Electrodiagnosis in Clinical Neurology, 5th Edition

ELSEVIER CHURClflLL LMNGSTONE The Curtis Center Independence Square West Philadelphia, PA 19106

ELECTRODlAGNOSIS IN CLINICAL NEUROLOGY Churchill Livingstone is an imprint of Elsevier, Inc. Copyright 2005, Elrievier (USA). 1999, 1992, 1986, 1980 All rights reserved.

ISBN 0-443-06647-7

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Medicine is an ever-changing field. Standard safety precautions must be followed but as new research and clinical experience broaden our knowledge, changes in treatment and drug therapy may become necessary or appropriate. Readers are advised to check the most current product information provided by the manufacturer of each drug to be administered to verify the recommended dose, the method and duration of administration, and contraindications. It is the responsibility of the treating physician, relying on experience and knowledge of the patient, to determine dosages and the best treatment for each individual patient. Neither the Publisher nor the author assumes any liability for any injury and/or damage to persons or property arising from this publication. Every effort has been made to ensure that the drug dosage schedules within this text are accurate and conform to standards accepted at time of publication. However, as treatment recommendations vary in the light of continuing research and clinical experience, the reader is advised to verify drug dosage schedules herein with information found on product information sheets. This is especially true in cases of new or infrequently used drugs. Recognizing the importance of preserving what has been written, Elsevier prints its books on acidfree paper whenever possible. Fifth Edition Library of Congress Cataloging-in-Publication Data Electrodiagnosis in clinical neurology/ [edited by] Michael J. Aminoff-5th ed. p. ;cm. Includes bibliographical references and index. ISBN 0-443-06647-7 1. Electrodiagnosis. 2. Nervous system-Diseases--Diagnosis. [DNLM: 1. Electrodiagnosis-methods. 2. Nervous System Diseases--diagnosis. E372005] I. Aminoff, Michael J. (Michaeljeffrey) RC349.E53E43 2005 616.8'047547-<1c22

WL 141

2004050320 The publisher offers special discounts on bulk orders of this book. For information, please contact: Manager of Special Sales Elsevier 170 Independence Square West Philadelphia, PA 19106 Tel: 215-238-7800 Fax: 215-238-8483 For information on all Elsevier publications available, contact our World Wide Web home page at: http://www.elsevierhealth.com Editor: Susan F. Pioli Editorial Assistant: Joan Ryan Production Editor: Mary Stermel MarketingManager: Michael Passanante Design: Melissa Olson

iO 9 8 7 6 5 4 3 2 1 Printed in the United States of America

Contributing Authors VIVIEN C. ABAD, M.D., M.B.A.

RAIN G. BOSWORTH, Ph.D.

Medical Director, Clinical Monitoring Sleep Disorders Center, Camino Medical Group, Cupertino, California Polysomnographic Evaluation of Sleep Disorders

Postdoctoral Fellow, Retina Foundation of the Southwest, Dallas, Texas Visual Evoked Potentials in Infants and Children

DAVID C. ADAMS, M.D.

MARY A. B. BRAZIER, Ph.D.

Associate Professor, Department of Anesthesia, University of Vermont College of Medicine, Burlington, Vermont Intraoperative Monitoring by Evoked Potential Techniques

Professor, Departments of Anatomy and Physiology, David Geffen School of Medicine, University of California, Los Angeles, California The Emergence ofElectrophysiology as an Aid to Neurology

(DECEASED)

JAMES W. ALBERS, M.D., Ph.D. Professor, Department of Neurology, University of Michigan Medical School, Ann Arbor, Michigan ElRctrophysiologic Techniques in the Evaluation of Patients with Suspected Neurotoxic Disorders

MARK B. BROMBERG, M.D.

MICHAEL J. AMINOFF, M.D., D.Sc.

JOHN A. CADWELL, B.S.E.E., M.D.

Professor, Department of Neurology, School of Medicine, University of California, San Francisco, California Electroencephalography: General Principles and Clinical Applications; Clinical Electromyography; Eoaluation of the Autonomic Nervous System; Somatosensory Evoked Potentials; f':lRctrophysiologic Techniques in the Evaluation of Patients with Suspected Neurotoxic Disorders

Director of Engineering, Cadwell Laboratories, Inc., Kennewick, Washington Electrophysiologic Equipment and Electrical Safety

Professor, Department of Neurology, University of Utah, Salt Lake City, Utah Qy,antitativeEMG

GREGORY D. CASCINO, M.D. Professor of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Intraoperative Electroencephalographic Monitoring During Carotid Endarterectomy and Cardiac Surgery

GASTONE G. CELESIA, M.D. LYNNE ADAMS BELL, M.D., Ph.D. Attending Neurologist, Oregon Nerve Center, Portland, Oregon Microneurography as a Clinical Research Tool

Professor of Neurology, Loyola University of Chicago Stritch School of Medicine, Chicago, Illinois Visual Evoked Potentials in Clinical Neurology

GIAN-EMILIO CHATRIAN, M.D. EILEEN E. BIRCH, Ph.D. Senior Research Scientist, Retina Foundation of the Southwest; Adjunct Professor of Ophthalmology, University of Texas Southwestern Medical Center, Dallas, Texas Visual Evoked Potentials in Infants and Children

Professor Emeritus, Department of Laboratory Medicine and Neurological Surgery, University of Washington School of Medicine, Seattle, Washington Electrophysiologic Evaluation of Brain Death: A CriticalAppraisal

CHARLES F. BOLTON, M.D.

JASPER R. DAUBE, M.D.

Professor of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Flectrophysiologic Evaluation of Patients in the Intensive Care Unit

Professor of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Nerve Conduction Studies

vii

viii

CONTRIBUTING AUTHORS

ANDREE DURIEUX-SMITH, Ph.D.

PETER GOURAS, M.D.

Vice-Dean, Professorial Affairs and Professor of Audiology, Faculty of Health Sciences, University of Ottawa, Ottawa, Ontario, Canada Brainstem AuditoryEvoked Potentials in Infants and Children

Professor, Department of Ophthalmology, Columbia University College of Physicians and Surgeons, New York, New York Electroretinography

JOHN S. EBERSOLE, M.D.

Professor, Sleep Medicine Program, Stanford University School of Medicine, Stanford, California Polysomnographic Evaluation of Sleep Disorders

CHRISTIAN GUILLEMINAULT, M.D., Biol.D. Professor, Department of Neurology, University of Chicago, Chicago, Illinois Ambulatory Electroencephalographic Monitoring

JIN S. HAHN, M.D. ANDREW EISEN, M.D., F.R.C.P.C. Professor Emeritus of Medicine (Neurology), University of British Columbia, Vancouver, British Columbia, Canada Somatosensory Evoked Potentials

Associate Professor, Department of Neurology and Neurological Sciences, Stanford University, Stanford, California Neonatal and Pediatric Electroencephalography

MARK HALLETT, M.D. RONALD G. EMERSON, M.D. Professor, Department of Neurology, Columbia University College of Physicians and Surgeons, New York, New York Intraoperative Monitoring by Evoked Potential Techniques

JEROME ENGEL, Jr., M.D., Ph.D. Jonathan Sinay Professor, Departments of Neurology and Neurobiology, David Geffen School of Medicine, University of California, Los Angeles, California Long-Term Monitoring for Epilepsy

MORRIS A. FISHER, M.D. Professor of Neurology, Loyola University of Chicago Stritch School of Medicine, Maywood, Illinois H-Reflex and F-Response Studies

Chief, Human Motor Control Section, National Institute of Neurological Disorders and Stroke, National Institutes of Health, Bethesda, Maryland Electrophysiologic Evaluation of Movement Disorders

JUN KIMURA, M.D. Professor of Neurology, University of Iowa College of Medicine, Iowa City, Iowa The Blink Reflex

KENNETH D. LAXER, M.D. Professor Emeritus, Department of Neurology, School of Medicine, University of California, San Francisco, California Invasive ClinicalNeurophysiology in Epilepsy and Movement Disorders

ALAN D. LEGATT, M.D., Ph.D. JOSEPH M. FURMAN, M.D., Ph.D. Professor, Department of Otolaryngology, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania Yestibular Laboratory Testing

Professor, Department of Neurology, Albert Einstein College of Medicine of Yeshiva University, Bronx, New York Brainstem Auditory Evoked Potentials: Methodology, Interpretation, and ClinicalApplication

ROBIN L. GILMORE, M.D.

WILLIAM J. MARKS, Jr., M.D.

Attending Neurologist, Department of Medicine, Maury Regional Hospital, Neurology Center of Middle Tennessee, Columbia, Tennessee Somatosensory Evoked Potentials in Infants and Children

Associate Professor, Department of Neurology, School of Medicine, University of California, San Francisco, California Invasive ClinicalNeurophysiology in Epilepsy and Movement Disorders

DOUGLAS S. GOODIN, M.D.

NICHOLAS M.F. MURRAY, M.B., Ch.B., F.R.C.P.

Professor, Department of Neurology, School of Medicine, University of California, San Francisco, California Event-Related Potentials

Consultant in Clinical Neurophysiology, The National Hospital for Neurology and Neurosurgery, London, United Kingdom Motor Evoked Potentials

Contributing Authors

ix

sass, M.D.

MARC R. NUWER, M.D., Ph.D.

JASON R.

Professor, Department of Neurology, David Geffen School of Medicine, University of California, Los Angeles, California Topographic Mapping, Frequency Analysis, and Other Digital Techniques in Electroencephalography

Assistant Professor, Department of Neurology, David Geffen School of Medicine, University of California, Los Angeles, California Long-Term Monitoringfor Epilepsy

WILLIAM W. SUTHERLlNG, M.D. JOSE L. OCHOA, M.D., Ph.D., D.Se. Clinical Professor, Department of Neurology and Neurosurgery, Oregon Health and Science University, Portland, Oregon Microneurography as a ClinicalResearch Tool

RICHARD K. OLNEY, M.D. Professor, Department of Neurology, School of Medicine, University of California, San Francisco California Use of Neurophysiologic Techniques in Clinical Trials

Director, Neuromagnetism Laboratory, Huntington Medical Research Institutes; Medical Director, Epilepsy and Brain Mapping Center, Huntington Hospital, Pasadena, California Magnetoencephalography

MARGOT J. TAYLOR, Ph.D. Directeur de Recherche, Centre de Recherche Cerveau et Cognition, Universite Paul Sabatier, Toulouse, France Brainstem Auditory Evoked Potentials in Infants and Children

TERENCE W. PICTON, M.D., Ph.D. Anne and Max Tanenbaum Professor of Cognitive Neuroscience, Departments of Medicine (Neurology) and Psychology, University of Toronto; Scientist, Rotman Research Institute, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada Brainstem Auditory Evoked Potentials in Infants and Children

SIMON PODNAR, M.D., D.Se. Assistant Professor, Department of Neurology, Medical School, University of Ljubljana, Ljubljana, Slovenia Electrophysiologic Evaluation of Sacral Function

DONALD B. SANDERS, M.D. Professor, Department of Medicine, Division of Neurology, Duke University Medical School, Durham, North Carolina Rlectrophysiologic Study of Disorders of Neuromuscular Transmission

BARRY R. THARP, M.D. Professor, Department of Neurology and Pediatrics, School of Medicine, University of California, Davis, California Neonatal and Pediatric Electroencephalography

RICHARD A. VILLARREAL, B.S.E. Principal Design Engineer, Cadwell Laboratories, Inc., Kennewick, Washington Elect'rophysiologic Equipment and Electrical Safety

DAVID B. VODUSEK, MD, D.Se. Professor, Department of Neurology, Medical School, University of Ljubljana, Ljubljana, Slovenia Electrophysiologic Evaluation of Sacral Function

FLORIS L. WUYTS, Ph.D. Professor, Department of Biomedical Physics, University of Antwerp, Antwerp, Belgium Vestibular Laboratory Testing

FRANK W. SHARBROUGH, III, M.D. Emeritus Professor, Department of Neurology, Mayo Clinic College of Medicine, Rochester, Minnesota Intraoperative Electroencephalographic Monitoring During Carotid Endarterectomy and Cardiac Surgery

G. BRYAN YOUNG, M.D. Professor, Department of Clinical Neurological Sciences, University of Western Ontario, London, Ontario, Canada Electrophysiologic Evaluation of Patients in the Intensive Care Unit

Preface to the Fifth Edition More than a quarter-century has passed since the first edition of this book was published by Churchill Livingstone. Over this time, the book has received a generous acceptance, for which I am grateful. However, the advances that have occurred in the field in the last five years have prompted the preparation of this new fifth edition. With successive editions, the book has grown considerably in size, and I have therefore attempted deliberately to prevent any further increase by eliminating older material that is now of only limited interest and, through the generosity of the publishers, arranging for many of the illustrations to be reformatted to improve their clarity and reproduction. Several new chapters have been added, and the other chapters have been updated to reflect advances in the field. As I emphasized in earlier editions, electrophysiologic techniques remain important as a means of evaluating the function of the nervous system in health and disease and for defining the clinical relevance of anatomic abnormalities that are now readily detected by neuroimaging procedures. Furthermore, electrophysiologic techniques help to define and distinguish disorders that clinically resemble one another, such as motor neuron disease and multifocal motor neuropathy. The advances that have occurred in the molecular biology of disease make it evident that similar clinical manifestations may reflect remarkable genetic heterogeneity and, probably, differing pathophysiologic mechanisms. In the years to come, electrophysiologic techniques will have a major role in clarifying such mechanisms and many other issues, such as the natural history of various disorders and their response to novel therapeutic approaches. In the meantime, clinical practice without access to modern electrophysiologic investigative techniques is almost inconceivable. The focus of this book remains on the clinical applications and role of electrophysiologic methods for evaluating the nervous system. It is my hope that it will continue to be useful for clinicians, clinical neurophysiologists, and trainees in these fields, by indicating the utility and limitations of the various techniques that are discussed. A., before, it has not been the intent to discuss techniques of

little or no clinical utility, and such topics have deliberately been omitted, as they were from earlier editions. I am grateful to the various authors who have contributed to this new edition. Most of them also contributed to earlier editions, but they have again taken the time to update their chapters, encompassing the latest developments in the field. Some authors are new to this volume and I am grateful to them for their contribution. The patience and understanding with which all the contributors responded to my requests, and their generosity in making time available to meet the deadlines imposed on them by the publication process, embody the spirit of collegiality that is such an important facet of professional life. I am grateful also to my publishers at Elsevier, and in particular to Ms. Susan Pioli, and her assistant, Ms.Joan Ryan, for their unfailing support and advice in the development of this new edition, and to Ms. Mary Stermel and, particularly, Mr. Naren Gupte for their help and guidance in the production process. The production of this volume was particularly burdensome as the improvement and relabeling of the artwork led to a variety of difficulties, and a series of other problems added to the complexities of the process. Mr. Gupte was meticulous in his attention to detail and devoted many extra hours to ensure the accuracy of the material and that deadlines were met. Finally, I must once more record my indebtedness to my family for their patience, tolerance, and understanding in giving me the time to work on this volume. My wife, Jan, has quietly and graciously supported and encouraged me in all my efforts, and it is to her that I dedicate this volume. Over the years since the first edition was published, our children have been a constant source of pleasure and pride to us both. Alexandra was a toddler when the first edition was published in 1980, and neither of our two sons had been born; she is now at medical school, Jonathan is a law student, and Anthony is an undergraduate at Berkeley. It is my hope that they will derive as much satisfaction from their own careers as I have from mine.

