convention in clinical electrophysiology. In the depolarized membrane, permeability to potassium ions also increases as a result of a molecular change, but only after a delay of about one millisecond. At about the same time, the increased permeability to sodium falls again to near the resting value with closure or inactivation of sodium channels. Inactivated sodium
channels fail to open for a few milliseconds even with depolarization above the critical level, giving rise to the refractory period (see Chapter 8-2). This inactivation of sodium conductance, together with increased potassium permeability, results in rapid recovery of the cell membrane from depolarization. After the potential falls precipitously toward the resting level, a transient increase in potassium conductance hyperpolarizes the membrane, which then returns slowly to the resting value, completing the cycle of repolarization. The amount of sodium influx and potassium efflux during the course of an action potential alters the concentration gradients of these two ions very little. Although repolarization primarily results from a delayed increase in potassium conductance in squid giant axon,28 this may not apply to mammalian peripheral or central myelinated axons.61 Voltage clamp ex-
Basics of Electrodiagnosis
32
periments indicate that sodium channels abound at the nodes of Ranvier, where potassium conductance may be minimal or absent in the intact 9mammalian peripheral myelinated axons8'40 '4950 or mammalian dorsal column axons. ' In contrast, potassium channels are distributed all along the internodes, although paranodal regions also contain some sodium conductance. Theoretically, the availability of potassium conductance facilitates repolarization, but at a cost of prolonging the refractory period. In mammalian fibers, the absence of potassium channels at the node of Ranvier, combined with the fast inactivation of sodium conductance, allows an increased rate of firing (see Chapter 8-2). Local Current An action potential initiated at one point on the cell membrane renders the inside
of the cell positive in that local region, reflecting elevated sodium conductance. Intracellular current then flows from the active area to the adjacent, negatively charged, inactive region. A return flow through the extracellular fluid from the inactive 10to active region completes the current. In other words, a current enters the cell at the site of depolarization (sink) and passes out to adjacent regions of the polarized membrane (source) (see Fig. 4-3). This local current tends to depolarize the inactive regions on both sides of the active area. When depolarization reaches the threshold, an action potential occurs, giving rise to a new local current further distally and proximally. Thus, an impulse, once generated in the nerve axon, propagates in both directions from the original site of depolarization, initiating orthodromic as well as antidromic volleys of the action potential (see Chapter 4-3).
Figure 2-3. Diagrammatic representation of an action potential in A fibers of the cat, with the spike and negative and positive afterpotentials drawn in their correct relative size and true relationships. [From Gasser,21 with permission.]
Electrical Properties of Nerve and Muscle
33
Afterpotentials In an extracellular recording, an action potential consists of an initial negative spike of about one millisecond duration representing the intracellular positive spike of depolarization, and two subsequent afterpotentials, negative, or depolarizing, and positive, or hyperpolarizing (Fig. 2-3). The negative afterpotential, an externally negative deflection grafted onto the declining phase of the negative spike, corresponds to a super-normal period of excitability. This phase results from sustained internodal positivity and the extracellular accumulation of potassium ions associated with the generation of an action potential. The subsequent positive afterpotential, a prolonged externally positive deflection signals a subnormal period of excitability. This phase reflects the elevated potassium conductance at the end of the action potential and an increased rate of the potassium-sodium pump to counter the internal sodium concentration (see Chapter 8-2).
4
VOLUME CONDUCTION AND WAVEFORM
Diphasic Recording of Action Potential A pair of electrodes placed on the surface of a nerve or muscle at rest register no difference of potential between them. If, in the tissue activated at one end, the propagating action potential reaches the nearest electrode (G1), then G1 becomes negative relative to the distant electrode (G2). This results in an upward deflection of the tracing according to the convention of clinical electrophysiology (although one could also set the oscilloscope to display negativity of G1 as a downward deflection as some investigators do against the general trend.) With further passage of the action potential, the trace returns to the baseline at the point where the depolarized zone affects G1 and G2 equally. When the action potential moves further away from G1 and toward G2, G2 becomes negative relative to G1, or G1 becomes posi-
Figure 2-4. Diphasic (top) and monophasic recording (bottom) of an action potential represented by the shaded area. As the impulse propagates from left to right in the top series, the two electrodes see no potential difference in a, c, and e. Relative to the reference electrode ( G2), the active electrode (G1) becomes negative in b, and positive in d, resulting in a diphasic potential. In the bottom tracing, the darkened area on the right indicates a killed end with permanent depolarization, making G1 positive relative to G2 in a', c', and d'. In b', G1 and G2 have no potential difference, causing upward deflection from the positive baseline to 0 potential.
tive relative to G2. Therefore, the trace now shows a downward deflection. It then returns to the baseline as the nerve activity becomes too distant to affect the electrical field near the recording electrodes. This produces a diphasic51action potential as shown in Figure 2-4.
Effect of Volume Conduction The above discussion dealt with a directly recorded action potential in animal ex-
34
periments with no external conduction medium intervening between the pick-up electrodes and the nerve or muscle. During a clinical study, however, connective tissue and interstitial fluid act as volume conductors surrounding the generator sources.10,16,22 Here, an electrical field spreads from a source represented as a dipole; that is, a pair of positive and negative charges.3 In a volume conductor, currents move along an infinite number of pathways between the positive and negative ends of the dipole, with the greatest number of charges passing per unit time through a unit area along the straight path. The current flow decreases in proportion to the square of the distance from the generator source. Thus, the effect of the dipole gives rise to a voltage difference between the active recording electrode in the area of high current density and a reference electrode at a distance. Whether the electrode records positive or negative potentials depends on its spatial orientation to the opposing charges of the dipole. For example, an active electrode located at a point equidistant from the positive and negative charges registers no potential. The factors that together determine the amplitude of a recorded potential at a given electrode include charge density, or the net charge per unit area, surface areas of the dipole, and its proximity to the recording electrode.12 The theory of solid angle approximation pertains to analyzing of an action potential recorded through a volume conductor. This theory states that the solid angle subtended by an object equals the area of its surface divided by the squared distance from a specific point to the surface.5,27 The resting transmembrane potential consists of a series of dipoles arranged with positive charges on the outer surface and negative charges on the inner surface. Thus, it increases in proportion to the size of the polarized membrane viewed by the electrode and decreases with the distance between the electrode and the membrane. Solid angle approximation closely predicts the potential derived from a dipole layer as schematically shown in Figure 2-5. The propagating action potential, visualized as a positively charged wave front, or
Basics of Electrodiagnosis
Figure 2-5. Potential recorded at P from a cell with active (dark area) and inactive regions. In a, total solid angle consists of 1, 0,2, and 3. Potential at P subtending solid angles 1 and 3 equals zero as, in each, the nearer and farther membranes form a set of dipoles of equal magnitude but opposite polarity. In 2, however, cancellation fails because these two dipoles show the same polarity at the site of depolarization. In b, charges of the nearer and farther membranes subtending solid angle 2 are placed on the axial section through a cylindrical cell. A dipole sheet equal in area to the cross-section then represents the onset of depolarization traveling along the cell from left to right with positive poles in advance. [Adapted from Patton.48]
leading dipole, represents depolarization at the cross-section of the nerve at which the transmembrane potential reverses.46 A negatively charged wave front, or trailing dipole, follows, signaling repolarization of the activated zone. Analysis of Triphasic Waveform Analyzing waveforms plays an important role in the assessment of nerve or muscle action potentials. A sequence of potential
Electrical Properties of Nerve and Muscle changes arise as two sufficiently close wave fronts travel in the volume conductor from left to right (Fig. 2-6). This results in a positive-negative-positive triphasic wave as the moving fronts of the leading and trailing dipoles, representing depolarization and repolarization, approach, reach, and finally pass beyond the point of the recording electrode. Thus, an orthodromic sensory action potential from a deeply situated nerve gives rise to a triphasic waveform in surface recording. The potentials originating in the
Figure 2-6. Triphasic potential characterized by amplitude, duration (A-D), and rise time (B-C). A pair of wave fronts of opposite polarity represent depolarization and repolarization. The action potential travels from left to right in a volume conductor with the recording electrode (G1)near the active region and the reference electrode (G2) on a remote inactive point. A. G1 initially registers the positivity of the first dipole, which subtends a greater solid angle ( d) than the second dipole of negative front ( r). B. The relationship shown in A reverses, with gradual diminution of d, compared with r, as the active region approaches GI. C. The maximal negativity signals the arrival of the impulse directly under G1, which now registers only negative ends of the dipoles. D. The negativity declines as G1 begins to register the positive end of the second dipole. E. The polarity reverses again as r exceeds d.F. The trace returns to the baseline when the active region moves further away. The last positive phase, though smaller in amplitude, lasts longer than the first, indicating a slower time course of repolarization.
35
region near the electrode, however, lack the initial positivity, in the absence of an approaching volley. A compound muscle action potential, therefore, appears as a negative-positive diphasic waveform when recorded with the active electrode near the end-plate region where the volley initiates. In contrast, a pair of electrodes placed away from the activated muscle registers a positive-negative diphasic potential indicating that the impulse approaches but does not reach the recording site. The number of triphasic potentials generated by individual muscle fibers summate to give rise to a motor unit potential recorded in electromyography (see Chapter 13-5). The waveform of the recorded potential varies with the location of the recording tip relative to the source of the muscle potential.6,19,23,57,59 Thus, the same motor unit shows multiple profiles depending on the site of the exploring needle. Moving the recording electrode short distances away from the muscle fibers results in an obvious reduction in amplitude. Additionally, the duration of the positive-to-negative rising phase, or rise time, becomes greater. The rise time gives an important clue in determining proximity to the generator source. Amplitude may not serve for this purpose, because it may decrease with smaller muscle fibers or lower fiber density. According to the volume conductor theory, the location of the needle dictates the waveform of recorded potentials. Thus, the same single fiber discharge may be registered as initially positive triphasic fibrillation potential, initially negative biphasic endplate spike, or initially positive biphasic positive sharp wave (see Chapter 14-4). Despite this prevailing unifying concept,14'15 an accurate description of the observed potential often provides clinically useful information.41'42 For example, positive sharp waves recorded in the absence of fibrillation potentials may imply subliminal hyperexcitability of single muscle fibers, that "spontaneously" fire only with mechanical irritation of the needle. If the tip of a needle blocks a propagating impulse, the recorded potential appears as a positive sharp wave signaling only the approach of the positive front of depolarization.
Basics of Electrodiagnosis
36
Near-Field and Far-Field Potentials The specific potential recorded under a particular set of conditions depends not only on the location of the recording electrodes relative to the active tissue at any instant in time but also on the physical characteristics of the volume conductor. 11,13,37,38,45,56 The near- and far-field potentials distinguish two different manifestations of the volume-conducted field.30,31,53 The near field represents recording of a potential as it propagates under a pair of usually closely spaced electrodes placed directly over the path of the impulse. A bipolar recording registers primarily, though not exclusively, the near field from the axonal volley along the course of the nerve. In contrast, the far field implies detection of a voltage step long before the signal arrives at the recording site, usually by a pair of widely separated electrodes located far from the traveling volleys. A referential montage preferentially records far-field potentials unless one of the electrodes lies near the passage of the traveling volley. A far-field derivation has become popular in the study of evoked potentials to detect voltage sources generated at a distance. Original work on short-latency auditory evoked potentials30,31,53 suggested that synaptically activated neurons in the brainstem gave rise to stationary peaks. Subsequent animal studies60 emphasized the role of a synchronized volley of action potentials within afferent fiber tracts as their source. Further work with the human peripheral nerve has documented that stationary peaks can result solely from the propagating impulse in the absence of synaptic discharge.20,35,37-39,43 Hence, stationary activities registered in far-field recording may represent a fixed neural source such as synaptic discharges or, alternatively, a nonpropagating peak from an advancing front of axonal depolarization. As for the second of the two possibilities discussed above, short sequential segments of the brainstem pathways may each summate in far-field recording, resulting in successive peaks of the recorded potentials.1,2,60 This mechanism by itself,
however, does not account for the standing peaks derived from the propagating volleys at certain points along the greater length of the afferent pathway. In short-latency somatosensory evoked potentials (SEP) of the median24,25,58 or tibial nerve54 a voltage step develops between the two compartments when the moving volley encounters a sudden geometric change38at the border of the conducting medium. The same principles apply in the analysis of motor unit action potential and spontaneous single-fiber discharge.17,18,23 Here, each volume conductor on the opposite side of the boundary, in effect, acts as a lead connecting any points within the respective compartment 11to32 36 the voltage source at the partition. ' ' '55 Consequently, the potential difference remains nearly, though not exactly, the same regardless of the distance between G1 and G2, thus allowing detection of the voltage step in far-field recording. The designation, junctional or intercompartmental potential, differentiates this type of stationary peaks from fixed neural generators and helps specify the mechanism of the voltage step generated by the travelling impulse at a specific location (see Chapter 20-3).
REFERENCES 1. Arezzo JC, Legatt AD, Vaughan HG Jr: Topography and intracranial sources of somatosensory evoked potentials in the monkey. I. Early components. Electroencephalogr Clin Neurophysiol 46:155-172, 1979. 2. Arezzo JC, Vaughan HG Jr.: The contribution of afferent fiber tracts to the somatosensory evoked potentials. In Bodis-Wollner I (ed), Evoked Potentials. Ann NY Acad Sci 388:679-682, 1982. 3. Boyd DC, Lawrence PD, Bratty P: On modeling the single motor unit action potential. IEEE Trans Biomed Eng 25:236, 1978. 4. Brazier MAB: Electrical Activity of the Nervous System, ed 4. Pitman Medical, Kent, Great Britain, 1977. 5. Brown BH: Theoretical and experimental waveform analysis of human compound nerve action potentials using surface electrodes. Med Biol Eng 6:375, 1968. 6. Buchthal F, Guld C, Rosenfalck P: Volume conduction of the spike of the motor unit potential investigated with a new type of multielectrode. Acta Physiol Scand 38:331-354, 1957. 7. Carpenter RHS: Neurophysiology, ed 3. Edward Arnold, London, 1995. 8. Catterall WA: Cellular and molecular biology of
Electrical Properties of Nerve and Muscle voltage-gated sodium channels. Physiol Rev 72: S15-48, 1992. 9. Chiu SY, Ritchie JM, Rogart RB, Stagg D: A quantitative description of membrane currents in rabbit myelinated nerve. J Physiol (Lend) 292: 149-166, 1979. 10. Clark J, Plonsey R: The extracellular potential field of the single active nerve fiber in a volume conductor. Biophys J 8:842-864, 1968. 11. Cunningham K, Halliday AM, Jones SJ: Stationary peaks caused by abrupt changes in volume conductor dimensions: potential field modelling. Abstract. Electroencephalogr Clin Neurophysiol 61:S100, 1985. 12. DeLisa JA, Kraft GH, Gans BM: Clinical electromyography and nerve conduction studies. Orthop Rev 7:75-84, 1978. 13. Desmedt JE, Huy NT, Carmeliet J: Unexpected latency shifts of the stationary P9 somatosensory evoked potential far field with changes in shoulder position. Electroencephalogr Clin Neurophysiol 56:623-627, 1983. 14. Dumitru D: Single muscle fiber discharges (insertional activity, end-plate potentials, positive sharp waves, and fibrillation potentials): A unifying proposal. Muscle Nerve 19:221-226, 1996. 15. Dumitru D: Issues & Opinion: Rebuttal. Muscle Nerve 19:229-230, 1996. 16. Dumitru D, DeLisa JA: AAEM Minimonograph #10: Volume Conduction. 1991. 17. Dumitru D, King JC, Rogers WE: Motor unit action potential components and physiologic duration. Muscle Nerve 22:733-741, 1999. 18. Dumitru D, King JC, Rogers WE, Stegeman DF: Positive sharp wave and fibrillation potential modeling. Muscle Nerve 22:242-251, 1999. 19. Dumitru D, King JC, van der Rijt W: The biphasic morphology of voluntary and spontaneous single muscle fiber action potentials. Muscle Nerve 17:1301-1307, 1994. 20. Eisen A, Odusote K, Bozek C, Hoirch M: Farfield potentials from peripheral nerve: generated at sites of muscle mass change. Neurology 36:815-818, 1986. 21. Gasser HS: The classification of nerve fibers. Ohio J Science 41:145-159, 1941. 22. Gath I, Stalberg E: On the volume conduction in human skeletal muscle: In situ measurements. Electroencephalogr Clin Neurophysiol 43:106-110, 1977. 23. Gootzen TH, Stegeman DF, Van Oosterom A: Finite limb dimensions and finite muscle length in a model for the generation of electromyographic signals. EEG Clin Neurophysiol 81:152162, 1991. 24. Hashimoto S, Kawamura J, Segawa Y, Yamamoto T, Nakamura M; Possible model for generation of Pg far-field potential. Muscle Nerve 15:106-110, 1992. 25. Hashimoto S, Segawa Y: Model of generation of Pg far-field potentials using an electric circuit diagram. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science, England, 1996, pp 251-254. 26. Hille B: Ionic basis of resting and action potentials. In Geiger SR (ed): Handbook of Physiology: Section 1: The Nervous System, Vol 1. American
37
27. 28.
29. 30.
31. 32.
33. 34. 35.
36.
37.
38.
39.
40. 41. 42. 43. 44.
Physiological Society, Bethesda, MD, 1977, pp 99-136. Hodgkin AL: The Conduction of the Nervous Impulse. The Sherrington Lectures, Vol 7. Liverpool University Press, Liverpool, 1965. Hodgkin AL, Huxley AF: A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol (Lond) 117:500-544, 1952. Hodgkin AL: Ionic movements and electrical activity in giant nerve fibers. Proc R Soc (Lond) Ser B 148:1-37, 1958. Jewett DL: Volume-conducted potentials in response to auditory stimuli as detected by averaging in the cat. Electroencephalogr Clin Neurophysiol 28:609-618, 1970. Jewett DL, Williston JS: Auditory-evoked far fields averaged from the scalp of humans. Brain 94:681-696, 1971. Jones SJ: Insights into the origin of subcortical SEPs gained from potential field models. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science, England, 1996, pp 255-259. Kandel E, Schwartz JH, (eds): Principles of Neural Science, ed 2. Elsevier, Amsterdam, 1985, pp 13-208. Katz B: Nerve, Muscle and Synapse. McGrawHill, New York, 1966. Kimura J: Field theory: The origin of stationary peaks from a moving source. In International Symposium on Somatosensory Evoked Potentials. Custom Printing, Rochester, MN, 1984, pp 39-50. Kimura J, Kimura A, Ishida T, Kudo Y, Suzuki S, Machida M, Matsuoka H, Yamada T: What determines the latency and the amplitude of stationary peaks in far-field recordings? Ann Neurol 19:479-486, 1986. Kimura J, Mitsudome A, Beck DO, Yamada T, Dickins QS: Field distributions of antidromically activated digital nerve potentials: model for farfield recording. Neurology (Cleveland) 33:11641169, 1983. Kimura J, Mitsudome A, Yamada T, Dickens QS: Stationary peaks from a moving source in farfield recording. Electroencephalogr Clin Neurophysiol 58:351-361, 1984. Kimura J, Yamada T, Shivapour E, Dickins QS: Neural pathways of somatosensory evoked potentials: Clinical implication. In Buser PA, Cobb WA, Okuma T (eds), Kyoto Symposium (EEG Suppl 36). Elsevier, Amsterdam, 1982, pp 328335. Kocsis JD, Waxman SG: Absence of potassium conductance in central myelinated axons. Nature 287:348-349, 1980. Kraft GH: Are fibrillation potentials and positive sharp waves the same? No. Muscle Nerve 19:216-220, 1996. Kraft GH: Issues & Opinions: Rebuttal. Muscle Nerve 19:227-228, 1996. Lin JT, Phillips LH II, Daube JR: Far-field potentials recorded from peripheral nerves. Electroencephalogr Clin Neurophysiol 50:174, 1980. Lorente De No R: Analysis of the distribution of the action currents of nerve in volume conduc-
Basics of Electrodiagnosis
38
45. 46.
47. 48.
49. 50.
51. 52. 53. 54.
55.
tors. In: Studies from the Rockefeller Institute for Medical Research: A Study of Nerve Physiology, Vol 132. The Rockefeller Institute for Medical Research, New York, 1947, pp 384-482. Nakanishi T: Origin of action potential recorded by fluid electrodes. Electroencephalogr Clin Neurophysiol 55:114-115, 1983. Patton HD: Special properties of nerve trunks and tracts. In Ruch HD, Patton HD, Woodbury JW, Towe AL (eds): Neurophysiology, ed 2. WB Saunders, Philadelphia, 1965, pp 73-94. Patton HD, Sundsten JW, Crill WE, Swanson PD: Introduction to Basic Neurology. WB Saunders, Philadelphia, 1976. Patton HD: Resting and action potentials of neurons. In Patton HD, Sundsten JW, Crill WE, Swanson PD (eds): Introduction to Basic Neurology. WB Saunders, Philadelphia, 1976. Ritchie JM: Physiology of axons. In: Waxman SG, Kocsis JD, Stys PK (eds): The Axon. 51 Oxford University Press, New York, 1995, pp 68-96. Rizzo MA, Kocsis DD, Waxman SG: Slow sodium conductances of dorsal root ganglion neurons: intraneuronal homogeneity and intemeuronal heterogeneity. J Neurophysiol 72:2796-2815, 1994. Rosenfalck P: Intra and extracellular potential fields of active nerve and muscle fibers. Acta Physiol Scand (Suppl 321): 1, 1969. Ruch TC, Fulton JF: Medical Physiology and Biophysics, ed 20. WB Saunders, Philadelphia, 1973. Sohmer H, Feinmesser M: Cochlear and cortical audiometry conveniently recorded in the same subject. Israel J Med Sci 6:219-223, 1970. Sonoo M: P15 in tibial nerve SEP as a simple example of the junctional potential. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science, England, 1996, pp 260-265. Stegeman DF, Roeleveld K, Dumitru D, Vinger-
56.
57.
58.
59.
60.
61.
62.
63.
hoets DM: Far-field potentials in surface EMG. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science, England, 1996, pp. 271-275. Stegeman D, Van Oosteron A, Colon E: Simulation of far field stationary potentials due to changes in the volume conductor. Abstract. Electroencephalogr Clin Neurophysiol 61:S228, 1985. Theeuwen MM, Gootzen TH, Stegman DF: Muscle electric activity. I: A model study on the effect of needle electrodes on single fiber action. Ann Biomed Engin 21:377-339, 1993. Urasaki E: A direct recording study of subcortical somatosensory evoked potentials. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science, England, 1996, pp 266-270. Van Veen BK, Wolters H, Wallinga W, Rutten WL, Boom HB: The bioelectrical source in computing single muscle fiber action potentials. BiophysJ 64:1492-1498, 1993. Vaughan HG Jr: The neural origins of human event-related potentials. In Bodis-Wollner I (ed): Evoked Potentials. Ann NY Acad Sci 388:125138, 1982. Waxman SG: Special Lecture: Adrian Lecture: From Lord Adrian to ion channels and beyond: the molecular basis of nerve transmission. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, pp 1-7. Woodbury JW: Action potential: Properties of excitable membranes. In Ruch TC, et al (eds): Neurophysiology, ed 2. WB Saunders, Philadelphia, 1965, pp 26-57. Woodbury JW: The cell membrane: Ionic and potential gradients and active transport. In Ruch TC, et al (eds): Neurophysiology, ed 2. WB Saunders, Philadelphia, 1965, pp 1-25.
Chapter 3 ELECTRONIC SYSTEMS AND DATA ANALYSIS 1. INTRODUCTION 2. ELECTRODES Preparation of Needle Electrodes Types of Available Electrodes 3. ELECTRODE AMPLIFIERS Differential Amplifiers Common Mode Rejection Ratio Means of Reducing Interference Input Impedance Frequency Response 4. VISUAL DISPLAYS Cathode-Ray Tube Delay Line Multiple Channel Recording Storage Oscilloscope 5. OTHER RECORDING APPARATUS Loudspeaker Magnetic Tape Recorder 6. ARTIFACTS Electrode Noise Amplifier Noise Defective Apparatus Movement Artifact Electrostatic and Electromagnetic Interference Radio and Mobile Phone Interference 7. STIMULATORS Electrical Stimulation Requirements Stimulus Isolation Constant-Voltage versus Constant-Current Magnetic Coil Stimulation 8. NORMATIVE DATA AND STATISTICS Control Values Statistical Analysis False-Positive versus False-Negative Results 9. EXPERT SYSTEMS AND QUALITY DEVELOPMENT KANDID ESTEEM MUNIN Interlaboratory Communication 39
Basics of Electrodiagnosis
40
1 INTRODUCTION The apparatus used in the performance of routine electrodiagnosis includes electrodes, amplifiers, displays, loudspeakers, and data storage devices. Muscle or nerve action potentials can be recorded by either a surface electrode placed on the skin over the target or a needle electrode inserted closer to the source. Surface electrodes register a summated electrical activity from many muscle or nerve fibers, whereas needle electrodes discriminate individual motor unit potentials discharging within a narrow radius from the recording tip. The electrical and physical characteristics of recording electrodes dictate the amplitude and other aspects of the potentials under study.43,44 Electromyographers analyze both the amplified waveform of action potentials on a visual display and the auditory characteristics of the signals heard through a loudspeaker. The kind of information desired and the type of activities under study determine the optimal amplifier settings. Devices for permanent recordings include photographs with Polaroid films, a fiber-optic system with sensitive papers, a magnetic tape recorder, and digital storage. Amplitude and time calibrations verify the accuracy of the stored signals. This chapter deals with practical aspects of instrumentation, 12,39,40,60 without a detailed discussion of electronics (see Appendix 2). 2
ELECTRODES
The signals recorded during voluntary muscle contraction depend to a great extent on the type of recording electrode used.15,38 Surface electrodes placed over the muscle record summated activities from many motor units. The use of a needle electrode allows recording of individual motor unit potentials during mild muscle contraction. With increased effort, synchronous activity in many adjacent motor units precludes the identification of single motor units. For routine purposes, clinical electromyographers use standard concen-
tric, bipolar concentric,2 or monopolar nee50 dles. Single-fiber electrodes have a leading edge small enough to allow recording of potentials derived from single muscle fibers in isolation.30,72 Less commonly used "special purpose" electrodes include the multielectrode, the flexible wire electrode, and the microelectrode placed intracellularly.14 Electrode lead wires should have protected pins to prevent inadvertent connection to a power source, 3causing shocks, burns and electrocutions.
Preparation of Needle Electrodes Sterilization of needle electrodes in boiling water for at least 20 minutes prior to use prevents the transmission of infection. Commercially available sterilizers bring the water temperature to 100° C and maintain it without excessive boiling. Only the metal and plastic components of needle electrodes will withstand the time and temperature of steam autoclaving, thus the need to detach nonautoclavable connectors and lead wires before the sterilization procedure. Gas sterilization also suffices, although the chemicals used may damage the plastic, causing defects in insulation. Thorough outgassing of electrodes reduces the amount of the agent retained in the plastic material. Electrode manufacturers provide instructions for optimal sterilization methods. With the advent of less costly disposables, it has now become a common practice in many laboratories to discard needle electrodes after use in each patient. The American Association of Electromyography and Electrodiagnosis42 recommends this practice to circumvent any concerns of possible transmission of diseases, especially after studying a patient with AIDS, hepatitis, or any other contagious disorder. Jakob-Creutzfeldt disease poses special problems because the transmissible agent responsible for the disease may resist conventional sterilization procedures.37,58 Further precaution before disposal, therefore, calls for incinerating used needles and blood-contaminated materials, or autoclaving them for one hour at 120° C and 15 PSI.8
Electronic Systems and Data Analysis Electrical properties of commercial needle electrodes vary considerably. Electrolytic treatment of reusable needles temporarily improves their performance.23 Periodic examination of needle electrodes ensures their structural integrity. The inner concentric shaft may become corroded. The Teflon coating of monopolar needles may peel off, exposing the insulated portion of the conductor. An increase in recording surface tends to reduce the amplitude and area of the recorded motor unit potentials21with relatively little effect on its duration. The use of a dissecting microscope with a magnifying factor of ten helps detect a slight bend in the shaft or a crack in the tip. To test needle insulation, one terminal of a battery can be connected to the lead of a needle and the other terminal to an ammeter with a small exploring metal hook or moist cotton. A current should flow only if the exploring hook touches the exposed tip of the needle. Any current, if registered while exploring the shaft of the needle, indicates defective insulation. An ammeter should register no current if connected to the battery through the two leads of standard or bipolar concentric needles unless there is a short circuit at the needle tip. A current will flow normally with immersion of the needle tip in water.
41
Types of Available Electrodes Figure 3-1 illustrates common electrodes used in electromyography. SURFACE ELECTRODES
Surface electrodes, square or round metal plates made of platinum or silver, come in different sizes with an average dimension of 1 X 1 cm. An adhesive tape suffices for applying them to the skin, although the use of collodion improves stability in long-term monitoring. Cleansing the skin with alcohol, scraping the calloused surface, and applying electrolyte cream under the electrode reduces impedance. Too much paste, however, can form a bridge between the two recording electrodes, cancelling the voltage difference. A short circuit between the stimulator and pick-up electrodes or ground introduces a large stimulus artifact. Perspiration can act in a similar manner. Time-efficient application of adhesive electrodes, including those marketed for electrocardiography, provides the same results as those obtained by the usual disc electrodes applied with adhesive tapes.11 Steady electrode offset voltage at the interface, not recorded by the amplifier, can
Figure 3-1. Schematic illustration of (a) standard or coaxial bipolar, (b) concentric bipolar, (c) monopolar, and (d,e) single-fiber needles. Dimensions vary, but the diameters of the outside cannulas shown resemble 26-gauge hypodermic needles (460 um) for a, d, and e, a 23-gauge needle (640 um) for b, and a 28-gauge needle 360 um for c. The exposed tip areas measure 150 x 600 um for a, 150 x 300 um with spacing between wires of 200 um center to center for b, 0.14 mm2 for c, and 25 um in diameter for d and e. A flat skin electrode completes the circuit with unipolar electrodes shown in c and d. [Modified from Stalberg and Trontelj.72]
42
give rise to an artifact if movement causes a sudden mechanical change in the metalelectrolyte interface. To reduce this type of potential, some surface electrodes allow most movement to occur between electrolyte and skin rather than at the metal-electrolyte interface. A surface electrode is best suited for monitoring voluntary muscle contraction during kinesiologic studies and recording evoked compound nerve or muscle action potentials. It registers electrical activities nonselectively from a wider region, covering the recording radius of some 20 mm compared to selective pickup from a 500 um radius by a needle electrode.10 The amplitude of compound muscle action potentials decreases with increasing electrode size.79 The surface electrode also serves well as a stimulating probe, a reference, or a ground lead in conjunction with the monopolar needle, but not as an active electrode to study motor unit potentials, because it fails to reproduce high-frequency components adequately. STANDARD OR COAXIAL CONCENTRIC NEEDLE
This electrode, introduced by Adrian and Bronk2 in 1929, has a stainless-steel cannula similar to hypodermic needles, with a wire in the center of the shaft. The wire, usually made of nichrome, silver, or platinum, measures 0.1 mm or slightly larger as compared to the external rim of the shaft, 0.3 mm in diameter. The pointed tip of the needle has an oval shape with an exposed area of about 150 um x 600 (um, and an impedance of around 50 kilohms. The wire and shaft, bare at the tip, form a spheric rather than hemispheric recording territory as might be anticipated by the direction.28 The needle, when near a source of electrical activity, registers the potential difference between the wire and the shaft, showing a restricted recording area. In fact, in the recording of a single motor unit discharge that extends at least 1 cm in diameter, only the muscle fibers located within about 500 um radius from the tip of the needle contribute to the amplitude, and those within 2.5 mm to the duration of the recorded potential.27 Thus, although recording characteristics
Basics of Electrodiagnosis vary from one type of needle to another, the pickup area, in general, constitutes a very small portion of the motor unit territory. A separate surface electrode, taped or applied with a suction cup, serves as the ground.26 Disposable concentric needles generally compare reasonably well with reusable electrodes, although electric or physical testing of the leads may not adequately 62predict their recording characteristics. BIPOLAR CONCENTRIC NEEDLE
The cannula contains two fine stainless steel or platinum wires. This electrode, therefore, has a larger diameter than the standard concentric needle for the same size wires embedded. The electrode registers the potential difference between the two inside wires, with the cannula serving as the ground. The bipolar electrode thus detects potentials from a much smaller volume than the standard needle. The three terminals in the connecting cable consist of two active leads and a ground connection. In this type of recording from a very localized area, only a small number of single muscle fibers contribute as the source for electrical activity.55 This restricted recording range provides selectivity, but at the risk of disregarding the overall activity of motor units. Concentric electrodes tend to detect more spontaneous potentials than monopolar needles probably because of increased tissue injury.71 MONOPOLAR NEEDLE
This electrode, made of stainless steel for its mechanical properties, has a fine point, insulated except at the distal 0.2 to 0.4 mm. The wire, covered by a Teflon sleeve, has an average diameter of about 0.8 mm. A surface electrode or a second needle in the subcutaneous tissue serves as a reference lead and a separate surface electrode, placed on the skin, as a ground. Its sharp tip causes less pain during insertion, but it is less stable electrically, hence 52,74 noisier than the concentric electrode. The average impedance ranges from 1.4 megohms at 10 Hz to 6.6 kilohms at 10 KHz.80 Presoaking the elec-
Electronic Systems and Data Analysis trodes with a small concentration of a wetting agent in saline solution reduces the impedance by 6- to 20-fold. This pretreatment improves the resolution of low amplitude signals. A monopolar needle records voltage changes between the tip of the electrode and the reference. The spatial recording characteristics,54 differ considerably from one type of needle to another. In general, a monopolar needle registers a potential 29,61,65, that is twice as large and more complex, from the same source, than a concentric needle, although duration and firing rate remain the same.49 SINGLE-FIBER NEEDLE
Single-fiber electromyography requires an electrode with a much smaller leading edge, to record from individual muscle fibers rather than motor units (see Chapter 16-2). A wire 25 um in diameter mounted on the side of a needle provides the maximal amplitude discrimination between near and distant muscle fiber potentials.31 As in concentric electrodes, single-fiber needles may contain two or more wires exposed along the shaft, serving as the leading edge. The most commonly used type has one wire inserted into a cannula with its end bent toward the side of the cannula, a few millimeters behind the tip.72 The spatial recording characteristics of single-fiber needles show specific asymmetries and a greater potential decline with radial distance compared with concentric or monopolar electrodes.53,72
43
along the shaft. Similar multi-lead electrodes may usefully 6serve for intraoperative nerve recording. FLEXIBLE WIRE
A flexible wire, usually introduced through a hypodermic needle, permits freedom of movement in kinesiologic examination. Some investigators prefer a bipolar electrode made of nylon-coated Evanohm alloy wire, 25 um in diameter.13 Although this type of electrode comes in different sizes, the most commonly used type has insulated platinum wires 50-100 /mi in diameter with the tip bare. A small hole made in the insulation of the wire may provide smaller lead-off surfaces on the order of 10-20 umi.45 These electrodes, however, lack the rigid standardization required for quantitative studies of action potentials.72 GLASS MICROELECTRODES
A glass microelectrode used for intracellular recording consists of fine glass tubing filled with potassium chloride solution. Because of its extreme fragility, one must use a cannula as a carrier to introduce the electrode through the skin, and a micromanipulator to insert it into the exposed muscle. The electrode has a very fine tip, less than 1 um in diameter, and consequently a very high impedance, on the order of 5 megohms. Therefore, recording from a glass microelectrode requires amplifiers of exceedingly high input impedance.14
MULTIELECTRODES
Multielectrodes contain three or more insulated wires, usually 1 x 1 mm in size, exposed through the side of the cannula. 16 One of the wires serves as the indifferent electrode; the outside cannula of the electrode, 1 mm in diameter, is connected to the ground. The separation between the leads along the side of the multielectrode determines the recording radius. The commonly used distances in measuring the motor unit territory include 0.5 mm for myopathy and 1.0 mm for neuropathy. The single-fiber needle may also contain multiple wires exposed
3
ELECTRODE AMPLIFIERS
Potentials assessed during electrodiagnostic examinations range in amplitude from microvolts to millivolts. With the oscilloscope display set at 1 V per cm, signals of 1 mV and 1 mV, if amplified 1 million times and 1000 times, respectively, cause a 1 cm deflection. To accomplish this range of sensitivity, the amplifier consists of several stages. One system uses a preamplifier with a gain of 500, followed by several amplifier and attenuator stages
Basics of Electrodiagnosis
44
to produce a variable gain of 2-2000. This arrangement increases the signal-to-noise ratio by allowing major amplification of the signal near the source prior to the emergence of noise that develops in the following circuits. To achieve this goal the preamplifier must have a high input impedance, a low noise level, and a large dynamic range.
Differential Amplifiers During electromyographic examination, a major source of interference conies from the coupled potential of the alternating current power line. The magnitude of this field can exceed that of biological potential by a million times. Proper assessment of the signal, therefore, requires its selective amplification without, at the same time, magnifying the noise. This would be impossible if the apparatus amplifies any voltage appearing between an input terminal and the ground terminal. Differential amplifiers used in most electromyography, therefore, amplify only the voltage difference between the two input terminals connected to the recording electrodes. This system effectively rejects common mode voltages, which appear between both input terminals and common ground. These include not only power line interference but also distant muscle action potentials that affect the two recording electrodes equally.
Common Mode Rejection Ratio Inherent imbalance in the electrical system of an amplifier renders rejection of the common mode voltage less than perfect. The common mode rejection ratio specifies the degree of differential amplification between the signal and the common mode voltage. Good differential amplifiers should have rejection ratios exceeding 100,000; that is, 100,000 times more amplification of the signals than unwanted potentials appearing as a common mode voltage. A very high rejection ratio, however, will not guarantee the complete elimination of external interference caused by undesired distant potentials, for two reasons. First,
electromagnetic interference affects the two recording electrodes almost, but not quite, equally depending on their relative positions. Second, the contact impedances inevitably differ between the two recording electrodes, leading to unequal distribution of the same common mode voltage. A common mode voltage too large to be perfectly balanced overloads the amplifier.
Means of Reducing Interference Other precautions for minimizing electromagnetic interference include reducing and balancing contact impedances of the two electrodes and the use of short, wellshielded electrode cables. The system must effectively ground not only the patient and the bed, but also the instrument and, if necessary, the examiner. Major interference may originate from unshielded power cords running to other appliances in the vicinity of the recording instrument. With adequate care, most modern equipment operates well without a shielded room. In the presence of electrical noise uncontrollable by ordinary means, a properly constructed Faraday shield can dramatically reduce the interference. To be effective it should enclose the examining room as one continuous conductor and be grounded at one point. The 50 or 60 Hz filter available in most instruments reduces power line interference at the expense of distorting electromyographic signals. Thus, only special situations, such as portable recording in an intensive care unit, may warrant their application when all other attempts have failed.
Input Impedance Analogous to the resistance in a DC circuit, the impedance in an AC circuit determines the current flow for a given alternating voltage source. For recording muscle or nerve action potentials, the tissue and electrode wires add only negligible impedances compared with those at the needle tip and at the input terminal of the amplifier. In this circuit, the needle
Electronic Systems and Data Analysis
45
tip and the input terminal act as a voltage divider with voltage changes occurring in proportion to the respective impedance. Thus, with the impedance equally divided between these two, only one-half of the original potential will appear across the input terminal. Increasing the input impedance of the amplifier to a level much higher than that of the electrode tip would minimize the loss. The input impedances of most amplifiers range from 100 kilohms to hundreds of megohms. An amplifier with a high input impedance also improves the common mode rejection ratio because the higher the input impedance, the smaller the effect of electrical asymmetry of the recording electrodes. Higher electrode impedances increase amplifier noise and external interference, although electrode impedances as high as 50 times usual values apparently cause no major waveform distortion.l
Here the new waveform approximates the first derivative (rate of change) of the original signal. Extending the frequency response too low causes instability of the baseline, which then shifts slowly in response to changing biopotentials. The analog filters also affect the peak latency of the recorded response because of phase shift. High frequency filtering increases, whereas low frequency filtering reduces, the apparent latency of peaks. The use of digital filtering, which introduces zero phase shift, circumvents this problem in clinical assessments.41,56'57 A square wave pulse of known amplitude and duration usually serves as a calibration signal to accurately determinine the amplitude and duration of the recorded potentials. The distortion seen in the square pulse results from the effects of high- and low-frequency filters. Its rise time indicates the high-frequency response, and the slope of the flat top, the low-frequency response (see Appendix Figs. A2-18 and A2-20). Other calibration signals include sine waves from the power line or discontinuous waveforms of known frequency and amplitude.
Frequency Response Most commercially available apparatuses have variable high- and low-bandpass filters to adjust frequency response according to39,40 the type of potentials under study. Fourier analysis of complex waveforms encountered in electromyography reveals sine waves of different frequencies as their harmonic constituents. The prominent sine wave frequencies of muscle action potentials, for example, range from 2 Hz to 10 KHz. For clinical electromyography, the frequency band of the amplifier ideally should cover this range.19'20 In the presence of interfering high-pitched noise or DC drift, however, a bandpass extending from 20 Hz to 5 KHz suffices. Filter settings must remain constant in serial studies. Their modification within the routine range results in statistically significant alteration of waveform.66 A high frequency filter (low pass), if set too low, reduces the amplitude of high frequency components disproportionately. Extending the high frequency response beyond the band required for proper recording results in an unnecessary increase in background noise. A low frequency filter (high pass), if set too high, distorts the slowly changing potential.
4
VISUAL DISPLAYS
Appropriate amplification ensures an optimal display of the waveform for visual analysis. The cathode-ray tube (CRT), with no mechanical limitations in dynamic high-frequency response, provides an excellent means to trace rapidly changing amplitude against time. Cathode-Ray Tube An electron gun discharges an electron beam internally toward the glass screen of a CRT. When struck by a beam of electrons, the phosphor coating on the inside surface of the screen emits light. The adjustable voltage between a pair of vertically placed plates (called horizontal deflection plates) determines the horizontal position of this bright spot. Applying a linearly increasing voltage to the plates makes the spot sweep at a constant speed.
Basics of Electrodiagnosis
46
A pair of horizontally placed plates (called vertical deflection plates), connected to the signal voltage from the amplifier, control the vertical position of the electron beam. The waveform displayed on the face of the screen, therefore, represents changing amplitude of the signal voltage in time. The vertical axis represents response amplitude, whereas the horizontal axis shows units of time. Electromyographic examination usually uses a free-running mode: when the spot reaches the end of the screen, it returns rapidly to the beginning to repeat. Most manufacturers now provide digital circuitry to process and store the potentials before displaying them on a monitor. Delay Line Instead of being free running, the horizontal sweep may initate on command. In this mode of operation, a motor unit potential itself can trigger the sweep. Thus, a given motor unit potential recurs successively at the beginning of each sweep for detailed analysis, although, by design, the portion of the waveform preceding the trigger point fails to appear on the screen. In an analog machine, an electronic delay circumvents this difficulty by storing the recorded motor unit potential for a short period. After a predetermined delay following the onset of a sweep triggered by the real-time potential, the stored signal leaves the delay line for display on the screen. With this arrangement, the potential in question occurs repetitively and in its entirety on the same spot of the screen for precise determination of its amplitude and duration.63 With digital circuitry, the computer begins displaying data at any desired point prior to the trigger, thus accomplishing the same objective. Multiple Channel Recording Some electromyographic instruments have multiple channels to allow simultaneous recording from two or more sets of electrodes. Typically, two or more channels share a beam from a single gun by switching the point vertically between the base-
lines of different traces as the beam sweeps horizontally across the screen. This electrical switching takes place so fast that each trace appears to be continuous despite the interruption from one trace to the next. Storage Oscilloscope Storage oscilloscopes have a different cathode ray tube that retains traces on the face of the screen for several hours. A second electron gun floods the screen to visualize the trace retained as electrostatic charges on a mesh behind the screen. Electrically discharging the mesh can quickly erase the stored pattern. The advent of digital storage and display techniques have made such storage oscilloscopes obsolete.
5
OTHER RECORDING APPARATUS Loudspeaker
Muscle or nerve action potentials have distinct auditory characteristics when played through a loudspeaker. For clinical analyses, electromyographers depend very heavily on the sounds produced by different kinds of spontaneous or voluntarily activated muscle potentials during needle examination. For example, fibrillation potentials sound like "rain on a tin roof" (see Chapter 14-4). Acoustic properties also help distinguish a nearby motor unit with a clear, crisp sound, reflecting a short rise time, from distant units with dull sound (see Chapter 13-5). In fact, an experienced examiner can detect the difference between near and distant units by sound better than by oscilloscope display. The acoustic cues often guide in properly repositioning the needle close to the source of the discharge.
Magnetic Tape Recorder Magnetic tape provides one means to store electrical potentials for later analysis. Am-
Electronic Systems and Data Analysis plitude modulation (AM) impresses the signal itself on the tape, whereas frequency modulation (FM) records the signal after converting it to a varying frequency of constant amplitude. The AM recording registers high frequency potentials well but not low-frequency responses below 10 to 15 Hz. In contrast, the FM method has a better low-frequency response, although it requires a very high tape speed to achieve the high-frequency response required for electromyography. The FM method reproduces the amplitude of potentials more accurately than the AM method. 6
ARTIFACTS
Not all electrical potentials registered during an electromyographic examination originate in muscle or nerve. Any voltage not attributable to the biologic potential sought represents an artifact, which usually causes a unique discharge pattern on
Figure 3-2. Artifacts induced by a cardiac pacemaker recorded by a monopolar needle electrode from (a, b) gluteus medius and (c, d) paraspinal muscle. Note opposite polarity of the sharp discharge at the two recording sites. The interval between the successive impulses of 800 ms corresponds to a discharge frequency of 75 impulses/minute. Trains in a and c show continuous recordings from top to bottom; those in b and d, interrupted tracings from one sweep to the next.
47
the oscilloscope and distinct sounds through the loudspeaker.40 Some noises, however, mimic biologic activity so closely that even a trained examiner may have difficulty in identifying them. Most artifacts unaffected by the position of the recording electrode originate outside the muscle. These exogenous activities may result from peculiarity of the patient, like those induced by a cardiac pacemaker (Fig. 3-2) or transcutaneous stimulator (Fig. 3-3). More commonly, they result from 60 Hz interference caused by the electrostatic or electromagnetic fields of electrical appliances. Improper or inadequate grounding results in electromagnetic interference from the nearby alternating current source. Different generator sources give rise to characteristic, though not specific, patterns for easy identification (Fig. 3-4). Artifacts may also originate in the recording instruments themselves or from a more remote generator, such as a hammer drill (Fig. 3-5). A loose connection in one or more parts of the recording circuit may generate electrical activity, similar to
Basics of Electrodiagnosis
48
Figure 3-3. Artifact induced by a transcutaneous stimulator. The 14 ms interval between the successive impulses (a, b, c) corresponds to an approximate discharge rate of 70 impulses per second; the 7 ms interval (d, e, f), a faster rate of 140 impulses per second.
the muscle action potential. Impedance variability within the muscle tissue may also cause electrical activity, depending on the location of the needle tip. Genuine biologic potentials generated in the muscle include end-plate noises and end-plate spikes (see Chapter 13-4). These artifacts may mimic the intended signals sought during electromyographic examination (see Figs. 13-3 and 13-4).
Electrode Noise Potentials may arise from two active metals or the metal-fluid junction at the needle electrode located intramuscularly. A constant electrode-fluid potential by po-
larization may distort the signals, whereas changing potentials will result in electrode noise. A small electrode tip, because of its high impedance, causes a greater voltage drop during the passage of current. Thus, the smaller the electrode surface, the greater the interference from its polarization or electrode noise. Therefore, the type of metal alters the recording characteristics of the needle electrode much more than those of the surface electrode. In fact, an electrode potential from active metals too small to affect surface recording could undermine the function of intramuscular studies. The use of relatively inert metals for needle electrodes, such as stainless steel or platinum, minimizes such adverse effects.
Electronic Systems and Data Analysis
49
Figure 3-4. Various types of 60 Hz interference induced by nearby electrical appliances. They include (a) common pattern, (b) spikes from high impedance of the recording electrode, (c) fluorescent light, 120 Hz interference from (d) diathermy unit and (e and /) heat lamp.
Amplifier Noise Electrical noise inherent in an amplifier originates from all components, including the resistors, transistors, and integrated circuits. Noise arising from the thermal agitation of electrons in a resistor increases with the impedance in the input stage. Microphonic noise results from the mechanical vibration of various components. The use of a high-pass filter suppresses low frequency noise from these and other sources in amplifier circuits. A low pass filter reduces high-frequency noise, which appears as a thickening of the baseline as it sweeps across the screen accompanied by a hissing noise on the
loudspeaker (Fig. 3-6). The level of amplifier noise as perceived on the oscilloscope increases in proportion to amplifier gain and frequency response. Thus, operating the system at lower gains and with narrower filter band widths substantially reduces this component of noise seen on the screen. Defective Apparatus By far the most likely cause of recording problems relates to a defect in one of the three recording electrodes or its application. A broken wire induces bizarre and unsuspected artifacts even if the insulat-
50
Basics of Electrodiagnosis
Figure 3-5. Effect of (a, b, c) a hammer drill operated nearby, and (d, e,f\ oscillation of the amplifier circuits probably induced by an excessively high impedance of the electrode tip. Both superficially resemble the complex repetitive discharge, but the recordings with a fast sweep speed (c and f) uncover a waveform and pattern of recurrence not usually associated with a biologic discharge.
ing cover appears intact. A partially severed conductor may generate very deceptive movement-induced potentials, which recur with muscle twitch, mimicking stimulus-locked evoked signals. Other common causes of artifacts include defective insulation of a monopolar needle or a concentric needle with a short-circuited tip. A 2-year study on durability revealed the feasibility of reusing monopolar electrodes on the average in 20-63 patients.59 Failure occurred, in order of frequency, as the result of Teflon retraction, a dull or burred tip, a break in a wire or pin, electrical artifacts, or a bend in the needle shaft. Inadvertent insulation of the electrode tip by blood protein "baked on" in
the process of autoclaving can also distort the potential. Careful cleaning of the needle tip prior to autoclaving will alleviate this problem. If necessary, application of an ultrasonic vibrator loosens dried material from the needle. The use of disposable needles precludes problems inherent to sterilization, but unused electrodes may manifest similar artifacts, caused by mechanical defects induced during the manufacturing process. Movement Artifact When a patient contracts a muscle, the surface electrode may slide over the skin.
Electronic Systems and Data Analysis
51
Figure 3-6. Amplifier noise superficially resembling positive sharp waves. Both traces were recorded with a disposable monopolar needle placed in the edematous subcutaneous tissue. The baseline thickness changed abruptly with slight relocation of the needle tip, probably altering the impedance, which is high when in contact with fatty tissue (top) and low when it is located elsewhere (bottom).
This causes a movement artifact primarily because of the change in impedance between the surface electrode and the skin. Movement-induced potentials also may result from existing fields near the surface of the skin, particularly those originating from sweat glands.73 Movement of electrode wires may produce artifacts resembling muscle activity, mostly reflecting changing capacitance. Rubbing the lead of the needle electrode with a finger or cloth sometimes produces friction artifacts from a static charge. Adequate insulation of the needle, ideally with the use of driven shields, reduces this type of interference.
Electrostatic and Electromagnetic Interference Sources of 50 or 60 Hz interference abound (Fig 3-4). They include electric fans, lamps,
fluorescent lights, cathode-ray tube screens, electric motors, light dimmers, and even unused power cords plugged into wall outlets. The use of an ungrounded wheelchair or metal examining table enhances this type of artifact. Appliances sharing the same circuit with the electromyographic instrument cause especially noticeable interference. Radio frequency electromagnetic waves can also "carry" alternating current. A strong field from a nearby diathermy apparatus produces a characteristc wave pattern. Federal regulations now restrict the amount of interference that such a unit can render to other equipment. Intermittent powerline load causes power-line voltage transient changes, which in turn give rise to artifact. In an examining room located near a driveway, auto ignition causes a popping sound. The examiner, if not properly grounded, acts as an antenna by touching the needle.
52
Basics of Electrodiagnosis
Simple but effective measures to reduce electromagnetic interference include bundling or weaving the lead wires from the pickup and ground electrodes to minimize the area susceptible to the field of interference, relocating them relative to the patient or recording apparatus, and repositioning the patient and recording apparatus within the room. With power cords near the patient, turning off power to the offending appliance does not necessarily eliminate the artifacts. To avoid interference pickup from a cathode-ray tube screen or monitor, the patient and operater should not come too close to the source. If these simple precautions fail, adequate control may require removing all electrical appliances from a room and shielding the examining area.
Radio and Mobile Phone Interference High-frequency interference or audio interference may appear on the screen of the oscilloscope from radio broadcasts, television, or radio paging systems. This type of artifact may escape detection because of its transient nature unless the sounds heard through the loudspeaker alert the examiner. Their elimination may require relocation or screening of the electromyographic instrument. An examining room located on the side of the building farthest from transmitting antennas has least interference. Screening the noise caused by power wiring may require the use of power-line radio frequency filters to remove the effect. A mobile phone in use near the electromyographic laboratory also can give rise to substantial artifacts, which may mimic high-frequency complex repetitive discharges.75 7 STIMULATORS
Electrical Stimulation Requirements Electrical stimulation of the nerve provides a clearly defined, reproducible response for nerve conduction studies. A po-
tential applied to electrodes, usually on the skin surface but sometimes inserted subcutaneously, induces a current of short duration, 50-1000 ms, in the fluid surrounding a nerve bundle. The stimulating current, directed primarily along the course, depolarizes the nerve under the cathode and hyperpolarizes it under the anode. Increasing the current to obtain a repeatable and maximal recorded response assures that essentially every nerve fiber in the bundle discharges. Surface electrode stimulation requires 50500 V to drive currents of 5-50 mA, depending on skin impedance. Higher shock intensities can usually, though not always, compensate for decreased nerve excitability seen in some neuropathic conditions (see Chapter 7-5, Chapter 25-3). Stimulation with subcutaneus needle electrodes, already in good fluid contact and closer to the nerve, uses much less intensity for adequate activation. Just a few volts may elicit a response in this case, requiring much tighter electrical control over stimulus values for consistent and safe practice than in the case of surface delivery. Effective depolarization displays an inverse relationship between stimulus intensity and duration. Thus, a lower intensity suffices if applied for a longer duration, but within limits. Generally, patients tolerate stimulus durations exceeding 1000 ms poorly. With durations of less than 50 mS, tissue capacitances limit the rate of rise, preventing the stimulus from reaching a full effective amplitude. The equipment must also provide control and timing of the stimuli for different types of measurements. In performing the paired-shock technique, the first shock with reduced intensity may subliminally excite the motor neuron pool, which then fires with the second shock delivered within a few milliseconds. Some collision techniques use two or three precisely timed stimuli, with individually adjustable intensities, durations, and latencies, delivered to the same or to different sets of electrodes. Train stimulus techniques deliver many shocks of identical intensity at rapid, adjustable rates of discharge. Such complex stimulus generators must have adequate programmability with fail-safe protection features.
Electronic Systems and Data Analysis
Stimulus Isolation Electrical stimulators are "isolated" from the recording amplifiers and other equipment circuits, for safety and artifact reduction. Thus, the stimulation circuits have no conductive path to other circuits except through the patient's body when stimulating and recording electrodes have been applied. This isolation ensures that stimulus current flows only in the loop provided by the two stimulating electrodes. If the stimulator circuit has any connection to the recording circuit, then the stimulus current distributed in the body can divide into additional paths, causing a large stimulus artifact, amplifier overload, or even spurious stimulation at unintended sites. Furthermore, under conditions of component failure, these additional paths might conduct hazardous levels of current. Stimulus isolation usually relies on magnetic coupling of energy to the stimulating circuits, although battery-powered stimulators may use optical coupling of the control signal.
Constant-Voltage versus Constant-Current "Constant-voltage" stimulators deliver an adjustable voltage across the stimulating electrodes, essentially independent of stimulus current. Adjusting the voltage varies the current through the stimulating electrodes to achieve a desired level of stimulation. At a fixed output voltage, changes in stimulating electrode impedance alter the stimulus current level. "Constantcurrent" stimulators deliver an adjustable current through the stimulating electrodes, essentially independent of their impedance. The voltage across the stimulating electrodes adjusts dynamically to maintain a constant stimulus current. Constant-current stimulators provide more consistent stimulus control, especially for techniques that require a train of stimuli or response averaging.
Magnetic Coil Stimulation Magnetic coil stimulation serves as an alternative means of nerve activation, more
53
widely used for excitation of the central rather9,22,46-48 than peripheral nervous system. A rapidly changing magnetic field of high intensity can induce sufficiently localized current in the body fluid to cause nerve excitation (see Chapter 21-3). The apparatus consists of a handheld, doughnut- or figure 8-shaped coil and a capacitive-discharge power unit, triggered from conventional electromyography equipment. The advantages of magnetic stimulation include the capability of exciting the brain non-invasively, a lower level of pain associated with the stimulus, and the elimination of stimulus electrode application. It might seem that magnetically inducing the stimulus would provide a high degree of isolation and thus reduce stimulus artifact, but in fact, the huge coil currents and high voltages couple substantial artifact into low-level recording circuits. The major disadvantages of magnetic stimulation include a greater uncertainty and variability as to the point of stimulation, and the greater expense of the stimulating equipment. Ordinary magnetic pulse generators require a few seconds for recharge between stimuli, eliminating the possibility of closely paired or train stimuli. Despite the advent of specially constructed devices for these purposes, safety considerations preclude routine application of train stimuli to the cortex. At this time, the U.S. Food and Drug Administration (FDA) has approved magnetic stimulation for human use only in studies of the peripheral nerve. The national review board has granted permission for some limited research applications conducted in the central nervous system. 8
NORMATIVE DATA AND STATISTICS
Most neurophysiological evaluation in the clinical setting makes comparison between a patient finding and some set of normative data. Thus, the quality of such database plays an essential25role for diagnostic accuracy and yields. Established control values should accompany a description of a new technique for clinical
Basics of Electrodiagnosis
54
use, even though testing large numbers of healthy subjects is tedious.18 The compilation of normative data7,17,67-70 must conform to established principles.
Control Values Normative data comprise a set of values derived from disease-free individuals. In contrast, the term "reference" usually indicates either a normative or disease control. Patients referred to the laboratory for evaluation of clinical signs or symptoms may have "normal results." Despite values within the "normal range," these patients do not belong to a normal group. To judge some patients normal on the basis of test results for inclusion into a normative database represents a circular argument that defies its own purpose. Similarly, patients with disease or injury unrelated to the study in question cannot serve as normal subjects because the apparently unaffected limbs may have subclinical involvement, and because systemic effects of treatments may influence the test outcome. Further, the population with illness may well contain a higher proportion than normal of preexisting conditions which, even if subclinical, may affect the test outcome.
Statistical Analysis In as much as population variables conform to a bell-shaped Gaussian distribution, statistical analysis shows an identical value for mean, median, and mode. Gaussian distribution, though generally symmetrical, tends asymmetrically to the baseline at both ends, reflecting a small proportion of extreme high and low values or outlyers. These values dictate "the range," which, unlike other methods for deriving normative data, critically depends on only two individual values, the lowest and the highest, essentially disregarding all other sample data. Extreme values may represent subclinical diseases or technical errors, making the range less useful as an index of normative limits. A non-Gaussian distribution, though not ideal, can sometimes be usefully trans-
formed for statistical manipulations. For example, the natural or base 10 logarithm, or square root will render positively skewed distributions more Gaussian. The mean and standard deviations of the transformed data may then be converted back to original units to set up normative limits for clinical application. Normative limits of the Gaussian distribution are customarily set at ±2 standard deviations about the mean, which include 95.44 percent of the entire population. About 5 percent of normative values falling outside these limits represent false-positive test results, half at either end of the range. Performing multiple independent tests on a single patient increases the likelihood of finding an "abnormal" value.33,76 The overall chance equals the sum of the probabilities in each of the individual tests.67,69 If each measurement allows a 2.5 percent rate of false-positivity using 2 standard deviations as the criterion, then an examination that consists of 10 independent electrophysiological measurements has a probability of more than 1 in 5 (20%) in turning up one or more abnormal values on the basis of chance alone.
False-Positive versus False-Negative Results False-positive outcomes present a major problem for clinical application.32 In general, therefore, we prefer to err on the side of false-negativity—that is, calling more borderline abnormalities normal than the reverse. The incidence of false-positiviry will decrease with the use of a broader limits, for example, mean ±2.5 standard deviations. In this case the false-positive rate falls to about 1 percent in aggregate, at the cost of a correspondingly higher falsenegativity rate.24 Excessive overlap between normative data and diseasereference values precludes the use of a broader normative range because falsenegativity increases to such an unacceptable level, so as to make the study useless. Despite considerable overlap between the two, powerful statistical tests may show a significant difference comparing, as a group, the values in normal subjects and diseased individuals. Such scientific con-
Electronic Systems and Data Analysis
55
elusions, though valid, provide only limited practical applications. In the clinical context, a single patient value must fall outside the established normative limits to declare its abnormality with reasonable confidence. Common sense must prevail in questioning an isolated borderline abnormality just outside the normal limit, a surprise result unrelated to the patient signs and symptoms, or a pattern of abnormalities inconsistent with each other and the clinical signs and symptoms. Unexpected findings that make little sense call for reevaluation of the patient, scrutinizing possible errors in the interpretation of clinical or electrophysiological data (or both) in an effort to resolve discrepancy.
nosis, or KANDID, runs on an IBMcompatible PC and assists clinical neurophysiologists during their examinations. The system processes the data in two steps: it converts raw data into a pathophysiological statement, and then matches this statement to a disorder knowledge base. To maintain an iterative cycle of planning, testing, and diagnosing, the clinician must provide data of sufficient quality and decide when to stop the electrodiagnostic examination. A prospective European multicenter field trial tested the validity of36 KANDID at seven independent laboratories. The agreement level among nine clinical neurophysiologists who participated in 159 electrodiagnostic examinations averaged 81 percent for pathophysiological conclusions and 61 percent for diagnostic categories. The pronounced inter-examiner variation reflected regional differences in epidemiology, examination techniques, reference values, interpretations and planning strategies.
9 EXPERT SYSTEMS AND QUALITY DEVELOPMENT Electromyographers face difficult challenges in considering a vast amount of constantly increasing knowledge in electrodiagnostic medicine. Computer-based methodology has helped the development of automated expert systems for use in some electrodiagnostic assessments. This type of automated analysis may complement the routine laboratory procedures, aiding the less-experienced examiner in time-efficient detection of abnormalities. Various expert systems, although still in the developmental stage, may eventually provide quick access to pertinent information that facilitates the decision-making process. The use of such a device can reduce interlaboratory variation, which results from differences in the quality of training and technical preference of investigators. This approach also helps standardize physiologic evaluations in formulating a diagnostic impression. Adherence to acceptable practice guidelines of electrodiagnosis ensures better quality control, which plays an essential role in the effective operation of an expert system.34 KANDID One such system, Knowledge Based Assistant for Neuromuscular Disorder Diag-
ESTEEM The experience with KANDID led to a multicenter project called European Standardized Telematic Tool to Evaluate EMG Knowledge Based Systems and Methods, or ESTEEM. This project used a multicenter database of neuromuscular cases to obtain diagnostic consensus by expert electromyographers and establish standards and guidelines of electrodiagnostic practice to develop an acceptable expert system. ESTEEM also served as a prototype for an electrophysiology platform that integrated different tools within the laboratory and for telematically communicated pertinent data at various posts within one hospital and also among different institutions Studies in 81 patients from the ESTEEM database established the degree of observer variation in interpreting individual tests. Despite a good overall agreement among physicians who assessed 735 muscles and 726 nerve segments, a considerable disagreement emerged in determining specific pathophysiology in general and in diagnosing demyelination in particular. For the consensus procedure of ESTEEM, the
Basics of Electrodiagnosis
56
moderator discarded all of the information except for electrodiagnostic data and related reference values.77 The selected experts then interpreted the data in each case with respect to pathophysiological conclusions and overall diagnosis. The experts must agree with the diagnosis before transferring the case to the consensus database. If not, the diagnosis given by the majority went back to the minority for a second interpretation, and when necessary, a panel discussion, leading to a consensus for nearly all cases. MUNIN Another EMG expert system, MUNIN, uses a causal probabilistic network in contrast to the rule-based 4KANDID.4,64 The microhuman prototype includes a limited "Microhuman" anatomy and a small number of nerve lesions. The system gives a detailed description of the most important groups of generalized disorders of muscle and nerve, as well as commonly used measures of electromyography and nerve conduction studies. For diagnostic purpose, a probabilistic inference engine "reasons" from test results to different aspects of pathophysiology to neuromuscular disorders. It can also provide causal reasoning in the opposite direction, from disorders to pathophysiology to expected test results. At the end of a 5-year project sponsored by the ESPRIT program, evaluation of its diagnostic performance revealed generally satisfactory results in 30 cases covering a wide range of neuromuscular disorders. The seven expert electromyographers who evaluated the system thought that MUNIN performed at a level similar to an experienced neurophysiologist.5 Compared to KANDID, MUNIN does not explicitly help in the planning of an examination. If such an interaction were available, the probabilities provided by the system would help direct the physician toward a proper course of action. Compared to 39 percent disagreement for KANDID in 159 cases collected in a field trial, electromyographers expressed no serious discrepancies between MUNIN and the majority opinion in any of the 11 cases
evaluated by peer review. MUNIN utilizes very few clinical findings. Thus, the system does not accept the cases with limited EMG examination performed only to confirm a clinical diagnosis. Methodological and population differences make it difficult, if not impossible, to compare MUNIN and KANDID regarding their diagnostic accuracy and dependability. Interlaboratory Communication The diversity of electrodiagnostic practices necessitates studying the differences between various existing techniques. For example, some physicians use quantitative muscle examination and near-nerve technique for nerve conduction studies, whereas others use qualitative muscle examination and surface electrodes. To improve the quality of studies, expert systems must consider these widely variable patterns of practices,35 and standardize terminology for pathophysiological interpretations and diagnoses. To facilitate interaction among different laboratories via the Internet, the ESTEEM project developed an electromyography communication protocol.51 It consists of general data, examination techniques, reference values, pathophysiological conclusions, and diagnoses. Its implementation of several computer programs allows an exchange of data among laboratories despite the use of different techniques and reference values. This consensus database may help develop an expert system, which integrates all tools concerned and generates a report independent of specific instrument and telematic programs.78 REFERENCES 1. Ackmann JJ, Lomas JN, Hoffmann RG, Wertsch J: Multifrequency characteristics of disposable and nondisposable EMG needle electrodes. Muscle Nerve 16:616-623, 1993. 2. Adrian ED, Bronk DW: The discharge of impulses in motor nerve fibres. Part II. The frequency of discharge in reflex and voluntary contractions. J Physiol (Lond) 67:119-151, 1929. 3. American Association Electrodiagnostic Medicine News and Comments: FDA Public health
Electronic Systems and Data Analysis Advisory: Unsafe electrode lead wires and patient cables used with medical devices. Muscle Nerve 17:565-566, 1994. 4. Andreassen S, Falk B, Olesen KG: Diagnostic function of the microhuman prototype of the expert system—MUNIN. Electroencephalogr Clin Neurophysiol 85:143-157, 1992. 5. Andreassen S, Rosenfalck A, Falck B, Olesen KG, Andersen SK: Evaluation of the diagnostic performance of the expert EMG assistant MUNIN. Electroencephalogr Clin Neurophysiol 101:129144, 1996. 6. Ashley RA, Wee AS: Construction of a simple multi-lead electrode for intraoperative nerve recording. J Neurosurg 70:962-964, 1989. 7. Bailer JC III, Mosteller F: Medical Uses of Statistics. Massachusetts Medical Society, 1986, Waltham, pp 160-162. 8. Baringer JR, Gajdusek DC, Gibbs CJ Jr, Masters CL, Stem WE, Terry RD: Transmissible dementias: Current problems in tissue handling. Neurology (New York) 30:302-303, 1980. 9. Barker AT, Freestone IL, Jalinous T, Merton PA, Morton HB: Magnetic stimulation of the human brain. J Physiol 369:3P, 1985. 10. Barkhaus PE, Nandedkar SD: Recording characteristics of the surface EMG electrodes. Muscle Nerve 17:1317-1323, 1994. 11. Barohn RJ, Jackson CE, Adams KK, Gronseth GS: Letters to the Editor: An electrophysiologic comparison of stainless steel and adhesive recording electrodes for nerve conduction studies. Muscle Nerve 18:1218-1219, 1995. 12. Barry DT: AAEM Minimonograph #36: Basic concepts of electricity and electronics in clinical electromyography. Muscle Nerve 14:937-946, 1991. 13. Basmajian JV: Muscles Alive: Their Functions Revealed by Electromyography, ed 4. Williams & Wilkins, Baltimore, 1978. 14. Beranek R: Intracellular stimulation myography in man. Electroencephalogr Clin Neurophysiol 16:301-304, 1964. 15. Buchthal F, Guld C, Rosenfalck P: Action potential parameters in normal human muscle and their dependence on physical variables. Acta Physiol Scand 32:200-218, 1954. 16. Buchthal F, Guld C, Rosenfalck P: Volume conduction of the spike of the motor unit potential investigated with a new type of multielectrode. Acta Physiol Scand 38:331-354, 1957. 17. Campbell WW, Robinson LR: Deriving reference values in electrodiagnostic medicine. Muscle Nerve 16:424-428, 1993. 18. Campbell MJ, Machin D: Medical Statistics. Wiley-Liss, New York, 1993. Council on Ethical and Judicial Affairs: Code of Medical Ethics. American Medical Association, 1994 Edition. 19. Chu J, Chan RC: Changes in motor unit action potential parameters in monopolar recordings related to filter settings of the EMG ampliflr. Arch Phys Med Rehabil 66:601-604, 1985. 20. Chu J, Chan RC, Bruyninckx F: Effects of the EMG amplifier filter settings on the motor unit potential paramaters recorded with concentric and monopolar needles. Electromyogr Clin Neurophysiol 26:627-639, 1986.
57 21. Chu J, Chan RC, Bruyninckx F: Progressive Teflon denudation of the monopolar needle: effects of motor unit potential parameters. Arch Phys Med Rehabil 68:36-40, 1987. 22. Day BL, Dick JPR, Marsden CD, Thompson PD: Differences between electrical and magnetic stimulation of the human brain. J Physiol 378: 36P, 1986. 23. Dorfman LJ, McGill KG, Cummins KL: Electrical properties of commercial concentric EMG electrodes. Muscle Nerve 8:1-8, 1985. 24. Dorfman LJ: Pitfalls in the interpretation of normative data in nerve conduction studies. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996. 25. Dorfman LJ, Robinson LR: AAEM Minimonograph #47: Normative data in electrodiagnostic medicine. Muscle Nerve 20:4-14, 1997. 26. Downey JM, Belandres PV, Di Benedetto M: Suction cup ground and reference electrodes in electrodiagnosis. Arch Phys Med Rehabil 70:64-66, 1989. 27. Dumitru D, King JC: Concentric needle recording characteristics related to depth of tissue penetration. EEG Clin Neurophysiol 109:124-134, 1998. 28. Dumitru D, King JC, Nandedkar SD: Motor unit action potentials recroded with concentric electrodes: physiologic implications. Electroencephalogr Clin Neurophysiol 105:333-339, 1997. 29. Dumitru D, King JC, Nandedkar SD: Concentric/monopolar needle electrode modeling: spatial recording territory and physiologic implications. Electroencephalogr Clin Neurophysiol 105:370-378, 1997. 30. Ekstedt J, Stalberg E: A method of recording extracellular action potentials of single muscle fibres and measuring their propagation velocity in voluntarily activated human muscle. Bull Am Assoc Electromyogr Electrodiagn 10:16, 1963. 31. Ekstedt J, Stalberg E: How the size of the needle electrode leading-off surface influences the shape of the single muscle fibre action potential in electromyography. Computer Prog Biomed 3:204-212, 1973. 32. Fagan TJ: Nomogram for Bayes' theorem. N Engl J Med 293:257, 1996. 33. Files JB, van Peenen HJ, Lindberg DAB: Use of 'normal range' in multiphasic testing. JAMA 205:94-98, 1968. 34. Finnerup NB, Johnsen B, Fuglsang-Frederiksen A, de Carvalho M, Fawcett P, Liguori R, Nix W, Schofleld I, Vila A: Can medical audit change electromyographic practice? EEG Clin Neurophysiol 109:496-501, 1998. 35. Fuglsang-Frederiksen A, Johnsen B, Vingtoft S, and the Clinical ESTEEM Group: Expert systems and quality development in electromyography. In Kimura, J and Shibasaki, H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science B.V., Amsterdam, pp 378-383, 1996. 36. Fuglsang-Frederiksen A, Ronager J, Vingtoft S: PC-KANDID: An expert system for electromyography. Artif Intell Med 1:1171-124, 1989. 37. Gajdusek DC, Gibbs JC Jr, Asher DM, Brown P, Diwan A, Hoffman P, Nemo G, Rohwer R,
Basics of Electrodiagnosis
58
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
White L: Precautions in medical care of, and in handling materials from, patients with transmissible virus dementia (Creutzfeldt-Jakob disease). N Engl J Med 297:1253-1258, 1977. Geddes LA, Baker LE, McGoodwin M: The relationship between electrode area and amplifier input impedance in recording muscle action potentials. Med Biol Engin 5:561-569, 1967. Giitter AJ, Stolov WC: AAEM Minimonograph #16: Instrumentation and measurement in electrodiagnostic medicine—Part I. Muscle Nerve 18:799-811, 1995. Giitter AJ, Stolov WC: AAEM Minimonograph #16: Instrumentation and measurement in electrodiagnostic medicine—Part II. Muscle Nerve 18:812-824, 1995. Green JB, Nelson AV, Michael D: Digital zerophase-shift filtering of short-latency somatosensory evoked potentials. Electroencephalogr Clin Neurophysiol 63:384-388, 1986. Guidelines in Electrodiagnostic Medicine. Professional Standards Committee, American Association of Electromyography and Electrodiagnosis, 1984. Guld C, Rosenfalck A, Willison RG: Report of the committee on EMG instrumentation. Technical factors in recording electrical activity of muscle and nerve in man. Electroencephalogr Clin Neurophysiol 28:399-413, 1970. Gydikov A, Gerilovsky L, Kostov K, Gatev P: Influence of some features of the muscle structure on the potentials of motor units, recorded by means of different types of needle electrodes. Electromyogr Clin Neurophysiol 20:299-321, 1980. Hannerz J: An electrode for recording single motor unit activity during strong muscle contractions. Electroencephalogr Clin Neurophysiol 37: 179-181, 1974. Hess CW, Mills KR, Murray WF: Percutaneous stimulation of the human brain: A comparison of electrical and magnetic stimuli. J Physiol 378:35P, 1986. Hess CW, Mills KR, Murray WF: Methodological considerations for magnetic brain stimulation. In Barber C, Blum T (eds): Evoked Potentials. III. The Third International Evoked Potential Symposium, Butterworths, London, 1987. Hess CW, Mills KR, Murray WF: Responses in small hand muscles from magnetic stimulation of the human brain. J Physiol 338:397-419, 1987. Howard JE, McGill KC, Dorftnan LJ: Properties of motor unit action potentials recorded with concentric and monopolar needle electrodes: ADEMG analysis. Muscle Nerve 11:1051-1055, 1988. Jasper H, Notman R: Electromyography in peripheral nerve injuries. National Research Council of Canada, Report C6121, N.R.C. Grant No. Army Med. 28 From the Montreal Neurological Institute, Vol IV. McGill University, Montreal Quebec, 1944. Johnsen B, Vingtoft S, Fuglsang-Frederiksen A, Barahona P, Fawcett P, Akobsen L, Liguori R, Nix W, Otte G, Schofield I, Sieben G, Talbot A, Veloso M, Vila A: A common structure for the
52. 53.
54. 55.
56.
57.
58.
59. 60. 61. 62.
63.
64.
65. 66. 67. 68.
69.
representation of data and diagnostic processes within clinical neurophysiology. In Barahona P, Velosa M, Bryant J (eds): Proceedings of the Twelfth International Congress of the European Federation for Medical Information MIF, 1994, pp 150-155. Joynt RL: The case for the monopolar needles: Muscle Nerve 21:1804-1806, 1998. King JC, Dumitru D, Nandedkar SD: Concentric and single fiber electrode spatial recording characteristics. Muscle Nerve 20:1525-1533, 1997. King JC, Dumitru D, Stegeman D: Monopolar needle electrode spatial recording characteristics. Muscle Nerve 19:1310-1319, 1996. Kohara N, Kaji R, Kimura J: Comparison of recording characteristics of monopolar and concentric needle electrodes. Electroencephalogr Clin Neurophysiol 89:242-246, 1993. Maccabee P, Hassan N, Cracco R, Schiff J: Short latency somatosensory and spinal evoked potentials: power spectra and comparison between high pass analog and digit filter. Electroencephalogr Clin Neurophysiol 65:177-187, 1986. Maccabee PJ, Hassan NF: AAEM minimonograph #39: Digital filtering: Basic concepts and application to evoked potentials. Muscle Nerve 15:865-875, 1992. Manuelidis EE, Angelo JN, Gorgacz EJ, Kim HA, Manuelidis L: Experimental Creutzfeldt-Jakob disease transmitted via the eye with infected cornea. N Engl J Med 296:1334-1336, 1977. Mikolich LM, Waylonis GW: Durability of monopolar Teflon-coated electromyographic needles. Arch Phys Med Rehabil 58:448-451, 1977. Misulis KE: Basic Electronics for Clinical Neurophysiology. J Clin Neurophysiol 6:41-74, 1989. Nandedkar SD, Sanders DB: Recording characteristics of monopolar EMG electrodes. Muscle Nerve 14:108-112, 1991. Nandedkar SD, Tedman B, Sanders DB: Recording and physical characteristics of disposable concentric needle EMG electrodes. Muscle Nerve 13:909-914, 1990. Nissen-Petersen H, Guld C, Buchthal F: A delay line to record random action potentials. Electroencephalogr Clin Neurophysiol 26:100-106, 1969. Olesen KG, Kjaerulff U, Jensen F, Jensen FV, Falck B, Andreassen S, Andersen SK: MUNIN network for the median nerve—a case study on loops. Appl Artif Intell 3:385-403, 1989. Pease WS, Bowyer BL: Motor unit analysis. Am J Phys Med Rehabil 67:2-6, 1988. Pease WS, Pitzer NL: Electronic filter effects on normal motor and sensory nerve conduction tests. Am J Phys Med Rehabil 69:28-31, 1990. Rimer MH: Statistical errors and their effect on electrodiagnostic medicine. Muscle Nerve 17: 811-814, 1994. Robinson LR, Temkin NR, Fujimoto WY, Stolov WC: Effect of statistical methodology on normal limits in nerve conduction studies. Muscle Nerve 14:1084-1090, 1991. Schoen I, Brooks SH: Judgment based on 95% confidence limits: a statistical dilemma involving multitest screening and proficiency testing
Electronic Systems and Data Analysis
70. 71.
72. 73. 74. 75. 76. 77.
of multiple specimens. J Clin Pathol 53:190193, 1970. Schulzer M: Diagnostic tests: a statistical review. Muscle Nerve 17:815-819, 1994. Sherman HB, Walker FO, Donofrio PD: Sensitivity for detecting fibrillation potentials: A comparison between concentric and monopolar needle electrodes. Muscle Nerve 13:1023-1026, 1990. Stalberg E, Trontelj JV: Single Fibre Electromyography. Raven Press, New York, 1994. Tarn HW, Webster JG: Minimizing electrode motion artifact by skin abrasion. IEEE Trans Biomed Engin 24:134-139, 1977. Trojaborg W: The case for the concentric needles: Muscle Nerve 21:1806-1808, 1998. Uludag B, Koklu F, On A: Letters to the editor: A new source of electromyographic artifact: Mobile phones. Muscle Nerve 121-122, 1997. Van Dijk JG: Multiple tests and diagnostic validity. Muscle Nerve 18:353-355, 1995. Veloso M, Vingtoft S, Fuglsang-Frederiksen A, Johnsen B, Fawcett P, Liguori R, Nix W, Otte G, Schofield I, Sieben G, Vila A, Sales-Luis M: Qual-
59
78.
79.
80. 81.
ity assurance in clinical neurophysiology: the ESTEEM project example. In Gordon, C and Christensen, JP (eds): Health Telematics for Clinical Guidelines and Protocols. IOS Press, Amsterdam, 1995, pp 125-133. Vingtoft S, Johnsen B, Fuglsang-Frederiksen A, Veloso M, Barahona P, Vila A, Fawcett P, Schofield I, Landegaard J, Otto G, Sieben G, Talbot A, Liguori R, Nix W. ESTEEM—a European telematic project for quality assurance within clinical neurophysiology. In Greene RA, Peterson HE, Protti DJ (eds): MEDINFO'95 IMIA, 1995, pp 1047-1051. Wee AS, Ashley RA: Relationship between the size of the recording electrodes and morphology of the compound muscle action potentials. Electromyogr Clin Neurophysiol 30:165-168, 1990. Wiechers DO, Blood JR, Stow RW: EMG needle electrodes: Electrical impedance. Arch Phys Med Rehabil 60:364-369, 1979. Yaar I: Computing normative ranges without recruiting normal subjects. Muscle Nerve 20: 1510-1514, 1997.
This page intentionally left blank
Part II NERVE CONDUCTION STUDIES
This page intentionally left blank
Chapter 4 ANATOMY AND PHYSIOLOGY OF THE PERIPHERAL NERVE
1. INTRODUCTION 2. ANATOMY OF PERIPHERAL NERVES Gross Anatomy Myelinated and Unmyelinated Fibers Axonal Transport 3. PHYSIOLOGY OF NERVE CONDUCTION Transmembrane Potential Generation and Propagation of Action Potential Factors Determining the Conduction Velocity 4. TYPES OF NERVE FIBERS AND IN VITRO RECORDING Classification of Nerve Fibers Modality Dependency of Nerve Conduction Velocity In Vitro Recording and Fiber Diameter Histogram of the Sural Nerve Analysis of Compound Nerve Action Potentials 5. CLASSIFICATION OF NERVE INJURIES Neurapraxia Axonotmesis Neurotmesis 6. INVOLVEMENT OF AXON VERSUS MYELIN IN NEUROPATHIC DISORDERS Axonal Degeneration Segmental Demyelination in Animal Models Pathophysiology of Demyelination Clinical Consequences of Demyelination Types of Abnormalities in the Clinical Domain
1 INTRODUCTION Histologic techniques have advanced our understanding of peripheral nerve pathology, especially through quantitative analysis of fiber diameter spectrum and single teased fiber preparations. Electrophysio-
logic methods have made equally important contributions in elucidating the pathophysiology of peripheral nerve disorders.32 In particular, in vitro recordings of compound nerve action potentials from the sural nerve have delineated the types of fibers predominantly affected in certain neuropathic processes. These studies also 63
Nerve Conduction Studies
64
demonstrated the close relationships between histologic and physiologic findings in many disease entities. Traumatic lesions of the nerve usually result in structural changes in the axon with or without separation of its supporting connective tissue sheath.45 Nontraumatic disorders of the peripheral nerve may affect the cell body, axon, Schwann cell, connective tissue, or vascular supply, singly or in combination. Electrophysiologic abnormalities depend on the kind and degree of nerve damage. Hence, the results of nerve conduction studies closely parallel the structural abnormalities of the nerve. Histologic changes in the nerve and the nature of conduction abnormalities allow subdivision of peripheral nerve lesions into two principal types: axonal degeneration and segmental demyelination. This chapter will deal with the basic anatomy and physiology of the peripheral nerve, and discuss types of conduction abnormalities. A number of excellent texts provide a more detailed review of the subject for interested readers. 56,183,199 chapters 23 through 26 will present the clinical aspects of peripheral nerve disorders.
2
ANATOMY OF PERIPHERAL NERVES Gross Anatomy
Nerves have a structure of considerable complexity and features of special relevance to nerve injury and regeneration.201 Three kinds of connective tissue—endoneurium, perineurium, and epineurium—surround the axons in the nerve trunks (Fig. 4-1). The endoneurium forms the supporting structure found around individual axons within each fascicle. The perineurium consists of collagenous tissue, which binds each fascicle with elastic fibers and mesothelial cells. This layer serves neither as a connective tissue nor as a simple supporting structure; rather, it provides a diffusion barrier to regulate intrafascicular fluid.209 Fascicular groups destined for the same endpoint remain localized within the nerve for long distances.218 The
epineurium, composed of collagen tissue, elastic fibers, and fatty tissue, tightly binds individual fascicles together providing a protective cushion against compression.201 This outermost layer of supporting structure for the peripheral nerve merges83 in the dura mater of the spinal roots. Paucity of endoneurial collagen at the roots as compared with the nerve trunk may explain why some disease processes selectively involve the root. The vasa nervorum, located in the epineurium, branch into arterioles that penetrate the perineurium to form capillary anastomoses in the fascicles. The perineurium probably acts as a blood-nerve barrier, but the elucidation of its detailed function needs further study.
Myelinated and Unmyelinated Fibers The nerve trunks contain myelinated and unmyelinated fibers. Certain inherent properties of the axon apparently determine whether or not myelination will eventually occur. In myelinated fibers, the surface membrane of a Schwann cell, or axolemma, spirals around the axon to form the myelin sheath (Fig. 4-2). Each myelinated axon has its own Schwann cell, which regulates myelin volume and thereby myelin thickness.189 The nodes of Ranvier, located at junctions between adjacent Schwann cells, represent uninsulated gaps along the myelinated fiber. In contrast, several unmyelinated axons share a single Schwann cell, which gives rise to many separate74 processes, each surrounding one axon. The spacing of the Schwann cells at the time of myelination determines the internodal distance. As the nerve grows in length, the internodal distance must increase, because Schwann cells do not proliferate. Thus, the fibers myelinated early achieve larger diameters and wider spacing between the nodes of Ranvier. In other words, the larger-diameter fibers have a greater internodal distance. In myelinated fibers, the action potentials propagate from one node of Ranvier to the next with the rate approximately proportional to the fiber diameter. In unmyelinated nerves,
Anatomy and Physiology of the Peripheral Nerve
65
Figure 4-1. Transverse (A) and longitudinal (B) sections of the sciatic nerve shown at low magnification. Vertical scales at lower right represent 20 um. In A, the epineurium (E) contains vessels, fibro-blasts, and collagen. The perineurium (P) surrounds fascicles of nerve fibers, which are separated by endoneurial connective tissue. The longitudinal section (B) includes a node of Ranvier (upper arrow), a Schwann cell nucleus (right arrowj), and Schmidt-Lantermann clefts (lower arrows). [From Webster,225 with permission].
conduction velocity varies in proportion to the square root of the fiber diameter. The largest and fastest conducting fibers include the sensory fibers transmitting proprioceptive, positional, and touch sensations and the alpha motor neurons. Small myelinated or unmyelinated fibers subserve pain and temperature sense and autonomic functions (see this chapter, part 4). Those found in the human epidermis apparently originate from nerve trunks in the dermis, subserving some sensory function.109 Axonal Transport In the peripheral nervous system, a small cell body with a diameter of 50-100 um regulates axons up to 1 m in length. A
complicated system of axonal transport provides the metabolic needs of the terminal axon segments. Hence, the axons not only conduct propagating electrical potential but also actively participate in conveying nutrient and other trophic substances. The velocity of transport varies from several hundred to a few millimeters per day. The majority flows centrifugally, though some particles seem to move centripetally. Axonal transport plays a complex role in maintaining the metabolic integrity of the peripheral nerve. Axonal flow of trophic substances, at least in part, dictates the histochemical and electrophysiologic properties of the muscle fibers. No particles other than acetylcholine (ACh), however, seem to transfer across the neuromuscular junction. Therefore, acetylcholine mol-
66
Nerve Conduction Studies
Figure 4-2. Fine structures of the peripheral nerve as visualized with the light microscope (A, B, D) and as reconstructed from electron micrographs (C, E). A. The epineurium covers the entire nerve, whereas the perineurium surrounds individual fascicles and endoneurium nerve fibers. B. The myelinated fiber consists of axis cylinder, myelin sheath, and Schwann (neurilemma) cells. The myelin sheath is absent at the node of Ranvier. C. The Schwann cell produces a helically laminated myelin sheath that wraps around an axon individually. D. Several unmyelinated nerve fibers share one Schwann cell. E. Several axis cylinders of unmyelinated fibers surround the nucleus of the Schwann cell. [From Noback,145 with permission.]
ecules may have a trophic influence on muscle in addition to their role as a neurotransmitter. Separation of the axon from the cell body first results in failure of the neuromuscular junction, followed by axonal degeneration and muscle fiber atro-
phy.35,136 Neuromuscular transmission fails, and the nerve terminals degenerate faster with distal than with proximal axonotmesis. Similarly, membrane changes in denervated muscle appear more rapidly after nerve injury close to the muscle.86
Anatomy and Physiology of the Peripheral Nerve
3
PHYSIOLOGY OF NERVE CONDUCTION
Transmembrane Potential Nerve axons have electrical properties common to all excitable cells (see Chapter 2-2). Measured transmembrane steady state potentials vary from about -20 to -100 mV in different tissues, despite the same basic physiologic mechanisms underlying the phenomenon. A smaller resting membrane polarization in the soma (-70 mV) as compared to the axon (-90 mV) probably reflects a partial depolarization from continuous synaptic influences. As in any excitable element, generation of a nerve action potential consists of two steps: graded subliminal excitation caused by any externally applied stimulus and suprathreshold activation as the result of increased sodium conductance. A local subliminal change in the transmembrane potential rapidly diminishes with distance. In contrast, threshold depolarization produces an all-or-none action potential determined by the inherent nature of the cell membrane, irrespective of the type of stimulus applied.
Generation and Propagation of Action Potential With application of a weak current to a nerve, negative charges from the negative pole, or cathode (so named because it attracts cation), accumulate outside the axon membrane, making the inside of the cell relatively more positive (cathodal depolarization). Under the positive pole, or anode, the negative charges tend to leave the membrane surface, making the inside of the cell relatively more negative (anodal hyperpolarization). The cell plasma resistance and membrane conductance and capacitance limit the subliminal local changes of depolarization or hyperpolarization only within a few millimeters of the point of origin. After about 10-30 mV of depolarization, the membrane potential reaches the critical level for opening the voltage dependent sodium (Na+) channels, leading to the generation of an all-or-none
67
action potential (see Chapter 2-3). Nerve excitability change seen after a single nerve impulse has three phases: the initial refractory period of a few milliseconds, supernormality lasting 30 ms or so, and subnormality extending up to 100-200 ms (see Chapter 8-2). An action potential initiated along the course of an axon propagates in both directions from its point of origin (Fig. 4-3). Intracellular current flows from the positively charged active area to the adjacent negatively charged inactive region. An opposing current flows through the extracellular fluid from the inactive to active region, allowing the recording of electric as well as magnetic fields.85 This local current depolarizes the inactive regions on both sides of the active area. When it attains the critical level, an action potential generated there initiates a new local current further distally or proximally. Hence, the nerve volleys always propagate bidirectionally from the site of external stimulation at one point along the axon. Physiologic impulses originating at the anterior horn cells or sensory terminal travel only orthodromically. In pathological situations, however, impulses may arise in the midportion of nerve fibers. For example, discharges occur in the middle of the spinal root axons in dystrophic mice, either spontaneously or as a result of ephaptic transmission (cross-talk) from neighboring fibers.165 In isolated nervemuscle and leg preparations, electric ac-
Figure 4-3. Saltatory conduction along the myelinated fiber. The myelin sheath effectively insulates the intemodal segment with the bare axon at the node of Ranvier, where the current flows between intracellular and extracellular fluid. A local current (broken arrows) produced by an action potential at one node (open arrow) depolarizes the axis cylinder at the adjacent nodes on either side, transmitting the impulse in both directions (solid arrows). This type of saltatory excitation propagates rapidly as it jumps from one node to the next.
Nerve Conduction Studies
68
tivity of frog muscle can show ephaptic influence on contiguous nerves under various conditions.14
Factors Determining the Conduction Velocity Various factors affect the time necessary for generating action potentials, which in turn determine the conduction velocity of an axon. Rapid propagation results from (1) faster rates of action potential generation, (2) increased current flow along the axons, (3) lower depolarization thresholds of the cell membrane, and (4) higher temperature. Warming up the body facilitates activation and inactivation of sodium conductance, thereby lowering the amplitude of action potential and increasing its rate of transmission. Conduction velocity increases nearly linearly about 5 percent per 1 °C from 29° to 38 °C.89 Thus, the change ranges 2-3 m/s per °C in a normal nerve conducting at 40-60 m/s. Other elements of clinical importance (see Chapter 5-6) include internodal length,31 variation among different nerves and segments, effect of age, and 187 metabolic factors such as hyperglycemia. In the myelinated fibers, action potentials occur only at the nodes of Ranvier. This induces a local current that, in effect, jumps from one node to the next, producing saltatory conduction (Fig. 4-3) instead of the continuous propagation observed in unmyelinated fibers. An increase in internodal distance allows a longer jump of the action potential, but at the same time causes greater loss of current through the internodal membrane. Typically, it takes approximately 20 JJLS for the local current to excite the next node. The conduction velocity would then be 50 m/s for an internode distance of 1 mm. The longitudinal resistance of axoplasm tends to inhibit the flow of the local current. The capacitance and conductance of the internodal membrane also have the same effect by the loss of the current before it reaches the next node. This in turn makes the time required to depolarize the adjacent nodal membrane longer, resulting in slower conduction. Both internodal
capacitance and conductance decrease with myelin thickness. Thus, for the same axon diameter, conduction velocity increases with myelin thickness up to a certain point. For a fixed total fiber diameter, an increase in axon diameter induces two opposing factors, smaller axoplasmic resistance on the one hand and greater membrane conductance and capacitance reflecting reduced myelin thickness on the other.220 Theoretical considerations indicate that the anatomic characteristics of myelinated fibers fulfill all the conditions required for maximal conduction velocity. Demyelinated or partially remyelinated segments have an increased internodal capacitance and conductance because of their thin myelin sheath. More local current is lost by charging the capacitors and by leaking through the internodal membrane before reaching the next node of Ranvier. Failure to activate the next node results in conduction block. If the conduction resolves, the impulse propagates slowly, because the dissipated current needs more time to generate an action potential.166 Thus, demyelinated axons characteristically exhibit conduction failure, decreased velocity, and temporal dispersion.219 After segmental demyelination, smaller diameter fibers may show continuous rather than saltatory conduction if the demyelinated region has 21a sufficient number of sodium channels. Reduction in length of the adjacent internodes tends to facilitate conduction past focally demyelinated zones.221 Conduction abnormalities do not necessarily imply demyelination. They can result from toxins or anesthetic agents.30 Reduced fiber diameter by focal compression decreases the capacitance of the internodal membrane, which tends to facilitate conduction. Concomitant increases in resistance of the axoplasm, however, more than offset this effect by delaying the flow of the local current to the next node. Most mechanisms known to influence nerve conduction velocity affect the cable properties of the internodal segments. Additionally, altered characteristics of the nodal membrane itself may interfere with generation of the action potential (see Chapter 8-3).