Michael] AminojJ, M.D., D.Sc., /':RC.P.

xi

Preface to the First Edition Fifty years have passed since Hans Berger's first paper on the human electroencephalogram. Over this time, electroencephalography has evolved into an investigative technique of undoubted practical value, and technologic advances have permitted the development of a number of new electrophysiologic approaches to neurologic diagnosis. These developments have led to certain difficulties for clinicians and neurophysiologists alike. On the one hand, the present-day physician is tempted to avail himself of investigative procedures that he does not entirely understand and that provide him with information which he is often unable to interpret. On the other hand, the neurophysiologist is commonly faced with clinical problems that he fails to appreciate or to which there is no ready solution by the means at his disposal. There is therefore a need for a conveniently sized monograph that provides a general introduction to the role of electrodiagnosis in neurology and is directed at the clinical relevance of the investigative procedures that are now within the province of the electrophysiologist. In preparing the present volume, it has therefore been my aim, and that of the other contributors, to provide in simple terms a comprehensive but concise account of the clinical application of various electrophysiologic methods of investigating the function of the central and peripheral nervous systems. Some of these methods, such as electroencephalography and electromyography, are admirably covered in encyclopedic detail in certain textbooks aimed at specialists or trainees in these fields. The chapters covering these topics in the present volume are in no way intended to take the place of such works; rather, they are directed at those who need to know the principles, uses and limitations of the meth-

ods, and who have to relate the information derived from such studies to the clinical context of individual cases. Certain quantitative aspects of these subjects have also been considered, however, because of their potential clinical utility. A number of the other electrophysiologic methods that are covered in this book-such as the various evoked potential techniques-have been developed comparatively recently, and their clinical applications are as yet incompletely defined. In view of the obvious interest shown by increasing numbers of clinicians and neurophysiologists in setting up facilities to undertake such studies for clinical purposes, the technical aspects of some of these subjects have been reviewed in somewhat greater detail, although the emphasis has remained on the practical relevance of the methods. Electrophysiologic techniques that are of more limited clinical utility at the present time, such as recording of the contingent negative variation, have deliberately not been considered. I am greatly indebted to the contributors to this book, all of whom have taken much time and trouble to survey developments in their own particular fields of interest. I am grateful also to those authors, editors, and publishers who have allowed us to reproduce illustrations previously published elsewhere, and whose permission is acknowledged in the text. The advice and understanding that I received from Ms. Carole Baker and Mr. Bill Schmitt of Churchill Livingstone, the publishers, are greatly appreciated. Finally, it is a pleasure to acknowledge the help, encouragement, and support that my wife, Jan, gave me during all stages of the preparation of this book. Michael J. Aminof/. M.D.

xiii

CHAPTER

2

Eledrophysiologic Equipment and Eledrical Safety JOHN A. CADWELL and RICHARD A. VILLARREAL

Filters Saturation Aliasing Quantization Instrument Malfunction Calibration Bad Electrodes DamagedAcoustic Transducers

MAJOR COMPONENTS OF AN ELECTRODIAGNOSTIC INSTRUMENT Electrodes Amplifier Gain and Sensitivity Analog Filters Analog-to-Digital Conversion Digital Circuitry Advantages of Digital Circuitry Digital FIlters Display Stimulators Electrical Stimulators Auditory Stimulators Visual Stimulators Magnetic Stimulators Software FACTORS THAT REDUCE SIGNAL FIDELITY Noise White Noise Impulse Noise Mains Noise In-bandNoise Source Synchronous Noise Signal-to-Noise Ratio (SNR)

The development and refinement of instrumentation has been a great asset in the diagnosis of neurologic diseases. With advances in instrumentation, however, physiologists are in danger of making more technologically advanced misinterpretations than previously, so it is important to have an understanding of the basic functions and limitations of modern instrumentation. This chapter will enhance the practitioner's ability to understand how the instrumentation and its limitations may influence the interpretation of signals recorded during routine electrophysiologic studies.

MAJOR COMPONENTS OF AN ELEORODIACiNOSTIC INSTRUMENT The major components of an electrodiagnostic instrument are shown in Figure 2-1.

SIGNAL-ENHANCING TECHNIQUES Common Mode Rejection Ratio Grounding Patient Grounding Instrument Grounding Isolation Interference Reduction Nonlinear Filtering Averaging Reject Stimulus Rate SAFETY Electrical and Mechanical Safety Electromagnetic Interference and Susceptibility Misuse of Equipment CONCLUDING COMMENTS

Eledrodes The electrode is the interface between the patient and the instrumentation. The proper application and use of electrodes is one of the most fundamental requirements for obtaining good signals, but it is often neglected. Electrode characteristics can affect the response. Electrodes can be classified into at least two different types: (1) surface and (2) needle. Surface electrodes applied with conductive gel form a battery. The voltage of the battery (usually less than 600 mV) depends primarily on the type of electrode material and secondarily on how good a contact is achieved at the microscopic level. The electrode metals are usually not homogeneous and consist of numerous microscopic or sometimes visible grains. Each grain produces a slightly different battery voltage. The electrolyte is assumed to be the gel, but sweat and serum change the 15

16

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

III Major components of an electrodiagnostic instrument. 1, Electrodes pick up electrical signals. 2, Differential input of amplifier removes noise sources common to both active and reference inputs but passes signals that are different at the two inputs. 3, Isolated ground serves as a reference to differential inputs to improve isolation mode rejection ratio (IMRR). 4, Synchronous switch prevents stimulus artifact from propagating through amplifier. 5, Gain stage of amplifier increases signal to a convenient amplitude for further processing. 6, Filters separate excess noise from signal. 7, Sampling (sample and hold) circuit captures and freezes the signal at frequent intervals. 8, Digitizer converts the continuously variable analog signal to a number proportional to its amplitude. 9, Storage provided for the digital values needed for one sweep. 10, Averager (lOa) adds the sweeps together and scales (lOb) the results. 11, Data are displayed (along with graticule, cursors, and numerical readouts of certain values). 12, A speaker provides auditory representation of the signal. 13, Timing circuitry (manual or automatic) generates the start-of-sweep, startof-stimulus, and amplifier switch control signals. 14, Stimulus generator produces electrical shock, auditory, or visual stimulus as appropriate.

FIGURE 2-1

concentrations of sodium and other ions, thus affecting the impedance and voltage. (This is the basis of the galvanic skin response, or GSR.) Most of the battery effect on the amplifier's active input is canceled by an equivalent battery on the reference input. Direct current blocking in the amplifier removes any imbalance. If the electrodes move or the patient sweats, however, the

small changes in potential can easily be larger than the signals of interest. Abrading the skin with a ground quartz suspension (OMNIPREP) or puncturing the skin completely eliminates the GSR and much of the movement artifact. Increasing the homogeneity of the material results in quieter electrodes. Silver electrodes can be corroded

Electrophysiologic Equipment and Electrical Safety

in a controlled manner to form a uniform silver chloride finish, which has noise and impedance characteristics better than those of the bare metal. If a silver electrode is abraded, its performance is reduced, and rechloriding or disposal should be considered. Aluminum electrodes form an aluminum oxide layer, which is very uniform and very quiet but also has a very high resistance. Aluminum electrodes are almost purely capacitive and do not pass low frequencies effectively. Tin, platinum, stainless steel, gold, and carbon are also used as fairly stable materials for electrodes.' Needle electrodes pose other problems. Microscopic burrs on the leading edge of the needle damage muscle as it penetrates, giving false evidence of injury in electromyography. These burrs can be detected by passing the needle through a cotton gauze. Monopolar needles have a thin layer of Teflon (polytetrafluoroethylene) or parylene insulating all but the tip. If the insulation cracks or abrades, or if there is a break in the insulation, the needle will be noisy and should be discarded. Materials and manufacturing problems and damage during transport and handling can also result in increased noise even in the absence of visible defects. Additional information on the care and testing of needles and electrodes is found elsewhere.f Electrode materials should not be mixed. The battery potentials created by two different materials will not cancel, and if direct-current (DC) blocking does not occur at the first stage of the amplifier, a large offset will be present. This offset may saturate the amplifier or decrease the headroom for saturation, so that clipping of the waveform occurs. The offset may contribute to unacceptable shock artifact. It may also change the operating point of the amplifier, which will degrade noise and performance. Fortunately, most modern amplifiers are designed to tolerate electrode offset.

Amplifier The amplifier increases the amplitude of the desired response while it rejects unwanted noise. The first and most crucial stage of the amplifier consists of a differential input. A differential input amplifies the difference in potential presented at its two inputs (active and reference), while rejecting any signal common to both of these inputs. Reference is often used to mean a neutral input, but this only refers to a location on the body that has very little signal compared with the site of the active electrode. Both active and reference inputs are equally effective at generating potentials, and there is no neutral input on a differential amplifier. An amplifier's ability to reject common signals is known as its common mode rejection ratio (CMRR). The higher the CMRR, the better the rejection. Another important parameter of the amplifier input is

17

the input impedance. Input impedance has resistive, capacitive, and inductive components. An input impedance of 10 Mohm or higher is desirable because a low input impedance attenuates the signal slightly and degrades the active-to-reference signal matching necessary for high CMRR. The higher frequency components ofa response are affected more by the input capacitance than by the input resistance. Input inductance is usually negligible. Another amplifier differential input characteristic of concern is input voltage noise and input current noise generated by the input circuitry itself. Input noise is added to the response signal.

Gain and Sensitivity Amplifier gain describes how much the input signal is increased in voltage. The units are volts per volt, and gains of 10 to 10,000 are common. Display sensitivity describes the visible waveform and is expressed as volts per division or volts per centimeter. Smaller numerical values represent increased sensitivity; thus, 1 mV/cm is more sensitive than 10mV/ cm. A graphic display shows a vertical deflection proportional to the voltage, and changing the gain alters the size of the display. A computer displays the digital representation of the analog signals, which maintains the concept of sensitivity at any convenient gain setting for which the amplifier is designed. In most newer systems, the amplifier gain is either fixed or has a few discrete steps, and the display system is changed digitally.

Analog Filters The stages following the differential input amplify and filter the response signal. Low-cut (low-frequency cutoff) and high-cut (high-frequency cutoff) filters are used to narrow the frequency range of the incoming signal (Fig. 2-2), and thus eliminate that portion of the noise outside the bandpass of the response signal. (Signal processing textbooks generally refer to highpass, lowpass, and bandpass filters for mathematical reasons. A bandpass from 10 to 1,000 Hz has a lO-Hz highpass and a 1,ODD-Hz lowpass. The designations lowcut and high-cutsidestep this confusing terminology and are used in this chapter.) The signal will also be affected if it has frequency components outside the bandpass; the filters are therefore adjustable to keep most of the signal and reject most of the noise. A component of the noise will always overlap the signal and cannot be reduced without distorting the signal (Fig. 2-3). Notch filters provide precise band-reject capability and are tuned for 50- or 50-Hz operation. The "Q" (quality) of a filter is a mathematical measure of its resonance. High-Q filters respond to a narrow but precise

18

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

III

1

"'0

0.7

j.....

;.;,;.r

i

"!

:

..

:

f0:: ~lr!llrifs 0.1

FIGURE 2-2

III

10 100 1k 10k 0.1

l:

: :

:

L :

:

1

~

j

.v.... :

~

......] l J

10 100 1k 10k

~

:

.

t .

~

1k

Low-cut filter High-cutfilter Bandpass filter Relationship of amplitude to frequency for single-pole filters at 1 Hz and 1 kHz.

range of frequencies. They are used for 50- and 60-Hz notches because signals that are only a few Hertz above or below the notch frequency are passed transparently. Low-Q filters respond to a wide range of frequencies and are used for bandpass applications. Comb filters function like multiple-notch filters, which are tuned to successive harmonics of the mains frequency (Fig. 2-4). Combs remove these harmonics, which make the "buzzing" sound commonly heard from 50- or 60-Hz in terference. The cutoff frequency of a filter is the frequency at which the output power is half of the input power (-6 dB) or the output voltage is 0.707 times the input voltage (-3 dB). Except for brickwall filters, the output

.....................l:

.

t :

,No.ts.e

;:

.

.......l~.N . . ·..······..··I·············.g.+~·50.HZ

power increases or decreases smoothly with frequency, as shown in Figure 2-2. Changes in frequency are measured in octaves (doubling or halving of the frequency) or decades (increases or decreases tenfold). The simplest filters have a single pole and will roll off, or attenuate, the signal by 6 dB for every doubling of the frequency (6 dB per octave or 10 dB per decade), but they will also cause attenuation and phase shift well away from the -3 dB point, as shown in Table 2-1. (The word pole is an engineering term used in describing transfer functions; one pole represents a single resistancecapacitance [RC] filter.) To separate noise and signal more effectively, steeper filters are used. Two-pole and four-pole filters are common. Each additional pole adds 6 dB per octave (10 dB per decade) attenuation. These filters are usually more than cascaded one-pole filters, having feedback and feedforward paths. Varying the amount of feedback and feedforward varies the overshoot, phase, roll off characteristics, and amplitude ringing. Special cases of filters have specific designations, but all are part of a continuum consisting ofjust a few distinct topologies (Fig. 2-5). Thus, the Butterworth has the flattest passband at the expense of poor rolloff; the Bessel has constant phase delay for all frequencies; the Chebyshev has maximal transition steepness at the expense of passband ripple; the elliptic filter has infinite rolloff with rebound in the stopband; the notch is a special elliptic filter that rebounds back to o dB; and a brickwall filter has infinite slope without rebound at the expense of maximum "ringing." (For

.

TABLE 2-1 • Effed of a I,ooo-Hz Single-Pole High-Cut Filter ~

~

••••••••• _.•• ••• - •••••••••••••••••• ••••••••••••••••••••• ;

iii

"~""""""T"

•• _ •••

,

A signal (5) with peak energy at 150 Hz and noise (N) with peak energy at 1 kHz are added and then highcut filtered at 1,000, 200, and 50 Hz. Optimal signal-to-noise ratio is achieved at 200 Hz. FIGURE 2-3

II

Input (Hz) 100 500 1,000 2,000 10,000 100,000

Amplitude

Percent Decrease

0.99 0.89 0.707 0.44 0.10 0.01

1 11 29 56 90 99

dB Decrease 0.04 1 3

7 20 40

Electrophysiologic Equipment and Electrical Safety

o

~Ir"""

__

High

"a" notch

Comb filter

o -20

dB -20 Low

"a" notch

Frequency (Hz)

..

-40

-40

60 FIGURE 1-4

19

II

240

300

Amplitude responses of notch filters and comb filter.

analog circuits, anything over 100 dB per decade is a brickwall.) Further discussion of analog filters is provided elsewhere."

Analog-to-Digital Conversion Analog-to-digital conversion requires a circuit to "freeze" the signal for a few microseconds, the sample and hold (SIR), and a circuit to convert the amplitude of the "frozen" signal to a digital value, the analog-todigital converter (ADC). Sometimes these circuits are implemented in one device, and so they will be referred to collectively as the ADC. The ADC is specified by its conversion rate and its resolution (number of bits).

Digital Circuitry The digital section consists of three major parts: processor, memory, and averager. The processor is the "brain" of the instrument: it coordinates all data flow and interface functions. Processors are classified by the number of bits processed in parallel and by the processing speed. Memory is used for processor instruction storage and for digitized signal storage. The amount of memory is expressed in bytes. The averager adds and scales synchronized signal responses to improve the signal-tonoise ratio and may be implemented by the processor. ADVANTAGES OF DIGITAL C,RCU,TRY

Operations on digital signals are precise. Adding two analog signals gives a result with a percentage error, and the errors are cumulative. Two digital signals added together will always give precisely the same result. The chip or the components that add two analog signals will shift or drift with time, temperature, humidity, power supply voltage, and other factors. They may require calibration or compensation, and sometimes they cannot be made to work at all. The chip that adds two digital signals always gives exactly the same answer

and is insensitive to its environment. Digital systems eliminate analog drift with time and temperature, and eliminate some of the need to recalibrate. Analog component values vary, but digital coefficients are absolute. A capacitor with ±0.1 percent tolerance in a filter is an exotic part, but a 16-bit digital system has 0.001 percent accuracy, is easy and inexpensive to build, and never changes with time or temperature. Every digital unit is also exactly like every other unit, so fabrication and characterization are simplified. Digital systems can perform functions that are not practical and sometimes not possible with analog systems. Almost all analysis is easier to perform digitally. Processor performance has increased tenfold since the last edition of this book, whereas the price and power requirements have fallen. There are now few problems or solutions that are not easier, less expensive, and more reliable to solve or implement digitally. DIGITAL FILTERS

In most electrodiagnostic instruments, analog filters are used sparingly and have been replaced by digital filters. Digital filters can duplicate analog filters, but they can also create classes of filters not readily implemented from analog components. If implemented correctly, multipole digital filters can be superior to the analog equivalent. Multipole analog filters require components that are subject to temperature and aging variations; these act to "detune" the filter. Two major types of digital filters are the infinite impulse response (IIR) and the finite impulse response (FIR) filters. In an IIR filter (Fig. 2-6), a portion of the output data is fed back to the input. If the output does not feed back, the filter is nonrecursive and is classified as an FIR filter. The feedback term in an IIR represents the contribution of previous data points to the output. Because only a few terms are needed (two terms for each two poles), efficient filters are realizable. IIRs act much like analog filters. The infinite means that like an RC network, the output approaches its final value asymptotically.

20

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

1.0

t

0.7

1.0

Brickwall ....................... ....................................................

t Q)

Q)

"'0

"'0

~ 0. E

~ 0. E

-c

-c

0

0 2

0 Frequency (Hz)

2

0.7 ....................... ...................................................

Q)

3 ~

1.0

Chebyshev

t

0.7

Q)

"'0

"'0

:E

='

~

0. E

0. E

-c

-c

0 2

0 Frequency (Hz)

0

3

2

0 Frequency (Hz)

~

1.0

t

1 Frequency (Hz)

~

1.0

t

0

3

~

1.0

t

0.7

Q)

3

Butterworth

0.7

Q)

"'0

"'0

~

~

='

='

0. E

0. E

«

-c

0

0

2

3

0 0

2

Frequency (Hz) ~ Frequency (Hz) FIGURE 2·5 • The relationship of amplitude to frequency for six different filter types.

FIRs (Fig. 2-7) compute each output from weighted portions of a limited number of past, present, and future input data points. Each point used in the computation is called a filter tap, and each tap requires a multiply-and-accumulate operation. If the FIR is symmetric (i.e., uses the same number of taps and matching coefficients into the future as into the past), then there is zero phase shift. Evoked potentials can be smoothed without changing their latency. (Three-point smoothing algorithms are FIRs.) Obviously, the future is never known in the real world, even for the upcoming few milliseconds. An FIR filter avoids this problem by delaying the output for half the number of taps and moving the time reference by the same amount. In a

3 ~

real-time system, all frequencies are delayed the same amount and only the relative phases have zero shift. This trick has a small price. Steep-skirted filters are accompanied by ringing whenever an edge or impulse occurs; this is called Gibbs' phenomenon. Because zero has been shifted out in time, half of the ringing occurs before the impulse and can sometimes be seen before the stimulus. The additional peaks created by Gibbs' phenomenon are artifact, as is their occurrence before the stimulus. Such peaks have been interpreted and presented as new responses previously hidden by the noise or as anticipatory potentials." Digital filter characteristics, including cutoff frequency and the number of poles, are determined by

Electrophysiologic Equipment and Electrical Safety

21

III A two-pole (biquad) infinite impulse response (IIR) filter uses five coefficients that determine the filter frequency and type and requires only two intermediate terms.

FIGURE 1-6

their coefficients. Adjustable frequencies do not require more resistors, capacitors, and analog switches; they only require changing a set of numbers in the processor. High-eut, low-eut, and notch filters for multiple channels are typically implemented with a single processor.

Fast Fourier transforms (FFT) are a class of algorithms that turn time data into frequency-phase data and vice versa. They are especially convenient for looking at a signal's frequency characteristics and for implementing brickwall or other arbitrary filters. Fast refers to algorithms implementing the Fourier transform by eliminating redundant calculations to speed up computation. A brickwall filter is implemented by performing the FFT, zeroing out all unwanted frequency terms, and then performing an inverse FFT. Brickwall filters show maximum Gibbs' phenomenon but are maximally effective at reducing noise outside the passband.

Display

FIGURE 1-7 • A five-tap finite impulse response (FIR) filter.

Computation of Qo is delayed until F-2 is loaded. If coefficient C2 equals Cv'2' and C1 equals C_l' phase shift is eliminated.

A response can be presented visually and audibly. Historically, electromyographic instruments (EMG) used analog oscilloscope displays, where the response signal vertically deflects an electron beam as it sweeps horizontally across the face of a phosphor-eoated tube. All modern systems use digital displays, which create pictures by illuminating individual pixels (dots) on a screen to create a picture. Factors that affect use-ability include resolution and screen size, and resolution is no longer a constraint. Digital displays have inherent persistence supplied by memory buffers instead of by long persistence phosphors. Other advantages that oscilloscopes once had can be simulated with clever programming, and the ability to combine waveforms and graphical and textual information on one display has eliminated oscilloscope technology.

22

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Auditory presentation of the response is useful not only in helping to classify a response but also in detecting and identifying noise. Types of interference from power lines, fluorescent lights, cathode ray tube (CRT) displays, biologic artifact such as EMG activity, electrode artifact, and sterilizing ovens are easily identified by their characteristic sounds.

..... 2k load

10mA conatam-

current stimulator

Stimulators ELECTRICAL STIMULATORS

Stimulation occurs when the voltage across the nerve membrane is decreased enough to initiate depolarization. The nerve has sodium ion pumps that normally maintain a resting potential, and the stimulating current must overwhelm the pumping capability. To do this requires a fairly constant charge per stimulus. The charge is the area under the curve of an amplitude-duration plot, and accounts for intensity, pulse width, and wave shape. The most common wave shape is a square wave because it is easy to generate. Other wave shapes will not change the charge requirements, although wild claims have been made to the contrary. If the charge is introduced very slowly with the use of a low-amplitude, long-duration pulse, the nerve is able to compensate partially for the stimulus and requires more total charge for depolarization. The strength-duration curve shows this relationship, which varies for different tissues. Several caveats and exceptions are well known. If the nerve is initially hyperpolarized and then depolarized, the total charge required can be reduced. A biphasic stimulus can also be used to achieve zero net charge transfer, which may eliminate electrolysis and possible tissue injury in direct nerve stimulation. In polysynaptic systems (the brain), both inhibition and potentiation are observed with paired pulses, depending on the interstimulus intervals. Constant, as in constant current, means that the output remains at the specified, adjustable level. Constantcurrent stimulators have high-output impedance and allow the output voltage to change to maintain the desired current. Constant-voltage stimulators have low output impedance and allow the output current to vary to maintain the desired voltage. Other stimulators have finite output impedance and allow both voltage and current to change as the load impedance changes (Fig. 2-8). Constant current has the theoretical advantage that the product of duration and current determines the stimulator's effectiveness. Some authors claim that constant current is less painful, but this is both subjective and sensitive to technique and methodology. Constantcurrent stimulators will deliver the same current as electrode gel dries out, an advantage for those who do not or cannot check electrodes.

50 volts constant-

voltage

stimulator 50 volt stimulator with 1k output resiatance

,

··,

··, 0.1 ~ ·, .c:J.

·,· 5k !oed · · ,

JVY'-

,,

5k

. ':it. . . '·fr. '" ...... . . jt+lt+············ .n,:::::::;::::.:: p: · ff1lLt:=$ '

,

·······················:········· ..·..·..···..l.. ~~n~~~~~~t~~~··· ,

.~~~~~~~t~ ,

~~ ~~

::.....: ....::::::r:: ,

,

'

,

,'

,,

FIGURE 2·8 • Current and voltage outputs of three stimulator types and three loads. Body impedance is similar to load 3, with both resistive and capacitive components. One vertical division represents 10 rnA (black line) and 20 volts (lighter line, offset slightly above and to the right).

Constant-voltage stimulators will deliver a stimulus to the nerve even if a small amount of gel or body fluid is shorting the leads. It will charge the body capacitance quickly and then deliver full current, and so it is a more powerful stimulator. Electrical stimulators produce large voltages that can introduce artifact into the waveform. Some of the artifact occurs if the amplifier saturates and has a long recovery time. In 1980 the authors began turning the amplifiers off during the stimulus to reduce amplifier recovery time. Clamping the stimulator output immediately after the stimulus or putting out a biphasic wave to remove the charge stored by the body, or both, is also useful. If the stimulus is completely isolated, very little current should be common to both stimulator and amplifier. Complete isolation is difficult to obtain; instead, the upgoing voltage on one stimulator electrode and the downgoing voltage on the other are designed to have equal capacitive coupling back to ground to cancel out amplifierstimulator leakage currents. AUDITORY STIMULATORS

The brainstem auditory evoked potential is generated by acoustic signals between 4 and 8 kHz. A 100-llsec square-wave click has most of its energy in this band. The click is amplitude-eontrolled between 0 and 130 dB, a range of over 3,000,000:1. Careful attention to noise and inadvertent feedthrough is needed for this dynamic range. Headphones capable of faithfully reproducing the electric pulse generate an auditory click. Both magnetic transducers and piezoelectric

Electrophysiologic Equipment and Electrical Safety

transducers are used. The magnetic transducers mounted on a headband generate a small electrically coupled artifact at the beginning of the sweep. Piezoelectric transducers are usually placed about a foot away from the ear on a hollow acoustic coupling tube, using spongy inserts to hold the tube in the ear canal and to suppress ambient noise. Because the click takes about 1 msec to traverse the tube, all components of the response are separated from the artifact by 1 rnsec. VISUAL STIMULATORS

Visual stimulators for eliciting evoked potentials depend on the rapidly changing contrast along the edges of checks to produce a response. Both raster scan (TV monitor) and light-emitting diode (LED) checkerboards are used. The LED checkerboard generators can reverse the on-off pattern almost instantly. The scanned displays take from 0 to 16.7 msec to change patterns, related to the time it takes for the beam to sweep the display. This stimulus lag shows up as a smearing and an 8-msec latency shift of the responses compared with the LED responses. Check size, contrast, intensity, and the subject's visual acuity also affect the response, as discussed in Chapter 21. MAt;NETIC STIMULATORS

Magnetic stimulators generate a 1- to 2-Tesla magnetic field in 50 to 100 usee, which induces a voltage in peripheral nerves or cerebral cortex sufficient to achieve depolarization. The principle is the same as that for a transformer, in which a changing magnetic field induces a voltage around it. The stimulator is the primary of the transformer, and the cortex is the singleturn secondary. The body is almost perfectly transparent to the magnetic field, and so currents can be induced below the skin with minimal or no pain. Magnetic stimulators allow measurement of motor evoked potentials, which are discussed in Chapter 27.

Software The algorithms used to control the instrument are known as the software. The resources allocated to write the software exceed the effort to design the modern electrodiagnostic instrument. To partition the design effort, most systems have multiple processors, each of which controls a portion of the system. The keyboard, the various stimulators, and the amplifier may each have its own dedicated processor and associated software or firmware. The software may reside in various formats in an instrument. Software that is programmed into nonvolatile memory is called firmware. Software may reside on flexible media (floppy disk) or rigid media

23

(hard disk) and is loaded into system memory by a small loader routine. The design and reliability of the software has a large influence on the utility of an instrument. As EMG instruments, and especially the reports and data they generate, become less stand-alone and more integrated into electronic records and connected computer systems, software design becomes more demanding. The user interface becomes more critical; it must appear simple and intuitive when, in fact, increased effort is needed to make it more intuitive. The underlying software components have to be cleanly partitioned so that maintenance and testing of the hardware, acquisition, and user interface can be verified. Awareness and adherence to software standards allows other systems to access, utilize, and display the results, and may allow collaboration in ways that proprietary solutions preclude.