Anatomy and Physiology of the Peripheral Nerve
4
TYPES OF NERVE FIBERS AND IN VITRO RECORDING
Classification of Nerve Fibers The compound nerve action potential elicited by supramaximal stimulation consists of several peaks, each representing a group of fibers with a different conduction velocity. Erlanger and Gasser61 in their original study of the A fibers designated successive peaks using the Greek letters, a, B, y, and 5 in order of decreasing velocity. Subsequent studies have revealed two additional components with very slow conduction velocity: B and C fibers. The mammalian peripheral nerves contain no B fibers. This designation, therefore, now indicates the preganglionic fibers in mammalian autonomic nerves. The original terminology for various peaks of the A fibers has created some confusion124; for example, referring to the initial peak as either A-alpha75 or A-beta,60 and the subsequent peak, now considered an artifact of recording, as A-gamma.75 Current practice designates the two peaks in the A potential of cutaneous nerves as A-alpha and A-delta. The three types of nerve fibers, A, B, and C, have histologically and electrophysiologically distinctive characteristics (Table 4-1). The A fibers are myelinated somatic axons, either afferent or efferent. The B fibers are myelinated efferent axons that constitute the preganglionic axons of the autonomic nerves. The unmyelinated C fibers consist of the efferent postganglionic axons of autonomic nerves and the small afferent axons of the dorsal root and peripheral nerves. Despite histologic resemblance, physiologic characteristics can differentiate B fibers from small A fibers. For instance, the B fibers lack negative afterpotentials and consequently a supernormal period of excitability after generation of an action potential. The negative spike lasts more than twice as long in B as in A fibers. The B fibers show smooth compound action potentials without discrete peaks, indicating an evenly distributed velocity spectrum. Several C fibers share a single Schwann cell, unlike
69
Table 4-1 Types of Nerve Fibers A fibers: myelinated fibers of somatic nerves Muscle nerve Afferent Group I: 12-21 um Group II: 6-12 um. Group III: 1-6 um Group IV: C fiber Efferent Alpha motor neuron Gamma motor neuron Cutaneous nerve Afferent Alpha: 6-17 um Delta: 1-6 um B fibers: myelinated preganglionic fibers of autonomic nerve C fibers: unmyelinated fibers of somatic or autonomic nerve sC fibers: efferent postganglionic fibers of autonomic nerve drC fibers: afferent fibers of the dorsal root and peripheral nerve
individually bound A or B fibers. This, and the absence of the myelin sheath, allow histologic identification of the C fibers. Physiologic features include high thresholds of activation, long spike duration, and slow conduction velocity. High-frequency stimulation of cutaneous afferents induces paresthesia attributable to hyperexcitability, followed by hypoesthesia that arises from stimulation-induced refractoriness at the central synaptic relays.29 As documented in the human median nerve, both myelinated and unmyelinated fibers show intrafascicular segregation by modality rather than random distribution.84 Despite specificity, two types of afferent input—for example, A-delta and C fibers—may interact at primary afferent level.229 Afferent fibers of the cutaneous nerves show a bimodal diameter distribution, with one component ranging between 6 and 17 um and the other between 1 and 6 um, or with the Greek letter designation, A-alpha and A-delta fibers. The muscle nerves comprise efferent and afferent A fibers. The efferent fibers consist of the axons of alpha and gamma motor neurons. In Lloyd's Roman numeral classification, the afferent fibers consist of groups I, II, and III, ranging in diameter from 12 to 21 um, from 6 to 12 um, and from 1 to 6 um, and group IV, represent-
Nerve Conduction Studies
70
ing small pain fibers. In this designation, the A-alpha fibers of cutaneous nerve correspond in size to groups I and II, the Adelta fibers to group III, and the C-fibers to group IV.
Modality Dependency of Nerve Conduction Velocity In cats and primates, muscle afferents transmit impulses at a considerably higher speed than cutaneous afferents, which in turn conduct faster than motor fibers. Thus, conduction characteristics distinguish various fiber populations in mammalian species. This relationship also holds in humans, although differences tend to be smaller. For example, direct recording from human nerves can differentiate A-alpha and A-delta peaks as shown in the intracranially recorded potentials from the169 sensory root of the trigeminal nerve.
In Vitro Recording and Fiber Diameter Histogram of the Sural Nerve An in vitro study of the sural nerve action potential complements the quantitative morphometric assessment of the excised nerve.54 The technique allows comparison between the fiber diameter spectrum and the range of conduction velocities for different components of the sensory nerve action potential. Some authors caution, however, that the sural nerve may 4occasionally contain some motor fibers. The
nerve biopsy consists of dissecting a bundle of several fascicles above the lateral malleolus for 148 a total length of approximately 10 cm. The distal half serves as the specimen for histologic studies and the proximal half for in vitro electrophysiologic evaluation. The segment used for conduction studies is immediately placed in cool Tyrode's solution and transferred to a sealed chamber filled with 5 percent carbon dioxide in oxygen and saturated with water vapor. Immersing the chamber in a warm water bath helps maintain a constant temperature at 37 °C. A series of silver electrodes support the nerve under slight tension by the pull of a 0.5-0.9 gm weight attached to each end. Stimulation at the distal end of the nerve allows recording of the compound nerve action potential with a pair of wire electrodes placed 20-30 mm proximally. A monophasic waveform results with the nerve crushed between the recording electrodes following application of 0.1 percent procaine at the distal electrode (see Fig. 2-4). The compound nerve action potential recorded in vitro consists of three distinct peaks: A-alpha, A-delta, and C components with an average conduction velocity of 60 m/s, 20 m/s, and 1-2 m/s (Fig. 4-4). Each component requires different supramaximal intensity for full activation. The gradual onset of A-delta and C peaks precludes accurate calculation of the maximal conduction velocity. Figure 4-5 shows a fiber diameter histogram for the A-alpha and A-delta components. Here, the fiber diameter increases from left to right on the abscissa, thus, the first peak on the left corresponds Figure 4-4. Compound nerve action potential of a normal sural nerve recorded in vitro from an 11-year-old boy who had an above-knee amputation for osteogenic sarcoma. The arrows, from left to right, indicate A-alpha, A-delta, and C components, measuring 2.6 mV, 0.22 mV, and 70 mV in amplitude and 42 m/s, 16 m/s, and 1 m/s in conduction velocity based on the peak latency. [Courtesy of E. Peter Bosch, M.D., Mayo Clinic, Scottsdale, AZ.]
Anatomy and Physiology of the Peripheral Nerve Case I
71 Control
Figure 4-5. Myelinated fiber size-frequency histogram plotting the number of fibers with increasing diameter from left to right. The first large peak on the left corresponds to A-delta and the second smaller peak to A-alpha. Note a bimodal distribution of myelinated fiber diameter in a normal subject (control). A patient (Case 1) with familial pressure-sensitive neuropathy had an abnormal unimodal pattern with preferential loss of the larger myelinated fibers. [Courtesy of E. Peter Bosch, M.D., Mayo Clinic, Scottsdale, AZ.]
to A-delta and the second smaller peak to A-alpha fibers. In the opposite arrangement with decreasing diameter plotted from left to right, fiber groups appear in order of decreasing conduction velocity, as in the tracings of compound action potentials. In normal fiber groups, fiber diameter histograms show a continuous distribution between the large and small myelinated fibers with no clear separation between the two. Similarly, A-alpha and Adelta peaks reflects a high concentration of fibers within the continuous spectrum.15 The largest fibers, with a diameter close to 12 um, conduct at an approximate rate of 60 m/s, indicating a 5:1 ratio between the two measurements. Morphological evaluation of the peripheral nerve must take into account the106,204 maturational and age related changes.
In one study of 51 normal sural nerve biopsies,179 the fiber diameter histogram changed gradually from unimodal to bimodal distribution between 7 and 13 months. Cross-sectional measurements showed a growth in diameter of the thickest fibers, an increase in peak of the larger fiber group, and separation between the smaller and the larger groups until the beginning of adult life. With age, total transverse fascicular area increased in the face of a stable number of nerve fibers, thus decreasing fiber density. Determining the internodal length spectra in teased fiber preparation also provides quantitative data in elucidating distribution of histologic abnormalities (Fig. 4-6).94 Statistical analyses show significant correlations between teased fiber changes and conduction abnormalities affecting both motor
Nerve Conduction Studies
72
CaseI
Control
Figure 4-6. Internodal length spectra of the same nerves shown in Figure 4-5. Each vertical line indicates internodal lengths measured on a given myelinated fiber. The marked variability of internodal length in the patient reflects the effects of chronic demyelination and remyelination. [Courtesy of E. Peter Bosch, M.D., Mayo Clinic, \Scottsdale, AZ.]
and sensory nerves in patients18 with sensory motor polyneuropathies.
Analysis of Compound Nerve Action Potentials The amplitude of a compound action potential, E, recorded over the surface of a nerve increases in proportion to current flow and external resistance. Ohm's law expresses this as E = IR, where I is current and R, resistance. Larger nerves have a greater number of fibers that would collectively generate larger currents, with each fiber contributing an approximately equal amount. Nerves with greater crosssectional areas, however, have a smaller total resistance. Large nerve size, therefore, may have a negligible overall effect on amplitude. In fact, a whole nerve composed of many fascicles does not necessarily give rise to an action potential larger than the one recorded from a single dissected fascicle.124 More current flows as the nerve fibers increase in number whereas the resistance falls in proportion to the square diameter of the nerve. Thus, fiber density or the number of fibers per unit cross-
sectional area determines the amplitude of an action potential. The factors that determine the waveform abnormality of a compound action potential include the magnitude of conduction block, diminution of current in individual nerve fibers and the degree of temporal dispersion. Selective involvement of different groups of fibers results in a major distortion of the recorded potential. In contrast, uniform involvement of all fibers reduces the amplitude with relative preservation of all the components. Hence, waveform analysis of compound nerve action potentials provides a means to assess fiber density and distribution spectrum (see Chapter 7-5). 5
CLASSIFICATION OF NERVE INJURIES
Seddon183 defined three degrees of nerve injury: neurapraxia, axonotmesis, and neurotmesis. In neurapraxia, or conduction loss without structural change of the axon, recovery takes place within days or weeks, following the removal of the cause. The conduction velocity, if initially slowed because of associated demyelination, re-
Anatomy and Physiology of the Peripheral Nerve turns to normal with remyelination. In axonotmesis, the axons lose continuity with subsequent wallerian degeneration along the distal segment. Recovery depends on regeneration of nerve fibers, a process that takes place slowly over months or years at a rate of 1-3 mm per day. In neurotmesis, an injury separates the entire nerve, including the supporting connective tissue. Without surgical intervention, regeneration proceeds slowly, resulting in an incomplete and poorly organized repair. This classification, originally proposed for external trauma such as superficial or penetrating nerve injuries, also applies to compression neuropathies such as carpal tunnel syndrome or tardy ulnar palsy. Neurapraxia The mildest form of nerve block results from local injection of procaine or the transient loss of circulation—for example, with leg crossing. These insults are immediately reversible and cause no structural changes of the axon. Tetrodotoxin has similar but more widespread reversible effects on myelinated nerve fibers throughout the entire length of the axon. It acts by lowering the conductance of sodium currents at the nodes of Ranvier.154 During transient paralysis experimentally induced in humans by an inflated cuff around the arm, conduction velocity may fall by as much as 30 percent. A complete conduction block usually occurs after 25-30 minutes of compression. Serial stimulation along the course of the nerve reveals normal excitability in the segment distal to such a neurapraxic lesion. In rat sciatic nerve, transcient conduction block developed within 10 minutes after femoral artery occlusion, reached a nadir at 45-60 minutes, and then improved to normal within 24 hours.158 The fall in amplitude with less than 15 percent slowing of conduction velocity implied relative sensitivity of slower conducting myelinated fibers to the effect of acute ischemia. These short-term changes in nerve conduction probably result from anoxia secondary to ischemia.105 Intraneural microelectrode
73
recordings show spontaneous activity in afferent fibers about half a minute after reestablishment of circulation. The perceived paresthesia also suggests ectopic impulses generated along the nerve fibers previously subjected to ischemia.153 In contrast to short-term effects, chronic nerve ischemia induced by a bovine shunt, for example, usually results in axonal degeneration of sensory fibers initially, and of motor fibers later.17 In experimental animals, partial infarction resulted in degeneration of fibers in the center of the nerve with no evidence of selective fiber vulnerability.157 Hypothermia, by reducing metabolic demands, rescues the 110 nerve from ischemic fiber degeneration. Chronic nerve ischemia may also play a role in the pathogenesis of some neuropathies such as multifocal motor conduction.149 In most acute compressive neuropathies, such as a Saturday night palsy or crutch palsy of the radial nerve, conduction across the affected segment returns within a few weeks.135 A neurapraxia causing incomplete and reversible paralysis could persist for a few months or longer, usually accompanied by demyelination. Similarly the prolonged application of a tourniquet causes sustained conduction block with paranodal demyelination.150 Conduction may return immediately after decompression or at times even under the prolonged9 compression, albeit markedly slowed. Complete recovery will ensue with remyelination. The degree of compression determines the severity of the initial conduction block, but not the 91 subsequent recovery rate of conduction. Although demyelination in these cases can result from anoxia secondary to ischemia, studies of experimental acute pressure neuropathy have stressed the importance of mechanical factors76,156,173 with the initial displacement of axoplasm and myelin in opposite directions under the edges of the compressed region (Figs. 4-7 and 4-8). Part of one myelin segment invaginates the next with occlusions of the nodal gaps. Demyelination of the stretched portions of myelin follows. A patient with documented pneumatic tourniquet paralysis had severe conduction block of sensory and motor fibers localized to the presumed lower margin of the compression.19,172
Figure 4-7. Diagram showing the direction of displacement of the nodes of Ranvier in relation to the cuff placed to induce a localized mechanical compression in experimental acute neuropathy. Proximal and distal paranodes are invaginated by the adjacent one. [From Ochoa, Fowler, and Gilliatt,151 with permission.]
Figure 4-8. A. Part of a single teased fiber showing an abnormal node. B. Electron micrograph of nodal region shown in A. a, terminal myelin loops of ensheathing paranode; b, terminal myelin loops of ensheathed paranode; c, myelin fold of ensheathing paranode cut tangentially; d, Schwann cell cytoplasm; e, microvilli indicating site of Schwann cell junction. Large arrows show length of ensheathed paranode (~20 urn). [From Ochoa, Danta, Fowler et al,150 with permission].
74
Anatomy and Physiology of the Peripheral Nerve Chronic entrapment states such as carpal tunnel syndrome or tardy ulnar palsy also show focal demyelination. 112,152,227 Like acute compression, local demyelination in chronic entrapment results from mechanical forces rather than ischemia, although microscopic findings of single fibers suggest different pathophysiology in the two types. Unexpected abnormalities of nerve conduction studies in asymptomatic subjects suggest a high incidence of subclinical entrapment neuropathy. Routine autopsies in patients without known disease of the peripheral nerve have also documented unpredicted focal anatomic abnormalites.144 Patients with demyelinating neuropathy develop paralysis primarily because of conduction block rather than slowed conduction velocity. The paralyzed muscles may show fibrillation potentials and positive sharp waves followng a prolonged lack of neural influence, despite the structural 210 integrity of the axons. In one study, 25 percent of 31 patients had spontaneous discharges solely on the basis of a conduction block lasting more than 14 days. In the remaining 75 per-
75
cent, spontaneous discharges accompanied an axonal change.
Axonotmesis In this condition, axonal damage results in loss of continuity and wallerian degeneration of the distal segment followed by denervation-induced muscle atrophy (Fig. 4-9).129 Conduction ceases immediately across the site of nerve injury followed by irreversible loss of excitability, first at the neuromuscular junction, then the distal nerve segment.35,136 The time course of such degeneration varies among different species but generally not until four80or five days following acute interruption. The proximal stump also undergoes relatively mild retrograde changes. Structurally, sodium (Na+) channel reorganization follows peripheral target disconnection involving not only the cutaneous 175afferent cell bodies but also their axons. In an experimental axotomy in cats, sensory fibers degenerated more quickly than did motor fibers of similar diameter, and velocities of the fast-conducting fibers
Figure 4-9. Schematic representation of nerve axon and myelin sheath. From left to right: normal structures, wallerian degeneration following transection of the fiber, segmental demyelination, and axonal degeneration secondary to disorders of the nerve cell. [From Asbury and Johnson,8 with permission.]
Nerve Conduction Studies
76
decreased at the most rapid rate.138 Permanent axotomy in cats produced by hind limb ablation results in sequential pathologic alteration of myelinated fibers of the proximal nerve stump, namely, axonal atrophy, myelin wrinkling, nodal lengthening and internodal demyelination and remyelination.53 In the baboon, the muscle response to nerve stimulation disappears four or five days after nerve section, but an ascending nerve action potential may persist in the segment distal to the section for two or three more days.78 Preceding conduction failure, the maximal conduction velocity remains the same whether calculated by the descending motor potential or ascending nerve action potential. Histologically, degeneration develops in the terminal portion of the intramuscular nerve at a time when the proximal parts of the same fibers show relatively little change. The central stump of a transected nerve fiber, though excitable, may show reduction in nerve action potentials and conduction velocity, possibly from retraction of the myelin sheath.99,207 Transverse section at this level reveals a marked reduction in the number of myelinated fibers.6 Chronic ligation at peripheral nerves initially induces a transient, focal conduction slowing or block at the site of constriction, followed by more protracted distal effects ranging from116-118 loss of excitability to slowed conduction. A persistent nerve constriction also results in axonal atrophy and a reduction in motor conduction velocity distal to the ligature. Studies in guinea pigs suggest that atrophic nerve fibers distal to a persistent constriction may become 184 particularly sensitive to local pressure. A tight constriction of the nerve distal to the crush site also adversely influences the process of regeneration as demonstrated in cat studies using special cuff electrodes suitable for repeated studies.118 The process of regeneration accompanies the transport of structural proteins newly synthesized in the cell body to the multiple sprouts derived from the parent axon. Once the axon successfully reaches the periphery and reestablishes the physiologic connections, an orderly sequence of maturation takes place and fiber diameter
progressively increases. The remaining sprouts that fail to make functional reconnection will eventually degenerate. The Schwann cell basement membrane and the remaining connective tissues, if intact, help the nerve axon to regenerate in an orderly manner along the nerve sheath. The axons grow at a rate of approximately 1-3 mm per day, eventually restoring nearly the normal number and size of fibers.
Available data lack detailed information to precisely characterize conduction abnormalities during wallerian degeneration in humans. In one series, muscle amplitudes fell 50 percent in 3-5 days, and abated completely by the ninth day after injury. Sensory amplitudes declined 50 percent in 7 days and disappeared by the eleventh day. Shorter distal stumps showed an earlier loss of amplitude.35 In two cases, serial studies revealed loss of action potentials as early as 185 hours in one case and 168 hours in the other after traumatic transection of the digital nerve. Conduction studies showed a normal velocity during wallerian degeneration prior to the loss of recorded response.161 During the first few days after nerve injury, studies of distal nerve excitability fail to distinguish axonotmesis from neuropraxia. Finger amputation177 resulted in permanent retrograde change of the digital nerve as evidenced by a reduction in amplitude of the digital nerve potential. Histologic studies revealed a decrease in axon diameter rather than the number of nerve fibers.37,39,40 Other types of axonotmesis include nerve injuries caused by injections or tourniquets,228 sustained high intensity electric stimulation,2 and extreme cooling.1,95 Severe compressive neuropathy may at times provide the opportunity to study a single motor axon showing a discrete abnormality.142 Otherwise, different types of changes coexist in the majority of nerve injuries and neuropathies. Thus, categorizing injuries of a nerve, as opposed to individual nerve fibers, depends on less precise definition. Nonetheless, electrophysiologic studies help elucidate the extent of axonal damage. Nerve stimulation above the site of the lesion reveals a reduced amplitude in proportion to the de-
Anatomy and Physiology of the Peripheral Nerve gree of conduction loss but falls to distinguish neurapraxia from axonotmesis. In either condition, unaffected axons, if present, conduct at a normal velocity across the segment in question. Stimulation of the nerve segment distal to the site of the lesion helps differentiate the two entities. Injured axons undergo wallerian degeneration after the first few days of injury. With the loss of distal excitability, the compound action potential elicited by distal stimulation declines steeply. Electromyography shows positive sharp waves one to two weeks and fibrillation potentials two to three weeks after axonotmesis. Rarely, distal nerve inexcitability may result from a proximal nerve lesion without frank axonal degeneration based on the time course of quick recovery.130 In these cases, changes+in the number or property of sodium (Na ) channels58 and impaired axoplasmic transport below the lesion may cause the initial inexcitability of the distal axons.
Neurotmesis Sunderland199 has proposed three subdivisions of Seddon's neurotmesis. In the first type, the injury damages the axon and surrounding connective tissue, preserving the perineurium and architecture of the nerve sheath. Unlike the central nervous system pathways,71,107 the peripheral nerve regenerates effectively after this type of injury though less completely than in axonotmesis. Misdirected sprouting leads to innervation of muscle fibers previously not supplied by the nerve. The clinical phenomenon of synkinesis probably indicates an antecedent nerve injury of at least this severity.115,202 In the second type, which involves the perineurium as well, the nerve barely maintains the continuity although it may look grossly intact on inspection. Some poorly oriented regeneration may occur for 125 myelinated as well as unmyelinated axons. Surgical intervention includes 47 sensory-motor differentiated nerve repair. The last type represents a complete separation of the nerve with loss of continuity. Surgical repair consists of suturing the stumps, usually with a nerve graft to bridge the gap, and the use
77
of immunosuppression.63,134,143 If the storage of nerve grafts becomes feasible, the use of peripheral nerve allografts may provide an attractive alternative to conventional nerve autografts.62 Sometimes, nerve anastomosis, for example, from spinal accessory nerve to facial nerve may achieve a better cosmetic and functional outcome.57 Despite the advent of microsurgical techniques, functional recovery following peripheral nerve lesion remains poor.137 The poor restoration of motor function primarily reflects the limited capacity of injured axons to regenerate across the lesion site and select the appropriate target to reinnervate.181,200 The complex factors that guide regenerating axons toward appropriate terminations involve such mechanisms as neurotrophic and neurite promoting factors, chemotactic influences, and properties of the extra cellular matrix. 164,181,208 The average axon diameter in the proximal segment of a transected and reconstructed peripheral nerve will decrease shortly after the transection and increase again when the regenerating axons make contact with their targets. As some axons reach their target organs and start to mature, a number of the axons which have not reached a proper target organ will lose their conduction. This loss of signal cancels out the expected maturational increase in compound nerve action potential.121 Misdirected regeneration is the rule, not the exception, with motor axons entering any muscles in an almost random fashion, sometimes even from the 67 homologous contralateral motor nucleus. Thus, after nerve repair, especially with proximal injury, aberrant reinnervation abounds, accounting for a poor quality of functional restoration.16,196 Misdirection of motor axons after proximal section of the nerve probably accounts for the absence205of orderly recruitment of motor units. Proprioceptors and other sensory axons may also reinnervate inappropriate end organs, sometimes giving rise to abnormal connections between sensory and motor fibers.119 Misdirected axon regrowth without central adaptation leads faulty tactile digit localization.55 Regeneration may progress poorly, with frequent formation of neuroma.128 Spontaneous discharges of nerve impulses re-
78
sult from changes in membrane properties as the source of pain associated with neuromas.217,223 The accumulation of sodium channels at injured axonal tips may account for ectopic axonal hyperexcitability and the resulting pain and paresthesias.58 Chronic axonal injury may also lead to intraneuronal heterogeneity of the populations of sodium channels in cutaneous afferents, one population activating the other, leading to membrane instability and170,224 possibly to ectopic impulse generation. Following re-anastomosis of the nerve, regenerating nerve fibers increase in number and in size gradually over many years, although they regain neither the original number nor diameter.141,155 The conduction velocity increases slowly, reaching 60 percent of the normal value within 4 years90 and a mean value of 85 percent after 16 years.194 Persistent prolongation of the distal latencies suggests the presence of limited number of fibers distally. Metabolic recoveries of the crush denervated muscle follow a similar time course as the sequentially tested conduction characteristics of the damaged nerve.123 The force produced by the reinnervated muscle depends on the length 68 of time the muscle remained denervated. In detailed sequential studies of the median 27 nerve after complete section and suture, the regeneration took place at an average rate of 1.5-2.0 mm per day in three patients. The sensory potential, when first recorded 3-4 months after the injury, propagated very slowly at a velocity between 10 percent and 25 percent of normal. The conduction velocity increased 3 percent per month during the first 2 years, and 10 times slower thereafter. The tactile sensibility had returned to normal by 40 months, when the sensory potential showed a normal amplitude but increased duration five times greater than normal. Conduction velocity reached 6575 percent of normal in the adults. In children, the same degree of recovery occurred 13 to 19 months after anastomosis. The sensory potential returned five times faster after a compressive nerve lesion than after section and repair. Few studies have dealt with the neurophysiologic recovery of human peripheral
Nerve Conduction Studies nerves after10,119,146 repair with an autogenous nerve graft. In one series,203 motor and sensory nerve conduction velocities showed sustained improvement after sural nerve grafts of the ulnar and median nerves. Two years after surgery, the motor conduction velocity across the graft itself reached, in most cases, 40-50 percent of the normal values obtained in the contralateral limb. Sensory nerve action potentials returned in 44 percent of the nerves after 18 months, though greatly reduced in amplitude and conduction velocity. In another study,48 based on experience with 67 injured nerves, voluntary motor unit activity returned 7 months after repair and 12 months after grafting. Nerve stimulation elicited a compound muscle action potential by 10 months after suture and 14 months after graft. Motor unit potentials steadily increased in amplitude with time, but sensory fibers showed poor recovery both clinically and electrophysiologically. Toe-to-digit transplantation provides an excellent model for study of patterns of nerve regeneration as it pertains to the donor and recipient nerves. In one series, the transplanted toe achieved 70-90 percent recovery for temperature, pinprick, light touch, and vibration, but to a lesser extent for two-point discrimination.41 In fact, the transplanted toe behaved more like a normal toe than a normal finger with38regard to current perception threshold. Conduction studies also showed incomplete recovery in toe-to-digit transplantation as compared with digit-to-digit replantation, which resulted in almost complete repair.41 The factors responsible for different recovery may include time interval from injury to surgery, size mismatch between recipient and donor nerves, retrograde effects on the recipient nerve and severity of tissue damage. In a study of transplanted autogenous muscles, function in motor endplate was restored in about half a year with the completion213of the myelination of the grafted nerve. Long-term alterations may persist or develop after regenerated axons have established connections with their targets.24 Electrophysiologic assessments can provide clinically important information about the localization, severity, and pathophysiology
Anatomy and Physiology of the Peripheral Nerve of nerve injuries, although available methods permit only inadequate quantitation of regeneration.49,162
6
INVOLVEMENT OF AXON VERSUS MYELIN IN NEUROPATHIC DISORDERS
The preceding section has dealt with types of conduction abnormalities associated with nerve injuries. These form the basis for assessing electrophysiologic features found in other disease processes, either localized as in entrapment syndromes, or more diffuse as in polyneuropathies. Histologic96 and electrophysiologic characteristics indicate the presence of three relatively distinct categories of peripheral nerve disorders (Fig. 4-9):(1)wallerian degeneration after focal interruption of axons as in vasculitis; (2) axonal degeneration with centripetal or dying back degeneration from metabolic derangement of the neuron; and (3) segmental demyelination with slowed nerve conduction.8,100,131,133 Of these wallerian and axonal degeneration cause denervation and reduce the amplitude of compound action potential, whereas demyelination slows the nerve conduction with or without conduction block.
Axonal Degeneration Axons may degenerate in neuropathies after mechanical compression of the nerve, exposure to vibration,44,127 application of toxic substances, or death of the cell body. Nerve ischemia also causes axonal degeneration, affecting large myelinated fibers first, followed by smaller myelinated fibers and unmyelinated axons.70 The extent of abnormality varies with location of occlusion.122 Electromyography reveals normal motor unit potentials that recruit poorly during the acute stage of partial axonal degeneration. Long-duration, high amplitude polyphasic potentials appear in the chronic phase. Fibrillation potentials and positive sharp waves develop in two to three weeks after the onset of illness. Axonal degeneration, if mild, affects
79
nerve conduction only minimally, especially if the disease primarily involves the small fibers.215 More commonly, selective loss of the large, fast-conducting fibers results in reduced amplitude and slowing of conduction below the normal range especially when recorded from distal as opposed to proximal muscles.167 In these cases, the reduction in size of the compound muscle or sensory action potential shows a correlation to the degree of slowing in nerve conduction. In milder cases with the amplitude of the recorded response greater than 80 percent of the expected value, conduction velocity should remain above 80 percent of the lower limits (80% rule).7,11,26,43 A greater loss of fast conducting fibers would result in further conduction slowing but not beyond 70 percent of the lower limits of the normal value. Thus, physiologic criteria rarely misclassifies a neuropathy with predominant axon loss on biopsy as demyelinating.126 Anterior horn cell diseases can also cause selective loss of the fastest fibers. In poliomyelitis, the motor nerve conduction velocity may fall below the normal value, usually in proportion to the decrease in amplitude. For the same reason, patients with motor neuron disease have slightly reduced motor conduction velocities. Slower conduction in patients with more severe atrophy, however, may in part reflect lowered temperature of the wasted extremities (see Chapter 5-4). Neuropathies with this type of abnormality include those associated with alcoholism, uremia, polyarteritis nodosa, acute intermittent porphyria, some cases of diabetes and carcinoma, and most cases of toxicity and nutritional deficiency (see Chapter 25). Most axonal neuropathies affect both sensory and motor fibers, as exemplified by uremic neuropathies and amyloidosis. Acute intermittent porphyria and hereditory motor sensory neuropathy Type II or neuronal type of Charcot-Marie-Tooth disease (CMT2),50 however, show prominent motor abnormalities. In contrast, sensory changes predominate in the majority of toxic or metabolic polyneuropathies, Friedreich's ataxia, and some cases of carcinomatous neuropathies. Histological studies in diabetic rats revealed paranodal axonal swellings
80
and nodal bulgings of the axon during the early metabolic phase of the distal symmetrical polyneuropathy.185 These alterations correlate with intra-axonal sodium accumulation and decreased sodium equilibrium potentials that account for the early nerve conduction defect. Nerve growth factors not visualized in normal adult nerves become readily demonstrable in nerves undergoing active axonal degeneration.191 In neuropathies secondary to chronic alcoholism, carcinoma, and uremia, axonal degeneration initially involves the most terminal segment of the longest peripheral nerve fibers. Thus, studies show a slower conduction velocity over the same nerve segment if calculated based on latencies to a distal muscle as compared with a proximal muscle.167 The distal predominance of pathology and its centripetal progression led to the term dying back neuropathy. Less commonly encountered conditions associated with the dying back phenomenon include thiamine deficiency,4633 tri-orthorcresyl phosphate neuropathy, acute intermittent porphyria,34 and 73,l63,197 experimental acrylamide neuropathies. In these conditions, impaired axoplasmic flow initially affects the segment of the nerve that is furthest from the perikaryon. Thus, primary involvement of the neurons leads to axonal degeneration in the distal segment, most removed from the trophic influence of the nerve cell.192 Single-unit recording in dying back axons has confirmed the earliest failure of impulse generation in the nerve terminal when impulse still propagates normally throughout the remainder of the axon.195 In acrylamide dying back neuropathy in rats, a sequential morphometric study of the end-plate region showed the initial enlargement of the nerve terminal area distended by neurofllaments.212 The postsynaptic regions eventually became denuded as more than half of all nerve terminals subsequently disappeared. Unlike the experimental acrylamide neuropathy with clear dying back phenomenon,93,178 not all the peripheral neuropathies with a distal predominance may qualify as truly dying back in type. Selective loss of the perikarya and axons of the longest and largest fibers
Nerve Conduction Studies can produce the same pattern of abnormality.25 Distally predominant symptoms do not necessarily indicate a distal pathologic process, according to probabilistic models that reproduce a distal sensory deficit on the basis of randomly distributed axonal lesions.222 In some neuropathies, studies fail to reveal the exact site of the primary damage responsible for axonal degeneration.
Segmental Demyelination in Animal Models In the second group, disturbance of the Schwann cells causes segmental demyelination associated with unequivocal reduction of nerve conduction velocity, commonly, though not always, substantially below the normal range.28 Axonal degeneration cannot account for this degree of slowing, even with selective loss of the fast-conducting fibers leaving only the slow conducting fibers relatively intact. Experimental allergic neuritis serves as one of the most useful animal models for pathogenetic 87,193,198 studies of demyelinative neuropathies. In animal experiments, demyelination blocks the transmission of nerve impulses through the affected zone in some fibers as well as dorsal root ganglion for sensory conduction.193 The slowed conduction results primarily from delayed nerve impulses passing through the lesion and not simply from selective block of transmission in the fastconducting fibers. Focal segmental demyelination gives rise to slowed conduction locally across the demyelinated segment but not below the lesion.88 In addition to various toxins,59 injection of proteinase K to the nerve226 or removal of a small piece of perineurium in amphibian nerve160 causes a lesion consistent with demyelination. Experimental conduction block may also results from serum containing antimyelin-associated glycoprotein (MAG), IgM M protein211 or anti-GM1 antibodies.176 These antibodies could affect sodium channels64'216 at the nodes of Ranvier where GM1 abounds.42 In an experimental study on demyelination induced by diphtheria toxin, conduction velocity began to decline one week
Anatomy and Physiology of the Peripheral Nerve after inoculation, reached a plateau during the sixth to eighth weeks, and recovered to the original level between the 18th and 20th weeks.97,98 The dose of toxin administered determined the degree of slowing and the severity of paralysis. The amplitude of the compound muscle action potential predicted the loss of strength more accurately than the conduction velocity. In antiserum-mediated focal demyelination of male Wistar rats, conduction block began between 30 and 60 minutes after198 injection, and peaked within a few hours. Paralysis of the foot muscles persisted until about the seventh day, when low-amplitude, long latency muscle action potentials returned for the first time. The strength gradually recovered thereafter, reaching a normal level by 16 days. Morphologic studies revealed evidence of remyelination with two to eight myelin lamellae around each axon coincident with the onset of clinical and electrophysiologic recovery. Conduction velocities returned to pre-injection values by the 37th day, when the myelin layer of remyelinating fibers averaged only about one third that of control nerves. Serial studies of an experimentally produced demyelinating lesion in cat spinal cord188 revealed the onset of conduction block during the initial phase. Remyelination commenced in the latter part of the second week concomitant with restoration of conduction through the lesion in the affected fibers. Within three months the initially prolonged refractory period returned to normal, even though the newly formed internodes remained abnormally thin. In the adult rat spinal cord, transplantation of cultured ravelin-forming cells from the peripheral nervous system resulted in the92,214 functional repair of demyelinated axons.
Pathophysiology of Demyelination Myelin normally provides high impedance and low capacitance, preventing leakage current through the internodal membrane to sustain saltatory conduction. Action current through sodium channels at the activated node of Ranvier produces "inward ionic current," which subsequently
81
causes "outward capacitative current" at the next node to be excited. This in turn depolarizes the nodal membrane to threshold, thus opening the sodium channels and initiating another cycle of "inward ionic current." The safety factor of transmission, defined as a ratio of action current available at a node to threshold current, must exceed unity for successful conduction through a node. In the presence of paranodal or segmental demyelination, the action current dissipates through the adjacent internode as a consequence of increased capacitance and decreased impedance of the demyelinated region. Reorganization of sodium channels also plays an important role in the pathophysiology of demyelination.132,147,180 Because it takes longer to charge the next nodal membrane to the threshold, this leakage current prolongs the internodal conduction time, slowing the conduction velocity. With advanced demyelination, the safety factor eventually falls below unity, and the conduction fails because the current no longer depolarizes the next node of Ranvier beyond threshold. Raising the body temperature reduces the amplitude of action potentials, further lowering the safety factor. Thus, demyelinated nerves suffer from temperature-induced inpulse blocking more than healthy nerves.36,69 It is possible to raise the safety factor above unity by prolonging the action current using an inhibitor +of the voltagedependent potassium (K ) channel such as scorpion toxin.23 Similarly, 4-aminopyridine partially reverses symptoms in patients with multiple sclerosis,22 but not in those with inflammatory demyelinating neuropathies.13 Ion channel blockers may also exert immunomodulatory effects, which may have implications for their therapeutic139 application in neuropathic disorders. Normal nerves can transmit impulses at high rates over several hours104,171 exceeding the maximal motor unit firing frequency of 50-70 Hz under physiologic conditions. By contrast, demyelinated nerve fibers, though capable of transmitting lower-frequency impulses faithfully, may block higher-frequency discharges.120 Rate-dependent failure also characterizes other neuropathies. For example, in rats
82
with acute streptozotocin-induced diabetes, insulin therapy preserves normal nerve conduction velocity and amplitude but not under the stress 5 produced by high-frequency stimulation. This type of block should impair information coded in frequencies up to 250 Hz or more in the central nervous system and peripheral sensory nerves. The severity of the physiologic defect dictates the critical frequency for block at the site of the lesion. Important factors include fiber geometry, sodium (Na+)-potassium (K+) pump activation and ion channel density in the demyelinated segment. After an action potential, sodium accumulates in axoplasm, more so after transmission of high-frequency impulses. This increase in sodium concentration surpasses the physiologic range in a demyelinated axon, which, to compensate for the current dissipation, must sustain more action current than in normal axons. High sodium concentration in turn activates the electrogenic sodiumpotassium pump, which removes sodium in exchange for potassium at a 3 to 2 ratio, thus raising the threshold of the nodal membrane by hyperpolarization.20 A raised threshold may lower the safety factor below unity, leading to rate-dependent conduction block. One new therapeutic strategy exploits digitalis, which specifically inhibits the sodium-potassium pump,102,103 thus circumventing rate-dependent conduction block by pump activation. In animal models with demyelination, digitalis not only reversed rate-dependent block but also normalized complete conduction block. The inhibition of the pump, lowering the resting membrane potential or the threshold of activation, also explains this additional action. Another experimental study81 confirmed the beneficial action of digitalis. In this study, the combined use of 4-aminopyridine and digitalis provided a more than additive action to reverse conduction block. These experimental data provided a rationale for the use of intravenous digitalis in selected patients with multiple sclerosis.101 Despite transient improvement of the symptoms clinically and as tested by physiologic means, the use of digitalis could not serve as a general therapeutic approach because of its very modest ac-
Nerve Conduction Studies tion and possible cardiac side effects. Perhaps a digitalis derivative with better penetrance through the blood-brain barrier would render a better therapeutic effect.
Clinical Consequences of Demyelination The pathophysiology of demyelination and its clinical consequences77,104,114.190 inelude (1) elevated thresholds and conduction block resulting in clinical weakness and sensory loss; (2) increased desynchronization of volleys causing temporal dispersion of waveforms, loss of reflexes, and reduced sensation; (3) prolonged refractory periods with frequency-dependent conduction block especially at very high firing rates, accounting for reduced strength during maximal voluntary effort; (4) exaggerated hyperpolarization after the passage of impulse, inducing conduction block even at low firing frequencies causing fatigue after mild but sustained effort; and (5) steady, ectopic discharges or sporadic bursts at sites of focal demyelination considered responsible for focal myokymia and spontaneous or mechanically induced paresthesias. A complete conduction block accompanies major loss of strength. In contrast, slowing of conduction by itself leads to few, if any, clinical symptoms, as long as all the impulses arrive at the target organ. Further, a prolonged refractory period for transmission, though helpful as a diagnostic indicator,79 causes no symptoms because the time intervals between voluntarily induced repetitive discharges in motor axons substantially exceed the refractory periods under physiologic conditions. Nonetheless, the identification of demyelination by these means offers potentially important clues to conditions that may reverse by pharmacologic, immunologic, or surgical measures. Slow nerve conduction resembling demyelination, however, can result from physiological sodium channel blockage by toxins.82 Nitric oxide168 also reversibly blocks axonal conduction. Other possibilities include excitability changes of the axons during hyperventilation and ischemia.140 Common demyelinating diseases of the
Anatomy and Physiology of the Peripheral Nerve peripheral nerve include the Guillain-Barr'e syndrome, chronic inflammatory demyelinating polyradiculoneuropathy, multifocal motor neuropathy, myelomatous polyneuropathy, hereditary motor sensory neuropathy type I or the hypertrophic type of Charcot-Marie-Tooth disease, hereditary neuropathy with susceptibility to pressure palsies, metachromatic leukodystrophy,72 and Krabbe's leukodystrophy (see Chapter 25-3 and 25-5). Some cases of diabetic and carcinomatous neuropathies also belong to this category, although most paraneoplastic syndromes show axonal degeneration rather than demyelination. Diphtheritic polyneuritis no longer affects humans very often. Alterations in nerve conduction in this condition resemble those seen in animal experiments, with marked reduction in conduction velocity diffusely or, in the case of focal demyelination, over a relatively resticted region. Despite the well-established concept of segmental demyelination in experimentally induced chronic lead intoxication, nerve conduction studies in human cases 186 show either normal or only mildly slowed values.66 In some hereditary neuropathies, the demyelinative process uniformly affects the nerve throughout its length, delaying saltatory conduction more or less at all the nodes of Ranvier. By contrast, segmental conduction block in certain parts of the nerve characterizes acquired demyelinating neuropathies with non-uniform involvement. Slowing of nerve conduction then accompanies a reduction of amplitude, indicating localized neurapraxia.65 Conduction block may also occur as an early sign of105 reversible injury in ischemic neuropathy. Thus, electrophysiologic evidence of conduction block does not always imply the presence of focal demyelination. 159,182 Conversely, conventional nerve conduction studies, basically designed for assessment of the distal segments, may fail to elucidate a focal proximal demyelinating lesion in the proximal segment.111 Increased range of conduction velocity results if the disease affects smaller fibers exclusively or out of proportion to its effect on larger fibers. The evoked action potential broadens, indicating a pathologically increased tempo-
83
ral dispersion. Desynchronization of the nerve volley may also result from repetitive discharges at the site of axonal injury after the passage of a single impulse. Unless damage of the myelin sheath results in secondary axonal degeneration, electromyography reveals little or no evidence of denervation. The motor unit potentials, though normal in amplitude and waveform, recruit poorly, indicating a conduction block in severely demyelinated fibers.
Types of Abnormalities in the Clinical Domain In the arbitrary division into axonal and demyelinating neuropathies, few cases fall precisely into one group or the other. A neuropathy with extensive demyelination often accompanies slight axonal degeneration.52,206 In a study of antiserum-mediated demyelination, the inflammatory reaction could account for the axonal degeneration seen174in 5-15 percent of myelinated fibers. Conversely, axonal atrophy proximal to a neuroma or distal to constriction may cause secondary paranodal demyelination in the presence of healthy Schwann cells. Other conditions that may belong to this mixed category include neuropathies associated with diabetes, uremia, myeloma, and Friedreich's ataxia. Axonal enlargemet can also cause axon-triggered demyelination as in giant axonal neuropathy or hexacarbone intoxication (see Chapter 25-4). In some cases, the slight loss of fibers or the mild degree of demyelination demonstrated histologically cannot account for the degree of slowing seen in nerve conduction studies.12 Despite the possibility of mixed abnormalities, the electrophysiologic finding of any true axonal or demyelinating component provides an important and major contribution in the differential diagnosis. Certain conduction abnormalities support the diagnosis of a predominantly demyelinating component even when superimposed upon moderate axonal degeneration as demostrated on needle electromyography 7,11,26,43,126 These include reduction of conduction velocity below 70-80 percent of the lower limit, prologation of dis-
84
tal motor or sensory latency and F wave latency above 120 percent of the upper limit, and the presence of unequivocal conduction block.3,108 In contrast, the absence of these criteria does not necessarily preclude an early demyelinating process. In fact, a substantial number of patients with the Guillain-Barr'e syndrome have no major slowing of conduction along the nerve trunk initially. Beyond such a broad classification, electrical studies have limited value in distinguishing one variety of neuropathy from another.51 In particular, conduction studies and electromyography rarely elucidate a specific etiology. Further, clinical electrodiagnosis assists only indirectly in the differential diagnosis of neuropathic pain caused primarily by diseases of small-caliber nerve fibers, which routine methods fail to assess adequately, and in the exclusion of a large number of patients with chronic pain syndrome who probably suffer from disorders other than neuropathy. These limitations notwithstanding, conduction studies can provide diagnostically pertinent information if used judiciously in appropriate clinical contexts.113
REFERENCES 1. Aflfl AK, Kimura J, Bell WE: Hypothermiainduced reversible polyneuropathy. Electrophysiologic evidence for axonopathy. Pediatr Neurol 4:49-53, 1988. 2. Agnew WF, McCreery DB, Yuen TGH, Bullara LA: Evolution and resolution of stimulationinduced axonal injury in peripheral nerve. Muscle Nerve 22:1393-1402, 1999. 3. Albers JW, Donofrio PD, McGonagle TK: Sequential electrodiagnostic abnormalities in acute inflammatory demyelinating _polyradiculoneuropathy. Muscle Nerve 8:528 539, 1985. 4. Amoiridis G, Schols L, Ameridis N, Przuntek H: Motor fibers in the sural nerve of humans. Neurology 49:1725-1728, 1997. 5. Andersen H, Nielsen JF, Nielsen VK: Inability of insulin to maintain normal nerve function during high-frequency stimulation in diabetic rat tail nerves. Muscle Nerve 17:80-84, 1994. 6. Anderson MH, Fullerton PM, Gilliatt RW, Hern JEC: Changes in the forearm associated with median nerve compression at the wrist in the guinea-pig. J Neurol Neurosurg Psychiatry 33: 70-79, 1970. 7. Asbury AK, Cornblath DR: Assessment of current diagnostic criteria for Guillain-Barre syndrome. Ann Neurol 27(suppl):S21-S24, 1990.
Nerve Conduction Studies 8. Asbury AK, Johnson PC: Pathology of Peripheral Nerve, Vol 9. In Bennington JL (ed): Major Problems in Pathology. WB Saunders, Philadelphia, 1978. 9. Baba M, Mastsunaga M: Recovery from acute demyelinating conduction block in the presence of prolonged distal conduction delay due to peripheral nerve constriction. Electromyogr Clin Neurophysiol 24:611-617, 1984. 10. Ballantyne JP, Campbell MJ: Electrophysiological study after surgical repair of sectioned human peripheral nerves. J Neurol Neurosurg Psychiatry 36:797-805, 1973. 11. Barohn RJ, Kissel JT, Warmolts JR, Mendell JR: Chronic inflammatory demyelinating polyradiculoneuropathy: Clinical characteristics, course, and recommendations for diagnostic criteria. Arch Neurol 46:878-884, 1989. 12. Behse F, Buchthal F: Sensory action potentials and biopsy of the sural nerve in neuropathy. Brain 101:473-493, 1978. 13. Bergin PS, Miller DH, Hirsch NP, Murray NMF: Failure of 3,4-diaminophyridine to reverse conduction block in inflammatory demyelinating neuropathies. Ann Neurol 34:406-409, 1993. 14. Biro G: Ephaptic influence of the electrical activity of muscle on the neighboring nerve. Electromyogr Clin Neurophysiol 32:425-434, 1992. 15. Bishop GH: My life among the axons. In Hall VE (ed): Annual Review of Physiology, Vol 27. Annual Reviews, Inc, Palo Alto, California, 1965, pp 1-18. 16. Bodine-Fowler SC, Meyer RS, Moskovitz A, Abrams R, Botte MJ: Inaccurate projection of rat soleus motoneurons: a comparison of nerve repair techniques. Muscle Nerve 20:29-37, 1997. 17. BoltonCF, DriedgerAA, Lindsay RM: Ischaemic neuropathy in uraemic patients caused by bovine arteriovenous shunt. J Neurol Neurosurg Psychiatry 42:810-814, 1979. 18. Bolton CF, Gilbert JJ, Girvin JP, Harm A: Nerve and muscle biopsy: Electrophysiology and morphology in polyneuropathy. Neurology (New York) 29:354-362, 1979. 19. Bolton CF, McFarlane RM: Human pneumatic tourniquet paralysis. Neurology (New York) 28:787-793, 1978. 20. Bostock H, Grafe P: Activity-dependent excitability changes in normal and demyelinated rat spinal root axons. I Physiol 365:239-257, 1985. 21. Bostock H, Sears TA: Continuous conduction in demyelinated mammalian nerve fibres. Nature 263:786-787, 1976. 22. Bostock H, Sears TA, Sherratt RM: The effects of 4-aminopyridine and tetraethylammonium ions on normal and demyelinated mammalian nerve fibers. J Physiol (Lond) 313:301-315, 1981. 23. Bostock H, Sherratt RM, Sears TA: Overcoming conduction failure in demyelinated nerve fibres by prolonging action potentials. Nature 274:385-387, 1978. 24. Bowe CM, Hildebrand C, Kocsis JD, Waxman SG: Morphological and physiological properties of neurons after long-term axonal regeneration:
Anatomy and Physiology of the Peripheral Nerve
25.
26.
27.
28.
29.
30. 31.
32. 33. 34.
35.
36. 37. 38. 39. 40.
41.
Observations on chronic and delayed sequelae of peripheral nerve injury. J Neurol Sci 91:259292, 1989. Bradley WG, Thomas PK: The pathology of peripheral nerve disease. In Walton JN (ed): Disorders of Voluntary Muscle, ed 3. Churchill Livingstone, Edinburgh, 1974, pp 234-273. Bromberg MB: Comparison of electrodiagnostic criteria for primary demyelination in chronic polyneuropathy. Muscle Nerve 14:968-976, 1991. Buchthal F, Kuhl V: Nerve conduction, tactile sensibility, and the electromyogram after suture or compression of peripheral nerve: A longitudinal study in man. J Neurol Neurosurg Psychiatry 42:436-451, 1979. Buchthal F, Rosenfalck A, Behse F: Sensory potentials of normal and diseased nerves. In Dyck PJ, Thomas PK, Lambert EH, Bunge R (eds): Peripheral Neuropathy, Vol 1. WB Saunders, Philadelphia, 1984, pp 981-1015. Burke D, Applegate C: Paraesthesia and hypaesthesia following prolonged high-frequency stimulation of cutaneous afferents. Brain 112:913-929, 1989. Cameron J, Flowers AE, Capra MF: Electrophysiological studies on ciguatera posisoning in man (Part II). J Neurol Sci 101:93-97, 1991. Caruso G, Massini R, Crisci C, Nilsson J, Catalano A, Santoro L, Battaglia F, Crispi F, Nolano M: The relationship between electrophysiological findings, upper limb growth and histological features of median and ulnar nerves in man. Brain 115:1925-1945, 1992. Caruso G, Santoro L: Nerve conduction and electromyography. Curr Opinions Neurol Neurosurg 5(5):689-696, 1992. Cavanagh JB: Peripheral nerve changes in ortho-cresyl phosphate poisoning in the cat. J Pathol Bacteriol 87:365-383, 1964. Cavanagh JB, Mellick RS: On the nature of the peripheral nerve lesions associated with acute intermittent porphyria. J Neurol Neurosurg Psychiatry 28:320-327, 1965. Chaudhry V, Cornblath DR: Wallerian degeneration in human nerves: Serial electrophysiological studies. Muscle Nerve 15:687-693, 1992. Chaudhry V, Crawford TO, DeRossett SE: Thermal sensitivity in demyelinating neuropathy. Muscle Nerve 16:301-306, 1993. Chu N-S: Retrograde effects of digital nerve severance on somatosensory evoked potentials in man. Muscle Nerve 17:313-319, 1994. Chu, N-S: Current perception thresholds in toe-to-digit transplantation and digit-to-digit replantation. Muscle Nerve 19:183-186, 1996. Chu N-S: Long-term effects of finger amputation on stump skin sensibility and digital nerve conduction. Muscle Nerve 19:1049-1051, 1996. Chu N-S: Toe-to-digit transplantation as a model to study nerve regeneration and functional recovery. In Kimura, J and Shibasaki, H (Eds): Recent Advances in Clinical Neurophysiology, Elsevier Science BV, Amsterdam, 1996, p 144. Chu N-S, Chu EC, Yu J-M: Conduction study
42.
43.
44.
45.
46. 47.
48.
49. 50.
51.
52.
53.
54. 55.
85
of digital nerve function recovery following toeto-digit transplantation and a comparison with digit-to-digit replantation. Muscle Nerve 18: 1257-1264, 1995. Corbo M, Quattrini A, Latov N, Hays AP: Localization of GM1 and Gal(Bl-3)GalNAc antigenie determinants in peripheral nerve. Neurology 43:809-814, 1993. Cornblath DR, Asbury AK, Albers JW, Feasby TE, Hahn AF, McLeod JH, Mendell JR, Parry GJ, Pollard JD, Thomas PK: Research criteria for diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP). Neurology 41: 617-618, 1991. Dahlin LB, Necking LE, Lundstrom R, Lundborg G: Vibration exposure and conditioning lesion effect in nerves: An experimental study in rats. J Hand Surg 17A:858-861, 1992. Dellingham TR, Spellman NT, Braverman SE, Zeigler DN, Belandres PV, Bryant PR, Salcedo VL, Schneider RL: Analysis of casualties referred to army physical medicine services during the Persian Gulf conflict. Am J Phys Med Rehabil 72:214-218, 1993. Denny-Brown D: The neurological aspects of thiamine deficiency. Federation Proc (suppl 2) 17:35-39, 1958. Deutinger M, Girsch W, Burggasser G, Windisch A, Mayr N, Freilinger G: Clinical and electroneurographic evaluation of sensory/ motor-differentiated nerve repair in the hand. J Neurosurg 78(5):709-713, 1993. Donoso RS, Ballantyne JP, Hansen S: Regeneration of sutured human peripheral nerves: An electrophysiological study. J Neurol Neurosurg Psychiatry 42:97-106, 1979. Dorftnan LJ: Quantitative clinical electrophysiology in the evaluation of nerve injury and regeneration. Muscle Nerve 13:822-828, 1990. Dyck PJ: Inherited neuronal degeneration and atrophy affecting peripheral motor, sensory, and autonomic neurons. In Dyck PJ, Thomas PK, Lambert EH, Bunge R (eds): Peripheral Neuropathy, Vol 2. WB Saunders, Philadelphia, 1984, pp 1600-1641. Dyck PJ: Invited review: limitations in predicting pathologic abnormality of nerves from the EMG examination. Muscle Nerve 13:371-375, 1990. Dyck PJ, Gutrecht JA, Bastron JA, Karnes WE, Dale AJD: Histologic and teased fiber measurements of sural nerve in disorders of lower motor and primary sensory neurons. Mayo Clinic Proc 43:81-123, 1968. Dyck PJ, Lais AC, Karnes JL, Sparks M, Hunder H, Low PA, Windebank AJ: Permanent axotomy, a model of axonal atrophy and secondary segmental demyelination and remyelination. Ann Neurol 9:575-583, 1981. Dyck PJ, Lambert EH: Compound nerve action potentials and morphometry. Electroencephalogr Clin Neurophysiol 36:573-574, 1974. Dyck PJ, Lambert EH, Wood MB, Linscheid RL: Assessment of nerve regeneration and adaptation after median nerve reconnection and digital neurovascular flap transfer. Neurology 38: 1586-1591, 1988.
86 56. Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo JF (eds): Peripheral Neuropathy, Vols I and II. WB Saunders, Philadelphia, 1993. 57. Ebersold MJ, Quast LM: Long-term results of spinal accessory nerve-facial nerve anastomosis. J Neurosurg 77:51-54, 1992. 58. England JD, Gamboni F, Ferguson MA, Levinson SR: Sodium channels accumulate at the tips of injured axons. Muscle Nerve 17:593598, 1994. 59. England JD, Rhee EK, Said G, Sumner AJ: Schwann cell degeneration induced by doxorubicin (Adriamycin). Brain 111:901-913, 1988. 60. Erlanger J: The interpretation of the action potential in cutaneous and muscle nerves. Am J Physiol 82:644-655, 1927. 61. Erlanger J, Gasser HS: Electrical Signs of Nervous Activity. University of Pennsylvania Press, Philadelphia, 1937. 62. Evans PJ, Awerbuck DC, Mackinnon SE, Wade JA, McKee NH: Isometric contractile function following nerve grafting: a study of graft storage. Muscle Nerve 17:1190-1200, 1994. 63. Evans PJ, Mackinnon SE, Best TJ, Wade JA, Awerbuck DC Makino AP, Hunter DA, Midha R: Regeneration across preserved peripheral nerve grafts. Muscle Nerve 18:1128-1138, 1995. 64. Feasby TE: Conduction block in demyelinating neuropathies. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, pp 115-116. 65. Feasby TE, Brown WF, Gilbert JJ, Afd H: The pathological basis of conduction block in human neuropathies. J Neurol Neurosurg Psychiatry 48:239-244, 1985. 66. Feldman RG, Haddow J, Chisolm JJ: Chronic lead intoxication in urban children. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 2. Karger, Basel, 1973, pp 313-317. 67. Fernandez E, Pallini R, Marchese E, Lauretti L, Bozzini V, Sbriccoli A: Reconstruction of peripheral nerves: The phenomenon of bilateral reinnnervation of muscles originally innervated by unilateral motoneurons. Neurosurgery 30: 364-369, 1992. 68. Finkelstein DI, Dooley PC, Luff AR: Recovery of muscle after different periods of denervation and treatments. Muscle Nerve 16:769-777, 1993. 69. Franssen H, Wieneke GH, Wokke JHJ: The influence of temperature on conduction Block. Muscle Nerve 22:166-173, 1999. 70. Fujimura H, Lacroix C, Said G: Vulnerability of nerve fibres to ischaemia, a quantitative light and electron microscope study. Brain 114: 1929-1942, 1991. 71. Fukuda Y, Watanabe M, Miyoshi T, Sawai H: Which types of retinal ganglion cell axons regenerate and transmit visual information in adult cats? In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, p 148. 72. Fullerton PM: Peripheral nerve conduction in metachromatic leucodystrophy (sulphatide lipidosis). J Neurol Neurosurg Psychiatry 27:100105, 1964.
Nerve Conduction Studies 73. Fullerton PM, Barnes JM: Peripheral neuropathy in rats produced by acrylamide. Br J Industr Med 23:210-221, 1966. 74. Gamble HJ, Eames RA: An electron microscope study of the connective tissues of human peripheral nerve. J Anat (Lond) 98:655-663, 1964. 75. Gasser HS: Effect of the method of leading on the recording of the nerve fiber spectrum. J Gen Physiol 43:927-940, 1960. 76. Gilliatt RW: Acute compression block. In Sumner A (ed): The Physiology of Peripheral Nerve Disease. WB Saunders, Philadelphia, 1980, pp 287-315. 77. Gilliatt RW: Electrophysiology of peripheral neuropathies. Muscle Nerve 5:S108-S116, 1982. 78. Gilliatt RW, Hjorth RJ: Nerve conduction during Wallerian degeneration in the baboon. J Neurol Neurosurg Psychiatry 35:335-341, 1972. 79. Gilliatt RW, Meer J: The refractory period of transmission in patients with carpal tunnel syndrome. Muscle Nerve 13:445-450, 1990. 80. Gilliatt RW, Taylor JC: Electrical changes following section of the facial nerve. Proc R Soc Med 52:1080-1083, 1959. 81. Gordon TR, Kocsis JD, Waxman SG: Electrogenie pump (Na/K-ATPase) activity in rat optic nerve. Neuroscience 37:829-837, 1990. 82. Gutmann L, Gutmann L: Axonal channelopathies: An evolving concept in the pathogenesis of peripheral nerve disorders. Neurology 47: 18-21, 1996. 83. Haller FR, Low FN: The fine structure of the peripheral nerve root sheath in the subarachnoid space in the rat and other laboratory animals. Am J Anat 131:1-20, 1971. 84. Hallin RG, Ekedahl R, Frank O: Segregation by modality of myelinated and unmyelinated fibers in human sensory nerve fascicles. Muscle Nerve 14:157-165, 1991. 85. Hari R, Hallstrom J, Tiihonen J, Joutsiniemi S-L: Multichannel detection of magnetic compound action fields of median and ulnar nerves. Electroencephalogr Clin Neurophysiol 72:277280, 1989. 86. Harris JB, Thesleff S: Nerve stump length and membrane changes in denervated skeletal muscle. Nature [New Biol] 236:60-61, 1972. 87. Hartung H-P, Schafer B, Heininger K, Stoll G, Toyka KV: The role of macrophages and eicosanoids in the pathogenesis of experimental allergic neuritis. Brain 111:1039-1059, 1988. 88. Harvey GK, Pollard JD, Schindhelm K, Antony J: Chronic experimental allergic neuritis. J Neurol Sci 81:215-225, 1987. 89. Henriksen JD: Conduction velocity of motor nerves in normal subjects and patients with neuromuscular disorders. Thesis, University of Minnesota, Minneapolis, 1966. 90. Hodes R, Larrabee MG, German W: The human electromyogram in response to nerve stimulation and the conduction velocity of motor axons. Studies on normal and injured peripheral nerves. Arch Neurol Psychiatry 60:340-365, 1948. 91. Hong C-Z, Yu J: Electrophysiologic recovery of
Anatomy and Physiology of the Peripheral Nerve acute conduction block of rat tall nerve. Arch Phys Med Rehabil 70:205-209, 1989. 92. Honmou O, Felts PA, Waxman SG, Kocsis JD: Restoration of normal conduction properties in demyelinated spinal cord axons in the adult rat by transplantation of exogenous Schwann cells. J Neurosci 16(10):3199-3208, 1996. 93. Hopkins AP, Gilliatt RW: Motor and sensory nerve conduction velocity in the baboon: Normal values and changes during acrylamide neuropathy. J Neurol Neurosurg Psychiatry 34: 415-426, 1971. 94. Jacobs JM, Love S: Qualitative and quantitative morphology of human sural nerve at different ages. Brain 108:897-924, 1986. 95. Jianping J, Pollock M: Cold nerve injury is enhanced by intermittent cooling. Muscle Nerve 22:1644-1652, 1999. 96. Johnson PC, Asbury AK: The pathology of peripheral nerve. Muscle Nerve 3:519-528, 1980. 97. Kaeser HE: Funktionsprufungen peripherer Nerven bei experimentellen Polyneuritiden und bei der wallerschen Degeneration. Deutsche Z Nervenheilk 183:268-304, 1962. 98. Kaeser HE: Zur Diagnose des Karpaltunnelsyndroms. Praxis 40:991-995, 1962. 99. Kaeser HE: Diagnostische Probleme beim Karpaltunnelsyndrom. Deutsche Z Nervenheilk 185:453-470, 1963. 100. Kaeser HE, Lambert EH: Nerve function studies in experimental polyneuritis. Electroenceph Clin Neurophysiol (suppl 22):29-35, 1962. 101. Kaji R, Happel L, Sumner AJ: Effect of digitalis on clinical symptoms and conduction variables in patients with multiple sclerosis. Ann Neurol 28:582-584, 1990. 102. Kaji R, Sumner AJ: Effect of digitalis on central demyelinative conduction block in vivo. Ann Neurol 25:159-165, 1989. 103. Kaji R, Sumner AJ: Ouabain reverses conduction disturbances in single demyelinated nerve fibers. Neurology 39:1364-1368, 1989. 104. Kaji R, Suzumura A, Sumner AJ: Physiologic consequences of antiserum-mediated experimental demyelination in CNS. Brain 111:675694, 1988. 105. Kaku DA, Malamut RI, Frey DJ, Parry GJ: Conduction block as an early sign of reversible injury in ischemic monomelic neuropathy. Neurology 43:1126-1130, 1993. 106. Kanda T, Tsukagoshi H, Oda M, Miyamoto K, Tanabe H: Morphological changes in unmyelinated nerve fibres in the sural nerve with age. Brain 114:585-599, 1991. 107. Kawaguchi S, Iwashita Y, Murata M: Spontaneous and graft-induced reconnection of CNS pathways in the rat. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam 1996, p 141. 108. Kelly JJ: The electrodiagnostic findings in peripheral neuropathy associated with monoclonal gammopathy. Muscle Nerve 6:504-509, 1983. 109. Kennedy WR, Wendelschafer-Crabb G: The innervation of human epidermis. J Neurol Sci 115:184-190, 1993. 110. Kihara M, Schmelzer JD, Kihara Y, Smithson
87
IL, Low PA: Efficacy of limb cooling on the salvage of peripheral nerve from ischemic fiber degeneration. Muscle Nerve 19:203-209, 1996. 111. Kimura J: F-wave velocity in the central segment of the median and ulnar nerves. A study in normal subjects and in patients with CharcotMarie-Tooth disease. Neurology (Minneap) 24: 539-546, 1974. 112. Kimura J: The carpal tunnel syndrome. Localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619-635, 1979. 113. Kimura J: Principles and pitfalls of nerve conduction studies. Ann Neurol 16:415-429, 1984. 114. Kimura J: Consequences of peripheral nerve demyelination: Basic and clinical aspects. Can J Neurol Sci 20:263-270, 1993. 115. Kimura J, Rodnitzky R, Okawara S: Electrophysiologic analysis of aberrant regeneration after facial nerve paralysis. Neurology (Minneap) 25:989-993, 1975. 116. Krarup C, Loeb GE: Conduction studies in peripheral cat nerve using implanted electrodes: I. Methods and findings in controls. Muscle Nerve 11:922-932, 1988. 117. Krarup C, Loeb GE, Pezeshkpour GH: Conduction studies in peripheral cat nerve using implanted electrodes: II. The effects of prolonged constriction on regeneration of crushed nerve fibers. Muscle Nerve 11:933-944, 1988. 118. Krarup C, Loeb GE, Pezeshkpour GH: Conduction studies in peripheral cat nerve using implanted electrodes: III. The effects of prolonged constriction on the distal nerve segment. Muscle Nerve 12:915-928, 1989. 119. Krarup C, Upton J, Creager MA: Nerve regeneration and reinnervation after limb amputation and replantation: Clinical and physiological findings. Muscle Nerve 13:291-304, 1990. 120. Kuwabara S, Nakajima Y, Hattori T, Toma S, Mizobuchi K, Ogwara K: Activity-dependent excitability changes in chronic inflammatory demyelinating polyneuropathy: A microneurographic study. Muscle Nerve 22:899-904, 1999. 121. Kuypers PDL, Walbeehm ET, Dudok V, Heel M, Godschalk M, Hovius SER: Changes in the compound action current amplitudes in relation to the conduction velocity and functional recovery in the reconstructed peripheral nerve. Muscle Nerve 22:1087-1093, 1999. 122. Lachance DH, Daube JR: Acute peripheral arterial occlusion: Electrophysiologic study of 32 cases. Muscle Nerve 14:633-639, 1991. 123. Lai K-S, Jaweed MM, Seestead R, Herbison GJ, Ditunno JF Jr, McCully K, Chance B: Changes in nerve conduction and P1/PCr ratio during denervation-reinnervation of the gastrocsoleus muscles of rats. Arch Phys Med Rehabil 73:1155-1159, 1992. 124. Lambert EH, Dyck P: Compound action potentials of sural nerve in vitro in peripheral neuropathy. In Dyck PJ, Thomas PK, Lambert EH, Bunge R (eds); Peripheral Neuropathy, Vol 1. WB Saunders, Philadelphia, 1984, pp 1030-1044. 125. Lisney SJW: Functional aspects of the regeneration of unmyelinated axons in the rat saphenous nerve. J Neurol Sci 80:289-298, 1987.
88 126. Logiglan EL, Kelly JJ, Adelman LS: Nerve conduction and biopsy correlation in over 100 consecutive patients with suspected polyneuropathy. Muscle Nerve 17:1010-1020, 1994. 127. Lundborg G, Dahlin LB, Hansson H-A, Kanje M, Necking L-E: Vibration exposure and peripheral nerve fiber damage. J Hand Surg ISA: 346-351, 1990. 128. Macias MY, Lehman CT, Sanger JR, Riley DA: Myelinated sensory and alpha motor axon regeneration in peripheral nerve neuromas. Muscle Nerve 21:1748-1758, 1998. 129. Maltin CA, Delday MI, Hay SM, Baillie ACS: Denervation increases clenbuterol sensitivity in muscle from young rats. Muscle Nerve 15:188-192, 1992. 130. McComas AJ, White CM: Distal dysfunction and recovery in ulnar neuropathy. Muscle Nerve 19:1617, 1996. 131. McDonald WI: Conduction in muscle afferent fibres during experimental demyelination in cat nerve. Acta Neuropathol 1:425-432, 1962. 132. McDonald WI: The pathological and clinical dynamics of multiple sclerosis. J Neuropathol Exp Neurol 53:338-343, 1994. 133. McLeod JG, Prineas JW, Walsh JC: The relationship of conduction velocity to pathology in peripheral nerves. A study of sural nerve in 90 patients. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 2. Karger, Basel, 1973, pp 248-258. 134. Midha R, Mackinnon SE, Evans PJ, BestTJ, Hare GM, Hunter DA, Falk-Wade JA: Comparison of regeneration across nerve allografts with temporary or continuous cyclosporin A immunosuppression. J Neurosurg 78:90-100, 1993. 135. Miller RG: Acute vs chronic compressive neuropathy. Muscle Nerve 7:427-430, 1984. 136. Miller RG: Injury to peripheral motor nerves. AAEE minimonograph #28. Muscle Nerve 10:698-710, 1987. 137. Millesi H: Progress in peripheral nerve reconstruction. World J Surg 14:733-747, 1990. 138. Milner TE, Stein RV: The effects of axotomy on the conduction of action potentials in peripheral, sensory and motor nerve fibers. J Neurol Neurosurg Psychiat 44:495-496, 1981. 139. Mix E, Olsson T, Solders G, Link H: Effect of ion channel blockers on immune response and course of experimental allergic neuritis. Brain 112:1405-1418, 1989. 140. Mogyoros I, Kiernan MC, Burke D, Bostock H: Excitability changes in human sensory and motor axons during hyperventilation and ischaemia. Brain 120:317-325, 1997. 141. Myles LM, Gilmour JA, Glasby MA: Effects of different methods of peripheral nerve repair on the number and distribution of muscle afferent neurons in rat dorsal root. J Neurosurg 77:457-462, 1992. 142. Nagendran K: Human models provide pathophysiological information about single motor axonal regeneration. Muscle Nerve 17:698-700, 1994. 143. Nakao Y, Mackinnon SE, Hertl MC, Miyasaka M, Hunter DA, Mohanakumar T: Monoclonal
Nerve Conduction Studies antibodies against ICAM-l and LFA-1 prolong nerve allograft survival. Muscle Nerve 18:93102, 1995. 144. Neary D, Ochoa J, Gilliatt RW: Sub-clinical entrapment neuropathy in man. J Neurol Sci 24:283-298, 1975. 145. Noback CR: The Human Nervous System. McGraw-Hill, New York, 1967. 146. Novak CB, Kelly L, Mackinnon SE: Sensory recovery after median nerve grafting. J Hand Surg 17A.-59-68, 1992. 147. Novakovic SD, Levinson SR, Schachner M, Shrager P: Disruption and reorganization of sodium channels in experimental allergic neuritis. Muscle Nerve 21:1019-1032, 1998. 148. Nukada H: Drug trials: To be biopsied or not. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, p 794. 149. Nukada H, Pollock M, Haas LF: Is ischemia implicated in chronic multifocal demyelinating neuropathy? Neurology 39:106-110, 1989. 150. Ochoa J, Danta G, Fowler TJ, Gilliatt RW: Nature of the nerve lesion caused by a pneumatic tourniquet. Nature 233:265-266, 1971. 151. Ochoa J, Fowler TJ, Gilliatt RW: Anatomical changes in peripheral nerves compressed by a pneumatic tourniquet. J Anat 113:433-455, 1972. 152. Ochoa J, Marotte L: The nature of the nerve lesion caused by chronic entrapment in the guinea-pig. J Neurol Sci 19:491-495, 1973. 153. Ochoa JL, Torebjork HE: Paraesthesiae from ectopic impulse generation in human sensory nerves. Brain 103:835-853, 1980. 154. Oda K, Araki K, Totoki T, Shibasaki H: Nerve conduction study of human tetrodotoxication. Neurology 39:743-745, 1989. 155. Pallini R, Fernandez E, Lauretti L, Draicchio F, Pettorossi VE, Gangitano C, Del-Fa A, OlMeriSangiacomo C, Sbriccoli A: Experimental repair of the oculomotor nerve: The anatomical paradigms of functional regeneration. J Neurosurg 77:768-777, 1992. 156. Parry GJ: Pathophysiological mechanisms in peripheral nerve injury. AAEM Plenary Session: Physical Trauma to Peripheral Nerves, Minneapolis, MN, 1996. 157. Parry GJ, Brown MJ: Selective fiber vulnerability in acute ischemic neuropathy. Ann Neurol 11:147-154, 1982. 158. Parry GJ, Cornblath DR, Brown MJ: Transient conduction block following acute peripheral nerve ischemia. Muscle Nerve 8:490-113, 1985. 159. Parry GJ, Linn DJ: Conduction block without demyelination following acute nerve infarction. J Neurol Sci 84:265-273, 1988. 160. Pencek TL, Schauf CL, Low PA, Eisenberg BR, Davis FA: Disruption of the perineurium in amphibian peripheral nerve: Morphology and physiology. Neurology (New York) 30:593-599, 1980. 161. Pilling JB: Nerve conduction during Wallerian degeneration in man. Muscle Nerve 1:81, 1978. 162. Pover CM, Lisney SJW: An electrophysiological and histological study of myelinated axon regeneration after peripheral nerve injury and repair in the cat. J Neurol Sci 85:281-291, 1988.
Anatomy and Physiology of the Peripheral Nerve 163. Prineas J: The pathogenesis of dying-back polyneuropathies. Part II. An ultrastructural study of experimental acrylamlde intoxication in the cat. J Neuropathol Exp Neurol 28:598621, 1969. 164. Rab M, Roller R, Haslik W, Neumayer C, Todoroff BP, Frey M, Gruber H: The impact of a muscle target organ on nerve grafts with different lengths—A histomorphological analysis. Muscle Nerve 21:618-627, 1998. 165. Rasminsky M: Ectopic generation of impulses and cross-talk in spinal nerve roots of "dystrophic" mice. Ann Neurol 3:351-357, 1978. 166. Rasminsky M: Physiological consequences of demyelination. In Spencer PS, Schaumburg HH (eds): Experimental Clinical Neurotoxicology. Williams & Wilkins, Baltimore, 1980, pp 257-271. 167. Raynor EM, Ross MH, Shefher JM, Preston DC: Differentiation between axonal and demyelinating neuropathies: identical segments recorded from proximal and distal muscles. Muscle Nerve 18:402-408, 1995. 168. Redford EJ, Kapoor R, Smith KJ: Nitric oxide donors reversibly block axonal conduction: Demyelinated axons are especially suseptible. Brain 120:2149-2157, 1997. 169. Ridderheim PA, Von Essen O, Blom S, Zetterlund B: Intracranially recorded compound action potentials from the human trigeminal nerve. Electroencephalogr Clin Neurophysiol 61:138-140, 1985. 170. Rizzo MA, Kocsis JD, Waxman SG: Mechanisms of paresthesiae, +dysesthesiae, and hyperesthesiae: role of Na channel heterogeneity. Eur Neurol 36:3-12, 1996. 171. Robinson LR, Nielsen VK: Limits of normal nerve function during high frequency stimulation. Muscle Nerve 13:279-285, 1990. 172. Rudge P: Tourniquet paralysis with prolonged conduction block. An electrophysiological study. J Bone Joint Surg 566:716-720, 1974. 173. Rudge P, Ochoa J, Gilliatt RW: Acute peripheral nerve compression in the baboon. J Neurol Sci 23:403-420, 1974. 174. Said G, Saida K, Saida T, Asbury AK: Axonal lesions in acute experimental demyelination: a sequential teased nerve fiber study. Neurology 31:413-421, 1981. 175. Sakai J, Honmou O, Kocsis JD, Hash! K: The delayed depolarization in rat cutaneous afferent axons is reduced following nerve transection and ligation, but not crush: Implications for injury-induced axonal Na+ channel reorganization. Muscle Nerve 21:1040-1047, 1998. 176. Santoro M, Uncini A, Corbo M, Staugaitis SM, Thomas FP, Hays AP, Latov N: Experimental conduction block induced by serum from a patient with anti-GM: antibodies. Ann Neurol 31:385-390, 1992. 177. Schady W, Braune S, Watson S, Torebjork HE, Schmidt R: Responsiveness of the somatosensory system after nerve injury and amputation in the human hand. Ann Neurol 36:68-75, 1994. 178. Schaumburg HH, Wisniewski HM, Spencer PS: Ultrastructural studies of the dying-back
179.
180.
181. 182.
183. 184.
185.
186.
187.
188. 189.
190. 191.
192.
193.
194.
89
process. I. Peripheral nerve terminal and axon degeneration in systemic acrylamide intoxication. J Neuropathol Exp Neurol 33:260-284, 1974. Schellens RLLA, van Veen BK, GabreelsFesten AAWM, Notermans SLH, van t Hof MA, Stegeman DF: A statistical approach to fiber diameter distribution in human sural nerve. Muscle Nerve 16:1342-1350, 1993. Schwarz JR, Corrette BJ, Mann K, Wietholter H: Changes of ionic channel distribution in myelinated nerve fibers from rat with experimental allergic neuritis. Neurosci Lett 122:205209, 1991. Seckel BR: Enhancement of peripheral nerve regeneration. Muscle Nerve 13:785-800, 1990. Sedal L, Ghabriel MN, He F, Allt G, Le Quesne PM, Harrison MJG: A combined morphological and electrophysiological study of conduction block in peripheral nerve. J Neurol Sci 60: 293-306, 1983. Seddon H: Surgical Disorders of the Peripheral Nerves, ed 2. Churchill Livingstone, Edinburgh, 1975. Shimpo T, Gilliatt RW, Kennett RP, Allen PJ: Susceptibility to pressure neuropathy distal to a constricting ligature in the guinea-pig. J Neurol Neurosurg Psychiatry 50:1625-1632, 1987. Sima AAF, BrismarT: Reversible diabetic nerve dysfunction: Structural correlates to electrophysiological abnormalities. Ann Neurol 18:2129, 1985. Simpson JA: Conduction velocity of peripheral nerves in human metabolic disorders. Electroencephalogr Clin Neurophysiol (suppl 22): 36-43, 1962. Sindrup SH, Ejlertsen B, Gjessing H, Froland A, Sindrup EH: Peripheral nerve function during hyperglycemic clamping in healthy subjects. Acta Neurol Scand 78:141-145, 1988. Smith KJ, Blakemore WF, McDonald WI: The restoration of conduction by central remyelination. Brain 104:383-404, 1981. Smith KJ, Blakemore WF, Murray JA, Patterson RC: Internodal myelin volume and axon surface area: a relationship determining myelin thickness? J Neurol Sci 55:231-246, 1982. Smith KJ, McDonald WI: Spontaneous and evoked electrical discharges from a central demyelinating lesion. J Neurol Sci 55:39-47, 1982. Sobue G, Yasuda T, Mitsuma T, Ross AH, Pleasure D: Expression of nerve growth factor receptor in human peripheral neuropathies. Ann Neurol 24:64-72, 1988. Spencer PS, Schaumburg HH: Central-peripheral distal axonopathy—The pathology of dyingback polyneuropathies. In Zimmerman HM (ed): Progress in Neuropathology, Vol III. Grune & Stratton, New York, 1976, pp 253-295. Stanley GP, McCombe PA, Pender MP: Focal conduction block in the dorsal root ganglion in experimental allergic neuritis. Ann Neurol 31: 27-33, 1992. Struppler A, Huckauf H: Propagation velocity in regenerated motor nerve fibres. Electroencephalogr Clin Neurophysiol (suppl 22): 58-60, 1962.
90 195. Sumner A: Physiology of dying-back neuropathies. In Waxman SG (ed): Physiology and Pathobiology of Axons. Raven Press, New York, 1978, pp 349-359. 196. Sumner AJ: Aberrant reinnervation. Muscle Nerve 13:801-803, 1990. 197. Sumner AJ, Asbury AK: Physiological studies of the dying-back phenomenon. Muscle stretch afferents in acrylamide neuropathy. Brain 98:91-100, 1975. 198. Sumner AJ, Saida K, Saida T, Silberberg DH, Asbury AK: Acute conduction block associated with experimental antiserum-mediated demyelination of peripheral nerve. Ann Neurol 11:469-477, 1982. 199. Sunderland S: Nerves and Nerve Injuries, ed 2. Churchill Livingstone, Edinburgh, 1978. 200. Sunderland S: The Anatomy and Physiology of Nerve Injury. AAEE International Symposium on Peripheral Nerve Regeneration, Washington, DC, 1989, pp 1-19. 201. Sunderland S: The anatomy and physiology of nerve injury. Muscle Nerve 13:771-784, 1990. 202. Swift TR, Leshner RT, Gross JA: Arm-diaphragm synkinesis: Electrodiagnostic studies of aberrant regeneration of phrenic motor neurons. Neurology (New York) 30:339-344, 1980. 203. Tallis R, Staniforth P, Fisher TR: Neurophysiological studies of autogenous sural nerve grafts. J Neurol Neurosurg Psychiatry 41:677-683, 1978. 204. Teixeira FJ, Aranda F, Becker LE: Postnatal maturation of phrenic nerve in children. Pediatr Neurol 8:450-454, 1992. 205. Thomas CK, Stein RB, Gordon T, Lee RG, Elleker MG: Patterns of reinnervation and motor unit recruitment in human hand muscles after complete ulnar and median nerve section and resuture. J Neurol Psychiatry 50:259-268, 1987. 206. Thomas PK: The morphological basis for alterations in nerve conduction in peripheral neuropathy. Proc R Soc Med 64:295-298, 1971. 207. Thomas PK: Motor nerve conduction in the carpal tunnel syndrome. Neurology (Minneap) 10:1045-1050, 1960. 208. Thomas PK: Invited review: Focal nerve injury: Guidance factors during axonal regeneration. Muscle Nerve 12:796-802, 1989. 209. Thomas PK, Olsson Y: Microscopic anatomy and function of the connective tissue components of peripheral nerve. In Dyck PJ, Thomas PK, Lambert EH, Bunge R (eds): Peripheral Neuropathy, Vol 1. WB Saunders, Philadelphia, 1984, pp 97-120. 210. Trojaborg W: Early electrophysiologic changes in conduction block. Muscle Nerve 1:400-403, 1978. 211. Trojaborg W, Galassi G, Hays AP, Lovelace RE, Alkaitis M, Latov N: Electrophysiologic study of experimental demyelination induced by serum of patients with IgM M proteins and neuropathy. Neurology 39:1581-1586, 1989. 212. Tsujihata M, Engel AG, Lambert EH: Motor end-plate fine structure in acrylamide dyingback neuropathy: A sequential morphometric study. Neurology (Minneap) 24:849-856, 1974.
Nerve Conduction Studies 213. Ugawa Y, Sakura M, Okutsu I, Kuroshima N, Ninomiya S: Free functional musculocutaneous transfer: Electrophysiological studies. Eur Neurol 28:241-245, 1988. 214. Utzschneider DA, Archer DR, Kocsis JD, Waxman SG, Duncan ID: Transplantation of glial cells enhances action potential conduction of amyelinated spinal cord axons in the myelindeflcient rat. Proc Natl Acad Sci USA 91:53-57, 1994. 215. Verdugo RJ, Ochoa JL: Use and misuse of conventional electrodiagnosis, quantitative sensory testing, thermography, and nerve blocks in the evaluation of painful neuropathic syndromes. Muscle Nerve 16:1056-1062, 1993. 216. Vincent A, et al: Plasma from patients with raised anti-GMl antibodies passively transfers conduction block. J Neuroimmunol 54:204, 1994. 217. Wall PD, Gutnick M: Properties of afferent nerve impulses originating from a neuroma. Nature 248:740-743, 1974. 218. Watchmaker GP, Gumucio CA, Crandall RE, Vannier MA, Weeks PM: Fascicular topography of the median nerve: A computer based study to identify branching patterns. J Hand Surg 16A:53-59, 1991. 219. Waxman SG: Conduction in myelinated, unmyelinated, and demyelinated fibers. Arch Neurol 34:585-589, 1977. 220. Waxman SG: Determinants of conduction velocity in myelinated nerve fibers. Muscle Nerve 3:141-150, 1980. 221. Waxman SG, Brill MH: Conduction through demyelinated plaques in multiple sclerosis: Computer simulations of facilitation by short internodes. J Neurol Neurosurg Psychiatry 41:408-416, 1978. 222. Waxman SG, Brill MH, Geschwind N: Probability of conduction deficit as related to fiber length in random distribution models of peripheral neuropathies. J Neurol Sci 29:39-53, 1976. 223. Waxman SG, Cummins TR, Dib-Hajj S, Fjell J, Black JA: Sodium channels, excitability of primary sensory neurons, and the molecular basis of pain. Muscle Nerve 22:1177-1187, 1999. 224. Waxman SG, Kocsis JD, Black JA: Type III sodium channel mRNA is expressed in embryonic but not adult spinal sensory neurons, and is reexpressed following axotomy. J Neurophysiol 72:466-470, 1994. 225. Webster H: Peripheral nerve structure. In Hubbard JI (ed): The Peripheral Nervous System. Plenum Press, New York, 1974, pp 3-26. 226. Westland K, Pollard JD: Proteinase induced demyelination. J Neurol Sci 82:41-53, 1987. 227. Wilbourn AJ: Electrodiagnosis with entrapment neuropathies. AAEM 1992 Plenary Session I: Entrapment Neuropathies. American Association of Electrodiagnostic Medicine, Rochester, MN, 1992, pp 23-27. 228. Wilbourn AJ: Nerve injuries caused by injections and tourniquets. AAEM Plenary Session: Physical Trauma to Peripheral Nerves, Minneapolis, MN, 1996. 229. Yarnitsky D, Ochoa JL: Release of cold-induced burning pain by block of cold-specific afferent input. Brain 113:893-902, 1990.
Chapter 5 PRINCIPLES AND VARIATIONS OF NERVE CONDUCTION STUDIES
1. INTRODUCTION 2. ELECTRICAL STIMULATION OF THE NERVE Cathode and Anode Types of Stimulators Stimulus Intensity and Duration Stimulus Artifact 3. RECORDING OF MUSCLE AND NERVE POTENTIALS Surface and Needle Electrodes Optimal Recording of Intended Signals Averaging Technique Display and Storage of Recorded Signals 4. MOTOR NERVE CONDUCTION Stimulation and Recording Calculation of Conduction Velocity Possible Sources of Error Types of Abnormalities 5. SENSORY NERVE CONDUCTION Stimulation and Recording Waveform, Amplitude, and Duration Latency and Conduction Velocity 6. NERVE CONDUCTION IN THE CLINICAL DOMAIN Physiologic Variation Among Different Nerve Segments Effects of Temperature Maturation and Aging Height and Other Factors Clinical Values and Limitations 7. STUDIES OF THE AUTONOMIC NERVOUS SYSTEM Heart-Rate Variation with Breathing Valsalva Ratio Response to Change in Posture Sympathetic Skin Response 8. OTHER EVALUATION OF NERVE FUNCTION Microneurography Thermal, Pain, Vibratory, and Tactile Sensation Thermography 91
Nerve Conduction Studies
92
1 INTRODUCTION Helmholtz (1850)133 originally recorded the mechanical response of a muscle to measure conduction velocity of motor fibers (see Appendix 1). Piper (1909)251 used the muscle action potential for this 21 purpose. Subsequent138 animal experiments and human studies popularized the technique as a clinical test. Eichler (1937)84 recorded nerve potentials percutaneously in man. Dawson and Scott (1949)58 introduced photographic superimposition and later electrical averaging for better resolution, making it possible to record sensory nerve action potentials through surface electrodes60 after stimulation of the digital nerves. With steady improvement of recording apparatus, nerve conduction studies have become a simple and reliable test of peripheral nerve function. With adequate standardization, the method now provides a means of not only objectively indentifying the lesion but also precisely localizing the site of maximal involvement.162 Electrical stimulation of the nerve initiates an impulse that travels along the motor or sensory nerve fibers. The assessment of conduction characteristics depends on the analysis of compound evoked potentials recorded from the muscle in the study of the motor fibers and from the nerve itself, in the case of the sensory fibers. The same principles apply in all circumstances, although the anatomic course and pattern of innervation dictates the exact201technique used for testing a given nerve. In addition to electrical shocks, used in most clinical studies, tactile stimulation can also elicit nerve action potentials.15,35,222 Assessment of mechanical characteristics helps delineate contractile properties of the isometric twitch induced by stimulation of the nerve.219 2
ELECTRICAL STIMULATION OF THE NERVE
Cathode and Anode Surface electrodes, usually made of silver plate, come in different sizes, commonly
in the range of 0.5-1.0 cm in diameter. Stimulating electrodes consist of a cathode (negative pole) and an anode (positive pole), so called because they attract cations and anions. As the current flows between them, negative charges that accumulate under the cathode depolarize the nerve. Conversely, positive charges under the anode hyperpolarize the nerve, although probably not to the extent of blocking74the conduction during routine studies. In bipolar stimulation, with both electrodes over the nerve trunk, placing the cathode closer to the recording site avoids anodal conduction block, if any. Alternatively, locating the anode away from the nerve trunk also prevents its hyperpolarizing effect. Accurate calculation of conduction velocities depends on proper measurements of the distance between the consecutive cathodal points used to stimulate the nerve at multiple sites. Clearly labeling the stimulating electrodes avoids inadvertent surface measurement from the cathode at one stimulus site to the anode at another, which would lead to an erroneous results.
Types of Stimulators Most commercially available stimulators provide a probe that mounts the cathode and the anode at a fixed distance, usually 2-3 cm apart. The intensity control located in the insulated handle, though bulky, simplifies the operation for a single examiner. The ordinary banana plugs connected by shielded cable also serve well as stimulating electrodes. Some electromyographers prefer a monopolar stimulation with a small cathode placed on the nerve trunk over the volar surface and a large anode over the dorsal surface in the same limb. The conduction velocities obtained in this fashion differ slightly, but randomly, from those determined by bipolar arrangements. The use of a needle inserted subcutaneously as the cathode reduces the current necessary to excite the nerve compared to surface stimulation. The maximum current during such stimulation causes neither248electric nor heat damage to the tissue. Either a surface electrode located on the skin nearby or a
Principles and Variations of Nerve Conduction Studies second needle electrode inserted in the vicinity of the cathode may serve as the anode. Electromyographers use two basically different kinds of electric stimulators in nerve conduction studies (see Chapter 3-7). Of these, constant-voltage stimulators regulate the output in voltage so that the actual current varies inversely with the impedance of the electrode, skin, and subcutaneous tissues. In constant-current units, the voltage changes according to the impedance, so that a specified amount of current reaches the nerve within certain limits of the skin resistance. Either type suffices for clinical use, provided that the stimulus output has an adequate range to elicit maximal muscle and nerve action potentials in all patients. A constant-current unit provides a better means of serially assessing the level of shock intensity as a measure of nerve excitability.
Stimulus Intensity and Duration The output impulse provides a square wave of variable duration, ranging from 0.05 to 2.0 ms. Surface stimulation of 0.1 ms duration and 100-300 V or 10-30 mA intensity usually activates a healthy nerve fully. A study of diseased nerves with decreased excitability may require a maximal output of 400-500 V or 40-50 mA. Electrical stimulation within the above intensity range causes no particular risk unless the patient is electrically sensitive. Any current, if delivered near the implantation site, for example, could inhibit a cardiac pacemaker.27 Special care to safeguard such patients includes proper grounding and placement of the nerve stimulator at sufficient distance from the pacemaker.1,179 Similarly, in patients with indwelling cardiac catheters or central venous pressure lines inserted directly into the heart, all the current may directly reach the cardiac tissue. This possibility makes routine nerve conduction studies contraindicated in such patients. Implanted cardioverters and defibrillators also pose safety hazards that usually preclude electric stimulation near the implantation site. Consultation
93
with a cardiologist with special expertise in this area should address feasibility of a nerve conduction study in any patient using such a medical device, and the need to turn the system off or on during the procedure. Placing the stimulator at least 6 inches away may minimize the233 chance of externally triggering the device. Electromyographers should always keep in mind these and other problems related to general electrical safety (see Appendix 3). It is common to qualify electrical stimuli on the basis of the magnitude of the evoked potential. A threshold stimulus barely elicits a response in some, but not all, of the axons contained in the nerve. Increasing the duration of stimulation decreases the threshold intensity, thereby prolonging the latency of nerve volleys whether tested in single motor axons, compound nerve potentials or H reflexes.227,228,243 A maximal stimulus activates the entire group of axons, so that further increase in shock intensity causes neither additional increase in the amplitude nor shortening in latency of the evoked potential. Unlike a threshold stimulus, a maximal shock activates the axon at or close to its rising edge, independent of its duration. The current required for maximal stimulation varies greatly from one subject to the next and from one nerve to another in the same individual. A supramaximal stimulus has an intensity greater than the maximal stimulus. Increasing the intensity of an already supramaximal stimulus continue to shorten the latency of nerve volleys because the spread of current tends to activate the nerve segment away from the cathodal point. If fibers with large diameters have the lowest threshold 89,300 in humans, as in experimental animals, then a submaximal stimulus should theoretically suffice for determining the onset latency of the fastest conducting fibers. Although this assumption usually holds, especially with sensory nerves,269 the exact order of excitation also depends on the spatial relationship of various fibers and the stimulating electrode.109,252 Further, in motor conduction studies, the length of the axon terminals, which partially determines the latency of the muscle response, varies
Nerve Conduction Studies
94
within a given nerve. Thus, with submaximal stimuli, the onset latency fluctuates considerably from one trial to the next, depending on the excited axons within a nerve. In extreme cases the first axons excited151 may in fact have the longest latencies. The use of supramaximal stimuli, which activate all of the axons, circumvents this uncertainty. Most commercial stimulators can provide a pair of stimuli at variable intervals and a train of stimuli of different rates and duration. Ideally, each of the paired stimuli should have independent controls of both duration and intensity. A trigger output for the oscilloscope sweep should precede each stimulus by a variable delay, to allow a clear marking of the exact stimulus point on the display.285
Stimulus Artifact The control of a stimulus artifact often poses a major technical challenge in nerve conduction studies. Most electrode amplifiers recover from an overloading input in 5 to 10 ms, depending on the amplifier design and the amount of overload. With the stimulus of sufficient magnitude, an overloading artifact interferes with accurate recording of short-latency responses. Better stimulus isolation from the ground through an isolating transformer serves to reduce excessive shock artifact.48 Not only does this eliminate amplifier overloading, but it also protects the patient from unexpected current leakage. The use of the transformer, however, makes it difficult to faithfully preserve the waveform of the original stimulus. A radio-frequency isolation also minimizes stimulus artifacts while maintaining the original shape of the stimulus better than the transformer. Unfortunately, high-frequency stimulus isolation units generally fail to deliver adequate intensity for supramaximal stimulation. Finally, the use of a fast-recovery amplifier circumvents the problem of stimulus artifacts.324 Even then, optimal recording of short-latency responses calls for adequate reduction of surface spread of stimulus current, as stated below. Shock artifacts increase with less separation between stimulus and recording
sites and greater distance between the active (G1) and reference (G2) electrodes. The stimulator leads, which have no shield, can also cause a large artifact if placed near the recording electrodes. With excessive surface spread, a square pulse of 0.1 ms duration can affect the active electrode for several milliseconds at the signal level of recording with high sensitivity. Thus, reduction in surface spread of stimulus current ensures an optimal recording of short-latency responses. Wiping with alcohol helps dry the moist skin surface before the application of the stimulus. Adequate preparation of the stimulating and recording sites reduces the skin resistance. Surface grease will dissolve if cleaned with ether. Callous skin needs gentle abrasion with a dull knife or fine sandpaper. Rubbing the skin with a cream or solvent of high conductance lowers the impedance between the electrode and the underlying tissue. Theoretically, placement of a ground electrode between the stimulating and recording electrode diminishes the stimulus artifact. In practice, an alternative location may suffice with the use of a modern fast-recovery amplifier.