FACTORS THAT REDUCE SIGNAL FIDELITY Noise Physiologic signals are mixed with noise. Low-level signals of all sorts are plagued with noise, and it is noise that limits the resolution and precision of the signal measurement. Noise usually refers to white noise, but several other types of noise, with multiple sources and varied solutions, are worth considering. Advances in technology have also introduced new noise sources. WHITE NOISE

Random noise or white noise sounds like a harsh "shhhhhh." It is generated by processes that are statistical in nature, and has uniform energy in all frequency bands (energy per band constant). A 10k-ohm resistor at room temperature generates about 0.3 llV of white noise across its leads just lying on a bench. This thermal noise, generated by agitated electrons, places an absolute lower limit on amplifier quietness unless the system is cooled to absolute zero or unless the input impedance is reduced to zero. Passing a current through the resistor creates additional pink noise (energy per band = l/frequency), which has a more musical "shhhhh" sound. Because cooling (of patients) is not practical, skin preparation and conductive gel are required to decrease impedance, and high-impedance amplifiers are necessary for decreasing current flow. The preamplifier has additional noise determined by engineering choices, and generally cannot be improved easily. The electroencephalogram (EEG) is nearly random noise (in evoked potential studies), as is weak background EMG activity; these are usually the dominant sources of white noise. =;

24

ElEaRODIAGNOSIS INCLINICAL NEUROLOGY

IMPULSE NOISE

SIGNAL-fo-NoISE RATIO (SNR)

Impulse noise sounds like a "pop," "crack," or "click" and includes transistor "popcorn" noise, static discharge, EMG artifact, artifact from metal dental fillings touching intermittently, and electrode movement. Impulse noiseas used here is present for a short time in only one epoch, unlike random noise, which is present uniformly throughout each epoch.

The relative size of the signal to the noise determines how well the signal can be visualized or even detected. The evoked potential signal present in anyone epoch is 1 to 100 times smaller than the background noise. The electrodiagnostic equipment must obtain an SNR that is better than 3:1 for reproducible testing. Figure 2-9 is a graphic presentation of SNR values. SNR is expressed as an integer ratio or in dB:

MAINS NOISE

Mains 50- or 60-Hz interference (assumed 60 Hz in this discussion) produces a continuous audible buzz if harmonics are present, but it is inaudible or barely audible otherwise. It is induced by magnetic induction and by capacitive coupling. Harmonics are present when ironcore transformers, dimmers, and fluorescent lights are nearby. The energy is all at 60 Hz, 120 Hz, 180 Hz, and so forth, and is usually biggest in the odd harmonics (e.g., 180 Hz, 300 Hz); the energy in high-order harmonics drops rapidly. IN-BAND NOISE SOURCE

Cellular telephones, high-efficiency fluorescent lights, switching power supplies for laptops, and blood pressure cuffs with digital readouts are examples of the profusion of noise sources. Regulatory mandates to control such sources are growing in response to the awareness of their adverse effect on sensitive measurements. Most of these emit electrical noise in the 1,000- to 100,000-Hz range, which either steps on the signal of interest or is poorly rejected by the amplifier. Awareness and avoidance are essential.

signal SNR=--.noise or Signal ) SNR db = 20 log (- - .noise Decibel notation is convenient because both large ratios (e.g., 100,000:1 = 100 dB) and small ratios (e.g., 0.25:1 = -12 dB) are easily represented, and because gain calculations use addition instead of multiplication (e.g., the product of a gain of 5 and a gain of 15 is a gain of 75, which is the same as 14 dB + 23.5 dB = 37.5 dB). The slope of analog filters is constant when plotted on a log-log scale, and the slope is then numerically the rolloff in dB per decade. Decibels are a relative scale given in logarithmic values. The reference (denominator of the ratio) must be specified for the

Signal

SYNCHRONOUS NOISE

Synchronous noise is time-locked with stimulation and averaging. It can be generated by numerous sources: 1. The instrument processor generating the stimulus executes the same instruction sequence and may radiate a characteristic burst of energy. 2. The timer used to generate the stimulus rate may radiate electrical noise. 3. Electrical stimulator recovery may have an abrupt turn-off after many milliseconds. 4. The power supply may be modulated by the slightly increased power demands during stimulation. 5. Headphones with a low-frequency resonance may ring down for several milliseconds. 6. The patient may blink or track the target used to elicit visual evoked potentials (VEPs), thereby introducing electroretinogram (ERG) signals.

10:1 SNR

FIGURE ]-9 • Signal, noise, and signal plus noise with signal-

to-noise ratios of 1:3, 1:1,3:1, and 10:1.

Electrophysiologic Equipment and Electrical Safety

decibel notation to be useful. For filters and amplifiers or attenuators, the reference is the input voltage. For SNR, the reference is the noise voltage. The audiometric reference uses an absolute pressure value of 0.0002 dyne zcm'' to define 0 dB SPL (sound pressure level). Most noise in electrodiagnostic instruments originates at the amplifier input and is measured at the output. Noise measurements are specified "referred to the input" by taking the noise on the display and dividing by the gain to get an equivalent noise at the amplifier jacks. Amplifier noise is measured with the inputs shorted (there will be additional noise with a patient connected) and is usually specified in llV RMS. The average or root-mean-square (RMS) noise relates to the heating capability of the signal. It has convenient mathematical properties and can be measured with an RMS voltmeter, but the heating capability is not intuitive on inspection. White noise has a gaussian distribution, and so a few large spikes will appear above a surface of more average spikes. Peak-to-peak noise voltage is seen directly on the display. To estimate peak-to-peak noise, exclude the largest single excursion in each direction to eliminate the right-sided tail of the gaussian curve (Fig. 2-10). To convert from peak-to-peak to RMS, multiply by 0.14 for white noise and by 0.35 for sine waves. Amplifier noise should be measured periodically or when problems are suspected. Static discharge (even without visible or sensible sparks) can damage sensitive input transistors, which have no input protection, decreasing the input impedance and adding enormous amounts of noise. The measurement is performed by

25

running an average with inputs shorted. Set gains to 10 llV per division (settings vary by system, but use the same settings each time and save a copy of the results for future reference); filters 10 to 3k; scale x 100; average count to 1,000 trials; and measure the peak-to-peak voltage. Channels with more than twice as much noise as the others are suspect.

Filters Fidelity means that the observed signal is the same as the originating signal. Fidelity requires that the bandwidth of the system be adequately wide, that amplitude is scaled linearly, and that phase relationships of the component sine waves are preserved. Filters can reduce fidelity, but in most cases fidelity has less value than good SNR. Moreover, the originating signal cannot be determined precisely unless excellent SNRs are obtainable. It is important to understand the results of intended or incidental distortion. Signals that are smoother have fewer high-frequency components and distort less (e.g., VEPs). Square waves and signals with fast rising and falling edges have the largest highfrequency components and distort more easily (e.g., potentials recorded by the needle EMG). Increasing the low-cut filters stabilizes the baseline, but it also removes low-frequency signals and adds a phase to the trailing edge of the signal (Fig. 2-11). Twopole and higher order low-cuts can add an additional small trailing phase. Digital low-cut filters can add a phase to the leading edge of the signal. Decreasing the high-cut filter frequency reduces white noise and results in smoother waves. Single-pole

..................... , .... .......... ......;.................. ···f················· ....

[

",J" ...

.~

... . ..... ...

..........

pop Noise

...........

........

.......,.-. ...........•.............. ..... . ....

~ ·M· ...

I\~"

......

··1'.. ·· ...... ....

~"" ......

..

.... .- .. -_ .....

FIGURE 2-10 .. Peak-to-peak noise measurement uses the typical largest peaks found by excluding the single largest upgoing and downgoing peak.

FIGURE 2-11 .. A frequency limited impulse is low-cut filtered at six different frequencies, showing undershoot and later an overshoot. The gain of the bottom waveform has been increased ten-fold.

26

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

high-cuts round the edges of fast signals but do not add overshoot. Most filters of two or more poles cause overshoot and undershoot, creating what appears to be an additional phase (Fig. 2-12). Digital high-cut filters can do worse, adding lightly damped ripple both preceding and following the edges (Fig. 2-13). Digital filters are being used more extensively and will replace analog filters in the future. Digital high-cuts can be made to act exactly like analog filters, but the steeper skirts available digitally allow better noise control, especially for evoked potential studies. Filters are a special source of synchronous noise. Steeply skirted filters and notch filters are resonant circuits that feed forward a signal equal in amplitude but opposite in phase to the noise presumed present. If the skirts or notch are in the signal passband, the resonance will create ringing artifact (Fig. 2-14). Large shock artifact, motor nerve conduction responses, and the expected responses of evoked potentials all can cause this ringing. Digital filters, as explained before, can cause ringing that precedes the response. VEPs and the compound muscle action potential have peak energy centered at about 60 Hz and are grossly distorted by 50or 60-Hz notch filters. All filters tend to smear sharp pulses and can obscure the signal.

Saturation Saturation is inherent in all systems when the output signal nears the power supply voltage. The incremental gain is then near zero, and further input causes no change in the output. If the amplifier gain is turned up, large noise-spikes will be clipped sooner, reducing the noise contribution. However, sometimes allowing the amplifier or filters to block can generate long, exponentially decaying artifact after the overvoltage as the amplifier comes back to life. Large power-supply currents and voltage changes can also couple into other circuits. By designing a voltage clamp, instead of waiting for saturation to occur, large-amplitude noise can be removed with immediate recovery after the spike. Clamping can be implemented with analog or digital circuits.

Aliasing Analog signals are continuous in time and amplitude. They are digitized by taking samples at intervals and recording a numerical value for each sample. Sampling is usually done at regular intervals. Events occurring

••••••••••••••••••••••• 1 •••••••••••••

•••

~

.

••• o., ••••••••••• : •••••••• 0

••••••••••••

°:.

.

0 • • • • • • • • _0 • • • • • • • • • • • • :

0

0

,......-'7'"--....,

,

.

••••••••••••

Unfiltered

10k

............................ ···..·..·....·..:··..···..·····......···:..···..·..3'1<"····..···

10~

1k •••••••••••••1.••••••••••••••••••••••• \ ••••••••••••••••••••••

500

1k:

200

.................199

330

. .

FIGURE 1·11 • A square-wave input is high-cut filtered at six different frequencies, showing increasing rounding of the leading and trailing edges.

;

;

:

,

.

FIGURE 1·13 III A 3-msec square wave is filtered at four different frequencies by a brickwall zero phase-shift digital filter. The amplitude of the overshoot is constant for all four settings.

27

Electrophysiologic Equipment and Electrical Safety

: Notch filter off

: Notch filter on

Q)

"0

:E a. E

-c

0

50

100

Frequency

•

FIGURE 2·14 • An 8-msecfrequency limited impulse has peak energy between 25 and 60 Hz, as shown by its frequency analysis.

When the 60-Hznotch is turned on, the impulse is grosslydistorted.

between samples are lost. If the sampling rate is too slow for a rapidly changing event that covers multiple data points, the event is grossly distorted and reappears at a different frequency: this is called aliasing. Aliasing appears on television when wagon wheels that are moving rapidly forward appear to be moving slowly backward. Limiting the frequency range of the signal limits the allowable complexity of an event. Sampling the frequency-limited event at twice the maximum signal frequency (known as the Nyquist sampling rate) just barely captures the event without aliasing. If a blip does occur between samples, it must have high-frequency components and the assumption about the maximum signal frequency is untrue (Fig. 2-15). White noise and other noise often have high frequencies. After sampling, noise frequency components above the halfNyquist rate will reappear below the half-Nyquist rate and will be added back in as increased noise at a lower frequency. If averaging is used, the new noise will average out as described earlier, although the noise can still

be excessive. In EEG, EMG, and nerve conduction studies that do not use averaging, the high-frequency noise appears as increased signal noise. Mains interference has continuous sinusoidal components that alias as new sinusoidal components at a lower frequency, possibly in the delta, theta, alpha, or beta bands of the EEG signal, causing confusion or misinterpretation. EMG artifact is several times worse in an EEG record without adequate antialiasing. To avoid aliasing, either the higher frequencies must be filtered out with analog filters or the sampling rate must be increased. Practical systems use sampling rates well above the Nyquist rate, usually 3 to 10 times higher than needed (i.e., 6 to 20 times the highest signal frequency). The higher sampling rate eases analog antialiasing filters requirements, and it makes waveform reconstruction possible with the use of a straightline segment to connect the samples.

Quantization

FIGURE 2·15 • Effect of sampling on various signals. In

Section 1, the pointer shows that the output is missing one serration present on the input EMG signal. Section 2 is adequately sampled (9 times the signal frequency). Section 3 is marginally sampled (4x). Section 4 is inadequately sampled (2x). Section 5 is aliased, and the output frequency is less than the input frequency.

Each analog sample is converted to a number by the ADC. The precision of that number determines how accurately it represents the original analog value. Whatever precision is chosen, the leftover is rounded up or down and the difference is called the quantizing error. The bit length of a word can be picked by choosing an acceptable quantizing error. One decimal place of precision equals 3.3 digital bits. If 1 percent accuracy is needed, the product of 2 decimal places and 3.3 (i.e., 6.6 or 7 bits) is used. Accuracy of 0.1 percent requires 10 bits, and so on. Nonlinearity, nonmonotonicity, and other ADC errors further decrease the digitizing accuracy. As in the decimal system, where the rightmost digit is the least significant digit, the rightmost bit is the least significant bit (lsb) in the binary system. The term lshas a percentage of full-scale error is used to describe the noise characteristics of the system. The quantization

28

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

error is by definition one-half lsb for an ideal ADC. For example, an 8-bit ADC has 0.4 percent resolution (i.e., 1 part in 256), and the largest quantizing error would be ±0.2 percent. Quantization error is a problem if the data are too "grainy" for the application because such error is a noise source. Quantization is also a problem when a small difference between two large digitized values is computed. This may occur internally as an intermediate step in a computation for digital filters, FFTs, and other algorithms. If quantization error is minimal and the data are sampled at an adequate rate, the analog and digital representations of the signal are equally valid. Further information is available elsewhere."

Instrument Malfundion

can cause anything from loss of signal to increased noise. Intermittent contact caused by a wire broken inside the insulation is difficult to detect. Using disposable electrodes or replacing electrodes on a regular basis will help to prevent electrode problems. DAMAGED ACOUSTIC TRANSDUCERS

It is difficult to tell whether the audible click of an auditory stimulator has the needed 4 to 8 kHz components. A good telephone connection, for example, has no frequencies above 2,700 Hz, and the loss of higher frequencies is hardly detectable. Headphones that are used at high stimulus intensities will degrade rapidly, producing subsequent brainstern auditory evoked potentials (BAEPs) of poor quality.