3 RECORDING OF MUSCLE AND NERVE POTENTIALS
Surface and Needle Electrodes Surface electrodes, in general, are better suited than needle electrodes for recording a compound muscle action potential in assessing contributions from all discharging units. Its onset latency indicates the conduction time of the fastest fibers, and its amplitude serves as a measure of the number of available axons. Averaging technique, though not usually required, may help14in evaluating markedly atrophic muscles. A needle electrode registers only a small portion of the muscle action potential, showing a more abrupt onset and less interference from neighboring discharges. Thus, its use may help in evaluating small atrophic muscles when surface recording fails. It also improves seg-
Principles and Variations of Nerve Conduction Studies regation of an action potential from a target muscle after proximal stimulation, which tends to excite many muscles simultaneously. Surface electrodes suffice for recording sensory and mixed nerve action potentials. Some electromyographers, however, prefer needle electrodes placed perpendicular to the nerve to improve the resolution. With this technique, the amplitude of the recorded potential increases by a factor of 2-3 times.269 The combination of the two effects enhances the signal-tonoise ratio by about 5 times and, when averaging, reduces the time required to reach the same resolution considerably. Many laboratories now use ring electrodes placed over the proximal and distal interphalangeal joints to record the antidromic sensory potentials from the digital nerves. Studies of the commonly tested nerves usually require no averaging because individual stimuli give rise to sensory potentials of sufficient amplitude. Unnecessary averaging is often a poor excuse for a bad technique.
Optimal Recording of Intended Signals The principles of amplification and display used in electromyography also apply to nerve conduction studies (see Chapter 3-3). Instead of continuous runs, a prepulse intermittently triggers the sweep, followed, after a short delay, by the stimulus. This arrangement allows precise measurement of the time interval between the stimulus and the onset of the evoked potential. The magnitude of the potential under study dictates the optimal amplifier sensitivity for determination of the amplitude and the latency. Overamplification results in truncation of the recorded response, whereas underamplification precludes accurate measurements of its take-off from the baseline. A 1.0 mV muscle action potential, if amplified 1000 times, causes a 1 cm vertical deflection on the oscilloscope at a display setting of 1 V/cm. A much smaller sensory or mixed nerve action potential, on the order of 10 ^tV, requires a total amplification of about 100,000 times. With
95
such a high gain, the amplifier must have a very low inherent noise level. The use of low-pass filters helps to further reduce such high-frequency interference. The electrode amplifier should provide differential amplification with a signal-to-noise discrimination ratio close to 100,000:1 and an input impedance greater than 1 megohm. It should respond to frequencies of wide bandwidth ranging from 2 Hz to 10 kHz without undue distortion.
Averaging Technique Conventional techniques fail to detect signals within the expected noise level of the system. Interposing a step-up transformer between the recording electrodes and the amplifier improves the signal-to-noise ratio,32 as does placing the first stage of the amplifier near the electrode site with a remote preamplifier box.324 The use of digital averaging represents a major improvement58over the photographic superimposition and early averager59 with its motor-driven switch and multiple storage capacitors. The electronic devices now in use are triggered by repetitive stimulation to sum consecutive samples of waveforms that are stored digitally after each sweep. The voltage from noise that randomly changes its temporal relationship to stimulation in successive tracings will average close to zero at each time point after stimulus onset. In contrast, signals timelocked to the stimulus will sum at a constant latency and appear as an evoked potential, distinct from the background noise within certain limits. In recording a sensory nerve action potential, for example, averaging can virtually eliminate the background noise up235 to 50 times but not 100 times the signal. Electrical division of the summated potential by the number of trials will provide an average value of the signal under consideration. Here, the degree of enhancement increases in proportion to the square root of the trial number. Thus, four trials give twice as large a response, whereas nine trials give three times the response. In other words, the signal-to-noise ratio improves by a factor of the square root of 2 every time the number of trials is doubled.
96
Nerve Conduction Studies
Display and Storage of Recorded Signals The use of an oscilloscope with digital storage capacity can display a series of responses with a stepwise vertical shift of the baseline, to facilitate the comparison of successively elicited potentials in waveform and latency. An automatic device digitally displays the latency based on mathematically defined take-off from the baseline. When necessary, manual positioning of the marker to the desired spot of the waveform improves accuracy. Conversely, minor adjustments inconsistent with the overriding definition may induce measurement errors. Modern oscilloscopes provide a very stable time base requiring no marking of calibration signals on the second beam. Consequently, a single channel suffices for most routine nerve conduction studies. Dual channels, however, have a distinct advantage in simultaneous recording of related events. Oscilloscopes with four or more channels allow multichannel analysis. A magnetic tape recorder can store the original potentials using either frequency modulation (FM) or amplitude modulation (AM). The FM mode has a limited highfrequency response, but can adequately
record the frequency range of the compound action potential, including DC changes. Further, in the analysis of evoked muscle or nerve potentials, the FM method preserves the amplitude of the recorded potential very accurately. In contrast, the AM modulation responds well to high-frequency bands but distorts the amplitude of the recorded response. The AM method preserves the high-frequency components better for recording motor unit potentials with needle electrodes (see Chapter 3-5). 4
MOTOR NERVE CONDUCTION
Stimulation and Recording Motor conduction studies consist of stimulating the nerve at two or more points along its course and recording muscle action potentials (Fig. 5-1) with a pair of surface electrodes: an active lead (G1) placed on the belly of the muscle and an indifferent lead (G2) placed on the tendon.168,330 Depolarization under the cathode results in the generation of a nerve action potential, whereas hyperpolarization under the anode tends to block the propagation of the nerve impulse (see Chapter 4-3). Although this
Figure 5-1. Compound muscle action potential recorded from the thenar eminence following stimulation of the median nerve at the wrist. The distal or terminal latency includes (1) nerve conduction from the stimulus point to the axon terminal and (2) neuromuscular transmission, including the time required for generation and propagation of the muscle action potential after depolarization of the endplate.
Principles and Variations of Nerve Conduction Studies poses theoretical rather than practical concern in the usual clinical setup, placing the anode 2-3 cm proximal to the cathode precludes the possibility of blocking the propagation of the nerve impulse. Pulses of moderate intensity help adjust the position of the cathode until further relocation causes no change in the size of the muscle action potential. With the cathode at the best stimulating site, increasing the intensity elicits a progressively larger response until it reaches a maximal potential. Increasing the stimulus further should result in no change in the size of the muscle potential. The use of a 20-30 percent supramaximal intensity guarantees the activation of all the nerve axons innervating the recorded muscle. With belly-tendon recording, the propagating muscle action potential, originating under GI, located near the motor point, gives rise to a simple biphasic waveform with the initial negativity (see Chapter 2-4). A small positive potential may precede the negative peak with inappropriate 77 positioning of the recording electrodes. If GI placed outside the motor point records a positivity from one part of muscle and a negativity from another, canceling effect makes the initial segment isoelectric with apparent delay of onset.298 This "false" motor point may also result from inadvertent recording from nearby muscles.63 The location of G2 substantially influences the waveform of recorded response.27 The compound muscle action potential consists of many motor unit action potentials within the recording radius of the active electrode in the range of 20 mm from the skin surface.16 A single shock applied to the nerve activates a group of motor units slightly asynchronously, reflecting the difference in conduction velocity and in terminal length of individual nerve axons (see Chapter 7-5). Temporally dispersed impulses result in a degree of phase cancellation depending on the nerve length under study and other multiple factors. The location of the pick up electrodes determines the spatial orientation to the constituent motor units and consequently the pattern of their contribution.30,161,168,184,221,318,319 The use of large electrodes tends to reduce site-induced variability of recorded poten305,317 tials.
97
The usual measurements include amplitude from the baseline to the negative peak or between negative and positive peaks; duration from the onset to the negative or positive peak or to the final return to the baseline; and latency, from the stimulus artifact to the onset of the negative response. Electronic integration can provide the area under the waveform, which shows a linear correlation to the product of the amplitude and duration measured by conventional means.108 Latency consists of three components: (1) nerve activation time from application of the stimulus to the generation of action potential, (2) nerve conduction time, from the stimulus point to the nerve terminal, and (3) neuromuscular transmission time, from the axon terminal to the motor end plate, including the time required for generation of muscle action potential. Onset latency in general provides a measure of the fast-conducting motor fibers, although the shortest, but not necessarily fastest, axons may give rise to the initial potential.
Calculation of Conduction Velocity The motor nerve conduction time equals the latency minus the time for nerve activation, and neuromuscular transmission including the generation of muscle action potential. The latency difference between the two responses elicited by stimulation at two separate points, in effect, excludes these components common to both stimuli. Thus, it represents the time necessary for the nerve impulse to travel from one point of stimulation to the next (Fig. 5-2). Dividing the distance between the stimulus points by the corresponding latency difference derives the conduction velocity. The reliability of results depends on accuracy in determining the length of the nerve segment, estimated with the surface distance along the course of the nerve, and the latency measurements at the two stimulus sites. To recapitulate, the nerve conduction velocity equals
98
Nerve Conduction Studies
Figure 5-2. Compound muscle action potential recorded from the thenar eminence folflowing the stimulation of the median nerve at the elbow. The nerve conduction time from the elbow to the wrist equals the latency difference between the two responses elicited by the distal and proximal stimulations. The motor nerve conduction velocity (MNCV), calculated by dividing the surface distance between the stimulus points by the subtracted times, concerns the fastest fibers.
where D is the distance between the two stimulus points in millimeters, and Lp and Ld, the proximal and distal latencies in milliseconds. Stimulation at multiple points along the length of the nerve allows calculation of segmental conduction velocities. Separation of the two stimulation points by at least several centimeters, and preferably more than 10, improves the accuracy of surface measurement and, consequently, determination of conduction velocity. In the case of restricted lesions, as in a compressive neuropathy, however, the inclusion of longer unaffected segments dilutes the effect of slowing and lowers the sensitivity of the test. Here, incremental stimulation across the shorter segment helps isolate the localized abnormality that may otherwise escape detection (see Chapters 6-2, 7-6). The latency from the most distal stimulus point to the muscle includes not only the nerve activation and conduction time but also neuromuscular transmission time. The inclusion of the additional factors precludes calculation of conduction velocity over the most distal segment. Here, meaningful comparison requires the use of either premeasured fixed distance or anatomic landmarks for electrode placement.223 Both approaches equally
improve the accuracy of latency determination.226 The actual conduction time in the terminal segment (Ld) slightly exceeds the calculated value (Ld' = D/CV) for the same distance (D) based on the conduction velocity (CV) of more proximal segments. The difference (Ld - Ld'), known as the residual latency, provides a measure of the conduction delay at the nerve terminal153,172 and at the neuromuscular junction. The ratio between the calculated and measured latency (Ld'/Ld), referred to as the terminal latency index, also relates to distal conduction delay (see Chapter 6-2).280 For example, a patient with a measured distal latency (Ld) of 4.0 ms for the terminal distance of 8 cm, and forearm conduction velocity (CV) of 50 m/s would have a calculated latency (Ld') of 1.6 ms (8 cm/50 m/s), residual latency of 2.4 ms (4.0-1.6 ms), and terminal index ratio of 0.4 (1.6 ms/4.0 ms).
Possible Sources of Error In normal subjects, shocks of supramaximal intensity elicit almost, but not exactly, identical compound muscle action potentials, depending on the nerve length between the stimulating and recording
Principles and Variations of Nerve Conduction Studies electrodes. The impulses of the slow conducting fibers lag progressively behind those of the fast conducting fibers over a longer conducting path. Hence, a proximal stimulus gives rise to an evoked potential of slightly longer duration and lower amplitude than a distal shock (see Chapter 7-5). This physiologic temporal dispersion does not drastically alter the waveform of the muscle action potentials, as predicted by analysis of durationdependent phase cancellation (see Fig. 7-11). The evoked potentials of dissimilar shapes preclude accurate calculation of conduction velocity, because the two onset latencies may represent motor fibers of different conduction characteristics. Distorted waveforms result from the use of an inappropriately low stimulus intensity, which activates only part of the nerve fibers. Conversely, an excessive stimulus intensity can cause an erroneously short latency because the spread of stimulus current depolarizes the nerve 250 a few millimeters away from the cathode. The surface length measured between the two cathodal points under these conditions does not precisely correspond to the conduction distance of the nerve segment under study.334 When recorded with a high sensitivity, a small negative peak sometimes precedes the main negative component of the muscle action potential.33,117,285 This small potential, disregarded in latency determination, probably originates from small nerve fibers near the motor point.76 A small nerve action potential recorded from the digital nerve by the G2 electrode has a longer latency and opposite polarity.118 In addition, G2 electrodes placed distal to the metacarpophalangeal junctions usually register a positive166 far-field potential (see Chapter 20-3), which may constitute the premotor potential recorded by GI as a small negativity preceding the main muscle response.64,78,244 Awareness of these possibilities helps one avoid miscalculation, especially if the nerve potential not seen with stimulation at one point appears at a second point with the use of a higher sensitivity for improved resolution. The use of the same amplifier sensitivity minimizes this type of error for comparison of successively elicited potentials with stimulation along the course of the nerve.
99
Types of Abnormalities In general, axonal damage or dysfunction results in loss of amplitude, whereas demyelination leads to prolongation of conduction time (see Chapter 4-5 and 4-6). Assessment of a nerve as a whole, as opposed to individual nerve fibers, usually reveals more complicated features because different types of abnormalities tend to coexist. Nonetheless, three basic types of abnormalities characterize motor nerve conduction studies when stimulating the nerve proximal to the presumed lesion (Fig. 5-3): (1) reduced amplitude with normal or slightly increased latency, (2) increased latency with relatively normal amplitude, and (3) absent response. In the first variety, a shock below the lesion may elicit a normal compound muscle action potential, even though proximal stimulation above the lesion evokes reduced amplitude (Fig. 5-4). This finding, if seen during the first few days of injury, fails to differentiate a partial nerve lesion causing neurapraxia or early axonotmesis before the onset of distal degeneration. Distinction between the two becomes possible by stimulating the nerve below the lesion several days after the injury, when degenerating axons will have lost their excitability. In partial neurapraxia, the distally evoked muscle response still exceeds the proximally elicited potential in amplitude. In contrast, stimulation above or below the lesion elicits an equally reduced amplitude in axonotmesis. Because the amplitude of the muscle response varies considerably from one normal subject to another, minor diminution in the recorded potential seen diffusely often escapes detection. In the second variety, slowed conduction accompanies relatively normal amplitude in stimulation above the lesion (Fig. 5-5). These changes generally imply segmental demyelination without conduction block affecting a majority of the nerve fibers. As shown in rabbits, incomplete proximal compressive lesions may also give rise to slowed conduction with a reduction in external fiber diameter distal to the site of constriction.12 The time course of recovery suggests that in these cases, conduction slowing along the dis-
100
Nerve Conduction Studies
Figure 5-3. Three basic types of alteration in the compound muscle action potential occur after a presumed nerve injury distal to the site of stimulation; mildly reduced amplitude with nearly normal latency (top), normal amplitude with substantially increased latency (middle), or absent response even with a shock of supramaximal intensity (bottom).
tal nerve segment results from distal paranodal demyelination.13 With neurapraxia, proximal stimulation above the lesion gives rise to a smaller compound muscle action potential than does a distal stimulation below the lesion (Figs. 5-6 and 5-7). A reduction in size of the compound muscle action potential may also result from
phase cancellation between peaks of opposite polarity based on pathologically increased temporal dispersion 165 in the absence of a conduction block. Such an excessive temporal dispersion commonly develops in acquired demyelinative neuropathies (Fig. 5-8). If the distal and proximal responses assume dissimilar wave-
Figure 5-4. Mild reduction in amplitude of the compound muscle action potential with a nearly normal latency. This type of abnormality indicates that a substantial number of axons remain functional. The affected axons, constituting only a small portion of the total population, have either neurapraxia or axonotmesis. The normal latency reflects the surviving axons that conduct normally. Because of inherent individual variability, minor changes in amplitude may escape detection as a sign of major abnormality.
Figure 5-5. Increased latency of the compound muscle action potential with normal amplitude. This type of abnormality indicates demyelination affecting the majority of nerve fibers, as in a compression neuropathy. Conduction block, if present during acute stages, will also diminish the amplitude of the recorded response.
Figure 5-6. A 67-year-old man with an acute onset of footdrop after chemotherapy and radiation treatment of prostate cancer. Although epidural metastasis was suspected clinically because of backache, nerve conduction studies showed evidence of a conduction block at the knee, indicating a compression neuropathy. [From Kimura,163 with permission.]
101
102
Nerve Conduction Studies
Figure 5-7. A 34-year-old man with selective weakness of foot dorsiflexors and low back pain radiating to the opposite leg. Nerve conduction studies revealed a major conduction block between the two stimulation sites, b and b', at the knee. The weakness abated promptly when the patient refrained from habitual leg crossing. [From Kimura,163 with permission.]
forms, their onset latencies may represent two groups of motor fibers with different conduction characteristics, precluding accurate calculation of velocity. A prolonged latency or slowing of the conduction velocity may also result from axonal neuropathy95 with loss of the fastconducting fibers. A major reduction in amplitude to less than 40-50 percent of the mean of the normal value usually accompanies this type of slowing. In fact, if the amplitude remains more than 80 percent of the control value, a reduction of the conduction velocity to less than 80 percent of lower limits of normal suggests the presence of demyelination. With a further diminution of amplitude to less than half the mean normal value, the conduction velocity may fall to 70 percent of the lower limit without demyelination. For the same reason, slowed motor conduction also results from loss of large anterior horn cells in myelopathies. Here, the motor conduction velocity can decrease to 70 percent of the mean normal value with diminution of amplitude to less than 10 percent of normal.180 Regardless of the amplitude, however, a conduction velocity reduced to less than 60 percent of the
mean normal value suggests 181 peripheral nerve disease, not myelopathy. Absent or very small proximal responses indicate that most of the nerve fibers fail to conduct across the site of the presumed lesion (Fig. 5-9). One must then differentiate a neurapraxic lesion from nerve transection. In either case, nerve stimulation distal to the lesion elicits an entirely normal muscle action potential for the first 4-7 days. During the second week, however, the normal distal excitability distinguishes neurapraxic changes from axonal abnormalities. With neurotmesis, stimulation below the point of the lesion produces no muscle action potentials, because of the initial failure at the neuromuscular junction (Fig. 5-10). The loss of nerve excitability follows during subsequent wallerian degeneration. Serial electrophysiologic studies help delineate progressive recovery from severe axonopathy on the basis of the amplitude of the evoked potential (Fig. 5-11). Single stimulation may also evoke various types of delayed responses usually representing focal reexcitation of hyperexcitable axons or abnormalities of the neuromuscular junction (see Chapter 183).316 Stimulation applied at different lev-
Principles and Variations of Nerve Conduction Studies Lt Ulnar Nerve
103 Rt Ulnar Nerve
Lt Tibial Nerve
Figure 5-8. A. A 31-year-old man with the Guillain-Barre syndrome. Stimulation of the ulnar nerves at the elbow or wrist elicited delayed, temporally dispersed compound muscle action potentials of the abductor digiti minimi bilaterally. Four consecutive trials at each stimulus site confirmed the consistency of the evoked potentials. B. Compound muscle action potential in the same patient as shown in A. Stimulation of the tibial nerve at the knee or ankle elicited delayed and very irregular compound muscle action potentials of the abductor hallucis.
els combined with the collision method helps clarify the origin of stimulus-induced high frequency discharges,246,271,288 Other causes of repetitive muscle action potentials316 include intramuscular nerve reex-
citation,270 excess acetylcholine or acetylcholinesterase inhibition88 at the neuromuscular junction (see Chapter 27-4) and neuromyotonia185 and related disorders (see Chapter 29-4).
104
Nerve Conduction Studies
Figure 5-9. No evoked potential with supramaximal stimulation of the nerve proximally. This type of abnormality indicates the loss of conduction in the majority of axons but fails to distinguish neurapraxia from axonotmesis or neurotmesis.
5
SENSORY NERVE CONDUCTION
Stimulation and Recording For sensory conduction studies in the upper limbs, stimulation of the digital nerves elicits an orthodromic sensory potential at a more proximal site. Alternatively, stimulation of the nerve trunk proximally evokes
the antidromic digital potential distally and mixed nerve potential proximally. For example, shocks applied to the ulnar or median nerve at the wrist give rise to an action potential along the nerve trunk at the elbow. Sensory fibers with large diameters have lower thresholds and conduct faster than motor fibers by about 5-10 percent.60 Thus, mixed nerve potentials allow determination of the fastest sensory nerve conduction velocity in healthy subjects and in
Figure 5-10. Nerve excitability distal to the lesion in neurotmesis or substantial axonotmesis. Distal stimulation elicits a normal compound muscle action potential during the first few days after injury, even with a complete separation of the nerve. Unlike neurapraxia, wallerian degeneration subsequent to transection will render the distal nerve segment inexcitable in 3 or 4 days.
Principles and Variations of Nerve Conduction Studies
105
Figure 5-11. A 21/2-year old boy with hypothermia-induced axonal polyneuropathy after prolonged exposure to severe freezing weather on a frigid winter night in Iowa. A. Compound muscle action potentials recorded over the thenar eminence after stimulation of the median nerve at the wrist (W), elbow (E), or axilla (A). The initial study on March 17, 1986, revealed no response on either side, followed by progressive return in amplitude and latency, with full recovery by January 8, 1987. B. Compound muscle action potential recorded from the abductor hallucis after stimulation of the tibial nerve at the ankle (A) or knee (K). The studies on March 17 and May 7, 1986, revealed no response on either side, with full recovery by January 8 1987.
patients with neuropathies affecting 207 motor fibers more than the sensory axons. This relationship, however, may not always hold in disease states that affect sensory fibers selectively. Such circumstances would preclude differentiation between the sensory and motor components of mixed nerve potentials.
For routine clinical recordings, surface electrodes provide adequate and reproducible information noninvasively.6,96 Some electromyographers prefer needle recording to improve the signal-to-noise ratio, especially in assessing temporal dispersion.141,142,173,176,242,269,311 Here a signal averager provides a sensitive mea-
106
Nerve Conduction Studies
Figure 5-11 (Cont.). C. Motor conduction studies of the tibial nerve on May 17, 1986. Stimulation at the knee elicited no response in the intrinsic foot muscle on either side (top three tracings), but a small compound action potential in the gastrocnemius bilaterally (bottom) as the result of early reinnervation. D. Antidromic sensory nerve action potential recorded from the second digit after stimulation of the median nerve at the wrist (W) or elbow (E). The studies on March 17 and 4May 7, 1986, showed no response on either side, with full recovery by January 8, 1987. [From Afifl et al, with permission.]
sure of early nerve damage by defining small late components that originate from demyelinated, remyelinated, or regenerated fibers.34,114,269 Minimum conduction velocity calculated from these late components normally, averages 15 m/s corresponding281to the fibers of about 4 mm in diameter. A reduction in minimum conduction velocity seves as a sensitive measure of neuropathy, often showing otherwise undetectable abnormalities.140 The technique of near nerve recording also provides unique opportunity to establish physiologic characteristics of var-
ious skin and muscle afferents in humans.44,45,47,173,175,284 For example, this method allows selective recording from nerve fibers with similar functional characteristics excited by a mechanical stimulus that46mostly activates Meissner's corpuscles or by vibration, which presumably drives pacinian corpuscles.126 In contrast, the conventionally recorded orthodromic compound sensory action potentials result from activation of all the large-diameter fibers excited by supramaximal electrical shocks that bypass the receptors and terminals axons.
Principles and Variations of Nerve Conduction Studies
107
ing the electrodes at this fixed distance. The antidromic digital potential recorded with a pair of ring electrodes has no iniWith the use of surface electrodes, the an- tial positive phase, clearly seen in the ortidromic potentials from digits generally thodromic potential. The lack of potential have a greater amplitude than the orthodifference between GI and G2 implies the dromic response from the nerve trunk, bestationary character of the positive phase cause the digital nerves lie nearer to the 'along the digit, as predicted by the farsurface.33 The relationship reverses with field theory (see Chapter 20-3). the use of needle electrodes placed near The amplitude of the sensory potential, the nerve. Some motor axons have thresh- measured either from the baseline to the olds similar to 112 those of large myelinated negative peak or between the negative and sensory axons. In studying the mixed positive peaks, varies substantially among nerve, therefore, superimposition of action subjects and to a lesser extent between the potentials from distal muscles may obtwo sides in the same individual. The same scure antidromically recorded sensory po- degree of variability occurs in recording tentials. Stimulation distal to the termina- with surface or needle electrodes.33 In adtion of the motor fibers selectively activates dition to the density of sensory innervasensory fibers of mixed nerves (see Chap- tion, body mass index, as a measure of ter 6-2). Moving more proximally, overlap- the depth of the nerve from the skin surping muscle action potentials, if any, be- face, determines the amplitude of the 36 come apparent by an abrupt change in nerve action potentials. Women tend to waveform.160 have greater sensory nerve action potenThe position of the recording electrodes tials than men for a yet undetermined 24,142,189 alters the waveform of a sensory nerve acreason, but possibly because the tion potential.8,256,263,335 An initially pos- nerves lie more superficially. Left-handers itive triphasic waveform characterizes the often have greater median sensory potenorthodromic potential recorded with an tials at214 the wrist on the right side, and vice active electrode (G1) on the nerve and a versa. Most electromyographers meareference electrode (G2) at a remote site. sure the duration of the negative-positive A separate late phase may appear in the diphasic antidromic potential from the initemporally dispersed response recorded at tial deflection to the intersection between a more proximal site. Placing G2 near the the descending phase and the baseline. nerve at a distance of more than 3 cm Some use the negative or positive peak as from G1 makes the recorded potential the point of reference, and still others resort to the less definable point where the tetraphasic, with addition of the final negativity.33 Bipolar recording register a sig- tracing finally returns to baseline. nal as the potential difference between GI The types of abnormalities described for and G2 when the impulse propagates un- motor conduction apply in principle to der the electrodes. Assuming a conducsensory conduction as well. Substantial tion velocity of 50 m/s and signal duraslowing in conduction velocity implies detion of 0.8 ms, a 4 cm interelectrode myelination of the sensory fibers, whereas distance allows the impulse to pass the axonotmesis results in reduced amplitude GI site before being picked up at the G2 of the compound nerve action potentials location. Thus, the waveform distortions with stimulation either distally or proxiare least 83,116,187 with the 4 cm interelectrode sep- mally to the site of the lesion. Sural nerve aration. Theoretical considera- potential serves as a sensitive measure for tion notwithstanding, some favour the use length-dependent distal axonal polyneuof a 3 cm over a 4 cm separation for two ropathy.5 In patients with neuropathy, practical reasons; less noise and easier sural to radial nerve sensory potential raapplication when recording from short tio often falls below 0.40, compared to the digits.323 We prefer a 2 cm separation, as mean of 0.71.272 The sensory fibers dedo many others, to be consistent with our generate only with a lesion distal to the sensory ganglion (Fig. 5-1 ID). Thus, the normative values established using commercially available recording bars mount- presence of distal sensory potential serves
Waveform, Amplitude, and Duration
Nerve Conduction Studies
108
as a criterion for differentiating preganglionic root avulsion from plexopathy.113 Intra-spinal canal lesions such as radiculopathy, however, could involve the ganglion or postganglionic portion of the157,192 root affecting the digital nerve potential. Distinction from plexopathy then depends on the distribution of sensory involvement. Plexopathy tends to affect multiple digits, whereas radiculopathy will show selective change of the first digit by C6, the second and third digits by C7, and the fourth and fifth digits by C8 root lesions.97 This type of assessment must take into account the relative amplitude values of the sensory action potential for each digit.55,56,231
Latency and Conduction Velocity Unlike motor latency, which includes neuromuscular transmission, sensory latency consists only of the nerve activation and conduction time from the stimulus point to the recording electrode. Therefore, stimulation of the nerve at a single site suffices for calculation of conduction velocity. The latency of activation, or a fixed delay of about 0.15 ms at the stimulus site174 makes the calculated conduction velocity slightly slower with the use of measured latency from a stimulus to a recording site compared to the latency difference between two recording sites flanking the same nerve segment. In measuring the latency of the orthodromic sensory potentials, some electromyographers use the initial positive peak and others the subsequent negative peak, as the point of reference.149 Sensory potentials elicited by stimulation at different sites vary in waveform because of temporal dispersion between fast and slow fibers. The interval between the positive and negative peaks also increases in proportion to the nerve length tested. Therefore, the conduction velocity calculated with the latency to the negative peak does not necessarily relate to the fastest conducting sensory fibers. The measurement to the negative peak circumvents the technical problems of identifying the preceding smaller positive peak, especially in diseased nerves.110 In
this practice, the conduction distance determined to the midpoint of GI and G2, rather than to GI itself, compensates for the discrepancy between the arrival of the impulse and the appearance of the negative peak.73 The use of modern amplifiers with high resolution now makes it feasible in most cases to measure the sensory latency to the initial positive peak. Determining the conduction distance from the stimulus point to G1 then allows accurate calculation of conduction velocity of the fastest fibers.33 With the biphasic digital potential recorded antidromically, the onset latency measured to the initial take-off of the negative peak corresponds to the conduction time of the fastest fibers from the cathode to GI. The use of the peak latency has some justification as a quick estimate of abnormal temporal dispersion, which increases the duration of the evoked potential. Mesuring both the onset latency and duration, however, provides more complete data especially with easily detectable antidromic digital potentials, which considerably exceed orthodromic potentials in amplitude. In one study, antidromic conduction times, despite identical mean values, showed slightly higher standard deviations than orthodromic measurements.33 In another study, the orthodromic recording revealed a shorter distal latency than the antidromic method in both median and ulnar nerves.51 For the same segment of the sensory nerve, however, the orthodromic and antidromic potentials recorded using the same interelectrode 53distance have the identical latencies.
6
NERVE CONDUCTION IN THE CLINICAL DOMAIN
The validity of the calculated nerve conduction velocity depends primarily on the accuracy in determining the latencies and the conduction distances. Sources of error in measuring latencies include unstable or incorrect triggering of the sweep, poorly defined take-off of the evoked response, inappropriate stimulus strength, and inaccurate calibration. Errors in es-
Principles and Variations of Nerve Conduction Studies timating the conduction distance by surface measurement result from uncertainty as to the exact site of stimulation and the nonlinear course of the nerve segments. Surface determination of the nerve length yields particularly imprecise results when the nerve takes an angulated path, as in the brachial plexus or across the elbow or knee. Because of these uncontrollable variables, the calculated values only approximate the true nerve conduction velocities. On repeated testing, the results may vary as much as 5-10 m/s because of the limitations inherent in the technique (see Chapter 7-6).171 Changes in limb temperature in part account for this variability.23 Strict adherence to the standard procedures minimizes the error, improves the reproducibility, and helps establish a small range of normal values, which justify the use of conduction studies as a diagnostic study. Unlike conduction velocity, latency comparison calls for a constant distance between the stimulating and recording electrodes. A number of factors, listed below, can modify the results of motor and sensory conduction studies. A combined index improves diagnostic classification over use of single test results.264 Analyzing multiple measurements, however, poses statistical problems, necessitating a technique for data reduction (see Chapter 3-8).265 The common assumption that conduction values follow a normal, bellshaped Gaussian distribution appears unwarranted.41 If so, calculation of reference values as the mean ± 2 (or 3) standard deviations, for example, must use the optionally transformed data to remove the effect of skew and unacceptable rate of misclassification (see Chapter 3-8).267
Physiologic Variation Among Different Nerve Segments Both motor and sensory fibers conduct substantially more slowly in the legs than in the arms. A small reduction in temperature cannot account for the recorded differences, ranging from 7 to 10 m/s.167,304 Longer nerves generally conduct more slowly than shorter nerves, as suggested by an inverse relationship between height
109
and nerve conduction velocity.40,196,336,340 Available data further indicate a good correlation between conduction velocity and estimated axonal length in peroneal and sural nerves, but not in motor or sensory fibers of the median nerve.290 These findings might suggest, without histologic proof, abrupt distal axonal tapering in the lower limbs. The other factors possibly responsible for the velocity gradient include progressive reduction in axonal diameter, shorter internodal distances, and lower distal temperatures. Statistical analyses of conduction velocities show no difference between median and ulnar nerves or between tibial and peroneal nerves. These measures also show a high degree of symmetry with only small 29side-to-side differences (see Chapter 6). The nerve impulse propagates faster in the proximal than in the distal nerve segments.111,134 For example, the most proximal motor nerve conduction velocity determined by F-wave latency clearly exceeds the conventionally54,85,158,164,169 derived most distal conduction velocity. Statistical analyses show no significant difference between cord-to-axilla and axilla-to-elbow segments.158 The F ratio (see Chapter 185) compares the proximal and distal motor nerve conduction time from the 159 stimulus site at the elbow or the knee. In healthy subjects, faster proximal conduction compensates for the difference in length between the cord-to-elbow and elbow-to-muscle segments or between the cord-to-knee and knee-to-muscle segments.164 Hence, approximately equal conduction time along the proximal and distal segments from the site of stimulation makes this ratio close to unity.
Effects of Temperature Lower temperatures slow down impulse propagation while at the same time augmenting the amplitude of nerve and muscle potential, as demonstrated in the squid axon,139 and in human studies.68,182,186,249 For example, distal latencies increase by 0.3 ms per degree for both median and ulnar nerves upon cooling the hand.42 These principles apply for both normal and demyelinated fibers as a straightforward
110
consequence of the temperature coefficients governing voltage-sensitive sodium (Na+) and potassium (K+) conductance. In particular, cold-induced slowing of sodium channel opening and a delay of its inactivation probably account for the slowing of conduction and the increase in amplitude. A parallel temperature-dependent202change also affects the refractory period. In contrast, nerve impulses conduct faster at a higher body temperature,61 as is seen, for example, after physical activity. 125 The conduction velocity increases almost linearly, by 2.4 m/s, or approximately 5 percent per degree, as the temperature measured near135,148 the nerve increases from 29° to 38 °C. Conduction velocity changes nonlinearly with increase in skin temperature, showing more pronounced effect in the lower temperature range.306 Very high temperatures, however, induce a pronounced effect, decreasing motor and sensory potentials by 27 percent and 50 percent in amplitude, and 19 percent and 26 percent in duration with warming
Nerve Conduction Studies of the limb from 32 °C to 42 °C.273 In demyelinated nerve fibers, conduction velocity increases as temperature rises until propagation ceases at a vulnerable site. The nerve conducts faster, reflecting quick activation of Na+ channels over a length of a fiber that is normal except for a short segment of demyelination. Conduction fails at a particular site with a low safety factor when the rising temperature reduces the 102,326 action potential below the critical level. Thus, the latency and amplitude measure two completely separate effects of change in temperature. Studies conducted in a warm room with ambient temperature maintained between 21° and 23 °C reduce this type of variability. Although impractical and unnecessary in clinical practice, a warmer room at 26°-28 °C or even 30 °C minimizes the temperature gradient along the course of a nerve.154 To check the intramuscular temperature, the insertion of a thermometer through the skin requires an additional puncture for each muscle tested. In prac-
Figure 5-12. Relation of age to conduction velocity of motor fibers in the ulnar nerve between elbow and wrist. Velocities in normal young adults range from 47 to 73 m/s, with most values between 50 and 70 m/s. Ages plotted indicate the month after the expected birth date based on calculation from the first day of last menstruation. [From Thomas and Lambert,302 with permission.]
111
Principles and Variations of Nerve Conduction Studies Table 5-1 Normal Motor Nerve Conduction Velocities (M/S) in Different Age Groups Age 0-1 week 1 week to 4 months 4 months to 1 year 1-3 years 3-8 years 8-16 years Adults
Ulnar 32 (21-39) 42 (27-53) 49 (40-63) 59 (47-73) 66 (51-76) 68 (58-78) 63 (52-75)
Peroneal 29 (19-31) 36 (23-53) 48 (31-61) 54 (44-74) 57 (46-70) 57 (45-74) 56 (47-63)
Median 29 (21-38) 34 (22-42) 40 (26-58) 50 (41-62) 58 (47-72) 64 (54-72) 63 (51-75)
Source: From Gamstorp,107 with permission.
tice, the skin temperature measured with a plate thermistor correlates linearly with the subcutaneous and intramuscular temperatures.123,124 A skin temperature of 34 °C or above indicates a muscle temperature close to 37 °C.69 A measured value falling below 32 °C calls for warming of the limb with an infrared heat lamp or by its immersion in warm water for a sufficient time in the order of 30 minutes.101 Alternatively, one may add 5 percent of the calculated conduction velocity for each degree below 32 °C to normalize the result. Such conversion factors, based on an average of many healthy subjects, however, may provide misleading interpretations in diseases of the peripheral nerve.11,19,40,71,237
Maturation and Aging Nerve conduction velocities increase rapidly as the process of myelination advances from roughly half the adult value in fullterm infants302 to the adult range at age 3-5 years (Fig. 5-12). Conduction velocity of slower fibers also show a similar time course of maturation.122 Table 5-1 summarizes the results of one series showing a steep increase in conduction of the peroneal nerve through infancy and a slower maturation of the median nerve during early childhood.107 Premature infants have even slower conduction velocities, ranging from 17 to 25 m/s in the ulnar nerve and from 14 to 28 m/s in the peroneal nerve.49 The
Table 5-2 Normal Sensory and Motor Nerve Conduction Velocities (M/S) in Different Age Groups Nerve Median nerve Digit-wrist Wrist-muscle Wrist-elbow Elbow-axilla Ulnar nerve Digit-wrist Wrist-muscle Wrist-elbow Elbow-axilla Common peroneal nerve Ankle-muscle Ankle-knee Posterior tibial nerve Ankle-muscle Ankle-knee H reflex, popliteal fossa
Age 1O-35 Years (3O Cases) Motor Sensory
67.5 ±4.7 67.7 ±4.4 70.4 ±4.8
Age 36-5O Tears (16 Cases) Sensory Motor
65.8 ±5.7 3.2 ±0.3* 59.3 ±3.5 65.9 ±5.0
65.8 ±3.1 70.4 ±3.4
64.8 ±3.8 69.1 ±4.3
2.7 ± 0.3* 58.9 ±2.2 64.4 ±2.6
53.0 ±5.9
4.3 ±0.9* 49.5 ±5.6
45.5 ±3.8 71.0 ±4.0 27.9 ±2.2*
'Latency in milliseconds. Values are means ± 1 standard deviation. Source: From Mayer220 with permission.
62.8 ±5.4 66.2 ±3.6
3..5 ± 0.2* 54..5 ± 4.0 63..6 ± 4.4
57.5 ±6.6
67.1 ±4.7 70.6 ±2.4
2.7 ± 0.3* 57.8 ±2.1 63.3 ± 2.0
56.7 ±3.7 64.4 ±3.0
3,.0 ± 0.35* 53.,3 ± 3.2 59..9 ± 0.7
50.4 ± 1.0
4.8 ± 0.5* 43.6 ±5.1
46.1 ±4.0
4..6 ± 0.6* 43.,9 ± 4.3
49.0 ±3.8
7.3 ± 1.7* 42.9 ± 4.9 64.0 ±2.1 28.2 ± 1.5*
5.9 ± 1.3*
56.9 ±4.4
59.4 ±4.9
3.7 ± 0.3* 55.9 ± 2.6 65.1 ±4.2
66.5 ±3.4
64.7 ±3.9
Age 51-80 Tears (18 Cases) Sensory Motor
48.9 ±2.6
6.,0± 1.2* 41..8 ±5.1 60..4 ± 5.0 32.,0±2.1*
Nerve Conduction Studies
112
Table 5-3 Comparison of Conduction Studies Between Younger Group (n = 52, 1O-40 Years) and Older Group (n = 52, 41-84 Years) Nerve Tested Peroneal M amplitude M latency MNCV F latency FWCV F number Tibial M amplitude M latency MNCV F latency FWCV F number H amplitude H latency Sural S amplitude S latency SNCV
No. of Nerves
Age 29.7 ± 6.9 Years (Mean ± SD)
No. of Nerves
Age 54.0 ± 10.5 Years (Mean ± SD)
P Value
(mV) (ms) (m/s) (ms) (m/s) (#)
104 104 104 44 44 10
5.4 3.7 49.5 47.1 60.6 8.5
± 1.5 ± 0.9 ± 5.4 ±5.3 ± 7.7 ± 1.7
98 98 98 42 42 29
5.0 ± 1.3 3.7 ± 0.7 47.8 ± 3.8 47.6 ± 4.9 59.9 ± 7.6 9.7 ±3.1
0.03* 0.98 0.01* 0.68 0.66 0.19
(mV) (ms) (m/s) (ms) (m/s) (#) (mV) (ms)
104 104 104 74 74 25 53 53
6.7 ± 2.0 3.5 ±0.6 48.6 ± 4.2 47.9 ±4.1 58.3 ± 6.2 11.6 ±3.4
1.4 ±0.8 29.8 ± 2.3
100 100 100 74 74 27 43 50
5.9 ± 1.5 3.6 ± 0.6 49.1 ±4.9 48.3 ± 4.6 57.5 ± 6.8 12.4 + 2.6 1.2 ± 0.8 30.7 ± 2.0
0.001* 0.23 0.52 0.63 0.49 0.39 0.20 0.04*
(^V) (ms) (m/s)
53 53 53
20.9 ± 8.0 2.7 ± 0.3 52.5 ± 5.6
50 50 50
17.2 ± 6.7 2.8 ± 0.3 51.1 ±5.9
0.01* 0.16 0.23
MMCV, motor nerve conduction velocity in the distal segment; FWCV, F-wave conduction velocity in the proximal segment; F number, number of responses out of 16 trials. *Amplitude was significantly reduced in the older group for all the nerves tested, whereas measures of conduction showed no changes except for peroneal MNCV and tibial H latency. Source: From Kumura,163 with permission.
values at 23-24 weeks of fetal life average roughly one third those of newborns of normal gestational age.57,225,286 In premature infants, motor and proprioceptive conduction show a different time course of maturation when studied on the expected date of birth.26 Fetal nutrition may alter peripheral nerve function by influencing myelin formation.268 In children and adolescents, from age 3 to 19 years, both motor and sensory conduction velocities tend to increase slightly in the upper limb and decrease in the lower limb183 as a function of age and growth in length. Conduction velocities begin to decline after 30-40 years of age, but the values normally change301,322 by less than 10 m/s by the sixtieth year or even the eightieth year.236 The most distal branches, such as the interdigital nerves, may degenerate earlier.188 Table 5-2 summarizes the results of one study220 showing a reduction in the mean conduction rate of about 10 percent at 60 years of age. Aging also causes a diminution in
amplitude and changes in the 94 shape of the evoked potential (Table 5-3), especially when recorded55,56 across the common sites of compression. The latencies of the F wave and somatosensory evoked potentials also72 gradually increase with advancing age, probably reflecting preferential loss of the largest and fastest conducting motor units.327
Height and Other Factors In addition to temperature and age, other factors that influence nerve conduction measures274,292 include anthropometric characteristics. For example, height shows negative association with sensory amplitude and positive association with distal latencies. Sural, peroneal and tibial nerve conduction velocities all have inverse correlation with height in normals261 and 106 in patients with diabetic neuropathy. Women have faster conduction velocity and greater amplitude for both motor and
Principles and Variations of Nerve Conduction Studies sensory studies than men.266 Most gender differences resolve when adjusted by height, whereas amplitude differences persist despite such correction. In one study312 dealing with sural nerve conduction velocity, the changes attributable to height fell within the experimental error of 2.3 percent expected from the method. Ischemia induced by a pneumatic tourniquet alters nerve excitability substantially, with progressive slowing in conduction velocity, decrease in amplitude, and increase in duration of the action potential.277 These changes affect the median nerve more rapidly in patients with carpal tunnel syndrome than in normal control subjects.105 Conversely, patients with diabetes or uremia or elderly subjects have a greater resistance to ischemia with regard to peripheral nerve function.43 Threshold tracking provides confirmatory evidence for the ischemic resistance in motor axons of diabetic subjects (see Chapter 8-3).331 In chronic hypoxemia and diabetes, reduction in amplitude of nerve potential during ischemia shows a time course correlated with the blood oxygen saturation. Thus, hypoxic exposures may induce resistance to ischemic conduction failure.127 Animal studies in rats suggest that both glucose and insulin also play an important role.245
Clinical Values and Limitations Over the years, nerve conduction studies have made major contributions to the understanding of peripheral nerve function in health and disease states.115 Such evaluations can precisely delineate the extent and distribution of the lesion, providing an overall distinction between axonal and demyelinating involvement.303 This dichotomy provides a simple and practical means of correlating conduction abnormalities with major pathologic changes in the nerve fibers. In support of this concept, in vitro recordings from the sural nerve have clearly delineated close relationships between histologic and physiologic findings.20 In addition to such a broad classification, the pattern of nerve conduction abnormalities can often characterize the general nature of the clinical disorder. For example,
113
hereditary demyelinating neuropathies commonly show diffuse abnormalities, with little difference from one nerve to another in the same patient and among different members in the same family.195 Approximately equal involvement of different nerve fibers limits the degree of temporal dispersion despite a considerably increased latency. In contrast, acquired demyelination tends to affect certain segments of the nerve disproportionately,159,169 giving rise to more asymmetric abnormalities and substantial increases in temporal dispersion. Pattern of distribution in sensory nerve conduction abnormalities also helps differentiate demyelinating and axonal polyneuropathies. For example, a reduced median amplitude compared with the sural amplitude supports the diagnosis of a primary demyelination.28 In contrast, a reduced sural amplitude compared with the radial amplitude implies axonal polyneuropathy.272 Optimal application of the nerve conduction study depends on an understanding of the principles and a recognition of the pitfalls of the technique. The conventional methods deal primarily with distal nerve segments in the four limbs. Special techniques enable assessment of nerve segments in less accessible anatomic regions for better evaluation of a focal lesion, and improved detection of subclinical abnormalities. Despite certain limitations, these methods can provide diagnostically pertinent information if used judiciously in appropriate clinical contexts. 7
STUDIES OF THE AUTONOMIC NERVOUS SYSTEM
Electrophysiologic evaluations of the sympathetic and parasympathetic pathways help confirm a clinical diagnosis of autonomic neuropathy.3,254 Some studies readily performed in a clinical neurophysiology laboratory complement invasive investigations required for precise localization of the site of the lesion. Autonomic functions change with age, requiring appropriately matched control values for comparison.75,98,203
Nerve Conduction Studies
114
Noninvasive studies for cardiovascular function include heart-rate (R-R intervals) variation with104 breathing, spectral analysis of heart rate, Valsalva ratio, blood pressure, and heart-rate response to change in posture and to 99eyeball pressure for vagal overreactivity. Studies of sudomotor function consist of sympathetic skin response (SSR), thermoregulatory sweat test, 293 quantitative sudomotor axon reflex test, and sweat imprint method.155,204 Studies routinely used in a clinical neurophysiology laboratory comprise R-R intervals and SSR.278,279 Some investigators advocate power spectral analysis.198,199
Heart-Rate Variation with Breathing The heart rate increases physiologically during inspiration and decreases during expiration. The R-R intervals recorded using an electrocardiogram (ECG) test the degree of this change during deep breathing. After a resting period of 5 minutes, the patient breathes deeply, in a recumbent position, at the rate of 6 breaths per minute for 1 minute. A standard electromyographic instrument suffices to display the ECG with a surface electrode placed at the midpoint of the left clavicle 131,291 and the other electrode over the sternum. The difference between the shortest and longest R-R intervals during 1 minute serves as the most reliable method, showing little intraindividual variation, whether tested manually or using automatic methods of analysis.291,332 With deep breathing, heart rates should normally change more than 15 beats per minute. Values of less than 10 beats usually indicate an abnormality, although the result depends on the age of the subject.144,205,232,291 The expiratory/inspiratory ratio (E/I ratio) provides another measure of R-R variation defined as the mean of the maximum R-R intervals during expiration over the mean of the minimum R-R intervals during inspiration.297 Subjects younger than 40 should have an E/I ratio above 1.2, which then decreases with age.144 Transfer function analysis may also provide an easy measure of103respiratory-induced heart rate variability.
The heart rates, determined mainly by vagal activity, reveal parasympathetic function. Atropine but not propranolol blocks its increase during inspiration.150,333 Heart rate variation during breathing decreases with age, and in diabetes and other disorders affecting autonomic pathways 144,199,255,278
Valsalva Ratio The Valsalva maneuver, or a brief period of forced expiration against a closed glottis or mouthpiece, increases the heart rate by stimulating the intrathoracic stretch receptors such as the carotid sinus and aortic arch baroreceptors. The subject lies in a semirecumbent position with a rubber clip over the nose and breathes forcefully into a mouthpiece for 10-15 s, maintaining an expiratory pressure of 40 mm Hg. The Valsalva ratio, calculated by dividing the longest R-R interval after the maneuver with the shortest R-R interval during the maneuver, measures the changes of heart rate resulting from the cardiac vagal efferent and sympathetic vasomotor activity.191 The highest ratio from three successive attempts, each separated by 2 minutes,206 normally exceeds 1.4 in subjects younger than 40. The Valsalva ratio reflects both parasympathetic and sympathetic function. In addition, patients with heart and lung disease may have low values for reasons unrelated to the autonomic system.
Response to Change in Posture When a person stands from the supine position, the heart rate increases usually from 10 to 20 beats per minute. After reaching a maximum at about the fifteenth heart beat, it declines to a relatively stable rate at about the thirtieth heart beat. The ratio of the R-R intervals corresponding to the 30th and 15th heart beats, termed the 30:15 ratio, measures parasympathetic function.90 Young adults should have a ratio of more than 1.04. Atropine blocks the effect, suggesting its dependency on vagal innervation of the
Principles and Variations of Nerve Conduction Studies
115
heart.90 Heart-rate responses measured on a tilt-table also normally increase 5 to 30 beats per minute, without the biphasic response seen on standing. This change also declines with age.144
or DC to 2-3 kHz). Effective stimuli comprise a surprise element such as a loud noise delivered unexpectedly. To trigger the osciloscope sweep for latency measurement, we apply an electrical shock 0.1 ms in duration, and 10-20 mA in intensity to the ipsilateral 279 or contralateral wrist, ankle, or any digit. The temperSympathetic Skin Response ature of the limbs is maintained at In the conventional nerve conduction 32°-34° C. Randomly timed electrical stimuli over studies, unmyelinated fibers do not contribute to the surface recorded responses. the median nerve elicit a biphasic potenRecording SSR using a non-invasive tech- tial with either the initial negativity or posnique119,279 provides a means to test these ax- itivity over the palmar surface of the hand ons. A surface electrode placed on and the plantar surface of the152foot. In one the palm of the hand (Fig. 5-13) or sole study of 35 healthy subjects, mean laof the foot (Fig. 5-14) serves best as the tencies increased from the wrist to the active electrode, G1, with the reference middle phalanx but then decreased to the electrode, G2, on the dorsal surface of the distal phalanx. This finding may reflect same limb. In contrast, G1, if placed on density difference of sweat glands, which the axilla, forearm, or dorsal surface of the dictates sympathetic sudomotor nerve achand or foot, usually fails to register a re- tivity.170In one study of 30 healthy subnormal values (mean ± SD) for sponse, probably reflecting the paucity of jects, sweat glands. Recording a long latency re- palmar and plantar responses consisted sponse (Fig. 5-15) with low frequency of the onset latency of 1.52 ± 0.135 s and components requires a very slow sweep 2.07 ± 0.165 s and amplitude of 479 ± (0.5-1 s per division), a high gain (100 /mV 105 mV and 101 ± 40 mV. In another per division), and a wide band-pass (0.16 study,86 measurements of the normal
Figure 5-13. Electrical stimulation of the index finger and recording of sympathetic skin response over the palm (G1) and the dorsal surface (G 2 ) of the same hand.
Figure 5-14. Electrical stimulation of the big toe and recording of sympathetic skin responses over the sole (G1) and the dorsal surface (G2) of the same foot.
Figure 5-15. Sympathetic skin responses recorded simultaneously in four limbs of a normal subject after electrical stimulation of the left wrist. A greater latency for the foot responses than the hand responses reflects the different lengths of the descending pathways. Oscilloscope settings consisted of very slow sweep (500 ms/division), high gain (500 mV/division), and a wideband pass (0.16-3 kHz).
116
Principles and Variations of Nerve Conduction Studies mean onset latency and amplitude were 1.50 ± 0.08 s and 3.1 ± 1.8 mV for the hands, and 2.05 ±0.10 s and 3141.4 ± 0.8 mV for the feet. Neither the site nor the type of stimulation275 alters the onset latency with any consistency, which reflects not only the peripheral C fiber function but also conduction in a long multineuronal pathway. In contrast, the density of spontaneously activatable sweat glands dictates the amplitude as a measure of peripheral sympathetic activity. Lower temperatures reduce the amplitude 62and prolong the latency. In one study, cooling the whole arm as compared to the hand induced a greater effect in latency but not in amplitude. Thus, amplitude change reflects only the neuroglandular junction, whereas latency modulation also involves the postganglionic194sympathetic C fibers. In another study, a change in temperature over 32°-34 °C range increased the amplitude by 8.5 percent and decreased the latency by 2.5 percent per degree. Although SSR can occur in the absence of normal sweat gland function,18 its abnormalities in general correlate 215 reasonably well with other sweat tests and certain other measures of autonomic function 215,278,289 Its variability and rapid habituation combined with a nonquantitative nature tend to limit clinical application. Some consider only its absence or major reduction in amplitude as a definite abnormality,309,310 whereas others regard a prolonged latency as a sign of neuropathy.66 Magnetic stimulation applied to the neck evokes easily recordable, highly reproducible sympathetic skin responses, refrecting strong afferent sensory in-puts proximally. The potentials thus recorded revealed an orderly latency gradient from proximal to distral sites of all limbs.200,217,218,315 Reported onset latencies include 1.0 ±0.1 s (mean ± SD) for the arm, 1.2 ± 0.1 s for the forearm, 1.1 ± 0.1 s for thigh, 1.5 ± 0.1 s for the calf, and 1.7 ± 0.2 s for the sole.313 Iontophoresis of atropine into the skin under the recording site abolishes194,289,339 the response. Patients with diabetes, 253 scleroderma, familial amyloid polyneu283 ropathy, or sympathectomy190 have absent or reduced response on the affected limbs. This contrasts to normal autonomic
117
function in Friedreich's ataxia, which primarily involves large143myelinated fibers, sparing smaller fibers. Ischemic conduction block of the arm abolishes the previously obtainable response.314 Disorders associated with delayed or absent responses 31 include lepromatous leprosy, hereditary motor329sensory neuropathy,67 chronic uremia, and palmar hyperhidrosis.197 The SSR also reflects preganglionic sympathetic activity, providing information different from the somatic pathway in evaluating myelopathy338 and other heterogeneous systemic diseases such as 87,199,338 multiple sclerosis, amyotrophic lateral sclerosis,70 Parkinson's disease,328 299 and rheumatoid arthritis. Patients show no detectable assymetry in the foot as the 7 result of L5 or SI radiculopathies or after sural nerve biopsy.247
8 OTHER EVALUATION OF NERVE FUNCTION Microneurography Conventional nerve conduction studies provide accurate measurement of the fastest conduction velocities, as well as an approximate number of volleys and the pattern of their distribution, based on the size and waveform of the evoked response. The technique usually relies on the application of an artificially synchronized electrical stimulus that the nervous system never experiences in the natural environment. Thus, despite the established diagnostic applications, such studies rarely help elucidate the exact physiologic mechanisms underlying the clinical signs and symptoms that concern the patients most. For example, the evaluation of pain and paresthesia falls outside the conventional stimulation methods, which only detect deficits in nerve function. Similarly, the conduction studies help assess the involvement of small fibers only indirectly by localizing focal abnormalities of large axons, which may have little to do with the patient symptoms. Thus, the lack of clinical correlation becomes particularly evident when the patient has positive
118
rather than negative signs and small rather than large fiber dysfunction. These and other concerns necessitate a different approach to explore the areas not easily accessible by the ordinary means of conduction assessments. Microneurography allows recordings of impulse activity in single nerve fibers within skin or muscle nerve fascicles through tungsten microelectrodes inserted percutaneously.121 Recording of this type in an alert human subject provides a great deal of physiologic information about various types of fiber populations.37,120 Most human studies have centered on post-ganglionic sympathetic fibers 22,91-93,211,282,296 innervating autonomic effector organs. '325 Other areas of possible interest include cutaneous afferents from mechano-, thermo- and nociceptors,239,280 and muscle afferents from spindles and Golgi tendon organs. Surface stimulation of the receptive field gives rise to evoked sensory action potentials with late components, representing either the high-threshold small-diameter fibers seen in normal subjects, or the abnormally lowthreshold regenerating or demyelinating fibers seen in patients with neuropathy.177 Studies of normal subjects have substantiated the association between complex high-frequency burst and sensation of paresthesia induced by nerve compression, hyperventilation, or prolonged tetanic stimulation of cutaneous afferents.145,209,241 The findings suggest that the abnormal sensation results from ectopic discharges of hyperexcitable cutaneous afferent. Combined with intraneural microstimulation,241 the method also helps establish the direct link between impulse propagation along various primary afferents and subjective somatosensory experiences. In fact, careful stimulation of single efferent axons give rise to distinctive perception correlated with the type of cutaneous receptor in question.210,241 Microstimulation of individual muscle afferents fails to evoke a coherent sensation, but stimulation of joint afferents evokes sensation of pressure or movement in 50 percent of cases.37 In addition to physiologic studies conducted in healthy subjects, this technique can explore the pathophysiologic mecha-
Nerve Conduction Studies nisms underlying various abnormalities of the somatosensory, motor, and autonomic systems.178 Spontaneous activity identified by this method in cutaneous afferent fibers shows a good correlation to paresthesia experienced in neuropathies, neuromas, entrapment syndromes, radiculopathies, thoracic outlet syndromes, and Lhermitte's signs.38,39,234,238 High-frequency discharges also originate at the site of nerve damage, spontaneously or during and after ischemia.9,25,156 A previous impalement of a nerve by a microelectrode gives rise to similar abnormalities from discharges generated ectopically at the site of injury. These recordings typically consist of brief bursts of 2-5 spikes occurring at a frequency of 7-10 Hz with peak instantaneous frequencies usually exceeding 300 Hz.208 In the clinical context, microneurographic techniques allow recording of 240,307,308 neural activity in single C fibers or autonomic fibers.211,212,294 Despite theoretical interest in correlating cutaneous pain with neural discharges and vasoconstriction with sympathetic activity, however, the technique has limited value for electrodiagnostic purposes, primarily because the nature of recording requires the expertise not generally available in an ordinary electromyography laboratory.
Thermal, Pain, Vibratory, and Tactile Sensation The cutaneous sensory tests usually include warm and cold thermal perception, vibration, touch-presure sensation and current threshold study.10,79,337 These quantitative measures have found a limited but useful role in the characterization and quantitation of cutaneous sensory function.80,82,213 As a noninvasive, nonaversive method, the test yields reliable results even in children as young as 4 years old.137 Like those of any psychophysiological tests, however, the findings vary among different control groups—for example, between paid volunteers and laboratory personnel familiar with the procedure.259 Automated tactile testers measure threshold values for light touch, high-frequency vibration, pinprick,129warming, and two-point discrimination. Weighted needle pinprick using in-
Principles and Variations of Nerve Conduction Studies expensive apparatus may give information on small-fiber dysfunction that compares with thermal threshold determination.50 These tests may allow documentation of abnormalities in a higher percentage of patients than do more traditional clinical evaluations. Thermal thresholds tests use either the method17 of limits or a forced-choice technique. A large-scale survey in patients with diabetes193 indicates that either approach serves as a simple, noninvasive tool to evaluate small-fiber neuropathy. Quantitative assessment of thermal sensitivity may detect early small-fiber dysfunction, even if conventional electrophysiologic studies reveal no abnormalities.146 Thus, some advocate that vibratory and thermal testing should constitute the primary screening test for diabetic neuropathy.321 Nerve conduction studies, however, provide better diagnostic value than quantitative sensory testing.257,258 As expected, thermal and vibratory threshold increases in 147 proportion to the severity of neuropathy. In addition to thermal hypoesthesia, the test may reveal hyperalgesia, or the perception of temperature-induced pain preceding cold or warmth sensation as a characteristic finding of small-fiber 230 damage.128,320 In one study on diabetes, thermal and sweating tests correlated significantly with the scores of abnormal temperature and pinprick sensation obtained by physical examination, but not with the duration of the illness. Thermal sensitivity but not sweat gland number predicted the degree of motor and sensory nerve conduction abnormalities. In one experiment measuring reaction times to stimuli at two sites on the lower limb, 10° the estimated conduction velocity (mean ± SD) for cooling (2.1 ± 0.8 m/s) exceeded that for warming (0.5 ± 0.2 m/s). These figures confirm the transmission of the sensation of warming via the unmyelinated peripheral nerve fibers and that of cooling via small myelinated peripheral nerve fibers. Compared to thermal discrimination thresholds, vibratory perception tests in general show a better reproducibility.65 Such quantitative measurements also help detect minor sensory signs of central origin in patients with multiple sclerosis.132
119
Current perception threshold testing uses constant current sine wave stimulator usually at three different frequencies of 5, 250, and 2,000 Hz, which may selectively activate three subsets of nerve fibers.260 Some studies have shown good correlation of high-frequency stimulation with largefiber function, and low-frequency216stimulation with small-fiber function, but2,262 its clinical usefulness remains uncertain. Measurement of alternating current perception thresholds may improve the quantitative assessment, as shown in grading the severity of diabetic sensory neuropathy,260 and the degree of sensory52function recovery after nerve transplant. Determination of the thresholds for heat pain in the foot may help evaluate disturbances of C-fiber-mediated sensibility in lumbosacral disc disease.295 The test may also provide a quantitative means to confirm elevated heat pain thresholds, or heat hypoalgesia, which indicates advanced stages of small fiber neuropathy.229 For this test, thermal stimulation must exceed 43 °C, bearing some risk of burn injury in patients with sensory loss.136 Thermography Despite the initial enthusiasm and favorable reports concerning thermography in the carpal tunnel syndrome and many other neurological disorders, more recent studies conclude that the technique offers only limited value as a test of neural function in the clinical context. For example, a well-controlled study224 documented thermographic alterations in 0 of 9 hands with mild nerve conduction abnormalities and 7 of 14 hands with marked nerve conduction changes. Similarly, thermography provides nonspecific findings of uncertain diagnostic or prognostic relevance in the evaluation of lumbosacral radiculopathy. 130,287
REFERENCES 1. AAEE: Guidelines in Electrodiagnostic Medicine. Professional Standard Committee, American Association of Electromyography and Electrodiagnosis, Rochester, Minnesota, 1984.
Nerve Conduction Studies
120 2. AAEM: Technology Review: the neurometer current perception threshold (CPT). Muscle Nerve 22:523-531, 1999. 3. AAS and AAN: The Consensus Committee of the American Autonomic Society and the American Academy of Neurology: Consensus statement on the definition of othostatic hypotension, pure autonomic failure and multiple system atrophy. Neurology 46:1470, 1996. 4. Afifi AK, Kimura J, Bell WE: Hyopthermiainduced reversible polyneuropathy: Electrophysiologic evidence of axonopathy. Pediatric Neurology 4:49-53, 1988. 5. Albers JW: Clinical neurophysiology of generalized polyneuropathy. J Clin Neurophysiol 10:149-166, 1993. 6. Albers JW: Principles of Sensory Nerve Conduction Studies. American Academy of Neurology, 49th Annual Meeting, Boston, 1997. 7. Andary MT, Stolov WC, Nutter PB: Sympathetic skin response in fifth lumbar and first sacral radiculopathies. Electromyogr Clin Neurophysiol, 33:91-99, 1993. 8. Andersen K: Surface recording of orthodromic sensory nerve action potentials in median and ulnar nerves in normal subjects. Muscle Nerve 8:402-408, 1985. 9. Applegate C, Burke D: Changes in excitability of human cutaneous afferents following prolonged high-frequency stimulation. Brain 112: 147-164, 1989. 10. Asbury AK, Porte D, Committee: Proceedings of a consensus development conference on standardized measures in diabetic neuropathy. Neurology 42:1823-1839, 1992. 11. Ashworth NL, Marshall SC, Satkunam LE: The effect of temperature on nerve conduction parameters in carpal tunnel syndrome. Muscle Nerve 21:1089-1091, 1998. 12. Baba M, Fowler CJ, Jacobs JM, Gilliatt RW: Changes in peripheral nerve fibres distal to a constriction. J Neurol Sci 54:197-208, 1982. 13. Baba M, Gilliatt W, Jacobs JM: Recovery of distal changes after nerve constriction by a ligature. J Neurol Sci 60:235-246, 1983. 14. Bamford CR, Rothrock RJ, Swenson M: Average techniques to define the low-amplitude compound motor action potentials. Arch Neurol 41:1307, 1984. 15. Bannister RG, Sears TA: The changes in nerve conduction in acute idiopathic polyneuritis. J Neurol Neurosurg Psychiatry 25:321-328, 1962. 16. Barkhaus PE, Nandedkar SD: Recording characteristics of the surface EMG electrodes. Muscle Nerve 17:1317-1323, 1994. 17. Bartlett G, Stewart JD, Tamblyn R, Abrahamowicz M: Normal distributions of thermal and vibration sensory thresholds. Muscle Nerve 21:367-374, 1998. 18. Base SM, Meer J, Polinsky RJ, Hallett M: Sudomotor function in autonomic failure. Neurology 41:1564-1566, 1991. 19. Baysal AI, Chang C-W, Oh SJ: Temperature effects on nerve conduction studies in patients with carpal tunnel syndrome. Acta Neurol Scand 88:213-216, 1993. 20. Behse F, Buchthal F: Sensory action potentials
21. 22.
23. 24.
25. 26.
27.
28.
29. 30.
31.
32.
33. 34.
35. 36. 37.
38.
and biopsy of the sural nerve in neuropathy. Brain 101:473-493, 1978. Berry CM, Grundfest H, Hinsey JC: The electrical activity of regenerating nerves in the cat. J Neurophysiol 7:103-115, 1944. Birkett CL, Ray CA, Anderson EA, Rea RF: A signal-averaging technique for the analysis of human muscle sympathetic nerve activity. J Appl Physiol 73:376-381, 1992. Bleasel AF, Tuck RR: Variability of repeated nerve conduction studies. Electroencephalogr Clin Neurophysiol 81:417-420, 1991. Bolton AF, Carter K: Human sensory nerve compound action potential amplitude: variation with sex and finger circumference. J Neurol Neurosurg Psychiatry 43:925-928, 1980. Bostock H, Bergmans J: Post-tetanic excitability changes and ectopic discharges in a human motor axon. Brain 117:913-928, 1994. Bougie D, Denise P, Yaseen H, Tranier S, Voirin J, Pettier M, Venezia R: Maturation of peripheral nerves in preterm infants. Motor and proprioceptive nerve conduction. Electroencephalogr Clin Neurophysiol 75:118-121, 1990. Brashear A, Kincaid JC: The influence of the reference electrode on CMAP configuration: Leg nerve observations and an alternative reference site. Muscle Nerve 19:63-67, 1996. Bromberg MB, Albers JW: Patterns of sensory nerve conduction abnormalities in demyelinating and axonal peripheral nerve disorders. Muscle Nerve 16:262-266, 1993. Bromberg MB, Jaros L: Symmetry of normal motor and sensory nerve conduction measurements. Muscle Nerve 21:498-503,1998. Bromberg MB, Spiegelberg T: The influence of active electrode placement on CMAP amplitude. Electroencephalogr Clin Neurophysiol 105: 385-389, 1997. Brown TR, Kovindha A, Wathanadilokkol U, Smith T, Kraft GH: Abnormalities of the sympathetic skin response in lepromatous leprosy. Muscle Nerve 19:1357-1358, 1996. Buchthal F, Rosenfalck A: Action potentials from sensory nerve in man: Physiology and clinical application. Acta Neurol Scand (suppl 13) 41:263-266, 1965. Buchthal F, Rosenfalck A: Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 3:1-122, 1966. Buchthal F, Rosenfalck A, Behse F: Sensory potentials of normal and diseased nerves. In Dyck PJ, Thomas PK, Lambert EH (eds): Peripheral Neuropathy, Vol 1. WB Saunders, Philadelphia, 1975, pp 442-464. Buchthal F: Action potentials in the sural nerve evoked by tactile stimuli. Mayo Clin Proc 55: 223-230, 1980. Buschbacher RM: Body mass index effect on common nerve conduction study measurement. Muscle Nerve 21:1398-1404, 1998. Burke D: Microneurography, impulse conduction and paresthesias. Seventeenth Annual Edward H. Lambert Lecture, American Association of Electrodiagnostic Medicine, October 1992. Burke D, Applegate C: Paresthesiae and hypaesthesia following prolonged high-frequency
Principles and Variations of Nerve Conduction Studies stimulation of cutaneous afferents. Brain 112: 913-929, 1989. 39. Campero M, Serra J, Marchettini P, Ochoa JL: Ectopic impulse generation and autoexcitation in single myelinated afferent fibers in patients with peripheral neuropathy and positive sensory symptoms. Muscle Nerve 21:1661-1667, 1998. 40. Campbell WW, Ward LC, Swift TR: Nerve conduction velocity varies inversely with height. Muscle Nerve 3:436-437, 1981. 41. Campbell WW, Robinson LR: Deriving reference values in electrodiagnostic medicine. Muscle Nerve 16:424-428, 1993. 42. Carpendale MTF: Conduction time in the terminal portion of the motor fibers of the ulnar, median, and peroneal nerves in healthy subjects and in patients with neuropathy. Thesis, University of Minnesota, Minneapolis, 1956. 43. Caruso G, Labianca O, Ferrannini E: Effect of ischemia on sensory potentials of normal subjects of different ages. J Neurol Neurosurg Psychiatry 36:455-466, 1973. 44. Caruso G, Massini R, Crisci C, Nilsson J, Catalano A, Santoro L, Battaglia F, Crispi F, Nolano M: The relationship between electrophysiological findings, upper limb growth and histological features of median and ulnar nerves in man. Brain 115:1925-1945, 1992. 45. Caruso G, Nilson J, Crisci C, Nolano M, Massini R, Lullo F: Sensory nerve findings by tactile stimulation of median and ulnar nerves in healthy subjects of different ages. Electroencephalogr Clin Neurophysiol 89:392-398, 1993. 46. Caruso G, Nolano M, Crisci C, Lullo F, D'Addio G: Mechanical vs. electrical stimulation of peripheral nerves. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, pp 125-131. 47. Caruso G, Nolano M, Lullo F, Crisci C, Bsee JN, Massini R: Median nerve sensory responses evoked by tactile stimulation of the finger proximal and distal phalanx in normal subjects. Muscle Nerve 17:269-275, 1994. 48. Casey EB, Le Quesne PM: Digital nerve action potentials in healthy subjects, and in carpal tunnel and diabetic patients. J Neurol Neurosurg Psychiatry 35:612-623, 1972. 49. Cerra D, Johnson EW: Motor nerve conduction velocity in premature infants. Arch Phys Med Rehabil 43:160-164, 1962. 50. Chan AW, MacFarlane IA, Bowsher D, Campbell JA: Weighted needle pinprick sensory thresholds: A simple test of sensory function in diabetic peripheral neuropathy. J Neurol Neurosurg Psychiatry 55:56-59, 1992. 51. Chodoroff G, Tashjian EA, Ellenberg MR: Orthodromic vs antidromic sensory nrve latencies in healthy persons. Arch Phys Med Rehabil 66:589-591, 1985. 52. Chu N-S: Current perception thresholds in toeto-digit transplantation and digit-to-digit replantation. Muscle Nerve 19:183-186, 1996. 53. Conn TG, Wertsch JJ, Pasupuleti DV, Loftsgaarden JD, Schenk VA: Nerve conduction studies: Orthodromic vs antidromic latencies. Arch Phys Med Rehabil 71:579-582, 1990.
121
54. Conrad B, Aschoff JC, Fischler M: Der Diagnostische Wert der F-Wellen-Latenz. J Neurol 210:151-159, 1975. 55. Cruz Martinez A, Barrio M, Perez Conde MC, Ferrer MT: Electrophysiological aspects of sensory conduction velocity in healthy adults. 2. Ratio between the amplitude of sensory evoked potentials at the wrist on stimulating different fingers in both hands. J Neurol Neurosurg Psychiatry 41:1097-1101, 1978. 56. Cruz Martinez A, Barrio M, Perez Conde MC, Gutierrez AM: Electrophysiological aspects of sensory conduction velocity in healthy adults. 1. Conduction velocity from digit to palm, from palm to wrist, and across the elbow, as a. function of age. J Neurol Neurosurg Psychiatry 41:1092-1096, 1978. 57. Cruz Martinez A, Ferrer MT, Martin MJ: Motor conduction velocity and H-reflex in prematures wih very short gestational age. Electromyogr Clin Neurophysiol 23:13-19, 1983. 58. Dawson GD, Scott JW: The recording of nerve action potentials through skin in man. J Neurol Neurosurg Psychiatry 12:259-267, 1949. 59. Dawson GD: A summation technique for the detection of small evoked potentials. Electroencephalogr Clin Neurophysiol 6:65-84, 1954. 60. Dawson GD: The relative excitability and conduction velocity of sensory and motor nerve fibres in man. J Physiol (Lond) 131:436-451, 1956. 61. De Jesus PV, Hausmanowa-Petrusewicz I, Barchi RL: The effect of cold on nerve conduction of human slow and fast nerve fibers. Neurology (Minneap) 23:1182-1189, 1973. 62. Deltombe T, Hanson P, Jamart J, Clerin M: The influence of skin temperature on latency and amplitude of the sympathetic skin response in normal subjects. Muscle Nerve 21:34-39, 1998. 63. Del Toro DR, Park TA: Abductor hallucis false motor points: electrophysiologic mapping and cadaveric dissection. Muscle Nerve 19:11381143, 1996. 64. Del Toro DR, Park TA, Wertsch JJ: Development of a model of the premotor potential. Arch Phys Med Rehabil 75:493-497, 1994. 65. de Neeling JND, Beks PJ, Bertelsmann FW, Heine RJ, Bouter LM: Sensory thresholds in older adults: reproducibility and reference values. Muscle Nerve 17:454-461, 1994. 66. Denisilic M, Meh D: Letters to the Editor: Reproducibility of sympathetic skin response. Muscle Nerve 20:1332-1333, 1997. 67. Denisilic M, Trontelj JV: Autonomic function in HMSN: evidence of slowed sudomotor conduction? Muscle Nerve 16:114-115, 1993. 68. Denys EH: AAEM minimonograph #14: The influence of temperature in clinical neurophysiology. Muscle Nerve 14:795-811, 1991. 69. Desmedt JE: The neuromuscular disorder in myasthenia gravis. 1. Electrical and mechanical response to nerve stimulation in hand muscles. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 241-304. 70. Dettmers C, Fatepour D, Faust H, Jerusalem
Nerve Conduction Studies
122
71.
72. 73. 74. 75. 76. 77. 78. 79. 80.
81.
82.
83.
84. 85.
86.
87. 88.
F: Sympathetic skin response abnormalities in amyotrophic lateral sclerosis. Muscle Nerve 16:930-934, 1993. Dioszeghy P, Stalberg E: Changes in motor and sensory nerve conduction parameters with temperature in normal and diseased nerve. Electroencephalogr Clin Neurophysiol 85:229235, 1992. Dorfman LJ, Bosley TM: Age-related changes in peripheral and central nerve conduction in man. Neurology (New York) 29:38-44, 1979. Downie AW, Newell DJ: Sensory nerve conduction in patients with diabetes mellitus and controls. Neurology (Minneap) 11:876-882, 1961. Dreyer SJ, Dumitru D, King JC: Anodal block V anodal stimulation: Fact or fiction. Ann J Phys Med Rehabil 72:10-18, 1993. Drory VE, Korczyn AD: Sympathetic skin response: age effect. Neurology 43:1818-1820, 1993. Dumitru D, Walsh NE, Ramamurthy S: The premotor potential. Arch Phys Med Rehabil 70: 537-540, 1989. Dumitru D, DeLisa JA: AAEM minimonograph #10: Volume conduction. Muscle Nerve 14:605-624, 1991. Dumitru D, King JC: Median/ulnar premotor potential identification and localization. Muscle Nerve 18:518-525, 1995. Dyck PJ: Quantitative sensory testing: a consensus report from the Peripheral Neuropathy Association. Neurology 43:1050-1052, 1993. Dyck PJ, Davies JL, Litchy WJ, O'Brien PC: Longitudinal assessment of diabetic polyneuropathy using a composite score in the Rochester Diabetic Neuropathy Study cohort. Neurology 49:229-239, 1997. Dyck PJ, Dyck PJB, Kennedy WR, Kesserwani H, Melanson M, Ochoa J, Shy M, Stevens JC, Suarez GA, O'Brien PC: Limitations of quantitative sensory testing when patients are biased toward a bad outcome. Neurology 50:1213, 1998. Dyck PJ, O'Brien PC: Quantitative sensation testing in epidemiological and therapeutic studies of peripheral neuropathy. Muscle Nerve 22:659-662, 1999. Eduardo E, Burke D: The optimal recording electrode configuration for compound sensory action potentials. J Neurol Neurosurg Psychiatry 51:684-687, 1988. Eichler W: Uber die Ableitung der Aktionspotentiale vom menschlichen Nerven in situ. Z Biol 98:182-214, 1937. Eisen A, Schomer D, Melmed C: The application of F-wave measurements in the differentiation of proximal and distal upper limb entrapments. Neurology (Minneap) 27:662-668, 1977. Elie B, Guiheneuc P: Sympathetic skin response: Normal results in different experimental conditions. Electroencephalogr Clin Neurophysiol 76:258-267, 1990. Elie B, Louboutin JP: Sympathetic skin responses (SSR) is abnormal in multiple sclerosis. Muscle Nerve 18:185-189, 1995. Engel AG: Myasthenic syndromes. In Engel AG,
89. 90.
91. 92.
93.
94.
95.
96.
97.
98.
99.
100.
101. 102. 103.
104.
105.
Franzini-Armstrong CF (eds): Amyology, ed 2. New York, McGraw-Hill, 1994, pp 1798-1835. Erlanger J, Gasser HS: Electrical Signs of Nervous Activity. University of Pennsylvania Press, Philadelphia, 1937. Ewing DJ, Campbell IW, Murray A: Immediate heart-rate response to standing: simple test for autonomic neuropathy in diabetes. Br Med J 1:145, 1978. Fagius J, Berne C: Increase in muscle nerve sympathetic activity in humans after food intake. Clin Sci 86:159-167, 1994. Fagius J, Beme C: Metabolic influences on sympathetic nerve activity, as seen by microneurography. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Amsterdam, 1996, pp 412-415. Faguis J, Wallin BG: Long term variability and reproducibility of resting human muscle nerve sympathetic activity, as reassessed after a decade. Clin Auton Res 3:201-205, 1993. Falco FJE, Hennessey WJ, Braddom RL, Goldberg G: Standardized nerve conduction studies in the upper limb of the healthy elderly. Am J Phys Med Rehabil 71:263-271, 1992. Feinberg DM, Preston DC, Shefner JM, Logigian EL: Amplitude-dependent slowing of conduction in amyotrophic lateral sclerosis and polyneuropathy. Muscle Nerve 22:937-940, 1999. Fen-ante MA, Olney R, Wilbourn AJ: Sensory nerve conduction study workshop. American Academy of Neurology, 49th Annual Meeting, Boston, 1997. Fen-ante MA, Wilbourn AJ: The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 18:879-889, 1995. Ferrer T, Ramos MJ, Perez-Sales P, PerezJimenez A, Alvarez E: Sympathetic sudomotor function and aging. Muscle Nerve 18:395-401, 1995. Flachenecker P, Miillges W, Wermuth P, Hartung HP, Reiners K: Eyeball pressure testing in the evaluation of serious bradyarrhythmias in Guillain-Barre syndrome. Neurology 47:102108, 1996. Fowler CJ, Sitzoglou K, All Z, Halonen P: The conduction velocities of peripheral nerve fibers conveying sensations of warming and cooling. J Neurol Neurosurg Psychiatry 51:1164-1170, 1988. Franssen H, Wieneke GH: Nerve conduction and temperature: necessary warming time. Muscle Nerve 17:336-344, 1994. Franssen H, Wieneke GH, Wokke JHJ: The influence of temperature on conduction block. Muscle Nerve 22:166-173, 1999. Freeman R, Cohen RJ, Saul JP: Transfer function analysis of respiratory sinus arrhythmia: A measure of autonomic function in diabetic neuropathy. Muscle Nerve 18:74-84, 1995. Freeman R, Saul JP, Roberts MS, Berger RD, Broadbridge C, Cohen RJ: Spectral analysis of heart rate in diabetic autonomic neuropathy. Arch Neurol 48:185-190, 1991. Fullerton PM: The effect of ischaemia on nerve conduction in the carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 26:385-397, 1963.
Principles and Variations of Nerve Conduction Studies 106. Gadia MT, Natori N, Ramos LB, Ayyar DR, Skyler JS, Sosenko JM: Influence of height on quantitative sensory, nerve conduction, and clinical indices of diabetic peripheral neuropathy. Diabetes Care 10/5:613-616, 1987. 107. Gamstorp I: Normal conduction velocity of ulnar, median and peroneal nerves in infancy, childhood and adolescence. Acta Paediatr (suppl 146):68-76, 1963. 108. Gans BM, Kraft GH: M-Response quantification: a technique. Arch Phys Med Rehabil 62:376-380, 1981. 109. Gassel MM: A study of femoral nerve conduction time. Arch Neurol 9:607-614, 1963. 110. Gilliatt RW, Sears TA: Sensory nerve action potentials in patients with peripheral nerve lesions. J Neurol Neurosurg Psychiatry 21:109118, 1958. 111. Gilliatt RW, Thomas PK: Changes in nerve conduction with ulnar lesions at the elbow. J Neurol Neurosurg Psychiatry 23:312-320, 1960. 112. Gilliatt RW, Melville ID, Velate AS, Willison RG: A study of normal nerve action potentials using an averaging technique (barrier grid storage tube). J Neurol Neurosurg Psychiatry 28:191-200, 1965. 113. Gilliatt RW, Le Quesne PM, Logue V, Sumner AJ: Wasting of the hand associated with a cervical rib or band. J Neurol Neurosurg Psychiatry 33:615-624, 1970. 114. Gilliatt RW: Sensory conduction studies in the early recognition of nerve disorders. Muscle Nerve 1:352-359, 1978. 115. Gilliatt RW: Electrophysiology of peripheral neuropathies—an overview. Muscle Nerve 5: S108-S116, 1982. 116. Gitter AJ, Stolov WC: AAEM Minimonograph #16: Instrumentation and measurement in electrodiagnostic medicine—part I. Muscle Nerve 18:799-811, 1995. 117. Gutmann L: The intramuscular nerve action potential. J Neurol Neurosurg Psychiatry 32: 193-196, 1969. 118. Gutmann L, Hopf HC, Roeder R: Origin of intramuscular nerve action potential. J Neurol Neurosurg Psychiatry 50:1669-1670, 1987. 119. Gutrecht JA: Sympathetic skin response. J Clin Neurophysiol 11:519-524, 1994. 120. Hagbarth K-E: Microneurography and applications to issues of motor control. Fifth Annual Stuart Reiner Memorial Lecture, American Association of Electrodiagnostic Medicine, October, 1992. 121. Hagbarth K-E, Vallbo AB: Single unit recordings from muscle nerves in human subjects. Acta Physiol Scand 76:321-334, 1969. 122. Hakamada S, Kumagai T, Watanabe K, Koike Y, Hara K, Miyazaki S: The conduction velocity of slower and the fastest fibres in infancy and childhood. J Neurol Neurosurg Psychiatry 45:851-853, 1982. 123. Halar EM, DeLisa JA, Brozovich FV: Nerve conduction velocity: Relationship of skin, subcutaneous and intramuscular temperatures. Arch Phys Med Rehabil 61:199-203, 1980. 124. Halar EM, Delisa JA, Soine TL: Nerve conduction studies in upper extremities: skin tern-
125. 126.
127.
128.
129.
130.
131.
132.
133.
134.
135.
136.
137.
138.
139.
123
perature corrections. Arch Phys Med Rehabll 64:412-16, 1983. Halar EM, Hammond MC, Dirks S: Physical activity: its influence on nerve conduction velocity. Arch Phys Med Rehabil 66:605-609, 1985. Hamano T, Kaji R, Diaz AF, Kohara N, Takamatsu N, Uchiyama T, Shibasaki H, Kimura J: Vibration-evoked sensory nerve action potentials derived from pacinian corpuscles. Electroencephalogr Clin Neurophysiol 89:278-286, 1993. Hampton KK, Alani SM, Wilson JI, Price DE: Resistance to ischemic conduction failure in chronic hypoxaemia and diabetes. J Neurol Neurosurg Psychiatry 52:1303-1305, 1989. Hansson P, Lindblom U, Lindstrom P: Graded assessment and classification of impaired sensibility in patients with diabetic polyneuropathy. J Neurol Neurosurg Psychiatry 54:527530, 1991. Hardy M, Jimenez S, Jabaley M, Horch K: Evaluation of nerve compression with the automated tactile tester. J Hand Surg 17A:838842, 1992. Harper CM, Low PA, Fealey RD, Chelimsky TC, Proper CJ, Gillen DA: Utility of thermography in the diagnosis of lumbosacral radiculopathy. Neurology 41:1010-1014, 1991. Harry JD, Freeman R: Determining heart rate variability: Comparing methodologies using computer simulations. Muscle Nerve 16:267277, 1993. Heijenbrok MW, Anema JR, Faes TJC, Bertelsmann FW, Heimans JJ, Polman CH: Quantitative measurement of vibratory sense and temperature sense in patients with multiple sclerosis. Electromyogr Clin Neurophysiol 32: 385-388, 1992. Helmholtz H: Vorlaufiger Bericht uber die Fortpflanzungsgeschwindigkeit der Nervenreizung. Arch Anat Physiol Wiss Med 71-73, 1850. Helmholtz H, Baxt N: Neue Versuche Uber die Fortpflanzungsgeschwindigkeit der Reizung in den motorischen Nerven der Menschen. Mber Konigl Preus Akad Wiss 184-191, 1870. Henriksen JD: Conduction velocity of motor nerves in normal subjects and in patients with neuromuscular disorders. Thesis, University of Minnesota, Minneapolis, 1966. Hilz MJ, Glaus D, Neundorfer B, Zimmermann P, Beric A: Is heat hypoalgesia a useful parameter in quantitative thermal testing of alcoholic polyneuropathy. Muscle Nerve 17: 1456-1460, 1994. Hilz M, Glorius SE, Schweibold G, Neuner I, Stemper B, Axelrod.FB: Quantitative thermal perception testing in preschool children. Muscle Nerve 19:381-383, 1996. Hodes R, Larrabee MG, German W: The human electromyogram in response to nerve stimulation and the conduction velocity of motor axons. Arch Neurol Psychiatry 60:340-365, 1948. Hodgkin AL, Katz B: The effect of temperature on the electrical activity of the giant axon of the squid. J Physiol (Lond) 109:240-249, 1949.
124 140. Horowitz SH: Correlation of near-nerve sural conduction and quantified sensory testing in patients with diabetic neuropathy. Muscle Nerve 18:1202-1204, 1995. 141. Horowitz SH: Correlation of near-nerve sural conduction and quantified sensory testing in polyneuropathies. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, p 362. 142. Horowitz SH, Krarup C: Conduction studies of the normal sural nerve. Muscle Nerve 15: 374-383, 1992. 143. Ingall TJ, McLeod JG: Autonomic function in Friedreich's ataxia. J Neurol Neurosurg Psychiatry 54:162-164, 1991. 144. Ingall TJ, McLeod JG, O'Brien PC: The effect of ageing on autonomic nervous system function. Aust NZ J Med 20:570, 1990. 145. Ingis JT, Leeper JB, Burke D, Gandevia SC: Morphology of action potentials recorded from human nerves using microneurography. Exp Brain Res 110:308-314, 1996. 146. Jamal GA, Hansen S, Weir AI, Ballantyne JP: The neurophysiologic investigation of small fiber neuropathies. Muscle Nerve 10:537-545, 1987. 147. Jensen TS, Bach FW, Kastrup J, Dejgaard A, Brennum J: Vibratory and thermal thresholds in diabetics with and without clinical neuropathy. Acta Neurol Scand 84:326-333, 1991. 148. Johnson EW, Olsen KJ: Clinical value of motor nerve conduction velocity determination. JAMA 172:2030-2035, 1960. 149. Joynt RL: Calculated nerve conduction velocity dependence upon the method of testing. Arch Phys Med Rehabil 64:212-216, 1983. 150. Julu POO, Hondo RG: Effects of atropine on autonomic indices based on electrocardiographic R-R intervals in healthy volunteers. J Neurol Neurosurg Psychiatry 55:31-35, 1992. 151. Kadrie HA, Yates SK, Milner-Brown HS, Brown WF: Multiple point electrical stimulation of ulnar and median nerves. J Neurol Neurosurg Psychiatry 39:973-985, 1976. 152. Kanzato N, Komine Y, Fukiyama K: Sympathetic skin responses of the hand in normal subjects: shorter latency at distal phalanx. Electroencephalogr Clin Neurophysiol 105: 165-170, 1997. 153. Kaplan P, Sahgal V: Residual latency: New applications of an old technique. Arch Phys Med Rehabil 59:24-27, 1978. 154. Kato M: The conduction velocity of the ulnar nerve and the spinal reflex time measured by means of the H-wave in average adults and athletes. Tohoku J Exp Med 73:74-85, 1960. 155. Kennedy WR, Navarro X: Evaluation of sudomotor function by sweat imprint methods. In Low PA (ed): Clinical Autonomic Disorders: Evaluation and Management. Boston, Little, Brown, 1993, pp 253-361. 156. Kiernan MC, Mogyoros I, Hales JP, Gracies JM, Burke D: Excitability changes in human cutaneous afferents induced by prolonged repetitive axonal activity. J Physiol (Lond) 500: 255-264, 1997.
Nerve Conduction Studies 157. Kikuchi S, Sato K, Konno S, Hasue M: Anatomic and radiographic study of dorsal root ganglia. Spine 19:6-11, 1994. 158. Kimura J: F-wave velocity in the central segment of the median and ulnar nerves: A study in normal subjects and in patients with CharcotMarie-Tooth disease. Neurology (Minneap) 24: 539-546, 1974. 159. Kimura J: A method for determining median nerve conduction velocity across the carpal tunnel. J Neurol Sci 38:1, 1978. 160. Kimura J: The carpal tunnel syndrome. Localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619-635, 1979. 161. Kimura J: Electrodiagnosis in Disease of Nerve and Muscle: Principles and Practice. FA Davis, Philadelphia, 1983. 162. Kimura J: Principles and pitfalls of nerve conduction studies. Ann Neurol 16:415-429, 1984. 163. Kimura J: Electromyography and nerve stimulation techniques: Clinical applications (Japanese). Igaku-Shoin, Tokyo, 1990. 164. Kimura J, Bosch P, Lindsay GM: F-wave conduction velocity in the central segment of the peroneal and tibial nerves. Arch Phys Med Rehabil 56:492-497, 1975. 165. Kimura J, Machida M, Ishida T, Yamada T, Rodnitzky R, Kudo Y: A relationship between the size of compound sensory and muscle action potentials and the length of the nerve segment under study. Neurology 36:647-652, 1986. 166. Kimura J, Mitsudome A, Yamada T, Dickins QS: Stationary peaks from a moving source in far-field recording. Electroencephalogr Clin Neurophysiol 58:351-361, 1984. 167. Kimura J, Yamada T, Stevland N: Distal slowing of motor nerve conduction velocity in diabetic polyneuropathy. J Neurol Sci 42:291302, 1979. 168. Kincaid JC, Brashear A, Markand ON: The influence of the reference electrode on CMAP configuration. Muscle Nerve 16:392-396, 1993. 169. King D, Ashby P: Conduction velocity in the proximal segments of a motor nerve in the Guillain-Barre syndrome. J Neurol Neurosurg Psychiatry 39:538-544, 1976. 170. Knezevic W, Bajada S: Peripheral autonomic surface potential: A quantitative technique for recording sympathetic conduction in man. J Neurol 67:239-251, 1985. 171. Kohara N, Kimura J, Kaji R, Goto Y, Ishii J: Multicenter analysis on intertrial variability of nerve conduction studies: healthy subjects and patients with diabetic polyneuropathy. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier Science BV, Amsterdam, 1996, pp 809-815. 172. Kraft GH, Halvorson GA: Median nerve residual latency: normal value and use in diagnosis of carpal tunnel syndrome. Arch Phys Med Rehabil 64:221-226, 1983. 173. Krarup C: Somatosensory changes in cisplatin neuropathy, with particular reference to studies of cutaneous mechanoreceptors. In Kimura J, Shibasaki H (Eds): Recent Advances in Clin-
Principles and Variations of Nerve Conduction Studies
174.
175.
176. 177.
178.
179.
180.
181. 182. 183.
184.
185. 186.
187.
188. 189.