CALIBRATION

System calibration IS Important. Most electrophysiologic systems have built-in signal generators that help to check calibration. More advanced systems have software-controllable signal generators and amplifier settings to allow automatic verification of all gain, filter, and montage setups. Overall system performance becomes more difficult to ascertain as the amount of data manipulation increases. For example, the Cadwell CIMON (cardiac intraoperative monitor) displays EEG relationships as bands of green, red, and yellow to indicate adequate, marginal, or inadequate cerebral perfusion. The amplitude and phase relationships of the input signals are critical. System verification must be performed on a regular basis. Such verification requires that prerecorded input data are processed correctly and that all gains and filter settings are correct. BAD ELECTRODES

Broken and damaged electrodes are one of the most common causes of poor responses. Electrodes are subject to repeated twisting, bending, and pulling, as well as to chemical attack from solvents and conductive gels. Broken or intermittent contact of the electrode wire

SIGNAL-ENHANCING TECHNIQUES Common Mode Rejedion Ratio Signals of interest may be as small as 0.1 !lV, and the ambient noise at 60 Hz may be a full volt (-120 dB SNR) or more. Averaging will add only 30 dB to the SNR, so a lot of help is needed to obtain a good response. The amplifier is responsible for most of the noise control. The 6o-Hz noise is common to both active and reference inputs and is called a common mode voltage (CMV). By subtracting reference from active, the CMV will disappear. Any "signal" common to both active and reference inputs will also disappear (Fig. 2-16). The efficacy of the differential amplifier at rejecting common mode signals is the common mode rejection ratio (CMRR) and is typically between 10,000:1 and 100,000:1 (80 to 100 dB). The high CMRR of an amplifier is produced by subtracting two extremely similar signals. Any signal mismatch between active and reference inputs causes a difference-error that decreases CMRR. One such mismatch results from unequal electrode impedance. Because the amplifier has finite input impedance, any

Differential component Common mode signal

Remaining common mode component

FIGURE 2-16 III A differential amplifier with a CMRR oflO: 1 willreduce the output 60-Hz signal by the same ratio. Any difference between active and reference inputs is amplified.

Electrophysiologic Equipment and Electrical Safety

mismatch in patient electrode impedance will produce different voltages at each input. In 1970, a typical l-Mohm impedance amplifier required electrode matching within Ik ohm to achieve 60 dB of CMRR. Modern (FET and MOSFET) amplifiers have typical inputs of 5 picofarad capacitance in parallel with 109 to 10 12 ohms resistance, which gives about 5 x 108 ohms impedance at 60 Hz. The new CMRR computation, assuming 10k ohms electrode impedance mismatch, is 100 dB. Thus, impedance mismatch is a smaller factor in determining actual CMRR for the modern instrument. CMRR drops rapidly with increasing frequency and will not effectively remove 15-kHz TV noise, AM radio stations, or similar signals. (Radio frequency filters and appropriate high-cut filters are used instead.) CMRR is usually specified with inputs shorted and at a frequency of 50 or 60 Hz.

Grounding PAnENT GROUNDING

By tying the body to earth ground, the 50- or 60-Hz mains interference can be reduced by 10 to 100 times. Earth ground is not a reference for each of the differential amplifiers; rather, it is used to drain off the excess common mode voltage. The voltage induced on the

Isolation

Power supply

29

body is capacitively coupled. It is induced by coupling between the exposed area of the body; the dielectric of free space; and the area of wire, lamp, or whatever equipment has large noise voltages. Ground forms a low (1 to lOOkohm) impedance shunt for these signals. Tying the body to earth is a potential safety hazard to the patient and is not recommended as a means of reducing common mode interference. Modern equipment connects the patient to an isolated ground, instead of to earth ground, to eliminate the potential safety hazard when earth ground is utilized. INSTRUMENT GROUNDING

The chassis of the electrophysiologic instrument is connected to earth ground by a ground wire in the power cord. If the electrical outlet ground lead is not connected, if the building has a poor connection to earth ground, or if the ground line is broken along its path from instrument to earth, a high-impedance connection results. Instrument noise will be excessive and noise generated by other devices will be carried back to the instrument and couple into the pickup electrodes. A dedicated ground wire from the power receptacle to earth is an optimal solution. Figure 2-17 shows the building ground, equipment ground, and potential leakage pathways of concern to the neurophysiologist.

Isolation

.-""'::::"O::::---"!'-"""

Line

(p.

............. 110

Ground.., 4 .... Primary current • Leakage current ..........., Leakage current pathways

- - EI~i~

Water pipes

................