190.
ical Neurophysiology, Elsevler Science BV, Amsterdam, 1996, p 132. Krarup C, Horowitz SH, Dahl K: The influence of the stimulus on normal sural nerve conduction velocity. Muscle Nerve 15:813-821, 1992. Krarup C, Trojaborg W: Compound sensory action potentials evoked by tactile and by electrical stimulation in normal median and sural nerves. Muscle Nerve 17:733-740, 1994. Krarup C, Trojaborg W: Sensory pathophysiology in chronic acquired demyelinating neuropathy. Brain 119:257-270, 1996. Kuwabara S, Nagase H, Arai K, Hattori T: Slowly conducting low-threshold components of sensory nerve potentials in peripheral neuropathy: a microneurographic study. Muscle Nerve 20:961-968, 1997. Kuwabara S, Nakajima Y, Hattori T, Toma S, Mizobuchi K, Ogawara K: Activity-dependent excitability changes in chronic inflammatory demyelinating polyneuropathy: A microneurographic study. Muscle Nerve 22:899-904, 1999. LaBan MM, Petty D, Hauser AM, Taylor RS: Peripheral nerve conduction stimulation: Its effect on cardiac pacemakers. Arch Phys Med Rehabil 69:358-362, 1988. Lambert EH: Neurophysiological techniques useful in the study of neuromuscular disorders. In Adams RD, Eaton LM, Shy GM (eds): Neuromuscular Disorders. Williams & Wilkins, Baltimore, 1960, pp 247-273. Lambert EH: Diagnostic value of electrical stimulation of motor nerves. Electroencephalogr Clin Neurphysiol (suppl 22) 9-16, 1962. Lang HA, Puusa A: Dual influence of temperature on compound nerve action potential. J Neurol Sci 51:81-88, 1981. Lang HA, Puusa A, Hynninen P, Kuusela V, Jantti V, Sillanpaa M: Evolution of nerve conduction velocity in later childhood and adolescence. Muscle Nerve 8:38-43, 1985. Lateva ZC, McGill K, Burgar CG: Anatomical and electrophysiological determinants of the human thenar compound muscle action potential. Muscle Nerve 19:1457-1468, 1996. Layzer RB: Neuromyotonia: an autoimmune disease. Ann Neurol 38:701, 1995. Lee HJ, DeLisa JA, Bach JR: The effect of temperature on antidromic median sensory conduction. Electromyogr Clin Neurophysiol 33: 125-128, 1993. Lee HJ, DeLisa JA, Bach JR: Physiologic considerations in the determination of optimum interelectrode distance for the anti-dromic recording of compound sensory nerve action potentials. Am J Phys Med Rehabil 72:99-100, 1993. Lee KW, Oh SJ: Early appearance of aging phenomenon in the interdigital nerves of the foot. Muscle Nerve 17:58-63, 1994. Lee Lim C, Lal H, Yiannikas C: The effect of wrist size on the orthodromic median sensory nerve action potential. Muscle Nerve 18:117119, 1995. Lefaucheur J-P, Fitoussi M, Becquemin J-P: Abolition of sympathetic skin responses fol-
191. 192.
193.
194.
195.
196.
197.
198. 199.
200.
201.
202.
203.
204.
205.
206.
125
lowing endoscopic thoracic sympathectomy. Muscle Nerve 19:581-586, 1996. Levin AB: A simple test of cardiac function based upon the heart-rate changes during the Valsalva maneuver. Am J Cardiol 18:90, 1966. Levin KH: L5 radiculopathy with reduced superficial peroneal sensory responses: intraspinal and extraspinal causes. Muscle Nerve 21:3-7, 1998. Levy D, Abraham R, Reid G: A comparison of two methods for measuring thermal thresholds in diabetic neuropathy. J Neurol Neurosurg Psychiatry 52:1072-1077, 1989. Levy DM, Reid G, Rowley DA, Abraham RR: Quantitative measures of sympathetic skin response in diabetes: Relation to sudomotor and neurological function. J Neurol Neurosurg Psychiatry 55:902-908, 1992. Lewis RA, Sumner AJ: The electrodiagnostic distinctions between chronic familial and acquired demyelinative neuropathies. Neurology 32:592-596, 1982. Liguori R, Krarup C, Trojaborg W: Determination of the segmental sensory and motor innervation of the lumbosacral spinal nerves. Brain 115:915-934, 1992. Lin T-K, Chee EC-Y, Chen H-J, Cheng M-H: Abnormal sympathetic skin response in patients with palmar hyperhidrosis. Muscle Nerve 18:917-919, 1995. Linden D, Diehl RR: Comparison of standard autonomic tests and power spectral analysis in normal adults. Muscle Nerve 19:556-562, 1996. Linden D, Diehl RR, Kretzschmar A, Berlit P: Autonomic evaluation by means of standard tests and power spectral analysis in multiple sclerosis. Muscle Nerve 20:809-814, 1997. Linden D, Weng Y, Glocker FX, Kretzschmar A, Diehl RR, Berlit P: Sympathetic skin responses evoked by magnetic stimulation of the neck: normative data. Muscle Nerve 19:1487-1489, 1996. Litchy WJ, Phillips LH, Shields RW: Motor nerve conduction studies. American Academy of Neurology, 49th Annual Meeting, Boston, MA, 1997. Louis AA, Hotson JR: Regional cooling of human nerve and slowed NA inactivation. Electroencephalogr Clin Neurophysiol 634:371375, 1986. Low PA, Denq J-C, Opfer-Gehrking TL, Dyck PJ, O'Brien PC, Slezak JM: Effect of age and gender on sudomotor and cardiovagal function and blood pressure response to tilt in normal subjects. Muscle Nerve 20:1561-1568, 1997. Low PA, Fealey RD: Sudomotor neuropathy. In Dyck PJ, Thomas PK, Asbury AK, (eds): Diabetic Neuropathy. WB Saunders, Philadelphia, 1986, p 140. Low PA, Opfer-Gehrking TL, Proper CJ, Zimmerman I: The effect of aging on cardiac 0autonomic and postganglionic sudomotor function. Muscle Nerve 13:152-157, 1990. Low PA, Walsh JC, Huang C-Y, McLeod JG: The sympathetic nervous system in diabetic neuropathy—a clinical and pathological study. Brain 98:341, 1975.
Nerve Conduction Studies
126 207. Luciano CA, Gilliatt RW, Conwit RA: Mixed nerve action potentials in acquired demyelinating polyneuropathy. Muscle Nerve 18:8592, 1995. 208. Macefield G: Spontaneous and evoked ectopic discharge recorded from single human axons. Muscle Nerve 21:461-468, 1998. 209. Macefield G, Burke D: Parasesthesaie and tetany induced by voluntary hyperventilation: Increased excitability of human cutaneous and motor axons. Brain 114:527-540, 1991. 210. Macefield G, Gandevia SC, Burke D: Perceptual responses to microstimulation of single afferents innervating the joints, muscles and skin of the human hand. J Physiol (Lond) 429:113-129, 1990. 211. Mano T: Sympathetic nerve mechanisms of human adaptation to environment: Findings obtained by recent microneurographic studies. Environ Med 34:1-35, 1990. 212. Mano T, Iwase S, Sugiyama Y: Double recording technique of sympathetic nerve traffic in microneurography. In Kimura J, Shibasaki H (Eds): Recent Advances in Clinical Neurophysiology, Elsevier, Amsterdam, 1996, pp 416-419. 213. Marlowe ES, Bonner FJ, Berkowitz AR: Correlation between two-point discrimination and median nerve sensory response. Muscle Nerve 22:1196-1200, 1999. 214. Martinez AC, Perez Conde MC, Del Campo F, Mingo P, Ferrer MT: Ratio between the amplitude of sensory evoked potentials at the wrist in both hands of left-handed subjects. J Neurol Neurosurg Psychiat 43:182-184, 1980. 215. Maselli RA, Jaspan JB, Soliven BC, Green AJ, Spire JP, Amason BG: Comparison of sympathetic skin response with quantitative sudomotor axon reflex test in diabetic neuropathy. Muscle Nerve 1989;12:420-423. 216. Masson EA, Boulton AJM: The Neurometer: validation and comparison with conventional tests for diabetic neuropathy. Diabet Med 8 (symposium):563-566, 1991. 217. Matsunaga K, Uozumi T, Tsuji S, Murai Y: Sympathetic skin responses evoked by magnetic stimulation of the neck. J Neurol Sci 128:188-194, 1995. 218. Matsunaga K, Uozumi T, Tsuji S, Murai Y: Sympathetic skin responses recorded from non-palmar and non-plantar skin sites: Their role in the evaluation of thermal sweating. EEG Clin Neurophysiol 108:482-489, 1998. 219. Maurer K, Hopf HC, Lowitzsch K: Isometric muscle contraction in endoctrine myopathies. Neurology 35:333-337, 1985. 220. Mayer RF: Nerve conduction studies in man. Neurology (Minneap) 13:1021-1030, 1963. 221. McGill KG, Lateva ZC: The contribution of the interosseous muscles to the hypothenar compound muscle action potential. Muscle Nerve 22:6-15, 1999. 222. McLeod JG: Digital nerve conduction in the carpal tunnel syndrome after mechanical stimulation of the finger. J Neurol Neurosurg Psychiatry 29:12-22, 1966. 223. Melvin JL, Schuchman JA, Lanese RR: Diagnostic specificity of motor and sensory nerve
224.
225. 226.
227. 228.
229.
230. 231. 232.
233.
234.
235.
236.
237.
238.
239. 240.
conduction variables in the carpal tunnel. Arch Phys Med Rehabil 54:69-74, 1973. Meyers S, Cros D, Sherry B, Vermeire P: Liquid crystal thermography: Quantitative studies of abnormalities in carpal tunnel syndrome. Neurology 39:1465-1469, 1989. Miller R, Kuntz N: Nerve conduction studies in infants and children. J Child Neurol 1:19-26, 1986. Mitz M, Gokulananda T, Di Benedetto M, Klingbeil GE: Median nerve determinations: analysis of two techniques. Arch Phys Med Rehabil 65:191-193, 1984. Mogyoros I, Kieman MC, Burke D: Strengthduration properties of human peripheral nerve. Brain 119:439-447, 1996. Mogyoros I, Kiernan MC, Grades J-M, Burke D: The effect of stimulus duration on the latency of submaximal nerve volleys. Muscle Nerve 19:1354-1356, 1996. NavarroX, Kennedy WR: Evaluation of thermal and heat pain sensitivity in type I diabetic patients. J Neurol Neurosurg Psychiatry 54:6064, 1991. Navarro X, Kennedy WR, Fries TJ: Small nerve fiber dysfunction in diabetic neuropathy. Muscle Nerve 1989:498-507, 1989. Newman M, Nelson N: Digital nerve sensory potentials in lesions of cervical roots and brachial plexus. Can J Neurol Sci 10:252-255, 1983 Nogues MA, Stalberg EV: Automatic analysis of hart rate variation: II. Findings in patients attending an EMG laboratory. Muscle Nerve 12:1001-1008, 1989. Nora LM: American Association of Electrodiagnostic Medicine guidelines in electrodiagnostic medicine: implanted cardioverters and defibrillators. Muscle Nerve 19:1359-1360, 1996. Nordin M, Nystrom B, Wallin U, Hagbarth K-E: Ectopic sensory discharges and paresthesaie in patients with disorders of peripheral nerves, dorsal roots and dorsal columns. Pain 20:231-245, 1984. Normand MM, Daube JR: Interaction of random electromyographic activity with averaged sensory evoked potentials. Neurology 42:16051608, 1992. Norris AH, Shock NW, Wagman IH: Age changes in the maximum conduction velocity of motor fibers of human ulnar nerves. J Appl Physiol 5:589-593, 1953. Notermans NC, Franssen H, Wieneke GH, Wokke JH: Temperature dependence of nerve conduction and EMG in neuropathy associated with gammopathy. Muscle Nerve 17:516-522, 1994. Nystrom B, Hagbarth K-E: Microelectrode recordings from transected nerves in amputees with phantom limb pain. Neurosci Lett 27:211216, 1981. Ochoa JL: The human sensory unit and pain: new concepts, syndromes, and test. Muscle Nerve 16:1009-1016, 1993. Ochoa JL: Microneurography and neuropathic pain. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Amsterdam, 1996, pp 122-124.
Principles and Variations of Nerve Conduction Studies 241. Ochoa JL, Torebjork HE: Sensations evoked by intraneural microstimulation of single mechanoreceptor units innervating the human hand. J Physiol (Lond) 342:633-654, 1983. 242. Odabasi Z, Oh SJ, Claussen GC, Kim DS: New near-nerve needle nerve conduction technique: Differentiating epicondylar from cubital tunnel ulnar neuropathy. Muscle Nerve 22:718-723, 1999. 243. Panizza M, Nilsson J, Bradley JR, Rothwell J, Hallett M: The time constants of motor and sensory peripheral nerve fibers measured with the method of latent addition. Electroencephalogr Clin Neurophysiol 93:147-154, 1994. 244. Park TA, Del Toro DR: Generators of the early and late median thenar premotor potentials. Muscle Nerve 18:1000-1008, 1995. 245. Parry GJ, Kohzu H: Acute changes in blood glucose affect resistance to ischemic nerve conduction failure. Neurology 40:107-110, 1990. 246. Pascoe MK, Silbert PL, Stolp-Smith KA: Stimulus-induced repetitive discharges of long latency: Axonal loop reflexes. Muscle Nerve 18:927-928, 1995. 247. Pavesi G, Medici D, Gemignani F, Lusvardi M, Tinchelli S, Mancia D: Sympathetic skin response (SSR) in the foot after sural nerve biopsy. Muscle Nerve 18:1326-1328, 1995. 248. Pease WS, Fatehi MT, Johnson EW: Monopolar needle stimulation: Safety considerations. Arch Phys Med Rehabil 70:412-414, 1989. 249. Phillips II LH: Pitfalls in nerve conduction studies: temperature and distance measurement. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Science BV, Amsterdam, 1996, p 674. 250. Pinelli P: Physical, anatomical and physiological factors in the latency measurement of the M response. Electroencephalogr Clin Neurophysiol 17:86, 1964. 251. Piper H: Weitere Mitteilungen uber die Geschwindigkeit der Erregungsleitung im markhaltigen menschlichen Nerven. Pflugers Arch Ges Physiol 127:474-480, 1909. 252. Preston DC, Venkatesh S, Shefner JM, Logigian ELL: Submaximal stimuli activate different verve fiber populations at different sites. Muscle Nerve 17:381-385, 1994. 253. Raszewa M, Hausmanowa-Petrusewicz I, Blaszczyk M, Jablonska S: Sympathetic skin response in scleroderma. Electromyogr Clin Neurophysiol 31:467-472, 1991. 254. Ravits JM: AAEM Minimonograph #48: Autonomic nervous system testing. Muscle Nerve 20:919-937, 1997. 255. Ravits J, Hallett M, Nilsson J, Polinsky R, Dambrosia J: Electrophysiological tests of autonomic function in patients with idiopathic autonomic failure syndromes. Muscle Nerve 19:758-763, 1996. 256. Raynor EM, Preston DC, Logigian EL: Influence of surface recording electrode placement on nerve action potentials. Muscle Nerve 20:361363, 1997. 257. Redmond JMT, McKenna MJ: Quantitative sensory testing (QST). Muscle Nerve 19:403 1996.
127
258. Redmond JMT, McKenna MJ, Feingold M, Ahmad BK: Sensory testing versus nerve conduction velocity in diabetic polyneuropathy. Muscle Nerve 15:1334-1339, 1992. 259. Redmond JMT, McKenna MJ, Feingold M, Ahmad BK: Sensory testing is different in laboratory personnel compared with paid volunteers. Muscle Nerve 18:351-352, 1995. 260. Rendell MS, Katims JJ, Richter R, Rowland F: A comparison of nerve conduction velocities and current perception thresholds as correlates of clinical severity of diabetic sensory neuropathy. J Neurol Neurosurg Psychiatry 52: 502-511, 1989. 261. Rivner MH, Swift TR, Grout BO, Rhodes KP: Toward more rational nerve conduction interpretations: The effect of height. Muscle Nerve, 13:232-239, 1990. 262. Ro LS, Chen ST, Tang LK, Hsu WC, Chang HS, Huang CC: Current perception threshold testing in Fabry's disease. Muscle Nerve 22:15311537, 1999. 263. Robinson LR: Interelectrode distance: A method for fixing electrode separation. Am J Phys Med Rehabil 71:122-123, 1992. 264. Robinson LR, Micklesen PJ, Wang L: Strategies for analyzing nerve conduction data: Superiority of a summary index over single tests. Muscle Nerve 21:1166-1171, 1998. 265. Robinson LR, Rubner DE, Wahl PW, Fujimoto WY, Stolov WC: Factor analysis. Am J Phys Med Rehabil 71:22-27, 1992. 266. Robinson LR, Rubner DE, Wahl PW, Fujimoto WY, Stolov WC: Influences of height and gender on normal nerve conduction studies. Arch Phys Med Rehabil 74:1134-1138, 1993. 267. Robinson LR, Temkin NR, Fujimoto WY, Stolov WC: Effect of statistical methodology on normal limits in nerve conduction studies. Muscle Nerve 14:1081-1090, 1991. 268. Robinson RO, Robertson .WC Jr: Fetal nutrition and peripheral nerve conduction velocity. Neurology 31:327-329, 1981. 269. Rosenfalck A: Early recognition of nerve disorders by near-nerve recording of sensory action potentials. Muscle Nerve 1:360-367, 1978. 270. Roth G: Double discharges of distal origin. J Neurol Sci 47:35-48, 1980. 271. Roth G: Letters to the Editor: Stimulusinduced repetitive discharge of long latency. Muscle Nerve 20:628-629, 1997. 272. Rutkove SB, Kothari MJ, Raynor EM, Levy ML, Fadic R, Nardin RA: Sural/radial amplitude ratio in the diagnosis of mild axonal polyneuropathy. Muscle Nerve 20:1236-1241, 1997. 273. Rutkove SB, Kothari MJ, Shefner JM: Nerve, muscle and neuromuscular junction electrophysiology at high temperature. Muscle Nerve 20:431-436, 1997. 274. Salerno DF, Franzblau A, Werner RA, Bromberg MB, Armstrong TH, Albers JW: Median and ulnar nerve conduction studies among workers: normative values. Muscle Nerve 21: 999-1005, 1998. 275. Satchell PM, Seers CP: Evoked skin sympathetic nerve responses in man. J Neruol Neurosurg Psychiatry 50:1015-1021, 1987.
128 276. Scranton PE Jr, Hasiba U, Gorenc TJ: Intramuscular hemorrhage in hemophiliacs with inhibitors. A medical emergency. JAMA 241: 2028-2030, 1979. 277. Seneviratne KN, Peiris OA: The effect of ischaemia on the excitability of human sensory nerve. J Neurol Neurosurg Psychiatry 31:338347, 1968. 278. Shahani BT, Day TJ, Cros D, Khalil N, Kneebone CS: R-R interval variation and the sympathetic skin response in the assessment of autonomic function in peripheral neuropathy. Arch Neurol 47:659-664, 1990. 279. Shahani BT, Halperin JJ, Boulu P, Cohen J: Sympathetic skin response: A method of assessing unmyelinated axon dysfunction in peripheral neuropathies. J Neurol Neurosurg Psychiatry, 47:536-542, 1984. 280. Shahani BT, Young RR, Potts F, Maccabee P: Terminal latency index (TL1) and late response studies in motor neuron disease (MND), peripheral neuropathies and entrapment syndromes. Acta Neurol Scand (suppl) 73:60, 118, 1979. 281. Shefner JM, Buchthal F, Krarup C: Slowly conducting myelinated fibers in peripheral neuropathy. Muscle Nerve 14:534-542, 1991. 282. Shimizu T, Takahashi Y, Kogawa S, Takahashi K, Kanbayashi T, Saito Y, Hishikawa Y: Muscle sympathetic nerve activity during apneic episodes in patients with obstructive sleep apnea syndrome. Electroencephalogr clin Neurophysiol 93:345-352, 1994. 283. Shivji ZM, Ashby P: Sympathetic skin responses in hereditary sensory and autonomic neuropathy and familial amyloid neuropathy are different. Muscle Nerve 22:1283-1286, 1999. 284. Simonetti S, Dahl K, Krarup C: Different indentation velocities activate different populations of mechanoreceptors in humans. Muscle Nerve 21:858-868, 1998. 285. Simpson JA: Fact and fallacy in measurement of conduction velocity in motor nerves. J Neurol Neurosurg Psychiatry 27:381-385, 1964. 286. Smit BJ, Kok JH, De Vries LS, Dekker FW, Ongerboer de Visser BW: Motor nerve conduction velocity in very preterm infants. Muscle Nerve 22:372-377, 1999. 287. So YT, Aminoff MJ, Olney RK: The role of thermography in the evaluation of lumbosacral radiculopathy. Neurology 39:1154-1158, 1989. 288. Soichot P, Roth G: High frequency discharge of a fraction (f) of motor unit action potential. Electroencephalogr Clin Neurophysiol 101: 201-205, 1996. 289. Soliven B, Maselli R, Jaspan J, Green A, Graziano H, Petersen M, Spire J-P: Sympathetic skin response in diabetic neuropathy. Muscle Nerve 10:711-716, 1987. 290. Soudmand R, Ward LC, Swift TR: Effect of height on nerve conduction velocity. Neurology 32:407-410, 1982. 291. Stalberg EV, Nogues MA: Automatic analysis of heart rate variation. I. Method and reference values in healthy controls. Muscle Nerve 12: 993-1000, 1989.
Nerve Conduction Studies 292. Stetson DS, Albers JW, Silverstein BA, Wolfe RA: Effects of age, sex, and anthropometric factors on nerve conduction measures. Muscle Nerve 15:1095-1104, 1992. 293. Stewart JD, Nguyen DM, Abrahamowicz M: Quantitative sweat testing using acetylcholine for direct and axon reflex mediated stimulation with silicone mold recording; controls versus neuropathic diabetics. Muscle Nerve 17:13701377, 1994. 294. Stjernberg L, Blumberg H, Wallin B: Sympathetic activity in man after spinal cord injury: outflow to muscle below the lesion. Brain 109: 695-715, 1986. 295. Strian F, Lautenbacher S, Karlbauer G, Galfe G: Disturbances of C-fiber-mediated sensibility in lumbosacral disc disease. J Neurol Neurosurg Psychiatry 54:1013-1014, 1991. 296. Sugiyama Y, Matsukawa T, Suzuki H, Iwase S, Shamsuzzaman ASM, Mano T: A new method of quantifying human muscle sympathetic nerve activity for frequency domain analysis. Electroencephalogr clin Neurophysiol 101: 121-128, 1996. 297. Sundkvist G, Aimer L-O, Lilja B: Respiratory influence on heart rate in diabetes mellitus. Br Med J 1:924, 1979. 298. Swenson MR, Villasana DR, Stigler J: The motor points and false motor points of intrinsic foot muscles. Muscle Nerve 13:862, 1990. 299. Tan J, Akin S, Beyazova M, Sepici V, Tan E: Sympathetic skin response and R-R interval variation in rheumatoid arthritis. Am J Phys Med Rehabil 72:196-203, 1993. 300. Tasaki I: Electric stimulation and the excitatory process in the nerve fiber. Am J Physiol 125:380-395, 1939. 301. Taylor PK: Nonlinar effects of age on nerve conduction in adults. J Neurol Sci 66:223-234, 1984. 302. Thomas JE, Lambert EH: Ulnar nerve conduction velocity and H-reflex in infants and children. J Appl Physiol 15:1-9, 1960. 303. Thomas PK: Morphological basis for alterations in nerve conduction in peripheral neuropathy. Proc R Soc Med 64:295-298, 1971. 304. Thomas PK, Sears TA, Gilliatt RW: The range of conduction velocity in normal motor nerve fibres to the small muscles of the hand and foot. J Neurol Neurosurg Psychiatry 22:175181, 1959. 305. Tjon-A-Tsien AML, Lemkes HHPJ, van der Kamp AJC, van Dijk JG: Large electrodes improve nerve conduction repeatability in controls as well as in patients with diabetic neuropathy. Muscle Nerve 19:689-695, 1996. 306. Todnem K, Knudsen G, Riise T, Nyland H, Aarli JA: The non-linear relationship between nerve conduction velocity and skin temperature. J Neurol Neurosurg Psychiatry 52:497-501, 1989. 307. Torebjork HE, LaMotte RH, Robinson CJ: Peripheral neural correlates of magnitude of cutaneous pain and hyperalgesia: simultaneous recordings in humans of sensory judgements of pain and evoked responses in nociceptors with C-fibers. J Neuropshyiol 51:325-339, 1984.
Principles and Variations of Nerve Conduction Studies 308. Torebjork HE, Schady W, Ochoa JL: A new method for demonstration of central effects of analgesic agents in man. J Neurol Neurosurg Psychiatry 47:862-869, 1984. 309. Toyokura M: Waveform and habituation of sympathetic skin response. Electroencephalogr Clin Neurophysiol 109:178-183, 1998. 310. Toyokura M, Murakami K: Reproducibility of sympathetic skin response. Muscle Nerve 19:1481-1483, 1996. 311. Trojaborg W: Sensory nerve conduction: nearnerve recording. Methods Clin Neurophysiol 3:17-40, 1992. 312. Trojaborg WT, Moon A, Andersen BB, Trojaborg NS: Sural nerve conduction parameters in normal subjects related to age, gender, temperature, and height: A reappraisal. Muscle Nerve 15:666-671, 1992. 313. Tsuji S, Uozumi T, Matsunaga K, Murai Y: Sympathetic skin responses and sudomotor potentials evoked by magnetic stimulation of the neck. In Kimura J, Shibasaki H (Eds): Recent Advances in Clinical Neurophysiology, Elsevier Science BV, Amsterdam, 1996, pp 644-648. 314. Uncini A, Pullman SL, Lovelace RE, Gambi D: The sympathetic skin response: Normal values, elucidation of afferent components and application limits. J Neurol Sci 87:299-306, 1988. 315. Uozumi T, Nakano S, Matusnaga K, Tsuji S, Murai Y: Sudomotor potential evoked by magnetic stimulation of the neck. Neurology 43: 1397-1400, 1993. 316. van Dijk JG, Lammers GJ, Wintzen AR, Molenaar PC: Repetitive CMAPs: Mechanisms of neural and synaptic genesis. Muscle Nerve 19:1127-1133, 1996. 317. van Dijk JG, Tjon-A-Tsien A, van der Kamp W: CMAP variability as a function of electrode site and size. Muscle Nerve 18:68-73, 1995. 318. Van Dijk JG, Van Benten I, Kramer CGS, Stegeman DF: CMAP amplitude cartography of muscles innervated by the median, ulnar, peroneal and tibial nerves. Muscle Nerve 22:378389, 1999. 319. van Dijk JG, van der Kamp W, van Hilten BJ, van Someren P: Influence of recording site on CMAP amplitude and on its variation over a length of nerve. Muscle Nerve 17:1286-1292, 1994. 320. Verdugo R, Ochoa JL: Quantitative somatosensory thermotest. A key method for functional evaluation of small calibre afferent channels. Brain 115:893-913, 1992. 321. Vinik AI, Suwanwalaikorn S, Stansberry KB, Holland MT, McNitt PM, Colen LE: Quantitative measurement of cutaneous perception in diabetic neuropathy. Muscle Nerve 18:574584, 1995. 322. Wagman IH, Lesse H: Maximum conduction velocities of motor fibers of ulnar nerve in human subjects of various ages and sizes. J Neurophysiol 15:235-242, 1952. 323. Walker FO: Letters to the Editor: Optimal interelectrode recording distances. Muscle Nerve 536, 1996.
129
324. Walker DD, Kimura J: A fast-recovery electrode amplifier for electrophysiology. Electroencephalogr Clin Neurophysiol 45:789-792, 1978. 325. Wallin BG, Burke D, Gandevia S: Coupling between variations in strength and baroreflex latency of sympathetic discharges in human muscle nerves. J Physiol (Lond) 474:331-338, 1994. 326. Wang AK, Raynor EM, Blum AS, Rutkove SB: Heat sensitivity of sensory fibers in carpal tunnel syndrome. Muscle Nerve 22:37-42, 1999. 327. Wang FC, de Pasqua V, Delwaide PJ: Agerelated changes in fastest and slowest conducting axons of thenar motor units. Muscle Nerve 22:1022-1029, 1999. 328. Wang SJ, Fuh JL, Shan DE, Liao KK, Lin KP, Tsai CP, Wu ZA: Sympathetic skin response and R-R interval variation in Parkinson's disease. Mov Disord 8:151-157, 1993. 329. Wang S-J, Liao K-K, Liou H-H, Lee S-S, Tsai C-P, Lin K-P, Kao K-P, Wu Z-A: Sympathetic skin response and R-R interval variation in chronic uremic patients. Muscle Nerve 17: 411-418, 1994. 330. Wee AS, Ashley RA: Where is the ideal reference site for recording the thenar compound muscle action potential. Electromyogr Clin Neurophysiol 28:249-252, 1988. 331. Weigl P, Bostock H, Franz P, Martius P, Miiller W, Grafe P: Threshold tracking provides a rapid indication of ischaemic resistance in motor axons of diabetic subjects. Electroencephalogr Clin Neurophysiol 73:369-371, 1989. 332. Weinberg CR, Pfeifer MA: An improved method for measuring heart-rate variability: assessment of cardiac autonomic function. Biometrics 40:855, 1984. 333. Wheeler T, Watkins PJ: Cardiac denervation in diabetes. Br Med J 4:584, 1973. 334. Wiederholt WC: Threshold and conduction velocity in isolated mixed mammalian nerves. Neurology (Minneap) 20:347-352, 1970. 335. Winkler T, Stalberg E, Haas LF: Uni-and bipolar surface recording of human nerve responses. Muscle Nerve 14:133-141, 1991. 336. Wu BJ, Neff J, Kingery WS, Date ES: Letters to the Editor: Conduction velocity is inversely related to axonal length in the median sensory nerve. Muscle Nerve 262,1998. 337. Yarnitsky D: Quantitative sensory testing. Muscle Nerve 20:198-204, 1997. 338. Yokota T, Matsunaga T, Okiyama R, Hirose K, Tanabe H, Furukawa T, Tsukagoshi H: Sympathetic skin response in patients with multiple sclerosis compared with patients with spinal cord transection and normal controls. Brain 114:1381-1394, 1991. 339. Zgur T, Vodusek DB, Krzan M, Vrtovec M, Denislic M, Sibanc B: Autonomic system dysfunction in moderate diabetic polyneuropathy assessed by sympathetic skin response and valsalva index. Electromyogr Clin Neurophysiol 33:433-439, 1993. 340. Zwarts MJ, Guechev A: The relation between conduction velocity and axonal length. Muscle Nerve 18:1244-1249, 1995.
Chapter 6 ASSESSMENT OF INDIVIDUAL NERVES
1. INTRODUCTION 2. COMMONLY TESTED NERVES IN THE UPPER LIMB Median Nerve—Motor Fibers Median Nerve—Sensory Fibers Multiple Stimulation Across the Carpal Ligament Ulnar Nerve Radial Nerve 3. OTHER NERVES DERIVED FROM THE CERVICAL OR THORACIC NERVE ROOTS Phrenic Nerve Greater Auricular Nerve Cervical Spinal Nerve and Brachial Plexus Musculocutaneous and Lateral Antebrachial Cutaneous Nerves Medial and Posterior Antebrachial Cutaneous Nerves Intercostal Nerves 4. COMMONLY TESTED NERVES IN THE LOWER LIMB
Tibial Nerve Common and Deep Peroneal Nerve Superficial Peroneal Nerve Sural Nerve 5. OTHER NERVES DERIVED FROM THE LUMBOSACRAL NERVE ROOTS Lumbosacral Plexus Femoral Nerve Saphenous Nerve Lateral Femoral Cutaneous Nerve Posterior Femoral Cutaneous Nerve Medial Femoral Cutaneous Nerve Pudendal Nerve Dorsal Nerve of the Penis 6. CRANIAL NERVES Mylohyoid, Deep Temporal, and Lingual Nerves Accessory Nerve Hypoglossal Nerve
130
Assessment of Individual Nerves 1 INTRODUCTION Nerve conduction studies consist of stimulating a nerve and recording the evoked potential either from the nerve itself or from a muscle innervated by the nerve. The basic principles outlined in the previous section apply to all studies, although the anatomic peculiarities dictate specific approaches to each of the commonly tested individual nerves. This section will describe the usual points of stimulation and recording sites together with the normal values as reported in the literature or established in our institution. To minimize the bias induced by different techniques, each laboratory should develop its own normal ranges, using a standardized method. Most electromyographers assess the latency and conduction velocity against the upper and lower limits of normal defined as a mean plus or minus two standard deviations in a healthy population. The same criterion does not apply to amplitude, which distributes in a non-Gaussian manner. In our experience, most individual measures of amplitude in healthy subjects exceed one half the mean of the control value, which thus serves as a lower limit of normal. An alternative approach uses a log transformation of the amplitude data to accomplish an equal distribution and then express the normal range in terms of plus or minus two standard deviation confidence intervals (see Chapter 3-8). The conduction studies commonly involve the readily available motor and sensory fibers of the median, ulnar, and radial nerves; the motor fibers of the accessory, peroneal, and tibial nerves; and the sensory fibers of the sural and superficial peroneal nerves. Less easily accessible structures include the phrenic nerve, brachial plexus, musculocutaneous, and other nerves of the shoulder girdle; lateral, medial, and posterior antebrachial cutaneous nerves; the dorsal sensory branch of the ulnar nerve; the lumbo-sacral plexus, femoral and sciatic nerves, lateral femoral cutaneous nerve, saphenous nerve, and lateral and medial plantar nerves. The ordinary nerve conduction studies provide limited information regarding the
131
central or most proximal nerve segment such as the radicular portion. Supplemental methods help evaluate the motor and sensory conduction in this region by measuring the F wave, H reflex, or somatosensory evoked potentials (see Chapters 18, 19, and 20). Studies of the facial nerve and blink reflex constitute an integral part of cranial nerve testing (see Chapters 7-3 and 17-2).
2
COMMONLY TESTED NERVES IN THE UPPER LIMB Median Nerve—Motor Fibers
The median nerve runs relatively superficially in its entire course from the axilla to the palm (Fig. 6-1A and B). The conventional sites of stimulation include Erb's point, axilla, elbow, and wrist. Stimulation at Erb's point or at the axilla tends to coactivate other nerves in close proximity.63 The use of the collision technique circumvents that problem (see Chapter 7-3). In our laboratory, we place the cathode over the brachial pulse near the volar crease at the elbow, and 3 cm proximal to the distal crease at the wrist. The anode is located 2 cm proximal to the cathode, with the ground electrode around the forearm between the stimulating and recording electrodes, if necessary to contain a stimulus artifact (Fig. 6-2A). Additionally, the nerve is accessible to92,137,149 percutaneous stimulation in the palm. Tables 6-1 and 6-2 summarize normal values in our laboratory. With stimulation at the wrist, elbow, or axilla, the convention calls for placing the cathode distally to the anode. This arrangement does not work in the palm, where the proximally placed anode could activate the thenar nerve if the distally located cathode has already passed the target point (see Chapter 7-3), thus concealing the actual site of nerve activation. With the reversal of electrode polarity (i.e., the anode located distally), the cathode placed mid-palm elicits no muscle response because neither electrode lies on the nerve. Moving 1-2 cm proximally, the cathode activates the palmar branch of the ulnar
132
Nerve Conduction Studies
Figure 6-1. A. Motor nerve conduction study of the median nerve. The sites of stimulation include Erb's point (A), axilla (B), elbow (C), wrist (D), and palm (E). Compound muscle action potentials are recorded with surface electrodes placed on the thenar eminence. B. Sensory nerve conduction study of the median nerve. The sites of stimulation include axilla (A), elbow (B), wrist (C), and palm (D). Antidromic sensory potentials are recorded with a pair of ring electrodes placed around the second digit.
nerve, adducting the thumb, and 1 cm further proximally, it activates the origin of the thenar nerve, abducting the thumb (Fig. 6-3). Unlike the sensory nerve, the motor axons take a recurrent course along the thenar nerve off the median nerve trunk. Thus, unless dealing with the exposed nerve for intraoperative monitoring,12 palmar stimulation may inadvertently activate the distal segment of the thenar nerve rather than the intended branching point (see Chapter 7-3). Specifically, surface stimulation aimed at the origin of the thenar nerve in the palm commonly depolarizes the distal branch near the motor point, resulting in an erroneously short latency. An unreasonably large latency increase between the wrist and palm then presents a false impres-
sion of carpal tunnel sydrome. Careful selection of the most distal point of palmar stimulation avoids this error, guided by an appropriate thumb twitch, indicating contraction of the abductor pollicis brevis. To further compound the problem, in rare instances the recurrent182branch may take an anomalous course. Recording leads consist of an active electrode (G1) over the belly of the abductor pollicis brevis and an indifferent electrode (G2) just distal to the metacarpophalangeal joint (Fig. 6-2A). Depending on the electrode positioning, the potentials from other intrinsic hand muscles innervated by the median nerve contribute to the evoked response. Comparison between the muscle action potentials from the second lumbrical innervated by
Assessment of Individual Nerves
133
Figure 6-2. A. Motor and sensory conduction studies of the median nerve. The photo shows stimulation at the wrist, 3 cm proximal to the distal crease, and recording over the belly (G1) and tendon (G2) of the abductor pollicis brevis for motor conduction, and around the proximal (G2) and distal (G2) interphalangeal joints of the second digit for antidromic sensory conduction. The ground electrode is located in the palm. B. Alternative recording sites for a sensory conduction study of the median nerve, with the ring electrodes placed around the proximal (G1) and distal (G2) interphalangeal joints of the third digit or the base (G1) and the interphalangeal joint (G2) of the first digit.
the median nerve and the volar interosseous innervated by the ulnar nerve provides an additional technique to evaluate the distal segment35,57,143,159,173,187 A
latency difference greater than 0.4-0.5 ms suggests an abnormal delay in conduction across the distal segment. Recording from the pronator quadratus
Site of Stimulation Motor fibers Palm
Amplitudef: Motor (mV) Sensory (uV)
Table 6-1 Median Nerve* Latency; to Difference Between Recording Site Right and Left (ms) (ms)
6.9 ± 3.2 (3.5)§
1.86 ±0.28 (2.4)'
0.19 ±0.17 (0.5)*
Wrist
7.0 ± 3.0 (3.5)
3.49 ± 0.34 (4.2)
0.24 ± 0.22 (0.7)
Elbow
7.0 ± 2.7 (3.5)
7.39 ± 0.69 (8.8)
0.31 ±0.24 (0.8)
7.2 ± 2.9 (3.5)
9.81 ±0.89 (11.6)
0.42 ±0.33 (1.1)
Palm
39.0 ± 16.8 (20)
1.37 ± 0.24 (1.9)
0.15 ±0.11 (0.4)
Wrist
38.5 ± 15.6 (19)
2.84 ± 0.34 (3.5)
0.18 ±0.14 (0.5)
Elbow
32.0 ± 15.5 (16)
6.46 ± 0.71 (7.9)
0.29 ±0.21 (0.7)
Axilla Sensory fibers Digit
Conduction Time Between Two Points (ms)
Conduction Velocity (m/s)
1.65 ± 0.25 (2.2)*
48.8 ± 5.3 (38)**
3.92 ± 0.49 (4.9)
57.7 ± 4.9 (48)
2.42 ± 0.39 (3.2)
63.5 ± 6.2 (51)
1.37 ±0.24 (1.9)
58.8 ± 5.8 (47)
1.48 ±0.18 (1.8)
56.2 ± 5.8 (44)
3.61 ± 0.48 (4.6)
61.9 ± 4.2 (53)
*Mean ± standard deviation (SD) in 122 nerves from 61 patients, 11 to 74 years of age (average, 40), with no apparent disease of the peripheral nerves. t Amplitude of the "evoked response, measured from the baseline to the negative peak. Latency, measured to the onset of the evoked response, with the cathode at the origin of the thenar in the palm. §Lower limits of normal, based on the distribution of the normative data. 'Upper limits of normal, calculated as the mean + 2 SD. **Lower limits of normal, calculated as the mean - 2 SD.
Assessment of Individual Nerves
Site of Stimulation Motor fibers Wrist Elbow Sensory fibers Palm Wrist
135
Table 6-2 Latency Comparison Between Two Nerves in the Same Limb* Median Nerve Ulnar Nerve (ms) (ms)
Difference (ms)
3.34 ± 0.32 (4.0) f 7.39 + 0.72 (8.8)
2.56 ± 0.37 (3.3)t 7.06 ± 0.79 (8.6)
0.79 ±0.31 (1.4)f 0.59 + 0.60 (1.8)
1.33 ±0.21 (1.8) 2.80 ± 0.32 (3.4)
1.19 ±0.22 (1.6) 2.55 ± 0.30 (3.2)
0.22 ±0.17 (0.6) 0.29 + 0.21 (0.7)
*Mean ± standard deviation (SD) in 70 nerves from 35 patients, 14 to 74 years of age (average, 37), with no apparent disease of the peripheral nerve. fUpper limits of normal, calculated as mean + 2 SD.
helps evaluate the lesion involving the anterior interosseous nerve.127,147 In the presence of an anomalous crossover from the median to ulnar nerve in the forearm, distal and proximal stimulation elicits compound muscle potentials of dissonant wave forms. The latencies of these responses represent two different nerves, precluding their comparison for calculation of the nerve conduction velocity (see Chapter 7-4).
The terminal latency index serves as a measure of the terminal latency adjusted to the terminal distance and expressed as a percentage of the proximal conduction velocity. Thus, it equals terminal distance divided by the product of terminal latency and conduction velocity.49,158,163,164 A value of 0.34 or less suggests disproportionate distal slowing as in the carpal tunnel syndrome and distally prominent polyneuropathy (see Chapter 5-4).
Figure 6-3. Stimulation of the median nerve with the cathode placed at the origin of the recurrent thenar nerve and the anode placed 2 cm distally, and recording of the muscle response over the belly (G1) and tendon (G2) of the abductor pollicis brevis, with the ground electrode placed between the stimulating and recordIng electrodes.
136
Nerve Conduction Studies
Figure 6-4. Stimulation of the median nerve at the wrist and palm with the cathode placed 3 cm proximal, and 5 cm distal, to the wrist crease, and the anode placed 2 cm proximally, and recording of the antidromic digital potential with the ring electrodes placed 2 cm apart around the proximal (G1) and distal (G2) interphalangeal joints of the ring finger (cf. Figure 6-11).
Median Nerve—Sensory Fibers Stimulation delivered at sites listed for the motor fibers also activates antidromic sensory action potentials of the first through fourth digits. Motor axons have a threshold similar to that of the large myelinated sensory axons. Thus, when one studies the mixed nerve, superimposition of action potentials from distal muscles may obscure the antidromically recorded sensory potential. Palmar stimulation distal to the origin of the recurrent motor fibers, however, selectively activates the sensory fibers of the median nerve. This helps identify muscle action potentials, if elicited with more proximal stimulation, by a change in waveform of the evoked response.93 Sensory fibers innervating the second digit originate more from C7 than C6 root and traverse the middle trunk rather than the upper trunk before entering the lateral cord. Thus, the second digit provides far less reliable results than the first digit in detecting upper trunk lesions.56 The sensory potentials recorded from the first or third digit (Fig. 6-2B), or the lateral half of the
fourth digit (Fig. 6-4) often reveal178abnormalities not otherwise detectable. The first digit provides asssessment of the C6 root, upper trunk, and lateral cord, whereas the third digit serves to evaluate the C7 root, middle trunk, and lateral cord. In contrast to postganglionic lesions, which cause degeneration of the sensory axons, preganglionic root avulsion results in no abnormalities of the sensory potential recorded from the anesthestic digits. Table 6-1 summarizes normal values for the digital potentials recorded with ring electrodes placed 2 cm apart around the proximal (G1) and distal (G2) interphalangeal joints of the second digit (Fig. 6-2A). For wrist and palm stimulation, we place the cathode 3 cm proximal and 5 cm distal to93 the distal crease of the wrist (Fig. 6-4). Alternative techniques use a fixed distance from the recording electrode, most commonly 12-14 cm.39 Because of mixed sensory innervation, stimulating the radial nerve also elicits a sensory nerve potential over the first digit; stimulating the ulnar nerve, over the fourth digit. Thus, inadvertent spread of stimulating current to the
Assessment of Individual Nerves other nerves may confuse the issue. Some investigators take advantage of this spread to gain an instantaneous comparison of the median nerve136 to the ulnar nerve79,178 or the radial nerve. Separate stimulation of the median and ulnar nerves at the wrist evokes a corresponding sensory potential of the fourth digit at nealy the identical latency for the same conduction distance (Fig. 6-5). Additional palmar stimulation at a fixed distance from the wrist, usually 8 cm, allows segmental latency calculation as one of the most sensitive, practical measures of comparison between the two nerves (Fig. 6-6). Unnecessarily strong shocks applied to the palm tend to coactivate the median and ulnar sensory fibers innervating the fourth digits. Selective stimulation of one or the other branch results from careful application of electrodes along the line connecting the medial or lateral aspect of the fourth digit and the ulnar or median nerve at the wrist. Slight twich of ulnar or median innervated muscle usually signals proper placement of the stimulator. In our series (Table 6-3), normal values consisted of the onset latency of 2.88 ± 0.35 ms (mean ± SD) after wrist stimulation and distal amplitude of 37.6 ± 17.2 uV after palm stimulation for the median nerve, and 2.86 ± 0.37 ms and 46.1 ± 24.3 uV for the ulnar nerve. The latency difference between the two nerves was 0.01 ± 0.17 ms with an upper limit of normal of 0.4 ms defined as the mean +2 SD. Unlike the compound muscle action potentials that maintain nearly the same amplitude irrespective of stimulus site, the antidromically activated digital potentials diminish substantially with increasing nerve length under study. Indeed, stimulation at Erb's point or the axilla may fail to elicit unequivocal digital potentials without the use of an averaging technique. Here, temporal dispersion between fast- and slow-conducting fibers results in durationdependent phase cancellation (see Chapter 7_5). 13,96 In addition, naturally recurring orthodromic sensory impulses may partially extinguish the antidromic impulse by collision. These tendencies favor a proximal stimulation over a more distal stimulation in proportion to the distance between the stimulating and recording electrodes. Recording of the antidromic sensory po-
137
tentials suffices for routine14,32 clinical purposes. Alternatively, digital or palmar stimulation31,45allows recording of the orthodromic sensory potential at the palm, wrist, or elbow with either surface electrodes or needle electrodes. This method demands a higher resolution to compensate for a smaller size of the orthodromic potential. The averaging technique offers a distinct advantage in detecting such small nerve potentials, especially in a diseased nerve. Women tend to have greater orthodromic median sensory nerve action potential at the wrist than men, possibly reflecting smaller wrist size.110 The palmar cutaneous branch of the median nerve usually arises about 5.5 cm proximal to the radial styloid and innervates skin of the thenar eminence. Antidromic stimulation of the median nerve elicits sensory potentials over the midthenar eminence. In one series, normal values over 10 cm segments included the onset latency of 2.6 ± 0.2 ms (mean ± SD) and amplitude of 12 ± 4.6 uV.111 This technique may help differentiate the carpal tunnel syndrome that spares the palmar cutaneous branch from a more proximal injury.
Multiple Stimulation Across the Carpal Ligament The use of palmar stimulation provides a simple means of identifying conduction abnormalities of sensory or motor fibers under the transverse carpal ligament or along its most terminal segment.108,168 This distinction differentiates the carpal tunnel syndrome from a distal neuropathy seen, for example, in digital nerves of diabetics.22,66 Stimulation of the median nerve at multiple sites across the wrist (Fig. 6-7) further localizes the point of maximal conduction delay within the distal segment of the median nerve.92,.93,128,148 Short segmental stimulation of the motor fibers poses a less technical challenge when recording from the lumbricals than from abductor pollicis brevis (see Chapter 7-3). Incremental stimulation provides the only way to precisely localize a motor lesion, which may deviate from the usual site of compression (see Chapter 26-5). The sensory axons normally show a pre-
Antidromic Sensory Conduction
Antidromic Sensory Conduction
Figure 6-5. A. Antidromic sensory potentials in a healthy subject recorded from the index (top) and ring finger (center) after stimulation of the median nerve at the palm and wrist, and from the ring finger (bottom) after stimulation of the ulnar nerve at the comparable sites (see Figures 6-4 and 6-11). Median and ulnar nerve responses showed nearly identical latencies with stimulation at the palm and at the wrist regardless of the recording fingers. B. The same arrangement as in A except for use of the middle finger (top) instead of the index finger for one of the median responses in another healthy subject.
Antidromic Sensory Conduction
Figure 6-6. A. The same arrangement as in Figure 6-5 in a patient with a mild carpal tunnel syndrome. Despite normal latency from the wrist to the index finger (3.2 ms) and to the ring finger (3.2 ms), the latency difference between median and ulnar nerve (0.7 ms) clearly exceeded the upper limit of normal value (0.4 ms). In contrast, median and ulnar responses showed nearly identical latencies with stimulation at the palm regardless of the recording finger, confirming a delay of median conduction between wrist and palm. B. Another patient with carpal tunnel syndrome showing a more pronounced latency difference (0.9 ms) between median and ulnar nerves and a reduced amplitude of median nerve response recorded from the ring finger. A normal median response elicited by palm stimulation suggests focal demyelination across the carpal ligament with no evidence of distal axonal degeneration.
Antidromic Sensory Conduction
Table 6-3 Distal Sensory Conduction Study Comparing Median and Ulnar Nerves* Recording Median Nerve 2nd Digit Median Nerve 4th Digit Ulnar Nerve 4th Digit Median & Ulnar Difference
Stimulation Palm
Measurement of Antidromic Sensory Potential Latency Amplitude (ms) (uV) 1.43 ±0.16 (1.7)* 49.8 ±21. 5 (25)§
Wrist
38.4 ± 15.6 (19)
2.87 ±0.31 (3.5)
Palm
37.6 ± 17.2 (19)
1.45 ±0.20 (1.9)
Wrist
22.3 ±8.2 (11)
2.88 ± 0.35 (3.6)
Palm
46. 1 ± 24.3 (23)
1.48 ± 0.26 (2.0)
Wrist
29.0 ± 14.8 (25)
2.86 ± 0.37 (3.6)
Palm
8.5 ± 20.7
0.02 ±0.1 7 (0.3)
Wrist
5.9 ± 10.1
0.01 ±0.17 (0.4)
Calculated Values for Wrist to Palm Segment Conduction Conduction time (ms) Velocity (m/s) 1.44 ± 0.20 (1.9)*
57.1 ± 8.3 (40)**
1.43 ± 0.22 (1.9)
57.4 ± 8.9 (40)
1.38 ± 0.30 (1.8)
59.1 ± 8.3 (43)
0.04 ± 0.20 (0.4)
*Mean ± standard deviation (SD) in 31 healthy subjects, 16 to 64 years of age (average 38), with no apparent disease of the peripheral nerve. tAmplitude of the evoked response, measured from the baseline to the negative peak. ^Latency, measured to the onset of the evoked response, with a standard distance of 8 cm between the stimulus sites at the wrist and palm. §Lower limits of normal, based on the distribution of the normative data. '"Upper limits of normal, calculated as the mean + 2 SD.
141
Assessment of Individual Nerves dictable latency change of 0.16-0.20 ms/cm with series of stimulation from mid-palm to distal forearm in 1 cm increments (Fig. 6-7B). A sharply localized latency increase across a 1 cm segment indicates focal abnormalities of the median nerve (Fig. 6-7, C, D, E). A nonlinear jump in latency usually accompanies an abrupt change in waveform showing abnormal temporal dispersion. A paradoxical increase in size of responses proximal to this point indicates the loss of physiologic phase cancellation because excessive desynchronization no longer superimposes fast and slow signals (see Chapter 7-5). Stimulation of the median nerve at the digit95 or at the elbow70 evokes orthodromic and mixed nerve po-
tentials simultaneously recordable at several sites across the carpal tunnel with multi-channel-recording electrodes. This technique provides instantaneous comparison of latencies but not amplitudes, which vary so much depending on the depth of the nerve at the site of recording.95,156 Ulnar Nerve Like the median nerve, the ulnar nerve takes a relatively superficial course along its entire length. Routine motor conduction studies consist of stimulating the nerve at multiple sites and recording the muscle potential from the hypothenar muscles with
Site of
Stimulation
Figure 6-7. A. Twelve sites of stimulation in 1 cm increments along the length of the median nerve. The "0" level at the distal crease of the wrist corresponds to the origin of the transverse carpal ligament. The photo shows a recording arrangement for sensory nerve potentials from the second digit and muscle action potentials from the abductor pollicis brevis. [From Kimura, with permission.] B. Sensory nerve potentials in a normal subject recorded after stimulation of the median nerve at multiple points across the wrist. The numbers on the left indicate the site of each stimulus (compare with A). The latency increased linearly with stepwise shifts of stimulus site proximally in 1 cm increments. [From Kimura,93 with permission.]
Figure 6-7. C. Sensory nerve potentials in a patient with the carpal tunnel syndrome. Both hands showed a sharply localized slowing from -2 to -1 with the calculated segmental conduction velocity of 14 m/s on the left (top) and 9 m/s on the right (bottom). Note a distinct change in waveform of the sensory potential at the point of localized conduction delay. Double-humped appearance at -2 on the left suggests sparing of some sensory axons at this level. Temporarily dispersed responses on the right at — 1 and beyond had greater negative and positive peaks in area compared to normal, more distal responses, presumably because of loss of physiologic phase cancellation (see Chapter 7-5). [From Kimura,93 with permission.] D. Sensory nerve potential in a patient with the carpal tunnel syndrome. Both hands show a sharply localized slowing from -3 to -2, with a segmental conduction velocity of 10 m/s on the left (top) and 7 m/s on the right (bottom). An abrupt change 93in waveform of the sensory potential also indicates the point of localized conduction delay. [From Kimura, with permission.] E. Sensory nerve potential in a patient with the carpal tunnel syndrome before (A) and after (B) surgery. Preoperative study showed a localized slowing from -4 to -3 with a calculated segmental conduction velocity of 8 m/s, which normalized in a repeat study conducted six months postoperatively.
142
Assessment of Individual Nerves
143
Figure 6-8. A. Motor and sensory conduction study of the ulnar nerve. The photo shows stimulation at the wrist, 3 cm proximal to the distal crease, and recording over the belly (G1) and tendon (G2) of the abductor digit minimi for motor conduction, and around the proximal (G1) and distal (G2) interphalangeal joints of the fifth digit for antidromic sensory conduction. B. Alternative recording sites for motor and sensory conduction studies of the ulnar nerve with the surface electrodes over the belly (G1) and tendon (G2) of the first dorsal interosseous muscle for motor conduction and around the proximal (G1) and distal (G2) interphalangeal joints of the fourth digit for antidromic sensory conduction. C. Sensory conduction study of the dorsal cutaneous branch of the ulnar nerve. The photo shows stimulation along the medial aspect of the forearm between the tendon of the flexor carpi ulnaris and the ulna, 14-18 cm from the active electrode, and recording over the dorsum of the hand between the fourth and fifth metacarpals (G1) and the base of the fifth digit (G2).
surface electrodes placed over the belly of the abductor digit minimi (G1) and its tendon (G2) 3 cm distally (Fig. 6-8A). Alternative recording sites include forearm
muscles such as flexor carpi ulnaris177 or flexor digitorum profundus.53 Common sites of stimulation include palm, wrist, axilla, and Erb's point (Fig. 6-9A,B). The use
144
Nerve Conduction Studies
Figure 6-9. A. Motor nerve conduction study of the ulnar nerve. The sites of stimulation include Erb's point (A), axilla (B), above the elbow (C), elbow (D), below the elbow (E), and wrist (F). Compound muscle action potentials are recorded with surface electrodes placed on the hypothenar eminence. B. Sensory nerve conduction study of the ulnar nerve. The sites of stimulation include axilla (A), above the elbow (B), elbow (C), below the elbow (D), wrist (E), and palm (F). The tracings show antidromic sensory potentials recorded with the ring electrodes placed around the fifth digit.
of a fixed distance from the distal crease of the wrist or from the recording electrode improves the accuracy of latency comparison between the two sides and among different subjects. In our laboratory, we place the cathode 3 cm proximal to the distal crease of the wrist and the anode 2 cm further, proximally. Spread of stimulus current at Erb's point or in the axilla causes less obvious problems in studying the ulnar nerve, as compared with the median nerve, because the hypothenar eminence contains only ulnar-innervated muscles. Nonetheless, coactivation of the median nerve gives rise to volume-conducted potentials from the thenar eminence, unless
eliminated by the collision technique.91 Tables 6-2 and 6-4 show the normal values in our laboratory. Stimulation of the motor fibers above and below the elbow helps document a tardy ulnar palsy and a cubital tunnel syndrome. For accurate determination of conduction velocity, the distance between the proximal and distal sites of stimulation should exceed 10 cm to minimize measurement error. The conventional studies often fail to uncover the abnormalities early because a focal slowing induces an insignificant delay when calculated over a longer segment. Segmental stimulation across the elbow in 1-2 cm
Table 6-4 Ulnar Nerve* Amplitudef: Motor (mV) Sensory
G*v)
Latency to Recording Site (ms)
Difference Between Right and Left (ms)
5.7 ± 2.0 (2.8)§
2.59 ± 0.39 (3.4)*
0.28 ± 0.27 (0.8)'"
Below elbow
5.5 ± 2.0 (2.7)
6. 10 ±0.69 (7.5)
0.29 ± 0.27 (0.8)
Above elbow
5.5 ± 1.9 (2.7)
8.04 ± 0.76 (9.6)
0.34 ± 0.28 (0.9)
Axilla
5.6 ±2.1 (2.7)
9.90 ±0.91 (11.7)
0.45 ±0.39 (1.2)
Wrist
35.0 ± 14.7 (18)
2.54 ±0.29 (3.1)
0.18 + 0.13 (0.4)
Below elbow
28.8 ± 12.2 (15)
5.67 ± 0.59 (6.9)
0.26 ± 0.21 (0.5)
Above elbow
28.3 ± 11.8 (14)
7.46 ± 0.64 (8.7)
0.28 ± 0.27 (0.8)
Site of Stimulation Motor fibers Wrist
Conduction Time Between Two Points (ms)
Conduction Velocity (m/s)
3.51 ± 0.51 (4.5)*
58.7 ± 5.1 (49)**
1.94 + 0.37 (2.7)
61.0 ± 5.5 (50)
1.88 ±0.35 (2.6)
66.5 ± 6.3 (54)
2.54 ±0.29 (3.1)
54.8 ± 5.3 (44)
3.22 ±0.42 (4.1)
64.7 ± 5.4 (53)
1.79 ± 0.30 (2.4)
66.7 ± 6.4 (54)
Sensory fibers Digit
*Mean ± standard deviation (SD) in 130 nerves from 65 patients, 13 to 74 years of age (average, 39), with no apparent disease of the peripheral nerves. fAmplitude of the evoked response, measured from the baseline to the negative peak. Latency, measured to the onset of the evoked response, with the cathode 3 cm above the distal crease in the wrist. §Lower limits of normal, based on the distribution of the normative data. '"Upper limits of normal, calculated as the mean + 2 SD. **Lower limits of normal, calculated as the mean — 2 SD.
146
Nerve Conduction Studies
Figure 6-1O. Stimulation of the ulnar nerve in the palm with the cathode placed over the palmar branch and the anode 2 cm distally, and recording of the muscle response over the belly of the adductor pollicis brevis (G1) referenced to the thumb (G2). Appropriate thumb twitch confirms activation of the deep palmar branch of the ulnar nerve as opposed to the recurrent thenar nerve, which usually lies 1 cm more proximally (cf. Figure 6-3).
increments detects an abrupt change in latency and waveform of the compound action potential at the site of localized compression.18,83,94 The ulnar nerve slides back and forth in the cubital tunnel with flexion and extension of the elbow joint.67 Thus, normal values vary depending on 8 the position of the elbow and, to a lesser degree, of the wrist.151 Holding the arm either at 135° or 90° flexion during stimulation and measurement minimizes the error.98,100 The study of the deep palmar motor branch depends on recording the muscle potential from the first dorsal interosseous or adductor pollicis after stimulation of the ulnar nerve at the wrist (Fig. 6-8B). The latency difference between the hypothenar and thenar responses provides a measure of conduction along the deep branch. In one series, the upper limit of the normal range based on 373 studies included 4.5 ms for the distal latency to the first dorsal interosseous, 2.0 ms for the latency difference between this muscle and adductor
digiti minimi and 1.3 ms for the latency difference between the two sides.132 In the assessment of the deep palmar branch, the size of muscle response elicited by stimulation in the palm distal to the site of the lesion provides a good measure of the number of remaining motor axons (Fig. 6-10). Lumbrical-interosseous comparison described for median nerve study (see above) also serves in assessing a distal ulnar nerve lesion, which typically causes a latency difference greater than 0.2 ms in the reverse direction.101,160 Stimulation of the ulnar nerve trunk elicits an antidromic sensory potential of the fourth and fifth digits (Fig. 6-8A,B). The common sites of cathodal points include above and below the elbow,54 3 cm proximal to the distal crease at the wrist, and 5 cm distal to the crease in the palm, with the anode located 2 cm further proximally (Fig. 6-11). These stimulus sites make the studies comparable to those of the median nerve (see Fig. 6-1). The fourth and fifth digits provide assessment of C8 and Tl
Assessment of Individual Nerves
147
Figure 6-11. Stimulation of the ulnar nerve at the wrist and palm with cathode placed 3 cm proximal, and 5 cm distal to the wrist crease and the anode placed 2 cm proximally, and recording of the antidromic digital potential with the ring electrodes placed 2 cm apart around the proximal (G1) and distal (G2) interphalangeal joints of the ring finger. This arrangement yields results directly comparable to the analogous study of the median nerve (cf. Figs. 6-4, 6-5, and 6-6).
roots, lower trunk, and medial cord. Stimulation of the digital nerve with ring electrodes placed around the interphalangeal joints of the fifth digit, cathode proximally, elicits orthodromic sensory potential at various sites along the course of the nerve. Stimulation of the nerve at the palm or wrist gives rise to a mixed nerve potential of the ulnar nerve proximally (Fig. 6-12). These studies help differentiate lesions of C8 and Tl roots from those of the lower trunk, medial cord of the brachial plexus, or ulnar nerve. Preganglionic C8 and Tl root avulsion should spare sensory potentials despite clinical sensory loss. The dorsal sensory branch, called the dorsal ulnar cutaneous nerve, leaves the common trunk of the ulnar nerve 5-8 cm proximal to the ulnar styloid.77,89 It becomes superficial between the tendon of the flexor carpi ulnaris and the ulna.10 Surface stimulation here selectively evokes antidromic sensory potentials over the dorsum of the hand, although anatomic variations may alter cutaneous innervation.138
Placing the active electrode (G1). between the fourth and fifth metacarpals optimizes the recording with the reference electrode (G2) at the base of the fifth digit (Fig. 6-8C). Stimulation of the ulnar nerve trunk more proximally elicits a mixed nerve potential that slightly precedes a large muscle action potential from the intrinsic hand muscles. The dorsal ulnar cutaneous nerve, like the ulnar nerve proper, derives from C8-T1 roots, the lower trunk and the medial cord, but it escapes compression at Guyon's canal. The normal values of the sensory potential established in one study77 include amplitude of 20 ± 6 uV with distal stimulation, distal latency of 2.0 ± 0.3 ms (mean ± SD) when recorded 8 cm from the point of stimulation and conduction velocity of 60 ± 4.0 m/s between elbow and forearm. This technique complements the conventional study of the ulnar nerve after a severe lesion at the wrist that has precluded the recording from the hypothenar muscles or digits. It also helps localize a lesion within
Nerve Conduction Studies
148
Figure 6-12. Stimulation of the ulnar nerve in the palm with the cathode placed 2 cm proximal to the anode, and recording of mixed nerve potential with the active electrode (G1) over the ulnar nerve trunk 8 cm proximal to the cathode and the reference electrode (G2) 2 cm further proximally.
the forearm in the segment proximal or distal to the take-off of this branch with its origin an average distance of 6.4 cm above the wrist.10 Its abnormality implies axonal degeneration with localization of the lesion to a more proximal site. Conversely, the presence of a normal response combined with abnormal digital ulnar sensory potential usually,68,89 though not always,179 localizes an ulnar neuropathy to the wrist.
Radial Nerve The radial nerve becomes relatively superficial at supraclavicular fossa, in the axilla near the spinal groove, above the elbow, and in the forearm (Fig. 6-13A.B). The optimal sites of electrical stimulation of the motor fibers therefore include (1) Erb's point, (2) between the coracobrachialis and medial edge of the triceps about 18 cm proximal to the medial epicondyle, (3) between the brachioradialis and the tendon of the biceps 6 cm proximal to the lateral epicondyle, and (4) between the extensor carpi ulnaris and ex-
tensor digiti minimi on the dorsal aspect of the ulna, 8 to 10 cm proximal to the styloid process. Either a needle electrode or surface electrodes suffice (Fig. 6-14A) when recording muscle action potentials from the extensor digitorum communis188 or the extensor indicis. In motor conduction studies, commonly encountered errors result from such technical problems as submaximal stimulation in an obese or muscular limb, coactivation of a number of extensors, and distortion of the waveform by volume-conducted potentials from distant muscles. Further, distal stimulation activates fewer muscles than does proximal stimulation, making a valid comparison between the two responses difficult. The use of needle electrodes for stimulation and recording helps circumvent some of these limitations.50 Needle electrodes also enable relatively selective recording from more proximal muscles such as the anconeus, brachioradialis, and triceps. In assessing the axilla to elbow segment, anterior surface tape measurement compares most favorably with the actual anatomic length.80
Assessment of Individual Nerves
149
V
Figure 6-13. A. Motor nerve conduction study of the radial nerve. The sites of stimulation include Erb's point (A), axilla (B), above the elbow (C), and mid-forearm (D). Compound muscle action potentials are recorded from the extensor indicis with a pair of surface electrodes. B. Sensory nerve conduction study of the radial nerve. The sites of stimulation include elbow (A) and distal forearm (B). Antidromic sensory potentials are recorded using the ring electrodes placed around the first digit.
The sensory branches run deep at the level of the elbow, where the posterior antebrachial cutaneous nerve emerges to innervate the dorsolateral aspect of the forearm.23 It then becomes more superficial about 10 cm above the lateral styloid process. The sensory fibers cross the extensor 41,42 pollicis longus at the base of the thumb and are palpable at this point. Percutaneous stimulation at the lateral edge of the radius in the distal forearm 10-14 cm proximal to the base of the thumb elicits an antidromic sensory potential recordable by a pair of ring electrodes placed around the thumb (Fig. 6-14B). Alternative arrangements combine the disc electrode (G1) over the first web space or slightly more proximally in the snuffbox, with the reference electrode (G2) near the first dorsal interosseous113,116
or between the second and third metacarpals.167 An additional stimulation at the elbow under the brachioradialis muscle lateral to the biceps tendon (see Fig. 6-13B) allows determination of conduction velocities in the segments between elbow and wrist and wrist and thumb.24,52,157,162 Sensory fibers innervating the thumb originate from C6 and C7 roots and traverse upper and middle trunk before entering the posterior cord. Preganglionic avulsion of the C6 and C7 roots results in a clinical sensory loss associated with no abnormalities of the sensory potentials. Stimulation of the radial nerve at the thumb or the wrist elicits orthodromic sensory potentials at the elbow or axilla. Spread of current to the median nerve, which partially supplies the thumb, accounts for 25 percent of the sensory po-
Figure 6-14. A. Motor and sensory conduction studies of the radial nerve. The photo shows stimulation in the forearm with the cathode at the lateral edge of the extensor carpi ulnaris muscle, 8 to 10 cm proximal to the styloid process. The monopolar needle electrode (G1) is inserted in the extensor indicis with a reference electrode (G2) over the dorsum of the hand laterally for motor conduction studies. The recording electrodes are placed around the base (G1) and interphalangeal joint (G2) of the first digit for antidromic sensory conduction. B. Alternative stimulation and recording sites for antidromic sensory nerve conduction study of the radial nerve. The photo shows the cathode placed at the lateral edge of the radius in the distal forearm, with the anode placed 2 cm proximally. The recording electrodes are placed either around the base (G1) and interphalangeal joint (G2) of the first digit or over the palpable nerve between the first and second metacarpals (G1) and 2-3 cm distally (G2).
151
Assessment of Individual Nerves Table 6-5 Radial Nerve Conduction Motor Axilla-elbow Elbow-forearm Forearm-muscle Sensory Axilla-elbow Elbow-wrist Wrist-thumb
n
Conduction Velocity (m/s) or Conduction Time (ms)
Amplitude: Motor (mV) Sensory (uV)
Distance (cm)
8 10 10
69 ± 5.6 62 ±5.1 2.4 ± 0.5
11 ±7.0 13 ± 8.2 14 ± 8.8
15.7 ± 3.3 18.1 ± 1.5 6.2 ± 0.9
16 20 23
71 ± 5.2 69 ± 5.7 58 ± 6.0
4± 1.4 5 ±2.6 13 ± 7.5
18.0 ± 0.7 20.0 ± 0.5 13.8 ± 0.4
Source: From Trojaborg and Sinrup,175 with permission.
tentlal recorded over the radial nerve at the wrist or elbow, and 175 50 percent of that recorded at the axilla. Stimulation at the wrist, especially with needle electrodes placed along the nerve, accomplishes more selective activation of the radial nerve. Table 6-5 summarizes the results in one series.175 Orthodromic potentials may be recorded from the snuffbox after stimulation of the third digit,81 indicating inconsistent anomalous innervation of this finger by the radial nerve.181
anode. With an optimally placed needle, shocks of very low intensity suffice for selective stimulation of the phrenic nerve, contracting the diaphragm as evidenced by hiccup or interruption of voluntarily sustained vocalization. Supramaximal stimulation may coactivate the brachial plexus located posteriorly behind the anterior scalene muscle. The diaphragmatic action potential gives rise to a strong positMry at the
3 OTHER NERVES DERIVED FROM THE CERVICAL OR THORACIC NERVE ROOTS Phrenic Nerve Conduction studies of the phrenic nerve, though described early0,33,130 have not gained popularity in part because surface stimulation in the cervical area requires shocks of a relatively high intensity. Moreover, some patients tolerate the esophageal electrode used for recording the diaphragmatic potentials poorly. As an alternative method, some investigators118 use a standard monopolar needle electrode inserted medially from the lateral aspect of the neck at the level of the cricoid cartilage (Fig. 6-15). After traversing the posterior margin of the sternocleidomastoid muscle, the needle tip comes to within a few millimeters of the phrenic nerve and adequately distant from the carotid artery anteriorly and the apex of the lung inferiorly. A metal plate placed on the manubrium serves as the
Figure 6-15. Motor conduction study of the phrenic nerve. The diagram shows stimulation with a needle inserted medially through the posterior margin of the sternocleidomastoid at the level of the cricoid cartilage. The recording electrodes are placed on the xiphoid process (G1) and at the eighth intercostal space near the costochondral junction (G2). [From MacLean and Mattioni,118 with permission.]
Nerve Conduction Studies
152 Table 6-6 Phrenic Nerve
Authors Newsom 78Davis (1967) Delhez 16 (1965) MacLean and Mattioni (1981)68
Stimulation Point
Recording Site
No. 18
Needle electrode placed posterior to sternocleidomastoid
Xiphoid process
Amplitude (uV)
Duration (ms)
Onset Latency (ms) 7.7 ± 0.80
Difference Between Sides (ms)
7.5 ± 0.53 30 on right 8. 2 ±0.71 30 on left 30 8.5 ± 40.5 48.1 ± 12.2 7.4 ±0.59 0.08 ± 0.42
7th or 8th intercostal space near the costochondral junction and a mild negativity at the xiphoid process.120 Paired surface electrodes placed over these recording sites, therefore, register the largest amplitude with summation of out-of-phase activities. Normal ranges established using 118 needle stimulation in 30 healthy subjects very closely approximate the earlier results obtained by surface stimulation (Table 6-6).130 Phrenic nerve conduction studies complement needle electromyography of the diaphragm by identifying the nature and site of disorder of the respiratory system.9 Diaphragmatic compound muscle action potentials show good intraindividual sideto-side agreement for latency but not for amplitude.171 Nonetheless, amplitude value serves as a better measure than the latency in predicting respiratory dysfunction.21 In one study of 50 phrenic nerves in 25 healthy subjects,25 normal values (mean ± SD) included the latency of 6.54 ± 0.77 ms and the amplitude of 660 ± 201 uV, with the right-left difference of 0.34 ± 0.27 ms and 66.3 ± 65.3 uV.
Greater Auricular Nerve The greater auricular nerve, derived mainly from the C2 and C3 roots, winds around the posterior border of the sternomastoid and ascends cephalad on the surface of that muscle from the neck to the ear. Stimulation with a pair of surface electrodes firmly placed against the lateral border of the sternocleidomastoid muscle elicits an orthodromic sensory potential easily detectable on the back of the ear lobe. Re-
ported values include latency of 1.7 ± 0.2 ms (mean ± SD) for the distance of 8 cm and conduction velocity 133 of 46.8 ± 6.6 m/s in 20 healthy subjects, and latency of 1.9 ± 0.2 ms, and amplitude of 22.4 ± 8.9 uV in 32 normal control subjects.97
Cervical Spinal Nerve and Brachial Plexus The brachial plexus comprises the anterior rami of the spinal nerves derived from the C5 through C8, and T1 roots. Surface stimulation at Erb's point (see Fig. 1-8) activates the 60 proximal muscles of the shoulder girdle. It also evokes action potentials in the distal muscles such as those of the thenar and hypothenar eminence. The volume-conducted potentials from a number of coactivated muscles interfere with the accurate recording of the intended signal even with the electrode placed over a specific intrinsic hand muscle. A collision technique circumvents this difficulty by blocking the unwanted impulse with a second stimulus applied distally to the nerve not under consideration (see Chapter 7-3). The use of needle electrodes accomplishes more selective stimulation but carries the risk of inducing pneumothorax.135 The triceps has the endplate zone vertically oriented with the distal portion of the muscle innervated by longer nerve branches. Thus, the latency of a recorded response increases with the distance from the stimulus point. The latency changes nonlinearly reflecting irregularly spaced points of innervation. The biceps and deltoid muscles have one or more horizontaly directed endplates mostly in the mid-
Assessment of Individual Nerves dle of the fibers.121,123,125 The point of recording does not affect the latency of the response in these muscles as much as in the triceps. The same probably applies to the infraspinatus and supraspinatus. Recording from the serratus anterior85 permits conduction studies of the long thoracic nerve.139,140 The needle electrodes register from a more limited area, providing a reliable measure of latencies, even with102simultaneous activation of many nerves. Intramuscular recordings, however, fail to reveal the true waveform of the compound muscle action potential because of restricted recording area. When testing a unilateral involvement of the brachial plexus, comparison between the affected and normal sides offers the most sensitive indicator (Table 6-7). The standard protocol calls for equalizing the distance between the stimulating and recording electrodes on both sides. This principle holds in the study of any muscle of the shoulder girdle, and particularly that of the triceps for the reasons stated previously. A localized stimulus applied through a needle electrode can directly activate the spinal nerve at the junction of the respective ventral and dorsal roots.7,86,118,119,126 The uninsulated tip comes to an optimal position when a standard 50-75 mm monopolar needle, inserted perpendicular to the skin surface, rests directly on the vertebral transverse process. Joint stimulation of the C5 and C6 spinal nerves by placing the needle 1-2 cm lateral to the C5 spinous process tests the upper trunk and lateral cord (Fig. 6-16A). Similarly,
Muscle Biceps Deltoid Triceps Supraspinatus Infraspinatus
153
positioning the needle slightly caudal to the C7 spinous process stimulates the C8 and Tl spinal nerves simultaneously for conduction across the lower trunk and medial cord (Fig. 6-16B). The needle inserted between these two points activates the C6, C7, and C8 spinal nerves simultaneously, for evaluation of the posterior cord. A metal plate or disk electrode on the skin surface or a second needle electrode serves as the anode. Alternatively, placing the anode over the T2 spinous process allows activation of the C8 and Tl with the stimulating cathode inserted at the C5-C6 level, minimizing the risk of pneumothorax.135,153 Recording from several muscles helps evaluate different portions of the brachial plexus—for example, biceps for the upper trunk and lateral cord, triceps for the posterior cord, and ulnar-innervated intrinsic hand muscles for the lower trunk and medial cord. Table 6-8 summarizes the conduction time across the brachial plexus calculated by subtracting the distal latency of the ulnar nerve.17 The sideto-side difference exceeding 0.6 ms indicates unilateral lesions, making it a more sensitive index than the absolute latency.
Musculocutaneous and Lateral Antebrachial Cutaneous Nerves Optimal sites of stimulation for motor conduction129,172 include the posterior cervical triangle 3 to 6 cm above the clavicle just behind the sternocleidomastoid
Table 6-7 Nerve Conduction Times From ERB's Point to Muscle n Distance (cm) 19 15 14 20 17 16 23 16 19 16 20 15
Source: Modified from Gassel,60 with permission.
20 24 28
15.5 18.5 21.5 26.5 31.5 8.5
10.5 14 17
Latency (ms) 4.6 ± 0.6 4.7 ± 0.6 5.0 ±0.5 4.3 ± 0.5 4.4 ± 0.4 4.5 ± 0.4 4.9 ± 0.5 5.3 ± 0.5 2.6 ± 0.3 2.7 ± 0.3 3.4 ± 0.4 3.4 ± 0.5
Nerve Conduction Studies
154
Figure 6-16. A. C5 and C6 root stimulation. The diagram shows the needle inserted perpendicular to the skin, 1-2 cm lateral to the C5 spinous process. B. C8 and Tl 1 root stimulation. The diagram shows the needle inserted slightly caudal to the C7 spinous process. [From MacLean,117 with permission.]
muscle (see Fig. l-9)61,102 and the axilla between the axillary artery medially and the coracobrachialis muscle laterally.145 Either surface electrodes or needle electrodes suffice to stimulate the nerve and to record the muscle action potentials from the biceps brachii (Table 6-9). The sensory branch runs superficially at the level of the elbow, just lateral to the tendon of the biceps. Stimulation of the nerve between the tendon of the biceps medially and the brachioradialis laterally elicits orthodromic sensory potentials recordable at
the posterior cervical triangle and axilla by the same electrodes positioned to stimulate motor fibers. The same stimulus also elicits antidromic sensory potential of the distal branch, the lateral antebrachial cutaneous nerve of the forearm (Fig. 6-17). The recording electrode is placed 12 cm distally over the course of the nerve in the forearm, along the straight line from the stimulus point to the radial artery at the wrist. Table 6-10 summarizes normal values reported in two series.74,166 Study of the musculocutaneous nerve provides evaluation of the
Table 6-8 Brachial Plexus Latency with Nerve Root Stimulation Plexus Brachial (upper trunk and lateral cord) Brachial (posterior cord) Brachial (lower trunk and medial cord)
Site of Stimulation C5 and C6 C6, C7, C8 C8 and Tl Ulnar nerve
Source: From MacLean,117 with permission.
Recording Site Biceps brachii Triceps brachii Abductor digiti quinti
Latency Across Plexus (ms) Range Mean SD 0.4 4.8-6.2 5.3 4.4-6.1 5.4 0.4 4.7 0.5 3.7-5.5
Table 6-9 Musculocutaneous Nerve
Age
n
15-24 25-34 35-^4 45-54 55-64 65-74
14 6 8 10 9 4
Motor Nerve Conduction Between Erb's Point and Axilla Range of Range of Amplitude (/iV) Conduction Velocity (m/s) Axilla Erb's Point 7-27 63-78 9-32 60-75 8-30 6-26 6-24 58-73 8-28 6-22 55-71 7-26 53-68 7-24 5-21 6-22 5-19 50-66
Source: From Trojaborg,172 with permission.
Orthodromic Sensory Nerve Conduction Between Erb's Point and Axilla
n
Range of Conduction Velocity (m/s)
14 6 7 10 9 4
59-76 57-74 54-71 52-69 49-66 47-64
Range of Amplitude (/iV) 3.5-30 3-25 2.5-21 2-18 2-15 1.5-12
Orthodromic Sensory Nerve Conduction Between Axilla and Elbow
n 15 8 8 13 10 6
Range of Conduction Velocity (m/s)
Range of Amplitude (/iV)
61-75 59-73 57-71 55-69 53-67 51-65
17-75 16-72 16-69 15-65 14-62 13-59
156
Nerve Conduction Studies
Figure 6-17. Sensory conduction study of the lateral cutaneous nerve of the forearm. The photo shows stimulation just lateral to the tendon of the biceps and recording from the nerve with the electrodes placed 12 cm distal to the cathode along the straight line to the radial artery (G1) and 2-3 cm further distally (G2).
C6 root, upper trunk, and lateral cord better than the median sensory potentials recorded from the second digit, which, more often than not, represent the C7 root and middle trunk.56
Medial and Posterior Antebrachial Cutaneous Nerves The medial antebrachial cutaneous nerve, like the ulnar nerve, originates from the C8 and Tl roots via the lower trunk and medial cord." It subserves the sensation over the medial aspect of the forearm, the area not affected by lesions of the ulnar nerve. The nerve pierces the deep fascia 4 cm above the elbow on a line bisecting the distance between the biceps tendon and the medial epicondyle. Surface stimulation at this point elicits antidromic sensory potentials best recorded over the course of its volar branch on the same line extended distally 8 cm from the elbow (Fig. 6-18). Table 6-10 shows the results of two studies.74,144 The posterior antebrachial cutaneous
nerve, derived from the C5 through C8 roots and the posterior cord, separates from the radial nerve in the spiral groove and innervates the skin of the lateral arm and the dorsal forearm. At its origin, it pierces the lateral head of the triceps, separating into proximal and distal branches. Surface stimulation above the lateral epicondyle, between the biceps and triceps brachii, elicits antidromic sensory potentials recordable with surface electrodes placed 12 cm distally along the line extended from the stimulus point to the wrist, midway between the ulnar and radial styloid processes (Fig. 6-19).
Intercostal Nerves Surface stimulation of this nerve elicits intercostal muscle action potentials with inconsistent latency. Recording from the rectus abdominis muscle improves reproducibilites of the waveform and allows calculation of conduction velocity after stimulating the nerve at two points.142
157
Assessment of Individual Nerves Table 6-10 Lateral and Medial Cutaneous Nerve (Mean ± SD)
Authors Spindler and Felsenthal166 Izzo et al.74
Reddy144
4
Number Latency . Conduction of Patients Age Distance Onset Peak Velocity Amplitude Nerve Seen (mean) (cm) (ms) (ms) (m/s) (uV) 24.0 ±7.2 1.8 ±0.1 2 .3 ±0.1 Lateral 12 65 ± 4 20-84 30 (35) cutaneous nerve 18.9 ±9.9 14 2 .8 ±0.2 62 ± 4 154 Lateral 17-80 (45) cutaneous nerve 11.4 ±5.2 2,,7 ±0.2 Medial 14 63 ± 5 17-80 155 (45) cutaneous nerve 15.4 ±4.1 66 ± 4 2.7 ±0.2 3 .3 ±0.2 Medial 18 23-60 30 (38) cutaneous nerve
COMMONLY TESTED NERVES IN THE LOWER LIMB
Tibial Nerve Motor conduction studies record the muscle response from one of the intrinsic foot
muscles after stimulation of the tibial nerve at the popliteal fossa and at the ankle posterior to the medial malleolus. The nerve bifurcates into two branches within 1 cm of the malleolar-calcaneal axis in 90 percent of feet.34 The usual choices for recording sites include the abductor hallucis and flexor pollicis brevis, innervated by the me-
Figure 6-18. Stimulation of the medial antebrachial cutaneous nerve of the forearm with the cathode placed medial to the brachial artery 4 cm above the elbow crease on a line drawn from the ulnar styloid process to a point halfway between the medial epicondyle and biceps brachii tendon, and recording of the antidromic sensory potential with the active electrode (G1) 8 cm distal to the elbow crease and the reference electrode (G2), 3-4 cm further distally along the same line. This arrangement yields results directly comparable to the analogous study of the lateral antebrachial cutaneous nerve (cf. Fig. 6-17).
158
Nerve Conduction Studies
Figure 6-19. Stimulation of the posterior antebrachial cutaneous nerve with the cathode placed just above the lateral epicondyle between the biceps brachii and triceps brachii, and recording of the antidromic sensory potentials with the active electrode (G1) 12 cm distally and the reference electrode (G2), 3-4 cm further distally along a line extended from the stimulus point to the mid-dorsum of the wrist, midway between the ulnar and radial styloid processes.
dial plantar nerve, and the abductor digiti quinti, supplied by the lateral plantar nerve (Figs. 6-20 and 6-21A,B). One study reports normal distal latencies (mean ± SD) of 4.9 ± 0.6 ms for medial and 6.0 ± 0.7 ms for lateral plantar nerves over a 12 cm segment.72 Stimulation of the tibial nerve above and below the medial malleolus determines the conduction characteristics of the motor fibers across the tarsal tunnel.55
Reported normal values across a 10 cm segment (mean ± SD) include 3.8 ± 0.5 ms for the medial and 3.9 ±0.5 ms for the lateral plantar nerves.58 Tables 6-11 and 6-12 summarize the normal values in our laboratory. Sensory conduction studies consist of stimulating the medial or lateral plantar nerves on the sole 11-13 cm distal to the G1 electrode184 and recording orthodromic
Figure 6-20. Motor conduction study of the tibial nerve. The sites of stimulation include the knee (A), above the medial malleolus (B) and below the medial malleolus (C). Compound muscle action potentials are recorded with surface electrodes placed over the abductor hallucis.
Figure 6-21. A. Motor conduction study of the medial plantar nerve. The photo shows stimulation of the tibial nerve posterior to the medial malleolus, 10 cm from the recording electrodes placed over the belly (G1) and tendon (G2) of the abductor hallucis. B. Motor conduction study of the lateral plantar nerve. The photo shows stimulation posterior to the medial malleolus and recording with surface electrodes placed on the belly (G1) and tendon (G2) of the abductor digiti quinti.
Nerve Conduction Studies
160 Table 6-11 Tibial Nerves* Site of Stimulation Ankle
Amplitude (mV) 5.8 ± 1.9 (2.9)§
Latency to Recording Site (ms) 3.96 ± 1.00 (6.0)'
Knee
5.1 ± 2 . 2 (2.5)
12.05 ± 1.53 (15.1)
Difference Between Two Sides (ms) 0.66 ±0.57 (1.8)*
0.79 ± 0.61 (2.0)
Conduction Time Between Two Points (ms)
Conduction Velocity (m/s)
8.09 ± 1.09 (10.3)
48.5 ± 3.6 (41)**
*Mean ± standard deviation (SD) in 118 nerves from 59 patients, 11 to 78 years of age (average, 39), with no apparent disease of the peripheral nerves. tAmplitude of the evoked response, measured from the baseline to the negative peak. Latency, measured to the onset of the evoked response, with a standard distance of 10 cm between the cathode and the recording electrode. §Lower limits of normal, based on the distribution of the normative data. .'Upper limits of normal, calculated as the mean + 2 SD. **Lower limits of normal, calculated as the mean - 2 SD.
sensory potentials with surface or needle electrodes placed just below the medial malleolus (Fig. 6-22 and 6-23).6,141,152 Alternative sites of stimulation include the first and fifth toes with a pair of ring electrodes. The medial plantar potentials have average latencies (mean ± SD) of 2.4 ± 0.2 ms, 3.2 ± 0.3 ms, and 4.0 ± 0.2 ms for 10, 14, and 18 cm segments, respectively. The lateral plantar latencies average 3.2 ± 0.3 ms and 4.0 ± 0.3 ms for 14 and 18 cm segments. As a modification of this method, selective stimulation of the interdigital nerve also gives rise to an orthodromic sensory potential for assessment of interdigital neuropathy or Joplin's neuroma.51,131 Stimulation on the medial aspect of the hallux selectively activates the terminal sensory branch of the medial plantar nerve, or medial plantar proper digital nerve, another uncommon site of Joplin's neuroma.28 The responses recorded at the knee after stimulation of the tibial nerve at the ankle comprise orthodromic sensory and antidromic motor potentials.124 Stimulation of the tibial nerve below the medial malleolus elicits the antidromic sensory
nerve potentials of the medial and lateral 73,90 plantar nerves at the first and fifth toes and of the medial calcaneal nerve at the heel.134 In these cases, the use of an averaging technique improves the resolution of small signals that would otherwise escape detection. The study of the plantar nerves helps evaluate the integrity of the postganglionic sensory fibers derived from the L4 and L5 roots,65for example, in patients with footdrop. Common and Deep Peroneal Nerve Stimulation of the common peroneal nerve above or below the head of the fibula or just above the ankle elicits muscle action potentials in the extensor digitorum brevis (Figs. 6-24 and 6-25). This muscle, primarily supplied by the deep peroneal nerve, may also receive an anomalous innervation from the superficial peroneal nerve. The communicating branch, called the accessory deep peroneal nerve, passes behind the lateral malleolus to reach the
Table 6-12 Latency Comparison Between Two Nerves in the Same Limb* Site of Stimulation Ankle Knee
Peroneal Nerve 3.89 ± 0.87 (5.6) 12.46 ± 1.38 (15.2)
Tibial Nerve 4.12 ± 1.06 (6.2) 12.13 ± 1.48 (15.1)
Difference 0.77 ±0.65 (2.1) 0.88 ± 0.71 (2.3)
*Mean ± standard deviation (SD) in 104 nerves from 52 patients, 17 to 86 years of age (average, 41), with no apparent disease of the peripheral nerve. fUpper limits of normal, calculated as the mean + 2 SD.
Assessment of Individual Nerves
161
Figure 6-22. Stimulation of the sensory branch of the medial and lateral plantar nerves with the cathode placed over the medial and lateral aspects in the mid-portion of the sole and the anode placed 2 cm further distally, and recording of the orthodromic sensory nerve potential with the active electrode (G1)placed immediately posterior to the medial malleolus 11-13 cm from the cathode, and reference electrode (G2) 3-4 cm further proximally.
Orthodromic Sensory Conduction
Figure 6-23. Orthodromic sensory nerve potentials of the medial (two top tracings) and lateral plantar nerves (two bottom tracings) recorded from the tibial nerve at the ankle following stimulation of each nerve on the sole in a 48-year-old healthy man (cf. Fig. 6-22).
lateral portion of the muscle. In the presence of this anomaly, stimulation of the deep peroneal nerve at the ankle evokes a much smaller compound muscle action potential than the shocks applied at the knee (see Chapter 7-4). For accurate determination of conduction velocity across the knee, the distance between the proximal and distal sites of stimulation should exceed 10 cm. A series of shocks applied in short increments, however, is better suited for delineating a focal conduction abnormality.82,94 In an advanced neuropathy, recording from the ex30 tensor digitorum longus or tibialis anterior36 instead of from the atrophic extensor digitorum brevis may facilitate the assessment. Stimulation of the peroneal nerve at the ankle elicits mixed nerve potentials at the fibula head.62 The use of needle electrode and averaging technique improves resolution in recording small potentials of the deep peroneal sensory nerve105 from the web between the first and second toes. Tables 6-12 and 6-13 summarize the normal values in our laboratory.
Nerve Conduction Studies
162
Figure 6-24. Motor conduction study of the common peroneal nerve. The sites of stimulation include above the knee (A), below the knee (B), and at the ankle (C). Compound muscle action potentials are recorded with surface electrodes over the extensor digitorum brevis.
Superficial Peroneal Nerve This mixed nerve, derived from the L5 root, originates below the fibular head as a branch of the common peroneal nerve. It gives rise to two sensory nerves in the lower third of the leg, the medial and intermedi-
Figure 6-25. Motor conduction study of the common peroneal nerve. The photo shows stimulation over the dorsum of the foot near the ankle, 7 cm from the recording electrodes over the belly (G1) and tendon (G2) of the extensor digitorum brevis.
ate dorsal cutaneous nerves. They innervate the skin of the dorsum of the foot and the anterior and lateral aspects of the leg. The medial dorsal cutaneous nerve pierces the superficial fascia at the anterolateral aspect of the leg about 5 cm above and 2 cm medial to the lateral malleolus.19,37 Stimulation at this point with the cathode adjusted to produce a sensation radiating into the toes elicits antidromic sensory potential over the dorsum of the foot medially. The averaging technique helps identify the potential with amplitude approximately half that of the sural nerve, especially in recording from a diseased nerve. In another method,75,78 stimulation of the intermediate dorsal cutaneous branch with the cathode placed against the anterior edge of the fibula elicits the antidromic sensory potential at the ankles just medial to the lateral malleolus (Fig. 6-26). Stimulation of the nerve at two points, 12-14 cm from the recording electrode and 8-9 cm further proximally, allows assessments of the distal and proximal segments. The study of this sensory nerve helps distinguish an L-5 radiculopathy from more distal lesions.78 The near nerve needle recording with signal averaging makes it possible to assess small sensory 134 action potential from interdigital nerves. Table 6-14 summarizes the normal values.
Sural Nerve This sensory nerve, primarily derived from the S1 root, originates in the popliteal
Assessment of Individual Nerves
163
Table 6-13 Common and Deep Peroneal Nerves* Site of Stimulation Ankle
Latency} to Amplitude! (mV) 5.1 ± 2.3 (2.5)§
Difference Between Recording Site (ms) 3.77 ± 0.86 (5.5)^
Conduction Time Right and Left (ms) 0.62 ± 0.61 (1.8)'
Below knee
5.1 ± 2.0 (2.5)
10.79 ± 1.06 (12.9)
0.65 ± 0.65 (2.0)
Above knee
5.1 ± 1.9 (2.5)
12.51 ± 1.17 (14.9)
0.65 ± 0.60 (1.9)
Conduction Between Two Points (ms)
Velocity (m/s)
7.01 ± 0.89 (8.8)*
48.3 ± 3.9 (40)**
1.72 ± 0.40 (2.5)
52.0 ± 6.2 (40)
*Mean ± standard deviation (SD) in 120 nerves from 60 patients, 16 to 86 years of age (average, 41), with no apparent disease of the peripheral nerves. tAmplitude of the evoked response, measured from the baseline to the negative peak. ^Latency, measured to the onset of the evoked response, with a standard distance of 7 cm between the cathode and the recording electrode. §Lower limits of normal, based on the distribution of the normative data. 'Upper limits of normal, calculated as the mean + 2 SD. **Lower limits of normal, calculated as the mean - 2 SD.
fossa as the medial sural branch of the tibial nerve. It becomes superficial at the junction of the mid and lower third of the leg, where it receives a communicating
Figure 6-26. Sensory conduction study of the superficial peroneal nerve. The photo shows stimulation against the anterior edge of the fibula, 12 cm from the active electrode (G1) located just medial to the lateral malleolus at the ankle with the reference electrode (G2) placed 2-3 cm distally.
branch of the common peroneal nerve. In some cases, the peroneal branch contributes more than the main trunk from the tibial nerve. Descending toward the ankle, it turns anterolaterally along the inferior aspect of the lateral malleolus. Its terminal branch, the lateral dorsal cutaneous nerve, supplies the lateral aspect of the dorsum of the foot. The sural nerves may contain some motor3 fibers in about 6 percent of individuals. Stimulation of the nerve in the lower third of the leg over the posterior aspect slightly lateral to the midline elicits antidromic sensory potentials, usually recorded around the lateral malleolus (Figs. 6-27 and 6-28), but at times more distally for the study106 of the lateral dorsal cutaneous branch. Sural potentials need no averaging for recording except perhaps in older population or patients with diseased nerve.15,17,38 Segmental studies dividing the nerve into three contiguous portions of 7 cm each have revealed a smaller mean velocity in the most distal segment than in the middle or proximal segment.176 Averaging technique facilitates the study of orthodromic potentials after stimulation of the nerve over the lateral aspect of the foot.4,5,69,87,161 Segmental studies depend on recording at the popliteal fossa and high at the ankle, 10-15 cm proximal to the lateral malleolus (Table 6-15). Near-nerve technique revealed a greater latency when measured from the stimulus to the recording sites than the true conduction time calculated as latency difference over the same
Table 6-14 Superficial Peroneal Nerve Stimulation Point 5 cm above, 2 cm medial to lateral malleolus
Recording Site Dorsum of foot
Amplitude
13.0 ± 4.6
Latency (ms) 1.22 ± 0.40
Conduction Velocity (m/s) 53. 1 ± 5.3 (Distal segment)
Over 15
13.9 ± 4.0
(Peak) 2.24 ± 0.49
47.3 ± 3.4
(Distal segment)
3-60
20.5 ±6.1
(Peak) 2.9 ± 0.3
65.7 ± 3.7
(Proximal segment)
18.3
(Peak) 2.8 ± 0.3
51.2 ± 5.7
(Proximal segment)
n 50
Age 1-15
50 Anterior edge of fibula, 12 cm above the active electrode
Medial border of lateral malleolus
50
Anterolateral aspect of leg, 14 cm above the active electrode
Medial border of lateral malleolus
80
G*v)
(Onset) 37
Source: Data from Di Benedetto,
78
Jabre,
75
and Izzo et al.
Assessment of Individual Nerves
165
Figure 6-27. Antidromic sensory nerve conduction study of the sural nerve. The diagram shows stimulation on the calf slightly lateral to the midline in the lower third of the leg, and recording with surface electrodes placed behind the lateral malleolus.
segment. This discrepancy results from the latency of activation at the stimulus site, or utilization time of about 0.15 ms, depending on the type of stimuli.103 The nearnerve potential recorded at midcalf showed a 32 percent higher amplitude in women than in men, probably reflecting different volume conductor properties.69
Sural nerve study conducted with care174 offers one of the most sensitive means of detecting electrophysiologic abnormalities in various types of neuropathies. In addition to absolute amplitude, sural to radial amplitude ratio may serve as a sensitive measure. In one study, 150 a ratio less than 0.40, as compared to the normal mean of
Figure 6-28. Sensory conduction study of the sural nerve. The photo shows stimulation along the posterior surface of the leg, slightly lateral to the midline and 7-10 cm from the ankle. The active electrode (G1) is placed above or immediately below and behind the lateral malleolus with the reference electrode 2-3 cm distally along the lateral dorsum of the foot (G2).
Table 6-15 Sural Nerve Authors Shiozawa161and Mavor DiBenedetto37 Behse and 4 Buchthal Wainapel et al.180 Truong 176 et al. Kimura (Unpublished)
Stimulation Point Foot
Recording Site High ankle
n 40
Lower third of leg
Lateral malleolus
38 62
15 cm above lateral malleolus Lower third of leg Distal 10 cm
Dorsal aspect of foot
71
Lateral malleolus
80
Middle 10 cm Proximal 10 cm 14 cm above lateral malleolus
Lateral malleolus Lateral malleolus
Amplitude Age
0*v)
13-41
6.3 (1.9-17)
1-15 Over 15 15-30 40-65
23.1 ±4.4 23.7 ± 3.8
20-79
18.9 ± 6.7
102 102 102 52
10-40 41-84
20.9 ± 8.0 17.2 ± 6.7
Latency (ms)
Conduction Velocity (m/s) 44.0 ± 4.7
1.46 ±0.43 2.27 ± 0.43 (Peak)
52.1 ±5.1 46.2 ± 3.3
3.7 ± 0.3 (Peak)
41.0 ± 2.5
2.7 ± 0.3 2.8 ± 0.3 (Onset)
51.2 ±4.5 48.3 ± 5.3
33.9 ± 3.25 51.0 51.6 52.5 51.1
±3.8 ±3.8 ± 5.6 ±5.9
Assessment of Individual Nerves 0.71, predicted axonal neuropathy. Sural nerve study also provides a unique opportunity for direct comparison between physiologic and histologic findings of the biopsied specimen (see Chapter 4-4).44 Preganglionic pathology consistently spares the sensory action potential despite the clinical symptoms. Thus, studies of the sural nerve help distinguish peripheral lesions from S1 or S2 radiculopathy or cauda equina involvement. 5
OTHER NERVES DERIVED FROM THE LUMBOSACRAL NERVE ROOTS Lumbosacral Plexus
The lumbosacral plexus consists of the lumbar plexus with fibers derived from the L2, L3, and L4 roots and the sacral plexus, which arises from the L5, S1, and
167
S2 roots. Conventional conduction studies fall short of adequately evaluating their integrity because of their inaccessibiity to percutaneous electrical stimulation. The use of the F wave and H reflex provides an indirect measure of impulses propagating across this region (see Chapters 18 and 19). An alternative method involves needle stimulation47,48,117,126 or percutaneous high voltage electrical stimulation71 of L4, L5, or S1 spinal nerve just proximal to the plexus and stimulation of the peripheral nerve just distal to the plexus. Conduction time through the plexus then equals the difference between the distal and proximal latencies. The study of the lumbar plexus involves the stimulation of the L4 spinal nerve by a 75 mm standard monopolar needle, placed so as to lie just below the level of the iliac crest. The needle inserted into the paraspinous muscle perpendicular to the skin surface must reach the periosteum of the articular process (Fig. 6-29A.B). With
Figure 6-29. A. Motor conduction study of the lumbar plexus. The diagram shows stimulation of the L4 root, with the needle inserted perpendicular to the skin just below the level of the iliac crest, and of femoral nerve distal to the inguinal ligament immediately lateral to the femoral artery. Muscle potentials are recorded with surface electrodes over the vastus medialis (G1) and patella (G2).B. Motor nerve conduction study of the sacral plexus. The diagram shows stimulation of the S1 root with the needle inserted at the level of the posterior iliac spine, of the L5 root halfway in between the L4 and S1 roots and of the sciatic nerve at the level of the gluteal skinfold midpoint between the ischial tuberosity and the greater trochanter of the femur. The recording electrodes (not shown) are placed on the belly (G1) and 112 tendon (G2) of the tibialis anterior for L5 and of the abductor hallucis for S1 root studies. [From MacLean, with permission].
Nerve Conduction Studies
168
an optimal needle position, a shock of very low intensity elicits the maximal compound muscle action potential of the vastus medialis. Stimulation of the femoral nerve just distal to the inguinal ligament, with either a surface or a needle electrode, provides the distal latency (Fig. 6-29A). The nerve lies immediately lateral to the readily palpable femoral artery, as discussed later in this section. The study of the sacral plexus involves inserting a needle between the spinous process and posterior iliac spine for the S1 spinal nerve and halfway in between the L4 and S1 spinal nerves for the L5 spinal nerve. At the level of the gluteal skin fold, the sciatic nerve bisects a line drawn between the ischial tuberosity and the greater trochanter of the femur. Needle stimulation here provides the distal latency required for calculation of conduction time across the sacral plexus (Fig. 6-29B). With careful adjustment of the needle position, shocks of very low intensity elicit a maximal compound muscle action potential of the tibialis anterior for the L-5 and of the abductor hallucis for the S1 spinal nerve. Inadvertent activation of the neighboring spinal nerves induces volume-conducted potentials from distant muscles. Without proper care to avoid such spread of stimulus, the recording electrodes placed over the tibialis anterior, for example, would regularly register a simultaneously activated action potential of the triceps surae. Table 6-16 summarizes the normal value in one series.117 The commercially available magnetic coils fail to optimally stimulate lumbosacral roots as diagnostic aids.47,48,115 Specially constructed large-diameter coils, placed flat on the skin surface, however, adequately excite the cauda equina lying
deep below the surface.114 Over proximal cauda equina, cranially directed induced current via vertically oriented coil junction preferentially activates root entry zone of the conus medullaris. Over distal cauda equina, horizontally oriented junction excites the lumbar roots—and vertically oriented junction, sacral roots—at or near the intervertebral foramina. The latency difference between proximal and distal stimulation typically yields the onset latency of 1.9 ms for vastus medialis, 2.3 ms for tibialis anterior, and 3.5 ms for abductor hallucis (see Chapter 21-5). High-voltage electrical stimulation given percutaneously can also activate the sciatic nerve for proximal and segmental nerve conduction measurements.71 This type of stimulation simultaneously excites the peroneal and tibial division of the sciatic nerve, requiring the collision technique 91to eliminate the unintended impulse. Femoral Nerve Shocks delivered to the femoral nerve above or below the inguinal ligament elicits the response recordable in the rectus femoris muscle at various distances from the point of stimulation. The latency of the response increases progressively with the distance reflecting vertical orientation of the endplate region.59 The femoral nerve conducts at an average rate of 70 m/s, based on the latency difference between the two responses recorded at 14 and 30 cm from the point of stimulation (Table 6-17). This calculation, however, does not hold unless all branches supplying proximal and distal parts of the muscle have similar and directly comparable electrophysiologic characteristics.
Table 6-16 Lumbosacral Plexus Plexus Lumbar Sacral
Site of Stimulation L2, L3, L4 Femoral nerve L5 and SI Sciatic nerve
From MacLean,117 with permission.
Recording Site Vastus medialis
Latency Across Plexus (ms) Range Mean SD 2.0-4.4 3.4 0.6
Abductor hallucis
2.5-4.9
3.9
0.7
Assessment of Individual Nerves
169
Table 6.-17 Femoral Nerve Stimulation Point Just below inguinal ligament
Recording Site 14 cm from stimulus point 30 cm from stimulus point
No. 42
8-79
Onset Latency (ms) 3.7 ± 0.45
42
8-79
6.0 ± 0.60
Age
Conduction Velocity (m/s) 70 ± 5.5 between the two recording sites
Source: Modified from Gassel,59 with permission.
Saphenous Nerve This largest and longest sensory branch of the femoral nerve lies deep along the medial border of the tibialis anterior tendon (Fig. 6-30). The nerve stimulation uses the surface electrodes pressed firmly between
the medial gastrocnemius muscle and tibia, usually 12-14 cm above the ankle. Signal averaging improves the resolution of small antidromic sensory potentials recorded just anterior to the highest prominence of the medial malleolus (Table 63-18). Orthodromic studies46,155,.169 consist of stimulating the nerve at two levels, anterior to the medial malleolus, and medial to the knee, and recording the evoked potential with a needle electrode placed near the femoral nerve trunk at the inguinal ligament. The orthodromic potentials average one half the size of the antidromic potentials in amplitude. The saphenous nerve may degenerate with postganglionic lesions such as lumbar plexopathy or femoral neuropathy. In contrast, preganglionic L3 or L4 radiculopathy spares the distal sensory nerve potentials despite clinical deficits. Lateral Femoral Cutaneous Nerve
Figure 6-30. Sensory conduction study of the saphenous nerve. The photo shows stimulation 14 cm above the active electrode (Gi) along the medial surface of the leg between the tibia and the gastrocnemius and recording at the ankle 2-3 cm above (G1) and just anterior to the medial malleolus (G2).
The nerve becomes superficial about 1012 cm below the anterior superior ilac spine, where it divides into large anterior and small lateral branches. Surface stimulation at this point elicits the orthodromic sensory potential recordable with a needle electrode inserted 1 cm medial to the112lateral end of the inguinal ligament. Alternative technique consists of stimulation at the inguinal ligament with a needle electrode and recording antidromic sensory potentials from the thigh (Fig. 6-31). In one study16 using a pair of specially constructed 1.2 x 1.9 cm lead strips fastened 4 cm apart, the normal
Nerve Conduction Studies
170 Table 6-18 Saphenous Nerve
Knee—Medial Malleolus
Inguinal Ligament—Knee Method
Age
Number
Amplitude (uV)
Orthodromic Orthodromic
33 28 41
4.2 ±2.3 5.5 ±2.6 5.1 ±2.7
Wainapel et al.180 Antidromic
17-38 40 20-79
Senden et al.155
18-56
71
Authors Ertekin46 Stohr et al.169
Orthodromic
Conduction Velocity
59.6 ± 2.3 58.9 ± 3.2 57.9 ± 4.0 Peak latency of 3.6 ± 1.4 for 14 cm
Number
Amplitude (uV)
Conduction Velocity (m/s)
10 22 32 80
4.8 ± 2.4 2.1 ± 1.1 1.7± 0.8 9.0 ± 3.4
52.3 ±2.3 51.2 ±4.7 50.2 ±5.0 41.7 ±3.4 54.8 ± 1.9
values (mean ± SD) in 25 healthy adults consisted of a latency of 2.6 ± 0.2 ms, an amplitude of 10-25 uV,V, and a calculated velocity165of 47.9 ± 3.7 m/s. In another study, the antidromic potentials recorded 25 cm distal to the stimulating electrode along the line connecting the stimulus site and the lateral edge of the patella showed onset conduction velocity of 62.3 ± 5.5 m/s (mean ± SD) and amplitude of 2.00 ± 1.0 uV.
placed 6 cm above the midpopliteal fossa register an antidromic sensory potential after stimulating the nerve 12 cm further proximally on a line drawn to the ischial tuberosity. Normal values (mean ± SD) obtained in 40 subjects43 with a mean age of 34 years included peak latency of 2.8 ± 0.2 ms (range, 2.3-3.4 ms) and amplitude of 6.5 ± 1.5 uV (range, 4.1-12.0 uV). This method may help evaluate the peripheral nerve in a patient with lower limb amputations.
Posterior Femoral Cutaneous Nerve
Medial Femoral Cutaneous Nerve
This sensory nerve originates from the anterior and posterior divisions of the S1, S2, and S3 roots, exits the pelvis distal to the piriformis muscle and proceeds distally between the medial and lateral hamstring muscles. Recording electrodes
Sensory abnormalities occasionally involve anterior medial thigh innervated by this nerve. Conduction studies107 may help distinguish this condition from radiculopathy Involving the L2 and L3 roots with overlapping dermatomal distribution.104
Figure 6-31. Sensory nerve conduction of the lateral femoral cutaneous nerve. The diagram shows stimulation above the inguinal ligament and recording over the thigh 12 cm below the anterior-superior iliac spine (G1) and 2-3 cm distally (G2). [From Butler, Johnson and Kaye,16 with permission.]
Assessment of Individual Nerves Pudendal Nerve The technique consists of stimulating the pudenda! nerve and recording compound muscle action potential from the external anal sphincter.170 A specially constructed disposable electrode, when properly mounted onto the gloved right hand, has the stimulating cathode at the tip of index finger. Locating the ischial spine and lateral margin of the sacrum with the fingertip inserted into the rectum helps place the cathode near the pudendal nerve. Methodical exploration then identifies the optimal location, which elicits maximal and reproducible muscle response. Latency values exceeding26,88,183 2.2 ms suggest pudendal neuropathy. Dorsal Nerve of the Penis Stimulation with a pair of electrodes placed at the base of the penis, cathode 2 cm distal to anode, gives rise to antidromic sensory nerve potential recordable at the distal shaft along the dorsal midline with G1 placed 2 cm proximal to G2.11,29 The latency measured to the peak of the negative wave after averaging the response 20 times yielded conduction velocity of 26. 9 m/s for flaccid and 29.7 m/s for stretched shaft.186 A specially constructed urinary catheter electrode placed in the urethra also registers sensory potential following stimulation of the dorsal nerve of the penis. 185 6 CRANIAL NERVES The most commonly tested cranial nerves in the electromyographic laboratory include the facial and trigeminal nerves (see Chapters 7-3 and 17-2). Mylohyoid, Deep Temporal, and Lingual Nerves The mandibular nerve comprises motor axons originating in the trigeminal motor nucleus in the mid pons,2 proprioceptive afferents having their cell bodies in the
171
mesencephalic nucleus of the midbrain,84 and sensory fibers arising from the gasserian ganglion.20 The jaw jerk reflex, elicited by tapping on the chin, evaluates jaw closure (see Chapter 19-3) whereas the blink reflex assesses the afferent trigeminal fibers and the efferent facial nerve (see Chapter 17-2). Less commonly used techniques include motor studies of40 the mylohyoid and deep temporal nerves and 109 sensory conduction of the lingual nerve. Intraoral surface stimulation of the mylohyoid nerve evokes the mylohyoid muscle potential under the chin in the anterior submandibular area. The cathode, taped to a tongue depressor, faces anteriorly in the pterygomandibular space at the level of the rear molars. The subject opens the mouth and pushes the tongue up against the front teeth to activate the muscle for placement of the active recording electrode. In one study of 42 healthy subjects,40 who all had the response bilaterally, the reported value included latency of 1.9 ± 0.2 ms (mean ± SD) and amplitude of 4.9 ±1.8 mV. For stimulation of the deep temporal nerve, the cathode, placed in the pterigomandibular fossa faces posteriorly near the upper rear molar. The patient activates the temporalis muscle by clenching the jaw for placement of the active recording leads. Only 60 percent of healthy subjects had bilateral responses, showing an average latency of 2.1 ± 0.3 ms and amplitude of 4.3 ± 2.0 mV. Stimulation of the mandibular nerve by a needle electrode inserted in the infratemporal fossa at the level of the foramen ovale elicits muscle action potentials of the masseter and mylohyoideus.109 The same needle registers sensory nerve action potentials elicited by stimulation of the lingual nerve along the inferolateral edge of the tongue and of the inferior alveolar nerve at the mental foramen. This method may prove useful in measuring the lingual and inferior alveolar nerve lesion subsequent to dental or orthognathic surgery.76,154 Accessory Nerve The accessory nerve runs superficially along the posterior border of the stern-
Nerve Conduction Studies
172
ocleidomastoid muscle. Surface stimulation at this point elicits a compound muscle action potential of the trapezius, usually recorded from the upper portion by an active electrode (G1) placed at the angle of the neck and shoulder and a reference electrode (G2) over the tendon near the acromion process. Some electromyographers prefer needle electrodes to stimulate the nerve.139,140 In one series of 25 subjects, 10-60 years of age, normal latencies to the upper27trapezius ranged from 1.8 ms to 3.0 ms. In another study of 21 nerves, the onset latency (mean ± SD) averaged 3.0 ± 0.2 ms to the middle trapezius 64and 4.6 ± 0.3 ms to the lower trapezius. Changes in amplitude also provide reliable information, with reduction to one half that of the response on the healthy side suggesting distal degeneration. Hypoglossal Nerve Submandibular surface stimulation of this nerve evokes glossal muscle action potential detectable over the anterior surface of the tongue. In one series of 30 normal subjects studied on both sides, reported values (mean ± SD) included latency of 2.2 ± 0.4 ms and amplitude of 3.8 ± 1.6 mV taking the best of five146responses measured baseline to peak. REFERENCES 1. American Association of Electrodiagnostic Medicine: Practice parameter for electrodiagnostic studies in ulnar neuropathy at the elbow: Summary statement. Muscle Nerve 22:408-411, 1999. 2. Adams RD, Victor M: In Lamsback WJ, Navrozov M (eds): Principles of Neurology. New York, McGraw-Hill, 1993, p 216. 3. Amoiridis G, Schols L, Ameridis N, Przuntek H: Motor fibers in the sural nerve of humans. Neurology 49:1725-1728, 1997. 4. Behse F, Buchthal F: Normal sensory conduction in the nerves of the leg in man. J Neurol Neurosurg Psychiatry 34:404-414, 1971. 5. Behse F, Buchthal F: Sensory action potentials and biopsy of the sural nerve in neuropathy. Brain 101:473-493, 1978. 6. Belen J: Orthodromic sensory nerve conduction
of the medial and lateral plantar nerves, a standardization. Am J Phys Med 64:17-23, 1985. 7. Berger AR, Busis NA, Logigian EL, Wierzbicka M, Shahani BT: Cervical root stimulation in the diagnosis of radiculopathy. Neurology 37:329332, 1987. 8. Bielawski M, Hallett M: Position of the elbow in determination of abnormal motor conduction of the ulnar nerve across the elbow. Muscle Nerve 12:803-809, 1989. 9. Bolton CF: AAEM minimonograph #40: Clinical neurophysiology of the respiratory system. Muscle Nerve 16:809-818, 1993. 10. Botte MJ, Cohen MS, Lavernia CJ, von Schroeder HP, Gellman H, Zinberg EM: The dorsal branch of the ulnar nerve: An anatomic study. J Hand Surg 15A:603-607, 1990. 11. Bradley WE, Lin JTY, Johnson B: Measurement of the conduction velocity of the dorsal nerve of the penis. J Urol 131:1127-1129, 1984. 12. Brown WF, Ferguson GG, Jones MW, Yates SK: The location of conduction abnormalities in human entrapment neuropathies. Can J Neurol Sci 3:111-122, 1976. 13. Buchthal F, Rosenfalck A: Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 3:1-122, 1966. 14. Buchthal F, Rosenfalck A, Trojaborg W: Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 37:340-360, 1974. 15. Burke D, Skuse NF, Lethlean AK: Sensory conduction of the sural nerve in polyneuropathy. J Neurol Neurosurg Psychiatry 37:647-652, 1974. 16. Butler ET, Johnson EW, Kaye ZA: Normal conduction velocity in the lateral femoral cutaneous nerve. Arch Phys Med Rehabil 55:31-32, 1974. 17. Bye A, Fagan E: Nerve conduction studies of the sural nerve in childhood. J Child Neurol 3:94-99, 1988. 18. Campbell WW, Pridgeon RM, Sahni KS: Short segment incremental studies in the evaluation of ulnar neuropathy at the elbow. Muscle Nerve 15:1050-1054, 1992. 19. Cape CA: Sensory nerve action potentials of the peroneal, sural and tibial nerves. Am J Phys Med 50:220-229, 1971. 20. Carpenter MB, Sutin J (eds): Human Neuroanatomy. Williams & Wilkins, Baltimore, 1983. 21. Carter GT, Kilmer DD, Bonekat HW, Lieberman JS, Fowler Jr WM: Evaluation of phrenic nerve and pulmonary function in hereditary motor and sensory neuropathy, type I. Muscle Nerve 15:459-462, 1992. 22. Casey EB, Le Quesne PM: Digital nerve action potentials in healthy subjects, and in carpal tunnel and diabetic patients. J Neurol Neurosurg Psychiatry 35:612-623, 1972. 23. Chang CW, Cho HK, Oh SJ: Posterior antebrachial cutaneous neuropathy: Case report. Electromyogr Clin Neurophysiol 29:109-111, 1989. 24. Chang CW, Oh SJ: Sensory nerve conduction study in forearm segment of superficial radial
Assessment of Individual Nerves nerve: Standardization of technique. Electromyogr Clin Neurophysiol 30:349-351, 1990. 25. Chen R, Collins S, Remtulla H, Parkes A, Bolton CF: Phrenic nerve conduction study in normal subjects. Muscle Nerve 18:330-335, 1995. 26. Cheong DMO, Vaccaro CA, Salanga VD, Waxner SD, Phillips RC, Hanson MR: Electrodiagnostic evaluation of fecal incontinence. Muscle Nerve 18:612-619, 1995. 27. Cherington M: Accessory nerve: conduction studies. Arch Neurol (Chicago) 18:708-709, 1968. 28. Cichy SW, Claussen GC, Oh SJ: Electrophysiological studies in Joplin's neuroma. Muscle Nerve 18:671-672, 1995. 29. Clawson DR, Cardenas DD: Dorsal nerve of the penis nerve conduction velocity: A new technique. Muscle Nerve 14:845-849, 1991. 30. Colachis III SC, Klejka JP, Shamir DY, Pease WS, Johnson EW: Amplitude of M responses: Side to side comparability. Am J Phys Med Rehabil 72:19-22, 1993. 31. Daube JR: Percutaneous palmar median nerve stimulation for carpal tunnel syndrome. Electroencephalogr Clin Neurophysiol 43:139-140, 1977. 32. DeLean J: Transcarpal median sensory conduction: Detection of latent abnormalities in mild carpal tunnel syndrome. Can J Neurol Sci 15:388-393, 1988. 33. Delhez L: Modalites, chez 1'homme normal, de la reponse electrique des piliers du diaphragme a la stimulation electrique des nerfs phreniques par des chocs uniques. Arch Int Physiol Biochim 73:832-839, 1965. 34. Dellon AL, Mackinnon SE: Tibial nerve branching in the tarsal tunnel. Arch Neurol 41: 645-646, 1984. 35. Desjacques P, Egloff-Baer S, Roth G: Lumbrical muscles and the carpal tunnel syndrome. Electromygr Clin Neurophysiol 20:443-450, 1980. 36. Devi S, Lovelace RE, Duarte N: Proximal peroneal nerve conduction velocity: Recording from anterior tibial and peroneus brevis muscles. Ann Neurol 2:116-119, 1977. 37. DiBenedetto M: Sensory nerve conduction in lower extremities. Arch Phys Med Rehabil 51:253-258, 1970. 38. DiBenedetto M: Evoked sensory potentials in peripheral neuropathy. Arch Phys Med Rehabil 53:126-131, 1972. 39. DiBenedetto M, Mitz M, Klingbeil G, Davidoff DD: New criteria for sensory nerve conduction especially useful in diagnosing carpal tunnel sydrome. Arch Phys Med Rehabil 67:586-589, 1986. 40. Dillingham TR, Spellman NT, Chang AS: Trigeminal motor nerve conduction: Deep temporal and mylohyoid nerves. Muscle Nerve 19:277-284, 1996. 41. Downie AW, Scott TR: Radial nerve conduction studies. Neurology (Minneap) 14:839-843, 1964. 42. Downie AW, Scott TR: An improved technique for radial nerve conduction studies. J Neurol Neurosurg Psychiatry 30:322-336, 1967. 43. Dumitru D, Nelson MR: Posterior femoral cu-
173
44.
45.
46. 47.
48.
49. 50.
51.
52. 53.
54.
55. 56.
57.
58.
taneous nerve conduction. Arch Phys Med Rehabil 71:979-982, 1990. Dyck PJ, Lambert EH, Nichols PC: Quantitative measurement of sensation related to compound action potential and number and sizes of myelinated and unmyelinated fibers of sural nerve in health, Friedreich's ataxia, hereditary sensory neuropathy, and tabes dorsalis. In Remond A (ed): Handbook of Electroencephalography and Clinical Neurophysiology, Vol 9. Elsevier, Amsterdam, 1972, pp 83-118. Eklund G: A new electrodiagnostic procedure for measuring sensory nerve conduction across the carpal tunnel. Uppsala J Med Sci 80:63-64, 1975. Ertekin C: Saphenous nerve conduction in man. J Neurol Neurosurg Psychiatry 32:530540, 1969. Ertekin C, Nejat RS, Sirin H, Selcuki D, Arac N, Ertas M: Comparison of magnetic coil and needle-electrical stimulation in diagnosis of lumbosacral radiculopathy. Muscle Nerve 17: 685-686, 1994. Ertekin C, Nejat RS, Sirin H, Selcuki D, Arac N, Ertas M, Colakoglu Z: Comparison of magnetic coil stimulation and needle electrical stimulation in diagnosis of lumbosacral radiculopathy. Clin Neurol Neurosurg 96:124-129, 1994. Evans BA, Daube JR: Comparison of three electrodiagnostic methods of diagnosing carpal tunnel syndrome. Muscle Nerve 7:565, 1984. Falck B, Hurme M: Conduction velocity of the posterior interosseus nerve across the arcade of Frohse. Electromyogr Clin Neurophysiol 23:567-576, 1983. Falck B, Hurme M, Hakkarainen S, Aarnio P: Sensory conduction velocity of plantar digital nerves in Morton's metatarsalgia. Neurology 34:698-701, 1984. Feibel A, Foca FJ: Sensory conduction of radial nerve. Arch Phys Med Rehabil 55:314-316, 1974. Felsenthal G, Brockman P, Mondell D, Hilton E: Proximal forearm ulnar nerve conduction techniques. Arch Phys Med Rehabil 67:440444, 1986. Felsenthal G, Freed MJ, Kalafut R, Hilton EB: Across-elbow ulnar nerve sensory conduction technique. Arch Phys Med Rehabil 70:668-672, 1989. Felsenthal G, Butler DH, Shear MS: Acrosstarsal-tunnel motor-nerve conduction technique. Arch Phys Med Rehabil 73:64-69, 1992. Ferrante MA, Wilboum AJ: The utility of various sensory nerve conduction responses in assessing brachial plexopathies. Muscle Nerve 18:879-889, 1995. Fitz WR, Mysiw WJ, Johnson EW: First lumbrical latency and amplitude: Control values and findings in carpal tunnel syndrome. Am J Phys Med Rehabil 69:198-201, 1990. Fu R, DeLisa JA, Kraft GH: Motor nerve latencies through the tarsal tunnel in normal adult subjects: Standard determinations corrected for temperature and distance. Arch Phys Med Rehabil 61:243-248, 1980.
174 59. Gassel MM: A study of femoral nerve conduction time. Arch Neurol 9:57-64, 1963. 60. Gassel MM: A test of nerve conduction to muscles of the shoulder girdle as an aid in the diagnosis of proximal neurogenic and muscular disease. J Neurol Neurosurg Psychiatry 27: 200-205, 1964. 61. Gassel MM: Sources of error in motor nerve conduction studies. Neurology (Minneap) 14: 825-835, 1964. 62. Gilliatt RW, Goodman HV, Willison RG: The recording of lateral popliteal nerve action potentials in man. J Neurol Neurosurg Psychiatry 24:305-318, 1961. 63. Ginzburg M, Lee M, Ginzburg J, Alba A: Median and ulnar nerve conduction determinations in the Erb's point-axilla segment in normal subjects. J Neurol Neurosurg Psychiatry 41:444-448, 1978. 64. Green RF, Brien M: Accessory nerve latency to the middle and lower trapezius. Arch Phys Med Rehabil 66:23-24, 1985. 65. Guiloff RJ, Sherratt RM: Sensory conduction in medial plantar nerve. Normal values, clinical applications, and a comparison with the sural and upper limb sensory nerve action potentials in peripheral neuropathy. J Neurol Neurosurg Psychiatry 40:1168-1181, 1977. 66. Hansson S: Segmental median nerve conduction measurements discriminate carpal tunnel syndrome from diabetic polyneuropathy. Muscle Nerve 18:445-453, 1995. 67. Harding C, Halar E: Motor and sensory ulnar nerve conduction velocities: Effect of elbow position. Arch Phys Med Rehabil 64:227-232, 1983. 68. Hoffman MD, Mitz M, Luisi M, Melville BR: Paired study of dorsal ulnar cutaneous and superficial radial sensory nerves. Arch Phys Med Rehabil 69:591-594, 1988. 69. Horowitz SH, Krarup C: Conduction studies of the normal sural nerve. Muscle Nerve 15:374383, 1992. 70. Imaoka H, Yorifuji S, Takahashi M, Nakamura Y, Kitaguchi M, Tarui S: Improved inching method for the diagnosis and prognosis of carpal tunnel syndrome. Muscle Nerve 15:318324, 1992. 71. Inaba A, Yokota T, Komori T, Hirose K: Proximal and segmental motor nerve conduction in the sciatic nerve produced by percutaneous high voltage electrical stimulation. Electroencephologr Clin Neurophysiol 101:100-104, 1996. 72. Irnai KD, Grabois M, Harvey SC: Standardized technique for diagnosis of tarsal tunnel syndrome. Am J Phys Med 61:26-31, 1982. 73. Iyer KS, Kaplan E, Goodgold J: Sensory nerve action potentials of the medial and lateral plantar nerve. Arch Phys Med Rehabil 65:529-530, 1984. 74. Izzo KL, Aravabhumi S, Jafri A, Sobel E, Demopoulous JT: Medial and lateral antebrachial cutaneous nerves: Standardization of technique, reliability and age effect on healthy subjects. Arch Phys Med Rehabil 66:592-597, 1985.
Nerve Conduction Studies 75. Izzo KL, Sridhara CR, Lemont H, Rosenholtz H: Sensory conduction studies of the branches of the superficial peroneal nerve. Arch Phys Med Rehabil 62:24-27, 1981. 76. Jaaskelainen SK: A new technique for recording sensory conduction velocity of the inferior alveolar nerve. Muscle Nerve 22:455-459, 1999. 77. Jabre JF: Ulnar nerve lesions at the wrist: New technique for recording from the sensory dorsal branch of the ulnar nerve. Neurology (New York) 30:873-876, 1980. 78. Jabre JF: The superficial peroneal sensory nerve revisited. Arch Neurol 38:666-667, 1981. 79. Joynt RL: Differences in sensory conduction velocity between different sensory branches and segments of the median and ulnar nerves. Am J Phys Med Rehabil 68:210-214, 1989. 80. Kalantri A, Visser BD, Dumitru D, Grant AE: Axilla to elbow radial nerve conduction. Muscle Nerve 11:133-135, 1988. 81. Kanakamedala RV, Fritch WL, Hong C-Z: Conduction of the dorsal digital branches of the radial nerve to the long finger. Arch Phys Med Rehabil 72:576-578, 1991. 82. Kanakamedala RV, Hong C-Z: Peroneal nerve entrapment at the knee localized by short segment stimulation. Am J Phys Med Rehabil 68:116-122, 1989. 83. Kanakamedala RV, Simons DG, Porter RW, Zucker RS: Ulnar nerve entrapment at the elbow localized by short segment stimulation. Arch Phys Med Rehabil 69:959-963, 1988. 84. Kandel ER, Schwartz JH, JessellTM (eds): Principles of Neural Science, Elsevier, New York, 1991. 85. Kaplan PE: Electrodiagnostic confirmation of long thoracic nerve palsy. J Neurol Neurosurg Psychiatry 43:50-52, 1980. 86. Kaplan PE: A motor nerve conduction velocity across the upper trunk and the lateral cord of the brachial plexus. Electromyogr Clin Neurophysiol 22:315-320, 1982. 87. Kayed K, Rosjo O: Two-segment sural nerve conduction measurements in polyneuropathy. J Neurol Neurosurg Psychiatry 46:867-870, 1983. 88. Kiff ES, Swash M: Slowed conduction in the pudenda! nerves in idiopathic (neurogenic) fecal incontinence. Br J Surg 71:615-616, 1984. 89. Kim DJ, Kalantri A, Guha S, Wainapel SF: Dorsal cutaneous ulnar nerve conduction: Diagnostic aid in ulnar neuropathy. Arch Neurol 38:321-322, 1981. 90. Kim W, Kim HJ, Blumenthal FS, Joynt RL: Antidromic sensory nerve conduction studies of medial and lateral plantar nerves in normals. Electromyogr Clin Neurophysiol 33:289-294, 1993. 91. Kimura J: Collision technique—Physiological block of nerve impulses in studies of motor nerve conduction velocity. Neurology (Minneap) 26:680-682, 1976. 92. Kimura J: A method for determining median nerve conduction velocity across the carpal tunnel. J Neurol Sci 38:1-10, 1978. 93. Kimura J: The carpal tunnel syndrome. Localization of conduction abnormalities within the
Assessment of Individual Nerves distal segment of the median nerve. Brain 102:619-635, 1979. 94. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principle and Practice, FA Davis, Philadelphia, 1983. 95. Kimura J: Principles and pitfalls of nerve conduction studies. Ann Neurol 16:415-429, 1984. 96. Kimura J, Machida M, Ishida T, Yamada T, Rodnitzky R, Kudo Y, Suzuki S: Relationship between size of commpound sensory or muscle action potentials, and length of nerve segment. Neurology 36:647-652, 1986. 97. Kimura I, Seki H, Sasao S, Ayyar DR: The great auricular nerve conduction study: A technique, normative data and clinical usefulness. Electromyogr Clin Neurophysiol 27:39-43, 1987. 98. Kincaid JC, Phillips II LH, Daube JR: The evaluation of suspected ulnar neuropathy at the elbow. Arch Neurol 43:44-47, 1986. 99. Kothari MJ, Macintosh K, Heistand M, Logigian EL: Medial antebrachlal cutaneous sensory studies in the evaluation of neurogenic thoracic outlet syndrome. Muscle Nerve 21:647-649, 1998. 100. Kothari MJ, Preston DC: Comparison of the flexed and extended elbow positions in localizing ulnar neuropathy at the elbow. Muscle Nerve 18:336-340, 1995. 101. Kothari MJ, Preston DC, Logigian EL: Lumbrical-interosseous motor studies localize ulnar neuropathy at the wrist. Muscle Nerve 19:170174, 1996. 102. Kraft GH: Axillary, musculocutaneous and suprascapular nerve latency studies. Arch Phys Med Rehabil 53:383-387, 1972. 103. Krarup C, Horowitz SH, Dahl K: The influence of the stimulus on normal sural nerve conduction velocity: A study of the latency of activation. Muscle Nerve 15:813-821, 1992. 104. Lee HJ: Sensory abnormalities in the medial thigh: An electrodiagnostic report of 2 cases. Muscle Nerve 19:1058-1059, 1996. 105. Lee HJ, Bach JR, DeLisa JA: Deep peroneal sensory nerve: Standardization in nerve conduction study. Am J Phys Med Rehabil 69:202-204, 1990. 106. Lee HJ, Bach JR, DeLisa JA: Lateral dorsal cutaneous branch of the sural nerve: Standardization in nerve conduction study. Am J Phys Med Rehabil 71:318-320, 1992. 107. Lee HJ, Bach JR, DeLisa JA: Medial femoral cutaneous nerve conduction. Am J Phys Med Rehabil 74:305-307, 1995. 108. Lesser EA, Venkatesh S, Preston DC, Logigian EL: Stimulation distal to the lesion in patients with carpal tunnel syndrome. Muscle Nerve 18:503-507, 1995. 109. Liguori R, Cevoli S, Montagna P: Electroneurographic investigation of the mandibular nerve in lingual neuropathy. Muscle Nerve 21:410-412, 1998. 110. Lim CL, Lal H, Yiannikas C: The effect of wrist size on the orthodromic median sensory nerve action potential. Muscle Nerve 18:117-119, 1995. 111. Lum PB, Kanakamedala RV: Conduction of the palmar cutaneous branch of the median nerve. Arch Phys Med Rehabil 67:805-806, 1986.
175 112. Lysens R, Vandendriessche G, Van Mol Y, Rosselle N: The sensory conduction velocity in the cutaneous femoris lateralis nerve in normal adult subjects and in patients with complaints suggesting meralgia paresthetica. Electromyogr Clin Neurophysiol 21:505-510, 1981. 113. Ma DM, Kim SH, Spielholz N, Goodgold J: Sensory conduction study of distal radial nerve. Arch Phys Med Rehabil 62:562-564, 1981. 114. Maccabee PJ, Lipitz ME, Desudchit T, Golub RW, Nitti VW, Bania JP, Wilier JA, Cracco RQ, Cadwell J, Hotson GC, Eberle LP, Amassian VE: A new method using neuromagnetic stimulation to measure conduction time within the cauda equina. Electroencephologr Clin Neurophysiol 101:153-166, 1996. 115. MacDonell RAL, Cros D, Shahani BT: Lumbosacral nerve root stimulation comparing electrical with surface magnetic coil techniques. Muscle Nerve 15:885-890, 1992. 116. Mackenzie K, DeLisa J: Distal sensory latency measurement of the superficial radial nerve in normal adult subjects. Arch Phys Med Rehabil 62:31-34, 1981. 117. MacLean 1C: Nerve root stimulation to evaluate conduction across the brachial and lumbosacral plexuses. Third Annual Continuing Education Course, American Association of Electromyography and Electrodiagnosis, September 25, 1980, Philadelphia, Pennsylvania. 118. MacLean IC, Mattioni TA: Phrenic nerve conduction studies: A new technique and its application in quadriplegic patients. Arch Phys Med Rehabil 62:70-73, 1981. 119. MacLean IC, Taylor RS: Nerve root stimulation to evaluate brachial plexus conduction. Abstracts of Communication of the Fifth International Congress of Electromyography, Rochester, Minnesota, 1975. 120. Markand ON, Kincaid JC, Pourmand RA, Moorthy SS, King RD, Mahomed Y, Brown JW: Electrophysiologic evaluation of diaphragm by transcutaneous phrenic nerve stimulation. Neurology 34:604-614, 1984. 121. Masuda T, Miyano H, Sadoyama T: The distribution of myoneural junctions in the biceps brachii investigated by surface electromyography. Electroencephologr Clin Neurophysiol 56: 597-603, 1983. 122. Masuda T, Miyano H, Sadoyama T: A surface electrode array for detecting action potential trains of single motor units. Electroencephalogr Clin Neurophysiol 60:435-443, 1985. 123. Masuda T, Sadoyama T: The propagation of single motor unit action potentials detected by a surface electrode array. Electroencephalogr Clin Neurophysiol 63:590-598, 1986. 124. Mavor H, Atcheson JB: Posterior tibial nerve conduction. Velocity of sensory and motor fibers. Arch Neurol 14:661-669, 1966. 125. McComas AJ, Keresh IS, Manzano G: Multiple innervation of human muscle fibers. J Neurol Sci 64:55-64, 1984. 126. Menkes DL, Hood DC, Ballesteros RA, Williams DA: Root stimulation improves the detection of acquired demyelinating polyneuropathies. Muscle Nerve 21:298-308, 1998.
Nerve Conduction Studies
176 127. Mysiw WJ, Colachis SC: Electrophysiologic study of the anterior interosseous nerve. Am J Phys Med Rehabil 67:50-54, 1988. 128. Nathan PA, Meadows KD, Doyle LS: Sensory segmental latency values of the median nerve for a population of normal individuals. Arch Phys Med Rehabil 69:499-501, 1988. 129. Nelson RM, Currier DP: Motor-nerve conduction velocity of the musculocutaneous nerve. Phys Ther 49:586-590, 1969. 130. Newsom Davis J: Phrenic nerve conduction in man. J Neurol Neurosurg Psychiatry 30:420426, 1967. 131. Oh SJ, Kim HS, Ahmad BK: Electrophysiological diagnosis of interdigital neuropathy of the foot. Muscle Nerve 7:218-225, 1984. 132. Olney RK, Wilbourn AJ: Ulnar nerve conduction study of the first dorsal interosseous muscle. Arch Phys Med Rehabil 66:16-18, 1985. 133. Palliyath SK: A technique for studying the greater auricular nerve conduction velocity. Muscle Nerve 7:232-234, 1984. 134. Park TA, del Toro DR: The medial calcaneal nerve: anatomy and nerve conduction technique. Muscle Nerve 18:32-38, 1995. 135. Peake JB, Roth JL, Schuchmann GF: Pneumothorax: a complication of nerve conduction studies using needle stimulation. Arch Phys Med Rehabil 63:187-188, 1982. 136. Pease WS, Cannell CD, Johnson EW: Median to radial latency difference test in mild carpal tunnel syndrome. Muscle Nerve 12:905a-909, 1989. 137. Pease WS, Cunningham ML, Walsh WE, Johnson EW: Determining neurapraxia in carpal tunnel syndrome. Am J Phys Med Rehabil 66:117-119, 1988. 138. Peterson AR, Giuliani MJ, McHugh M, Shipe CC: Variations in dorsomedial hand innervation, electrodiagnostic implications. Arch Neurol 49:870-873, 1992. 139. Petrera JE, Trojaborg W: Conduction studies along the accessory nerve and follow-up of patients with trapezius palsy. J Neurol Neurosurg Psychiatry 47:630-636, 1984. 140. Petrera JE, Trojaborg W: Conduction studies of the long thoracic nerve in serratus anterior palsy of different etiology. Neurology (Cleveland) 34:1033-1037, 1984. 141. Ponsfor SN: Sensory conduction in medial and lateral plantar nerves. J Neurol Neurosurg Psychiatry 51:188-191, 1988. 142. Pradhan S, Taly A: Intercostal nerve conduction study in man. J Neurol Neurosurg Psychiatry 52:763-766, 1989. 143. Preston DC, Logigian EL: Lumbrical and interossei recording in carpal tunnel syndrome. Muscle Nerve 15:1253-1257, 1992. 144. Reddy MP: Conduction studies of the medial cutaneous nerve of the forearm. Arch Phys Med Rehabil 64:209-211, 1983. 145. Redford JWB: Conduction time in motor fibers of nerves which innervate proximal muscles of the extremities in normal persons and in patients with neuromuscular diseases. Thesis, University of Minnesota, Minneapolis, 1958. 146. Redmond MD, Di Benedetto M: Hypoglossal
147. 148. 149. 150.
151.
152.
153.
154.
155.
156. 157. 158.
159.
160. 161. 162.
163. 164.
nerve conduction in normal subjects. Muscle Nerve 11:447-452, 1988. Rosenberg JN: Anterior interosseous/median nerve latency ratio. Arch Phys Med Rehabil 71:228-230, 1990. Ross MA, Kimura J: AAEM case report #2: the carpal tunnel syndrome. Muscle Nerve 18:567573, 1995. Roth G: Vitesse de conduction motrice du nerf median dans le canal carpien. Ann Med Phys 13:117-132, 1970. Rutkove SB, Kothari MJ, Raynor EM, Levy M, Fadic R, Nardin RA: Sural/Radial amplitude ratio in the diagnosis of mild axonal polyneuropathy. Muscle Nerve 20:1236-1241, 1997. Rutkove SB, Kothari MJ, Sampson C, Preston DC: The effect of wrist position on the conduction velocity of the ulnar nerve. Muscle Nerve 19:657-658, 1996. Saeed MA, Gatens PF: Compound nerve action potentials of the medial and lateral plantar nerves through the tarsal tunnel. Arch Phys Med Rehabil 63:304-307, 1982. Sander HW, Menkes DL, Triggs WJ, Chokroverty S: Cervical root stimulation at C5/6 excites C8/T1 roots and minimizes pneumothorax risk. Muscle Nerve 22:766-768, 1999. Schultze-Mosgau S, Reich RH: Assessment of inferior alveolar and lingual nerve disturbances after dento-alveolar surgery, and of recovery of sensitivity. Int J Oral Maxillofac Implants 22:214-217, 1993. Senden R, Van Mulders J, Ghys R, Rosselle N: Conduction velocity of the distal segment of the saphenous nerve in normal adult subjects. Electromyogr Clin Neurophysiol 21:3-10, 1981. Seror P: Orthodromic inching test in mild carpal tunnel syndrome. Muscle Nerve 21: 1206-1208, 1998. Shahani B, Goodgold J, Spielholz NI: Sensory nerve action potentials in the radial nerve. Arch Phys Med Rehabil 48:602-605, 1967. Shahani BT, Young RR, Potts F, Maccabee P: Terminal latency index and late response studies in motor neuron disease, peripheral neuropathies and entrapment syndromes. Acta Neurol Scand 60(suppl 73): 118, 1979. Sheean GL, Houser MK, Murray NMF: Lumbrical-interosseous latency comparison in the diagnosis of carpal tunnel syndrome. Electroencephalogr Clin Neurophysiol 97:285-289, 1995. Sheean GL, Kanabar G, Murray NMF: Lumbrical-interosseous comparison in a distal ulnar nerve lesion. Muscle Nerve 19:673-674, 1996. Shiozawa R, Mavor H: In vivo human sural nerve action potentials. J Appl Physiol 26:623629, 1969. Shirali CS, Sandier B: Radial nerve sensory conduction velocity masurement by antidromic technique. Arch Phys Med Rehabil 53:457460, 1972. Simovic D, Weinberg DH: Terminal latency index in the carpal tunnel syndrome. Muscle Nerve 20:1178-1180, 1997. Simovic D, Weinberg DH: The median nerve
Assessment of Individual Nerves
165. 166. 167. 168. 169. 170.
171. 172. 173.
174.
175.
176.
terminal latency index in carpal tunnel syndrome: A clinical case selection study. Muscle Nerve 22:573-577, 1999. Spevak MK, Prevec TS: A noninvasive method of neurography in meralgia paraesthetica. Muscle Nerve 18:601-605, 1995. Spindler HA, Felsenthal G: Sensory conduction in the musculocutaneous nerve. Arch Phys Med Rehabil 59:20-23, 1978. Spindler HA, Felsenthal G: Radial sensory conduction in the hand. Arch Phys Med Rehabil 67:821-823, 1986. Stevens JC: AAEE Minimonograph #26: The electrodiagnosis of carpal tunnel syndrome. Muscle Nerve 2:99-113, 1987. Stohr M, Schumm F, Ballier R: Normal sensory conduction in the saphenous nerve in man. Electroenceph Clin Neurophysiol 44:172-178, 1978. Swash M, Snooks SJ: Motor nerve conduction studies of the pelvic floor innervation. In Henry MM, Swash M (eds): Coloproctology and the Pelvic Floor, ed 2. Butterworth-Heinemann, Oxford, 1992, pp 196-206. Swenson MR, Rubenstein RS: Phrenic nerve conduction studies. Muscle Nerve 15:597-603, 1992. Trojaborg W: Motor and sensory conduction in the musculocutaneous nerve. J Neurol Neurosurg Psychiatry 39:890-899, 1976. Trojaborg W, Grewal RP, Sheriff P: Value of latency measurements to the small palm muscles compared to other conduction parameters in the carpal tunnel syndrome. Muscle Nerve 19:243-245, 1996. Trojaborg WT, Moon A, Andersen BB, Trojaborg NS: Sural nerve conduction parameters in normal subjects related to age, gender, temperature, and height: A reappraisal. Muscle Nerve 15:666-671, 1992. Trojaborg W, Sindrup EH: Motor and sensory conduction in different segments of the radial nerve in normal subjects. J Neurol Neurosurg Psychiatry 32:354-359, 1969. Truong XT, Russo FI, Vagi I, Rippel DV: Con-
177
177.
178.
179.
180. 181.
182.
183. 184. 185. 186. 187.
188.
duction velocity in the proximal sural nerve. Arch Phys Med Rehabil 60:304-308, 1979. Uchida Y, Sugioka Y: The value of electrophysiological examination of the flexor carpi ulnaris muscle in the diagnosis of cubital tunnel syndrome. Electromyogr Clin Neurophysiol 33: 369-373, 1993. Uncini A, Lange DJ, Solomon M, Soliven B, Meer J, Lovelace RE: Ring finger testing in carpal tunnel syndrome: A comparative study of diagnostic utility. Muscle Nerve 12:735-741, 1989. Venkatesh S, Kothari MJ, Preston DC: The limitations of the dorsal ulnar cutaneous sensory response in patients with ulnar neuropathy at the elbow. Muscle Nerve 18:345-347, 1995. Wainapel SF, Kim DJ, aEbel A: Conduction studies of the saphenous nerve in healthy subjects. Arch Phys Med Rehabil 59:316-319, 1978. Wee AS, Ashley RA: Radial sensory innervation to index and middle fingers: Electrophysiologic considerations. Electromyogr Clin Neurophysiol 29:13-15, 1989. Werschkul JD: Anomalous course of the recurrent motor branch of the median nerve in a patient with carpal tunnel syndrome. Case report. J Neurosurg 47:113-114, 1977. Wexner SD, Marchetti F, Salanga VA, et al: Neurophysiologic assessment of the anal sphincters. Dis Colon Rectum 34:606-612, 1991. Wilbourn AJ: Sensory nerve conduction studies. J Clin Neurophysiol 11:584-601, 1994. Yang CC, Bradley WE: Innvervation of the human anterior urethra by the dorsal nerve of the penis. Muscle Nerve 21:514-518, 1998. Yang CC, Bradley WE, Berger RE: The effect of pharmacologic erection on the dorsal nerve of the penis. Muscle Nerve 20:1439-1444, 1997. Yates SK, Yaworski R, Brown WF: Relative preservation of lumbrical versus thenar motor fibres in neurogenic disorders. J Neurol Neurosurg Psychiatry 44:768-774, 1981. Young AW, Redmond MD, Hemler DE, Belandres PV: Radial motor nerve conduction studies. Arch Phys Med Rehabil 71:399-402, 1990.
Chapter
7
FACTS, FALLACIES, AND FANCIES OF NERVE STIMULATION TECHNIQUES
1. INTRODUCTION 2. COMMON TECHNICAL ERRORS
Stimulating System Recording System 3. SPREAD OF STIMULATION CURRENT Stimulation of the Facial Nerve Axillary Stimulation and Collision Technique Palmar Stimulation of the Median and Ulnar Nerves
4. ANOMALIES AS SOURCES OF ERROR Martin-Gruber Anastomosis Anomalies of the Hand Accessory Deep Peroneal Nerve Anomalous Communication Between Peroneal and Tibial Nerve 5. PRINCIPLES AND PITFALLS OF WAVEFORM ANALYSIS Physiologic and Pathologic Temporal Dispersion Detection of Conduction Block Distribution of Conduction Velocities Collision Technique to Block Fast- or Slow-Conducting Fibers 6. STUDIES OVER SHORT AND LONG DISTANCES Segmental Stimulation in Short Increments Late Responses for Evaluation of Long Pathways Reproducibility of Various Measures Clinical Considerations
1 INTRODUCTION Nerve conduction studies help delineate the extent and distribution of the neural lesion and distinguish two major categories of peripheral nerve disease: demyelination and axonal degeneration. 178
With steady improvement and standardization of methods,42,89 such studies have become a reliable means of testing not only for precise localization of a lesion but also for accurate characterization of peripheral nerve function.25,73 This chapter reviews the fundamental principles and changing concepts of nerve stimulation
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques techniques and their proper application in the differential diagnosis of peripheral nerve disorders. Despite simple basic principles that dictate the method in theory, pitfalls abound in practice.74,77,78-104 Commonly encountered sources of error that are often overlooked include intermittent failure in the stimulating or recording system, excessive spread of stimulation current, anomalous innervation, temporal dispersion, and inaccuracy of surface measurement. Stationary far-field peaks may result from a moving source not only in a referential but also a bipolar derivation, usually selected for recording near-field potentials (see Chapter 20-3). Lack of awareness of these possibilities can cause confusion in the interpretation of the results. All these factors limit the reproducibility of conduction studies, making it imperative to maintain good quality control.87,168 Conventional studies deal primarily with evaluation of the fastest conducting fibers, based on the latency measured to the onset of the evoked potential. In some clinical entities, special techniques may help in evaluating other aspects, such as conduction velocity of the slower fibers and the time course of the absolute and relative refractory periods. The phenomenon of collision provides a useful means of assessing these features of nerve conduction.53,56,70,71,82 Here, a second stimulus delivered distally to the nerve blocks the unwanted impulses not under study. Other areas of interest include studies of threshold electrotonus and motor unit number estimates, described in Chapter 8. Although these methods supplement the conventional technique in theory, their clinical application in practice awaits further clarification. 2
COMMON TECHNICAL ERRORS
Technical problems often account for unexpected observations during routine nerve conduction studies. The failure to appreciate this possibility will lead to an incorrect diagnosis, especially if the findings mimic expected manifestations of the disease under consideration. Commonly
179
unidentified yet easily correctable problems include malfunction of the stimulating electrodes or the recording system.
Stimulating System Absent or unusually small responses result from inappropriately low shock intensity or from a stimulus that is misdirected despite adequate current strength. The amplitude should improve with relocation of the stimulating electrode, pressing it firmly closer to the nerve, and, if necessary, increasing shock intensity or duration. The use of monopolar or concentric needles may help, especially in obese patients. Profuse perspiration or an excessive amount of cream over the skin surface may shunt the cathode and anode, rendering the otherwise sufficient stimulating current ineffective. Inadvertent reversal of the anode and cathode would, in theory, block the propagating impulse; however, in the usual clinical setup with a 2-3 cm separation, anodal hyperpolarization abates too quickly to render any detectable effects. More important, misidentifying the cathodal and anodal positions would invalidate the relationship between the measured distance and latency. Regardless of the responsible technical fault, submaximal activation of the nerve proximally may erroneously suggest a conduction block, especially if a distal stimulus elicits a full response. In some neuropathic states with an abnormally elevated threshold, an ordinarily sufficient intensity may fail to excite the nerve at the site of pathology, necessitating more proximal stimulation to confirm propagation of impulses across the lesion (see this chapter, part 5).
Recording System Even optimal stimulation elicits a small response if a faulty connection hampers recording. Common problems include inappropriate placement of the pick-up electrodes; breaks in the electrode wires; use of a disconnected preamplifier; loss of power supply; and incorrect oscilloscope settings for sensitivity, sweep, or filters. A
Nerve Conduction Studies
180
broken recording electrode may escape detection because it shows no change in appearance if the insulating sheath remains intact. With partial damage to the wire, stimulus-induced muscle twitches cause movement-related potentials, which can mimic a compound muscle action potential. A quick check of the recording system should be the first step: ask the patient to contract the muscle with the electrode in position and the amplifiers turned on. Deficiencies at any step of the recording circuit would prevent a normal display of muscle action potentials on the oscilloscope. An initial positivity preceding the major negative peak of the compound muscle action potential usually results from incorrect positioning of the active electrode away from the end-plate region. Alternatively, it may represent a volume-conducted potential from distant muscles, activated by anomalous innervation or by spread of stimulation to other nerves. The compound muscle action potential reverses its polarity with an inadvertent switch of the active (G1) and reference (G2) electrodes. Similarly, any deviation from the standard belly (G1) and tendon (G2) placement of recording electrodes distorts the waveform. 3
SPREAD OF STIMULATION CURRENT
With an inappropriately high shock intensity, stimulating current can spread to a nerve or muscle not being tested. Failure to recognize this possibility may result in false determination of latencies to the onset of a volume-conducted potential from unintended muscles. Under these circumstances, visual inspection of the contracting muscle, rather than the waveform on the oscilloscope, will identify the generator source. In some such cases, the collision technique (see this chapter, part 5) can, in effect, activate the intended nerve selectively by blocking the unwanted nerve.70 The use of needle pick-up also restricts the recording to limited target areas for such special purposes as studying innervation of individual motor branches, patterns of anomaly, and function of atrophic muscles that may escape surface detection. This
type of recording, by design, fails to provide the most important information on the total size of muscle response.
Stimulation of the Facial Nerve The facial nerve becomes accessible to surface or needle stimulation as it exits from the stylomastoid foramen (see Chapter 17-2) (see Figs. 17-2 and 17-3). The distal segment, tested by stimulating the nerve here and recording compound muscle action potentials from various facial muscles, remains normal for a few days after complete separation of the nerve at a proximal site. The loss of distal excitability by the end of the first week coincides with the onset of nerve degeneration, which generally implies poor prognosis. With shocks of very high intensity, stimulating current may also activate the motor point of the masseter muscle. A volume-conducted potential then erroneously suggests a favorable prognosis, when in fact the facial nerve has already degenerated (Fig. 7-1). As stated before, visual inspection would verify that the contraction involved the masseter, not the facial, muscle. Surface stimulation of the facial nerve may also activate cutaneous fibers of the trigeminal nerve, causing reflexive contraction of the orbicularis oculi (see Chapter 17-2). The reflex response may mimic a late component of the compound muscle action potential or recurrent response from antidromic activation of motor neurons.
Axillary Stimulation and Collision Technique With the use of ordinary shock intensity, stimulation of the median or ulnar nerve activates only the nerve in question at the wrist or elbow, but not at the axilla, where the two nerves lie in close proximity.70 If the current intended for the median nerve spreads to the ulnar nerve, the electrodes placed on the thenar eminence register not only median but also ulnar innervated muscle potential. The measured latency will then indicate normal ulnar conduction if the median nerve conducts more slowly, as in carpal tunnel syndrome (Fig. 7-2). In
Figure 7-1. Compound muscle action potential from the orbicularis oculi after stimulation of the facial nerve in a patient with traumatic facial diplegia. A. Left side. B. Right side. Shocks of ordinary intensity (top three tracings) elicited no response but with a much higher intensity, a definite muscle response appeared (bottom three tracings). Close observation of the face revealed contraction of the masseter rather than the orbicularis oculi.
Figure 7-2. A 39-year-old man with carpal tunnel syndrome. The stimulation of the median nerve at the wrist (S1) or elbow (S2) elicited a muscle action potential with increased latency in the thenar eminence. Spread of axillary stimulation (S3) to the ulnar nerve (third tracing from top) activated ulnar-innervated thenar muscles with shorter latency. Another stimulus (S4) applied to the ulnar nerve at the wrist (bottom tracing:) blocked the proximal impulses by collision. The muscle action potential elicited by S4 occurred much earlier. The diagram on the left shows collision between the orthodromic (solid arrows) and antidromic (open arrows) impulses. [From Kimura,70 with permission.]
181
182
the same case, a stimulus at the elbow activates only the median nerve, revealing a prolonged latency. The calculated conduction time between the axilla and elbow would then suggest an erroneously fast conduction velocity. In extreme cases, the latency of the median response after stimulation at the elbow exceeds that of the ulnar component elicited with shocks at the axilla. The reverse discrepancy can occur in a study of tardy ulnar palsy, with spread of axillary stimulation to the median nerve. In this case, the surface electrodes on the hypothenar eminence register the volumeconducted response from thenar muscles or lumbricals as a small positive potential of 1-5 mV in amplitude and 10-20 ms in duration. This positivity, though usually buried in a much larger ulnar response occurring simultaneously, becomes obvious if the ulnar nerve conducts slower than the median nerve as in tardy ulnar
Nerve Conduction Studies palsy (Fig. 7-3). The earlier median component from thenar muscles then obscures the onset of the ulnar response originating from hypothenar muscles. The short latency measured to the onset of the median component fails to correctly reflect a delayed ulnar response. A stimulus at the elbow in the same case activates only the ulnar nerve with a prolonged latency, leading to misculculation of an erroneously fast conduction velocity from axilla to elbow. A physiologic nerve block with collision allows selective recording of the median or ulnar component despite coactivation of both nerves proximally.70 In studies of the median nerve, for example, a distal stimulus delivered to the ulnar nerve at the wrist generates the antidromic impulse, which collides with the orthodromic ulnar impulse from the axilla. Thus, only the median impulse reaches the muscle (Fig. 7-2). The ulnar response induced by
Figure 7-3. A 29-year-old man with tardy ulnar palsy. Stimulation at the wrist (S1) or elbow (S2) selectively activated the ulnar nerve giving rise to an abnormally delayed muscle action potential over the hypothenar eminence. Spread of axillary stimulation (S3) to the median nerve (third tracing from top) elicited an additional short latency median response with initial positivity. This potential, registered through volume conduction, obscurred the onset (arrowhead) of the muscle response under study. Another stimulus (S4) applied to the median nerve at the wrist (bottom tracing) blocked the proximal impulses by collision. The positive median potential elicited by S4 clearly preceded the ulnar component under study. [From Kimura,70 with permission.]
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques the distal stimulus precedes the median response under study, usually without obscuring it. If necessary, delivering the distal stimulus a few milliseconds before the proximal stimulation accomplishes a greater separation. This time interval should not exceed the conduction time between the distal and proximal points of stimulation, lest the antidromic impulse from the wrist pass the stimulus site at the axilla without collision. The same principles apply for the use of a distal stimulus to block the median nerve in selective recording of the ulnar response after coactivation of both nerves at the axilla (Fig. 7-3). The collision technique can clarify otherwise confusing results of motor nerve conduction studies in patients with carpal tunnel syndrome or tardy ulnar palsy. In each of the illustrated cases (see Figs. 7-2 and 7-3), spread of the stimulus caused obvious distortion in waveform of the proximally evoked potential. Less apparent discrepancies escape detection unless the collision methods block unwanted nerve impulses, uncovering the true response from the intended muscle. The collision technique provides a simpler, noninvasive means than the procaine nerve block previously employed to identify the origin of the recorded muscle potentials. Optimal studies of motor nerve conduction depend on either selective activation of the nerve in question or isolated recording from the target muscle. The collison method improves latency and waveform determination even under circumstances that preclude selective stimulation of the nerve at a proximal point. As an alternative method, the use of needle electrodes renders reliable latencies even after coactivation of more than one nerve, but its restricted recording area precludes assessment of the size of the compound muscle action potential.
Palmar Stimulation of the Median and Ulnar Nerves Palmar stimulation provides a unique contribution in evaluating the distal segment of the median nerve, although studies of the motor conduction in this region pose
183
some technical problems. 10,24,72,126,165 With serial stimulation in 1 cm increments from palm to wrist, the sensory latency increases linearly (see Fig. 6-7B). The motor study, when recorded from the thenar eminence, sometimes shows unexpected latency changes reflecting the recurrent course of the motor fibers. For example, a stimulus directed to the branching point of the thenar nerve in the palm could accidentally activate a terminal portion near the motor point. If another stimulus, delivered 1 cm proximally, excites only the median nerve trunk, the latency difference between the two stimulus points becomes unreasonably large, erroneously suggesting a focal slowing (Fig. 7-4A). Thus, a disproportionate latency change indicates a localized pathology only if serial stimulation shows a linear latency increase in the segment proximal and distal to the presumed site of lesion (Fig. 7-4B). Placing the pick-up leads over the second lumbrical, in lieu of the abductor pollicis brevis, circumvents this problem because tracking the terminal branch innervating this muscle poses no technical difficulty (Fig. 7-5 A and B). As an additional advantage, the same pair of electrodes may also be used to register muscle action potentials from the first volar interosseous muscle for inching study of the ulnar nerve along the course of the palmar branch and across the wrist (Fig. 7-5 C and D). The pattern of muscle twitch, rather than recorded waveforms, should be used to confirm selective activation of the intended nerve; for example, thumb abduction with stimulation of the recurrent thenar branch of the median nerve, and thumb adduction with stimulation of the deep palmar branch of the ulnar nerve. The same sort of error occurs in the calculation of motor latency over the wrist-topalm segment unless palmar stimulation activates the median nerve precisely at the origin of the thenar nerve as intended. Incremental stimulation from the wrist toward the digit with the cathode placed distally to the anode can activate the thenar nerve at the anodal point (acting as a floating cathode), even when the actual cathode lies clearly distal to the origin of the nerve.
184
Nerve Conduction Studies
Site of Stimulation
Figure 7-4. A. Compound muscle action potentials in a normal subject recorded after stimulation of the median nerve at multiple points across the wrist. On the initial trial (left), the latency decreased with the cathode inching proximally from -4 to -2, indicating inadvertent spread of stimulating current to a distal portion of the thenar nerve. An apparent steep latency change from -2 to -1 gave an erroneous impression of a focal slowing at this level. A more careful placement of the cathode (righfi eliminated unintended activation of the thenar nerve. The zero level at the distal crease of the wrist corresponds to the origin of the transverse ligament (cf. Figure 6-3). B. Sensory nerve (top) and muscle action potentials (bottom) in a symptomatic hand with the carpal tunnel syndrome. Serial stimulation showed a linear motor latency increase from —4 to -2 and from -1 to 5 with a localized slowing between -2 and — 1. A temporally dispersed, double-peaked sensory nerve potential indicates the point of localized conduction delay from -4 to -3 (cf. Figure 6-7). [From Kimura,72 with permission.]
A surface distance measured to the cathodal point would then overestimate the nerve length, thereby making the calculated conduction velocity erroneously fast. Proceeding from the distal palm toward the wrist with reversal of the electrode position, that is, cathode proximally to the anode, circumvents this problem. In this approach, palmar stimulation initially fails to produce a twitch, then causes thumb adduction, activating the deep branch of the ulnar nerve
and, about one cm more proximally, thumb abduction, signaling the arrival of the cathode just over the origin of the thenar nerve. In most subjects, this point lies 3 to 4 cm from the distal crease of the wrist, near the edge of the transverse carpal ligament.64 The use of needle stimulation renders more precise increments with less shock intensity, facilitating the process, especially when testing the palm, with its thick skin surface.
Figure 7-5. An inching study of median (A and B) and ulnar nerve (C and D, overleaf) across the wrist in 1 cm increments at eight sites of stimulation along the course of the nerve. The zero level at the distal crease of the wrist corresponds to the origin of the transverse carpal ligament and Guyon's canal. The photographs show a recording arrangement for muscle action potentials from the second lumbrical after stimulation of the median nerve (A), and the first volar interosseous after stimulation of the ulnar nerve (C). The latency increases linearly with stepwise shifts of stimulus site proximally in 1 cm increments for both median (B) and ulnar study (D)
185
Figure 7-5—Continued. An inching study of the ulnar nerve (C and D) across the wrist in 1 cm increments at eight sites of stimulation.
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques
4 ANOMALIES AS SOURCES OF ERROR Martin-Gruber Anastomosis The anatomic studies of Martin103 and 45 Gruber demonstrated frequent communication from the median to the ulnar nerve at the level of the forearm a few centimeters 157 distal to the medial humeral epicondyle. This anastomosis, often originating from the anterior interosseous nerve, predominantly consists of motor axons with rare sensory contribution, which19,160 may follow a different distribution. The communicating branch usually, though not always,93 supplies ordinarily ulnar-innervated intrinsic hand muscles, most notably the first dorsal interosseous, adductor pollicis, and abductor digiti minimi.100,143,166 The number of axons taking the anomalous course varies widely.2 A properly adjusted electrical stimulus delivered at the elbow may activate the anomalous fibers, maximally and
Figure 7-6. A 46 year old woman with the carpal tunnel syndrome and the Martin-Gruber anomaly. Stimulation at the elbow (S2) activated not only the median nerve but also communicating fibers, giving rise to a complex compound muscle action potential. With proper adjustment of electrode position and shock intensity, another stimulus at the elbow (S2) excited the median nerve selectively without activating the anastomosis. Another stimulus (S3) applied to the ulnar nerve at the wrist (bottom tracing) achieved the same effect by blocking the unwanted impulse transmitted through the communicating fibers. [From Kimura,76 with permission.]
187
selectively, without exciting the median nerve proper or vice versa (Fig. 7-6).82 This observation suggests a grouping of the nerve fibers forming the anastomosis in a separate bundle, rather than being scattered within the median nerve. The anomaly occurs, often bilaterally, in 15-32 percent of subjects in an unselected population.2,82 The higher incidence of this anomaly reported among congenitally abnormal fetuses in general and those with trisomy 21 in particular indicates its phylogenetic origin.143 The communicating fibers rarely cross from the ulnar to the median nerve in the forearm,44,113 occasionally involving only the sensory axons.55 Other anomalies associated with Martin-Gruber anastomosis include innervation of the ulnar aspect of the dorsum of the hand by the superficial radial sensory nerve.105 Careful analysis of the compound muscle action potentials readily reveals the presence of a Martin-Gruber anomaly during routine nerve conduction studies. This anastomosis, in effect, represents a
188
small bundle of the ulnar nerve, which accompanies the median nerve as it descends from the axilla to the elbow before separating from it in the forearm to join the ulnar nerve proper above the wrist. Thus, stimulation of the median nerve at the elbow excites the small bundle of the ulnar nerve, activating not only the median-innervated thenar muscle but also the anomalously innervated thenar and hypothenar muscles. In contrast, stimulation of the median nerve at the wrist elicits a smaller response lacking the ulnar component. If proximal stimulation elicits a larger response as compared to distal stimulation, it always implies the presence of an anomalous communication or technical problem. The reverse discrepancy may pose difficulty, mimicking a conduction block. For example, in the presence of Martin-Gruber anastomosis, studies of the ulnar nerve show a smaller amplitude of thenar or hypothenar compound muscle action potentials elicited by proximal rather than distal stimulation.86 Here stimulation at the wrist activates the additional anomalous fibers, giving rise to a full response, whereas stimulation at the elbow spares the communicating
Nerve Conduction Studies branch still attached to the median nerve, sometimes mimicking ulnar neuropathy at the elbow.102 When recording from the first dorsal interosseous, adductor pollicis, or hypothenar muscles after stimulation of the median nerve at the elbow, volume-conducted potentials from distant median-innervated muscles may mimic an anomalously activated response. Under this circumstance, a careful comparison between distal and proximal stimulation usually clarifies the ambiguity.11,39,70,89,140 In difficult cases, recording with a needle electrode may localize the origin of the recorded response, although distant activities, if present, may still confuse the issue. The collision technique70,132 provides selective blocking of unwanted impulses transmitted via the communicating fibers (Fig. 7-7). Normally, antidromically directed impulses from the distal stimulation will completely block the orthodromic impulses from the proximal stimulation in the same nerve.53,152 The orthodromic impulses traveling through an anastomotic branch to the ulnar nerve, however, would bypass the antidromic impulses and escape collision.70
Figure 7-7. Muscle action potentials recorded from the hypothenar eminence after stimulation of the median nerve at the wrist (S1) or elbow (S2). The top tracing shows a volume conducted potential from thenar muscles (U-shaped wave of positive polarity). The middle tracing reveals a small negative potential superimposed upon the thenar component. In the bottom tracing, collision technique clearly separated the anomalous response (bracket), with Si preceding S2 by 4 ms. [From Kimura,76 with permission.]
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques
189
Figure 7-8. Muscle action potentials recorded from the thenar eminence after stimulation of the median nerve at the wrist (S1) or elbow (S2) as in Figure 7-7. In the middle tracing, a large compound action potential buried a small anomalous response mediated by the anastomosis. In the bottom tracing, a collision technique separated the anomalous response (bracket) with S1 preceding S2 by 4 ms. [From Kimura,76 with permission.]
The technique helps identify and characterize the anomalous response, although a thenar, as opposed to hypothenar, response elicited via anastomosis tends to overlap with a large median potential elicited distally (Fig. 7-8). Delay of the proximal stimulation by a few milliseconds usually achieves its satisfactory separation from the distally elicited response. The time interval must not exceed the latency difference between the two stimulus sites, lest the orthodromic impulse escape antidromic collision in the absence of an anomalous route of transmission. If this anastomosis accompanies the carpal tunnel syndrome, stimulation of the median nerve at the elbow evokes two temporally dispersed potentials, a normal ulnar component and a delayed median component. The latency of the initial ulnar response erroneously suggests the presence of normal-conducting median fibers. In contrast, stimulation of the median nerve at the wrist evokes a delayed response without an ulnar component.62,70,89 The discrepancy between proximal and distal stimulation would
lead to an unreasonably fast conduction velocity from the elbow to the wrist.1l,70,89 The anomalously innervated ulnar muscles usually lie at some distance from the recording electrodes placed on the thenar eminence. Thus, the ulnar component commonly, though not always, displays an initial positive deflection.47 As mentioned earlier, a collision technique can block impulses in the anomalous fibers without affecting those transmitted along the median nerve proper (Fig. 7-9). Severence or substantial injury of the ulnar nerve at the elbow ordinarily results in wallerian degeneration and inexcitability of the distal segment. In the presence of this anomaly, stimulation at the wrist will excite the communicating fibers that bypass the lesion to evoke a small but otherwise normal muscle action potential. In extreme cases, separation of the ulnar nerve at the elbow may not appreciably affect the intrinsic hand muscles because all or nearly all ulnar fibers attached to the median nerve escape injury. In this rare condition, called all-median hand, the intrinsic hand muscles ordinarily sup-
Nerve Conduction Studies
190
Figure 7-9. A 55-year-old man with the carpal tunnel syndrome and the Martin Gruber anastomosis. Stimulation at the elbow (S2) spread to the ulnar nerve through the anomalous communication (middle tracing). Another stimulus (S3) applied to the ulnar nerve at the wrist (bottom tracing) blocked the impulses transmitted through the communicating fibers. In the bottom tracing, S3 preceded S2 by 4 ms to avoid the overlap of the 76 muscle responses. [From Kimura, with permission.]
plied by the ulnar nerve receive innervation via the communicating fibers.101 Electromyography may reveal normal motor unit potentials in the ulnar-innervated muscles, despite severe damage to the ulnar nerve at the elbow. Conversely, an injury to the median nerve at the elbow could lead to the appearance of spontaneous discharges in the ulnar-innervated intrinsic hand muscles. Hence, an anomaly of this type, if undetected, gives rise to considerable confusion in the interpretation of electrophysiologic findings.48,161
Anomalies of the Hand Common anomalies of the peripheral nerves include variations in innervation of the intrinsic hand muscles.136 Although not as widely recognized as the medianto-ulnar communication, they too constitute sources of error in the evaluation of nerve conduction velocity and electromyography. Electrophysiologic tech-
niques often hint at the presence of such anastomoses, although precise characterization and delineation of the extent of the anomaly call for anatomic studies.147 Various communications may link the recurrent branch of the median and the deep branch of the ulnar 14,32,98,122,128 nerve in the lateral portion of the hand. Any of the intrinsic hand muscles, the flexor pollicis brevis in particular, may receive median, ulnar, or dual innervation.135 In a small percentage of cases, thenar muscles, including the adductor pollicis, may derive their supply exclusively from the median or ulnar nerve.38,130 In addition to neural anastomoses, skeletal anomalies of the upper limb may confuse the clinical picture. The congenital absence of thenar muscles, for example, may suggest a false15diagnosis of carpal tunnel syndrome. The posterior interosseous nerve may innervate accessory hand muscles consistent with extensor digitorum brevis manus.107 The deep branch of the ulnar nerve may form a motor neural loop, caus-
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques ing an atypical clinical presentation after penetrating injuries or compression neuropathy at the wrist.123
Accessory Deep Peroneal Nerve The most frequent anomaly of the lower limb involves the innervation of the extensor digitorum brevis, the muscle commonly used in conduction studies of the peroneal nerve. This muscle usually derives its supply from the deep peroneal
191
nerve, a major branch of the common peroneal nerve. In 20-28 percent of an unselected population, the superficial peroneal nerve also contributes via a communicating fiber. This branch, called the accessory deep peroneal nerve (Fig. 7-10), descends on the lateral aspect of the leg after arising from the superficial peroneal nerve, then passes behind the lateral malleolus and proceeds anteriorly to innervate the lateral portion of the extensor digitorum brevis.58,90,162 Occasionally, the extensor digitorum brevis
Figure 7-10. A. Compound muscle action potentials recorded from surface electrodes over the extensor digitorum brevis after a maximal stimulus to the common peroneal nerve at the knee (A), deep peroneal nerve on the dorsum of the ankle (B), accessory deep peroneal nerve posterior to the lateral malleolus (C and D), and tibial nerve posterior to the medial malleolus (K) at the ankle. Left and right panels show responses before and after block of the accessory deep peroneal nerve with 2 percent lidocaine posterior to the lateral malleolus. Diagram of the foot indicates the site of block (X) and the points of stimulation (B, C, and D) and recording (R). B. Course of the accessory deep peroneal nerve and action potentials recorded with coaxial needle electrode (R) in the lateral belly of the extensor digitorum brevis muscle following stimulation of the common peroneal nerve at the knee (A), just below the head of fibula (B), superficial peroneal nerve (C), accessory deep peroneal nerve posterior to the lateral malleolus (D) and deep peroneal nerve on the dorsum of the ankle (E). The volume-conducted potential from the medial bellies of the extensor digitorum brevis (£5 reduces amplitude of action potential of the lateral belly with simultaneous stimulation of the common peroneal nerve at A or B. [From Lambert,90 with permission.]
Nerve Conduction Studies
192
may receive exclusive supply from this communication.111 The anomaly, when inherited, shows a dominant trait.21 In patients with the anastomosis, stimulation of the deep peroneal nerve at the ankle elicits a smaller compound muscle action potential than stimulation of the common peroneal nerve at the knee. Stimulation of the accessory deep peroneal nerve behind the lateral malleolus activates the anomalously innervated lateral portion of the muscle. Injury to the deep peroneal nerve ordinarily causes weakness of the tibialis anterior, extensor digitorum longus, extensor hallucis longus, and extensor digitorum brevis. In the presence of the anastomosis, however, such a lesion would spare the lateral portion of the extensor digitorum brevis. Overlooking this possibility would, therefore, lead to an erroneous interpretation.28,46 The collision technique70 may help identify isolated abnormalities of the accessory deep peroneal nerve.133
Anomalous Communication Between Peroneal and Tibial Nerve The sural nerve, ordinarily a sensory branch of the tibial nerve, may arise from the common peroneal nerve, which in turn receives anastomosis from the tibial nerve.119 Although the nerve usually consists purely of sensory fibers, its anomalous motor branch may innervate96the abductor digiti quihti of the foot. Rare motor anastomosis between the peroneal and tibial nerves, if undetected, may give rise to an erroneous conclusion by showing patterns of waveform change similar to those seen in Martin-Gruber anomaly.142 In rare anomalies, the tibial nerve may supply all the intrinsic foot muscles.97 In documenting the innervation pattern of this and other rare anastomoses, volume-conducted responses often confuse the issue.3,99 A pair of surface electrodes placed anywhere in the foot register a muscle action potential after stimulation of peroneal or tibial nerve. Needle studies may also fall short of selectively recording from individual intrinsic foot muscles. In questionable cases, a
collision technique70 or a nerve block usually provides conclusive evidence.
5
PRINCIPLES AND PITFALLS OF WAVEFORM ANALYSIS
Physiologic and Pathologic Temporal Dispersion In nerve conduction studies, latency measure of the fastest fibers allows calculation of the maximal motor or sensory velocities. In addition, waveform analyses of compound muscle and sensory nerve action potentials help estimate the range of the functional units.25,74,118 This aspect of the study provides an equally, if not more, important assessment, especially in the study of peripheral neuropathies with segmental block, in which surviving axons may conduct normally.41,51,95,109,145,151 In clinical tests of motor and sensory conduction, the size of the recorded response approximately parallels the number of excitable fibers. Any discrepancy between responses to proximal and distal shocks, however, does not necessarily imply an abnormality. The impulses of slow-conducting fibers lag increasingly behind those of fast-conducting fibers over a long conduction path.9,22,89 With increasing distance between stimulating and pickup electrodes, the recorded potentials become smaller in amplitude and longer in duration; and, contrary to the common belief, the area under the waveform also diminishes. Thus, the size of the recorded response depends to a great extent on the site of stimulation. In fact, stimulation proximally in the axilla or Erb's point may normally give rise to a small or undetectable digital potential, despite a large response elicited73,117,164 by stimulation at the wrist or palm. For the same number of conducting fibers activated by the stimulus, the size of sensory potentials changes linearly with the length of the nerve segment.79,85 A physiologic reduction both in amplitude and in the area under the waveform may erroneously suggest a conduction abnormality between the proximal and the distal sites of stimulation.
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques With short-duration diphasic sensory spikes, a slight physiologic latency difference could line up the positive peaks of the fast fibers with the negative peaks of the slow fibers, canceling both (Figs. 7-11 and 7-12A). According to computer simulation,83,121 this phenomenon alone can reduce the normal sensory nerve action potential to below 50 percent in area as well as in amplitude, a conservative figure based on computation of a limited number of nerve fibers for analysis. Thus, a major reduction in the size of the compound sensory action potential can result solely from physiologic phase cancellation. In contrast, the same temporal dispersion has less effect35,116 on compound muscle action potential because motor unit potentials of longer duration superimpose nearly in phase rather than out of phase, despite the same latency shift, resulting in less cancellation compared to sensory potentials (Fig. 7-12B). In support of this view, the duration change of the sensory potential, expressed as a percentage of the respective baseline values, far exceeds that of the muscle response.79 As expected from the term, duration-de-
Figure 7-11. Simultaneous recordings of compound muscle action potentials (CMAP) from the thenar eminence and sensory nerve action potentials (SNAP) from index and middle fingers after stimulation of the median nerve at palm, wrist, elbow and axilla. With progressively more proximal series of stimuli elicited nearly the same CMAP but progressively smaller SNAP from the wrist to the79axilla. [From Kimura et al., with permission.]
193
pendent phase cancellation,79 a physiological temporal dispersion, also reduces substantially the amplitudes of short-duration muscle action potentials such as those recorded from intrinsic foot muscles. The degree of overlap between peaks of opposite polarity depends on the separation between GI and G2, which dictates the duration and waveform of unit discharges.9 A maximal cancellation results when a waveform contains negative and positive phases of comparable size. In a triphasic orthodromic sensory potential, as compared with biphasic antidromic digital potentials, the initial positivity provides an additional probability for phase cancellation. Changes in temperature also affect the temporal dispersion, influencing the fast- and slow-conducting fibers more or less equally in percentage and therefore differently in absolute terms.129 The equations for the best fit lines between nerve length and other measurements in one study may not necessarily apply to another unless the recording technique conforms to the particular specifications.
194
Nerve Conduction Studies
Figure 7-12. A. Sensory action potentials. A model for phase cancellation between fast (F] and slow (S) conducting sensory fibers. With distal stimulation two unit discharges summate in phase to produce a sensory action potential twice as large. With proximal stimulation, a delay of the slow fiber causes phase cancellation between the negative peak of the fast fiber and the positive peak of the slow fiber, resulting in a 50% reduction in size of the summated response. [From Kimura et al.,79 with permission.] B. Compound muscle action potentials. Same arrangements as in A to show the relationship between fast [F] and slow (S) conducting motor fibers. With distal stimulation, two unit discharges representing motor unit potentials summate to produce a muscle action potential twice as large. With proximal stimulation, long duration motor unit potentials still superimpose nearly in phase despite the same latency shift of the slow motor fiber as the sensory fiber shown in A. Thus, a physiologic temporal dispersion alters the size of the muscle action potential only minimally, if at all. [From Kimura et al.,79 with permission.]
If the latency difference between fast- and slow-conducting motor fibers increases substantially, as might be expected in demyelinating neuropathy, muscle responses also diminish dramatically based solely on phase cancellation as predicted by our
model83 and computer simulation with a broader spectrum of motor nerve conduction velocities.92 This type of phase cancellation reduces the amplitude of muscle response well beyond the usual physiologic limits in the absence of conduction block.
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques Thus, in pathologic temporal dispersion associated with segmental demyelination, focal phase cancellation of the muscle action potential could give rise to a false impression of motor conduction block. This phenomenon explains an occasionally encountered discrepancy between severe reduction in amplitude of the compound muscle action potential, on the one hand, and relatively normal recruitment of the motor units and preserved strength, on the other. Thus, sustained reduction in size of compound muscle action potential may result from a pathological temporal disper-sion rather than a prolonged neurapraxia.83,109,121 A simple model provides an excellent means to test the effects of desynchronized inputs.83 A shock applied to the median (S1) or ulnar (S2) nerve at the wrist evokes a sensory potential of the fourth digit and a muscle potential over the thenar eminence. Hence, a concomitant
Recorded response
195
application of S1 and S2 with varying interstimulus intervals simulates the effect of desynchronized inputs (Fig. 7-13). In 10 hands, an inter stimulus interval on the order of 1 ms between S1 and S2 caused a major reduction in sensory potential by as much as 50 percent but little change in muscle action potential. With further separation of S1 and S2, the muscle response began to decrease in amplitude and area, reaching a minimal size at interstimulus intervals of 5-6 ms. The duration also increased in proportion to the latency shift, although a gradual return of the response to the baseline obscured the magnitude of this aspect of change in waveform. A latency difference slightly less than one half the total duration of unit discharge maximized the phase cancellation between the two components and consequently the loss of area under the waveform. Further increase in latency dif-
Algebraic sum
Difference
Figure 7-13. A. Antidromic sensory potentials of the fourth digit elicited by stimulation of the median (S1) or ulnar (S2) nerve (top two tracings), or by both S1 and S2 at interstimulus intervals ranging from 0 to 2.0 ms (left). Algebraic sums of the two top tracings (middle) closely matched the actual recording at each interval as evidenced by small difference shown in computer subtraction (right). The area under the negative peak reached 83 a minimal value at 0.8 ms in actual recordings as well as in calculated waveforms. [From Kimura et al., with permission.]
Nerve Conduction Studies
196 Recorded response
Algebraic sum
Difference
Figure 7-13. B. Compound muscle action potentials from the thenar eminence elicited by stimulation of the median (S1) and ulnar (S2) nerve (top two tracings), or by both S1 and S2 at interstimulus intervals ranging from 0 to 10 ms (lefi). Algebraic sums of the top two tracings almost, but not exactly, equaled the actual recordings as shown by computer subtraction at each interstimulus interval (right). The area under the negative peak 83 reached a minimal value at 5 ms in actual recordings as well as in calculated waveform. [From Kimura et al., with permission.]
ference results in complete separation of the two potentials, precluding phase cancellation. As an inference, pathological temporal dispersion may decrease the size of the compound sensory or muscle action potentials or conversely counter physiologic phase cancellation, causing paradoxical increase of the responses (see Figure 6-7). Comparison between distally and proximally elicited responses often fails to differentiate physiologic, as opposed to pathologic, temporal dispersion, not to mention conduction block. Many variables, such as electrode position and distance, make the commonly held criteria based on percentage reduction nearly untenable 91except in entirely standardized studies. A simpler, more practical approach relies on a linear relationship seen in physiological phase cancellation between the latency and the size of the
recorded responses78 (see Fig. 7-11). Although this calls for segmental stimulation at more than two sites to test the linearity of observed changes, it enjoys the distinct advantage of having a built-in internal control for all recording variables such as inter-electrode spacing. A nonlinear reduction in amplitude or area, often associated with waveform changes, indicates either a pathological temporal dispersion or conduction block. The distinction between the two possibilities must in part depend on clinical cue as stated below (Fig. 7-14 and 7-15). In summary, physiologic as well as pathologic temporal dispersion can effectively reduce the area of diphasic or triphasic evoked potentials recorded in bipolar derivation. The loss of area under the waveform seen in the absence of conduction block implies a duration-dependent phase cancellation of unit discharges
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques
197
Figure 7-14. Ulnar nerve conduction study with segmental stimulation at the wrist, below elbow, above elbow, and at the axilla. Traces show a complete conduction block in a 42-year-old patient (right), a partial conduction block in a 40-year-old patient (center), and a normal study (left) for comparison.
within the compound action potential. Segmental studies provide the best means of detecting pathologic nonlinear changes as opposed to physiologic linear regres-
sion in amplitude and area of compound action potential. An awareness of this possibility helps analyze dispersed action potentials in identifying various patterns of
Figure 7-15. Segmental study of the ulnar nerve in 1 cm increments from below elbow (1) to above elbow (7). A. A 41-year-old man with distal ulnar neuropathy had prolonged terminal latency (3.7 ms) with normal conduction across the elbow. B. A 52-year-old woman with a tardy ulnar palsy showed normal distal latency with an abrupt drop in amplitude above the elbow (5 to 6 and 6 to 7), indicating a partial conduction block as the cause of weakness.
198
neuropathic processes.74 Area difference between negative and positive peaks in each unit discharge provides a unique measure, which sums without phase cancellation irrespective of desynchronization among different units. A composite of this calculated value should also remain the same between proximal and distal sites of stimulation. Thus, in analyzing compound muscle or nerve action potentials, subtraction of the positive peak from the negative peak should theoretically yield an identical value regardless of the site of stimulation along the course of the nerve. This approach, therefore, can circumvent the ambiguity of waveform changes caused by temporal dispersion and phase cancellation. In practice, however, muscle and nerve action potentials tend to have negative and positive peaks of similar size and consequently a small area difference between the two, making its precise determination difficult. Also, a baseline shift or other electrical interference poses a major technical problem for reliable measure of area for this purpose. Referential derivation of a monophasic waveform in a "killed-end" arrangement also conserves the area irrespective of stimulus sites showing no phase cancellation. This type of recording, however, may register a stationary far-field potential generated by the propagating impulse crossing 80,81 the partition of the volume conductor. Such a steady potential could, in turn, distort the waveform of the near-field activity (see Chapter 20-3).
Detection of Conduction Block In a demyelinating polyneuropathy, slowing of nerve conduction often accompanies a reduction of amplitude associated with a partial conduction block.8,34,63,155 Conversely the evidence of conduction block usually implies the presence of focal demyelination,134 although other conditions such as ischemic neuropathy can cause similar reversible changes.41,54 Increased ranges of conduction velocities result in pathological temporal dispersion broadening the evoked action potential. Desynchronization of the nerve volley may also result from repetitive discharges at
Nerve Conduction Studies the site of axonal injury after the passage of a single impulse. Unless secondary axonal degeneration is induced by damage of the myelin sheath, electromyography reveals little or no evidence of denervation. The motor unit potentials, though normal in amplitude and waveform, show poor recruitment because some fibers fail to transmit the impulse. The usual criteria for conduction block in motor fibers revolve around the comparison of compound muscle action potentials elicited by proximal versus distal stimulation, expressed in the ratio of their amplitudes or areas.1,114,146,158 This ratio remains normal in axonal neuronopathy, which reduces distal and proximal responses equally. Generally accepted diagnostic clues used for motor conduction block comprises a reduction in amplitude ratio greater than 20-50 percent, with less than 15 percent increase in duration of the compound muscle action potential elicited by proximal stimulation. These criteria, however, do not neccessarily apply in all studies because the effects of temporal dispersion vary depending on the electrode placement. A triple stimulation method with double collisions allows identification of motor conduction block in the face of desynchronization.127 The technique, however, fails if the lesion is too proximal or if it compromises nerve excitability at stimulus sites as the consequence of demyelination or degeneration. In documenting motor conduction block, the combination of clinical and electrophysiologic finding usually circumvents the ambiguity of the criteria based purely on waveform analysis.77 In the presence of conduction block, a shock applied distally to the nerve lesion in question elicits a vigorous twitch and a large distal amplitude despite disproportionately severe clinical weakness75 associated with paucity of voluntarily activated motor unit potentials.20 As an exception, the same finding also characterizes any weakness attributable to upper motor neuron involvement or hysteria or during the first few days of axonal lesion before the distal stum loses its excitability.106 In equivocal cases, inability to distinguish focal pathological temporal dispersion
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques from conduction block poses no major practical problem because either finding usually suggests demyelination, leading to an appropriate treatment. The absence of F waves complements conventional nerve conduction studies to document conduction block in the proximal segment.36,69 Several other factors play an important role in the clinical assessment of conduction block. The use of insufficient stimulus intensities at the proximal site erroneously reduces the proximal amplitude. Likewise, increased threshold for excitation in regenerated or chronically demyelinated nerves may account for a reduced proximal response.108 In some cases of multifocal motor neuropathy, failure to maximally excite the involved segment calls for near-nerve stimulation using a needle electrode. Alternatively, stimulation of more proximal, unaffected nerve segment may give rise to a normal response, indicating the passage of impulse across the lesion site despite its abnormally elevated threshold for local excitation (Figs. 7-16). During the course of wallerian degeneration, the distal stump of the nerve remains viable for several days at a time when its proximal part fails to transmit the signal across the injury site. In this situation, conduction studies performed soon after nerve severance show a decreased proximal-distal amplitude ratio. Unexpected excitation of anomalous branches such as Martin-Gruber anastomosis may lead to a confusing discrepancy in amplitude, as does inadvertant70 current spread to a neighboring nerve. In addition, lesions selectively affecting smaller myelinated fibers may not result in major loss of the proximal-to-distal ratio. In antiserum-mediated experimental demyelination,88 smaller fibers underwent conduction block first. If this holds in the acute phase of demyelinating neuropathies such as Guillain-Barre syndrome, normal conduction studies do not necessarily rule out such selective involvement that might account for weakness. More slowly conducting fibers, however, belong to motor units generating relatively small twitches, whose contributions, if lost through conduction block, may cause only limited weakness.