.............. volts

Neutral

pa::::///.. . .

~~~: 7."': 7."': 7."'~"'~"~"':':":'": 7.' ': 7.' ': 7.' ': -: ' ' ::-::,..f-I Grounding rod

FIGURE 1-17 • Grounding techniques and leakage current pathways. Leakage current is generated by the mains voltage (1)

applied across the inherent capacitance of the line cord and power supply (4). If the ground lead is intact, the leakage current returns to ground by that route (3). If the ground is faulty, current can flow through the operator if one hand (5) contacts the instrument and another touches a sink, radiator, or other grounded equipment (6). The amplifier is secondarily isolated but still has a certain unavoidable capacitance across the isolation. A smaller leakage current (7) can flow through this capacitance and through the patient to earth ground (8) or through other attached equipment (9), especially if the other equipment is nonisolated. If line and neutral at the outlet or power cord are reversed (2), the leakage usually increases but must still be within agency specifications. Transcardiac current and possible fibrillation dictate the maximum leakage currents allowed.

30

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

Isolation A safer method of reducing the effect of common mode voltage (and so reducing the need for high CMRR and balanced electrode impedances) is to float the amplifier from ground and connect it solely to the patient. The amplifier circuitry will electrically ride the common mode voltage on the body. This is comparable to a fishing bobber in the ocean, which rises and falls with the waves. The measurement used to describe the rejection of common mode noise is the isolation mode rejection ratio (IMRR) and is typically above 100 dB. Isolation requires coupling the power, the signals, and control lines across a very low capacitance barrier. Transformers, optocouplers, and capacitive couplers using frequency modulation, pulse width modulation, or linear modulation techniques are typically used. Isolation also increases patient safety because fault currents cannot flow through the amplifier and the patient. An IMRR of 120 dB means that a I-volt signal should be reduced to IllV. Measured values would probably be 10 or more times higher. The ground electrode impedance forms a voltage divider with the amplifier's capacitive coupling to earth ground, so the amplifier does not exactly float with the common mode voltage (the fishing bobber sinks a little). Also, small currents that flow through the body generate voltage drops (the product of current and resistance). The resulting voltage is a differential signal, not a common mode signal, and is amplified instead of rejected (Fig. 2-18). Induced

Isolation

110 v

"v

···1 FIGURE 1-18 III Capacitive coupling causes small induced cur-

rents to flow through the body. The current creates a voltage drop between the electrodes, which is a differential signal. Increased IMRRand CMRR have no effect on this residual 60Hz interference.

currents from magnetic fields also produce differential signals. As a result, a specified IMRR above 100 dB produces little or no improvement in rejecting 50-Hz noise. CMRR and IMRR have the same units and the same effect, but describe different processes and should not be used interchangeably.

Interference Redudion Interference from external noise sources couples to the patient and electrode wires by magnetic induction and capacitive coupling. The best remedy for the interference problem is to remove the noise source or to move the instrument to a different location. Some of the more common noise waveforms are shown in Figure 2-19. If the source of noise cannot be removed, other measures can be taken to minimize the interference. Twisting active and reference electrode wires together minimizes the loop area. The larger the loop area, the more interference is "caught." Shielding the electrode wires with a continuous metal foil will also reduce external pickup. The shield must be tied to the amplifier's ground to be effective. Placing amplifiers on the electrodes optimally reduces loop area and can sometimes improve noise performance.

Nonlinear Filtering Nonlinear filters can improve SNR by selectively attenuating data that are likely to be noise. Slew-rate limiting eliminates large fast transients. The slew-rate is how fast the output signal is allowed to change, and is expressed in volts per second. For small signals, slew-rate limiting has no effect. For large signals (which are mostly noise), it limits the output excursion, decreasing the amount of signal passed. Sometimes a small signal is riding on a large signal. Slew-rate limiting flattens the leading and trailing edges of the noise, creating triangles on the screen. A small signal on the side of the triangle is squashed and lost, and so the averaged value will be smaller. High slew-rates are not especially useful, and low slew-rates distort the signal excessively. For evoked potential studies, 20 to 50 llV per msec is reasonable.

Averaging Averaging is the most useful technique for improving the SNR. Averaging is performed by adding successive traces and dividing the result by the number (n) of samples. If the noise is unrelated to the signal, the SNR will be improved in proportion to the square root of the number of trials (vn) independent of the noise spectrum (Table 2-2). The buried signal is often assumed to have the same size and shape on each sweep. This

Eleclrophysiologic Equipment and Electrical Safety

.................;

:

. . ......... -.. -..-..............................•............. --...

................:

:

;

,

;

;

.

.

'

..

.

;

~

;

;

;

31

·········~···CRT·vertic~i·reiiace······

• • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • __ • • • • • _ . _ • • • • • • • • • • • • • • • • • • • • • • 0# • • • • • • • • • • • • • • • • • • • • • • _ • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • • •

CRT horlzontal retrace

;

;

·Aci'bC'·m~to·r·

.

III Electrical artifact arising from five types of sources is shown. Sinusoidal 60-Hz activity is generated by incandescent bulbs, heaters, and ungrounded equipment. Fluorescent lights and some electrical motors generate characteristic artifact ~t exact 16.6-m~ec i.ntervals. The CRT vertical retrace may be close to 50 or 60 Hz or may range from 40 to 75 Hz. The CRT horizontal retrace IS ahased, appearing as shown, but may have other envelopes.

FIGURE 2-19

assumption is good for monosynaptic or oligosynaptic responses and becomes less valid for polysynaptic cortical potentials, whose average is the average of numerous different responses. Each trace is added to the preceding traces, and the result is divided by the total number. The signal value increases with the number of trials because the response in each trace is assumed to be in-phase and, when normalized by "n," remains constant. The noise component is variable in phase and amplitude; each point is independent of the preceding points. Large positive values tend to cancel with large negative values, medium with medium, small with small. The total increases by the square root of the number of trials but, when normalized by "n," tends to zero at the rate of l/~.

( ~n

=

_1 )

~

REJECT

Impulse noise can be considered a single event that is unrelated to the stimulus. Because the event is not present except in one sweep, it averages out faster than random noise and tends to zero at the rate of lin. However, most impulses are huge compared with the signal, and so even lin is not good enough. A better solution is to delete sweeps that have large artifacts. A voltage-sensitive trigger detects these large amplitude impulses and inhibits the averager for that sweep. The stimulus artifact can also trigger the reject circuitry, and so rejects are implemented to ignore the stimulus artifact. By selecting the reject level and delay appropriately, the troublesome impulses are effectively controlled. Electrocautery can generate huge artifacts that affect multiple sweeps, and another version of artifact rejection is to stop the averager for a short period after any rejected signal. STIMULUS RATE

TABlE 2-2 • Effects of Averaging with 10-mV RMS Noise Present

in the Signal Number of Averages 1

10 1,000 10,000

Noise (ll-V) 10 3

0.3 0.1

Time to Average at 3 Hz NA 3 sec 5 min 1 hr

The choice of stimulus rate affects the averaging process. Mains interference is neither random nor unrelated to a fixed stimulus rate. The 60-Hz artifact can be made random (relative to the stimulus) by using a random stimulus rate. In evoked potential studies, a stimulus rate that is an exact submultiple of 60 Hz (e.g., 2.00 or 3.00 Hz) becomes synchronous with mains noise, and the 60-Hz noise is averaged in instead of out. For nerve conduction

32

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

studies, only a single unaveraged sweep is desired, but several sequential stimuli are applied until the desired response is obtained. Having an exact submultiple then makes the 60-Hz activity in the baseline appear to stand still, and small responses are easier to identify. EMG that is made synchronous with the 60Hz noise is easier to look at for the same reason. Common EMG sweep speeds for EMG are 100 msec and 200 msec, and both are exactly synchronous at 50 or 60 Hz. When averaging is performed, selecting irrational ratios such as 3.17:60 will spread the noise uniformly across the trace. The resulting SNR improvement is generally better than 1/~. Zero noise is conceivable, but small variations in stimulus rate and in the actual 60-Hz frequency, and variations in noise from sweep to sweep, mean that zero noise is not practical to obtain. If the stimulus can be made to fire randomly between desired limits (e.g., 2.5 to 3.5), the mains noise will be uncorrelated and will reduce at about 1/~. Random stimulus rates are very effective at reducing noise in an averaged response when the noise is caused by multiple uncorrelated sources. Synchronous noise cannot be averaged out. Placing the electrodes and ground in saline or water and averaging may reveal noise sources creating nonphysiologic responses. Solutions include eliminating the noise source (if possible), making it asynchronous by adding jitter to the source, or subtracting the noise from the signal. Of these, the last solution works well only if the average of the noise is repeatable. The extracted noise reference will inevitably have additional white noise, which will then add back into the signal when the synchronous noise is subtracted. Subtracting such noise templates may create new and more interesting problems.

SAFETY A patient connected to an electrodiagnostic instrument is potentially at risk for excessive electric current if special precautions are not taken. Normally the skin's high impedance (lOOk ohms) offers some protection from electric shock; however, electrodes applied to abraded skin with conductive gel can lower the impedance to below 1,000 ohms. There are two major sources of shock hazard: (l) leakage current and (2) dielectric breakdown. Leakage current may potentially flow from the instrument itself or from another source connected to the patient. In the operating room, a patient connected to a variety of other monitoring and life-support instruments could conduct leakage current if any of the instruments fails. Dielectric breakdown (an electric arc through

or around the insulation) usually requires thousands of volts. Two sources are large transients on the mains (e.g., from a lightning strike) and intentional cardiac defibrillation. The primary safety concerns are unintentional cardiac arrest or fibrillation and burns. Regulatory agencies have created standards for safety. Medical equipment has its own safety standards, and they give substantial assurance that inadvertent injury to the patient or operator, especially because of electrical leakage, is unlikely even with improper grounding or multi-instrument configurations. Most countries have a national agency or agencies that specify or create the safety standards as dictated by national law. States and municipalities may also have special requirements, as may hospital accreditation bodies. Even individual hospitals may have their own safety requirements. Fortunately, most agencies are harmonizing their requirements to conform to the International Electrotechnical Commission (IEC) standards. IEC601 is the standard applied to medical equipment. The IEC601 standard is a substantial collection of requirements for medical equipment including electrical and mechanical safety, electromagnetic radiation, electromagnetic susceptibility, software design, and equipment accuracy. The standard also stipulates requirements for EMG/EP and EEG equipment.

Eledrical and Mechanical Safety The maximum allowable leakage current varies with the type of medical instrument. The maximum leakage current for patient-connected parts is chosen to give a safety margin by a factor of 10, assuming that a conductive line from ground to the sinoatrial node has been inserted. The IEC601 standard has three different values based on three different classifications: (1) type B (nonisolated); (2) type BF (isolated); and (3) type CF (ultra-isolated for direct cardiac contact). Loss of earth ground to the instrument is considered a possible fault condition, and the safety standards limit the leakage current that could possibly flow from the instrument through the operator or patient to earth ground. The safety standards also deal with a variety of other possible hazards, including flammability of the instrument, maximal temperatures, mechanical stability and strength, and warning labels. The protection of the patient from inadvertent stimuli during power-up and power-down and from an instrument malfunction is also considered. The standards require that exposed metal be connected to earth ground via the power cord to prevent inadvertent coupling from electrical sources inside the system (e.g., a broken wire) to the patient or operator.

Electrophysiologic Equipment and Electrical Safety

33

Eledromagnetic Interference and Susceptibility The proliferation of electrical devices leads to an increase in the possibility that the electromagnetic radiation produced by these devices may interfere with electrodiagnostic studies. The potential for interference depends on factors such as the distance from the interference source; carrier frequency, modulation frequency, and strength of the electromagnetic radiation; electrode wire orientation; electrode impedance; and the design of the amplifier input circuitry. The majority of electrical devices are regulated to produce low electromagnetic radiation in certain frequency bands. However, some devices (e.g., electrosurgical units and portable communication devices) are allowed to radiate significant radio frequency energy. Because electrodiagnostic instruments are used increasingly as monitors in the operating room, the ability of the instrument to continue to operate correctly after being subjected to severe conditions becomes important. The amplifier or stimulator inputs should survive static discharges of 8,000 to 12,000 volts. A static discharge to any part of the instrument should not cause the instrument to malfunction. Electromagnetic radiation must not couple to the instrument's internal circuitry and cause disruption. A large transient on the input power must also not cause any disruption.

Misuse of Equipment Despite all the safety standards, instrumentation safety and the interpretation of results can still be compromised by operator misuse. During stimulation, electrical stimulators must generate large currents to depolarize the nerves. Normally, the current is confined to a small volume close to the electrodes. However, attaching stimulator leads across the chest or between contralateral sites across the chest (e.g., to both median nerves) creates a pathway that includes the heart. This is often done inadvertently, as shown in Figure 2-20. Electrical stimulators can cause burns if the power dissipation, as determined by the pulse width, current, and repetition rate, is too great or is applied for too long. Other hazards include possible chemical injury from metal ion transport similar to iontophoresis, which is used to carry beneficial drugs across the skin. Overabraded skin may develop scar tissue. Connecting two electrical stimulators in parallel by jumpering them together may not work because the outputs of stimulators are commonly clamped to decrease shock artifact.

FIGURE 2·20 • Use of one stimulator for both median nerves allows transcardiac current to flow.

The clamped stimulator will then short the active stimulator, and no current will flow in the patient. The screen display will appear normal if a constant-current stimulator is used. Auditory stimulators can also cause hearing damage if high stimulus intensities are employed for extended durations.

CONCLUDING COMMENTS Accurate and reliable electrophysiologic recordings require an understanding of the principles and pitfalls of the technology used. A knowledge of the characteristics of the equipment allows the user to eliminate or avoid problems and to obtain technically superior results. Equipment will change and technology advances, but principles remain the same.

REFERENCES 1. Geddes LA, Baker LE: Principles of Applied Biomedical Instrumentation. 3rd Ed. John Wiley & Sons, New York, 1989 2. Reiner S, Rogoff JB: Instrumentation. p. 498. In Johnson EW (ed): Practical Electromyography. 2nd Ed. Williams & Wilkins, Baltimore, 1988 3. Horowitz P, Hill W: The Art of Electronics. 2nd Ed. Cambridge University Press, New York, 1989 4. Sheingold DH (ed): Analog-Digital Conversion Handbook. Prentice-Hall, Englewood Cliffs, NJ, 1986

CHAPTER

Electroencephalography: General Principles and Clinical Applications

3

MICHAEL J. AMINOFF

PRACTICAL CONSIDERATIONS Recording Arrangements Activation Procedures Artifacts Bioelectric Artifacts Instrumental Artifacts EEG Interpretation ACTIVITY RECORDED IN THE EEG Alpha Activity Alpha Rhythm Other Rhythms of Alpha Frequency Beta Activity Theta Activity Delta Activity Polymorphic Delta Activity Intermittent Rhythmic Delta Activity Breach Rhythm Lambda Waves Triphasic Waves Spike Discharges Paroxysmal Activity Spike-Wave Activity . .. Intermittent RhythmIc Delta ActIvIty Burst-Suppression Pattern Periodic Complexes Periodic Lateralized Epileptiform Discharges Low-Voltage Records Other EEG Patterns 14· and 6·Hz Positive Spikes Small Sharp Spikes or Benign Epileptiform Transients of Sleep 6-HzSpike-Wave Activity Wicket Spikes Rhythmic Temporal Theta Bursts of Drowsiness (Psychomotor Variant) Rhythmic Theta Discharges Hereditary Factors and the EEG EEG Changes with Aging EEG RESPONSES TO SIMPLE ACTIVATING PROCEDURES Hyperventilation Photic Stimulation Natural Sleep Sleep Deprivation

EEG FINDINGS IN PATIENTS WITH NEUROLOGIC DISORDERS Epilepsy .. .. Primary (idiopathic) Genetolized Eptlepsy Secondary (Symptomatic) Generalized Epilepsy Partial EpIlepsy Anticonvulsant Drugs, Seizure Control, and the EEG Status Epilepticus Long-Term EEG Monitoring for Epilepsy Value of the EEG in the Management of Patients with Epilepsy Syncope Infections Herpes Simplex Encephalitis Subacute Sclerosing Panencephalitis Creutzfeldt-Jakob Disease Abscess Vascular Lesions Subarachnoid Hemorrhage Subdural Hematoma . Intracranial Aneurysms and Arteriovenous Malformations Sturge-Weber Syndrome Headache Tumors Tuberous Sclerosis Pseudotumor Cerebri Dementia Metabolic Disorders Hepatic Encephalopathy Renal Insufficiency Anoxic Encephalopathy Effects of Drugs and Alcohol on the EEG Multiple Sclerosis Trauma Coma De-efferented State (Locked-In Syndrome) Miscellaneous Disorders Spinocerebellar Degeneration Parkinson's Disease Paroxysmal Choreoathetosis Huntington's Disease Sydenham's Chorea Hepatolenticular Degeneration Progressive Supranuclear Palsy Transient GlobalAmnesia Psychiatric Disorders

37

38

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