199
Contrary to motor studies, which rely heavily on clinical assessment of weakness to define conduction block, sensory studies usually depend solely on waveform analysis of antidromic response elicited by short incremental stimulation (see this chapter, part 6). Surface stimulation applied at multiple sites may not necessarily indicate the exact point of nerve activation. An alternate method consists of stimulating the digital nerve and recording the orthodromic sensory potential at multiple points with a series of electrodes mounted 1 cm apart on a specially constructed flexible strap.57,74 Though applicable to any other superficially located sensory or mixed nerve, the method suffers from a major limitation. Using surface recording, the depth of the nerve from the skin surface greatly influences the amplitude of the evoked potential. Thus, a small potential derived from a deeply located nerve segment under the area in question may erroneously suggest a conduction block. In contrast to peripheral study, segmental recording registers comparable spinal somatosensory evoked potentials in intraoperative spinal cord monitoring. All recording electrodes are nearly equidistant to the spinal cord115,139,148,150,156 if placed in the subdural or epidural space, the ligamentum flavum, or the intervertebral disc. Figure 7-17 show unipolar recording from the ligamentum flavum at multiple levels after epidural stimulation of the cauda equina in a patient with cervical spondylotic myelopathy. The combination of an abrupt loss of the negative peak at one level, augmentation of the negative peaks in the leads closely caudal to that level, and monophasic positive waves at more rostral levels constitutes a typical pattern of waveform changes, indicating a complete focal conduction block. Paradoxically enhanced negative peak results from resynchronization of physiologically desynchronized signals because the leading impulses stop traveling when they reach the site of involvement, whereas the trailing impulses continue to propagate until they arrive at the same point. In addition, the fast-conducting fibers lose their terminal-positive phases, which would have reduced the
Figure 7-16. A. Motor and sensory conduction studies of the left median nerve in a patient with multifocal motor neuropathy. The left diagram illustrates the consecutive slices of MR images in relation to the sites of stimulation at the wrist crease (Al) and at 2 cm increments more proximally. One horizontal division equals 5 ms (motor) or 2 ms (sensory), and one vertical division corresponds to the gain indicated at the end of each trace, together with stimulus intensity. Note the complete and selective motor conduction block across the segment between A2 and A3, corresponding to the site of maximal nerve enlargement. [From Kaji et al,66 with permission.] B. A repeat study in the same patient as shown in A after return of strength of the median-innervated intrinsic hand muscles. High-intensity stimulation failed to excite the nerve along the affected segments, A5-A8, mimicking a conduction block. More proximal stimulation at the elbow applied to the presumably normal nerve segments, A9-A11, however, induced a series of temporally dispersed muscle responses associated with thumb abduction, indicating recovery of conduction. [From Kimura,77 with permission.]
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques negative phases of the slower fibers by physiologic phase cancellation. Even when only some of the fibers sustain a conduction block, the identical mechanism enhances the negative peak at the points immediately preceding an incomplete lesion. Thus, the response consists of positive-negative diphasic waves with enhanced negativity at points immediately preceding the block, a diphasic wave with reduced negativity at the point of the block, and initialpositive waves alone or abolition of any wave at points beyond the block.78,149
Distribution of Conduction Velocities In contrast to the onset latency of the action potential, which relates only to the fastest conducting fibers, its waveform reveals the functional status of the remaining slower conducting fibers. With the loss
201
of nerve fibers, a smaller range of conduction velocity reduces the duration of the compound action potential. Conversely, disproportionate slowing of slower conducting fibers will result in increased temporal dispersion. The greater the range between the fastest and slowest nerve fibers, the longer the duration of the evoked potential. Temporal dispersion also increases with more proximal stimulation in proportion to the distance to the recording site.35,71,80 Near-nerve recordings uncover the late components of sensory action potentials not detectable by surface electrodes. The minimum conduction velocity thus determined for the slower fibers may serve as a sensitive measure of both axonal and demyelinating peripheral nerve pathology.138 The use of needle electrodes improves the selectivity of recording in measuring conduction velocity of different motor units within a given muscle. A wide range of mo-
Figure 7-17. A. A T1-weighted MR image (TR 400 ms; TE 13 ms) (left) and a recording of spinal somatosensory evoked potentials (right) obtained from a 65-year-old patient with cervical myelopathy. Epidural stimulation at L2 elicited a series of potentials recorded unipolarly from the ligamentum flavum of C7 to Tl through C12. Note the progressive increase in size of the negative component (arrows pointing up) from C7 to Tl (-3) through C5 to C6 (-1) with the abrupt reduction at C4 to C5 (0) followed by a monophasic positive wave at C3 to C4 (+1). The negative wave doubled in amplitude and quadrupled in area at '-1' compared to '-3'. The '0' corresponded to the level of the spinal cord, showing the most prominent compression on the MR image. [From Tani, Ushida, Yamamoto, et al,149 with permission.]
202
Nerve Conduction Studies
Figure 7-17. B. AT1-weighted MR image (TR 600 ms; TE 90 ms) (left) and a recording of spinal somatosensory evoked potentials (right!) obtained from a 36-year-old patient with cervical spondylotic myelopathy. Epidural stimulation at T11 elicited a series of potentials recorded unipolarly from the ligamentum flavum of Tl to C2 through C3 to C4 after epidural stimulation at T11. Note the progressive increase in size of the second negative component (arroius pointing up) from Tl to T2 (-3) through C6 to C7 (-1) with the abrupt reduction at C5 to C6 (0). The zero corresponded to the level of149 the spinal cord, showing a moderate compression on the MR image. [From Tani, Ushida, Yamamoto, et al, with permission.]
tor fibers with different conduction characteristics sampled by this means show a close correlation to the twitch tension and recruitment threshold.27 The technique has limited clinical application because patients tolerate poorly the multiple needle insertions required for isolation of the slowest-conducting fibers. A number of publications have dealt with mathematical models for studying the waveform.23,94,110,144,163 The method allows estimation of conduction velocity distribution in a nerve bundle based on a detailed model of the compound action potential as a weighted sum of asynchronous single-fiber action potentials.31 The technique has provided some interesting, though unconfirmed, results. Distribution of conduction velocity may reflect the pathologic changes as reported in the study of the sural nerve affected by nhexane neuropathy.169 Nerve conduction
velocity of the large myelinated axons, which contribute to the surface recorded response, may vary by as much as 25 m/s between fast and slow sensory fibers but over a much narrower range of 11 m/s for motor fibers.30 This observation, although not universally accepted,29 would in part explain the different effect of temporal dispersion on sensory and motor fibers for a given length of nerve segment.117 Decomposition techniques in general suffer from the inherent limitation of identifying individual elements no longer retained within the compound action potential because of phase cancellation. Any sophistication in technology cannot retrieve the information, if already lost. Besides, some of the assumptions derived from normal distributions may not necessarily apply in various types of neuropathy.26,153 In the analysis of compound muscle action potential, the length of
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques axon, rather than the conduction characteristics, may dictate the order of line-up of motor unit potentials. Thus, in motor conduction, unlike in sensory conduction, short latencies do not necessarily imply fast-conducting elements. This explains why the use of peak latencies does not necessarily yield a slower conduction velocity compared to the conventional calculations based on the onset latencies. Careful attention to the waveform of each evoked potential improves the accuracy of interpretation in any electrophysiologic study. If the responses have dissimilar shapes when elicited by distal and proximal stimuli, the onset latencies probably represent fibers of different conduction characteristics. This discrepancy results, for example, from the use of a submaximal stimulus at one point and a supramaximal stimulus at a second site. In diseased nerves, the impulse from a proximal site of stimulation may fail to propagate in some fibers because of conduction block even with an adequate shock intensity. In addition, apparently supramaximal stimuli may not activate a bundle of regenerating or severely demyelinated axons if local structural changes or nerve pathology per se effectively prevent the excitation of the nerve segment. The impulses, once generated voluntarily or reflexively at a proximal site, however, may propagate along these fibers, giving rise to a confusing set of electrophysiologic findings. Any of these circumstances preclude the calculation of conduction velocity with the conventional formula.
Collision Technique to Block Fast- or Slow-Conducting Fibers The duration of the compound action potentials, although useful as an indirect es-
Authors Thomas et al.152120 Poloni53 and Sala Hopf 141 Skorpil
timate, falls short of providing a precise measure of slow fibers. Different methods devised for a more quantitative assessment commonly employ the principle of collision.152 A distal stimulus of submaximal intensity initially excites the large-diameter, fast fibers with low thresholds. A shock of supramaximal intensity given simultaneously at a proximal site, then, allows selective passage of impulses in the slower fibers, because antidromic activity from the distal stimulation blocks the fast fibers. This assumption, however, does not always hold, because the order of activation with threshold stimulation depends in part on the position of the stimulating electrode 65 in relation to the different fascicles. An alternative method utilizes a series of paired shocks of supramaximal inten43,49,50,53,59,60,112,125,129 This technique, in essence, consists of incremental delay of proximal shock after distal stimulation without varying stimulus intensity. Shocks applied simultaneously cause collision to occur in all fibers. With increasing intervals between the two stimuli, the fastest fibers escape collision before the slow fibers. Measurement of the minimal interstimulus interval sufficient to produce a full muscle action potential provides an indirect assessment of the slowest conduction (Table 7-1). Direct latency determination of the slowest fibers requires blocking of the fast conducting fibers, leaving the activity in the slower fibers unaffected. The use of two sets of stimulating electrodes, one placed at the axilla and the other at the wrist, allows delivery of two stimuli, S(A1) and S(A2), through the proximal electrodes and another shock, S(W), through the distal electrodes. The antidromic impulse of S(W) blocks the orthodromic impulse of S(A1), provided the distal shock precedes the arrival of the proximal im-
Table 7-1 Range of Conduction Velocity in Motor Fibers of the Ulnar Nerve Fastest Fibers Slowest Fibers 60.0 ± 3.2 61.1 ±4.5
203
37.7 ±7.1
Range 30-40% 35-39% 4-7 m/s 22.4 m/s
204
pulse. With an appropriate adjustment of the interstimulus interval between S(A1) and S(W), the collision takes place only in the slow fibers, sparing the antidromic activity from S(W) in the fast fibers. Thus, the impulse of the subsequent proximal stimuli, S(A2), collides with the antidromic activity only in the fast fibers. In this way, the muscle action potential elicited by S(A2) corresponds to the remaining slow conducting fibers that selectively transmit the orthodromic impulses (Fig. 7-18). This technique allows direct determination of the amplitude and latency of the slowest-conducting fibers. The muscle action potential elicited by S(A2) shows progressive diminution of amplitude as the
Nerve Conduction Studies antidromic impulse of S(W) eliminates an increasing number of fast-conducting fibers. The latency changes, however, do not always coincide exactly with the values expected from the time interval between S(A1) and S(W), presumably because the impulses in the slowest conducting fibers do not necessarily arrive at the motor end-plate last. The conduction time must depend not only on the speed of the propagated impulse but also, and perhaps more importantly, on the length of fine terminal branches that characteristically lack myelin sheath. Even though the branches vary in length only on the order of a few millimeters, this degree of difference can still give rise to a
Figure 7-18. A. Compound muscle action potential recorded by surface electrode placed over the abductor digiti minimi after stimulation of the ulnar nerve. The diagrams on the left show orthodromic (solid line) and antidromic (dotted line) impulses generated by three stimuli, S(A1), S(W), and S(A2) delivered at the axilla, wrist, and axilla, respectively. Note the collision between the orthodromic impulse from S(A1) and antidromic impulse of S(W) in slow conduction fibers (S), and between the orthodromic impulse of S(A2) and antidromic impulse of S(W) in the fast conducting fibers (F]. The orthodromic impulse of S(A2) propagates along the slow conducting fibers and elicits the second compound muscle action potential. B. Paired axillary shocks of supramaximal intensity combined with a single shock at the wrist (cf. bottom tracing in A). The first axillary stimulation, S(A1)preceded the wrist stimulation, S(W), by intervals ranging from 6.0 to 8.0 ms in increments of 0.2 ms. Adjusting the second axillary shock, S(A2), to recur always 6.0 ms after S(W) automatically determined interstimulus interval between S(A1) and S(A2). The figures on the left shows the entire tracing with a slow sweep triggered by S(A1) for amplitude measurement. The ftgwes on the right illustrate latency determination with a fast sweep triggered by S(A2) and displayed after a predetermined delay of 6.0
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques substantial latency change at this level, where the impulse normally conducts at a very slow rate.
6
STUDIES OVER SHORT AND LONG DISTANCES
Segmental Stimulation in Short Increments Ordinary conduction studies suffice to approximate the site of involvement in entrapment neuropathies.25 More precise localization requires inching the stimulus in short increments along the course of the nerve 72in order to isolate the affected segment. In the evaluation of a focal lesion such as compressive neuropathy, inclusion of the unaffected segments in calculation dilutes the effect of slowing at the site of lesion and lowers the sensitivity of the test. Therefore, incremental stimulation across the shorter segment helps isolate a localized abnormality that may otherwise escape detection (see Chapter 6-2). Thus, the study of short segments provides better resolution of restricted lesions. For example, assume a nerve impulse conducting at a rate of 0.2 ms/cm (50 m/s) except for a 1 cm segment where demyelination has doubled the conduction time to 0.4 me/cm. In a 10 cm segment, normally covered in 2.0 ms, a 0.2 ms increase would constitute a 10 percent change, or approximately one standard deviation, well within the normal range of variability. The same 0.2 ms increase, however, would represent a 100 percent change in latency if measured over a 1 cm segment signaling a clear abnormality. The large per unit increase in latency more than compensates for the inherent measurement error associated with multiple stimulation in short increments.12,40 This technique is best suited for assessing a possible compressive lesion, such as in carpal tunnel syndrome,57,72,124,137 ulnar neuropathy at the elbow,13,68 or peroneal nerve entrapment at the knee.67 With stimulation of a normal median nerve in 1 cm increments across the wrist, the latency changes approximately 0.16-0.21 ms/cm from mid-palm to dis-
205
tal forearm. A sharply localized nonlinear latency increase across a 1 cm segment indicates a focal abnormality. An abrupt change in waveform nearly always accompanies a latency72increase across the site of compression. In fact, waveform analysis often localizes a focal lesion, unequivocally confirming the validity of excessive latency change that might have resulted from inaccurate advances of the stimulating electrodes or inadvertent spread of stimulus current, activating a less affected and consequently more excitable neighboring segments. If technical difficulties preclude a complete study across the presumed site of the lesion, incremental stimulation of the more proximal and distal segments suffices to delineate the abnormality. In these cases, the waveform analysis shows abrupt changes together with nonlinear shift of the onset (or peak) latencies of successive responses above and below the affected zone, forming two parallel lines rather than one. These findings confirm a focal lesion within the short interval in question encompassed by normal segments proximally and distally.
Late Responses for Evaluation of Long Pathways Nerve stimulation studies commonly used in an electromyographic laboratory apply mainly to the distal, relatively short segments of the peripheral nerves. In assessing a more diffuse or multi-segmental process as might be seen in polyneuropathies, the longer the segment under study, the more evident the conduction delay. In other words, this approach has an advantage in accumulating all the segmental abnormalities, which individually might not show a clear deviation from the normal range. If a nerve impulse conducts at a rate of 0.2 ms/cm (50 m/s), for example, a 20 percent delay for a 10 cm segment is only 0.4 ms, whereas the same change for a 100 cm segment amounts to 4.0 msec, an obvious increase that is easily detectable. In addition, evaluating a longer as compared to shorter segment improves the accuracy of latency and distance measurement be-
Nerve Conduction Studies
206
cause the same absolute error constitutes a smaller percentage. Measuring the surface distance (carelessly) in a 10 cm segment, the actual value may vary between 9.5 and 10.5 cm. A 1 cm difference constitutes a 10 percent error. Thus, the calculated conduction velocity based on this measurement could vary between 50 m/s and 55 m/s. The same 1 cm error in a 100 cm segment represents only 1 percent error, resulting in the range of calculated conduction velocity between 50 m/s and 50.5 m/s. The same argument applies in determining the effect of possible error in latency measurement. Consequently, the study of a longer path offers a better sensitivity and accuracy and, as stated later, improved reproducibility in serial studies. A number of neurophysiological methods supplement the conventional techniques for the assessment of longer pathways.37 The selection of such techniques necessarily reflects the special orientation of each laboratory. Those of general interest include the F wave and the H reflex (see Chapters 18-6 and 19-2). Reproducibility of Various Measures In the assessment of polyneuropathy, nerve conduction studies serve as a measure of abnormality to document serial changes during the clinical course.84 Although the method provides a sensitive and objective indicator, its accuracy primarily depends on the adherence to technical details.17 Any deviations from the standards result in inconsistencies of the results. The awareness of this possibility plays an important role in designing a multicenter clinical trial, which involves many investigators of different backgrounds and training. Nonetheless, few studies have emphasized technical factors influencing the reproducibility of nerve conduction measurements in the evaluation of polyneuropathy.159 Several investigators5-17,52,131 reported on the reliability of nerve conduction velocity in normal subjects and 18,33,159 patients with diabetic polyneuropathy. A
study of median and peroneal nerves18,33 in patients with diabetic polyneuropathy revealed good reproducibility in nerve conduction velocity but not in amplitude. A few studies4,33,159 of diabetic polyneuropathy yielded excellent reproducibility of the median and peroneal F-wave latencies. In contrast, amplitude varied considerably for both the motor and sensory nerves although the use of large electrodes improved reproducibility 154 of compound muscle action potentials. Of a few 159 reported studies of F waves, all but one dealt with the experience at a single laboratory, showing variations of up to 10 m/s in conduction velocity.6,52 We also conducted a multicenter analysis on intertrial variability of nerve conduction studies to determine the confidence limits of the variations for use in future drug assessments for diabetic polyneuropathy.77,87 All measurements, repeated twice at a time interval of 1-4 weeks, followed a standardized method. In all, 32 centers participated in the study of 132 healthy subjects (63 men) and 65 centers in the evaluation of 172 patients with diabetic polyneuropathy (99 men). Motor nerve conduction studies consisted of stimulating the left median and tibial nerves and measuring amplitude, terminal latency, and minimal F-wave latency and calculating motor conduction velocity and F-wave conduction velocity. Sensory nerve conduction studies comprised antidromic recording of latency and amplitude after distal stimulation of the left median and sural nerves and calculation of sensory conduction velocities over the distal segment. In both the control group and the patient group, amplitude varied most, followed by terminal latency, and motor and sensory conduction velocity. In contrast, minimal F-wave latency showed the least change, with the range of variability only 10 percent for the study of the median nerve and 11 percent for the tibial nerve in normal subjects and 12 percent and 14 percent, respectively, in patients with diabetic polyneuropathy. These results support the contention that minimal F-wave latency serves as the most stable and consequently reliable measure for a sequen-
Reproducibility of Neurophysiological Measurements healthy volunteers
Figure 7-19. Reproducibility of various measures in (A) healthy volunteers and (B) patients with diabetic neuropathy. All studies were repeated twice at a time77 interval of 1-4 weeks to calculate relative intertrial variations as an index of comparison. [From Kimura, courtesy of Nobuo Kohara, M.D. et al, data from a multicenter reliability study sponsored by Fujisawa Pharmaceutical Co., Ltd.]
208
tial nerve conduction study of individual subjects. The same does not hold, however, when evaluating single patients against a normal range established in a group of subjects. Here F-wave conduction velocity suits better, minimizing the effect of limb length. Alternatively, some prefer the use of a nomogram, plotting the latency against the height as a simple, albeit indirect, measure of limb length. In the assessment of reproducibility, we use two independent indices, relative intertrial variation (RTV) and intertrial correllation coefficiency (ICC). Of the two, RIV directly represents a variation of measurements expressed as the percentages of the difference between V1 and V2 over the mean value of the two. Thus, RIV(%) = 100*(V2 - V1)/0.5(V1 + V2) where V1 and V2 represent the values of the first and the second measurements of the pair. The ranges of RIV within ±10 percent usually indicate a higher precision. Measures having larger interindividual differences are expected to show a greater infra-individual variability as well. The calculation of ICC takes this into consideration as follows to partially offset the ef-
Nerve Conduction Studies fects of a large variability among different subjects. Thus, where os2 and oe2 represent among-subject variance and experimental error. The value exceeding 0.9 indicates a reliable measure although, as seen from the formula, this may indicate a large amongsubject variance rather than a small experimental error. Figure 7-19 shows the 5th to 95th percentiles of RTVs and ICC in both groups; Figure 7-20 illustrates some examples of the individual data from the patients. The measures showing the range of RIV within ±10 percent included F-wave latency and F-wave conduction velocity of both median and tibial nerves, and sensory conduction velocity of the median nerve. In general, amplitudes showed a greater variation than latencies or nerve conduction velocities. Similarly, ICC exceeded 0.9 for F-wave latency of the median and tibial nerves in both the healthy subjects and the patients. Median nerve sensory nerve potential and median and tibial compound muscle action potentials had a large range of RIV despite a high ICC. In
Figure 7-2O. Comparison between the first and the second measures of (A) median nerve motor conduction velocity and (B) F-wave latency. Individual values plotting the first study on the abscissa and the second study on the ordinate show a greater reproducibility of the F-wave latency compared to the motor nerve conduction velocity (cf. Fig. 7-19). [From Kimura,77 courtesy of Noboru Kohara, M.D. et al, data from a multicenter reliability study sponsored by Fujisawa Pharmaceutical Co., Ltd.]
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques these amplitude measurements, a large among-subject variance of the amplitudes made os2 much greater than oe2, leading to a high ICC despite a considerable intertrial variability. Although a high ICC indicates a statistical correlation between two measurements,33,167 it does not necessarily imply a good reproducibility. Thus, to achieve an optimal comparison, a sequential study must exclude any measurements with a wide RIV regardless of ICC values. The calculation of RIV in addition to ICC helps detect the indices with an acceptable degree of reproducibility. In our data, F-wave latencies of the median and tibial nerves qualified as a reliable measure showing a large 1 ICC (>0.9) with a small RIV (±10 percent).
Clinical Considerations The main factors contributing to intertrial variability include inadequate control of skin temperature, insufficient stimulus intensity, errors in determining the latency and the surface distance, and difficulty in placing recording electrodes exactly at the6,16,17 same place on two separate occasions. Amplitudes vary most probably because of a shift in the recording site. A question often posed in regard to the accuracy and sensitivity of latency or velocity measurements relates to the length of the segment under study. Other factors being equal, should one study shorter or longer segment for better results? Although both approaches have merits and demerits, the choice depends entirely on the pattern of the conduction change. Short segmental approaches uncover a focal lesion involving a very restricted zone better than evaluating across a longer distance, which tends to obscure the abnormality. In contrast, studies of a longer segment detect diffuse or multisegmental abnormalities better, increasing sensitivity and decreasing measurement errors, which, in percentage, diminish in proportion to the overall latency and surface distance. The increased accuracy of the techniques in turn improves the repro-
209
ducibility of the results. In summary, short distances magnify focal conduction abnormalities despite increased measurement error, and long distances, though insensitive to focal lesions, provide better yields and reliability for a diffuse or multisegmental process.
REFERENCES 1. American Association of Electrodiagnostic Medicine: Consensus criteria for the diagnosis of partial conduction block. Muscle & Nerve (Suppl 8):S225-S229, 1999. 2. Amoiridis G: Median-ulnar nerve communications and anomalous innervation of the intrinsic hand muscles: An electrophysiological study. Muscle Nerve 15:576-579, 1992. 3. Amoiridis G, Schols L, Meves S, Przuntek H: Fact and fallacy in clinical and electrophysiological studies of anomalous innervation of the intrinsic foot muscles. Muscle Nerve 19:12271228, 1996. 4. Andersen H, Stalberg E, Falck B: F-wave latency, the most sensitive nerve conduction parameter in patients with diabetes mellitus. Muscle Nerve 20:1296-1302, 1997. 5. Bergmans J: On variability of conduction velocity in measurements in normal subjects. Electromyography 11:143-148, 1971. 6. Bleasel A, Tuck R: Variability of repeated nerve conduction studies. Electroencephalogr Clin Neurophysiol 81:417-420, 1991. 7. Bril V, Ellison R, Ngo M, Bergstrom B, Raynard D, Gin H, Roche Neuropathy Study Group: Electrophysiological monitoring in clinical trials. Muscle Nerve 21:1368-1373, 1998. 8. Brown WF, Snow R: Patterns and severity of conduction abnormalities in Guillain-Barre syndrome. J Neurol Neurosurg Psychiatry 54: 768-774, 1991. 9. Buchthal F, Rosenfalck A: Evoked action potentials and conduction velocity in human sensory nerves. Brain Res 3:1-122, 1966. 10. Buchthal F, Rosenfalck A: Sensory potentials in polyneuropathy. Brain 94:241-262, 1971. 11. Buchthal F, Rosenfalck A, Trojaborg W: Electrophysiological findings in entrapment of the median nerve at wrist and elbow. J Neurol Neurosurg Psychiatry 37:340-360, 1974. 12. Campbell WW: The value of inching techniques in the diagnosis of focal nerve lesions. Muscle Nerve 21:1554-1566, 1998. 13. Campbell WW, Pridgeon RM, Sahni KS: Short segment incremental studies in the evaluation of ulnar neuropathy at the elbow. Muscle Nerve 15:1050-1054, 1992. 14. Cannieu JMA: Note sur une anastomose entre la branche profonde du cubital et le median. Bull Soc D'Anat Physiol Bordeaux 18:339-340, 1897.
210 15. Cavanagh NPC, Yates DAH, Sutcliffe J: Thenar hypoplasia with associated radiologic abnormalities. Muscle and Nerve 2:431-436, 1979. 16. Chaudhry V, Cornblath DR, Mellits ED, Avila O, Freimer ML, Glass JD: Inter- and intra-examiner reliability of nerve conduction measurements in normal subjects. Ann Neurol 30: 841-843, 1991. 17. Chaudhry V, Corse AM, Freimer ML, Glass JD, Mellits ED, Kuncl RW: Inter- and intraexaminer reliability of nerve conduction measurements in patients with diabetic neuropathy. Neurology 44:1459-1462, 1994. 18. Claus D, Mustafa C, Vogel W, Herz M, Neundorfer B: Assessment of diabetic neuropathy definition of norm and discrimination of abnormal nerve function. Muscle Nerve 16:757-768, 1993. 19. Claussen GC, Ahmad BK, Sunwood IN, Oh SJ: Combined motor and sensory median-ulnar anastomosis: report of an electrophysiologically proven case. Muscle Nerve 19:231-233, 1996. 20. Cornblath DR, Sumner AJ, Daube J, Gilliat RW, Brown WF, Parry GJ, Albers JW, Miller RG, Petajan J: Conduction block in clinical practice. Muscle Nerve 14:869-871, 1991. 21. Crutchfleld CA, Gutmann L: Hereditary aspects of accessory deep peroneal nerve. J Neurol Neurosurg Psychiatry 36:989-990, 1973. 22. Cummins KL, Dorfman LJ, Perkel DH: Nerve fiber conduction—velocity distributions. II. Estimation based on two compound action potentials. Electroencephalogr Clin Neurophysiol 46:647-658, 1979. 23. Cummins KL, Perkel DH, Dorfman LJ: Nerve fiber conduction-velocity distributions. I. Estimation based on the single-fiber and compound action potentials. Electroencephalogr Clin Neurophysiol 46:634-646, 1979. 24. Daube JR: Percutaneous palmar median nerve stimulation for carpal tunnel syndrome. Electroencephalogr Clin Neurophysiol 43:139-140, 1977. 25. Daube JR: Nerve Conduction Studies. Churchill Livingstone, New York, 1980, pp 229-264. 26. Del Toro DR, Park TA: Abductor hallucis false motor points: electrophysiologic mapping and cadaveric dissection. Muscle Nerve 19:11381143, 1996. 27. Dengler R, Stein RB, Thomas CK: Axonal conduction velocity and force of single human motor units. Muscle Nerve 11:136-145, 1988. 28. Dessi F, Durand G, Hoffmann JJ: The accessory deep peroneal nerve: A pitfall for the electromyographer. J Neurol Neurosurg Psychiatry 55:214-215, 1992. 29. Dominque J, Shahani BT, Young RR: Conduction velocity in different diameter ulnar sensory and motor nerve fibers. Electroencephalogr Clin Neurophysiol 50:239P-245P, 1980. 30. Dorfman LJ: The distribution of conduction velocities (DCV) in peripheral nerves: A review. Muscle & Nerve 7:2-11, 1984. 31. Dorfman LJ, Cummins KL, Reaven GM, Ceranski J, Greenfield MS, Dobeme L: Studies of diabetic polyneuropathy using conduction velocity distribution (DCV) analysis. Neurology 33:773-779, 1983.
Nerve Conduction Studies 32. Dumitru D, Walsh NE, Weber CF: Electrophysiologic study of the riche-cannieu anomaly. Electromyogr Clin Neurophysiol 28:27-31, 1988. 33. Dyck PJ, Kratz KM, Lehman KA, Karnes JL, Melton LJ, O'Brien PC: The Rochester Diabetic Neuropathy Study: design, criteria for types of 1991. 34. Feasby TE, Brown WF, Gilbert JJ, Hahn AFD: The pathological basis of conduction block in human neuropathies. J Neurol Neurosurg Psychiatry 48:239-244, 1985. 35. Felsenthal G, Teng CS: Changes in duration and amplitude of the evoked muscle action potential (EMAP) over distance in peroneal, median, and ulnar nerves. Am J Phys Med 62: 123-134, 1983. 36. Fisher MA: Whither F waves. In Kimura, J and Shibasaki, H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science BV, 1996, pp 752-755. 37. Fraser JL, Olney RK: The relative diagnostic sensitivity of different F-wave parameters in various polyneuropathies. Muscle Nerve 15:912-918, 1992. 38. Ganes T: Complete ulnar innervation of the thenar muscles combined with normal sensory fibres in a subject with no peripheral nerve lesion. Electromyogr Clin Neurophysiol 32:559563, 1992. 39. Gassel MM: Sources of error in motor nerve conduction studies. Neurology (Minneap) 14: 825-835, 1964. 40. Geiringer SR: Inching techniques are of limited use. Muscle Nerve 21:1557-1559, 1998. 41. Gilliatt RW: Acute Compression Block: The Physiology of Peripheral Nerve Disease. WB Saunders, Philadelphia, 1980, pp 287-315. 42. Gilliatt RW: Electrophysiology of peripheral neuropathies: An overview. Muscle Nerve 5: S108-S116, 1982. 43. Gilliatt RW, Hopf HC, Rudge P, Baraitser M: Axonal velocities of motor units in the hand and foot muscles of the baboon. J Neurol Sci 29:249-258, 1976. 44. Golovchinsky V: Ulnar-to-median anastomosis and its role in the diagnosis of lesions of the median nerve at the elbow and the wrist. Electromyogr Clin Neurophysiol 30:31-34, 1990. 45. Gruber W: Ueber die Verbindung des Nervus medianus mit dem Nervus ulnaris am Unterarme des Menschen und der Saugethiere. Arch Anat Physiol Med, Leipzig, 1870, pp 501-522. 46. Gutmann L: Atypical deep peroneal neuropathy in presence of accessory deep peroneal nerve. J Neurol Neurosurg Psychiatry 33:453-456, 1970. 47. Gutmann L: Median-ulnar nerve communications and carpal tunnel syndrome. J Neurol Neurosurg Psychiatry 40:982-986, 1977. 48. Gutmann L: AAEM Minimonograph #2: Important anomalous innervations of the extremities. Muscle Nerve 16:339-347, 1993. 49. Hakamada S, Kumagai T, Watanabe K, Koike Y, Hara K, Miyazaki S: The conduction velocity of slower and the fastest fibres in infancy and childhood. J Neurol Neurosurg Psychiatry 45:851-853, 1982.
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques 50. Harayama H, Shinozawa K, Kondo H, Miyatake T: A new method to measure the distribution of motor conduction velocity in man. Electroencephalogr Clin Neurophysiol 81:323-331, 1991. 51. Harrison MJG: Pressure palsy of the ulnar nerve with prolonged conduction block. J Neurol Neurosurg Psychiatry 39:96-99, 1976. 52. Honet J, Jebson R, Perrin E: Variability of nerve conduction velocity determinations in normal persons. Arch Phys Med Rehabil 49:650-654, 1968. 53. Hopf HC: Untersuchungen uber die Unterschiede in der leitgeschwindigkeit motorischer Nervenfasern beim Menschen. Deutsche Zeitschrift Fur Nervenheilkunde 183:579-588, 1962. 54. Homberg V, Reiners K, Toyka KV: Reversible conduction block in human ischemic neuropathy after ergotamine abuse. Muscle Nerve 15:467-470, 1992. 55. Hopf HC: Forearm ulnar-to median nerve anastomosis of sensory axons. Muscle Nerve 13:654-656, 1990. 56. Hopf HC, Lowitzsch K: Relative refractury periods of motor nerve fibers. In Kunzesk and Desmedt (eds): Studies on Neuromuscular Diseases. Proceedings of the International Symposium (Giessen), Karger, Basel, 1975, pp 264267. 57. Imaoka H, Yorifuji S, Takahashi M, Nakamura Y, Kitaguchi M, Tarui S: Improved inching method for the diagnosis and prognosis of carpal tunnel syndrome. Muscle Nerve 15:318-324, 1992. 58. Infante E, Kennedy WR: Anomalous branch of the peroneal nerve detected by electromyography. Arch Neurol 22:162-165, 1970. 59. Ingram DA, Davis GR, Swash M: The double collision technique: a new method for measurement of the motor nerve refractory period distribution in man. Electroencephalogr Clin Neurophysiol 66:225-234, 1987. 60. Ingram DA, Davis GR, Swash M: Motor conduction velocity distributions in man: Results of a new computer-based collision technique. Electroencephalogr Clin Neurophysiol 66:235243, 1987. 61. Isch F, Isch-Treussard C, Buchheit F, Delgado V, Kirchner JP: Measurement of conduction velocity of motor nerve fibres in polyneuritis and polyradiculoneuritis (abstr). Electroencephalogr Clin Neurophysiol 16:416, 1964. 62. Iyer V, Fenichel GM: Normal median nerve proximal latency in carpal tunnel syndrome: A clue to coexisting Martin-Gruber anastomosis. J Neurol Neurosurg Psychiatry 39:449-452, 1976. 63. Jamieson PW, Giuliani MJ, Martinez AJ: Necrotizing angiopathy presenting with multifocal conduction blocks. Neurology 41:442-444, 1991. 64. Johnson RK, Shrewsbury MM: Anatomical course of the thenar branch of the median nerve—Usually in a separate tunnel through the transverse carpal ligament. J Bone Joint Surg 52A:269-273, 1970. 65. Kadrie H, Yates SK, Milner-Brown HS, Brown
66.
67.
68.
69.
70.
71. 72.
73. 74. 75. 76. 77.
78.
79.
80.
81.
82.
83.
211
WF: Multiple point electrical stimulation of ulnar and median nerves. J Neurol Neurosurg Psychiatry 39:973-985, 1976. Kaji R, Oka N, Tsuji S, Mezaki T, Nishio T, Akiguchi I, Kimura J: Pathological findings at the site of conduction block in multifocal motor neuropathy. Ann Neurol 33:152-158, 1993. Kanakamedala RV, Hong C-Z: Peroneal nerve entrapment at the knee localized by short segment stimulation. Am J Phys Med Rehabil 68:116-122, 1989. Kanakamedala RV, Simons DG, Porter RW, Zucker RS: Ulnar nerve entrapment at the elbow localized by short segment stimulation. Arch Phys Med Rehabil 69:959-963, 1988. Kimura J: F-wave velocity in the central segment of the median and ulnar nerves: A study in normal subjects and in patients with Charcot-Marie-Tooth disease. Neurology (Minneap) 24:539-546, 1974. Kimura J: Collision technique. Physiologic block of nerve impulses in studies of motor nerve conduction velocity. Neurology (Minneap) 26:680-682, 1976. Kimura J: Electrical activity in voluntarily contracting muscle. Arch Neurol 34:85-88, 1977. Kimura J: The carpal tunnel syndrome: Localization of conduction abnormalities within the distal segment of the median nerve. Brain 102:619-635, 1979. Kimura J: Electrodiangosis in Diseases of Nerve and Muscle: Prinicples and Practices. FA Davis, Philadelphia, 1983. Kimura J: Principles and pitfalls of nerve conduction studies. Ann Neurol 16:415-429, 1984. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle, Principles and Practice, Edition 2. Philadelphia: F.A. Davis, 1989. Kimura J: Electromyography and Nerve Stimulation Techniques: Clinical Applications (Japanese). Igakushoin, Tokyo, 1990. Kimura J: Facts, fallacies, and fancies of nerve conduction studies: Twenty-First annual Edward H. Lambert lecture. Muscle Nerve 20: 777-787, 1997. Kimura J: Kugelberg lecture: Principles and pitfalls of nerve conduction studies. Electroencephalogr Clin Neurophysiol, 106:470-476, 1998. Kimura J, Machida M, Ishida T, Yamada T, Rodnitzky RL, Kudo Y, Suzuki S: Relation between size of compound sensory or muscle action potentials and length of nerve segment. Neurology 36:647-652, 1986. Kimura J, Mitsudome A, Beck DO, Yamada T, Dickins QS: Field distribution of antidromically activated digital nerve potentials: model for farfield recording. Neurology 33:1164-1169, 1983. Kimura J, Mitsudome A, Yamada T, S DQ: Stationary peaks from a moving source in far-field recording. Electroencephalogr Clin Neurophysiol 58:351-361, 1984. Kimura J, Murphy JM, Varda DJ: Electrophysiological study of anomalous innervation of intrinsic hand muscles. Arch Neurol 33: 842-844, 1976. Kimura J, Sakimura Y, Machida M, Fuchigami
Nerve Conduction Studies
212 Y, Ishlda T, Claus D, Kameyama S, Nakazuml Y, Wang J, Yamada T: Effect of desynchronized inputs on compound sensory and muscle action potentials. Muscle Nerve 11:694-702, 1988. 84. Kimura J, Yamada T, Stevland NP: Distal slowing of motor nerve conduction velocity in diabetic polyneuropathy. J Neurol Sci 42:291-302, 1979. 85. Kincaid JC, Minnick KA, Pappas S: A Model of the differing change in motor and sensory action potentials over distance. Muscle Nerve 11:318-323, 1988. 86. Kingery WS, Wu PBJ, Date ES: An unusual presentation of a traumatic ulnar mononeuropathy with a Martin-Gruber anastomosis. Muscle Nerve 19:920-922, 1996. 87. Kohara N, Kimura J, Kaji R, Goto Y, Ishii J: Multicenter analysis on intertrial variability of nerve conduction studies: healthy subjects and patients with diabetic polyneuropathy. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology, Elsevier Science BV, 1996, pp 809-815. 88. Lafontaines S, Rasminsky M, Saida T, Sumner AJ: Conduction block in rat myelinated fibers following acute exposure to anti-galactocerebroside serum. J Physiol 323:287-306, 1982. 89. Lambert EH: Diagnostic value of electrical stimulation of motor nerves. Electroencephalogr Clin Neurophysiol (suppl 22):9-16, 1962. 90. Lambert EH: The accessory deep peroneal nerve. A common variation in innervation of extensor digitorum brevis. Neurology (Minneap) 19:1169-1176, 1969. 91. Lateva ZC, McGill KC, Burgar CG: Anatomical and electrophysiological determinants of the human thenar compound muscle action potential. Muscle Nerve 19:1457-1468, 1996. 92. Lee R, Ashby P, White D, Aguayo A: Analysis of motor conduction velocity in human median nerve by computer simulation of compound action potentials. Electroencephalogr Clin Neurophysiol 39:225-237, 1975. 93. Leibovic SJ, Hastings H: Martin-Gruber revisited. J Hand Surg 17A:47-53, 1992. 94. Leifer LJ: Nerve-Fiber Conduction Velocity Distributions of the Human Median Nerve: Comparison of Methods. Alan R. Liss. 1981, pp 233-263. 95. Lewis RA, Sumner AJ, Brown MJ, Asbury AK: Multifocal demyelinating neuropathy with persistent conduction block. Neurology (NY) 32:958-964, 1982. 96. Liguori R, Trojaborg W: Are there motor fibers in the sural nerve? Muscle Nerve 13:12-15, 1990. 97. Linden D, Berlit P: The intrinsic foot muscles are purely innervated by the tibial nerve ("all tibial foot")—an unusual innervation anomaly. Muscle Nerve 17:560-561, 1994. 98. LoMonaco M, Pauda L, Gregori B, Valente EM, Tonali P: Ulnar innervation of the thenar eminence with preservation of median innervation of first lumbrical muscle. Muscle Nerve 629, 1997. 99. Magistris MR, Truffert A: Extensor digitorum brevis innervated by the tibial ("all tibial foot"):
100.
101. 102.
103. 104. 105.
106. 107. 108.
109. 110.
111. 112.
113.
114.
115.
116.
Anomalous innervation or technical pitfall?" Muscle Nerve 20:906-908, 1997. Mannerfelt L: Studies on the hand in ulnar nerve paralysis: A clinical-experimental investigation in normal and anomalous innervation. Acta Orthopaedica Scand (suppl 87):23-176, 1966. Marinacci AA: Diagnosis of "all median hand." Bull LA Neurol Soc 29:191-197, 1964. Marras C, Midroni GYL: Proximal Martin-Gruber anastomosis mimicking ulnar neuropathy at the elbow. Muscle Nerve 22:1132-1235, 1999. Martin R: Tal om Nervers Allmanna Egenskapper i Manniskans Kropp. L Salvius, Stockholm, 1763. Maynard FM, Stolov WC: Experimental error in determination of nerve conduction velocity. Arch Phys Med Rehabil 53:362-372, 1972. McCluskey LF: Anomalous superficial radial sensory innervation of the ulnar dorsum of the hand: a cause of "paradoxical" preservation of ulnar sensory function. Muscle Nerve 19:923-925, 1996. McCluskey F, Feinberg D, Cantor C, Bird S: "Pseudo-conduction block" in vasculitic neuropathy. Muscle Nerve 22:1361-1366, 1999. McManis PG, Daube JR: Electromyographic evaluation of an accessory hand muscle. Muscle Nerve 12:460-463, 1989. Meulstee J, Darbas A, van Doom PA, van Briemen L, van der Meche FGA: Decreased electrical excitability of peripheral nerves in demyelinating polyneuropathies. J Neurol Neurosurg Psychiatry 62:398-400, 1997. Miller RG, Olney RK: Persistent conduction block in compression neuropathy. Muscle Nerve 5:S154-S156, 1982. Milner TE, Stein RV, Gillespie J, Hartley B: Improved estimates of conduction velocity distributions using single unit action potentials. J Neurol Neurosurg Psychiatry 44:476-484, 1981. Neundorfer B, Seiberth R: The accessory deep peroneal nerve. J Neurol 209:125-129, 1975. Nix WA, Liider G, Hopf HC, Luth G: A computerized re-evaluation of the collision technique. Electromyogr Clin Neurophysiol 29: 391-397, 1989. Oh SJ, Claussen GC, Ahmad BK: Double anastomosis of median-ulnar and ulnar-median nerves: Report of an electrophysiologically proven case. Muscle Nerve 18:1332-1334, 1995. Oh SJ, Kim DE, Kuruoglu HR: What is the best diagnostic index of conduction block and temporal dispension? Muscle Nerve 17:489-493, 1994. Ohmi Y, Harata S, Ueyama K, Okamura Y, Sasaki H, Iwaya D: Level diagnosis using spinal cord evoked potentials in cervical myelopathy. In Shimoji K, Kurokawa T, Tamaki T and Wills Jr WD (eds): Spinal Cord Monitoring and Electrodiagnosis. Springer, Berlin, 1991, pp 454— 460. Olney RK, Budingen HJ, Miller RG: The effect of temporal dispersion on compound action potential area in human peripheral nerve. Muscle Nerve 10:728-733, 1987.
Facts, Fallacies, and Fancies of Nerve Stimulation Techniques 117. Olney RK, Miller RG: Pseudo-conduction block in normal nerves. Muscle Nerve 6:530, 1983. 118. Olney RK, Miller RG: Conduction block in compression neuropathy: Recognition and quantification. Muscle Nerve 7:662-667, 1984. 119. Phillips II LH, Morgan RF: Anomalous origin of the sural nerve in a patient with tibial-common perioneal nerve anastamosis. Muscle Nerve 16:414-417, 1993. 120. Poloni AE, Sala E: The conduction velocity of the ulnar and median nerves stimulated through a twin-needle electrode. Electroencephalogr Clin Neurophysiol (Suppl 22): 17-19, 1962. 121. Rhee EK, England JD, Sumner AJ: A computer simulation of conduction block: Effects produced by actual block versus interphase cancellation. Ann Neurol 28:146-156, 1990. 122. Richie P: Le nerf cubital et les muscles de 1'eminence thenar. Bull Mem Soc Anat Paris 72:251-252, (Series 5), 1897. 123. Rogers MR, Bergfield TG, Aulicino PL: A Neural loop of the deep motor branch of the ulnar nerve: An anatomic study. J Hand Surg 16A: 269-271, 1991. 124. Ross MA, Kimura J: AAEM case report #2. The carpal tunnel syndrome. Muscle Nerve 18:567573, 1995. 125. Rossi B, Sartucci F, Stefanini A: Measurement of motor conduction velocity with Hopf s technique in the diagnosis of mild peripheral neuropathies. J Neurol Neurosurg Psychiatry 44:168-170, 1981. 126. Roth G: Vitesse de conduction motrice du nerf median dans le cana carpien. Ann Med Physique 13:117-132, 1970. 127. Roth G, Magistris MR: Identification of motor conduction block despite desyncheronisation: A method. Electromyogr Clin Neurophysiol 29:305-313, 1989. 128. Russomano S, Herbison GJ, Baliga A, Jacobs SR, Moore J: Riche-cannieu anastomosis with partial transection of the median nerve. Muscle Nerve 18:120-122, 1995. 129. Rutten GJM, Gaasbeck RDA, Franssen H: Decrease in nerve temperature: a model for increased temporal dispersion. Electroencephalogr Clin Neurophysiol 109:15-23, 1998. 130. Sachs GM, Raynor EM, Shefner JM: The all ulnar motor hand without forearm anastomosis. Muscle Nerve 18:309-313, 1995. 131. Salerno DF, Werner RA, Albers JW, Becker MP, Armstrong TJ, Franzblau A: Reliability of nerve conduction studies among active workers. Muscle Nerve 22:1372-1379, 1999. 132. Sander HW, Quinto C, Chokroverty S: Medianulnar anastomosis to thenar, hypothenar, and first dorsal interosseous muscles: collision technique confirmation. Muscle Nerve 20: 1460-1462, 1997. 133. Sander HW, Quinto C, Chokroverty S: Accessory deep peroneal neuropathy: collision technique diagnosis. Muscle Nerve 21:121-123, 1998. 134. Sedal L, Ghabriel MN, He F, Allt G, Le Quesne PM, Harrison MJG: A combined morphological and electrophysiological study of conduction
213
block in peripheral nerve. J Neurol Sci 60: 293-306, 1983. 135. Seddon H: Surgical Disorders of the Peripheral Nerves, ed 2. Churchill Livingstone, Edinburgh, 1975, pp 203-211. 136. Seradge H, Seradge E: Median innervated hypothenar muscle: Anomalous branch of median nerve in the carpal tunnel. J Hand Surg 15A:356-359, 1990. 137. Seror P: Orthodromic inching test in mild carpal tunnel syndrome. Muscle Nerve 21: 1206-1208, 1998. 138. Shefner JM, Dawson DM: The use of sensory action potentials in the diagnosis of peripheral nerve disease. Arch Neurol 47:341-348, 1990. 139. Shimqji K, Higashi H, Kano T: Epidural recording of spinal electrogram in man. Electroencephalogr Clin Neurophysiol 30:236-239, 1971. 140. Simpson JA: Fact and fallacy in measurement of conduction velocity in motor nerves. J Neurol Neurosurg Psychiatry 27:381-385, 1964. 141. Skorpil V: Conduction velocity of human nerve structures. Roszpr Cesk Akad Ved 75:1-103, 1965. 142. Spillane K, Nagendran K, Kunzru KM: Anomalous communication in anterior tarsal tunnel syndrome. Muscle Nerve 20:395-396, 1997. 143. Srinivasan R, Rhodes J: The median-ulnar anastomosis (Martin-Gruber) in normal and congenitally abnormal fetuses. Arch Neurol 38:418-419, 1981. 144. Stegeman DF, De Weerd JPC, Notermans SLH: Modelling compound action potentials of peripheral nerves in situ. III. Nerve propagation in the refractory period. Electroencephalogr Clin Neurophysiol 55:668-679, 1983. 145. Sumner AJ: The physiological basis for symptoms in Guillain-Barre syndrome. Ann Neurol (suppl) 9:28-30, 1981. 146. Sumner AJ: Consensus criteria for the diagnosis of partial conduction block and multifocal motor neuropathy. In Kimura J, Kaji R (eds): Physiology of ALS and Related Diseases, Elsevier Science BV, 1997, pp 221. 147. Sunderland S: Nerves and Nerve injuries, ed 2. Churchill Livingstone, Edinburgh, 1978. 148. Tamaki T, Tsuji H, Inoue S, Kobayashi H: The prevention of iatrogenic spinal cord injury utilizing the evoked spinal cord potential. Int Orthop 4:313-317, 1981. 149. Tani T, Ushida T, Yamamoto H, Okuhara Y: Waveform changes due to conduction block and their underlying mechanism in spinal somatosensory evoked potential: a computer simulation. J Neurosurg 86:303-310, 1998. 150. TaniT, Yamamoto H, Kimura J: Cervical spondylotic myelopathy in elderly people: A high incidence of conduction block at C3-4 or C4-5. J Neuro Neurosurg Psychiatry 66:456-464, 1999. 151. Thomas PK: The morphological basis for alterations in nerve conduction in peripheral neuropathy. Proc R Soc Med 645:295-298, 1971. 152. Thomas PK, Sears TA, Gilliatt RW: The range of conduction velocity in normal motor nerve fibres to the small muscles of the hand and foot. J Neurol Neurosurg Psychiatry 22:175181, 1959.
214 153. Tjon-A-Tsien AML, Lemkes HHPJ, Callenbach PMC, Van Dijk JG: CMAP variation over a length of nerve in diabetic neuropathy. Muscle Nerve 18:907-909, 1995. 154. Tjon-A-Tsien AML, Lemkes HHPJ, van der Kamp-Huyts AJC, van Dijk JG: Large electrodes improve nerve conduction repeatability in controls as well as in patients with diabetic neuropathy. Muscle Nerve 19:689-695, 1996. 155. Triggs WJ, Cros D, Gominack SC, Zuniga A, Beric A, Shahani BT, Ropper AH, Roongta SM: Motor nerve inexcitability in Guillain-Barre syndrome: The spectrum of distal conduction block and axonal degeneration. Brain 115: 1291-1302, 1992. 156. Tsuyama N, Tsuzuki N, KurokawaT, Imai T: Clinical application of spinal cord action potential measurement. Int Orthop 2:39-46, 1978. 157. Uchida Y, Sugioka Y: Electrodiagnosis of Martin-Gruber connection and its clinical importance in peripheral nerve surgery. J Hand Surg 17A:54-59, 1992. 158. Uncini A, Di Muzio A, Sabatelli M, Magi S, Tonali P, Gambi D: Sensitivity and specificity of diagnostic criteria for conduction block in chronic inflammatory demyeinating polyneuropathy. 89:161-169. Electroencephalogr Clin Neurophysiol 89:161-169, 1993. 159. Valensi P, Attali J-R, Gagant S: Reproducibility of parameters for assessment of diabetic neuropathy. Diabet Med 10:933-939, 1993. 160. Valls-Sole J: Martin-Gruber anastomosis and unusual sensory innervation of the fingers: Report of a case. Muscle Nerve 14:1099-1102, 1991.
Nerve Conduction Studies 161. Van Dijk JG, Bouma PAD: Recording of the Martin-Gruber anastomosis. Muscle Nerve 20:887-889, 1997. 162. Van Dijk JG, Van der Hoeven BJ: Compound muscle action potential cartography of an accessory peroneal nerve. Muscle Nerve 21:13311333, 1998. 163. van Veen BK, Schellens RLLA, Stegeman DF, Schoonhoven R, Gabreels-Festen AAWM: Conduction velocity distributions compared to fiber size distributions in normal human sural nerve. Muscle Nerve 18:1121-1127, 1995. 164. Wiechers D, Fatehi M: Changes in the evoked potential area of normal median nerves with computer tehniques. Muscle Nerve 6:532, 1983. 165. Wiederholt WC: Median nerve conduction velocity in sensory fibers through carpal tunnel. Arch Phys Med Rehabil 51:328-330, 1970. 166. Wilbourn AJ, Lambert EH: The forearm median-to-ulnar nerve communication: Electrodiagnostic aspects. Neurology (Minneap) 26:368, 1976. 167. Winer BJ: Statistical Principles in Experimental Design. 2nd ed. New York: McGraw-Hill, 1972, pp 283-289. 168. Xiang X, Eisen A, MacNeil M, Beddoes MP: Quality control in nerve conduction studies with coupled knowledge-based system approach. Muscle Nerve 15:180-187, 1992. 169. Yokohama K, Feldman RG, Sax DS, Salzsider BT, Kucera J: Relation of distribution of conduction velocities to nerve biopsy findings in n-hexane poisoning. Muscle Nerve 13:314-320, 1990.
Chapter
8
OTHER TECHNIQUES TO ASSESS NERVE FUNCTION
1. MOTOR UNIT NUMBER ESTIMATES Compound Muscle Action Potential Sampling of Single Motor Unit Potential Methods for Quantitative Assessments Normal Values and Clinical Application 2. ASSESSMENT OF NERVE EXCITABILITY Refractory Period Paired Shock and Collision Technique Changes in Amplitude versus Latency Excitability Changes after Passage of an Impulse 3. THRESHOLD TRACKING Strength-Duration Curve Threshold Measurement of Strength-Duration Time Constant Latent Addition and Accommodation Electrotonus and Threshold Electrotonus Techniques to Measure Threshold Electrotonus Applications of Threshold Measurements Clinical Assessments
1 MOTOR UNIT NUMBER ESTIMATES In neuromuscular disorders characterized by a loss of lower motor neurons, a patient's strength depends primarily on the number of remaining motor units in a group of muscles. A variety of techniques provide the means for calculating motor unit number estimates (MUNE).31,104,132 Each method relies on dividing an average size of a single motor unit potential into a maximal compound muscle action potential that represents the sum of all motor units. All the methods have certain as-
sumptions relating to the adequacy of sampling in estimating average size. Technical limitation in achieving unbiased selection constitutes a major source of error.
Compound Muscle Action Potential The amplitude of a maximal compound muscle response directly relates to the total number and size of muscle fibers, providing a rough estimate,73 although phase cancellation may distort the pattern of summation.4,54 Supramaximal stimulation of a peripheral nerve activates all the 215
Nerve Conduction Studies
216
muscles innervated by that nerve, eliciting a muscle response as a sum of activity from multiple sources rather than from a single source. For example, a thenar response elicited by stimulation of the median nerve represents the activity of all the intrinsic hand muscles innervated by this nerve. In a strict sense, therefore, the motor unit number estimate relates to a group of muscles that contribute to the measure to a greater or lesser extent, depending on their spatial relationship to the recording electrodes. Another technical concern centers on the intensity of stimulation that must activate all the excitable motor axons. During the process of demyelination or regeneration an ordinarily adequate stimulus may fail for the nerve with an abnormally elevated threshold. The use of submaximal stimulation would underestimate the motor units number. Although the maximal amplitude is usually proportional to the number of axons, abnormally large motor unit potentials after reinnervation partially restore the size, thus concealing the loss of axons. In fact, despite the loss of over one half of its motor innervation, a muscle may maintain its normal amplitude. Therefore, a reliable motor unit number estimate requires the knowledge about the average size of individual motor unit potentials in addition to a measure of the total response. The accuracy of the estimated number depends, among other factors, on the adequacy in sampling the representative population of single unit size, which varies considerably in normal subjects55 and, to a much greater extent, in patients with neuromuscular diseases.
Sampling of Single Motor Unit Potential A severe neurogenic process may reduce the number of axons to a level that allows identification of all the existing motor units individually. In general, direct counts provide a reliable, reproducible result up to a maximum of 10 units. With a greater number of units, an overlap precludes a complete count of all the units, necessitating the selection of a subset for
calculation of an average size of single motor unit potentials. If all the motor units in the muscle give rise to nearly identical potential, then sampling a subset constitutes a valid approach. Variation among different motor units causes sampling error, especially with a non-Gaussian distribution. Thus, sampling a greater population128leads to more reproducible results. In chronic neurogenic processes, the ease of measuring the reduced number of larger potentials compensates for the inaccuracy resulting from an increased size variation of individual units. The same motor unit potential may vary in size from one stimulus to the next, with defects of neuromuscular transmission requiring special interpretation. These include myasthenia gravis, amyotrophic lateral sclerosis, and neurogenic processes with ongoing reinnervation. With a decrement of compound muscle action potentials on slow repetitive stimulation, for example, the calculated value falls short of the actual number of motor units.
Methods for Quantitative Assessments The methods described for obtaining single motor unit action potential values include (1) all-or-none increments of compound muscle action potential, (2) F-wave measurements, (3) spike-triggered averaging, and (4) statistical estimates. Of these, spike-triggered averaging relies on voluntary recruitment whereas the remaining three measures use nerve stimulation to record individual elements.52,104,133 Different methods place varying technical emphasis on meeting the underlying assumptions mentioned above, although basic principles remain the same. All these techniques, when properly executed, yield the same order of estimates.32 The original incremental method103,104 provides the easiest and most direct approach to counting of single motor unit potentials. Based on the all-or-none characteristic of the activation of motor axons, application of finely controlled current in very small steps allows measurement of successively recruited individual motor units. The maximal muscle action poten-
Other Techniques to Assess Nerve Function tial derived by the average size of the stepwise increments yields the estimated number of motor units. In incremental methods, a selection bias for more easily activated larger motor units could result in an overestimation of the size of individual elements and consequently underestimation of motor unit numbers. This technique may also fail to identify the increments by very small motor unit potentials, such as nascent units or those seen in severe myopathies. Several modifications introduced to minimize these errors tend to favor the low-threshold units, with a selection9,49,51,60,61 bias against the highthreshold units. As a variation, stimulating the nerve at several points with very low intensity yields only the first recorded single motor unit potentials.51,52,56,78,144 The average sizes of the units obtained with stimulation along the nerve divided into the maximal compound muscle action potential yields the number of motor units. The firing threshold for an individual axon varies in time. Thus, at any given stimulus intensity, different axons may discharge according to their probability of firing. If two motor axons have similar excitability, a threshold stimulus may activate them together or alternately. This possibility, termed alternation, constitutes another source of error. In the presence of two units, for example, three distinct potentials would be recognized, one each or both together, giving the count of three rather than two. Similarly, in the case of three motor units, alternation could result in an erroneous count of seven instead of the actual number of three. As mentioned later, the stochastic approach127 avoids such an error by using cluster analysis to sort out the templates of the individual elements from all potentials recorded at a fixed intensity. The F-wave method relies on the assumption that repeater F waves represent single rather than multiple motor units. If so, dividing the maximal muscle response by their average size yields the number of motor units.90,131 Alternation can occur as described above. The mistaken inclusion of F waves activated by multiple instead of single motor units inflates the average size of individual elements, lowering
217
the estimated number.43 Automated use of submaximal stimulation and template matching reduce the risk and improve the accuracy. Spike-triggered averaging uses a twochannel recorder to isolate voluntarily activated motor units as a measure of single motor unit potentials.33,44 The technique consists of detecting single units by a needle electrode on the first channel, and averaging its size using a pair of surface electrodes on the second channel. An amplitude trigger window selects the units recorded by single-fiber, bipolar concentric, standard concentric, or fine-wire electrodes. Their average size divided into the maximal muscle potential recorded from the same surface electrode yields the number of motor units. As a variant, motor units recruited at three levels of effort and recorded at two locations on the surface provide a broader sampling.125 The sources of error unique to this method include recording29with a spurious and erroneous trigger and missing some motor units at the surface, unless studying the muscle located superficially.10 Further, voluntary activation preferentially recruits smaller motor units, without recruiting larger units. Despite these concerns, the method provides values comparable to those expected from histologic studies and those obtained with other methods of recording. Microstimulation of nerve terminals in the endplate region may activate the full range of motor units, thus reducing the selection bias characteristic of voluntary activation.107 In contrast to all the other methods, the statistical approach makes no attempt to identify individual motor units. Instead, it takes advantage of intermittent firing of individual motor units near threshold that results in variation in the size of a submaximal compound muscle action potential.45,127 It relies on Poisson statistics to calculate the size of the individual steps based on their known relationship to the variance of multiple measures of step functions. In this type of analysis, the sizes of a series of measurements are multiples of the size of a single component and the variance of their distribution provides an estimate of the average size of the individual components making up
Nerve Conduction Studies
218
each measurement. Obtaining adequate estimates of motor units calls for testing the axons with different thresholds at multiple stimulus intensities. In the interest of brevity, an initial scan of the compound muscle response identifies large steps with a series of 30 stimuli increasing in equal increments. The scan thereby defines appropriate stimulus intensity levels to recognize the representative single units for the particular nerve under study. In neurogenic disorders, for example, the axons with large motor unit potentials may have a higher threshold than the axons of smaller potentials, necessitating stimulation at higher intensities. The statistical method has the advantage of not requiring identification of individual components producing increments too small to isolate at gains used to record high-amplitude compound muscle action potentials. It also circumvents the possible miscalculation caused by alternation with activation of the same units in different combinations. The technical problems include the need for a larger sample size, requiring patient cooperation to undergo over 100 low-intensity stimuli. The remaining motor units not tested at the stimulus intensities used are assigned a motor unit size estimate made at any stimulus strength. Thus, this stimulus strength influences the final result excessively.126 Defective neuromuscular transmission also causes inaccuracies in this measurement, from varying sizes of motor unit potentials. A shift from Poisson to normal distributions can produce errors of up to 10 percent, necessitating a display of the histogram of the individual responses.
Normal Values and Clinical Application Normal values, though they vary among authors using different techniques,45,52 range from 200 to 350 for the thenar muscles tested with stimulation of the median nerve and from 150 to 220 for the extensor digitorum brevis tested with stimulation of the peroneal nerve. According to histological estimation, the flexor digit minimi has about 130 motor units.111 This result is in agreement with 411 motor units estimated for the four hypothenar muscles 61 by an automated incremental method. Few studies report on proximal muscles because of the technical difficulty. The number of motor units remains stable for a given muscle except for a mild decrease in the elderly.30,60 Table 8-1 summarizes normal MUNE values obtained by the statistical method for distal muscles innervated by median, ulnar, and tibial nerves tested at different stimulus intensities.45 Earlier clinical studies used near-threshold methods,9,52,103 which are best suited to test a muscle with a reduced number of motor units, allowing individual recognition of each unit with successive increments of stimulus current. In addition, reproducibility improves in absolute values with a smaller number of motor units in the muscle. In contrast, the method tends to underestimate the number of motor units in myopathies, which render some of the increments too small to identify. A 20 percent accuracy gives estimates in the range of 16-24 for 20 motor units and 160-240 for 200. Thus, a larger number of units makes a small loss harder to detect. Stim-
Table 8-1 Statistical MUNE in 30 Normal Subjects Stimulus Level 5-10% 15-20% 40-50% 70-90% Multipoint
Median Thenar 210/90 185/85 153/70 175/85 234/95
Ulnar Hypothenar 285/105 223/110 154/70 213/115 256/115
Peroneal EDB 154/52 137/45 135/38 105/35 158/58
Tibial Abductor Hallucis 310/195 250/167 195/154 202/115 285/187
Statistical motor unit number estimate (MUNE) in 30 normal subjects tested at different stimulus intensities. The mean and lower limit (XX/YY) is shown for each stimulus level for each nerve. Multipoint recordings measured MUNE at45 5-10 percent, and at 15-20 percent at two distal sites 1 cm apart along the nerve. Source: From Daube, with permission.
219
Other Techniques to Assess Nerve Function ulus currents above 15 percent of threshold also yield unreliable results, even in normal subjects. The technique supplements conventional studies in documenting the loss of motor units in patients with a normal compound muscle action potential amplitude. These include congenital brachial palsy,123 tetraplegia64 and amyotrophic lateral sclerosis.92 It also serves to quantitate the number of motor units for follow-up studies, documenting the rate of loss in patients with motor neuron disease and other neurogenic processes,5,6,29,46,57,58,144,147 although it sheds no light on the functional status of the surviving motor units.41
2
ASSESSMENT OF NERVE EXCITABILITY
This section reviews the modulation of axonal excitability following a single action potential.38,138 The behavior of a single axon in this regard remains poorly understood. The altered excitability of many fibers collectively determines the size of a compound potential that will yield even more complex, yet important biophysical information. Axonal excitability also undergoes profound change after subthreshold stimulation, as discussed in subsequent sections.
Refractory Period After passage of an impulse, an axon becomes totally inexcitable for a fraction of a millisecond during the absolute refractory period, then gradually recovers its prestimulus excitability within the ensuing few milliseconds during the relative refractory period. Direct measurement of the nerve action potentials in experimental animals13,67,142 substantiates the results in human studies, mostly tested in the sensory nerves or mixed nerves.15,35,63,71,139 When measured by muscle response, the refractory period depends not only on the excitability of the nerve but also the excitability of the neuromuscular junction, as implied by the term refractory period of transmission.62,105 Modified paired-shock techniques, however, make the study of the
refractory characteristics possible as they pertain to the motor fibers per se.70,87,91,121 Although a considerable amount of data has accumulated, its clinical value and limitations await clarification.86 The physiologic mechanism underlying the refractory period centers on inactivation of sodium (Na+) conductance (see Chapter 2-3). After the passage of an impulse, sodium channels will close to initiate repolarization. Once closed, or inactivated, they cannot open immediately, regardless of the magnitude of depolarization by a subsequent impulse. This constitutes the absolute refractory period, lasting 0.5-1.0 ms. During the subsequent relative refractory period, lasting 3-5 ms, only an excessive depolarization, far beyond the ordinary range, can reactivate sodium conductance. Here, the impulse propagates more slowly than usual because it takes longer to reach the elevated critical level required to generate the action potential. The refractory period is prolonged with low temperature,35,40,113,114 advanced age,50 slow conduction velocity,113,114 and after experimental demyelination 48,96,129,130
Paired Shock and Collision Technique A second shock delivered at a varying time interval after the first reveals excitability changes induced by the preceding impulse. In this method, called the paired-shock or conditioning and testing technique, the first shock conditions the nerve and the second impulse tests the effect. The test stimulus, given during the refractory period of the conditioning stimulus, elicits no response. During the relative refractory period that ensues, the test response shows reduced amplitude and increased latency. After extensive investigation in experimental animals,16,142 the paired-shock technique has found its way to the study of human sensory potentials35,137,14° and mixed-nerve potentials.63,99 In testing the motor fibers with the short interstimulus interval required for the study of the refractory period, the muscle responses elicited by the first and second stimuli overlap. A computerized subtraction technique circumvents this problem
220
Nerve Conduction Studies
by separating the test stimulus from the of antidromic activity (Fig. 8-1). Its magconditioning muscle response.15,91 The size nitude and speed depends solely on the of the test response measured, however, neural excitability after passage of the still depends on the excitability change of conditioning stimulus, S(A1). The S(A1)-toS(W) time interval dictates the point of colnot only the motor axons but also the neuromuscular junction and muscle fibers.34 lision and consequently the length of the Therefore, this technique, based on suc- nerve segment made refractory by S(A1), cessively evoked muscle responses, fails to before it is eliminated by the antidromic measure the nerve refractory period per se. activity of S(W). Changing the S(A1)-toA collision technique originally devised to S(A2) time interval defines the range of the avoid this difficulty determines the refracabsolute refractory periods of the differtory period of antidromic motor impulses ent motor fibers by demonstrating the seby paired distal stimuli followed by an aprial recovery of the test response amplipropriately timed single proximal stimulus, tude (Fig. 8-2A). In contrast, the latency of the test response elucidates the durawhich measures the test volley.70 Alternatively, paired proximal stimuli, tion of the relative refractory period of the combined with a single distal stimulus, almost excitable fibers (Fig. 8-2B). Table 8-2 summarizes the results in 20 ulnar low assessment of orthodromic motor impulses, eliminating the effects of 84 the musnerves from 10 healthy subjects studied cle and neuromuscular junction. In this in our laboratory.87 arrangement, the descending impulse generated by the first of the paired axilChanges in Amplitude lary shocks, S(A1), eliminates the antidromic impulse from the distal shock at versus Latency the wrist, S{W). The impulse of the second axillary stimulus, S(A2), will propa- The amplitude changes of the test regate distally along the motor fibers cleared sponse obtained with shocks of maximal
Figure 8-1. Compound muscle action potentials recorded by surface electrodes placed over the abductor digiti minimi after stimulation of the ulnar nerve. The diagrams on the left show the collision between orthodromic (solid arrows) and antidromic (dotted arrows) impulses. Axillary stimulation, S(A), given 6.0 ms after the stimulus at the wrist, S(W), triggered sweeps on the oscilloscope. With single stimulation at the axilla and at the wrist (middle tracing), the orthodromic impulse elicited by S(A) collided with the antidromic impulse of S(W) from the wrist. With paired shocks at the axilla (bottom tracing), M(A2) appeared because the first axillary stimulus, S(A1), cleared the path for the second stimulus, S(A2). [From Kimura, Yamada, and Rodnitzky,87 with permission.]
Figure 8-2. A. Paired axillary shocks, S(A1)and S(A2), of just maximal intensity combined with a single shock at the wrist, S(W). Interstimulus intervals between S(A1) and S(A2) ranged from 0.6 to 3.0 ms. S(A2) always occurred 5.0 ms after S(W), which triggered sweeps on the oscilloscope. In the normal subject, M(A2) first appeared (small arrows) at an interstimulus interval of 0.8 ms and recovered completely by 3.0 ms. The patient with the Guillain-Barre syndrome showed delayed and incomplete recovery. [From Kimura,85 with permission.] B. Paired axillary shocks, S(A1) and S(A2), of just maximal intensity combined with a single shock at the wrist, S(W) (cf. bottom tracing in A). Delivering S(A1) 6.0 ms after S(W) allowed collision to occur 1.5 ms after S(A1). The interstimulus intervals between S(A1) and S(A2) ranged from 1.2 to 3.0 ms in increments of 0.2 ms. The figures on the left show amplitude measurements with a slow sweep triggered by S(W). The figures on the right illustrate latency determination with a fast sweep triggered by S(A2) and displayed after a predetermined delay of 11.0 ms. [From Kimura,87 with permission.]
Table 8-2 Interstimulus Intervals of the Paired Shocks and Conduction Velocity of the Test Response (Mean ± SD)
Length of Refractory Segment A distance normally covered in 0.5 ms A distance normally covered in 1.5 ms
Initial Recovery in Amplitude (Test Response >5% of Unconditioned Response) Interstimulus Conduction Interval Between Velocity of Paired Shocks Test Impulse (ms) (% of normal)
Full Recovery in Amplitude (Test Response >95% of Unconditioned Response) Interstimulus Conduction Interval Velocity of Between Paired Shocks Test Impulse (% of normal) (ms)
Full Recovery in Velocity (>95%) Interstimulus Interval Between Paired Shocks (ms)
1.16 ±0.18
55.3 ± 19.2
2.11 ±0.50
81.2 ± 17.4
2.65 ± 0.65
1.18 ±0.16
70.3 ± 13.5
2.16 ±0.52
87.3 ± 14.2
2.36 ± 0.45
87
Source: From Kimura et al. with permission.
221
222
Nerve Conduction Studies
Figure 8-3. The pattern of recovery in amplitude of M(A2) during the refractory period in 10 healthy subjects. The return of M(A2) followed the identical time course with the passage of impulse along the shorter (0.5 ms) or longer refractory segment (1.5 ms). The gradual increase of M(A2) indicates the range of the absolute refractory periods of different motor fibers. [From Kimura, Yamada, and Rodnitzky,87 with permission.]
intensity follow a nearly identical course irrespective of the length of the refractory segment (Fig. 8-3). Therefore, reduction in amplitude of the test response must result from failure of nerve activation at the site of stimulation, rather than cessation of propagation along the course of the nerve. The impulse conducts at a slower speed than normal, if transmitted at all, during the relative refractory period, showing the greatest delay near the absolute refractory period (Fig. 8-4). Thereafter, the conduction progressively recovers to normal as the interstimulus interval between the conditioning and test stimuli increases. The length of the refractory segment, which hardly influences the recovery of the amplitude, substantially alters the time course of the latency. The longer the refractory segment, the greater the change in latency of the test response. The delay, however, does not increase linearly in proportion to the length of the refractory segment; in fact, a change in latency per unit length decreases for a longer conduction distance. Therefore, the average conduction velocity improves as the refractory segment increases (Fig. 8-5). These findings confirm the results of an
animal study142 that indicate (1) that a delay of the test impulse during the refractory period allows an increasing interval between conditioning and test impulses as they travel further distally and (2) that an increasingly longer interval between the two impulses, in turn, leads to progressive recovery of the test impulse conduction velocity. Because of this regressive process, the test impulse conducts at a relatively normal speed by the time it reaches the end of the refractory segment, especially for a longer nerve.85 Electrophysiologic studies of human sensory fibers,35 as well as computer simulation, have shown the same relationship between the refractory period and the length of the nerve segment.145 Human studies of the refractory period suffer from technical limitation in precisely measuring the amplitude and latency of the test response. Specific problems include small signals, unstable baseline, gradual onset of the evoked response, and partial overlap of the test response with the preceding events, despite the use of a collision technique. A computerized cross-correlation analysis helps improve numeric quantification of the
Figure 8-4. The pattern of recovery in latency of M(A2) in the same subjects as shown in Figure 8-3. The curve shows the latency difference between the response to a single axillary shock M(A), and the response to the second axillary shock M(A2) of the pair. The passage of impulse across the longer refractory segment (1.5 ms) showed significantly slower recovery as compared with the shorter refractory segment (0.5 ms). The bottom curve (triangles) plots the difference in delay of latency between 1.5 ms and 0.5 ms segments. The values so calculated represent the delay attributable to the last 1.0 ms of the 1.5 ms segment. [From Kimura, Yamada, and Rodnitzky,87 with permission.]
Figure 8-5. The time course of recovery in conduction velocity of M(A2) in the same subjects shown in Figures 8-3 and 8-4. The conduction velocities were calculated assuming that the delay of M(A2) occurred primarily in the segment proximal to the point of collision. In contrast to the pattern of recovery in latency (compare Figure 8-4), the conduction velocity returned significantly faster for the passage of impulse across the longer (1.5 ms) than the shorter (0.5 ms) refractory segment. The top curve (triangles) shows the estimated velocity of M(A2) over the last 1.0 ms of the 1.5 ms segment. [From Kimura, Yamada, and Rodnitzky,87 with permission.]
223
Nerve Conduction Studies
224
compound muscle potential in shape and latency.53 In this method, the height of the peak in the correlation curve gives a shape-weighted measure of the size of the test response, and the time lag of the peak indicates the delay of the test response as compared with an averaged unconditioned muscle response. Another technique, called the double-collision method, alleviates the transient changes in nerve and muscle fiber conduction3,74,75 that can distort test muscle responses. A number of studies have shown prolongation of the refractory period of sensory and mixed nerve fibers in diseases of the peripheral nerve.98,99,140 Patients with alcoholic neuropathy had an increased refractory period of the median sensory fibers.2 In patients with chronic renal failure, the initially abnormal relative refractory period97reverted to normal after hemodialysis. An increased refractory period of median nerve sensory fibers in patients with multiple sclerosis suggested the possible involvement 68of peripheral nerve fibers in this disorder. Conversely, hypokalemia of various origins102shortened the relative refractory period. Most previous studies in humans have dealt with the sensory or mixed-nerve fibers, but similar alterations probably occur in the refractory characteristics of motor fibers. In fact, the absolute and relative87 refractory periods affect motor fibers, sensory fibers, and mixed fibers63 alike. For example, full recovery in the amplitude of the test response precedes full recovery of the conduction velocity,35,63,69,87 regardless of the type of nerve fiber tested. Determining the refractory period of individual motor fibers requires recording of single motor unit potentials after delivery of paired stimuli to the nerve.17 Studies of the whole nerve lack precision because fibers with different conduction characteristics contribute to the absolute and relative refractory period. Furthermore, in contrast to amplitude, which follows a predictable time course, small, often variable changes in latency provide limited value in clinical assessment. These uncertainties make the measurement of the refractory period less useful than might be expected on theoretical grounds as a clinical test in diagnosing diseases of the motor fibers and in elucidating their pathophysiology.
Excitability Changes after Passage of an Impulse Studies of the myelinated axons reveal superexcitable and late subexcitable phases of excitability changes (see Chapter 4-3) after absolute and relative refractory periods.12,63,138 Superexcitability reflects negative, or depolarizing, afterpotential from long-lasting depolarization of the internodal axon.11 Activation of fast potassium channels terminates this phase by regulating the conductance of the internodal axon membrane. Thus, blocking these channels by 4-aminopyridine breaks down the normal relationship between Superexcitability and membrane potential.8 The late subexcitability results from a positive, or hyperpolarizing, afterpotential that reflects two very different mechanisms: opening of slow potassium channels8,11,138 and activation of an electrogenic sodium pump triggered12,24,82,83 by intracellular sodium accumulation. These hyperpolarizing effects intensify after the passage of a train of impulses, probably contributing to the rate-dependent conduction failure in demyelinating neuropathies. Long, high-frequency trains, however, lead to an opposite, hyperexcitatory state, causing posttetanic13repetitive activity and ectopic discharges. These paradoxical hyperexcitability and spontaneous discharges may account for neuropathic sensory disturbance and neuromyotonia.14 Threshold tracking study of a single motor axon during posttetanic hyperexcitability24 revealed a+ build-up of extracellular potassium (K ) ions. Rat axons show similar phenomena after injection of potassium ions into or under a myelin sheath.47,80 In either case, a reversal of the electrochemical gradient causes the influx of potassium ions across the internodal axolemma into the axon, resulting in depolarization and further opening of potassium channels, accelerating inward potassium current.
3
THRESHOLD TRACKING Strength-Duration Curve
The threshold intensity just capable of exciting the axons varies according to the
Other Techniques to Assess Nerve Function duration of the current; the shorter the duration, the greater the intensity to achieve the same depolarization. The strength-duration curve plots this relationship with a motor point stimulation that elicits a constant muscle response (Fig. 8-6 A and B). A long-duration shock excites both nerve and muscle, whereas a short-duration stimulus activates the nerve more effectively than the muscle. The excitability characteristics expressed by this curve, therefore, can differentiate a normally innervated muscle from a partially or totally denervated one. To formulate numerical indices of excitability, rheobase is defined as the minimal current strength below which no response occurs even if the current lasts infinitely or at least 300 ms. Chronaxie is the minimal duration of a current required to excite the cell at twice the rheobase strength. The same prinicple applies to the study of sensory fibers as a measure of sensory deficit in peripheral neuropathy. Although of historical interest, neither rheobase nor chronaxie has proven94satisfactory as a test in clinical practice. The strength-duration curve itself has fallen into disrepute because of the excessive time required for its determination and the complexity of its interpretation, but the test of nerve excitability remains an area of considerable theoretical and possibly clinical interest. Threshold tracking techniques test nerve excitability to assess the membrane potential, properties of ion26 channels, and electrogenic ion pumps. Changing the environment may alter the threshold—for example, by inducing ischemia or applying preceding currents. As described in the previous section, a single shock or a train of supramaximal shocks given as a conditioning stimulus tests refractoriness and superexcitability that follow the passage of an action potential (Fig. 8-7). In contrast, a brief or prolonged subthreshold current assesses subliminal excitability changes, which latent addition and threshold electrotonus measures. The threshold measurements all test the membrane properties of the nerve at a point of stimulation, thus complementing the conventional studies that measure the conduction characteristics of the axon along its length. The technique, therefore, is better suited for studying diffuse axonal properties, as in
225
Figure 8-6. A. The strength-duration curve showing the stimulus intensity (ordinate) required for each duration of stimulus (abscissa). In this example, the stimulus intensity required remains the same after the stimulus duration reaches 1.8 ms. This strength is the rheobase (R). Chronaxie (Q is determined (arrow) as that duration along the strength-duration curve 112 at a strength twice the rheobase (2 x R). [From Ochs, with permission.] B. The normal strengthduration curve from a motor point of the abductor pollicis brevis compared to the dashed line plotting curves from the denervated muscle of the other hand. Determinations at different times during reinnervation showed the return of the strength-duration curve toward normal. [From Ochs,112 with permission.]
metabolic or toxic neuropathy, than focal abnormalities, as in demyelinating neuropathies. Although these methods provide important insights into the physiology and pathophysiology of neuronal properties, their clinical utility awaits confirmation.
Threshold Measurement of Strength-Duration Time Constant In the simplest type of threshold tracking, only test stimuli delivered alone determine
226
Figure 8-7. Nerve excitability changes following action potential. [From Ochs,112 with permission.]
the nerve excitability change brought about by ischemia, hyperventilation, anesthetic agents, or other drugs.59,146 Baseline studies consist of the application of a series of stimuli, stepped up and down, at regular intervals to determine the intensity required to activate a standard fraction (e.g., 40 percent of the maximum muscle response). A repeated procedure then evaluates the new threshold compared to the control value after altering the environment. The changes detected by these means, if expressed in percentages, apply equally to both single-fiber and multi-fiber preparations. As shown in the strength-duration curve, increased duration reduces the current strength needed to excite the same fraction of a compound action potential. Threshold
Nerve Conduction Studies tracking tests this relationship in human peripheral nerve.108 The old term, chronaxie, corresponds to the strength-duration time constant defined from the thresholds for just two pulses of different duration.108 The sensory fibers with more prominent, persistent sodium conductance27 have longer strength-duration time constants than the motor fibers.108,115 Abnormalities may result from changes in resting membrane potential, sodium conductance, or myelination. An increase by depolarization and a decrease by hyperpolarization24,27 will reflect the +voltage-dependent behavior of sodium (Na ) conductance. For example, patients with acquired neuromyotonia, a condition of peripheral nerve hyperexcitabiliry, have an increased strength-duration time constant of motor but not sensory axons.101 This finding suggests relative axonal depolarization, greater persistent sodium conductance or enlarged nodal area as a result of paranodal demyelination. When the nodes under the stimulating electrode have a very high value in threshold, inadvertent excitation of the intact nodes further away may show a normal value in strength-duration time constant. For example, studies may remain normal in carpal tunnel syndrome, despite abnormally high rheobase.109 This limitation applies to all threshold-tracking techniques, making them unsuitable for studying focal neuropathies, especially if a pathologic segment shows hypoexcitabiliry instead of hyperexcitability.
Latent Addition and Accommodation Very brief subthreshold conditioning pulses produce a membrane potential called local response, which is confined to the node of Ranvier and shows a decay regulated by the membrane time constant. It simply adds to the changes induced by a subsequent test stimulus if given within a certain time, as implied by the term latent addition.141 In contrast, currents longer than a few milliseconds affect not only the nodes but also the myelin sheath, altering the potential difference across the internodal axon mem-
Other Techniques to Assess Nerve Function
227
Figure 8-8. A. Latent addition for the motor axons of a human ulnar nerve, plotting percentage threshold changes (ordinate) against time delay (abscissa). From top to bottom, the traces show time course of recovery after three sets each of hyperpolarizing and depolarizing conditioning stimuli of 60 us duration. The intensities used equal -90, -60, and -30 percent (top half) and 30, 60, and 90 percent (bottom half] of the control stimulus established by threshold tracking to maintain a 30 percent amplitude of maximal hypothenar compound muscle action potential. Changing membrane excitability measured by test stimuli of 60 us duration delivered every 20 us indicated a slower recovery of excitability following depolarizing (lower half) than hyperpolarizing (upper half) conditioning pulses. [Courtesy of Shouchan Lin, M.D., Department of Neurology, Cheng-Kung University Hospital, Tainan, Taiwan.] B. Latent addition for the sensory axons of a human ulnar nerve, using the same arrangements as for the motor axons in A, except for the use of the target threshold to maintain 30 percent amplitude of maximal fifth digit compound sensory potential. Compared with motor fibers, sensory fibers show a slower time course of recovery after a hyperpolarizing conditioning stimulus (top half) and, to a lesser extent, a depolarizing conditioning stimulus (bottom half). [Courtesy of Shouchan Lin, M.D., Department of Neurology, Cheng-Kun University Hospital, Tainan, Taiwan.]
brane. Activation of a variety of nodal and internodal ion channels regulates this type of change of membrane potential, termed electrotonus. The threshold also changes in association with electrotonus, as implied by the term threshold electrotonus.8,19,21 A brief subthreshold depolarizing current increases nerve excitability (or de-
creases its threshold) because it brings the membrane potential that much closer to the critical level of activation. In other words, a second stimulus generates an action potential more easily if applied to an already depolarized membrane. Brief hyperpolarizing currents show the opposite effect on membrane excitability, elevating its threshold to the test stimulus (Fig.
228
8-8). One study of latent addition estimated the sensory fibers to have about three times larger average time constants of a local response than motor fibers with depolarizing conditioning stimuli.116 This difference dropped to about one and a half with116hyperpolarizing 27 conditioning stimuli. Another study, using automatic threshold tracking, found a slower recovery from hyperpolarizing pulses than from depolarizing pulses in sensory fibers, although both motor and sensory fibers had a similar membrane time constant of about 45 us. These findings suggest greater resting activation or persistent sodium conductance in the sensory fibers, which adds a slow component to the recovery of threshold from hyperpolarizing pulses and increases the strength-duration time constant.27,37 Latent addition allows in vivo study of persistent sodium conductance, which may explain the mechanism underlying some forms of axonal hyperexcitability. A prolonged subthreshold current may not increase the excitability as much as expected because the voltage-dependent channels tend to oppose depolarization in the process known as accommodation. Similarly, opposing actions of voltagedependent ion channels tend to modify the effect induced by hyperpolarizing current. Testing the change of membrane excitability in this context, therefore, can uncover function and dysfunction of the ion channels regarding their rectifying properties. In particular, this method holds promise in assessing the role of potassium channels, which probably play a key role in the accommodative process under ordinary circumstances.7,21,22,23,100 Capacitative and resistive membrane properties11 determine the internodal potential changes in the axons induced either by a nerve impulse or by externally applied currents.18,95 Various rectifying channels in the nodal and internodal axon membranes alter electrotonic potentials recorded from the axon. A slow and fast potassium conductance, gKs and gKf, activated by prolonged subthreshold depolarization, relates to the currents induced by the specific channel types identified in voltage-clamp and patch-clamp studies; gKs to Ks currents via S channels, and gKf
Nerve Conduction Studies to Kf1 currents via I channels. Subthreshold electrotonus probably does not involve Kf2 currents related to F channels, which respond to a greater depolarization compared to I channels. Subthreshold hyperpolarization activates inward rectification, gIR. The contributions of gKs, gKf, and gIR were inferred from the effects of the channel blockers tetraethyl ammonium (TEA), 4-amino pyridine (4-AP), and Cs+, respectively.8
Electrotonus and Threshold Electrotonus A study of threshold electrotonus determines the time course of membrane excitability change induced by a rectangular subthreshold current pulse based on the intensity of the test shock necessary to evoke a26defined fraction of the maximal response. Multiunit recording enables direct comparisons between the changes in threshold determined by this method and the changes in membrane potential7 measured by extracellular recordings. According to these studies, the change in threshold normally follows the electrotonic changes in membrane potential caused by the subthreshold polarizing currents.21 The channel blockers seem to affect these two measures in the same way, confirming the close causal correspondence between electrotonic and threshold changes.21 The threshold measurements usually parallel electrotonic potentials; thus, the term threshold electrotonus8 defines the threshold changes corresponding to electrotonic changes. This technique, measuring the threshold noninvasively, estimates changes of membrane excitability after subthreshold polarization. Threshold electrotonus, like electrotonus, can be used to study the effect of depolarizing as well as hyperpolarizing current pulses. A family of accommodation curves thus generated will provide information about the subthreshold electrical properties of the axon or the nerve. The slow changes in threshold in response to depolarizing currents occur mainly in the direction of accommodation, or less excitability than expected. Hyperpolarizing currents induce
Other Techniques to Assess Nerve Function the response mainly in the opposite direction or less suppression than expected, as implied by the term negative accommodation.7,21 A normally very close relationship between membrane potential and threshold, and therefore between electrotonus and threshold electrotonus, breaks down in a few situations, where a fast component of accommodation not reflected in the membrane potential causes threshold electrotonus to deviate from electrotonus. Such separations occur with DC depolarizing currents, raised extracellular potassium concentrations, or ischemia. Inactivation of closed (unactivated) sodium channels probably underlies the most important accommodative process that manifests without altering the membrane potential per se, as has been shown in isolated toad fibers.143 Mammalian fibers rapidly accommodate only when they are depolarized by 15-20 mV.26 The insensitivity to potassium channel blockers of this fast accommodation supports the hypothesis that sodium channel inactivation plays a role.7
Techniques to Measure Threshold Electrotonus Threshold electrotonus21 tests the effect of standardized subthreshold depolarizing or hyperpolarizing currents on "threshold," defined as the current required to just excite a standard, submaximal response. Subthreshold depolarizing currents lasting 100 ms adequately activate the slow potassium channels responsible for Ks currents inducing accommodation. Hyperpolarizing current pulses, usually 300 ms in duration, activate IH, an inwardly rectifying current causing negative accommodation. A test shock applied to measure thresholds ordinarily has a 1 ms duration; that value is chosen to be long compared with the time constant of the nodes of Ranvier but short compared with the time constants of the internodal axon and slowly activating ion channels. Normalizing both the polarizing currents and threshold measurements as percentages of the unconditioned threshold current minimizes the effect of tissue impedance.
229
Under computer control, 1 ms test pulses, delivered alone at 1Hz, determine the "threshold" current that is just sufficient to maintain a constant response in amplitude of a predetermined size. The value usually chosen equals 40 percent of the maximal response established by a supramaximal shock prior to the study. Depolarizing and hyperpolarizing conditioning current pulses of 100 ms duration usually have ± 20 percent and ± 40 percent of "threshold" current. The procedure consists of alternating test pulses on their own and test pulses superimposed on 100 ms depolarizing and hyperpolarizing conditioning pulses. The interval between the start of the test and conditioning shocks is slowly advanced from +2 to -98 ms over a period of 10 minutes. The increase in excitability produced by a depolarizing current, expressed upward as percentage reductions in threshold, cannot exceed the line at the top for 100 percent threshold reduction (Fig. 8-9). The start of the current pulse immediately depolarizes the node, resulting in a step increase in excitability. Subsequent depolarization of the node, as well as of the internodal part of the axon, causes a further increase in excitability, but more slowly, for about 20 ms. Accommodation follows,19 with a partial repolarization of the nodal membrane, caused mainly by the activation of slow potassium channels21 present in the nodal and internodal axon membrane.8 Hyperpolarization gives rise to only two phases of response, the fast component with changes in the nodal potential and prominent slow changes affecting both the node and the internode together. Longer and stronger hyperpolarizing currents lead to a late depolarization or negative accommodation by inward rectification,21 a phenomenon more prominent in the sensory than the motor fibers.25 A computer model of a node and an internode gives a reasonable account of the time course of threshold electronus, taking into consideration one type of sodium channel and three types of potassium channels.26 For example, increased activation of potassium channels would decrease the axonal membrane resistance, resulting in "fanning-in" or flattening of
Nerve Conduction Studies
230
Figure 8-9. Membrane potential and threshold electrotonus showing increased excitability followed by accommodation to depolarizing subthreshold currents (top half) and decreased excitability followed by negative accommodation to hyperpolarizing currents (bottom half). Prior membrane depolarization or hyperpolarization shifted the response curves toward the baseline (fanning-in) or away from the baseline 79(fanning-out). [From Kaji et al, with permission.]
the excitability curve. The opposite abnormalities would result in 'fanning out' of the threshold electrotonus.
Applications of Threshold Measurements The two threshold-tracking techniques, latent addition and threshold electrotonus, test human nerve excitability in vivo, providing better understanding of any channel abnormalities. According to the 27 experimental data on latent addition, the axonal responses to brief current pulses depend for the mostpart on a small, persistent sodium conductance. Thus, any models of human nerve excitability should incorporate persistent as well as transient nodal sodium channels, in addition to fast and slow potassium channels and inward rectification, as described above. The classical theory based on nodal currents suffices to analyze the normal waveform of an action potential. Modern approaches emphasize internodal mechanisms to account for pathologic nerve activity, as seen in11 Barrett and Barrett's equivalent circuit derived from the electrical interaction between nodes and internodes.122 This model can explain many conditions in which threshold electrotonus closely parallels electrotonus:23,25,134,135 for example, pathophysiology of postis-
chemic ectopic discharges23 and mechanisms underlying the difference in inward rectification 25between motor and sensory nerve fibers. The model must be modified in reproducing abnormal features when threshold electrotonus deviates from electrotonus in such28conditions as amyotrophic lateral7 sclerosis or prolonged depolarized state. Motor and sensory axons show very similar depolarizing responses but different hyperpolarizing responses. Hyperpolarization deactivates potassium channels in the internodal axon and later activates the axonal inward rectifier, IH, an excitatory channel with permeability to sodium as well as potassium ions. A difference in expression of the inward rectifier helps to explain the characteristic behaviors of the motor and sensory axons on release of experimentally induced ischemia25,36 and on the cessation of prolonged tetanization.81,83 Applying a pneumatic tourniquet to a limb induces substantial ischemia, which inhibits the +electrogenic sodium (Na+)potassium (K ) pump, causing membrane depolarization. The extracellular accumulation of potassium 12ions also reduces membrane potential. On release of the cuff, hyperactivity of the electrogenic sodium pump rapidly hyperpolarizes the axons. In tests of these changes, threshold tracking of a constant fraction of the compound muscle action potential shows
Other Techniques to Assess Nerve Function results similar to those12,22,59 obtained from tracking of a single fiber. Ischemia, like depolarization, causes a fanning-in of the threshold electrotonus, reflecting increased activation of fast and slow potassium channels. The pattern reverses after release of ischemia, showing a fanningout, mimicking the trend seen during hyperpolarization. These findings indicate that the ischemic fall in threshold primarily reflects depolarization; the postischemic rise, hyperpolarization. This close relationship breaks down, however, if the axons become so depolarized that sodium channel inactivation becomes a major determinant of excitability.7 This occurs during prolonged ischemia, and in a few patients with amyotrophic lateral sclerosis. During ischemia, motor latency increases despite depolarization, reflecting inactivation of sodium channels.109 Postischemia, latency stays prolonged, reflecting the hyperpolarization of axons with a threshold increase exceeding 200 percent. Studies of sensory25,106 fibers have shown similar observations. Threshold tracking studies have also elucidated the mechanism of postischemic ectopic discharges in motor axonsm22,23 as well as postischemic paresthesias originating from cutaneous afferents.25 Patients with diabetic 136 neuropathy show resistance to ischemia, as indicated by a deviation of threshold changes from the normal pattern 146 within 5 minutes of arterial occlusion and an even greater dissociation during postischemic hyperpolarization.136 A similar study in patients with amyotrophic lateral sclerosis, however, failed to confirm110previous reports of ischemic resistance.