The electroencephalogram (EEG) represents the electrical activity of the brain as recorded from electrodes placed on the scalp. Many clinical neurologists and neurosurgeons do not fully appreciate the potential value or the limitations of electroencephalography, and this lack of information is reflected in the manner in which they use this technique in clinical practice. On the one hand, patients are often referred indiscriminately for study with little, if any, information provided about their clinical background. On the other hand, patients for whom electroencephalography might be expected to provide clinically useful information are not investigated by this means at all. Electroencephalography is most useful in the investigation and management of patients with epilepsy. The presence of "epileptiform" activity (p. 49) in the EEG of a patient with suspected epilepsy does not establish the diagnosis beyond doubt because similar activity may occasionally be found in patients who have never had a seizure. It is, however, one more factor that must be taken into account when patients are evaluated clinically. In patients with behavioral or other disturbances that could be epileptic in nature, but about which there is some uncertainty, the presence of such activity increases considerably the likelihood that the attacks are indeed epileptic. In patients with an established seizure disorder, the EEG findings may help to classify the disorder, identify a focal or lateralized epileptogenic source, indicate the most appropriate medication that should be prescribed, provide a guide to prognosis, and follow the course of the disorder. Electroencephalography also provides a noninvasive means of localizing structural abnormalities, such as brain tumors. Localization is generally by an indirect means, however, depending on the production of abnormalities by viable brain in the area of the lesion. Moreover, it is sometimes disappointingly inaccurate, and the findings themselves provide no reliable indication of the type of underlying pathology. In this context, it is hardly surprising that advances in neuroimaging techniques for the detection of structural abnormalities in the brain-in particular, by the development of computed tomography (CT) scanning and magnetic resonance imaging (MRI)-have led to a reduction in this use of electroencephalography as a screening procedure. Nevertheless, the EEG reflects the function of the brain and is therefore a complement to, rather than an inconsequential alternative to, these newer procedures. The third major use of electroencephalography is in the investigation of patients with other neurologic disorders. Certain of these disorders produce characteristic EEG abnormalities that, although nonspecific, help to suggest, establish, or support the diagnosis. These abnormalities are exemplified well by the repetitive slow-wave complexes sometimes seen in herpes simplex

encephalitis, which should suggest this diagnosis if the complexes are found in patients with an acute cerebral illness. The electrical findings are best regarded as one more physical sign, however, and as such should be evaluated in conjunction with the other clinical and laboratory data. A further use of electroencephalography-one that may increase in importance with the development of quantitative techniques for assessing the data that are obtained-is the screening or monitoring of patients with metabolic disorders, because it provides an objective measure of the improvement or deterioration that may precede any change in the clinical state of the patient. Electroencephalography is also an important means of evaluating patients with a change in mental status or an altered level of consciousness. Finally, it is used for studying natural sleep and its disorders, and as help in the determination of brain death. Further comment on these aspects is deferred to Chapters 32 and 34, and the clinical utility of the EEG in the investigation of infants and children is considered separately in Chapter 4.

PRAOICAL CONSIDERATIONS Recording Arrangements The EEG is recorded from metal electrodes placed on the scalp. The electrodes are coated with a conductive paste, then applied to the scalp and held in place by adhesives, suction, or pressure from caps or headbands. Alternatively, needle electrodes can be inserted directly into the scalp. The placement of recording electrodes is generally based on the international 10-20 system, in which the placement of electrodes is determined by measurements from four standard positions on the head: the nasion, inion, and right and left preauricular points (Fig. 3-1). When this system is used, most electrodes are about 5 to 7 ern from the adjacent electrodes, in an adult. If closer spacing is required (e.g., to define the site of an epileptogenic focus), electrodes can be placed in intermediate positions. The potential differences between electrodes are amplified and then recorded on continuously moving paper by a number of pen-writers or displayed on an oscilloscope screen. Relatively inexpensive digital systems are now widely available commercially, and these have many advantages over the older analog systems. 1 They permi t reconstruction of the EEG with any desired derivation or format and also permit data manipulation for added analysis. They also facilitate access to any desired portion of the record and obviate storage problems. Further discussion of such systems is provided in Chapter 8. Electrodes are connected with amplifiers in predetermined patterns, or montages, to permit the electrical

Electroencephalography: General Principles and Clinical Applications

Inion

FIGURE J-I !IIi The International 10-20 system of electrode placement. A, earlobe; C, central; F, frontal; Fp, frontal polar; P. parietal; Pg, nasopharyngeal; T, temporal; 0, occipital. Right-sided placements are indicated by even numbers, leftsided placements by odd numbers, and midline placements by z.

activity of various areas to be recorded in sequence. The recording arrangements can be varied, or the EEG can be reconstructed after digital recording, so that the potential difference is measured either between pairs of scalp electrodes (bipolar derivation) or between individual electrodes and a common reference point. In the latter arrangement, the reference point can be either a relatively inactive site on the scalp or elsewhere (e.g., the vertex or the linked-ears) or a point connected to all the electrodes in use so that it reflects the average of the potentials at these electrodes. Each technique has its own advantages and drawbacks, but for routine purposes at least two of these methods for deriving the EEG should be used. Montages are generally selected so that recordings are made from rows of equidistant electrodes running from the front to the back of the head or transversely across it. In most North American laboratories, traces from the left side of the head are displayed above those from the right, and those from anterior regions are displayed above those from more posterior areas. The more detailed technical and practical aspects of EEG recording are beyond the scope of this chapter; the interested reader is referred to Cooper and colleagues.f There is, however, one general point that must be made clear. As already indicated, the potential difference between pairs of electrodes-or between an electrode and its reference point-is amplified before being displayed on moving paper or an oscilloscope

39

screen. The input leads of the individual amplifiers are designated as black (input terminal 1, or Gl) and white (input terminal 2, or G2). They are arranged so that when the electrode connected with the black lead is relatively more negative than that connected with the white one, an upward deflection of the trace occurs. The relationship between the two inputs, then, determines the direction in which the trace is deflected, not the absolute value of any discharge that is recorded. With bipolar derivations, the conventional American recording arrangement is for the most anterior electrode of each pair to be connected with the black lead when recording from the front to the back of the head, and for the left-hand electrode of any pair to be connected with the black lead when recording across the head. With the reference derivations, each of the active scalp electrodes is connected to the black lead of an amplifier, and all the white leads are connected with the common reference point. Interpretation of specific events in the EEG derived with a common reference point may be confounded if the reference electrode itself lies within the active field, whereas a voltage peak involving two adjacent electrodes to the same extent may not be detected in the bipolar derivation because no potential difference will be recorded between these electrodes. Both the bipolar derivation and the use of a common reference permit abnormalities to be localized, but the former method is less satisfactory for localizing widespread changes or for demonstrating areas in which activity is suppressed. With linked bipolar derivations, the source of localized EEG abnormalities is determined by locating the common electrode at which the deflections of the traces show a reversal of phase or polarity when recordings are made simultaneously from at least two rows of electrodes at right angles to each other (i.e., from rows of electrodes in the anteroposterior and transverse axes of the head). Changes in amplitude may be misleading in bipolar recordings. For example, a greater amplitude in one channel does not necessarily reflect the origin of a particular discharge but signifies only that a larger potential difference exists between the two electrodes connected to that amplifier; a lower amplitude may reflect inactivity or equal activity at the two electrode sites. By contrast, with common reference derivations, localization of abnormalities is accomplished primarily on the basis of amplitude. Various specialelectrodes have been devised for recording the activity of inaccessible regions of the brain, because electrodes placed on the scalp may not detect such activity. A nasopharyngeal electrode, consisting of a flexible, insulated rod or wire with a small silver electrode at its tip, can be inserted into the nostril and advanced until the terminal electrode is in contact with the mucosa of the posterior nasopharynx. This permits recording of electrical activity from the anteromedial

40

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

surface of the temporal lobe. Insertion of these electrodes is generally undertaken by technologists without the application of any local anesthetic, and not all patients tolerate the procedure. Furthermore, recordings from them may be contaminated by pulse, muscle, and respiratory movement artifact, and only rarely do they reveal abnormalities that are not seen with the usual scalp electrodes. Sphenoidal electrodes can record activity from the anteroinferior portion of the temporal lobe. These electrodes are less likely to lead to artifacts than are nasopharyngeal electrodes, but their application is more difficult. The electrode consists of a sterile needle or fine wire that is insulated except at its tip, and a physician inserts it percutaneously under local anesthesia so that it lies acljacent to the sphenoid bone, a little in front of the foramen ovale. Other accessory electrodes that are commonly used in the electrophysiologic evaluation of patients with complex partial seizures are surface sphenoidal electrodes, minisphenoidal electrodes, and anterior temporal electrodes. The relative merits of these different electrodes have been considered elsewhere.v' but there is no consensus of opinion in this regard. The EEG examination usually is undertaken in a quiet, relaxed environment, with the patient seated or lying comfortably with the eyes closed. Recordings initially are made for up to about 5 minutes from each of several different standard montages, for a total of about 30 minutes. Depending on the findings, the examination can then be continued by recording with less conventional montages. During the recording of activity from each montage, the patient should be asked to lie with the eyes open for about 20 seconds before closing them again, so that the responsiveness of the background activity can be assessed. When this routine part of the examination has been completed, recording continues while activation procedures are undertaken in an attempt to provoke abnormalities.

Adivation Procedures Hvperoentilation for 3 or 4 minutes is a generally welltolerated method of provoking or accentuating EEG abnormalities, but it should not be performed in patients who have recently had a stroke, transient ischemic attack, or subarachnoid hemorrhage; or in those with moyamoya disease, significant cardiac or respiratory disease, or sickle cell disease or trait. The patient is asked to take deep breaths at a normal respiratory rate until instructed to stop. The resultant fall in arterial PC0 2 leads to cerebral vasoconstriction and thus to mild cerebral anoxia. This anoxic state is generally held to be responsible for bringing out the EEG abnormalities, although the validity of this mechanism

has been questioned," Certain quantitative techniques have been suggested in an endeavor to establish a more uniform procedure, but these methods are not in routine use. The EEG is recorded during the period of hyperventilation and for the following 2 minutes, with the use of a montage that encompasses the area where it is suspected, based on clinical or other grounds, that abnormalities may be found; or, in the absence of any localizing clues, an area that covers as much of the scalp as possible. The eyes are generally kept closed during the procedure, apart from a brief period at its conclusion to evaluate the responsiveness of any induced activity. Hyperventilation usually causes more prominent EEG changes in children than in adults. There is considerable variation, however, in the response of individual subjects, and this variation makes it difficult to define the limits of normality. It may also produce or enhance various bioelectric artifacts. Recording during sleep or after a 24-hour period of sleep deprivation may also provoke EEG abnormalities that might otherwise be missed, and this technique is similarly harmless. This approach has been used most widely in the investigation of patients with suspected epilepsy. Abnormalities may also be elicited with an electronic stroboscope to cause rhythmic photic stimulation while the EEG is recorded with the use of a bipolar recording arrangement that covers particularly the occipital and parietal regions of the scalp. The flash stimulus is best monitored on one channel either with a photocell or directly from the stimulator. Abnormalities are more likely to be elicited when the patient is awake for the procedure. At any given flash rate, the EEG is recorded with the patient's eyes open for about 5 seconds and then while the eyes are closed for a further 5 seconds. Flash rates of up to 30 Hz are generally used, but an even wider range of frequencies is employed in some laboratories. The manner by which abnormalities are produced is unknown. Various auditory stimuli may also precipitate EEG abnormalities in patients with epilepsy, but they are not used routinely in the EEG laboratory except in the evaluation of comatose patients (p. 77). Other stimuli that may induce paroxysmal EEG abnormalities include tactile stimuli and reading. A number of different pharmacologic activating procedures have been described. These procedures are not in general use, however, and may carry some risk to the patient; thus, they are not discussed in this chapter.

Artifads A variety of artifacts may arise from the electrodes, recording equipment, and recording environment. Examples are the so-called electrode "pop" resulting

Electroencephalography: General Principles and Clinical Applications

from a sudden change in impedance (seen as an abrupt vertical deflection of the traces derived from a particular electrode, superimposed on the EEG tracing, as shown in Fig. 3-2); distorted waveforms resulting from inappropriate sensitivity of the display; excessive "noise" from the amplifiers; and environmental artifacts generated by currents from external devices, electrostatic potentials (as from persons moving around the patient), and intravenous infusions (generating sharp transients coinciding with drops of the infusion, possibly caused by electrostatic charges). Bioelectric artifacts are noncerebral potentials that arise from the patient and include ocular, cardiac, swallowing, glossokinetic, muscle, and movement artifacts. With reference derivations, artifacts may be introduced because of the location of the reference electrode or because the reference electrode is within the cerebral field under study. No single site is ideal as a reference point. The ear or mastoid placements are commonly used but may be contaminated by muscle, electrocardiogram (ECG), or temporal spike discharges. The vertex, which is also widely used, is very active during sleep and is sometimes contaminated by vertical eye movements during wakefulness. In general, artifacts are recognized because of their temporal relationship to extracerebral monitors such as the ECG, because of their unusual appearance, or

A

because the electrical field of the event is hard to interpret in a biologically plausible manner (Fig. 3-3). BIOELECTRIC ARTIFACTS

Eye-movement artifacts are generated by the corneoretinal potential, which is on the order of 100 mV and has been likened to a dipole with the positive pole at the cornea and the negative pole at the retina. Eye movement leads to a positive potential recorded by the electrodes closest to the cornea. Thus, with upward movement of the eyes (such as occurs during a blink), a positive potential is recorded at the frontopolar (supraorbital) electrodes relative to more posteriorly placed electrodes, and thus a downward deflection occurs at these electrodes. Such eye movement is easily distinguished from frontal slow EEG activity by recording from an infraorbital electrode referenced to the mastoid process: the former leads to activity that is out of phase between the supra- and infraorbital electrodes, whereas frontal EEG activity is in phase. Similarly, horizontal eye movements lead to a positivity at the frontotemporal electrode on the side to which the eyes are moved and a corresponding negativity on the opposite side (Fig. 3-4). Nystagmus or eyelid flutter, for example, produces a rhythmic discharge in the frontal electrodes (see Fig. 3-4). Oblique eye movements and asymmetric

c

B

41

0

Fp1-F7

Fp1-A1

~

F7-T3

C3-A1

~

T3-TS

T3-A1 ~

TS-01

01-A1 ~

Fp2-F8

Fp2-A2 ~

F8-T4~~

C4-A2

~

T4-T6~~

T4-A2

----

T6-02~

02-A2

~

EKG

~ I I [souv

I

I 1 sec

[75~V

1 sec

1 sec

-~

1 sec FIGURE :S-1 • A, Electrode artifact arising at the T5 electrode; B, EMG artifact in the left temporal region; C, Chewing artifact;

D, ECG artifact.

42

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

A

phone

phone

phone

Fp1-F3 F3-C3 C3-P3 P3-01 Fp2-F4 F4-C4 C4-P4 P4-02 L...-J

1 sec

B Fp1-F7 ~~~"""'~"""""~~~WI\IIo.M'~"""""""""'''''''''\oIWI_~''''''''''''''''''/IfII' F7- T3 1<'J'I\~"""MI"IAJ""'I'\MfVI\J"vAJ''''''''MJ~~''''''''''{'IfJlrV'lr'V''IW\J''vA,~WIM.

T3-T5 --......,..,...,...............,...--"""""',.,...........,.,..""""'-_........~-......""""'--..,.,.. T5-01

-.-.'

Fp2-F4 ~~'f/rII#o"'/V'~~'Y¥'I4fft/~~W\,~~""'V"M~~~"I¥t~ F4-C4

""'NY.' -

.~

+

..,,,,,

"

C4-P4 WI..Nv-M..W,.M.M""""'"""",,MJ"'-MIJv(I,M'o-Jv..II-JWI-NAJ'vN'"NvI'v{--J'v/l./I,(I,f,.""" P4-02

i"IM"J"vI\f"V'Ir¥v"'V"V'V'V'Il"V"VvJ'orlf"lt'\rvvYV'\f'Vi,.-..r.,"I"v"vy"\r'>(V'(V"~r'>('vW"'" L.-..J

[75 IlV

1 sec

FIGURE J.J II A, EEG showing a repetitive artifact that coincided with the ringing of the telephone ("phone"). B, EEG showing a rhythmic discharge with an unusual field, which related to continuous venous-to-venous hemodialysis. When hemodialysis wasbrieflystopped, the rhythmic activity ceased.

eye movements may be confusing, but are usually easily distinguished by experienced electroencephalographers and are described in technical texts.' Cardiac artifacts are related to the ECG or ballistocardiogram and are especially conspicuous when monitoring for electrocerebral inactivity in brain-death suspects or in referential recordings involving the ears (see Fig. 3-2). The ECG can be monitored by electrodes placed on the chest and recorded on a separate channel, thereby facilitating recognition of such artifacts in the scalp recording, which may otherwise be misinterpreted as sharp waves or spike discharges. Conspicuous ECG artifact may obscure underlying low-voltage electrocerebral activity, a point of concern when comatose patients are being examined for possible brain death. Pulse artifact may occur at any site but is typically localized to a single electrode and appears as a recurrent slow wave that sometimes has a saw-toothed appearance and is time-locked to the ECG. It occurs when an electrode is placed over or close to an artery, and movement of the electrode will eliminate it. Pacemaker artifacts consist of spike discharges that precede the ECG.

Muscle artifact is composed of brief-duration spike discharges that are too rapid to be cerebral in origin. Chewing artifacts are electromyographic (EMG) artifacts produced by the temporalis muscles (see Fig. 3-2). Sucking artifact (in infants) is characterized by bitemporal sharp activity that may be difficult to identify without careful observation of the baby. Lateral rectus spikes are recorded from the anterior temporal electrodes and relate to horizontal eye movements; they are out of phase on the two sides of the head. Attempting to relax or quieten the patient should reduce muscle artifact. In most instances, high-frequency filters should not be used for this purpose except as a last resort because they simply alter the appearance of the artifact (sometimes so that it looks more like EEG fast activity) and also influence the background EEG. Movement produces artifact in addition to muscle activity. Any movements may produce such artifacts, and these will vary in appearance depending on the nature and site of the movement. Movements arising at a consistent site may be monitored by a pair of electrodes and recorded on a separate channel. Hiccups