Clinical Assessments The first clinical studies of amyotrophic lateral sclerosis showed two kinds of findings during depolarization:28,88 type 1, abnormally reduced threshold or loss of physiologic accommodation, probably reflecting an imbalance between sodium and potassium currents, and type 2, sharply increased threshold, indicating sodium channel inactivation. In another series72 many responses fell within the
231
normal range, but the average showed significant fanning-out, resembling the effects of hyperpolarization. The results of these studies suggest deactivation of potassium channels and, consequently, reduced potassium conductances. Later series had more equivocal results, with the mean responses not significantly different from those of the controls, although the depolarizing responses showed distinctive changes in some patients. When divided into subgroups, the "definite amyo-trophic lateral sclerosis" and "progressive muscular atrophy" groups—not the "bulbar" and "primary lateral sclerosis" groups—exhibited these abnormalities. Threshold electrotonus cannot test the abnormal membrane properties related to fasciculations20,79 if the change primarily involves the motor nerve terminals.93 In addition to motor fibers, cutaneous sensory axons may show excitability change in patients with amyotrophic lateral sclerosis.39 In one study of diabetic polyneuropathy in which the motor and sensory axons were tested at the wrist, only a minority of responses lay outside the normal range.72 The group means, however, showed a highly significant difference when compared to those of normal control subjects or patients with amyotrophic lateral sclerosis. The abnormalities, seen only in response to hyperpolarization, implied a deficit in inward rectification involving both motor and sensory nerves.118 The inward rectification depends on the level of intracellular cyclic adenosine monophosphate,1,76 a substance77reportedly lacking in diabetic nerves. Interestingly, threshold electrotonus applied to biopsied human sural nerve in vitro has shown the most prominent inward rectification in C fibers,65 often most severely affected in diabetic neuropathy. This method has also demonstrated the reversal of the pathologic resistance to ischemia after therapy in patients with diabetes.120 Threshold electrotonus showed a marked symmetrical fanning-in of the responses66,124 in rapidly developing, predominantly large-fiber sensory neuropathy induced by combination42chemotherapy of Taxol and cisplatin. These
232
findings, seen before any clinical or neurological signs of neuropathy,124 indicate disturbances in membrane excitability caused by depolarization or increased conductance of the internodal axon membrane. Taxol also119 depolarized human sural nerves in vitro. Patients with carpal tunnel syndrome show no abnormalities, probably because stimulation at the point of involved sites preferentially excites adjacent, normal nodes, or other more normal fibers.109 The other conditions tested by threshold electrotonus include multifocal motor neuropathy with conduction block, showing abnormalities restricted to the site of the lesion,79 and monomeric amyotrophy with spinal hemiatrophy.89 Threshold tracking is a powerful tool for investigating excitable membranes. A single stimulus or a train of suprathreshold stimuli causes refractoriness and superexcitability. Brief and prolonged subthreshold currents induce excitability changes that latent addition and threshold electrotonus can delineate. In particular, threshold electrotonus can serve as an index of membrane potentials, which under most circumstances closely correspond to the changes in membrane excitability. It provides a more sensitive indicator of changes in membrane potential than simple threshold tracking. This approach, though in theory well suited for studying human peripheral nerves in vivo, has so far found little use in practice because of its inherent limitations in the clinical context. The method tests the excitability of only a small population of axons with thresholds close to the level chosen for tracking, omitting the remaining, more or less excitable, fibers. Abnormalities also go undetected for degenerated axons or for demyelinated fibers with conduction block that lie between the stimulation site and the recording site. Furthermore, the technique relates only to the point of stimulation, making it less applicable for focal lesions because stimuli tend to activate the more excitable neighboring segments. In contrast, these measures provide important insights into membrane properties in normal and diffuse neuropathies that affect the axons uniformly. Their usefulness as a diagnostic test awaits clarification.
Nerve Conduction Studies
REFERENCES 1. Akasu T, Shoji S: cAMP-dependent inward rectifier current in neurons of the rat suprachiasmatic nucleus. Pfiugers Arch 429:117-125, 1994. 2. Alderson MK, Petajan JH: Relative refractory period: A measure to detect early neuropathy in alcoholics. Muscle Nerve 10:323-328, 1987. 3. Arasaki K: Maximal and minimal motor nerve conduction velocities determined by a collision method: correlation with axonal conduction velocity of type-identified motor units. J Neurol Sci 110:131-138, 1992. 4. Arasaki K, Tamaki M, Hosoya Y, Kudo N: Validity of electromyograms and tension as a means of motor unit number estimation. Muscle Nerve 20:552-560, 1997. 5. Armon C, Brandstater ME: Motor unit number estimate-based rates of progression of ALS predict patient survival. Muscle Nerve 22:15711575, 1999. 6. Armon C, Brandstater ME, Peterson GW: Motor unit number estimates and quantitative muscle strength measurements of distal muscles in patients with amyotrophic lateral sclerosis. Muscle Nerve 20:499-501, 1997. 7. Baker M, Bostock H: Depolarization changes the mechanism of accommodation in rat and human motor axons. J Physiol (Lond) 411: 545-561, 1989. 8. Baker M, Bostock H, Grafe P, Martius P: Function and distribution of three types of rectifying channel in rat spinal root myelinated axons. J Physiol (Lond) 383:45-67, 1987. 9. Ballantyne JP, Hansen S: A new method for the estimation of the numbers of motor units in muscles. 1. Control subjects and patients with myasthenia gravis. J Neurol Neurosurg Psychiatry 37:907-915, 1974. 10. Barkhaus PE, Nandedkar SD: Recording characteristics of the surface EMG electrodes. Muscle Nerve 17:1317-1323, 1994. 11. Barrett EF, Barrett JN: Intracellular recording from vertebrate myelinated axons: mechanism of depolarizing after potential. J Physiol (Lond) 323:117-144, 1982. 12. Bergmans J: The Physiology of Single Human Nerve Fibres. Vander, University of Louvain, Belgium, 1970. 13. Bergmans J: Physiological observations on single human nerve fibres. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiolsiology, Vol 2. Karger, Basel, 1973, pp 264-267. 14. Bergmans J: Repetitive activity induced in single human motor axons: a model for pathological activity. In Culp WJ, Ochoa J (eds): Abnormal Nerves and Muscles as Impulse Generators. Oxford University Press, New York, 1982, pp 393-418. 15. Betts RP, Johnston DM, Brown BH: Nerve fibre velocity and refractory period distributions in nerve trunks. J Neurol Neurosurg Psychiatry 39:694-700, 1976.
Other Techniques to Assess Nerve Function 16. Bishop GH, Heinbecker P: Differentiation of axon types in visceral nerves by means of the potential record. Am J Physiol 94:170-200, 1930. 17. Borg J: Refractory period of single motor nerve fibres in man. J Neurol Neurosurg Psychiatry 47:344-348, 1984. 18. Bostock H: Lorente de No and nerve physiology. News Physiol Sci 6:235-237, 1991. 19. Bostock H: Mechanisms of accommodation and adaptation in myelinated axons. In Waxman SG, Stys PK, Koosis JD (eds): The Axon. Oxford University Press, 1995, pp 311-327. 20. Bostock H: Abnormal excitability of motor axons in ALS. In Kimura J, Kaji R (eds): Physiology of ALS and Related Diseases. Elsevier, Amsterdam, 1997, pp 133-142. 21. Bostock H, Baker M: Evidence for two types of potassium channel in human motor axons in vivo. Brain Res 462:354-358, 1988. 22. Bostock H, Baker M, Grafe P, Reid G: Changes in excitability and accommodation of human motor axons following brief periods of ischaemia. J Physiol (Lond) 441:513-535, 1991. 23. Bostock H, Baker M, Reid G: Changes in excitability of human motor axons underlying post-ischaemic fasciculations: Evidence for two stable states. J Physiol (Lond) 441:537-557, 1991. 24. Bostock H, Bergmans J: Post-tetanic excitability changes and ectopic discharges in a human motor axon. Brain 117:913-928, 1994. 25. Bostock H, Burke D, Hales JP: Differences in behaviour of sensory and motor axons following release of ischemia. Brain 117:225-234, 1994. 26. Bostock H, Cikurel K, Burke D: Threshold tracking techniques in the study of human peripheral nerve. Muscle Nerve 21:137-158, 1998. 27. Bostock H, Rothwell JC: Latent addition in motor and sensory fibres of human peripheral nerve. J Physiol (Lond) 498:277-294, 1997. 28. Bostock H, Sharief MK, Reid G, Murray NMF: Axonal ion channel dysfunction in amyotrophic lateral sclerosis. Brain 118:217-225, 1995. 29. Bromberg MB, Abrams JL: Sources of error in the spike-triggered averaging method of motor unit number estimation (MUNE). Muscle Nerve 18:1139-1146, 1995. 30. Brown WF: A method for estimating the number of motor units in thenar muscles and the change in motor unit count with aging. J Neurol Neurosurg Psychiatry 35:845-852, 1972. 31. Brown WF, Fasby TE: Estimates of functional motor axon loss in diabetes. J Neurol Sci 23:275-293, 1974. 32. Brown WF, Chan KM: Quantitative methods for estimating the number of motor units in human muscles. Muscle Nerve 20:S70-S73, 1997. 33. Brown WF, Strong MJ, Snow R: Methods for estimating numbers of motor units in bicepsbrachialis muscles and losses of motor units with aging. Muscle Nerve 11:423-432, 1988. 34. Buchthal F, Engbaek L: Refractory period and conduction velocity of the striated muscle fibre. Acta Physiol Scand 59:199-220, 1963. 35. Buchthal F, Rosenfalck A: Evoked action po-
36.
37.
38. 39.
40.
41.
42.
43.
44.
45. 46.
47.
48.
49.
50.
233 tentials and conduction velocity in human sensory nerves. Brain Res 3:1-122, 1966. Burke D: Paraesthesiae and ectopic impulse actibity. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Amsterdam, 1996, pp 28-33. Burke D, Kiernan MC, Mogyoros I, Bostock H: Susceptibility to conduction block: Differences in the biophysical properties of cutaneous afferents and motor axons. In J Kimura, R Kaji (eds): Physiology of ALS and Related Diseases. Elsevier, Amsterdam, 1997, pp 43-53. Burke D, Miller TA, Kiernan MC, Mogyoros I: Activity-dependent modulation of excitability. Muscle Nerve 18:675-676, 1995. Burke D, Mogyoros I, Kiernan MC, Bostock H: Excitability of cutaneous sensory axons in amyotrophic lateral sclerosis. In Kimura J, Kaji R (eds): Physiology of ALS and Related Diseases. Elsevier, Amsterdam, 1997, pp 145-154. Burke D, Mogyoros I, Vagg R, Kiernan MC: Temperature dependence of excitability indices of human cutaneous afferents. Muscle Nerve 22:51-60, 1999. Chan KM, Doherty TJ, Andres LP, Porter MM, Brown T, Brown WF: Longitudinal study of the contractile and electrical properties of single human thenar motor units. Muscle Nerve 22:839-849, 1998. Chaudhry V, Rowinsky EK, Sartorius SE, Donchower RC, Cornblath DR: Peripheral neuropathy from taxol and cisplatin combination chemotherapy: Clinical and electrophysiological studies. Ann Neurol 35:304-311, 1994. Chroni E: Reservations on the motor unit number estimates based on the automated analysis of F-responses. Muscle Nerve 18:10741075, 1995. Conwit RA, Tracy B, Jamison C, McHugh M, Stashuk D, Brown WF, Metter EJ: Decomposition-enhanced spike-triggered averaging: contraction level effects. Muscle Nerve 20:976-982, 1997. Daube JR: Estimating the number of motor units in a muscle. J Clin Neurophysiol 583594, 1995. Daube JR: Motor unit number estimates in ALS. In Kimura J, Kaji R (eds): Physiology of ALS and Related Diseases. Elsevier, Amsterdam, 203-227, 1997. David G, Barrett JN, Barrett EF: Activation of internodal potassium conductance in rat myelinated -axons. J Physiol (Lond) 472:177-202, 1993. Davis FA: Impairment of repetitive impulse conduction in experimentally demyelinated and pressure-injured nerves. J Neurol Neurosurg Psychiatry 35:537-537, 1972. de Koning P, Wieneke GH, van der Most van Spijk D, Van Huffelen AC, Gispen WH, Jennekens FGI: Estimation of the number of motor units based on macro-EMG. J Neurol Neurosurg Psychiatry 51:403-411, 1988. Delbeke J, Kopec J, Mccomas AJ: The effects of age, temperature, and disease on the refractoriness of human nerve and muscle. J Neurol Neurosurg Psychiatry 41:65-71, 1978.
234 51. Doherty TJ, Brown WF: The estimated numbers and relative sizes of thenar motor units as selected by multiple point stimulation in young and older adults. Muscle Nerve 16:355-366, 1993. 52. Doherty TJ, Simmons Z, O'Connell B, Felice KJ, Conwit R, Ming Chan K, Brown T, Stashuk DW, Brown WF: Methods for estimating the numbers of motor units in human muscles. J Clin Neurophysiol 12:565-584, 1995. 53. Faisst S, Meyer M: A non-invasive computerized measurement of motor neurone refractory period and subnormal conduction in man. Electroencephalogr Clin Neurophysiol 51:548558, 1981. 54. Fang J, Shahani BT, Graupe D: Motor unit number estimation by spatial-temporal summation of single motor unit potentials. Muscle Nerve 20:461-468, 1997. 55. Feasby TE, Brown WF: Variation of motor unit size in the human extensor digitorum brevis and thenar muscles. J Neurol Neurosurg Psychiatry 37:916-926, 1974. 56. Felice KJ: Thenar motor unit number estimates using the MPS technique: Reproducibility studies in ALS patients and normal subjects. Muscle Nerve 18:1412-1416, 1995. 57. Felice KJ: A longitudinal study comparing thenar motor unit number estimates to other quantitative tests in patients with amyotrophic lateral sclerosis. Muscle Nerve 20:179-185, 1997. 58. Felice KJ: Nerve conduction velocities of single thenar motor axons based on the automated analysis of F waves in amyotrophic lateral sclerosis. Muscle Nerve 21:756-761, 1998. 59. Franz P, Weigl P, Grafe P, Baker M, Bostock H: Changes in excitability of human motor axons during ischaemia. Pflugers Archiv 411: R152, 1988. 60. Galea V: Changes in motor unit estimates with aging. J Clin Neurophysiol 13:253-260, 1996. 61. Galea V, de Bruin H, Cavasin R, McComas AJ: The numbers and relative sizes of motor units estimated by computer. Muscle Nerve 14:11231130, 1991. 62. Gilliatt RW, Meer J: The refractory period of transmission in patients with carpal tunnel syndrome. Muscle Nerve 13:445-450, 1990. 63. Gilliatt RW, Willison RG: The refractory and supernormal periods of the human median nerve. J Neurol Neurosurg Psychiatry 26:136-147, 1963. 64. Gorman PH, Kikta DG, Peckham PH: Neurophysiologic evaluation of lower motor neuron damage in tetraplegia. Muscle Nerve 21:13211323, 1998. 65. Grafe P, Quasthoffs, Grosskreutz J, Alzheimer C: Function of the hyperpolarization-activated inward rectification in nonmyelinated peripheral rat and human axons. J Neurophysiol 77:421-426, 1997. 66. Hanauske AR, Schilling T, Heinrich B, Kau R, Herzog M, Quasthoff S, Bochtler H, Diergarten K, Rastetter J: Clinical phase I study of paclitaxel followed by cisplatin in advanced head and neck squamous cell carcinoma. Semin Oncol 22:35-39, 1995.
Nerve Conduction Studies 67. Hodgkin AL: The Conduction of the Nervous Impulse. The Sherrington Lectures VII, Liverpool University Press, Liverpool, 1965. 68. Hopf HC, Eysholdt M: Impaired refractory periods of peripheral sensory nerves in multiple sclerosis. Ann Neurol 4:499-501, 1978. 69. Hopf HC, LeQuesne PM, Willison RG: Refractory periods and lower limiting frequencies of sensory fibres of the hand. In Kunze K, Desmedt JE (eds): Studies on Neuromuscular Diseases. Karger, Basel, 1975, pp 258-263. 70. Hopf HC, Lowitzsch K: Relative refractory periods of motor nerve fibres. In Kunze K, Desmedt JE (eds): Studies on Neuromuscular Diseases. Proceedings of the International Symposium (Giessen), Karger, Basel, 1975, pp 264-267. 71. Hopf HC, Lowitzsch K, Galland J: Conduction velocity during the supernormal and late subnormal periods in human nerve fibres. J Neurol 211:293-296, 1976. 72. Horn S, Quasthoff S, Grafe P, Bostock H, Renner R, Schrank B: Abnormal axonal inward rectification in diabetic neuropathy. Muscle Nerve 19:1268-1275, 1996. 73. Hughes AR, Colebatch JG: Surface potentials generated by synchronous activation of different fractions of the motor pool. Muscle Nerve 19:836-842, 1996. 74. Ingram DA, Davis GR, Swash M: The double collision technique: a new method for measurement of the motor nerve refractory period distribution in man. Electroencephalogr Clin Neurophysiol 66:225-234, 1987. 75. Ingram DA, Davis GR, Swash M: Motor conduction velocity distributions in man: results of a new computer-based collision technique. Electroencephalogr Clin Neurophysiol 66:235243, 1987. 76. Ingram SL, Williams JT: Modulation of the hyperpolarization-activated current (Ih) by cyclic nucleotides in guinea-pig primary afferent neurons. J Physiol (Lond) 492:97-106, 1996. 77. Ito H, Kanazawa A, Ohno A, Tanaka T, Miwa T, Fukuda T, Ueki A, Notoya Y: Effect of prostaglandin El, E2 and I2 derivatives and methyl-B12 on peripheral neuropathy in diabetic rats. In Ward J, Goto Y, (eds): Diabetic Neuropathy. John Wiley & Sons, Chichester, 1990, pp 517-523. 78. Kadrie H, Yates SK, Milner-Brown HS, Brown WF: Multiple point electrical stimulation of ulnar and median nerves. J Neurol Neurosurg Psychiatry 39:973-985, 1976. 79. Kaji R: Physiological and technical bases of peripheral nerve and motoneuron testing. In Kimura J, Kaji R (eds): Physiology of ALS and Related Diseases. Elsevier, Amsterdam, 1997, pp 15-41. 80. Kapoor R, Smith KJ, Felts PA, Davies M: Internodal potassium currents can generate ectopic impulses in mammalian myelinated axons. Brain Res 611:165-169, 1993. 81. Kiernan MC, Hales JP, Gracies J-M, Mogyoros I, Burke D: Paraesthesiae induced by prolonged high frequency stimulation of human cutaneous afferents. J Physiol (Lond) 501(Pt 2): 461-471, 1997.
Other Techniques to Assess Nerve Function 82. Kiernan MC, Mogyoros I, Burke D: Changes in excitability and impulse transmission following prolonged repetitive activity in normal subjects and patients with a focal nerve lesion. Brain 119:2029-2037, 1996. 83. Kiernan MC, Mogyoros I, Hales JP, Gracies JM, Burke D: Excitability changes in human cutaneous afferents induced by prolonged repetitive activity. J Physiol (Lond) 500:255-264, 1997. 84. Kimura J: Collision technique. Physiologic block of nerve impulses in studies of motor nerve conduction velocity. Neurology (Minneap) 26:680-682, 1976. 85. Kimura J: A method for estimating the refractory period of motor fibers in the human peripheral nerve. J Neurol Sci 28:485-490, 1976. 86. Kimura J: Refractory period measurement in the clinical domain. In Waxman SA, Ritchie JM (eds): Demyelinating Disease: Basic and Clinical Electrophysiology. Raven Press, New York, 1981, pp 239-265. 87. Kimura J, Yamada T, Rodnitzky RL: Refractory period of human motor nerve fibres. J Neurol Neurosurg Psychiatry 41:784-790, 1978. 88. Kodama M, Kaji R, Kojima Y, Hirota N, Kohara N, Shibasaki H, Bostock H, Kimura J: Threshold electrotonus in patients with amyotrophic lateral sclerosis: Further experience with Japanese subjects. Electroencephalogr Clin Neurophysiol 97:S172, 1995. 89. Kojima Y, Kaji R, Hirota N, Kohara N, Kimura J, Murray NMF, Bostock H: Threshold electrotonus in monomeric amyotrophy with spinal hemiatrophy. Electroencephalogr Clinical Neurophysiol 97:S172, 1995. 90. Komori T, Brown WF: Motor unit estimate with F-response. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology E1sevier, Amsterdam, 1996, pp 568-571. 91. Kopec J, Delbecke J, Mccomas AJ: Refractory period studies in a human neuromuscular preparation. J Neurol Neurosurg Psychiatry 41: 54-64, 1978. 92. Kuwabara S, Mizobuchi K, Ogawara K, Hattori T: Dissociated small hand muscle involvement in amyotrophic lateral sclerosis detected by motor unit number estimates. Muscle Nerve 22:870-873, 1999. 93. Layzer RB: The origin of muscle fasciculations and cramps. Muscle Nerve 17:1243-1249, 1994. 94. Ljubin C: A modern representation of neuromuscular excitability in the form of intesity-duration curve or line. Electromyogr Clin Neurophysiol 33:341-346, 1993. 95. Lorente de No R: A study of nerve physiology. In: Studies from the Rockefeller Institute for Medical Research, Vols 131 and 132. New York, Rockefeller Institute, 1947. 96. Low PA, Mcleod JG: Refractory period, conduction of trains of impulses, and effect of temperature on conduction in chronic hypertrophic neuropathy: Electrophysiological studies on the trembler mouse. J Neurol Neurosurg Psychiatry 40:434-447, 1977. 97. Lowitzsch K, Gohring U, Hecking E, Kohler H: Refractory period, sensory conduction velocity and visual evoked ptentials before and after
98.
99.
100. 101.
102.
103. 104. 105.
106.
107.
108. 109.
110.
111.
112. 113. 114.
235 haemodialysis. J Neurol Neurosurg Psychiatry 44:121-128, 1981. Lowitzsch K, Hopf HC: Refractory periods and propagation of repetitive mixed nerve action potentials in severe and mild neuropathy. In Hausmanowa-Petrusewicz I, Jedrzejowska H (eds): Structure and Function of Normal and Diseased Muscle and Peripheral Nerve. Proceedings of the Symposium (Kazimierz upon Vistula, Poland) Polish Medical Publisher, 1972. Lowitzsch K, Hopf HC, Schlegel HJ: Conduction of two or more impulses in relation to the fibre spectrum in the mixed human peripheral nerve. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 2. Karger, Basel, 1973, pp 272-278. Mackel RG, Brink EE: Accommodation in single human nerve fibers in vivo. Muscle Nerve 18:469-471, 1995. Maddison P, Newsom-Davis J, Mills KR: Strength-duration properties of peripheral nerve in acquired neuromyotonia. Muscle Nerve 22: 823-830, 1999. Maurer K, Hopf HC, Lowitzsch K: Hypokalemia shortens relative refractory period of peripheral sensory nerves in man. J Neurol 216:67-71, 1977. McComas AJ: Motor unit estimation: Methods, results, and present status. Muscle Nerve 14: 585-597, 1991. McComas AJ: Motor-unit estimation: The beginning. J Clin Neurophysiol 12:560-564, 1995. McDonald WI, Sears TA: The effects of experimental demyelination on conduction in the central nervous system. Brain 93:583-598, 1970. Miller TA, Kiernan MC, Mogyoros I, Burke D: Activity-dependent changes in impulse conduction in a focal nerve lesion. Brain 119:429437, 1996. Milner-Brown HS, Brown WF: New methods of estimating the number of motor units in a muscle. J Neurol Neurosurg Psychiatry 39: 258-265, 1976. Mogyoros I, Kiernan MC, Burke D: Strengthduration properties of human peripheral nerve. Brain 119:439-447, 1996. Mogyoros I, Kiernan MC, Burke D: Strengthduration properties of cutaneous and motor axons in carpal tunnel syndrome. Muscle Nerve 20:508-510, 1997. Mogyoros I, Kiernan MC, Burke D, Bostock H: Ischemic resistance of cutaneous afferents and motor axons in patients with amyotrophic lateral sclerosis. Muscle Nerve 21:1692-1700, 1998. Neto HS, de Carvalho VC, Marques MJ: Estimation of the number and size of human flexor digiti minimi muscle motor units using histological methods. Muscle Nerve 21:112-114, 1998. Ochs S: Elements of Neurophysiology. John Wiley & Sons, Inc., New York, 1965. Paintal AS: Block of conduction in mammalian myelinated nerve fibres by low temperatures. J Physiol (Lond) 180:1-19, 1965. Paintal AS: Effects of temperature on conduction in single vagal and saphenous myelinated
Nerve Conduction Studies
236 nerve fibres of the cat. J Physiol (Lond) 180: 20-49, 1965. 115. Panizza M, Nilsson J, Roth BJ, Basser PJ, Hallett M: Relevance of stimulus duration for activation of motor and sensory fibers: Implications for study of H-reflexes and magnetic stimulation. Electroencephalogr Clin Neurophysiol 85:22-29, 1992. 116. Panizza M, Nilsson J, Roth BJ, Rothwell J, Hallett M: The time constants of motor and sensory peripheral nerve fibers measured with the method of latent addition. Electroencephalogr Clin Neurophysiol 93:147-54, 1994. 117. Panizza M, Nilsson J, Roth BJ, Grill SE, Demirci M, Hallett M: Differences between the time constant of sensory and, motor peripheral nerve fibers: Further studies and considerations. Muscle Nerve 21:48-54, 1998. 118. Quasthoff S: The role of axonal ion conductances in diabetic neuropathy: A review. Muscle Nerve 21:1246-1255, 1998. 119. Quasthoffs, Grosskreuts J, Kuhn M, Schilling T, and Hanauske A: Taxol-cisplatin neuropathy caused by membrane depolarisation: An in vivo and in vitro study. J Neurol 242(suppl 2): 145, 1995. 120. Quasthoff S, Horn S, Grosskreutz J, Grafe P: Effects of ischaemia on threshold electrotonus of peripheral nerve in diabetic patients. J Neurol 242: S51, 1995. 121. Reitter BF, Johannsen S: Neuromuscular reaction to paired stimuli. Muscle Nerve 5:593603, 1982. 122. Ritchie JM: Physiology of axons. In Waxman SG, Stys PK, Kocsis JD (eds): The Axon. Oxford University Press, Oxford, 1995, pp 68-96. 123. Scarfbne H, McComas AJ, Pape K, Newberry R: Denervation and reinnervation in congenital brachial palsy. Muscle Nerve 600-607, 1999. 124. Schilling T, Heinrich B, Kau R, Herzog M, Quasthoff S, Diergarten K, Rastetter J, Hanausk A-R: Paclitaxel administered over 3 h followed by cisplatin in patients with advanced head and neck squamous cell carcinoma: A clinical phase I study. Oncology 54:89-95, 1997. 125. Shahani BT, Jang J, Dhand UK: A new approach to motor unit estimation with surface EMG triggered averaging technique. Muscle Nerve 18:1088-1092, 1995. 126. Shefner JM, Jillapalli D, Bradshaw DY: Reducing intersubject variability in motor unit number estimation. Muscle Nerve 22:14571460, 1999. 127. Slawnych M, Laszlo C, Hershler C: Motor unit estimates obtained using the new "MUESA" method. Muscle Nerve 19:626-636, 1996. 128. Slawnych M, Laszlo C, Hershler C: Motor unit number estimation: Sample size considerations. Muscle Nerve 20:22-28, 1997. 129. Smith KJ: A sensitive method for detection and quantification of conduction deficits in nerve. J Neurol Sci 48:191-199, 1980. 130. Smith KJ, Hall SM: Nerve conduction during peripheral demyelination and remyelination. J Neurol Sci 48:201-219, 1980. 131. Stashuk DW, Doherty TJ, Kassam A, Brown
132.
133. 134.
135.
136.
137. 138.
139. 140.
141. 142. 143.
144.
145.
146.
147.
WF: Motor unit number estimates based on the automated analysis of F-responses. Muscle Nerve 17:881-890, 1994. Stein RB: Measuring the numbers of motor units in muscle during health and disease. In Kimura J, Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Amsterdam, 1996, pp 558-563. Stein RB, Yang JF: Methods for estimating the number of motor units in human muscles. Ann Neurol 28:487-495, 1990. Stephanova DI, Bostock H: A distributedparameter model of the myelinated human motor fibre: Temporal and spatial distributions of action potentials and ionic currents. Biol Cybem 73:275-280, 1995. Stephanova DI, Bostock H: A distributed-parameter model of the myelinated human motor nerve fibre: temporal and spatial distributions of electrotonic potentials and ionic currents. Biol Cybern 74:543-547, 1996. Strupp M, Bostock H, Weigl P, Piwernetz K, Renner R, Grafe P: Is resistance to ischaemia of motor axons in diabetic subjects due to membrane depolarization? J Neurol Sci 99: 271-280, 1990. Stys PK, Ashby P: An automated technique for measuring the recovery cycle of human nerves. Muscle Nerve 13:750-758, 1990. Stys PK, Waxman SG: Activity-dependent modulation of excitability: Implications for axonal physiology and pathophysiology. Muscle Nerve 17:969-974, 1994. Tackmann W, Lehmann HJ: Refractory period in human sensory nerve fibres. Eur Neurol 12:277-292, 1974. Tackmann W, Lehmann HJ: Relative refractory period of median nerve sensory fibres in the carpal tunnel syndrome. Eur Neurol 12:309316, 1974. Tasaki I: Nature of the local excitatory state in the nerve fiber. Jpn J Physiol 126:367-379, 1950. Tasaki I: Nervous Transmission. Charles C Thomas, Springfield, 111, 1953. Vallbo AB: Accommodation related to inactivation of the sodium permeability in single myelinated nerve fibres from Xenopus laevis. Acta Physiol Scand 61:429-444, 1964. Wang FC, Delwaide PJ: Number and relative size of thenar motor units estimated by an adapted multiple point stimulation method. Muscle Nerve 18:969-979, 1995. Waxman SG, Kocsis JD, Brill MH, Swadlow HA: Dependence of refractory period measurements on conduction distance: A computer simulation analysis. Electroencephalogr Clin Neurophysiol 47:717-724, 1979. Weigl P, Bostock H, Franz P, Martius P, Muller W, Grafe P: Threshold tracking provides a rapid indication of ischaemic resistance in motor axons of diabetic subjects. Electroencephalogr Clin Neurophysiol 73:369-371, 1989. Yuen EC, Olney RK: Longitudinal study of fiber density and motor unit number estimate in patients with amyotrophic lateral sclerosis. Neurology 49:573-578, 1997.
Part III
Assessment of Neuromuscular Transmission
This page intentionally left blank
Chapter
9
ANATOMY AND PHYSIOLOGY OF THE NEUROMUSCULAR JUNCTION
1. INTRODUCTION 2. ANATOMY OF THE NEUROMUSCULAR JUNCTION End Plate Synaptic Vesicles Acetylcholine Receptors 3. ELECTRICAL ACTIVITY AT THE END PLATE Miniature End-Plate Potential Events Related to Nerve Action Potential End-Plate Potential 4. EXCITATION-CONTRACTION COUPLING Generation of Muscle Action Potential Transverse and Longitudinal Tubules and Triad Role of Calcium Ions 5. ABNORMALITIES OF NEUROMUSCULAR TRANSMISSION Postsynaptic Defect in Myasthenia Gravis Experimental Models in Animals Presynaptic Defect in Lambert-Eaton Myasthenic Syndrome Heterogeneous Pathophysiology of Congenital Myasthenic Syndromes Effect of Toxins and Chemicals 6. TIME COURSE OF NEUROMUSCULAR TRANSMISSION Enhanced Excitability Causing Repetitive Discharges Effects of Paired or Repetitive Stimulation Neuromuscular Depression and Facilitation Normal Recovery Cycle Effects of Disease States Posttetanic Potentiation and Exhaustion
1 INTRODUCTION The neuromuscular junction is a synaptic structure consisting of the motor nerve terminal, junctional cleft, and muscle end
plate. Its chemical mode of transmission has properties that are fundamentally different from the electrical propagation of impulses along the nerve and muscle. For example, the release of acetylcholine (ACh) ensures unidirectional conduction 239
240
Assessment of Neuromuscular Transmission
from the axon terminal to the muscle end plate. The same principle applies to synaptic transmission in a sequence of neurons. In contrast, the nerve axons conduct an impulse bidirectionally from the point of stimulus unless the impulse originates at the cell body or axon terminal, as expected for any physiologic activation. The muscle fibers also show bidirectional propagation of impulses initiated at the motor point. Other characteristics common to the nerve synapse and neuromuscular junction include synaptic delay of a fraction of a millisecond and the nonpropagating nature of end-plate potentials (EPPs). These local potentials cause no refractoriness, unlike the all-or-none response of the nerve or muscle action potential. The graded responses summate temporally as well as spatially after subliminal stimuli, thereby providing greater flexibility and adaptability. As in a synapse, the mobilization store must continuously replenish the liberated transmitters. Otherwise, the neuromuscular junction would fail, with depletion of immediately available molecules. This section provides a simplified overview of the complex physiology in preparation for a subsequent more detailed clinical discussion (see Chapter 27). The presynaptic ending contains many minute vesicles, each with up to 10,000 ACh molecules. At rest, these vesicles randomly migrate into the junctional cleft. At the muscle end plate, they produce small depolarizations of the postsynaptic membrane. These miniature EPPs (MEPPs) do not attain the critical level for generation of a muscle action potential. Depolarization of the presynaptic ending at the axon terminal triggers an influx of calcium (Ca2+), initiating the calcium-dependent release of immediately available vesicles into the junctional cleft. The greatly enhanced and synchronized ACh activity gives rise to a nonpropagated EPP from summation of multiple MEPPs. When the EPP exceeds the excitability threshold of the muscle cell, opening of the voltagedependent sodium (Na+) channels leads to the generation of an action potential. Propagation of the muscle potential activates the contractile elements through excitation-contraction coupling.
2
ANATOMY OF THE NEUROMUSCULAR JUNCTION
End Plate Nerve and muscle become dependent on each other during the course of embryogenesis. The formation of the neuromuscular junction follows differentiation of presynaptic nerve terminals, innervation of postsynaptic components, and elimination of the remaining multiple axons.112 The name motor end plate originally implied the specialized efferent endings that terminate on a striated muscle as a whole. Most authors, however, now use the term to describe the postsynaptic membrane of the muscle alone. Each muscle fiber usually has only one end plate, and each branch of a motor axon innervates one end plate. The motor nerve fiber loses the myelin sheath at the nerve terminals. Distal to the myelin sheath, therefore, only the Schwann cells separate the nerve terminals from the surrounding tissue. Thus, the neuromuscular junction consists of the motor nerve ending, Schwann cell, and muscle end plate (Fig. 9-1). At the junctional region the nerve ending also loses the Schwann cells, forming a flattened plate lying within a surface depression of the end plate. This indentation of the muscle fiber, called a synaptic gutter or a primary synoptic cleft, measures about 200-500 A deep. The thickened postsynaptic membrane in this region has narrow infoldings called junctional folds or secondary clefts. A large number of mitochondria, nuclei, and small granules accumulate close to the secondary clefts. Many mitochondria and synaptic vesicles also lie in the axon terminals, just proximal to the presynaptic membrane. Electron-microscopic studies have delineated the ultrastructural features of the end plates in human external intercostal muscles.45 The presynaptic nerve terminal contains clear, round synaptic vesicles, lying mostly clustered in the regions called active zones, where acetylcholine (ACh) release into the synaptic cleft takes place. On average, a nerve terminal that
Anatomy and Physiology of the Neuromuscular Junction
241
Figure 9-1. Motor end plate as seen in histologic sections in the long axis of the muscle fiber (A) and in surface view (B) under the light microscope, and a section through the motor end plate (area in the rectangle in A) under the electron microscope (C). The myelin sheath ends at the junction at which the axon terminal fits into the synaptic cleft. The Schwann (teloglial) cells cover the remaining portion without extending into the primary cleft. The plasma membrane of axon (axolemma) forms the presynaptic membrane and that of muscle fiber (sarcolemma), the postsynaptic membrane of the end plate. Interdigitation of the sarcolemma gives rise to the subneural or secondary clefts. The axon terminal contains synaptic vesicles and mitochondria. [From Bloom and Fawcett,15 with permission.]
occupies an area close to 4 um2 contains approximately 50 synaptic vesicles per square micrometer. The synaptic basal lamina interposed between the nerve terminal and the muscle cell has a special composition containing, among other molecules, acetylcholinesterase.91 The postsynaptic membrane, 10 times longer than the presynaptic membrane, forms elaborate imaginations known as junctional folds, containing a concentration of ACh receptors.109 The postsynaptic folds cover an area about two and a half times that of the terminal itself. Diseases of neuromuscular transmission alter the end-plate profile (Fig. 9-2). In myasthenia gravis, the terminal occupies less area, and postsynaptic folds appear simplified. In contrast, in the myasthenic30,78 syndrome or Lambert-Eaton syndrome the termi-
nal, though normal in area, contains an elongated and sometimes markedly hypertrophic postsynaptic membrane. Neither disease is characterized by significant alteration in the mean synaptic vesicle diameter or the mean synaptic vesicle count per unit nerve terminal area. Clinically unaffected limb muscles may show the ultrastructural changes of the motor end plate in117patients with ocular myasthenia gravis.
Synaptic Vesicles Minute intracellular structures, 300-500 A in diameter, encapsulate ACh molecules inside the presynaptic axoplasm. In addition to the synaptic vesicles, the nerve endings contain high concentrations of choline
Assessment of Neuromuscular Transmission
242
Figure 9-2. Schematic representation of the motor endplates in control, myasthenia gravis, and myasthenic syndrome drawn to the scale of the mean figure. The diagram shows an oversimplification of the postsynaptic membrane in myasthenia gravis and marked hypertrophy in myasthenic syndrome. [From Engel and Santa,44 with permission.]
acetyltransferase, which synthesizes ACh, and acetylcholinesterase, which hydrolyzes ACh. The proximal portions of neurons also possess the neurotransmitter and the two enzymes, although to a much lesser extent. This finding suggests that enzymatic synthesis takes place in the cell body 86before transport to the nerve terminals. Each vesicle contains 500010,000 molecules of ACh or a quantum.64 Some quanta (about 1000) located adjacent to the cell membrane are immediately available for release; many more (10,000), contained in the mobilization store, move toward the membrane to continuously replace liberated ACh. The remaining and largest portion of quanta (300,000) forms the main store as a reserve supply for the mobilization store.
Acetylcholine Receptors The nicotinic acetylcholine receptor, a transmembrane glycoprotein, comprises five subunits, a x 2, b, y, and 5 in the fetus and a x 2, b, e, and d in the adult, forming an ion channel. Binding of two ACh molecules to two specific sites of a subunits opens the ACh channel, allowing cations to flow through the postsynaptic membrane, with the net result of depolarization.88 Patch-clamp studies have shown bursts of ACh channel activation alternating open intervals and brief closures.20 Synaptic maturation with switching from
the y to the e subunit results in change of channel open time, and consequently, conductance. Studies of the kinetic properties of the normal ACh receptor94 help elucidate pathologic alterations seen in47 some congenital myasthenic syndromes.
3
ELECTRICAL ACTIVITY AT THE END PLATE Miniature End-Plate Potential
Many resting muscle fibers show a spontaneous subliminal electrical activity, miniature end-plate potential (MEPP). It represents a small depolarization of the postsynaptic membrane induced by sustained but random release of a single quantum of acetylcholine (ACh) from the nerve terminal.49 An ordinary needle electrode placed near the end plate of the muscle fibers can record the MEPP (see Chapter 13-4). A microelectrode inserted directly into the end-plate region achieves a higher resolution for quantitative analysis. Each ACh quantum liberated from the nerve terminal contains a nearly equal number of ACh molecules, irrespective of external factors such as temperature or ionic concentration. In contrast, the frequency of the MEPP varies over a wide range. It increases with elevated temperatures and upon depolarization of the motor nerve terminals. It decreases with de-
Anatomy and Physiology of the Neuromuscular Junction ficiency of calcium (Ca2+), the ion known to enhance quantal release by increasing fusion of the ACh vesicles with the membrane of the nerve terminal. The factors that dictate the amplitude of the MEPP or quantum size include the number of ACh molecules in a vesicle, diffusion properties of the liberated molecules, structural characteristics of the end plate, and sensitivity of the ACh receptors. In normal human intercostal muscles, an MEPP recurs roughly every 5 seconds, measuring approximately 1 mV in amplitude when recorded intracellularly.32 Hence, the MEPP falls far short of the excitability threshold of the muscle fiber, averaging about 2-4 percent of the normal end-plate potential (EPP) generated by a volley of nerve impulses. A small dose of curare greatly reduces the amplitude of the MEPP, whereas an equivalent dose of neostigmine (Prostigmin) increases it.73 The MEPP ceases after denervation or nerve anesthesia. In myasthenia gravis, receptor insensitivity results in reduced amplitude of the MEPP, despite normal discharge frequency. Conversely, defective release of ACh reduces the rate of firing in the myasthenic syndrome and in botulism, although the MEPP remains normal in amplitude (see Chapter 27-2 and 3).
Events Related to Nerve Action Potential In the resting state, the interior of the muscle fibers is negative relative to the exterior by about 90 mV. This transmembrane potential primarily results from an unequal distribution of inorganic ions across the membrane, with +a high concentration of potassium (K ) intracellularly and of sodium (Na+) and chloride (Cl-) extracellularly (see Chapter 2-2). It also depends on differential permeability across the muscle membrane, with a high conductance for potassium and chloride and low conductance for sodium. The energy-dependent sodium-potassium pump compensates for a slight inward movement of sodium and outward movement of potassium at steady state to maintain the electrochemical potential equilibrium (see Fig. 2-1).
243
As mentioned earlier, spontaneous release of a single quantum of ACh induces a MEPP that falls far below the critical level necessary for generation of a muscle action potential. With the arrival of a nerve impulse, depolarization of the motor nerve ending initiates an influx of calcium into the motor axons. The increased amount of calcium accelerates fusion of the vesicle membrane with the nerve terminal membrane, thereby producing a large increase in the rate of quantal release. Massive synchronized release of ACh triggered by the arrival of a nerve action potential results in summation of many MEPPs, giving rise to a localized EPP. Thus, the number of immediately available ACh quanta and the voltage-dependent concentration of calcium within the axon terminal, together, determine the size of the EPP. The number of quanta emitted per nerve impulse, or quantum content, averages 25-50, based on the amplitude ratio, EPP/MEPP.
End-Plate Potential Like MEPPs, EPPs result from depolarization of the motor end plate by ACh. The opening of ACh receptors by the synaptic transmitter increases the conductance of various diffusible ions, principally those of sodium and potassium. Therefore, these ions move freely down their electrochemical gradients, resulting in depolarization of the motor end plate. The rise time, amplitude, and duration characterize this nonpropagated local response, which declines rapidly with distance from the end plate. It normally begins about 0.5 ms after the release of ACh, reaches its peak in about 0.8 ms, and decreases exponentially with a half decay time of about 3.0 ms. The EPP, a graded, rather than all-or-none, response, increases in proportion to the number of ACh quanta liberated from the nerve terminal. The sensitivity of the end plate to the depolarizing action of ACh also affects the degree of depolarization. Like the excitatory post-synaptic potential (EPSP), two or more subthreshold EPPs generated in near synchrony can summate to cause a depolarization exceeding the critical level for generation of an action potential.
Assessment of Neuromuscular Transmission
244
4 EXCITATION-CONTRACTION COUPLING
Generation of Muscle Action Potential An end-plate potential (EPP) exceeding the threshold or the critical level of depolarization to open the sodium channel generates an all-or-none muscle action potential. A molecular change of the depolarized membrane results in selective increase of sodium conductance, followed by an increase in potassium conductance. As long as depolarization reaches the critical value, this phenomenon, inherent in the muscle membrane, occurs irrespective of the nature of the stimulus. In contrast to the all-or-none characteristic of the amplitude dictated by sodium channel kinetics, the latency of the action potential changes depending on the speed of initial depolarization. This variability forms the source of jitter in single-fiber studies (see Chapter 16-5), which serves as a sensitive measure of subtle alteration of end plate. For example, even healthy muscle shows reversible changes of neuromuscular transmission after a period of disuse.56 Once generated at the end plate, the action potential propagates bidirectionally to the remaining parts of the fiber. The impulse conducts only in the range of 3-5 m/s along the muscle membrane, compared with 60 m/s over the nerve (see Chapter 12-2). A neuromuscular block results when the EPP fails to reach the critical level. A subliminal EPP may imply insufficient liberation of acetylcholine from the axon terminal or reduced sensitivity of the muscle end plate. In contrast to the all-or-none generation of a muscle action potential in each muscle fiber, the compound muscle action potential shows a graded response in proportion to the number of activated muscle fibers.
Transverse and Longitudinal Tubules and Triad The spread of action potential from the motor end plate to the transverse tubules initiates muscle contraction. This process, called excitation-contraction coupling, links
electrical and mechanical activity.77,102 Electrical activity of a muscle fiber consists of two temporally separate components attributable to different structures within the fiber.22 The first portion originates at the motor end plate and spreads along the outer surface of the muscle fiber. The second part occurs within a complex tubular system that surrounds and interpenetrates the muscle fiber. This network, called the transverse tubules because of its orientation relative to the axis of the muscle fiber, lies at the junctions of the A and I bands in humans (see Fig. 12-1). These tridimensional tubules, though structurally internal to the cell, contain extracellular fluid. Consequently, the inside of the tubule is electropositive relative to the outside, surrounded by intracellular fluid. Muscle action potentials propagate along the tubules into the depth of the muscle. A second tubular system, called the longitudinal tubule or sarcoplasmic reticulum, surrounds the myofibrils of a muscle fiber (Fig. 9-3). These tubules have a longitudinal orientation with respect to the myofibrillar axis and, unlike transverse tubules, form a closed system devoid of continuity with either extracellular fluids or sarcoplasm. They appear as fenestrated sacs surrounding the myofibrils. The longitudinal tubules expand to form bulbous terminal cisterns on both sides of the transverse tubules, where they come into close contact. The two terminal cisterns and one interposed transverse tubule form a triad in longitudinal sections of the muscle.
Role of Calcium Ions Propagated action potentials invade the muscle fibers along the transverse tubules to come into contact with the terminal cisterns of the longitudinal tubules at the triad. This coupling to the sarcoplasmic reticulum gives rise to a small electrical potential referred 19 to as intramembranous charge movement The action potential crossing the terminal cistern initiates the release of calcium (Ca2+) from the longitudinal tubules into the sarcoplasm that surrounds the myofilaments. The presence of calcium there triggers a chemical interaction that leads to the formation of bridges between thin and thick filaments.
Anatomy and Physiology of the Neuromuscular Junction
245
Figure 9-3. Anatomic relationship between the perpendicularly oriented longitudinal and transverse tubules. Propagating muscle action potentials initiate electromechanical coupling at the triad of the reticulum, which consists of two terminal cisterns of the longitudinal tubules and one transverse tubule between them. [From Bloom and Fawcett,15 with permission.]
Sliding of thin filaments against thick filaments results in contraction of the myofibril (see Chapter 12-2). At the end of the muscle action potential, rapid resequestering of calcium into the longitudinal tubules lowers its concentration in the sarcoplasm. The myoflbers relax as adenosine triphosphate breaks the existing bridges between filaments.
5
ABNORMALITIES OF NEUROMUSCULAR TRANSMISSION
Postsynaptic Defect in Myasthenia Gravis In myasthenia gravis (see Chapter 27-2) intracellular recordings from the inter-
costal muscles have revealed reduced amplitude of miniature end-plate potential (MEPP) or small quantum size but normal or nearly normal discharge frequency.32 Consequently, the end-plate potential (EPP) elicited by a nerve impulse also shows a reduced amplitude, despite a normal number of acetylcholine (ACh) quanta liberated by a single volley or normal EPP quantum content. On repetitive stimulation, the number of quanta released falls gradually, as it does in normal muscle, causing a further decrease in the amplitude of the initially small EPP. With successive stimuli, the EPP becomes insufficient to bring the membrane potential to the critical level in a progressively greater number of fibers, thus causing reduction in amplitude of compound muscle action potential. Neuromuscular transmission fails first in small motor units, perhaps
246
Assessment of Neuromuscular Transmission
because they have a lower margin of safety than the large motor units.69,70 Reduction in amplitude of MEPP suggests (1) decreased numbers of ACh molecules per quantum, (2) diffusional loss of ACh within the synaptic cleft, or (3) reduced sensitivity of the ACh receptor. In early studies, postsynaptic sensitivity to carbachol and decamethonium added to the bath solution appeared to be normal.33 A presynaptic abnormality proposed on the basis of this finding, however, has subsequently received neither morphologic nor electrophysiologic confirmation. Indeed, micro-iontophoretic application of ACh at the end-plate region has since disclosed impaired postsynaptic sensitivity to ACh.1 The observed electrophysiologic changes may also imply diffusional ACh loss resulting from alterations in postsynaptic membrane structure. Ultrastructural histometric studies in myasthenic intercostal muscles have shown a distinct end-plate profile, indicating postsynaptic membrane abnormalities.45 Another experiment has revealed three types of neuromuscular junctions in the surface fibers of internal and external intercostal muscles of myasthenics.1 One group with mild morphologic alterations had EPPs of sufficiently large amplitude to trigger an action potential. A second group with a grossly altered postjunctional membrane showed marked reduction not only in amplitude but also in frequency of the MEPP and in amplitude of the EPP. The last group had totally degenerated endplates showing neither MEPPs nor EPPs. Not every myasthenic end plate shows morphologic alterations, despite diminished MEPP amplitude demonstrated uniformly. Therefore, changes in end-plate geometry per se may not totally explain the physiologic defect. Myasthenic muscles have decreased functional receptor sites detected by radioactively labeled alpha-bungarotoxin, a snake venom that binds to the ACh receptor.31,48,57 Further, the number of functional ACh receptors, when counted by this technique, shows positive correlation 66with the mean amplitude of the MEPP. These findings indicate the presence of an ACh receptor ab-
normality in myasthenia gravis. Partial blocking of the ACh receptors with curare produces a similar physiologic defect. Studies using plasma exchange have revealed an inverse relationship between clinical muscle strength and antibody titers. This finding supports the view that the auto antibody against nicotinic acetylcholine receptor plays the most important role in impairing neuromuscular transmission in myasthenia gravis and experimental autoimmune myasthenia gravis.4,28,55,96,97,126 Cytokines produced by CD4+ and CD8+ T helper cells mediate the production of anti-ACh receptor antibodies,125 sometimes induced by an external stimulus.7 ACh receptor subunits found in the thymus alone, however, do not produce myasthenia gravis.71 Antibodies mediate obstruction of the ACh receptor, presumably by binding with complement to the receptor zone of the postsynaptic membrane.40 Intercostal muscle biopsies show reduced numbers of ACh receptors and binding of antibodies to many of the remaining receptors in patients with myasthenia gravis.84 Patients with thymoma often have striational antibodies in addition to 67 anti-acetylcholine receptor antibodies. This may interfere with calcium (Ca2+) release from the sarcoplasmic reticulum, resulting in a defect of excitation-contraction coupling and contractility reported in myasthenic muscle.100,101 Autoantibody also appears to mediate seronegative myasthenia gravis, a heterogeneous disorder that can be passively transferred to mice.17
Experimental Models in Animals Experimental autoimmune myasthenia gravis shares the morphologic and physiologic abnormalities of the disease in humans 28,29,46,51,61,83,105,107 Studies in rats
showed reduced receptor content and increased receptor-bound antibody. Thus, defective neuromuscular transmission seems to result from a reduced number of fully active receptors.84 Typical histologic and electrophysiologic myasthenic features develop in mice after passive transfer of human serum fractions obtained from patients with myasthenia gravis.116
Anatomy and Physiology of the Neuromuscular Junction Decamethonium causes paralysis by persistent depolarization of the end-plate region in normal muscle.93 Reduction of postsynaptic sensitivity to depolarization renders myasthenic muscles resistant to this type of neuromuscular blocking. Antibodies to the ACh receptor do not impair the ionophore, an ion-conductance modulator protein thought to control the permeability change following a reaction of ACh with ACh receptor. Experimental autoimmune myasthenia gravis improves by administration of dantrolene sodium, which induces accumulation 115 of free calcium in the subcellular store. Intracellular recordings from muscle end plates of immunized rabbits show reduced amplitude of MEPPs but a normal number 35 of ACh quanta released per nerve impulse. Rats with chronic experimental myasthenia have reduced amplitude of MEPPs despite normal ACh output at rest and during stimulation.74 After passive transfer of human myasthenia gravis to rats, reduction of MEPP amplitude does not develop immediately, but occurs after the first 24 hours, reaching minimum levels by 6 days.63 The delayed development of reduced MEPP amplitude suggests a more complex mechanism by IgG antibodies63 than a simple block of ACh receptors like that caused by curare. Similarly, despite a precipitous drop of antibody titers, electrophysiologic findings usually improve with a delay of at least 7 days from the start of plasmapharesis in men.18
Presynaptic Defect in LambertEaton Myasthenic Syndrome In the Lambert-Eaton myasthenic syndrome (see Chapter 27-3), myasthenia of skeletal muscles and autonomic symptoms result from an autoimmune mechanism against the voltage-gated calcium channel located in the motor nerve terminal79,82,98,103,111 and parasympathetic nerve.62,121,122 In contrast to the receptor insensitivity of myasthenia gravis, defective release of ACh quanta 30characterizes the myasthenic syndrome. Microelectrode recordings from excised intercostal muscles reveal no abnormality in ampli-
247
tude of the MEPPs, and consequently in quantum size, or the sensitivity of the muscle end plate to ACh. The discharge frequency of the MEPP, however, does not increase as expected in response to depolarization of the motor nerve terminal.78 Thus, a single nerve impulse releases a smaller number of ACh quanta than normal or decreased quantum content. The EPP then fails to trigger an action potential in some muscle fibers, which leads to a reduced amplitude of the compound muscle action potential.80 The defect improves immediately with various 34 maneuvers to prime the nerve terminals. For example, the EPP augments progressively with repetitive stimulation of the nerve. Postexercise augmentation lasts longer after cooling, which 2+ reduces the rate of removal of calcium (Ca ) from the nerve terminal.87 An increase of external calcium or the addition of quinidine also enhances the EPP. These findings suggest a normal number of quanta available in the presynaptic store, despite a low probability of quantum release at the nerve terminal. Indeed, ultrastructural studies have revealed no alteration in the mean nerve terminal area or in the synaptic vesicle count per unit.45
Heterogeneous Pathophysiology of Congenital Myasthenic Syndromes Congenital myasthenic syndromes result from different39 types of pre- or postsynaptic mechanisms caused by one or more specific genetic abnormalities (see Chapter 27-4).8,37,43,53,54,72,108 They comprise a number of myasthenic disorders not associated with detectable anti-ACh receptor antibodies. These entities, presenting at birth or in early life, share many common clinical features, despite distinct etiologies identified by physiologic, ultrastructural, and cytochemical studies. Typical patients have such features as deficient muscle acetylcholinesterase, decreased frequency but normal amplitude of the MEPP, decreased number of quanta liberated per nerve impulse, small nerve terminals, and focal degeneration of the postsynaptic membrane. In some types, a low number
Assessment of Neuromuscular Transmission
248
of quanta released per EPP primarily reflects a reduced store of ACh vesicles, rather than a low probability of release, as in the case of the classic myasthenic syndrome. A congenital defect in the molecular assembly of acetylcholinesterase or its attachment to the postsynaptic membrane also represents a basic abnormality. A familial congenital myasthenic syndrome shows deficient synthesis of ACh.60 The syndromes adequately characterized to date include acetylcholinosterase deficiency,36,65 defective resynthesis or vesicular packaging of ACh,38,95 ACh receptor deficiency such as congenital paucity of secondary synaptic clefts,81,110,124 kinetic dysfunction of ACh receptor, such as slow channel syndrome,52,99 high-conductance, fast channel syndrome42,47 and other abnormalities of interaction with ACh,118 and familial limb-girdle myasthenia with tubular aggregates.50,92
lar blocking agents can also cause prolonged muscle weakness.6 Aminoglycoside antibiotics such as neomycin and kanamycin not only interfere with ACh release directly3,75 but also inhibit the transmission by postsynaptic block.24 A number of other drugs induce dysfunction of the neuromuscular junction.3 These include the HIV protease inhibitor ritonavir,104 D-penicillamine, used to treat rheumatoid arthritis27 and Wilson's disease,2 21and cocaine.9,23 In addition, (368 blockers ' and calcium channel blockers113,119,123 may aggravate myasthenia gravis or induce a myasthenic syndrome.
Effect of Toxins and Chemicals
The amount of acetylcholine (ACh) in the immediately available store and the concentration of calcium (Ca2+) at the nerve terminal, together, determine the number of ACh molecules released by a nerve action potential. Single nerve shocks may excite muscle fibers twice or, rarely, three times or more if enough ACh molecules remain after the first discharge, as in congenital myasthenia 39 with acetylcholinesterase deficiency (see Chapter 27-4 and 6). Excess amounts of ACh may result from the use of anticholinesterase as therapy for myasthenia gravis37 or after organophosphate poisoning.10-13 Reactivation of muscle response results, despite the normal amounts of ACh molecules, in the slow channel syndrome with prolonged depolarization.41,99 In this entity, as in organophosphate poisoning, repetitive stimulation of the nerve show a rate-dependent decrement of all muscle potentials, although secondary responses diminish first.58,59,120
2+
Abnormalities in calcium (Ca )-dependent ACh release also reduce the amplitude of the EPP in a number of other conditions, including a neuromuscular block by botulinum toxin89,106 (see Chapter 27-5 and 6). The neuromuscular insufficiency in botulism results neither from blockage of calcium entry into the nerve nor from reduced storage of ACh vesicles. The toxin interferes with the ACh release process itself, by blocking exocytosis at the release sites by cleaving synaptic protein 25 (see Chapter 27-5). Thus, the reduced frequency of the MEPP, not affected by the addition of calcium, recovers after the administration of a spider venom known to neutralize the toxin. High concentrations of magnesium (Mg2+16,114 ) block neuromuscular transmission. Lowering the temperature increases transmitter release and reactivates previously paralyzed muscle in botulinum paralysis, but not in normal muscle blocked by high magnesium concentration.85 Experimental evidence indicates an inhibitory effect of manganese (Mn2+) on transmitter release at the neuromuscular junction.5 The long-term use of various nondepolarizing neuromuscu-
6
TIME COURSE OF NEUROMUSCULAR TRANSMISSION
Enhanced Excitability Causing Repetitive Discharges
Effects of Paired or Repetitive Stimulation Repetitive stimulation affects the release of ACh and the end-plate potential (EPP)
Anatomy and Physiology of the Neuromuscular Junction in two opposing manners. On the one hand, the first shock utilizes a portion of the store, partially depleting the amount of ACh available for subsequent stimuli, until the mobilization store has refilled the loss. On the other hand, calcium accumulates in the nerve terminal after each shock, enhancing ACh release. These two competing phenomena, though initiated by the same stimulus, follow different time courses.26 Influx of calcium into the terminal axons takes place immediately after depolarization of the nerve, but the ion diffuses out of the axon over the next 100-200 ms. Hence, paired or repetitive stimulation with a shorter interstimulus interval causes accumulation of calcium. Such fast rates of stimulation, therefore, tend to facilitate release of ACh, despite concomitant reduction of its immediately available store. In contrast, slower rates of repetition result in suppression, because the negligible electrosecretory facilitation at such stimulus intervals can no longer compensate for the loss of ACh stores. The dichotomy between the fast and slow rates of stimulation, however, does not always hold. For example, even at high rates of stimulation, ACh depletion far exceeding its mobilization will lead to reduced release of the transmitter. The partially depleted ACh store recovers exponentially in 5-10 seconds through the slow reloading of ACh ejection sites. Neuromuscular Depression and Facilitation Reduction in the number of ACh quanta released by the second nerve impulse results in a smaller EPP, which no longer reaches the threshold in some muscle fibers. The amplitude of the second compound muscle action potential decreases, or shows a decrement, compared with the first response. Conversely, an increase in the number of quanta released by the second nerve impulse gives rise to a larger EPP. Such true facilitation is based on the neurosecretory potentiation rather than on summation of two EPPs elicited by paired shocks with a very short interstimulus interval.26
249
Both facilitation and summation result in larger compound muscle action potentials through recruitment, provided that the initial stimulus failed to activate all the muscle fibers. The greater amplitude and area under the waveform in recruitment imply the discharge of additional muscle fibers. An increased amplitude may also result from better synchronization of different muscle fibers without recruitment. In this phenomenon, called pseudofactiitation, the area under the waveform, which approximates the number of active muscle fibers, shows no major changes. Increased+ activation of the electrogenic sodium (Na )-potassium (K+) pump triggered by preceding shocks also potentiates the amplitude of the subsequent single action potentials as the result of hyperpolarization.90 Normal Recovery Cycle Studies of the recovery cycle consist of recording the muscle action potentials after delivering paired stimuli to the nerve at various interstimulus intervals. A second shock delivered a few milliseconds after the first falls in the refractory periods of the muscle and nerve (Fig. 9-4). For intervals of 10-15 ms, an overlap between the first and second muscle responses precludes accurate measurement of the individual potentials. Thereafter, the second compound muscle potential recovers to the size of the first in the normal muscle. This finding, however, does not necessarily imply that the first and second stimuli elicit the same EPPs. At interstimulus intervals of 100-200 ms, the second shock may normally evoke a greater EPP than the first through neurosecretory potentiation. If the EPP by the first stimulus exceeds the threshold of excitation in all muscle fibers, however, enhanced EPP by the second stimulus recruits no additional fibers. A slow rate of stimulation depresses the number of ACh quanta released successively, even in normal muscles. Because of a large margin of safety, however, the decreased amount of ACh suffices to cause an EPP well above the critical level of excitation in all muscle fibers. In normal muscles, therefore,
Assessment of Neuromuscular Transmission
250
Figure 9-4. Compound action potentials from the thenar muscles elicited by paired shocks delivered to the median nerve at the wrist. Time intervals ranged from 2 to 30 ms between conditioning (arrow) and test stimuli. The top tracing on the left shows a response to a single stimulus. The bottom tracing on the right is a composite picture superimposing 20 paired responses. The conditioning response of each pair appeared in the same spot, whereas the test responses shifted to the right in proportion to the interstimulus interval. An imaginary line connecting the peaks of the sequential test responses represents the time course of neuromuscular excitability change following the conditioning stimulus. [From Kimura,76 with permission.]
changes in the amount of ACh do not alter the size of compound muscle action potential elicited by the second or subsequent stimuli.
Effects of Disease States Partially curarized mammalian muscle, with a reduced margin of safety, serves as a good model for25studying the recovery cycle of the EPP. With paired stimuli, the second muscle response equals or exceeds the first for the interstimulus intervals of 100-200 ms that accompany calcium-dependent neurosecretory facilitation.14,25 With longer intervals, the second response falls below the first, because depleted stores of available ACh quanta can no longer overcome the receptor insensitivity. The maximal depression at interstimulus intervals ranging from 300 to 600 ms is followed by a slow recovery. Full return to the control value in about 10 s implies restoration of releasable ACh through replenishment of the stores. In myasthenia gravis, a reduced amount of ACh also fails
Figure 9-5. Composite pictures superimposing 20 paired responses from the thenar muscles (compare bottom tracing on right in Figure 9-4.) A patient with myasthenia gravis (A) and a normal control (B) showed the same recovery course for the interstimulus intervals ranging from 1 to 30 ms.
Anatomy and Physiology of the Neuromuscular Junction
251
to activate some muscle fibers with receptor insensitivity. Hence, the recovery cycle of the muscle action potential shows a great resemblance to that of curarized muscle (Figs. 9-5 through 9-8). In either case, the maximal depression results from repetitive stimulation at 2-3 Hz, the rate fast enough for the depletion of ACh but slow enough for the diffusion of calcium out of the axon. In the myasthenic syndrome, characterized by a defective release of ACh, the EPP elicited by a single stimulus falls short of activating many muscle fibers. With the second stimulus given in less than a few milliseconds, the summated EPPs will recruit additional muscle fibers. With stim-
Figure 9-7. Composite pictures similar to those shown in Figures 9-5 and 9-6. Unlike the previous tracings, both conditioning and test stimuli consist of paired shocks with interstimulus intervals of 10 ms. The paired test stimuli followed the paired conditioning stimuli (open arrow) by the interval of 200-400 ms. The double-peaked conditioning responses appeared in the same spot of each tracing (paired arrows). The second peak of the pair, though displaced downward, had the same amplitude as the first. In myasthenia gravis (A), depletion of acetylcholine (ACh) by the conditioning stimuli reduced the first peak of each test response. The second peak of each test response, elicited 10 ms after the first, recovered to a normal level indicating the summation of the two endplate potentials (EPP). In each test response of the normal muscle, the maximal size of the first peak precluded any amplitude increase of the second peak.
Figure 9-6. Composite pictures of 16 paired responses from the thenar muscles arranged in the same manner as in Figure 9-5. The interstimulus intervals of paired shocks ranged from 30 to 400 ms. The conditioning response of each pair appeared in the same spot of each tracing (arrows pointing down), whereas the test responses shifted to the right successively. The test response showed a mild but definite reduction in amplitude at the interstimulus intervals of 150-250 ms in the myasthenic muscle (A), but not in the normal muscle (B).
uli delivered at a longer interval of 100-200 ms, EPPs no longer summate, but the electrosecretory facilitation partially overcomes the defective release of ACh (Fig. 9-8). Increased EPPs will in turn recruit some of the muscle fibers not activated by the first stimulus, which will lead to an increase in amplitude of the second compound muscle action potential.26 This finding, though characteristic, reveals only a nonspecific abnormality seen whenever the first stimulus evokes less than the maximal response, including some cases of
Assessment of Neuromuscular Transmission
252
Decrement: Reduced Quantum Content Below Safety Margin. Increment: Neurosecretory Potentiation (Ca++ dependent?) Figure 9-8. Typical changes in quantum size and quantum content as determined by intracellular recordings in myasthenia gravis (MG) and myasthenic syndrome (MS). The compound muscle action potential show a decrement to repetitive nerve stimulation with dropout of individual muscle fibers and an increment with recruitment of additional fibers.
myasthenia gravis (Fig. 9-7). At a slower traction, the immediately available store rate separated by more than 200 ms, the of ACh may increase as a result of a greater mobilization rate. This increase of second EPP diminishes because calcium no longer accumulates to compensate for ACh storage, coupled with the accumulation of calcium in the axon, enhances the depletion of available ACh stores. Limited release of ACh and, consequently, the EPP release of ACh by the first stimulus, however, may preclude major decremental for 1-2 minutes, causing posttetanic potentiation. Subsequent stimuli release muscle responses in most patients. fewer ACh quanta for up to 15 minutes, Defective release of ACh also underlines the electrophysiologic abnormality in botprobably because of metabolic changes in the nerve terminal, leading to posttetanic ulism. With paired stimuli, summation of exhaustion. These findings resemble the the EPPs augments the second response experimentally induced block by hemiat intervals of less than 10 ms. Increased cholinium, which interferes with ACh synnumber of quanta released by the second impulse also causes facilitation at interthesis.25 stimulus intervals of 100-200 ms. As expected, paired shocks of longer intervals REFERENCES usually cause depression of the second response, though not as consistently as in myasthenia gravis. 1. Albuquerque EX, Rash JE, Mayer RF, Satter-
Posttetanic Potentiation and Exhaustion With prolonged repetitive stimulation, or after a sustained voluntary muscle con-
field JR: An electrophysiological and morphological study of the neuromuscular junction in patients with myasthenia gravis. Exp Neurol 51:536-563, 1976. 2. Anlar B, Kuruoglu R, Varli K: Neuromuscular transmission and acetycholine receptor antibodies in penicillamine-treated Wilson's disease patients. Muscle Nerve 19:676, 1996.
Anatomy and Physiology of the Neuromuscular Junction 3. Argov Z, Wirguin I: Drugs and the neuromuscular junction: Pharmacotherapy of transmission disorders and drug-induced myasthenic syndromes. In Lisa RP (ed): Handbook of Myasthenia Gravis and Myasthenic Syndromes. Marcel Dekker, New York, 1994, pp 295-319. 4. Asthana D, Fujii Y, Huston GE, Lindstrom J: Regulation of antibody production by helper T cell clones in experimental autoimmune myasthenia gravis. Clin Immunol Immunopathol, 67:240-248, 1993. 5. Balnave RJ, Gage PW: The inhibitory effect of manganese on transmitter release at the neuromuscular junction of the toad. Br J Pharmacol 47:339-352, 1973. 6. Barohn RJ, Jackson CE, Rogers SJ, Ridings LW, McVey AL: Prolonged paralysis due to nondepolarizing neuromuscular blocking agents and corticosteroids. Muscle Nerve 17:647-654, 1994. 7. Batocchi AP, Evolo A, Servidei S, Palmisani MT, Apollo F, Tonali P: Myasthenia gravis during interferon alfa therapy. Neurology 45:382-383, 1995. 8. Beeson D, Newland C, Croxen R, Newsom-Davis J: Mutations in the muscle acetylcholine receptor a subunit gene in slowchannel congenital myasthenic syndrome (abstract). Ann Neurol 40:487-488, 1996. 9. Berciano J, Oterino A, Rebello M, Pascual J: Myasthenia gravis unmasked by cocaine use (Letter). N Engl J Med 325:892, 1991. 10. Besser R, Gutmann L: A quantitative study of the pancuronium antagonism at the motor endplate in human organophosphorus intoxication. Muscle Nerve 18:956-960, 1995. 11. Besser R, Gutmann L, Dillmann R, Weilemann LS, Hopf HC: End-plate dysfunction in acute organophosphate intoxication. Neurology 39: 561-657, 1989. 12. Besser R, Vogt R, Gutmann L: Pancuronium improves the neuromuscular transmission defect of human organophosphate intoxication. Neurology 40:1275-1277, 1990. 13. Besser R, Weilemann LS, Gutmann L: Efficacy of obidoxime in human organophosphorus poisoning: Determination by neuromuscular transmission studies. Muscle Nerve 18:15-22, 1995. 14. Betz WJ: Depression of transmitter release at the neuromuscular junction of the frog. J Physiol (Lond) 206:629-644, 1970. 15. Bloom W, Fawcett DW: A Textbook of Histology, ed 10. WB Saunders, Philadelphia, 1975. 16. Branisteanu DD, Miyamoto MD, Voile RL: Effects of physiologic alterations on binomial transmitter release at magnesium-depressed neuromuscular junctions. J Physiol (Lond) 254: 19-37, 1976. 17. Burges J, Vincent A, Molenaar PC, NewsomDavis J, Peers C, Wray D: Passive transfer of seronegative myasthenia gravis to mice. Muscle Nerve 17:1393-1400, 1994. 18. Campbell WW Jr, Leshner RT, Swift TR: Plasma exchange in myasthenia gravis: Electrophysiologic studies. Ann Neurol 8:584-589, 1980. 19. Chandler WK, Rakowski RF, Schneider MF: A
253
non-linear voltage dependent charge movement in frog skeletal muscle. J Physiol (Lond) 254: 245-283, 1976. 20. Colquhoun D, Hawkes AG: On the stochastic properties of single ion channels. Proc R Soc London (Ser B) 211:205-235, 1981. 21. Confavreux C, Charles N, Aimard G: Fulminant myasthenia gravis soon after initiation of acebutolol therapy. Eur Neurol 30:279-281, 1990. 22. Costantin L: The role of sodium current in the radial spread of contraction in frog muscle fibers. J Gen Physiol 55:703-715, 1970. 23. Daras M, Samkoff LM, Koppel BS: Exacerbation of myasthenia gravis with cocaine use. Ann Neurol 46:271, 1996. 24. Daube JR, Lambert EH: Post-activation exhaustion in rat muscle. In Desmedt, JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 343-349. 25. Desmedt JE: Presynaptic mechanisms in myasthenia gravis. Ann NY Acad Sci 135:209-246, 1966. 26. Desmedt JE: The neuromuscular disorder in myasthenia gravis. 1. Electrical and mechanical response to nerve stimulation in hand muscles. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol. 1. Karger, Basel, 1973, pp 241-304. 27. Dominkus M, Grisold W, Albrecht G: Stimulation single fiber EMG study in patients receiving a long-term D-penicillamine treatment for rheumatoid arthritis. Muscle Nerve 1:13001301, 1992. 28. Drachman DB: Myasthenia gravis. N Engl J Med 330:1797-1810, 1994. 29. Drachman DB: Immunotherapy in neuromuscular disorders: Current and future strategies. Muscle Nerve 19:1239-1251, 1996. 30. Eaton LM, Lambert E: Electromyography and electric stimulation of nerves in diseases of motor units: Observations on myasthenic syndrome associated with malignant tumors. JAMA 163:117-1124, 1957. 31. Elias SB, Appel SH: Acetylcholine receptor in myasthenia gravis: Increased affinity for t-bungarotoxin. Ann Neurol 4:250-252, 1978. 32. Elmqvist D: Neuromuscular transmission defects. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 229-240. 33. Elmqvist D, Hofmann WW, Kugelberg J, Quastel DMJ: An electrophysiological investigation of neuromuscular transmission in myasthenia gravis. J Physiol (Lond) 174:417-434, 1964. 34. Elmqvist D, Lambert EH: Detailed analysis of neuromuscular transmission in a patient with the myasthenic syndrome sometimes associated with bronchogenic carcinoma. Mayo Clin Proc 43:689-713, 1968. 35. Elmqvist D, Mattsson C, Heilbronn E, Londh H, Libelius R: Acetylcholine receptor protein: Neuromuscular transmission in immunized rabbits. Arch Neurol 34:7-11, 1977. 36. Engel AG: Congenital disorders of neuromuscular transmission. Semin Neurol 10:12-26, 1990.
254
Assessment of Neuromuscular Transmission
37. Engel AG: Myasthenic syndromes. In Engel AG, Franzinl-Armstrong C (eds): Myology: Basic and Clinical ed 2. McGraw-Hill, New York, 1994, pp 1798-1835. 38. Engel AG, Lambert EH: Congenital myasthenic syndromes. Electroencephalogr Clin Neurophysiol 39(suppl):91-102, 1987. 39. Engel AG, Lambert EH, Gomez MR: A new myasthenic syndrome with end-plate acetylcholinesterase deficiency, small nerve terminals, and reduced acetylcholine release. Ann Neurol 1:315-330, 1977. 40. Engel AG, Lambert EH, Howard FM Jr: Immune complexes (IgG and C3) at the motor end-plate in myasthenia gravis. Mayo Clin Proc 52:267-280, 1977. 41. Engel AG, Lambert EH, Mulder DM, Torres CG, Sahashi K, Bertorini TE, Whitaker JN: A newly recognized congenital myasthenic syndrome attributed to a prolonged open time of the acetylcholine-induced ion channel. Ann Neurol 11:553-569, 1982. 42. Engel AG, Nagel A, Walls TJ, Harper CM, Waisburg HA: Congenital myasthenic syndromes: I. Deficiency and short open-time of the acetylcholine receptor. Muscle Nerve 16:1284-1292, 1993. 43. Engel AG, Ohno K, Bouzat C, Sine SM, Griggs RC: End-plate acetylcholine receptor deficiency due to nonsense mutations in the € subunit. Ann Neurol 40:810-817, 1996. 44. Engel AG, Santa T: Histometric analysis of the ultrastructure of the neuromuscular junction in myasthenia gravis and in the myasthenic syndrome. Ann NY Acad Sci 183:46-63, 1971. 45. Engel AG, Santa T: Motor endplate fine structure. In Desmedt, JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 196-228. 46. Engel AG, Tsujihata M, Lambert EH, Lindstrom JM, Lennon VA: Experimental autoimmune myasthenia gravis: A sequential and quantitative study of the neuromuscular junction ultrastructure and electrophysiologic correlations. J Neuropathol Exp Neurol 35:569-587, 1976. 47. Engel AG, Uchitel O, Walls TJ, Nagel A, Harper CM, Bodensteiner J: Newly recognized congenital myasthenic syndrome associated with high conductance and fast closure of the acetylcholine receptor channel. Ann Neurol 34:3847, 1993. 48. Fambrough DM, Drachman DB, Satyamurti S: Neuromuscular junction in myasthenia gravis: Decreased acetylcholine receptors. Science 182:293-295, 1973. 49. Fatt P, Katz B: Spontaneous subthreshold activity at motor nerve endings. J Physiol (Lond) 117:109-128, 1952. 50. Furui E, Fukushima K, Sakashita T, Sakato S, Matsubara S, Takamori M: Familial limb-girdle myasthenia with tubular aggregates. (Case of the Month) Muscle Nerve 20:599-603,1997. 51. Gold R, Schrnied M, Gregerich G, Breitschopf H, Hartung HP, Toyka KV, Lassmann H: Differentiation between cellular apoptosis and necrosis by the combined use of in situ tailing
52.
53.
54.
55.
56. 57. 58.
59. 60.
61.
62.
63.
64.
65.
66.
and nick translation techniques. Lab Invest 71:219-225, 1994. Gomez CM, Bhattacharyya BB, Charnet P, Day JW, Labarca C, Wollmann RL, Lambert EH: A transgenic mouse model of the slow-channel syndrome. Muscle Nerve 19:79-87, 1996. Gomez CM, Gammack JT: A leucine-to-phenylalanine substitution in the acetylcholine receptor ion channel in a family with the slowchannel syndrome. Neurology 45:982-985, 1995. Gomez CM, Maselli R, Gammack J: A b-subunit mutation in the acetylcholine receptor channel gate causes severe slow-channel syndrome. Ann Neurol 39:712-723, 1996. Gotti C, Balestra B, Mantegazza R, Tzartos S, Moretti M, dementi F: Detection of antibody classes and subpopulations in myasthenia gravis patients using a new nonradioactive enzyme immunoassay. Muscle Nerve 20:800-808, 1997. Grana EA, Chiou-Tan F, Jaweed MM: Endplate dysfunction in healthy muscle following a period of disuse. Muscle Nerve 19:989-993, 1996. Green DPL, Miledi R, Perez de la Mora M, Vincent A: Acetylcholine receptors. Phil Trans R Soc (Lond) B 270:551-559, 1975. Harper CM: Neuromuscular transmission disorders in childhood. In Jones RH, Bolton CF, Harper CM (eds): Pediatric Clinical Electromyography. Lippincott-Raven, Philadelphia, 1996, pp 353-385. Harper CM, Engel G: Quinidine sulfate therapy for the slowchannel congenital myasthenic syndrome. Ann Neurol 43:480-484, 1998. Hart ZH, Sahashi K, Lambert EH, Engel AG, Lindstrom JM: A congenital familial myasthenic syndrome caused by a presynaptic defect of transmitter resynthesis or mobilization. Neurology 29:556-557, 1979. Hoedemaekers AC, Verschuuren JJ, Spaans F, Graus YF, Riemersma S, van Breda Vriesman PJ, Baets MH: Age-related susceptibility to experimental autoimmune myasthenia gravis: immunological and electrophysiological aspects. Muscle Nerve 20:1091-1101, 1997. Houzen H, Hattori Y, Kanno M, Kikuchi S, Tashiro, K, Motomura M, Nakao Y, Nakamura T: Functional evaluation of inhibition of autonomic transmitter release by autoantibody from Lambert-Eaton myasthenic syndrome. Ann Neurol 43:677-680, 1998. Howard Jr JF, Saunders DB: Passive transfer of human myasthenia gravis to rats. 1. Electrophysiology of the developing neuromuscular block. Neurology 30:760-764, 1980. Hubbard JI: Neuromuscular transmission: Presynaptic factors. In Hubbard JI (ed): The Peripheral Nervous System. Plenum Press, New York, 1974, pp 151-180. Hutchinson DO, Walls TJ, Nakano S, Camp S, Taylor P, Harper CM, Groover RV, Peterson HA, Jamieson DG, Engel AG: Congenital endplate acetylcholinesterase deficiency. Brain 116:633653, 1993. Ito Y, Miledi R, Vincent A, Newsom Davis J: Acetylcholine receptors and end-plate electro-
Anatomy and Physiology of the Neuromuscular Junction
67. 68.
69. 70. 71.
72.
73.
74.
75.
76.
77.
78. 79.
80.
81.
physiology in myasthenia gravis. Brain 101: 345-368, 1978. Iwasa K: Striational autoantibodies in myasthenia gravis mainly react with ryanodine receptor. Muscle Nerve 20:753-756, 1997. Jonkers I, Swerup C Pirskanen R, Bjelak S, Matell G: Acute effects of intravenous injection of beta-adrenoreceptor- and calcium channel antagonists and agonists in myasthenia gravis. Muscle Nerve 19:989-965, 1996. Kadrie HA, Brown WF: Neuromuscular transmission in human single motor units. J Neurol Neurosurg Psychiatry 41:193-204, 1978a. Kadrie HA, Brown WF: Neuromuscular transmission in myasthenic single motor units. J Neurol Neurosurg Psychiatry 41:205-214, 1978b. Kaminski HJ, Fenstermaker RA, Abdul-Karim FW, dayman J, Ruff RL: Acetylcholine receptor subunit gene expression in thymic tissue. Muscle Nerve 16:1332-1337, 1993. Kaminski HJ, Ruff RL: The myasthenic syndromes. In Schultz SG (ed): Molecular Biology of Membrane Transport Disorders. Plenum, New York, 1996, pp 565-593. Katz B: Microphysiology of the neuro-muscular junction: A physiological "quantum of action" at the myoneural junction. Bull Johns Hopkins Hosp 102:275-312, 1958. Kelly JJ Jr, Lambert EH, Lennon VA: Acetylcholine release in diaphragm of rats with chronic experimental autoimmune myasthenia gravis. Ann Neurol 4:67-72, 1978. Khella SL: Management of critically ill patients with myasthenia gravis, the Guillain-Barre syndrome and the inflammatory myopathies. In Mandell BF (ed): Management of Critically 111 Patients with Immunological and Rheumatic Disease. Marcel Dekker, New York, 1994, pp 475-504. Kimura J: Electrodiagnostic study of pesticide toxicity. In Xintaras C, Johnson BL, De Groot I (eds): Behavioral Toxicology. U.S. Department of Health, Education and Welfare, U.S. Government Printing Office, Washington, DC, 1974, pp 174-181. Lamb GD, Stephenson DG: Excitation-contraction coupling in skeletal muscle fibres of rat and toad in the presence of GTPrS. J Physiol 444:65-84, 1991. Lambert EH, Elmqvist D: Quantal components of end-plate potentials in the myasthenic syndrome. Ann NY Acad Sci 183:183-199, 1971. Lambert EH, Lennon VA: Selected IgG rapidly induces Lambert-Eaton myasthenic syndrome in mice: complement independence and EMG abnormalities. Muscle Nerve 11:1133-1145, 1988. Lambert EH, Okihiro M, Rooke ED: Clinical physiology of the neuromuscular junction. In Paul WM, Daniel EE, Kay CM, Monckton G (eds): Muscle. Proceedings of the Symposium, The Faculty of Medicine, University of Alberta, Pergamon Press, London, 1965, pp 487-499. Lecky BRF, Morgan-Hughes JA, Murray NMF, Landon DN, Wray DW, Prior C: Congenital myasthenia: Further evidence of disease heterogeneity. Muscle Nerve 9:233-242, 1986.
255
82. Lennon VA, Kryzer TH, Griesmann GE, O'Suilleabhain PE, Windebank AJ, Woppmann A, Miljanich GP, Lambert EH: Calcium-channel antibodies in the Lambert-Eaton syndrome and other paraneoplastic syndromes. N Engl J Med 332:1467-1474, 1995. 83. Lennon VA, Lambert EH, Leiby KR, Okama TB, Talib S: Recombinant human acetylcholine receptor a-subunit induces chronic experimental autoimmune myasthenia gravis. J Immunol 22:2449-2552, 1991. 84. Lindstrom JM, Lambert EH: Content of acetylcholine receptor and antibodies bound to receptor in myasthenia gravis, experimental autoimmune myasthenia gravis, and Eaton Lambert syndrome. Neurology 28:130-138, 1978. 85. Lundh H: Antagonism of botulinum toxin paralysis by low temperature. Muscle Nerve 6:56-60, 1983. 86. Macintosh FC: Formation, storage, and release of acetylcholine at nerve endings. Can J Biochem Physiol 37:343-356, 1959. 87. Maddison P, Newsom-Davis J, Mills KR: Decay of postexercise augmentation in the LambertEaton myasthenic syndrome. Neurology 50: 1083-1087, 1998. 88. Magleby KL: Neuromuscular transmission. In Engel, AG, Franzini-Armstrong, C (eds): Myology, ed 2. McGraw-Hill, New York, 1994, pp 442-463. 89. Maselli R, Ellis W, Mandler P, Sheikh F, Senton G, Knox S, Salari-Namin H, Agius M, Wollmann RL, Richman DP: Cluster of wound botulism in California: Clinical, electrophysiologic and pathologic study. Muscle Nerve 20: 1284-1295, 1997. 90. McComas AJ, Galea V, Einhorn RW: Pseudofacilitation: A misleading term. Muscle Nerve 17:599-607, 1994. 91. McMahan UJ, Sanes JR, Marshall LM: Cholinesterase is associated with the basal lamina at the neuromuscular junction. Nature 271:172-174, 1978. 92. McQuillen MP: Familial limb-girdle myasthenia. Brain 89:121-132, 1966. 93. Meadows JC, Ross-Russell RW, Wise RP: A reevaluation of the decamethonium test for myasthenia gravis. Acta Neurol Scand 50:248-256, 1974. 94. Milone M, Hutchinson DO, Engel AG: Patchclamp analysis of the properties of acetylcholine receptor channels at the normal human endplate. Muscle Nerve 17:1364-1994, 1994. 95. Mora M, Lambert EH, Engel AG: Synaptic vesicle abnormality in familial infantile myasthenia. Neurology 37:206-214, 1987. 96. Nakano S, Engel AG: Myasthenia gravis: Quantitative immunocytochemical analysis of inflammatory cells and detection of complement membrane attack complex at the end-plate in 30 patients. Neurology 43:1167-1172, 1993. 97. Newsom-Davis J, Pinching AJ, Vincent A, Wilson SG: Function of circulating antibody to acetylcholine receptor in myasthenia gravis: Investigation by plasma exchange. Neurology 28:266-272, 1978.
256
Assessment of Neuromuscular Transmission
98. Newsom-Davis J: Antibody-mediated presynaptic disorders of neuromuscular transmission. Eighteenth Annual Edward H. Lambert Lecture, AAEM Plenary Session, 1993. 99. Oosterhuis HJGH, Newsom-Davis J, Wokke JHJ, Molenaar PC, Weerden TV, Oen BS, Jennekens FGI, Veldman H, Vincent A, Wray DW: The slow channel syndrome: Two new cases. Brain 110:1061-1079, 1987. 100. Pagala MKD, Nandakumar NV, Venkatachari SAT, Ravindran K, Namba T, Grob D: Responses of intercostal muscle biopsies from normal subjects and patients with myasthenia gravis. Muscle Nerve 13:1012-1022, 1990. 101. Pagala MKD, Nandakumar NV, Venkatachari SAT, Ravindran K, Amaladevi B, Namba T, Grob D: Mechanisms of fatigue in normal intercostal muscle and muscle from patients with myasthenia gravis. Muscle Nerve 16:911-921, 1993. 102. Patterson MF, Mould J, Dulhunty AF: Depolarization accelerates the decay of K+ contractures in rat skeletal muscle fibers. Muscle Nerve 19:1025-1036, 1996. 103. Protti DA, Reisin R, Mackinley TA, Uchitel OD: Calcium channel blockers and transmitter release at the normal human neuromuscular junction. Neurology 46:1391-1396, 1996. 104. Saadat K, Kaminski HJ: Ritonavir-associated myasthenia gravis. Muscle Nerve 21:680-681, 1998. 105. Satyamurti S, Drachman DB, Slone F: Blockade of acetylcholine receptors: A model of myasthenia gravis. Science 187:955-957, 1975. 106. Shapiro BE, Soto O, Shafqat S, Blumenfeld H: Adult botulism. (Short Report) Muscle Nerve 20:100-102, 1997. 107. Shi F-D, Bai X-F, Li H-L, Link H: Macrophage apoptosis in muscle tissue in experimental autoimmune myasthenia gravis. Muscle Nerve 21:1071-1074, 1998. 108. Sieb JP, Dorfler P, Tzartos S, Wewer UM, Ruegg MA, Meyer D, Baumann I, Lindemuth R, Jakschik J, Ries F: Congenital myasthenic syndromes in two kinships with end-plate acetylcholine receptor and utrophin deficiency. Neurology 50:54-61, 1998. 109. Slater CR, Lyons PR, Walls TH, Fawcett PRW, Young C: Structure and function of neuromuscular junctions in the vastus lateralis of man. Brain 115:451-478, 1992. 110. Smit LME, Hageman G, Veldman H, Molenaar PC, Oen BS, Jennekens FGI: A myasthenic syndrome with congenital paucity of secondary synaptic clefts. Muscle Nerve 11:337-348, 1988. 111. Smith DO, Conklin MW, Jensen PJ, Atchison WD: Decreased calcium currents in motor nerve terminals of mice with Lambert-Eaton myasthenic syndrome. J Physiol 487:115-123, 1995. 112. Sohal GS: Sixth annual Stuart Reiner memor-
113. 114. 115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125. 126.
ial lecture: Embryonic development of nerve and muscle. Muscle Nerve 18:2-14, 1995. Swash M, Ingram DA: Adverse effect of verapamil in myasthenia gravis. Muscle Nerve 15:396-398, 1992. Swift TR: Weakness from magnesium-containing cathartics: Electrophysiologic studies. Muscle Nerve 2:295-298, 1979. Takamori M, Sakato S, Matsubara S, Okumura S: Therapeutic approach to experimental autoimmune myasthenia gravis by dantrolene sodium. J Neurol Sci 58:17-24, 1983. Toyka KV, Drachman DB, Griffin DE, Pestronk A, Winkelstein JA, Fischbeck K Jr, Kao I: Myasthenia gravis: Study of humoral immune mechanisms by passive transfer to mice. N Engl J Med 296:125-131, 1977. Tsujihata M, Hazama R, Ishii N, Ide Y, Mori M, Takamori M: Limb muscle endplates in ocular myasthenia gravis: Quantitative ultrastructural study. Neurology 29:654-661, 1979. Uchitel O, Engel AG, Walls TJ, Nagel A, Atassi ZM, Bril V: Congenital myasthenic syndromes: II. A syndrome attributed to abnormal interaction of acetylcholine with its receptor. Muscle Nerve 16:1293-1301, 1993. Ueno S, Hara Y: Lambert-Eaton myasthenic syndrome without anti-calcium channel antibody: adverse effect of calcium antagonist diltiazem. J Neurol Neurosurg Psychiatry 55:409410, 1992. van Dijk JG, Lammers GJ, Wintzen AR, Molenaar PC: Repetitive CMAPs: Mechanisms of neural and synaptic genesis. Muscle Nerve 19:1127-1133, 1996. Waterman SA: Multiple subtypes of voltagegated calcium channel mediate transmitter release from parasympathetic neurons in the mouse bladder. J Neurosci 16:4155-4161, 1996. Waterman SA, Lang B, Newson-Davis J: Effect of Lambert-Eaton myasthenic syndrome antibodies on autonomic neurons in the mouse. Ann Neurol 42:147-156, 1997. Wirguin I, Brenner T, Sicsic C, Argov Z: Variable effect of calcium channel blockers on the decremental response in experimental autoimmune myasthenia gravis. Muscle Nerve 17: 523-527, 1994. Wokke JHJ, Jennekens FGI, Molenaar PC, Van den Oord CJM, Oen BS, Busch HFM: Congenital paucity of secondary synaptic clefts (CPSC) syndrome in 2 adult sibs. Neurology 39:648-654, 1989. Zhang G-X, Navikas V, Link H: Cytokines and the pathogenesis of myasthenia gravis. Muscle Nerve 20:543-551, 1997. Zhang GX, Xiao BG, Bakhiet M, van der Meide PH, Wigzell H, Link H, Olsson T: Both CD4+ and CD8+ T cells are essential to induce experimental autoimmune myasthenia gravis. J Exp Med 184:349-356, 1996.
Chapter
10
TECHNIQUES OF REPETITIVE STIMULATION
1. INTRODUCTION 2. METHODS AND TECHNICAL FACTORS Belly-Tendon Recording Movement-Induced Artifacts Temperature and Other Factors
3. COMMONLY USED NERVES AND MUSCLES
4.
5.
6.
7.
8.
Distal Versus Proximal Muscle Upper Limb and Shoulder Girdle Lower Limb Face RECOVERY CURVES BY PAIRED STIMULATION Short Interstimulus Intervals Long Interstimulus Intervals DECREMENTAL RESPONSE AT SLOW RATES OF STIMULATION Normal Muscles Myasthenia Gravis Other Neuromuscular Disorders INCREMENTAL RESPONSE AT FAST RATES OF STIMULATION Normal Muscles Lambert-Eaton Myasthenic Syndrome and Botulism Other Neuromuscular Disorders EFFECT OF TETANIC CONTRACTION Use of Prolonged Stimulation Posttetanic Potentiation Posttetanic Exhaustion CHANGES IN MYOGENIC DISORDERS Muscle Glycogenosis Myotonia Paramyotonia Congenita and Periodic Paralysis Proximal Myotonic Myopathy
257
258
Assessment of Neuromuscular Transmission
1 INTRODUCTION Nerve stimulation techniques as tests for neuromuscular transmission began with Jolly (1895),43 who applied faradic current repeatedly at short intervals. Using a kymographic recording and visual inspection of skin displacement, he found that the size of the muscle response deteriorated rapidly in patients with myasthenia gravis during the faradization. Faradic current failed to elicit a response in the volitionally fatigued muscle prior to testing. Conversely, after faradization, muscle responded poorly to subsequent volitional contraction. Based on these findings, Jolly concluded that the myasthenics had motor failure of the peripheral, rather than central, nervous system, a remarkable insight, considering the technical limitations at the time. His equipment consisted of a double-coil stimulator capable of eliciting only submaximal responses and a mechanical, rather than electrical, recorder. The use of supermaximal stimulation and the recording of the muscle action potential have increased the reliability and sensitivity of nerve stimulation techniques considerably. In 1941 Harvey and Masland37 noted that in myasthenia a single muscle response induced a prolonged depression, during which a second maximal motor nerve stimulus excited a reduced number of muscle fibers, and that a train of impulses resulted in a progressive decline in amplitude of compound muscle potential. Later studies have established optimal frequency of stimulation, proper control of temperature, appropriate selection of muscles, and various activation procedures to enhance an equivocal neuromuscular block.25 Microelectrode studies provide direct recording of end-plate potentials from muscle in vitro. All other electrophysiologic methods assess the neuromuscular junction only indirectly. Nonetheless, such an approach allows quantitation of the motor response to paired stimuli, tetanic contraction, or repetitive 22,51,85,96 stimulation at fast and slow rates. Transmission defects affect a variety of disease states, such as myasthenia gravis, myasthenic syndromes, botulism, amy-
otrophic lateral sclerosis, poliomyelitis, and multiple sclerosis. This chapter deals with the physiologic techniques for elucidating decremental or incremental responses in the differential diagnosis of clinical disorders (see also Chapter 27).
2
METHODS AND TECHNICAL FACTORS
Belly-Tendon Recording Belly-tendon recording consists of stimulating the nerve with supramaximal intensity and recording the muscle action potential with the active electrode (G1) placed over the motor point and the reference electrode (G2) on the tendon. The initially negative potential thus recorded represents the summated electrical activity from the entire muscle fiber population, discharging relatively synchronously. The area under the negative phase changes primarily with the number of muscle fibers activated. The magnitude of the unit discharge from individual muscle fibers also alters the size of the compound muscle potential, especially in myogenic disorders. In clinical studies, measurement of the amplitude suffices in a train of responses that shows the same duration and waveform.
Movement-Induced Artifacts Movement-related artifacts abound during repetitive stimulation of the nerve. The recording electrode may continuously slide away from the muscle belly, or the stimulating electrodes may gradually slip from the nerve, causing subthreshold activation. In either case, a progressively smaller amplitude of a train mimics a decremental response. Changes in limb position alter the shape of the volume conductor and the spatial relationship of muscle and recording electrodes, leading to a misleading alteration in amplitude of the recorded response (Fig. 10-1). Firm immobilization of the limb together with visual inspection of the contracting muscle under study minimizes the movementinduced variability.
Techniques of Repetitive Stimulation
259
Figure 10-1. A train of responses recorded from the thenar muscle with stimuli delivered one per second to the median nerve at the wrist in a healthy subject. Intentional stepwise alteration in thumb position from abduction to adduction after each shock gave rise to a smooth reduction in amplitude with concomitant increase in duration of successive potentials. The area under the waveform showed relatively little change from the first to the fifth response.
In most instances, technical problems cause abrupt, irregular changes in the amplitude or shape of the evoked response. Some movement artifacts, however, induce a smooth, progressive alteration of amplitude that closely mimics the myasthenic response. Nevertheless, close scrutiny often discloses accompanying changes in duration or other aspects of waveform, usually attributable to an alteration in the shape of the volume conductor. In our experience, even gradually changing waveforms represent artifacts if repetitive stimulation induces excessive movement. Repeated trials help establish the reproducibility of the finding, increasing the reliability of the results. Intertrial intervals should exceed 30 s to avoid the effect of subnormality of neuromuscular transmission that lasts for a few seconds after a single stimulus and a greater time period after repetitive impulses.
Temperature and Other Factors Exposure to warm sunlight may precipitate ptosis and diplopia in myasthenic patients.8,90 Similarly, electrophysiologic abnormalities of weak muscles may appear only after local warming. Physiologic mechanisms that account for the improved neuromuscular transmission with cooling include (1) facilitated transmitter
replacement in the presynaptic terminal,38,39 (2) reduced amount of transmitter release at the neuromuscular junction by the first of a train of impulses, leaving more quanta available for subsequent stimuli,23 (3) decreased hydrolysis of acetylcholine (ACh) by acetylcholinesterase, allowing sustained action of the transmitter already released from the axon terminal,32,79,80 (4) increased 36postsynaptic receptor sensitivity to ACh, and (5) reduced rate of removal of calcium ions (Ca2+) from the nerve terminal after stimulation.60 Elevated body temperature up to 42° C causes no abnormality in healthy subjects,80 but enhances the decrement on repetitive nerve stimulation in patients with myasthenia gravis.81 Lowering the intramuscular temperature from 35° to 28° C increases the amplitude of the compound muscle action potential and enhances the force of the9 isometric twitch and tetanic contraction. Patients with the myasthenic syndrome also experience distinct improvement after cooling,72,99 as do 24 those with100amyotrophic lateral sclerosis, 19 botulism, or tick paralysis. Cooling reduces the decrement to repetitive nerve stimulation (Fig. 10-2). Paradoxically, brief stimulation at high rates may produce a decremental response in normal muscles cooled below 32° C.52 Prior immersion of the limb in warm water or the
Assessment of Neuromuscular Transmission
260
Figure 10-2. Decremental response of the hypothenar muscle with stimulation of the ulnar nerve at two per second in a patient with myasthenia gravis. On the left, stimulation at 36 °C and on the right, after cooling of the hand to 30 °C. Note the reduction in the decrement from 15 percent to 6 percent, and the increase in amplitude after cooling of the hand. [From Denys,23 with permission.]
use of an infrared heat lamp helps maintain the recommended skin temperature over the tested muscle, above 34° C in most laboratories for diagnostic application. The effect of cholinesterase inhibitors also influences the results of repetitive stimulation. Administration of anticholinesterase drugs within a few hours before the test reduces the probability of obtaining a decremental response. Discontinuance of the short-acting medication for several hours improves the sensitivity of the test. The patient must withhold a long-acting medication for a longer period, if clinically feasible. With an overdose of anticholinesterase drugs, a single nerve impulse may cause a repetitive muscle response, and repetitive stimuli at a high rate give rise to a decremental response (see Chapter 9-6).
3
COMMONLY USED NERVES AND MUSCLES Distal Versus Proximal Muscle
Patients with myasthenia gravis rarely have a decremental response in clinically
unaffected muscle. Thus, isolated bulbar or respiratory muscle weakness may pose a diagnostic challenge.61 Weak proximal or facial muscles show a higher incidence of electrical abnormality than stronger distal muscles.92 In a series of experiments, electrical and mechanical responses to repetitive stimuli revealed substantially greater decrement and posttetanic potentiation in the platysma than in the adductor pollicis.48,49 Also, the trapezius has proven more sensitive than the distal hypothenar muscles in detecting abnormalities of neuromuscular transmission in amyotrophic lateral sclerosis.45 Similarly, electrophysiologic findings in botulism may involve only weak muscles of the clinically affected limbs. In contrast, patients with the myasthenic syndrome usually have prominent abnormalities not only in the proximal muscles but also in distal muscles, albeit less severely.13 In principle, the method consists of applying repetitive stimulation to a motor or mixed nerve and recording a train of responses from the innervated muscle. Although less sensitive, studies of the distal musculature provide technically more reliable results than those of more proximal muscles in the limb or facial muscles.
Techniques of Repetitive Stimulation Stimulation of the ulnar nerve at the elbow allows simultaneous recordings from one proximal muscle and three distal muscles: the flexor carpi ulnaris, abductor digiti quinti, first dorsal interosseous, and adductor pollicis.26 A negative result with distal muscles should prompt examination of the proximal muscles, such as the deltoid, biceps, and upper trapezius. Stimulation of the brachial plexus at the supraclavicular fossa tends to activate many muscles simultaneously. In contrast, stimulation of the accessory nerve selectively excites the trapezius without contamination from other muscles.59,87 Studies of the lower limb pose greater technical difficulty, yielding a wider67normal range compared to the upper limb. Wise choice of the nerve and muscle based on distribution of weakness increases test sensitivity.
261
line. The patient upright in a chair holds on to the handle with the arm flexed approximately 130 degrees and exercises the muscle by extending the elbow. BICEPS
The musculocutaneous nerve is stimulated at the axilla with G1 on the belly of the biceps and G2 over the tendon. The position of the arm depends on the type of mechanical board available. A handlebar attached under a solid table can serve as an excellent restraint. The patient, upright in a chair, holds on to the handle from below with the arm flexed approximately 45 degrees in the adducted and supinated position. Pulling up against the handlebar with flexion at the elbow exercises the muscle. DELTOID
Upper Limb and Shoulder Girdle HYPOTHENAR MUSCLES
The ulnar nerve is stimulated at the wrist with G1 placed over the belly of abductor digiti quinti and G2 on the tendon. Binding the four fingers together with a bandage or Velcro strap prevents interference from movement. The use of a restraining metal bar also helps hold the hand flat, palm down, on the examining table. The patient exercises by abducting the fifth digit against the restraint.
The brachial plexus is stimulated at Erb's point with GI on the belly of the muscle and G2 on the acromion. The patient sits upright with the arm adducted, flexed at the elbow, and internally rotated to place the hand in front of abdomen for selfrestraint by the opposite hand, and exercises by abducting the arm against his own resistance. Weak or uncooperative patients do better with a Velcro strap applied firmly against the trunk holding the arm adducted at the side. TRAPEZIUS
THENAR MUSCLES
The median nerve is stimulated at the wrist with G1 placed on the belly of the abductor pollicis brevis and G2 placed 2 cm distally. The hand, held palm up by the restraining metal bar, lies flat on the board with the thumb in the adducted position. The patient exercises the muscle by abducting the thumb against the bar.
The spinal accessory nerve is stimulated along the posterior border of the sternocleidomastoid muscle, with G1 on the upper trapezius muscle at the angle of neck and shoulder and G2 over the tendon, near the acromion process. The patient, upright in a chair, with the arms adducted and extended, holds on to the bottom of the chair and exercises, shrugging the shoulders against his own resistance.
ANCONEUS
The distal branch of the radial nerve is stimulated 4 cm above the elbow on the line bisecting the line connecting the olecranon and lateral epicondyle, with G1 placed 4 cm below the elbow on the same
Lower Limb ANTERIOR TIBIAL
The peroneal nerve is stimulated at the flbular head, with G1 on the belly of the
Assessment of Neuromuscular Transmission
262
muscle and G2 a few centimeters distally. The patient sits in a chair with the thigh restrained firmly by Velcro straps and exercises by dorsiflexing the foot held on a restraining foot board. QUADRICEPS
The femoral nerve is stimulated at the groin, just lateral to the femoral artery, with G1 placed on the rectus femoris and G2 on the patellar tendon. The patient sits in a chair with the thigh and the leg fastened to the chair with Velcro straps and exercises by extending the leg against the restraint. The patient may also lie supine with the thigh bound to the bed by a Velcro strap and exercise by lifting the foot off the bed.
Face ORBICULARIS OCULI, ORBICULARIS ORIS, AND NASALIS
A branch of the facial nerve is stimulated in front of the ear as distally as technically feasible. This usually allows nearly selective recording from the target muscle with G1 placed on its belly and G2 on the opposite side or on the bridge of the nose (see Figs. 1-3 and 17-2). The patient, lying supine, exercises by contracting the muscle as vigorously as possible without the benefit of a restraining device to immobilize facial muscles.
4
RECOVERY CURVES BY PAIRED STIMULATION
Short Interstimulus Intervals Paired stimuli applied at various intervals reveal the time course of recovery of neuromuscular transmission (see Chapter 9-6). In normal muscles, the first supramaximal stimulus activates the entire group of muscle fibers. A second stimulus delivered within a few milliseconds evokes a smaller response, indicating refractoriness of the nerve and muscle (see Fig. 9-4). The second potential then progressively recovers, with some overlap of the two responses at intervals of less than 15 ms. In typical cases of myasthenia gravis, the first stimulus elicits a maximal or near-maximal muscle response. The recovery curve also follows a normal pattern for short interstimulus intervals up to 15 ms. The curve deviates from normal in the Lambert-Eaton myasthenic syndrome, where the first stimulus elicits a submaximal response; a second shock given at very short interstimulus intervals evokes a larger response with the amplitude one and a half to two times that of the first. The increment represents recruitment, based on summation of two end-plate potentials (EPPs), of those fibers activated only subliminally by the first stimulus. Most patients with botulism (Fig. 10-3)17 and, occasionally patients
Figure 10-3. The effect of paired shocks given at interstimulus intervals of 2.5 ms (top), 15 ms (middle), and 25 ms (bottom). On the top tracings, the reduced test response in a healthy subject (A) indicates the effect of the refractory period, whereas the increased test response in a patient with botulism (B) suggests summation of two closely elicited end-plate 15potentials. [From Cherington, with permission.]
Techniques of Repetitive Stimulation with myasthenia gravis who have less than maximal initial responses also show the same phenomenon. Long Interstimulus Intervals Two EPPs no longer summate at interstimulus intervals exceeding 15 ms. Potentiation of the second response here represents true facilitation, resulting from an increased number of quanta liberated by the second stimulus. Despite the release of a greater amount of acetylcholine (ACh), the second muscle potential normally shows no increment from the already maximal first response. Most patients with myasthenia gravis or botulism also have minimal change at this interstimulus range. In contrast, patients with myasthenic syndrome show an increment at interstimulus intervals ranging from 15 to 100 ms as one of the most characteristic electrophysiologic features. The decremental response in myasthenia gravis begins at intervals of about 20 ms but becomes more definite at intervals between 100 and 700 ms. The response reaches the trough at an interstimulus interval of about 300-500 ms (see Fig. 9-6). At shorter intervals, concomitant facilitation attributable to the electrosecretory mechanism obscures the depression. The response slowly returns to the baseline in about 10 s. The results of paired stimuli predict that a train of stimuli produces the maximal decrement at the rate of 2-3 Hz.25
5
DECREMENTAL RESPONSE AT SLOW RATES OF STIMULATION Normal Muscles
Repetitive stimulation at a rate of 1-5 Hz depletes the immediately available acetylcholine (ACh) store, without superimposed facilitation from neurosecretory mechanisms (see Fig. 9-8). At slow rates of stimulation, movement-related artifacts are minimal because the muscle returns close to its original relaxed position before
263
the next stimulus. Most patients tolerate a train at faster rates poorly. Moreover, continuous muscle contraction alters the geometry of the volume conductor, which in turn affects the waveform of the successive responses. Random or irregular variations in amplitude or waveform suggest artifacts. Occasionally, inadvertent movement may cause smooth, reproducible changes erroneously suggesting abnormality of neuromuscular transmission. Even when the amplitude measures show a deceptive change, careful evaluation of the waveform and close visual inspection of the contracting muscle usually disclose the source of artifacts. Most modem equipment automatically calculates the percentage reduction for the smallest of the initial five to seven responses, compared with the first in the same train. Accepting the computed results without verification of the waveform may lead to an erroneous conclusion. In normal muscles, decrement at stimulation of 2-3 Hz, if present, does not exceed 5-8 percent.91 In fact, an optimal train comprises practically identical responses from the first to the last. Thus, the presence of any reproducible decrement should raise suspicion in a tracing free of any technical problems. Myasthenia Gravis In myasthenia gravis, the amplitude drops maximally between the first and second responses of a train, followed by a further but lesser decline up to the fourth or fifth potential (Fig. 10-4). Subsequent responses in the series then level off or, more typically, reverse the course by regaining some of the lost amplitude. Occasionally, the recovery may even exceed the original value by 10-20 percent, especially after several seconds of repetitive stimulation. More characteristically, continued stimulation induces a long, 42slow decline after a transient increment. To avoid a false-positive result, most electromyographers use a conservative criterion of abnormality: A reprducible decrement of 10 percent or more between the first response and the smallest of the next four to six responses.40 In
Figure 10-4. Thenar muscle potentials elicited by a train of stimuli of three per second to the median nerve before and after 1 minute of exercise in a patient with generalized myasthenia gravis. Amplitude comparison between the first and fifth responses revealed a decrement of 25 percent at rest, 12 percent immediately after exercise, and 50 percent 4 minutes later.
Figure 10-5. A 25-year-old woman with double vision of 11/2 months duration. A train of shocks of one, two, and three per second to the median nerve revealed no detectable abnormalities in the abductor pollicis brevis. Stimulation of the facial nerve elicited decrementing responses in the nasalis. Note greater change within the train as the rate of stimulation increased from one to three per second. [From Kimura,46 with permission.]
264
Techniques of Repetitive Stimulation addition to the changes in amplitude, the latency may progressively increase in some myasthenic muscles. In equivocal cases, sampling several muscles improves the chance of documenting localized myasthenic weakness. In particular, a negative result in the distal limb muscles by no means precludes electrical abnormalities detectable in the proximal or facial musculature (Fig. 10-5). The administration of edrophonium (Tensilon) or neostigmine (Prostigmin) helps further delineate the characteristics of defective neuromuscular transmission. These agents potentiate the action of ACh by blocking acetylcholinesterase (AChE) in patients with postjunctional abnormalities. Therefore, a partial or complete reversal of the decrement by anticholinesterase agents tends to confirm the diagnosis of myasthenia gravis.
Other Neuromuscular Disorders A train of stimuli at a slow rate causes decrementing responses not only in myasthenia gravis but also in a number of other
Figure 10-6. Thenar muscle potential elicited by a train of stimuli 1 through 20 per second to the median nerve in a patient with myasthenic syndrome. Note decremental responses to slow rates of stimulation up to five per second and incremental responses to faster rates of stimulation at 10 per second and to a much greater degree at 20 per second.
265 conditions with reduced margins of safety. These include the Lambert-Eaton myasthenic syndrome (Figs. 10-6 and 10-7), congenital myasthenic syndromes, botulism, multiple sclerosis,4,30,45,66 motor neuron disease, and regenerating nerve.33 A partially curarized muscle will develop a similar decrement to a train of stimuli. In the patient with the myasthenic syndrome or botulism, single stimuli typically elicit very small muscle action potentials. A decremental tendency with a slow rate of repetitive stimulation, though present in most cases, does not constitute an essential feature of these disorders, characterized by defective release of ACh. In depolarizing block seen in slow-channel congenital myasthenic syndrome or end-plate AChE deficiency syndrome, as in organophosphate poisoning,5,7 markedly prolonged end-plate potential remains excitatory beyond the refractory period of neuromuscular junction. Thus, single stimuli of the motor nerve typically elicit more than one compound muscle action potential, an initial main response followed by one or more smaller recurrent responses, which appear at 3-7 ms inter-
Assessment of Neuromuscular Transmission
266
Figure 10-7. A repeat study in the same patient as in Figure 10-6 using the same recording arrangements. Note further diminution in amplitude of the compound muscle action potentials compared with the earlier study, slight decrement at slow rates of stimulation up to five per second, and progressively more prominent increment at faster rates of 10-30 per second.
vals.34,98 With repetitive stimulation both responses show a rate-dependent decrement, although the recurrent potentials diminish more rapidly and disappear after brief exercise (Fig. 10-8). Quinidine sulfate therapy reverses this abnormality 35 concomitant with clinical improvement. A low dose of pancuronium, an ACh receptor antagonist, repairs the decrement seen in organophosphate intoxication, countering prolonged depolarization at the end-plate (Fig. 10-9).6
6
INCREMENTAL RESPONSE AT FAST RATES OF STIMULATION Normal Muscles
Supramaximal stimulation normally activates all muscle fibers innervated by the nerve. This precludes any increment in
response to greater amounts of acerylcholine (ACh) released by subsequent stimuli. The recruitment of muscle fibers not activated by the first stimulus underlines the incremental tendency seen in the myasthenic syndrome, botulism, and, occasionally, myasthenia gravis. Muscles stimulated repetitively at a high rate tend to discharge with increased synchrony without recruitment of additional muscle fibers. The compound muscle action potential may then increase in amplitude, but not in area under the waveform, as implied by the term pseudofactiitation. In normal adults, muscle action potentials remain stable during repetitive stimulation at a rate of up to 20-30 Hz.70 Some healthy infants, however, may show a progressive decline in amplitude at this rate.18 In adults, trains of 50 Hz may cause apparent decremental or incremental responses. At such a fast rate, however, inherent movement artifacts render the measurement unreliable.
Figure 10-8. A 29-year-old woman with congenital myasthenic syndrome from acetylcholinesterase deficiency. Generalized weakness began in her childhood, peaking at the age of 15-16 years with increasing fatigability of truncal and proximal limb muscles and development of scoliosis on standing. Administration of anticholinesterase worsened the symptoms. A. Single shocks of the ulnar nerve elicited two compound muscle action potentials in the first dorsal interosseous muscle, M1 and M2, but not in the deltoid, the weakest muscle of the limb. B. A train of stimuli at 3 Hz caused a decrement of M2 but not M1 in the first dorsal interosseous and a clear decrement of M1 in the deltoid. [Courtesy of Nobuo Kohara, M.D., Department of Neurology, Kyoto University School of Medicine.]
267
Figure 10-9. A 25-year-old woman with organophosphate intoxication after attempting suicide by ingestion of phenitrothion. Severe cholinergic crisis resulted in a respiratory failure necessitating mechanical ventilation for 70 hours. The patient remained comatose for a week, followed by gradual improvement; she returned to normal in 17 days. A: On day 2 (a) single shocks of the median nerve elicited three compound muscle action potentials, M1, M2 (underline), and M3 in the thenar muscle, and (b) a train of stimuli at 20 Hz showed a decrement of M1 and M2 followed by an increment of M1 with absent M2 and M3. B: On day 5 (c) the same train resulted in complete abolition of all responses after the third train of stimuli, and (d) administration of acetylcholine receptor antagonist, pancuronium, in low dosage (1000 mg) repaired the deficit completely, as expected in depolarization block. [Courtesy of Nobuo Kohara, M.D., Department of Neurology, Kyoto University School of Medicine.]
268
Techniques of Repetitive Stimulation Lambert-Eaton Myasthenic Syndrome and Botulism In the myasthenic syndrome29 single stimuli typically elicit a strikingly small compound muscle action potential (Fig. 10-10). The amplitude varies over a wide range among different subjects. Thus, a decrease by as much as 50 percent of the maximal response in some individuals may still remain above the lower limit of a population norm. An apparent lack of reduction in amplitude, therefore, does not necessarily rule out the syndrome. A marked potentiation following a brief voluntary exercise would disclose the subnormality of the initial amplitude and confirm the diagnosis. A slow rate of sustained stimulation also facilitates the response if superimposed on voluntary contraction.56 A train of stimulation at high rates, despite its theoretical interest,69 has seen only limited clinical application. Because
269
of the discomfort it causes, most patients tolerate the procedure poorly. Besides, voluntary contraction usually induces greater potentiation.96 Repetitive stimulation given at 20-50 Hz induces a remarkable increment of successive muscle action potentials to a normal or near normal level (see Figs. 10-6 and 10-7). A slight initial decrement may precede the increment, but the last response of a train at the end of 1 minute usually 54exceeds the first response several times. The electrophysiologic abnormalities often improve in parallel to the clinical course after the administration of guanidine or 3,4-diaminopyridine.68,86 Patients with botulism may have entirely normal electrical responses in early stages of the illness or have a small muscle potential in response to a single stimulus.16 An initially small response usually potentiates after voluntary exercise or with a train of stimuli (Fig. 10-11). Incrementing responses, though smaller in
Figure 10-10. Relationship between clinical estimate of weakness and the amplitude of muscle action potential in patients with myasthenia gravis and myasthenic syndrome. The histogram plots the amplitude of the hypothenar muscle potential elicited by single maximal stimuli to the ulnar nerve. The scale on the abscissa denotes normal strength (0), 75 percent (1), 50 percent (2), 25 percent (3), and complete paralysis (4). [From Lambert, Rooke and Eaton,55 with permission.]
Assessment of Neuromuscular Transmission
270
Figure 10-11. Muscle action potentials to a train of stimulation applied to the motor nerve at 50 per second in a patient with botulism. Note incremental responses when the patient received a 7 mg/kg daily dose of guanidine (A) and electrophysiologic15recovery after the dosage was increased to 35 mg/kg (B). Vertical calibration is 2 mV. [From Cherington, with permission.]
range, resemble those found in the myasthenic syndrome.62,89 Tetanic and posttetanic facilitation, the most characteristic abnormality of infantile botulism, persists for a number of minutes.20,31,84
Other Neuromuscular Disorders An incremental response, though characteristic of the myasthenic syndrome and botulism, by no means excludes other disorders of the neuromuscular junction (see Chapter 27-6). Patients with myasthenia gravis not infrequently show such a pattern, either during a progressive phase of the disease or during steroid therapy.21,63,88 In contrast to the marked potentiation in the myasthenic syndrome, however, changes rarely exceed the initial value by more than 40 percent at the end of 1 minute. Other disorders associated with depressed neuromuscular transmission and incremental tendency by a train of stimuli include antibiotic toxicity,71 94 hypocalcemia, hypermagnesemiam,11, 44 and snake venom poisoning. Again, a limited degree of potentiation seen in these conditions stands in sharp contrast to the multifold increase characteristic of the myasthenic syndrome.
7 EFFECT OF TETANIC CONTRACTION
Use of Prolonged Stimulation A short train of several shocks at a slow rate suffices for routine evaluation of neuromuscular transmission. Prolonged
stimulation at a rapid rate adds diagnostic information in the evaluation of infantile botulism. Otherwise, clinical yields seldom justify subjecting the patient to this painful procedure. Further, sustained muscle contraction causes excessive movement artifacts that often interfere with accurate assessments of the waveform. As a research tool, a long train helps elucidate the time course of the mechanical force of contraction. The force of muscle twitch increases during prolonged stimulation in healthy subjects, but not in patients with myasthenia gravis. This phenomenon, called a positive staircase, has no diagnostic specificity as a clinical test.27,91 Whatever the purpose, clinicians must resort to a train of rapid stimulation judiciously to avoid inflicting unnecessary discomfort. Tetany develops after electrical stimulation of a 20-30 s train at 50 Hz or a continuous run for a few minutes at 3 Hz. Most subjects tolerate these procedures poorly. Fortunately, voluntary muscle contraction accomplishes the same effect, discharging motor fibers up to 50 Hz during maximal effort. A typical postactivation cycle after voluntary or involuntary tetanic contraction consists of two phases: Posttetanic potentiation,42 lasting for about252 minutes, and posttetanic exhaustion, lasting up to 15 minutes.
Posttetanic Potentiation Tetanic contraction not only causes calcium (Ca2+) to accumulate inside the axon but also mobilizes acetylcholine (ACh) vescicles from the main store. Subsequent nerve stimulation gives rise to a larger
Techniques of Repetitive Stimulation
271
Figure 10-12. Thenar muscle potential elicited by a train of stimuli three per second to the median nerve before and after 10 s of exercise in a patient with the myasthenic syndrome. Note a posttetanic potentiation of 70 percent on the left and 160 percent on the right immediately after the exercise and a posttetanic exhaustion, 11/2 minutes later.
EPP, thus recruiting additional muscle fibers not previously activated in the Lambert-Eaton myasthenic syndrome or related disorders with defective release of ACh (Figs. 10-12 and 10-13). In physiologic experiments of single nerve fiber stimulation, four types of short term synaptic enhancement follow the end of tetanic activation, fast-decaying facilitation, slow decaying facilitation, augmentation, and posttetanic potentiation.60 In practice, a simple procedure consists of delivering single shocks of supramaxlmal intensity to the nerve and comparing the size of the muscle response measured before and after exercise. A striking increase in amplitude, usually reaching a level more than twice the baseline value, indicates a presynaptic defect of neuromuscular transmission.55 Posttetanic augmentation lasts about 20 s, showing less decay after cooling, reflecting slower removal 60 of calcium ions from the nerve terminal. Duration of exercise should
not exceed 15 s, to minimize depletion of ACh during voluntary contraction. In general, a posttetanic potentiation greater than twice the preactivation response suggests the diagnosis of Lambert-Eaton syndrome. The magnitude of potentiation, however, varies considerably from one subject to another and during the course of the illness within the same patient (Fig. 10-14). A lesser degree of facilitation also implies a presynaptic disturbance seen not only in myasthenic syndrome but also in congenital myasthenic syndromes, botulism, and occasional cases of myasthenia gravis. The use of a train of stimuli at 3 Hz instead of a single shock allows simultaneous evaluation of the decremental trends. The procedure consists of repeating the same train before and immediately after the exercise and then every 30 s thereafter for a few minutes. In this arrangement, posttetanic potentiation partially compensates for depletion of ACh during
Assessment of Neuromuscular Transmission
272
Figure 10-13. A repeat study in the same patient as in Figure 10-12 using the same recording arrangements. Compared with the earlier study, the patient had further diminution in amplitude of the compound muscle action potentials and a greater posttetanic potentiation on both sides.
each train, repairing the deficit caused by the slow rate of stimulation (see Fig. 10-4). Thus, the characteristic decrement seen within a train in myasthenia gravis tends to normalize immediately after the tetanic stimulation.
Posttetanic Exhaustion Decreased excitability of the neuromuscular junction follows a transient potentiation in 2-4 minutes after exercise. The underlying physiologic mechanism for this phenomenon probably relates to the depletion of the immediately available store of ACh during prolonged contraction, despite an increased rate of ACh mobilization. In normal subjects with a large margin of safety, the reduced amount of ACh released during posttetanic exhaustion will still generate an adequate EPP in each individual muscle fiber. In premature infants and some newborns with lim-
ited neuromuscular reserves,47 however, the amplitude of the compound muscle action potential progressively declines at high rates of stimulation. In myasthenia gravis, neuromuscular block worsens during posttetanic exhaustion, indicating a reduced margin of safety. Some patients showing an equivocal decrement at rest may develop a definite abnormality after exercise (see Fig. 10-4). In the myasthenic syndrome, a reduced EPP after exercise results in further diminution of the originally small compound muscle action potential (see Figs. 10-12 and 10-13). Thus, the use of exercise increases the sensitivity of the nerve-stimulation technique as a test of neuromuscular transmission. In the evaluation of posttetanic exhaustion, a 1 minute period of voluntary contraction results in optimal depletion of the ACh store. In contrast, a shorter exercise, for a period ranging from 10 to 15 seconds, suffices for assessment of the posttetanic
273
Techniques of Repetitive Stimulation
Figure 10-14. A 63-year-old woman with proximal weakness of all four extremities since October 1982. Thenar muscle potentials were elicited by stimuli applied to the median nerve at the wrist at three per second before and after 15 s of exercise. Notice the gradual reduction in the magnitude of posttetantic potentiation from 1983 through 1987. In the last study, the exercise induced only an incrementing tendency within the train, rather than the absolute increase in amplitude considered mandatory for the diagnosis of myasthenic syndrome. [From Kimura,46 with permission.]
potentiation to avoid excessive depletion of ACh, which would mask the expected change. 8
CHANGES IN MYOGENIC DISORDERS
A train of stimuli causes an apparent decrement of the compound muscle action potentials in a number of myogenic disorders, such as McArdle's disease, myotonia, pararnyotonia congenita, and periodic paralysis, but not in proximal myotonic myopathy (see Chapter 27-6).93
Muscle Glycogenosis In McArdle's disease and other disorders of muscle glycogenesis, painful, electrically
silent muscle contracture develops after exercise (see Fig. 12-3). With rapid repetitive stimulation of a motor nerve, the amplitude of the compound muscle action potential progressively declines, eventually leading to the development of contracture.10,75 Lowrate nerve stimulation during regional ischemia also gives rise to abnormal reduction of muscle response in patients with muscle glycogenoses as compared with control subjects.57
Myotonia In myotonic muscles, repetitive nerve stimulation produces commonly but not invariably decrementing responses.2,12,28,53,68 Unlike the responses in myasthenia gravis, a train fails to show a repair, or leveling off, after the fourth or fifth stimulus. In-
274
Assessment of Neuromuscular Transmission
stead, progressive decline continues for the initial few seconds followed by gradual recovery during subsequent stimulation for many seconds. In general, the higher the rate of stimulation, the greater the change in amplitude and the shorter the latent periods. The presence of clinical weakness also favors the possibility of finding prominent electrical decrement. The change occurs at a lower stimulation frequency in myotonia congenita than in myotonia dystrophica. The decremental changes in myotonia may result from prolonged afterdepolarization, induced by accumulation of potassium (K+) in the transverse tubules.1 Direct stimulation of the muscle evokes decreasing response, suggesting an excitability change of the muscle, rather
than the neuromuscular junction.12 Direct stimulation of single muscle fibers at 10-20 Hz gives rise to a progressive decline of single muscle fiber action potential associated with either increasing or decreasing propagation velocity.50,65,97 Intracellular recording of a myotonic discharge also shows a progressive decline in amplitude.77 Myotonic bursts may render some of the muscle fibers refractory to subsequent stimuli. In contrast to muscles in myasthenia gravis, myotonic muscles show neither posttetanic potentiation nor exhaustion. Indeed, the amplitude of the muscle response is less than the baseline value immediately after exercise. The decremental tendency also worsens after exercise, gradually restoring the resting value in about two to three minutes.
Figure 1O-15. A 27-year-old woman with a 10-year history of hyperkalemic periodic paralysis occurring two or three times a year. Stimulation of the ulnar nerve at the wrist elicited a normal compound muscle action potential (CMAP) of the abductor digiti minimi (9.7 mV). After a 5 minute exercise alternating 20 S maximal contraction and 3 S rest, CMAP initially increased in amplitude, peaking at 30 s post-exercise (13.5 mV); it then declined progressively throughout the test to a value below the baseline, reaching a trough at 40 minutes (4.7 mV). Repetitive stimulation of the median or facial nerve at 3 Hz revealed no change in CMAP amplitude of the target muscle. [Courtesy of Mark Ross, M.D., Department of Neurology, University of Kentucky.]
Techniques of Repetitive Stimulation
Paramyotonia Congenita and Periodic Paralysis In paramyotonia congenita, cooling worsens both the weakness and electrical abnormalities. Thus, patients characteristically show a decremental response on repetitive stimulation, especially following cold exposure, and an equally typical cold-sensitive decrease in amplitude after exercise.41 In vitro study has shown decreased excitability of the muscle membrane in hypokalemic periodic paralysis.78 This abnormality probably underlies the decline of the compound muscle action potential amplitude elicited after prolonged exercise3,64 and decrementing response on repetitive stimulation, associated with increasing muscle membrane refractoriness.76 During a paralytic episode, a single stimulus elicits a small compound muscle action potential, which may progressively increase with sustained or intermittent repetitive stimulation of the nerve at high14rates, although it falls again during rest. In hyperkalemic periodic paralysis, a compound muscle action potential elicited by nerve stimulation usually shows an initial increase after a long exercise, followed by progressive decline to a level below the baseline value (Fig. 10-15).
Proximal Myotonic Myopathy In proximal myotonic myopathy, which clinically resembles myotonic dystrophy,73,74,82,95 exercise tests of the distal muscle resulted in no diminution of response in one small series.83 This finding stands in sharp contrast to the postexercise depression seen in myotonic dystrophy (see Chapter 29-2).
REFERENCES 1. Adrian RH, Bryant SH: On the repetitive discharge in myotonic muscle fibers. J Physiol (Lond) 240:505-515, 1974. 2. Aminoff MJ, Layzer RB, Satya-Murti S, Faden AI: The declining electrical response of muscle to repetitive nerve stimulation in myotonia. Neurology 27:812-816, 1977.
275 3. Arimura Y, Arimura K, Suwazono S, Imamura H, Sonoda Y, Maruyama Y, Nakano K, Osame M: Predictive value of the prolonged exercise test in hypokalemic paralytic attack, (short report) Muscle Nerve 18:472-474, 1995. 4. Bernstein LP, Antel JP: Motor neuron disease: Decremental responses to repetitive nerve stimulation. Neurology 31:202-204, 1981. 5. Besser R, Gutmann L: A quantitative study of the pancuronium antagonism at the motor endplate in human organophosphorus intoxication. Muscle Nerve 18:956-960, 1995. 6. Besser R, Vogt T, Gutmann L: Pancuronium improves the neuromuscular transmission defect of human organophosphate intoxication. Neurology 40:1275-1277, 1990. 7. Besser R, Vogt T, Gutmann L, Hopf H, Wessler I: Impaired neuromuscular transmission during partial inhibition of acetylcholinesterase: The role of stimulus induced antidromic backfiring in the generation of the decrement-increment phenomenon. Muscle Nerve 15:10721080, 1992. 8. Borenstein S, Desmedt JE: Temperature and weather correlates of myasthenic fatigue. Lancet 2:63-66, 1974. 9. Borenstein S, Desmedt JE: Local cooling in myasthenia: Improvement of neuromuscular failure. Arch Neurol 32:152-157, 1975. 10. Brandt NJ, Buchtal F, Ebbesen F, Kamieniecka Z, Krarup C: Post-tetanic mechanical tension and evoked action potentials in McArdle's disease. J Neurol Neurosurg Psychiatry 40:920925, 1977. 11. Branisteanu DD, Miyamoto MD, Voile RL: Effects of physiologic alterations on binomial transmitter release at magnesium-depressed neuromuscular junctions. J Physiol (Lond) 254:19-37, 1976. 12. Brown JC: Muscle weakness after rest in myotonic disorders: An electrophysiological study. J Neurol Neurosurg Psychiatry 37:1336-1342, 1974. 13. Brown JC, Johns RJ: Diagnostic difficulties encountered in the myasthenic syndrome sometimes associated with carcinoma. J Neurol Neurosurg Psychiatry 37:1214-1224, 1974. 14. Campa JF, Sanders DB: Familial hypokalemic periodic paralysis: Local recovery after nerve stimulation. Arch Neurol 31:110-115, 1974. 15. Cherington M: Botulism: Electrophysiologic and therapeutic observations. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1, Karger, Basel, 1973, pp 375-379. 16. Cherington M: Botulism: Ten-year experience. Arch Neurol 30:432-437, 1974. 17. Cherington M, Ginsberg S: Type B botulism: Neurophysiologic studies. Neurology 21:43-46, 1971. 18. Churchill-Davidson HC, Wise RP: Neuromuscular transmission in the newborn infant. Anesthesiology 24:271-278, 1963. 19. Cooper BJ, Spence I: Temperature-dependent inhibition of evoked acetylcholine release in tick paralysis. Nature 263:693-695, 1976. 20. Cornblath DR, Sladky JT, Sumner AJ: Clinical
276
21. 22.
23.
24. 25.
26. 27.
28.
29.
30. 31. 32. 33. 34.
35. 36.
Assessment of Neuromuscular Transmission
electrophysiology of infantile botulism. Muscle Nerve 6:448-452, 1983. Dahl DS, Sato S: Unusual myasthenic state in a teen-age boy. Neurology 24:897-901, 1974. Daube JR: Minimonograph #8: Electrophysiologic testing for disorders of the neuromuscular junction. American Association of Electromyography and Electrodiagnosis, Rochester, Minn, 1978. Denys EH: Minimonograph 14: The role of temperature in electromyography. American Association of Electromyography and Electrodiagnosis, Rochester, Minn, 1980. Denys EH, Norris FH Jr: Amyotrophic lateral sclerosis: Impairment of neuromuscular transmission. Arch Neurol 36:202-205, 1979. Desmedt JE: The neuromuscular disorder in myasthenia gravis. 1. Electrical and mechanical response to nerve stimulation in hand muscles. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 241-304. Desmedt JE, Borenstein S: Diagnosis of myasthenia gravis by nerve stimulation. Ann NY Acad Sci 274:174-188, 1976. Desmedt JE, Emeryk B, Hainaut K, Reinhold H, Borenstein S: Muscular dystrophy and myasthenia gravis: Muscle contraction properties studied by the staircase phenomenon. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 380-399. Deymeer F, Cakirkaya S, Serdaroglu P, Schleithoff L, Lehmann-Horn F, Rudel R, Ozdemir C: Transient weakness and compound muscle action potential decrement in myotonia congenita. Muscle Nerve 21:1334-1337, 1998. Eaton LM, Lambert EH: Electromyography and electric stimulation of nerves in diseases of motor unit: Observations on myasthenic syndrome associated with malignant tumors. JAMA 163:1117-1124, 1957. Eisen A, Yufe R, Trop D, Campbell I: Reduced neuromuscular transmission safety factor in multiple sclerosis. Neurology 28:598-602, 1978. Fakadej AV, Gutmann L: Prolongation of posttetanic facilitation in infant botulism. Muscle Nerve 5:727-729, 1982. Foldes FF, Kuze S, Vizi ES, Deny A: The influence of temperature on neuromuscular performance. J Neurol Transm 43:27-45, 1978. Gilliatt RW: Nerve conduction in human and experimental neuropathies. Proc R Soc Med (Lond) 59:989-993, 1966. Harper CM: Neuromuscular transmission disorders in childhood. In Jones RH, Bolton CF, Harper CM (eds): Pediatric clinical electromyography. Lippincott-Raven, Philadelphia, 1996, pp 353-385. Harper CM, Engel AG: Quinidine sulfate therapy for the slow-channel congenital myasthenic syndrome. Ann Neurol 43:480-484, 1998. Harris JB, Leach GDH: The effect of temperature on end-plate depolarization of the rat diaphragm produced by suxamethonium and acetylcholine. J Pharm Pharmacol 20:194-198, 1968.
37. Harvey AM, Masland RL: The electromyogram in myasthenia gravis. Bull Johns Hopkins Hosp 48:1-13, 1941. 38. Hofmann WW, Parsons RL, Feigen GA: Effects of temperature and drugs on mammalian motor nerve terminals. Am J Physiol 211:135-140, 1966. 39. Hubbard JI, Jones SF, Landau EM: The effect of temperature change upon transmitter release, facilitation and post-tetanic potentiation. J Physiol (Lond) 216:591-609, 1971. 40. Jablecki CK: AAEM case report #3: Myasthenia gravis. Muscle Nerve 14:391-397, 1991. 41. Jackson CE, Barohn RJ, Ptacek LJ: Paramyotonia congenita: Abnormal short exercise test, and improvement after mexiletine therapy. Muscle Nerve 17:763-768, 1994. 42. Johns RJ, Grob D, Harvey AM: Studies in neuromuscular function. 2. Effects of nerve stimulation in normal subjects and in patients with myasthenia gravis. Bull Johns Hopkins Hosp 99:125-135, 1956. 43. Jolly F: Myasthenia gravis pseudoparalytica. Berliner Klinische Wochenschrift 32:33-34, 1895. 44. Kamenskaya MA, Thesleff S: The neuromuscular blocking action of an isolated toxin from the elapid (oxyuranus scutellactus). Acta Physiol Scand 90:716-724, 1974. 45. Killian JM, Wilfong AA, Burnett L, Appel SH, Boland D: Decremental motor responses to repetitive nerve stimulation in ALS. Muscle Nerve 17:747-754, 1994. 46. Kimura J: Electromyography and Nerve Stimulation Techniques: Clinical Applications (Japanese). Igaku Shoin, Tokyo, 1990. 47. Koenigsberger MR, Patten B, Lovelace RE: Studies of neuromuscular function in the newborn. 1. A comparison of myoneural function in the full term and the premature infant. Neuropadiatrie 4:350-361, 1973. 48. Krarup C: Electrical and mechanical responses in the platysma and in the adductor pollicis muscle: In normal subjects. J Neurol Neurosurg Psychiatry 40:234-240, 1977a. 49. Krarup C: Electrical and mechanical responses in the platysma and in the adductor pollicis muscle: In patients with myasthenia gravis. J Neurol Neurosurg Psychiatry 40:241-249, 1977b. 50. Lagueny A, Marthan R, Schuermans P, Ferrer X, Julien J: Single fiber EMG and spectral analysis of surface EMG in myotonia congenita with or without transient weakness. Muscle Nerve 17:248-250, 1994. 51. Lambert EH: Electromyographic responses to repetitive stimulation of nerve. American Academy of Neurology, Special Course #16, April 23-28, 1979. 52. Lambert EH: Neuromuscular transmission studies. American Association of Electromyography and Electrodiagnosis, Third Annual Continuing Education Course, September 25, 1980. 53. Lambert EH, Millikan CH, Eaton LM: Stage of neuromuscular paralysis in myotonia (abstr). Am J Physiol 171:741, 1952.
Techniques of Repetitive Stimulation 54. Lambert EH, Rooke ED: Myasthenic state and lung cancer. In Brain, WR, Norris, FH, Jr (eds): Contemporary Neurology Symposia, Vol 1, The Remote Effects of Cancer on the Nervous System. Grune & Stratton, New York, 1965, pp 67-80. 55. Lambert EH, Rooke ED, Eaton LM, Hodgson CH: Myasthenic syndrome occasionally associated with bronchial neoplasm: Neurophysiologic studies. In Viets HR (ed): Myasthenia Gravis, The Second International Symposium Proceedings. Charles C Thomas, Springfield, 111, 1961, pp 362-410. 56. LoMonaco M, Milone M, Padua L, Tonali P: Voluntary contraction in the detection of compound muscle action potential facilitation in Lambert-Eaton myasthenic syndrome. Muscle Nerve 20:1207-1208, 1997. 57. Lomonaco M, Milone M, Valente EM, Padua L, Tonali P: Low-rate nerve stimulation during regional ischemia in the diagnosis of muscle glycogenosis. Muscle Nerve 19:1523-1529, 1996. 58. Lundberg PO, Stalberg E, Thiele B: Paralysis periodica paramyotonica: A clinical and neurophysiological study. J Neurol Sci 21:309-321, 1974. 59. Ma DM, Wasserman EJL, Giebfried J: Repetitive stimulation of the trapezius muscle: Its value in myasthenic testing. Muscle Nerve 3:439-440, 1980. 60. Maddison P, Newsom-Davis J, Mills KR: Decay of postexercise augmentation in the LambertEaton myasthenic syndrome: Effect of cooling. Neurology 50:1083-1087, 1998. 61. Maher J, Grand'Maison F, Nicolle MW, Strong MJ, Bolton CF: Diagnostic difficulties in myasthenia gravis. Muscle Nerve 21:577-583, 1998. 62. Maselli RA, Ellis W, Mandler RN, Sheikh F, Senton G, Knox S, Salari-Namini H, Agius M, Wollmann RL, Richman DP: Cluster of wound botulism in California: Clinical, electrophysiologic, and pathologic study. Muscle Nerve 20:12841295, 1997. 63. Mayer RF, Williams IR: Incrementing responses in myasthenia gravis. Arch Neurol 31:24-26, 1974. 64. McManis PG, Lambert EH, Daube JR: The exercise test in periodic paralysis. Muscle Nerve 9:704-710, 1986. 65. Mihelin M, Trontelj JV, Stalberg E: Muscle fibre recovery functions studied with double pulse stimulation. Muscle Nerve 14:739-747, 1991. 66. Mulder DW, Lambert EH, Eaton LM: Myasthenic syndrome in patients with amyotrophic lateral sclerosis. Neurology 9:627-631, 1959. 67. Oh SJ, Head T, Fesenmeier J, Claussen G: Peroneal nerve repetitive nerve stimulation test: Its value in diagnosis of myasthenia gravis and Lambert-Eaton myasthenic syndrome. Muscle Nerve 18:867-873, 1995. 68. Oh SJ, Kim DS, Head TC, Claussen GC: Lowdose guanidine and pyridostigmine: relatively safe and effective long-term symptomatic therapy in Lambert-Eaton myasthenic syndrome. Muscle Nerve 20:1146-1152, 1997.
277 69. Oh SJ, Kim DE, Kuruoglu R, Brooks J, Claussen G: Electrophysiological and clinical correlations in the Lambert-Eaton myasthenic syndrome, (short report) Muscle Nerve 19:903906, 1996. 70. Ozdemir C, Young RR: The results to be expected from electrical testing in the diagnosis of myasthenia gravis. Ann NY Acad Sci 274: 203-225, 1976. 71. Pittinger C, Adamson R: Antibiotic blockade of neuromuscular function. Ann Rev Pharmacol 12:169-184, 1972. 72. Ricker K, Hertel G, Stodieck S: The influence of local cooling on neuromuscular transmission in the myasthenic syndrome of Eaton and Lambert. J Neurol 217:95-102, 1977. 73. Ricker K, Koch MC, Lehmann-Hom F, Pongratz D, Otto M, Heine R, Moxley RT III: Proximal myotonic myopathy: A new dominant disorder with myotonia, muscle weakness and cataracts. Neurology 44:1448-1452, 1994. 74. Ricker K, Koch MC, Lehmann-Horn F, Pongratz D, Speich N, Reiners K, Schneider C, Moxley RT III: Proximal myotonic myopathy: Clinical features of a multisystem disorder similar to myotonic dystrophy. Arch Neurol 52:25-31, 1995. 75. Ricker K, Mertens HG: Myasthenic reaction in primary muscle fibre disease. Electroencephalogr Clin Neurophysiol 25:413-414, 1968. 76. Ricker K, Samland O, Peter A: Elektrische und mechanische Muskelreaktion bei Adynamia episodica und Paramyotonia congenita nach Kaltecinwirkung und Kaliumgabe. J Neurol 208:95-108, 1974 77. Rudel R, Keller M: Intracellular recording of myotonic runs in dantrolene-blocked myotonic muscle fibres. In Bradley WG, GardnerMedwin D, Walton JN (eds): Recent Advances in Myology. Proceedings of the Third International Congress on Muscle Diseases, Excerpta Medica, Amsterdam, 1975, pp 441-445. 78. Rudel R, Lehmann-Horn F, Ricker K, Kuther G: Hypokalemic periodic paralysis: In vitro investigation of muscle fiber membrane parameters. Muscle Nerve 7:110-120, 1984. 79. Rutchik JS, Rutkove SB: Effect of temperature on motor responses in organophosphate intoxication, (short report) Muscle Nerve 21:958960, 1998. 80. Rutkove SB, Kothari MJ, Shefner JM: Nerve, muscle, and neuromuscular junction electrophysiology at high temperature. Muscle Nerve 20:431-436, 1997. 81. Rutkove SB, Shefner JM, WangAK, Ronthal M, Raynor EM: High-temperature repetitive nerve stimulation in myasthenia gravis. Muscle Nerve 21:1414-1418, 1998. 82. Sander HW, Tavoulareas G, Chokroverty S: Heart sensitive myotonia in proximal myotonic myopathy. Neurology 47:956-962, 1996. 83. Sander HW, Tavoulareas GP, Quinto CM, Menkes DM, Chokroverty S: The exercise test distinguishes proximal myotonic myopathy from myotonic dystrophy. (short report) Muscle Nerve 20:235—237, 1997. 84. Sanders DB: Clinical neurophysiology of disor-
278
85.
86.
87. 88.
89. 90. 91.
92.
Assessment of Neuromuscular Transmission
ders of the neuromuscular junction. J Clin Neurophysiol 10:167-180, 1993. Sanders DB: Electrodiagnostic testing of neuromuscular transmission. AAN 49th Annual Meeting: Principles and Pitfalls in the Practice of EMG and NCS, 327-337, 1997. Sanders DB, Howard JF, Massey JM: 3,4-Diaminopyridine in Lambert-Eaton myasthenic syndrome and myasthenia gravis. Ann NY Acad Sci 681:588-590, 1993. Schumm F, Stohr M: Accessory nerve stimulation in the assessment of myasthenia gravis. Muscle Nerve 7:147-151, 1984. Schwartz MS, Stalberg E: Myasthenia gravis with features of the myasthenic syndrome: An investigation with electrophysiologic methods including single-fiber electromyography. Neurology 25:80-84, 1975. Shapiro BE, Soto O, Shafgat S, Blumenfeld H: Adult botulism. Muscle Nerve 20:100-102, 1997. Simpson JA: Myasthenia gravis: A new hypothesis. Scott Med J 5:419-436, 1960. Slomic A, Rosenfalck A, Buchthal F: Electrical and mechanical responses of normal and myasthenic muscle with particular reference to the staircase phenomenon. Brain Res 10:1-78, 1968. Somnier FE, Trojaborg W: Neurophysiological evaluation in myasthenia gravis: A comprehensive study of a complete patient population.
93. 94. 95. 96. 97.
98.
99.
100.
Electroencephalogr Clin Neurophysiol 89:7387, 1993. Streib E: AAEE Minimonograph #27: Differential diagnosis of myotonic syndromes. Muscle Nerve 10:603-615, 1987. Swift TR: Weakness from magnesium-containing cathartics: Electrophysiologic studies. Muscle Nerve 2:295-298, 1979. Thornton CA, Griggs R, Moxley RT: Myotonic dystrophy with no trinucleotide repeat expansion. Ann Neurol 35:269-272, 1994. Tim RW, Sanders DB: Repetitive nerve stimulation studies in the Lambert-Eaton syndrome. Muscle Nerve 17:995-1001, 1994. Trontelj JV, Stalberg EV: Single fiber EMG and spectral analysis of surface EMG in myotonia congenita with or without transient weakness, (letters to the editor) Muscle Nerve 18:252-253, 1995. Van Dijk JG, Lammers GJ, Wintzen AR, Molenaar PC: Repetitive CMAPs: Mechanisms of neural and synaptic genesis. Muscle Nerve 19:1127-1133, 1996. Ward CD, Murray NMF: Effect of temperature on neuromuscular transmission in the EatonLambert syndrome. J Neurol Neurosurg Psychiatry 42:247-249, 1979. Wright GP: The neurotoxins of clostridium botulinum and clostridium tetani. Pharmacol Rev 7:413-465, 1955.
Chapter 11 ACTIVATION PROCEDURES AND OTHER METHODS
1. INTRODUCTION 2. PROVOCATIVE TECHNIQUES
Ischemic Test Regional Curare Test 3. ELECTROMYOGRAPHY
Varying Motor Unit Potentials Jitter and Blocking 4. OTHER TECHNIQUES Miniature End-Plate Potentials Tonography Stapedius Reflex Tests for Oculomotor Function
1 INTRODUCTION Not all muscles show electrophysiologic abnormalities in diseases of the neuromuscular junction. Thus, conventional nerve stimulation techniques may fail to substantiate the clinical diagnosis. Imposing metabolic stress with regional ischemia or application of curare reduces the safety factor of the system, thus increasing the yields of detection of neuromuscular transmission abnormalities. Obviously, such activation maneuvers used to improve the diagnostic yields must avoid false-positive results. A number of other electrodiagnostic methods supplement the technique of paired or repetitive electrical stimuli. Measuring the fatigability of a few accessible muscles serves to document a decremental tendency. Electromyographic analyses of motor unit potentials can also elucidate
the stability of neuromuscular transmission. In particular, assessment of the extraocular muscles provides useful information in patients with ocular myasthenia. Single-fiber electromyography has added the calculation of jitter as a very sensitive and most useful measure of neuromuscular transmission. The in vitro analysis of miniature end-plate potentials (MEPPs) of intercostal muscles allows a direct quantitative measure of neuromuscular transmission.6 This technique, however, is not feasible as a routine clinical test. Other laboratory methods of diagnostic value include tonography and infrared optokinetic nystagmography for ocular myasthenia,26 1and stapedius reflex for facial weakness. Electrophysiologic methods also help in quantitating edrophonium (Tensilon) and neostigmine (Prostigmin) tests. Measurement of serum antibodies against the muscle end plate detects immunologic abnormalities in most patients 279
Assessment of Neuromuscular Transmission
280
with myasthenia gravis.13,14 As a part of the clinical evaluation, pulmonary function tests provide useful objective criteria in documenting neuromuscular fatigue.9 2
PROVOCATIVE TECHNIQUES
Ischemic Test A double-step test may increase the diagnostic sensitivity of nerve stimulation techniques in some patients with mild, generalized, or ocular myasthenia gravis.3,5 The first step consists of continuous supramaximal stimulation of the ulnar nerve at 3 Hz for 4 minutes and recording the compound muscle action potential from the ulnar-innervated intrinsic hand muscles or the flexor carpi ulnaris in the forearm. Thereafter, a train of 3 Hz stimulation for several shocks at 30 s intervals determines the amount of decrement within each three-per-second trial. In cases of a negative or equivocal result, the second step consists of the same procedure under ischemia induced by a cuff inflated above arterial pressure proximal to the stimulation site. The double-step test has helped elucidate different degrees of myasthenic involvement in the same patient. How much additional help this procedure provides in the early diagnosis of myasthenia gravis remains unclear.19 Regional Curare Test The amount of acetylcholine (ACh) released with each nerve impulse normally produces an end-plate potential (EPP) that substantially exceeds the critical level for excitation of all the muscle fibers. This margin of safety protects the neuromuscular transmission with a latent deficit, rendering clinical and electrophysiologic evaluation difficult. Curare causes a nondepolarizing block by competing with ACh for the end-plate receptors. Its administration, therefore, reduces or eliminates the functional reserve and elucidates the defect of neuromuscular transmission that otherwise escapes detection.7,8 In one
series of 600 patients, the repetitive stimulation at the wrist and shoulder11verified the diagnosis in 320 (53 percent). In the remaining 280 patients, the regional curare test revealed abnormality in 72 (26 percent), including 52 (74 percent) of 70 patients with definite generalized myasthenia gravis, 13 (10 percent) of 136 patients with possible generalized myasthenia gravis, and 7 (14 percent) of 49 with ocular myasthenia. The concentration of curare that reaches the muscle depends on diffusion through the volume of tissue, which probably varies from one case to the next. Thus, titrated dosages of curare do not always differentiate normal and pathologic responses.10 Studies have revealed undue sensitivity to 23 curare not only in myasthenia gravis18 but also in amyotrophic lateral sclerosis and muscular weakness after administration of antibiotics.21 Therefore, an abnormal regional curare test indicates a defect in neuromuscular transmission but not necessarily myasthenia gravis. These uncertainties notwithstanding, the regional curare test, as a measure of last resort, supplements conventional nerve stimulation techniques in difficult cases. Patients with established myasthenia gravis should not undergo the procedure, to avoid the possible risk. 3
ELECTROMYOGRAPHY
In normal muscles, a motor unit firing repetitively under voluntary control gives rise to identical potentials in waveform and amplitude every time it discharges, as long as the needle electrode remains in the same location relative to the generator source. This does not hold in myasthenic muscles, because nerve impulses may not always depolarize all the individual muscle fibers to the critical level, as the result of a reduced margin of safety. Intermittent failure of some muscle fibers innervated by the same axon in response to successive nerve impulse causes amplitude variability of the recurring motor unit potentials. Blocking at the neuromuscular junction also explains diminished mean amplitude and duration of
Activation Procedures and Other Methods motor unit potentials in myasthenic muscles.
Varying Motor Unit Potentials Electromyography can assess the stability of isolated motor unit potentials by slowly advancing or retracting the needle for optimal display of the repetitive discharges. During minimal contraction of the muscle, the amplitude variability of an isolated potential, heard over the loudspeaker, alerts the examiner to search for unstable motor unit discharges. This method does not necessarily provide accurate quantitative assessment of neuromuscular transmission. The needle examination, applicable to any muscles, including those not tested by the stimulation technique, has the added advantage of not requiring muscle immobilization. For example, electromyography helps establish the diagnosis of ocular myasthenia (see Chapter 15-3). The administration of anticholinesterase reverses the abnormalities of motor unit potentials in patients with myasthenia gravis. Thus, the injection of edrophonium (Tensilon) increases motor unit potentials recorded in the extensor digitorum communis by 30-130 percent in amplitude and 10-25 percent in duration.20 In patients with ocular myasthenia gravis, a progressive decrease in amplitude and frequency during a prolonged period of voluntary contraction partially reverses after intravenous administration of edrophonium.24
Jitter and Blocking The single-fiber recording has proven useful in early detection of neuromuscular disturbances as an important adjunct technique 27 in the evaluation of myasthenia gravis. -28 As described in detail later (see Chapter 16-5) the method consists of recording a pair of single-fiber potentials simultaneously and measuring fluctuation of the neuromuscular transmission by the stability of the interpeak intervals. Either blocking or increased jitter characterizes neuromuscular disturbances.
281
The end plate potential (EPP) generated by voluntary contraction normally reaches the threshold in all muscle fibers. If the EPP falls short of this critical level at some neuromuscular junction, those muscle fibers fail to discharge. This blocking affecting only some fibers of a motor unit reduces the size of the motor unit potential on standard needle recordings. If an EPP barely reaches the necessary level, its rate of rise falls below the normal range, delaying, rather than blocking, the action potential. This abnormality escapes detection in conventional electromyography because, unlike blocking, delayed discharge of muscle fibers alters the motor unit potential very little. On single-fiber recording of a pair of potentials, an intermittent delay of the action potential in either fiber increases the jitter or the variability of interpotential interval. This finding, as the first sign of neuromuscular instability, precedes blocking of transmission. Thus, increased jitter heralds variation of motor unit potentials or decrementing response to repetitive stimulation of the nerve. The practice of single-fiber electromyography has added a new dimension to the assessment of neuromuscular transmission, although it requires additional training. Most commercially available instruments have the capability of computerizing the method, which surpasses the manual calculation of the recorded responses. 4
OTHER TECHNIQUES
Minature End-Plate Potentials Microelectrode recordings from single intercostal muscles provide the only means of measuring the size of minature endplate potentials (MEPP) and the number of the acetylcholine (ACh) quanta released per volley of nerve impulses. These determinations in turn can precisely characterize the abnormality of neuromuscular transmission. The method helps elucidate the specific pathophysiology underlying the deficits in production or mobilization of ACh. It also measures the sensitivity of the motor end plate by quantitative as-
Assessment of Neuromuscular Transmission
282
sessment. This method, which depends on intercostal muscle biopsy, lies beyond the scope of routine clinical tests. In selected cases that pose a diagnostic challenge, however, it helps differentiate myasthenia gravis, the myasthenic syndrome, and other disorders involving the neuromuscular junction. A spectral analysis of endplate noise recorded by a conventional monopolar needle may help evaluate ACh receptor ion channel kinetics, but its clinical value waits further clarification.15 Tonography Other techniques not ordinarily used in an electromyographic laboratory include edrophonium (Tensilon) tonography. The intraocular pressure results in part from contraction of the extraocular muscles. Thus, measurements of the pressure with an electronic tonometer reveal the effect of anticholinesterase on ocular motility. Some investigators advocate simultaneous recording of muscle action potentials with needle electrodes placed in the extraocular muscles. In normal subjects, intraocular pressure may fall, on average, 1.6-1.8 mm Hg over a 1 minute period after an intravenous injection of edrophonium up to 10 mg.4 Patients with decreased extraocular tone, as in ocular myasthenia, have low intraocular pressures. The administration of edrophonium produces changes in tonography coincident with a moderate increase in electrical activity in the extraocular muscles. A sudden increase in extraocular muscle tone alters intraocular pressure by a mean of 1.6 mm Hg within 35 s of injection. This phenomenon does not necessarily imply ocular myasthenia, being also seen, for example, in ocular myositis without other features of myasthenia gravis.30 Intraocular pressure may also rise with the Valsalva maneuver. In this case, a control injection of saline can identify a false-positive result.
Stapedius Reflex The stapedius muscles contract bilaterally in response to unilateral sound stimula-
tion. This contraction in turn dampens the acoustic sensitivity of the middle ear and prevents hyperacusis. Thus, impedance audiometry can measure the function of the stapedius muscle. In normal subjects, a sound stimulus 70-100 dB above the hearing threshold elicits the stapedius reflex. It shows no decay during sustained contraction for up to 1 minute with stimulus frequencies of 250-1000 Hz. In patients with myasthenia gravis, weakened stapedius muscles enhance transmission of sound in the 1-4 kHz range, resulting in hyperacusis. Here, only high-intensity sound can induce the acoustic reflex.17 In addition, reflex contraction of the stapedius muscle shows a rapid decrement, analogous to the similar response of the limb muscles to repetitive electrical stimulation of the nerve.1 The administration of edrophonium enhances the acoustic reflex, diminishes hyperacusis, and improves the decay of the stapedius reflex in response to repetitive sound stimulation. In some patients with myasthenia gravis, testing the stapedius reflex may reveal 29 the only electrophysiologic abnormality. In one study, stapedial reflex showed clear abnormalities in 84 percent of the patients with myasthenia gravis as compared with 56 percent by repetitive stimulation and 91 12percent by single-fiber electromyography.
Tests for Oculomotor Function Electronystagmography provides quantitative measurements of amplitude, velocity, and frequency of optokinetic nystagmus to document fatigue of extraocular muscles. In patients with ocular myasthenia, edrophonium induces an increase in previously reduced oculomotor function.2 In one series, electrooculography revealed neuromuscular fatigue in 50 percent of myasthenic patients.4 The infrared reflection technique improves the sensitivity of the test with the use of numeric criteria in grading neuropharmacologic effects on oculomotor fatigue.25 For example, velocity of saccade measured by this means increases after administration of edrophonium.16 The Lancaster red-green
Activation Procedures and Other Methods test also detects oculomotor fatigue, which improves after rest or with administration of edrophonium in patients with myasthenia gravis.22
REFERENCES 1. Blom S, Zakrisson JE: The stapedius reflex in the diagnosis of myasthenia gravis. J Neurol Sci 21:71-76, 1974. 2. Blomberg LH, Persson T: A new test for myasthenia gravis. Preliminary report. Acta Neurol Scand 41 (Suppl 13): 363-364, 1965. 3. Borenstein S, Desmedt JE: New diagnostic procedures in myasthenia gravis. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 350-374. 4. Campbell MJ, Simpson E, Crombie AL, Walton JN: Ocular myasthenia: Evaluation of Tensilon tonography and electronystagmography as diagnostic tests. J Neurol Neurosurg Psychiatry 33:639-646, 1970. 5. Desmedt JE: The neuromuscular disorder in myasthenia gravis. 1. Electrical and mechanical response to nerve stimulation in hand muscles. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 241-304. 6. Elmqvist D: Neuromuscular transmission defects. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology, Vol 1. Karger, Basel, 1973, pp 229-240. 7. Feldman SA, Tyrrell MF: A new theory of the termination of action of the muscle relaxants. Proc R Soc Med 63:692-695, 1970. 8. Foldes FF, Klonymus DH, Maisel W, Osserman KE: A new curare test for the diagnosis of myasthenia gravis. JAMA 203:649-653, 1968. 9. Griggs RC, Donohoe KM, Utell MJ, Goldblatt D, Moxley RT III: Evaluation of pulmonary function in neuromuscular disease. Arch Neurol 38:9-12, 1981. 10. Hertel G, Ricker K, Hirsch A: The regional curate test in myasthenia gravis. J Neurol 214:257-265, 1977. 11. Horowitz SH, Sivak M: The regional curare test and electrophysiologic diagnosis of myasthenia gravis: Further studies. Muscle Nerve 1:432434, 1978. 12. Kramer LD, Ruth RA, Johns ME, Sanders DB: A comparison of stapedial reflex fatigue with repetitive stimulation and single fiber EMG in myasthenia gravis. Ann Neurol 9:531-536, 1981. 13. Lefvert AK, Bergstrom K, Matell G, Osterman PO, Pirskanen R: Determination of acetylcholine receptor antibody in myasthenia gravis: Clinical usefulness and pathogenetic implications. J Neurol Neurosurg Psychiatry 41:394-403, 1978. 14. Lindstrom JM, Lennon VA, Seybold ME, Whit-
283
tingham S: Experimental autoimmune myasthenia gravis and myasthenia gravis: Biochemical and immunochemical aspects. Ann NY Acad Sci 274:254-274, 1976. 15. Maselli RA: End-plate electromyography: Use of spectral analysis of end-plate noise. Muscle Nerve 20:52-58, 1997. 16. Metz HS, Scott AB, O'Meara DM: Saccadic eye movements in myasthenia gravis. Arch Ophthalmol 88:9-11, 1972. 17. Morioka WT, Neff PA, Boisseranc TE, Hartman PW, Cantrell RW: Audiotympanometric findings in myasthenia gravis. Arch Otolaryngol 102:211-213, 1976. 18. Mulder DW, Lambert EH, Eaton LM: Myasthenic syndrome in patients with amyotrophic lateral sclerosis. Neurology 9:627-631, 1959. 19. Ozdemir C, Young RR: The results to be expected from electrical testing in the diagnosis of myasthenia gravis. Ann NY Acad Sci 274:203-235, 1976. 20. Pinelli P: The effect of anticholinesterases on motor unit potentials in myasthenia gravis. Muscle Nerve 1:438-441, 1978. 21. Pittinger C, Adamson R: Antibiotic blockade of neuromuscular function. Ann Rev Pharmacol 12:169-184, 1972. 22. Retzlaff, JA, Kearns, TP, Howard, FM, Jr, and Cronin, ML: Lancaster red-green test in evaluation of edrophonium effect in myasthenia gravis. Am J Ophthalmol 67:13-21, 1969. 23. Rowland LP, Aranow H Jr, Hoefer PFA: Observations on the curare test in the differential diagnosis of myasthenia gravis. In Viets HR (ed): Myasthenia Gravis, The Second International Symposium Proceedings. Charles C Thomas, Springfield, 111, 1961, pp 411-434. 24. Sears ML, Walsh FB, Teasdall RD: The electromyogram from ocular muscles in myasthenia gravis. Arch Ophthalmol 63:791-798, 1960. 25. Spector RH, Daroff RB: Edrophonium infrared optokinetic nystagmography in the diagnosis of myasthenia gravis. Ann NY Acad Sci 274:642651, 1976. 26. Spector RH, Daroff RB, Birkett JE: Edrophonium infrared optokinetic nystagmography in the diagnosis of myasthenia gravis. Neurology 25:317-321, 1975. 27. Stalberg E, Trontelj JV: Single fiber electromyography. In: Healthy and Disease Muscle. Raven Press, New York, 1994. 28. Stalberg E, Trontelj JV, Schwartz MS: Singlemuscle-fiber recording of itter phenomenon in patients with myasthenia gravis and in members of their families. Ann NY Acad Sci 274: 189-202, 1976. 29. Warren WR, Gutmann L, Cody RC, Flowers P, Segat AT: Stapedius reflex decay in myasthenia gravis. Arch Neurol 34:496-497, 1977. 30. Wray SH, Pavan-Langston D: A reevaluation of edrophonium chloride (Tensilon) tonography in the diagnosis of myasthenia gravis: With observations on some other defects of neuromuscular transmission. Neurology 21:586-593, 1971.
This page intentionally left blank
Part IV ELECTROMYOGRAPHY
This page intentionally left blank
Chapter 12 ANATOMY AND PHYSIOLOGY OF THE SKELETAL MUSCLE
1. INTRODUCTION 2. FUNCTIONAL ANATOMY Gross Anatomy of Muscle Excitability and Conductivity Myofibrils and Myofilaments Mechanism of Contraction 3. TYPES OF MUSCLE FIBERS Type I and Type II Fibers Fast and Slow Twitch Fibers Fast and Slow Muscles Effect of Muscle Injury, Denervation, and Innervation 4. STRETCH-SENSITIVE RECEPTORS Anatomy of Muscle Spindles Function of Muscle Spindles Golgi Tendon Organ 5. ANATOMY OF THE MOTOR UNIT Innervation Ratio Distribution of Muscle Fibers 6. PHYSIOLOGY OF THE MOTOR UNIT Animal Experiments Recruitment Twitch Characteristics Rate Coding
1 INTRODUCTION The skeletal muscles comprise the extrafusal and intrafusal fibers, which show distinct anatomic and physiologic features. The alpha motor neurons innervate the extrafusal fibers that occupy the bulk of muscle mass as contractile elements. The skeletal muscles usually have innervation zones with abundant motor end plates in
the middle length of the muscle, though the detailed configurations vary among different subjects.110 Many muscles in the limbs have multiple motor points169 that in part dictate their contraction characteristics.101 Mammalian skeletal muscles may consist of two or more separate subdivisions known as neuromuscular compartments, with unique anatomic and functional characteristics.44,98,100,152 The gamma motor neurons subserve the stretch-sensitive in-
287
Electromyography
288
trafusal fibers that constitute the muscle spindles found in parallel with the extrafusal fibers, which contract to generate force. The Golgi tendon organs, aligned in series with the tendon of the extrafusal fibers, also respond to stretch. The spindles and Golgi tendon organ continuously monitor and regulate the tonus of the reflexive or volitional muscle contraction. The motor unit, the smallest contractile element, consists of a single motor neuron and all the muscle fibers innervated by its axon. A nerve impulse initiates muscle contraction in two distinct steps: neuromuscular transmission and electromechanical coupling (see Chapter 9-3 and 9-4). Acetylcholine (ACh), released at the neuromuscular junction, depolarizes the end-plate region generating an action potential, which then propagates along the muscle membrane. As the impulse reaches the triad, depolarization of the transverse tubules releases ionized calcium (Ca2+) into the sarcoplasm. The interaction between calcium and the thin filaments triggers electromechanical coupling, leading to the formation of bridges between the thin and thick filaments. The sliding of thin filaments between the thick filaments shortens the muscle fibers. This section will present a description of the anatomy of the contractile elements, the mechanism underlying the shortening of the muscle fibers, and the anatomy and physiology of motor units. 2
FUNCTIONAL ANATOMY
Gross Anatomy of Muscle A connective tissue called epimysium covers the surface of each muscle. Inside this sheath are many fascicles bound by the coarse sleeves of the connective tissue perimysium (Fig. 12-1). Individual fascicles contain many muscle fibers, each surrounded by a delicate network of fine connective tissue, called endomysium. A muscle fiber, the smallest anatomic unit capable of contraction, averages in diameter 10 /AHI in a newborn and 50 um in an adult.22 Individual muscle fibers range from
Figure 12-1. Anatomic composition of the skeletal muscle. The epimysium surrounds the entire muscle (a), which consists of many fascicles bound by perimysium (b). Individual muscle fibers (c) in the fascicle are covered by endomysium. Each muscle fiber contains many bundles of myofibrils (d), which in turn consist of thin (e) and thick (f) myofilaments. Thin actin filaments slide past thick myosin filaments during muscle contraction (g).
2 to 12 cm in length, some extending the entire length of the muscle and others only through a short segment of the total length. The sarcolemma on the surface membrane of a muscle fiber contains multiple nuclei distributed beneath the thin sheath.
Excitability and Conductivity The muscle membrane has functional properties of excitability and conductivity similar to those of an axon. Thus, a myoelectric signal originating from a neuromuscular junction propagates in109both the proximal and distal directions. An ordinary electromyography instrument suf-
Anatomy and Physiology of the Skeletal Muscle fices to measure muscle fiber conduction velocity using needle recording. With surface studies, computerized data analyses of frequency and time domain give rise to an average estimate from many motor units at different contraction levels.89,196 An averaging method with arrays of surface electrodes shows a high correlation of conduction velocity with twitch and threshold forces but not with rise time. 132 The measurements with electric stimulation of single or a bundle of fibers, in general, yield lower values than those obtained during voluntary contraction.4,58,201 When measuring electrically elicited contraction of single motor units, the higher the stimulation rate, the133 greater the velocity within certain ranges. During submaximal contractions, two opposing factors influence the average values of muscle fiber conduction velocity: Increase with recruitment of fresh motor units, and decrease with fatigue of already active motor units.5 On average, the conduction velocity increases with the level of contraction force either measured with surface electrodes or needle electrode.128,150 Muscle fiber conduction velocity also changes with length. 19,91 In one study180using single fiber electromyography, propagation velocity increased by 33 percent on shortening and decreased by 22 percent on elongating the muscle fiber. These length-dependent changes may contribute to the supernormal phase of muscle fiber propagation velocity and interdischarge interval-dependent myogenic jitter seen in single fiber studies. Compared to the nerve axons, muscle fibers conduct considerably more slowly, with an estimated rate of 3 to 5 22,53,78,107,108,134,153,161,179,195
propagation velocities increase with age, body height, and muscle diameter in the growing normal child.106 Surface and needle recording reveal a reduced muscle fiber conduction velocity in myopathies,201 although a needle study may show the change more138 clearly.184 Peripheral vascular diseases and high-dose methylprednisolone therapy183 also alter surface myoelectric signals. Despite some encouraging results, muscle fiber conduction studies have found only limited clinical value as a diagnostic measure.
289
Myofibrils and Myofilaments The semi-fluid intracellular content of a muscle fiber, called sarcoplasm, contains many bundles of cylindrical myofibrils. They appear as a thin, threadlike substance with light and dark bands of striations under the light microscope. Myofibrils consist of two types of myofilaments, which represent the basic substrates for the contraction of muscle fibers. The transverse striations seen by light microscopy result from their specific arrangements. The structural subunit, called the sarcomere, extends between two adjacent Z lines. The center of the sarcomere contains the longitudinally oriented thick myosin myofilaments. The thin actin filaments extend from either side of the Z line into the two adjacent sarcomeres to interdigitate with the myosin filaments. The thick filaments consist of only myosin molecules, which form parallel elongated rods. The thin filaments contain not only actin molecules but also two other proteins, troponin and tropomyosin. Globular-shaped troponins are attached to each end of the elongated tropomyosin molecule, which, in turn, is intimately bound to several actin molecules along its interwoven course (Fig. 12-2). During muscle fiber contraction, actin filaments slide relative to the myosin filaments. This brings the adjacent Z lines closer together, shortening the sarcomere, rather than individual filaments.25
Mechanism of Contraction The mechanism of the sliding begins with the formation of calcium (Ca2+)-dependent bridges that link the actin and myosin filaments. At rest, tropomyosin physically blocks the formation of bridges between myosin and actin. The propagation of the action potential into the sarcoplasmic reticulum via the transverse tubules releases calcium from the terminal cistern of the longitudinal tubules. The free calcium binds to troponin, the only calcium-receptive protein in the contractile system. This interaction shifts the position of tropomyosin relative to the actin molecule,
290
Electromyography
Figure 12-2. Fine structure of the thin actin filament with actin molecules attached to globe-shaped troponin and rod-shaped tropomyosin in an orderly arrangement. [From Ebashi, Endo, and Ohtsuki,52 with permission.]
allowing the globular heads of myosin to gain access to the actin molecules. Myosin-actin cross-bridges pull the actin filaments past the myosin filaments. The tension develops in proportion to the number of cross-bridges formed by this chemical interaction. The dissociation of actin and myosin by adenosine triphosphate (ATP) shears old bridges to allow further sliding with new bridges. Without a sustained muscle action potential, ATP-dependent active transport sequesters calcium into the sarcoplasmic reticulum. The removal of calcium from troponin allows tropomyosin to return to the resting position, and the muscle relaxes. Muscle contractility depends in part on extracellular calcium concentration.103,104 In McArdle disease, characterized by deficiency of muscle phosphorylase, this initial step of muscle relaxation
does not occur, presumably because of an insufficient amount of ATP. Failure of relaxation results in persistent shortening of the muscle in the absence of ongoing muscle action potentials. This condition, called contracture, typically develops when patients exercise under ischemic conditions (Fig. 12-3). In porcine malignant hyperthermia, a mutation of the calcium channel in the skeletal muscle sarcoplasmic reticulum causes excessive release of calcium into the myoplasm, leading to contracture.156 Although the degree of muscle contraction determines strength and endurance, the amount of force generated serves only partially as an index of motor skill. Functional alteration, for example, occurs with sarcopenia or loss of lean tissue with aging, and its metabolic and physiologic consequences.51 ,154
Figure 12-3. Contracture during ischemic exercise in a 66-year-old man with McArdle disease. With an inflated pressure tourniquet placed around the arm and a concentric needle electrode inserted into the flexor digitorum profundus, the patient exercised the forearm flexors. Contracture began 45 seconds after the start of ischemic exercise (arrow in A), and persisted (B). Electrical activity returned 15 minutes after the release of the cuff (arrow in C). [Courtesy of E. Peter Bosch, M.D., Mayo Clinic, Scottsdale.]
291
Anatomy and Physiology of the Skeletal Muscle
IIA, IIB, and IIC, emerge according to their ATPase reactions (Table 12-1) after preincubation at different pH values.17,18,56 The subdivision of muscle fibers depends Type IIC fibers constitute fetal precursor cells, rarely seen in adult muscles. on their histologic and physiologic proThe myosine ATPase content dictates files. Important determining factors inthe speed of contraction,6 which forms the clude enzymatic properties demonstrated basis for the physiologic subdivision of by histochemical reactions; rate of rise in muscle fibers. Thus, in general, physiotwitch tension, regulating the speed of logic data correlate the slow twitch fibers contraction; degree of fatigability; and the to histochemical 36,94 type I, and fast twitch nature of motor innervation. 16° Table fibers to type II, though exceptions 12-1 summarizes the commonly used abound. For example, histochemically classification of muscle fibers into type I mixed extensor digitorum longus of the rat and type II according to histochemical recontains only fast fibers;41 slow soleus actions;18,50,56 slow (S), fast resistant muscle of eels shows greater myosin (FR), and fast fatiguing (FF), based on ATPase activity than fast gastrocnemius twitch and fatigue characteristics;35 or slow oxidative (SO), fast oxidative gly- muscle. Therefore, the intensity of histochemical ATPase reaction cannot serve as colytic (FOG), and fast glycolytic (FG), by the sole criterion in distinguishing fast twitch and enzymatic properties.141 and slow twitch fibers.37 The growth of muscle cross-sectional Type I and Type II Fibers area from childhood to adult age reflects an increase in mean fiber size from 10-12 Histochemical reactions (Fig. 12-4) reveal fjLW. shortly after birth to 40-60 um at age two types of human muscle fibers. Type I 15-20 years.137 Accompanying functional fibers react strongly to oxidative enzymes development includes a change of the such as nicotinamide adenine dinucleotide fiber population with an increase of type dehydrogenase (NADH) and reduced di2 fibers from about 35 percent at the age phosphopyridine nucleotide (DPNH) and of 5, to 50 percent at the age of 20, most weakly to both phosphorylase and mylikely by a transformation of type 1 to type ofibrillar adenosine triphosphatase (AT2 fibers.95 Aging atrophy or sarcopenia bePase). Type II fibers show the reverse regins around age 25 and then acceleractivity.50 Three subtypes of type II fibers, ates,51,96,154 mainly reflecting a loss of 3
TYPES OF MUSCLE FIBERS
Table 12-1 Types of Muscle Fibers Commonly used designations Fiber types18 Twitch and fatigue characteristics35 Twitch and enzymatic properties141 Properties of muscle fibers Resistance to fatigue Oxidative enzymes Phosphorylase (glycolytic) Adenosine triphosphate Twitch speed Twitch tension Characteristics of motor units Size of cell body Size of motor unit Diameter of axons Conduction velocity Threshold for recruitment Firing frequency Frequency of miniature end-plate potentials
Type I Slow (S) Slow oxidative (SO)
Type IIA Fast resistant (FR) Fast oxidativeglycolytic (FOG)
Type II B Fast fatigue (FF) Fast glycolytic (FG)
High High Low Low Low Low
High High High High High High
Low Low High High High High
Small Small Small Low Low Low Low
Large Large Large High High High High
Large Large Large High High High High
292
Electromyography
Figure 12-4. Cross-section of a normal skeletal muscle pH 9.4 in A, and with nicotinamlde adenine dinucleotide resent type II in A and type I in B. [Courtesy of Linda partment of Pathology, University of Iowa Hospitals and
fibers of all types, and to a lesser extent, reduction in fiber size mostly of type 2 fibers.
Fast and Slow Twitch Fibers Muscle fibers differ in their contraction time, force-velocity curves, and rates of decay.141 Slow fibers (S) with high oxidative properties (SO) resist fatigue. Fast resistant (FR) fibers with high oxidative and glycolytic properties (FOG) also resist fatigue, whereas the fast fatigue (FF) fibers with high glycolytic activity35but low oxidative enzyme (FG) do not. These findings suggest that glycolytic capacity generally relates to twitch characteristics, and oxidative capability dictates fatigability. Intracellular recordings have shown that compared with slow fibers, fast glycolytic fibers have greater resting membrane potential, a larger amplitude of the
stained with adenosine triphosphatase (ATPase) at dehydrogenase (NADH) in B. The darker fibers repAnsbacher, M.D., and Michael N. Hart, M.D., DeClinics.]
action potential, higher maximum rates of depolarization and repolarization, and a more variable shape of the repolarization phase.188 The slow twitch fibers have higher antioxidative capacity than the fast twitch fibers.83 The production of lipid peroxides parallels the exercise-induced increase of oxygen uptake in the muscle, showing higher values in more oxidative and better perfused, oxygen-consuming muscle fibers.93
Fast and Slow Muscles In animals, most muscles consist mainly of one muscle fiber type. Slow muscles appear deeper red in color, reflecting a higher myoglobin content, whereas fast muscles tend to show a whitish hue. Functionally, slow muscles have a tonic postural role, like that of the soleus in the cat, whereas fast muscles provide willed
Anatomy and Physiology of the Skeletal Muscle phasic movements, like those of the wing muscles of a chicken. This distinction, however, blurs in humans because most human limb muscles consist of slow and fast twitch motor units in various combinations.30 For example, the slow fibers with contraction times longer than 60 ms constitute a majority in triceps surae, one half in tibialis anterior, one third in biceps brachii, and a small percentage in triceps brachii.27 Slow oxidative fibers occupy 38 and 44 percent of superficial and deep areas in the vastus lateralis and 47 and 61 percent in the vastus medialis.84 Further, fibers of the same types do not necessarily share the same contractile speed in different muscles.147
Effect of Muscle Injury, Denervation, and Innervation Focal injury to a long multinucleated muscle fiber could destroy it totally unless repair takes place immediately at the site of the lesion, sealing the remainder of its length. The satellite cell-derived myoblasts fuse with the injured muscle fiber to undertake such localized repair9,159 without affecting the major gene expression in the uninjured parts of the fiber.200 Transient loss of functional innervation has a permanent effect on the myosin composition.81 After denervation, the pattern of phosphorylation in fast muscle changes to resemble that of slow muscle, a finding consistent with denervation-induced changes observed using other phenotypic markers.118,131,184 Denervation usually causes irreversible muscle atrophy unless denervated muscles receive reinnervation promptly.80 In one study, muscles grafted with nerve implants had a higher mass and generated twice the force compared with denervated muscle receiving only nerve implants without muscle graft.8 Functional recovery also depends critically on the duration of denervation before nerve repair.90 The rate of stimulation dictates the contractile characteristics of muscle fibers in animals,82,102,142,143,144,145,151,170,187 as well as in humans.39 Brachial plexus palsy at birth alters isometric contraction time and half relaxation time of the af-
293
fected muscles.167 The finding suggests that denervation during infancy impairs normal development of muscle contractile properties. In patients with chronic neuromuscular diseases, normal muscle fiber histochemistry persists as long as motor neuron differentiation remains. In patients with long-term spastic hemiplegia, some motor units show greater fatigability and longer twitch contraction times than normal. Thus, the dynamic properties of the muscle seem to change even in upper motor neuron lesions.198 Alterations in histochemical properties may reflect the firing pattern and axonal conduction velocity of the motor neurons.12 Athletes engaged in endurance training have a greater number of slow fibers,63 whereas weight lifters have more fast fibers.176 Exercise training alone, however, induces little change in basic muscle contractility in humans.3,63,176 Hence, motor neuron activity does not suffice in itself to alter the distribution of fast and slow fibers in a muscle. The findings in favor of additional neurotrophic influences66 include effects of neurons on muscle in tissue cultures199 and the inverse relationship of nerve length on the time interval before the development of muscle membrane changes after nerve section.38 The hypertrophy with type I fiber predominance seen in some patients with neuromyotonia may represent conversion of type II fibers based on similar physiologic mechanisms as described in animal models.67,185 Reactive hypertrophy of the masticatory muscle, induced by increased workload, also accompanies progressive type I fiber predominance and type II fiber atrophy.73 After the transplantation of the nerve normally innervating a fast fiber to a slow fiber, the originally slow muscle fiber will acquire the properties of a fast muscle fiber.31,85,146,197 Such a relationship between the type of innervation and muscle activity also determines the mechanical characteristics of contraction in some fibers,28 but not others.190 For example, motor neurons innervating fast twitch muscles have shorter afterhyperpolarization than those supplying slow twitch muscles.54 A study of the effect of cross-innervation in patients with muscle transfer for
294
facial palsies have shown considerably less plasticity than in animal models in the conversion of slow to fast twitch fibers.74 Also, the minimal changes in the spatial distribution of fiber types following selfreinnervation in adults suggests a limited degree of conversion of muscle fibers to myosin heavy chain phenotype matching the motor neuron characteristics.182 Studies using inactivity with and without cross-reinnervation have shown that electrically silent motor neurons can influence type-related skeletal muscle properties.148 Further, activity-dependent fiber-type modulation differs substantially among fibers in a relatively homogeneous muscle.171 Thus, the driving forces for this regulation, though not yet elucidated, probably include not only the discharge pattern of the motor neuron but also the axoplasmic transplantation of trophic substances from the nerve to the muscle. Many other factors influence the twitch and other characteristics of muscle fibers. In one study,20 capsaicin treatment, which selectively eliminated fibers belonging to the group III and IV muscle afferents,43,192 induced muscle fiber transformation from fast contracting fatiguing fibers to slowly
Electromyography contracting nonfatiguing ones. Muscle fiber types also correlate with innervation topography, as shown in the rat serratus anterior muscle.65
4
STRETCH-SENSITIVE RECEPTORS
Anatomy of Muscle Spindles Muscle spindles consist of small specialized muscle fibers encapsulated by connective tissue. The intrafusal fibers measure only 4-10 mm in length and 0.2-0.35 mm in diameter, in contrast to the much larger extrafusal fibers of striated muscle.57,86 The connective tissue capsule surrounding the intrafusal fiber joins the sarcolemma of the extrafusal fibers attached to the origin and insertion of the muscle. The muscle spindles lie parallel to the striated muscle fibers. The nuclear arrangement in their equatorial region distinguishes two types of intrafusal fibers, nuclear bag and nuclear chain (Fig. 12-5). Both dynamic and static bag fibers expand near their midpoint over a short
Figure 12-5. Simplified diagram of the central region (about 1 mm) of the nuclear bag fiber (top) and nuclear chain fiber (bottom) showing two types of motor endings, two types of afferent fibers, and two types of gamma motor neurons. [From Matthews,111 with permission.]
Anatomy and Physiology of the Skeletal Muscle length of about 100 um by a collection of some 50 nuclei. The smaller nuclear chain fibers contain a linear array of nuclei along the center of the fiber. The afferent and efferent nerves that supply muscle spindles each have two different kinds of endings: primary (annulospiral) and secondary (flower-spray) sensory endings; and plate (single, discrete) and trail (multiple, diffuse) motor endings. The primary sensory ending spirals around the center of the bag and chain fibers. In contrast, the secondary ending terminates more peripherally and chiefly on nuclear chain fibers. The largediameter fast-conducting group IA afferent nerve fibers from the primary endings subserve the monosynaptic stretch reflex. In contrast, the secondary ending gives rise to group II afferent nerve fibers that terminate on the interneurons in the
295
spinal cord. Although both kinds of motor endings can innervate either type of intrafusal fibers, the plate endings tend to supply preferentially the nuclear bag; the trail endings, the chain fibers.
Function of Muscle Spindles The dynamic afferent fibers respond to the velocity of the actively stretching spindles. The static afferent fibers detect a sustained change in the length. The primary ending has both dynamic and static function, but the secondary ending mainly mediates static changes (Fig. 12-6). The dynamic and static axons of the fusimotor system influence the dynamic and static muscle spindles respectively.29,111 The trail endings mediate static changes, whereas the plate endings primarily con-
Figure 12-6. Responses of primary and secondary endings to a rapidly applied stretch before (top) and after (bottom) cutting the ventral root. Spontaneous fusimotor discharge maintained a steady intrafusal contraction in the decerebrate cat. The primary endings show a greater sensitivity to stretch112than the secondary endings, but both types respond equally to changes in muscle length. [From Matthews, with permission.]
Electromyography
296
trol dynamic changes.113 The bag fibers receive a sufficiently distinctive motor innervation to subserve preferentially dynamic fusimotor effects, and the chain fibers, static fusimotor effect.13,14 Muscle spindles, using the contraction-dependent discharge pattern, monitor activity of motor units in the vicinity.32,116 Receptor feedback, however, has a negligible effect on the motor neuron pool, compared with the excitatory drive during voluntary contraction.117 During prolonged, sustained contractions, afferent spindle discharges decline, whereas motor unit discharges of the parent muscles increase. This may give the impression of an a-y dissociation, although the decline in spindle discharge may result from a progressive failure in the peripheral mechanisms by which the fusimotor system normally excites the spindle endings.68,69,70,105 Table 12-2 shows a simplified summary of sensory endings found in muscle spindles. The basic structural elements comprise two types of intrafusal fibers, nuclear bag and nuclear chain; two types of sensory receptor endings, primary and secondary, giving rise to group IA and group II afferent fibers; and two types of fusimotor endings, plate and trail, which preferentially subserve dynamic and static function. The dynamic bag fibers receive innervation from the fusimotor fibers with plate endings and modulate dynamic function via the primary sensory endings. The static bag fibers and chain fibers, innervated mainly by fusimotor fibers with trail endings, give rise to both types of sensory afferents to regulate static muscle length. Muscle receptors play a role in proprioception, as evidenced by sensory effects of pulling or vibrating exposed tendons in humans,114 although cutaneous afferents may also provide a dominant input.129
Golgi Tendon Organ The Golgi tendon organ, arranged in series with extrafusal striated muscle fibers, monitors not only active muscle contraction but also passive stretch. The group IB afferent fibers originating herein subserve disynaptic inhibition of the motor neurons that innervate the stretched muscle. According to the traditional view, this inhibitory mechanism provides a safety function to prevent excessive muscle tension when motor neuron firing reaches a certain level. The threshold tension, much less than previously believed, however, excites the tendon organ, especially during active stretch.79 The activation of group IB afferent fibers during mild tension helps continuously monitor and adjust the magnitude of muscle activity for smooth contraction even at a low level of tension. 5
ANATOMY OF THE MOTOR UNIT
As defined by Liddell and Sherrington,97 the motor unit consists of a motor neuron and the few hundred muscle fibers that it supplies. A single discharge of a motor neuron gives rise to synchronous contraction of all muscle fibers innervated by the axon. Hence, even though individual muscle fibers represent the anatomic substrate, the motor unit constitutes the smallest functional element of contraction.158
Innervation Ratio The innervation ratio relates to the average size of a motor unit expressed as a ratio between the total number of extrafusal
Table 12-2 Sensory Endings of Muscle Spindles Location Sensitivity Fusimotor system Form of ending Length of ending Type of afferent fiber Diameter of afferent fiber
Primary Sensory Ending Both bag and chain fibers Both length and velocity Both dynamic and static Half rings in annulospirals About 300 /Am Group IA 12-20 /urn
Secondary Sensory Ending Mainly chain fibers Mainly length Mainly static Spirals and flower sprays About 400 um Group II 6-12 Aum
Anatomy and Physiology of the Skeletal Muscle
Material