Electroencephalography: General Principles and Clinical Applications

A

43

Fp1-F7----~-,/

FP1-F7~

F7- T3--~--~, I

F7-T3.~ T3-T5~

T5-01

----------....,..-v---....,"""--'~---_~_~

Fp2-F8 ------.....-..,

T5-01~

FP2-F8~

F8-T4 - - - - - - , I

F8-T4~ T6-02------~------

T4-T6~

L10-A1 - - - - - - - . , T6-02~

Submental--------'

L10-A1~

L---J [ 50 IlV

1 sec FIGURE ]-5 • EEG showing rhythmic slow waves in the temporal regions relating to glossokinetic artifact. It can be seen

LOC-A1~

~ [ 751lV

that the cerebral activity is in phase with activity recorded infraorbitaIIy (indicating that it is not eye movement artifact) and out of phase with activity recorded submentally.

1 sec

c Fp1-F7 F7-T3

-------..,.,...

T3-T5 T5,-01 Fp2-F8 F8-T4 T4-T6

_~

~

N-..

T6-02 L10,-A1 LOC-A1 l--J

[50 IlV

1 sec

FIGURE ]-4 • Eye movements. A, Eye closure. B, Lateral eye movement. C, Eyelid flutter. LIO, left infraorbital; LOC, left outer canthus.

produce a brief generalized movement artifact that may have a pseudo-periodic quality; this is not confined to the EEG but is seen also in, for example, the ECG channel. Glossokinetic artifact relates to the difference in potential between the tip of the tongue and its base; tougue movements generate slow waves that are recorded over one or both temporal regions (Fig. 3-5). When suspected, tongue movements may be monitored by a submental electrode. Sweat artifact is common and is characterized by high-amplitude potentials of very low frequency. Lowfrequency filters can attenuate these slow waves. INSTRUMENTAL ARTIFACTS

Background 50-Hz noise in a restricted number of channels is often related to mismatched electrical impedance

of electrode pairs or to the poor application of electrodes so that slight movement alters transiently the impedance of an individual electrode. A 50-Hz interference signal is normally common to the pair of electrodes connected to an amplifier; the differential amplifier essentially discards this common signal. If the electrical impedance of one of the electrodes is altered, however, the current flowing across that electrode-skin interface will be altered, thereby leading to voltage differences between one electrode and the other of the pair. The differential amplifier will magnify these differences so that the 50-Hz artifact then becomes obtrusive. In addition, other artifacts related to movement arise at that electrode and are limited to the channel with which the faulty electrode is connected, causing "mirror-image" phase reversals when these channels are part of a linked bipolar montage. Widespread 50-Hz artifact is of concern because it may indicate a safety problem warranting attention as discussed in Chapter 2, where the various artifacts that arise in digital equipment are also considered.

EEG Interpretation Evaluation of the EEG for clinical purposes involves definition of the frequency, amplitude, and distribution of the electrical activity that is present, and of its response to external stimulation such as eye opening. The degree of synchrony and symmetry over the two sides of the head is noted. The presence of any focal activity is determined and its nature characterized. The findings must be interpreted in relation to the patient's age and level of arousal. For descriptive purposes, EEG activity is usually characterized on the basis of frequency.

44

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

ACTIVITY RECORDED IN THE EEG The mechanisms responsible for the presence of widespread rhythmic activity in the brain are not known. Some means of generating rhythmic activity must be involved, as also must some method of synchronizing the activity of different cerebral regions. Experiments on animals have produced considerable evidence to suggest that the rhythmic activity normally recorded from the scalp has a cortical origin, being derived from the postsynaptic potentials of cortical neurons. In particular, it is the pyramidal neurons-cells that are vertically oriented with regard to the cortex and have a large apical dendrite extending toward the surface-that are important in this respect. The origin of the scalprecorded EEG primarily from postsynaptic potentials, rather than action potentials, is in keeping with the former's longer duration and synchronous occurrence over a large area of cell membrane. Potentials arising from neuronal activity in subcortical structures or from horizontally oriented cortical cells contribute little, if anything, to the normal scalp-recorded EEG. The factors that determine whether a cortical potential is recorded over the scalp include its voltage, the extent to which the generator cells are discharging synchronously, the area of cortex involved, and the site of cortical involvement with respect to the sulcal convolutions. Potentials arising in the sulcal depths are less likely to be recorded over the scalp by conventional EEG than are those arising at the surface. The cortical activity has a regular rhythmicity that seems to depend on the functional integrity of subcortical mechanisms. It has generally been accepted that the thalamus serves as the pacemaker of certain of the cortical rhythms that are recorded during electroencephalography, but intracortical circuitries may also be involved significantly" The precise details are beyond

Eyes open

Alpha Adivity ALPHA RHYTHM

Alpha rhythm may have a frequency of between 8 and 13 Hz, but in most adults it is between 9 and 11 Hz. This rhythm is found most typically over the posterior portions of the head during wakefulness, but it may also be present in the central or temporal regions. Alpha rhythm is seen best when the patient is resting with the eyes closed. Immediately after eye-closure, its frequency may transiently be increased (the "squeak" phenomenon). The alpha rhythm is not strictly monorhythmic but varies over a range of about 1 Hz even under stable conditions. It is usually sinusoidal in configuration, may wax and wane spontaneously in amplitude, and sometimes has a spiky appearance; a spindle configuration denotes a beating phenomenon that results from the presence of two (or more) dominant frequencies. The alpha rhythm is attenuated or abolished by visual attention (Fig. 3-6) and affected transiently by other sensory stimuli and by other mental alerting activities (e.g., mental arithmetic) or by anxiety. The term paradoxical alpha rhythm refers to the appearance of alpha rhythm on eye-opening in drowsy subjects; this represents an alerting response. Alpha activity is well formed and prominent in many normal subjects but is relatively inconspicuous or absent in about 10 percent of instances. Its precise frequency is usually of

Eyes shut

----------'r----...

-,~-,

Fp1-F3 - - - ,

the scope of this chapter. The physiologic basis of the abnormal rhythms that are encountered at electroencephalography is even less clearly defined. It also remains unclear whether it is possible to record at the scalp the EEG from sources deeper than the most superficial cortex.I? but source area is clearly important in this regard.

F3-C3 - - - - . . . . - - - - - - - - - - - - - - . . . -" - -~. -----C3-P3 'il

....

,

WINto_-------------_/oIiNN-_..,,;

P3-01 Fp2-F4 F4-C4 C4-P4

_,."w.I'-----------------""""'----

--""'--l----------r----

-----f---------------..r-----; ,.

~

P4-02~,

L....-J

1 sec

r= .

t'--

~_'Mt'

,•.,.i'MJ.\JiM 200 IJ.V [

FIGURE :S·I .. Normal 9- to IO-Hz alpha rhythm recorded in the EEG of a 64-year-oldman. The posterior distribution and responsiveness to eye-opening of the rhythm can be seen.

Electroencephalography: General Principles andClinical Applications

little diagnostic significance unless information is available about its frequency on earlier occasions. In children, the dominant, posterior responsive rhythm reaches about 8 Hz by the age of 3 years and reaches 10 Hz by approximately 10 years of age. Slowing occurs with advancing age; as a consequence of certain medication (e.g., anticonvulsant drugs); and in patients with clouding of consciousness, metabolic disorders, or virtually any type of cerebral pathology. The alpha activity may increase in frequency in children as they mature and in older subjects who are thyrotoxic. A slight asymmetry is often present between the two hemispheres with regard to the amplitude of alpha activity and the degree to which it extends anteriorly. In particular, alpha rhythm may normally be up to 50 percent greater in amplitude over the right hemisphere, possibly because this is the nondominant hemisphere or because of variation in skull thickness. A more marked asymmetry of its amplitude may have lateralizing significance but is difficult to interpret unless other EEG abnormalities are present, because either depression or enhancement may occur on the side of a hemispheric lesion. Similarly, a persistent difference in alpha frequency of more than 1 to 2 Hz between the two hemispheres is generally regarded as abnormal. The side with the slower rhythm is more likely to be the abnormal one, but it is usually difficult to be certain unless other abnormalities are also found. Unilaterally attenuated or absent responsiveness of the alpha rhythm sometimes occurs with lesions of the parietal or temporal lobe (Bancaud's phenomenorn.!" Asymmetric attenuation of the alpha rhythm during mental alerting procedures with the eyes closed may also be helpful for lateralizing any impairment of cerebral function. 10 Some normal adults have an alpha rhythm that is more conspicuous centrally or temporally than posteri-

orly, or has a widespread distribution. Care must be taken not to misinterpret such findings as evidence of abnormality. The so-called slow alpha variant resembles normal alpha rhythm in distribution and reactivity but has a frequency of about 4 to 5 Hz, which approximates one-half that of any alpha rhythm in the same record. This variant is of no pathologic significance. OTHER RHYTHMS OF ALPHA FREQUENCY

Not all activity having a frequency of 8 to 13 Hz is necessarily an alpha rhythm. Alpha-frequency activity that is widespread in distribution and unresponsive to external stimulation is found in some comatose patients (p. 78). Temporal alpha activity is sometimes found in elderly subjects and may be asynchronous, episodic, and persistent during drowsiness. Runs of activity in the alpha range of frequencies are occasionally found frontally in children immediately after arousal from sleep (frontal arousal rhythm) and are not of pathologic significance. Mu rhythm has a frequency that usually is in the alpha range, is seen intermittently over the central region of one or both hemispheres, is unaffected by eye-opening, and is blocked unilaterally or bilaterally by movement or the thought of movement (Fig. 3-7). It is also blocked by sensory stimulation. Bilateral mu rhythm is often asynchronous and may exhibit amplitude asymmetries between the two hemispheres. The negative portions of the waves are sharpened, and the positive portions are generally rounded. Mu rhythm is often associated with a centrally located beta rhythm that is also attenuated by contralateral movement. In most instances, mu rhythm has no diagnostic significance. It is found in about 20 percent of young adults. When recorded over a skull defect, it may be mistaken as a potential epileptogenic abnormality. I I

Clench R fist

Fp1-F3

--..,-....~

F3-C3 - " _ •• C3-P3.·...

oAo"

__ - - - - - - - - -......- - - - - - - - - - - - - -

P3-01 -/tIIdIfiI"-' --.,.........- ----.-.....~----------------Fp2-F4 F4-C4 -

r.......'' ' ' /11,.,. ---v-

~ _. ",~.

45

t----

r---------

C4-P4 -....... , . - - - -.........- - - - - - - - - - - - - - - - - - Ii' '"

P4-02~r" -;~o fLY [ FIGURE J·7 • Bilateral mu rhythm recorded in the EEG of a 26-year-old woman with no neurologic disorder. The effect on the rhythm of clenching the right fist can be seen.

46

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Beta Adivity Any rhythmic activity that has a frequency greater than 13 Hz is referred to as beta activity. Activity of this sort is present anteriorly in the EEG of normal adults. Beta activity, responsive to eye-opening, is sometimes found over the posterior portions of the hemispheres and is then best regarded as a fast variant of the alpha rhythm. Beta activity that fails to respond to eye-opening is a common finding and usually has a generalized distribution, but in some instances it is located centrally and is attenuated by tactile stimulation or contralateral movements. It usually has an amplitude of less than about 30 flV. The amount of such activity varies considerably among normal subjects. Activity having a frequency between 18 and 25 Hz is usually more conspicuous during drowsiness, light sleep, and rapid eye movement (REM) sleep than during wakefulness. It may also be augmented by cognitive tasks. Beta activity may be induced by a number of different drugs, particularly barbiturates and the benzodiazepine compounds, but also neuroleptics, antihistaminics, n-amphetamine, methylphenidate, and cocaine.'! Drug-induced fast activity is typically diffuse and symmetric over the two hemispheres. Focal or lateralized spontaneous beta activity, or asymmetric drug-induced fast activity, raises the possibility of localized cerebral pathology. In such circumstances, however, it must be borne in mind that the amplitude of beta activity may be increased either ipsilateral or contralateral to a lesion involving one cerebral hemisphere; and that an amplitude asymmetry is common in normal subjects, with beta activity being up to 30 percent lower on one side than the other. Beta activity is increased in amplitude over the area of a skull defect owing to the greater proximity of the recording electrodes to the surface of the brain and the low-impedance pathway. It is reduced in amplitude over a subdural collection of fluid or localized swelling or edema of the scalp, and transiently in either a localized or lateralized manner after a partial seizure. Generalized paroxysmal fast activity is a rare finding, occurring in less than 1 percent of EEG recordings; it may be mistaken for muscle artifact, drug effect, or sleep spindles, depending on the circumstances. In one study of 20 patients with such activity, all had seizure disorders, usually with seizures of more than one type, and most were mentally retarded.F The paroxysmal EEG disturbance occurred almost always during sleep, often with associated clinical seizures that were commonly of the tonic variety. 12

Theta Adivity Activity with a frequency between 4 and 7 Hz is referred to as theta activity. Theta and slower activity is usually

very conspicuous in children but becomes less prominent as they mature. Some theta activity is often found in young adults, particularly over the temporal regions and during hyperventilation, but in older subjects theta activity with amplitude greater than about 30 flV is seen less commonly except during drowsiness. Focal or lateralized theta activity may be indicative of localized cerebral pathology. More diffusely distributed theta activity is a common finding in patients with a variety of neurologic disorders, but it also may be caused by nothing more than a change in the patient's state of arousal. Rhythmic trains of midline theta activity, occurring especially at the vertex and having an arciform, sinusoidal, or spiky configuration, have been described as a nonspecific finding in patients with many different disorders. Such activity may be persistent or intermittent; may be present during wakefulness or drowsiness; and shows variable reactivity to eye-opening, movement, alerting, and tactile stimulation.!" It is not present during sleep. Its origin is uncertain.

Delta Adivity Activity that is slower than 4 Hz is designated delta activity.Activity of this sort is the predominant one in infants and is a normal finding during the deep stages of sleep in older subjects. When present in the EEG of awake adults, delta activity is an abnormal finding. Delta activity, responsive to eye opening, is commonly seen posteriorly (intermixed with alpha activity) in children and sometimes in young adults; it is then designated posteriorslow waves ofyouth. The spontaneous occurrence interictally of posterior, rhythmic slow waves is well described in patients with absence seizures. The slow activity has a frequency of about 3 Hz, is present during wakefulness, is responsive to eye-opening, and may be enhanced by hyperventilation. The symmetric or asymmetric occurrence of rhythmic delta activity over the posterior regions of the head after eyeclosure is rare. Such activity usually lasts for no more than 2 or 3 seconds and is a nonspecific finding that has been described in a number of different neurologic disorders. 10 POLYMORPHIC DELTA ACTIVITY

Polymorphic delta activity is continuous, irregular, slow activity that varies considerably in duration and amplitude with time; it persists during sleep, and shows little variation with change in the physiologic state of the patient (Fig. 3-8). It has been related to deafferentation of the involved area of the cortex and to metabolic factors. Such activity may be found postictally and in patients with metabolic disorders. It is commonly seen, with a localized distribution, over destructive cerebral

Electroencephalography: General Principles and Clinical Applications

47

T3-Cz Cz-T4

L-J 1 sec

200 !LV [

FIGURE ]·8 • EEG of a 19-year-old patient with encephalitis, showing a background of diffuse, irregular theta and delta activity.

lesions involving subcortical white matter (Fig. 3-9), but it generally is not found with lesions restricted to the cerebral cortex itself. 13 ,14 It may be found either unilaterally or bilaterally in patients with thalamic tumors or lesions of the midbrain reticular formation, but its distribution in such circumstances is somewhat variable. Thus, although diffuse irregular slow activity is found over one hemisphere in some cases,'" in others it has a more restricted distribution. Gilmore and Brenner correlated the EEG finding of focal polymorphic delta activity with the CT scan appearance in 100 patients.l'' Focal CT abnormalities were present in 68, and nonfocal abnormalities (e.g., diffuse atrophy or cerebral edema) were found in 10. The maximal delta focus did not always overlie the CT abnormality, but it never appeared in the opposite hemisphere. No CT abnormalities whatsoever were found in 22, most of whom had seizure disorders, cerebral trauma, or ischemic strokes. The experience of oth-

ers has been similar. 16 Schaul and colleagues found that the amplitude, frequency, and distribution of focal slow waves did not relate to lesion size or to mass effect. 17 Diffuse polymorphic delta activity occurs in patients with white matter encephalopathies" and following acute or extensive lesions of the upper brainstem. Unilateral lesions of the midbrain tegmentum-unlike bilateral lesions-did not, however, produce delta activity in the experimental studies performed by Gloor and co-workers in cats.!" INTERMITTENT RHYTHMIC DELTA ACTIVITY

Intermittent rhythmic delta activity is paroxysmal, has a relatively constant frequency, and is usually synchronous over the two hemispheres. It is often more prominent occipitally in children or frontally in adults (Fig. 3-10), may be enhanced by hyperventilation or drowsiness, and usually is attenuated by attention. Its origin is unclear, but

F3-A1 "W"......".,.,..."rW"1,(\rI'''''-INfoIo1A,N\t...fV\JoI'\f'JvI,'VIvv~wr.~~IfJ''\'''''''''roMNv..-''''''..J'VV\ C3-A 1 "vv-~N"vAJl,..rv-.""""'¥oNV''VNv..".N''''''''''iJv''M,"""",,,,,,,,,,~~'"M't.Ja.J~'''-''~ P3-A 1 ............J"\M-~A/I,No-..~V"V.M!"'''-I.JJI/W'w...''''fJfvI''fo'J\\~''''''V'Vof'v~ItY'''''I\V'WofIY/v"O/, 01-A1 ~NIv""""'~"JIV-'~HN'VV"JV.J""""",~~N"o~''''W'W!''~ F4-A2

~,V'-··\~V>/'(\".~./V~~.Jvt

C4-A2 P4-A2 02-A2

'\NI.;...,I<..J ....I\fV'.N .... AJ"''IN'''I

.Jo!'w,.,t\t'\,I'v'I.,.rv.'I.M,M/'~"__'"

L-.-J

1 sec

300 !LV[

FIGURE ]., • EEG of a 52-year-old man who had a right parietal glioma. Note the polymorphic slow-wave focus in the right central region and the diffusely slowed background.

48

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp2-F4 F4-C4 C4-P4 """""'--""'" \./"'v""""''''''''"~~-.r'I_''-'..N\jV P4-02 '''''-""f"M..!VI,""'r-.J'''''''-----\..I'-t'v-I'\"f'vJ'v.J\f'V'''''-'

".,....''''''V.........._

Fp2-F8 F8-T4 __ f " _ "V T4-T6 _ _"V'I"""""'_ _.........

~-v"......,...,....,.~"V

T6-02 _;WW""",,,,' \1''''1

"""'"v~"'...JI"""'--.i'wJ'v.-J''''rV

\../'V-.A._""''''''''''''W-VVW'---

Fp1-F3~ F3-C3 _-""-""","\I C3-P3<-""""....,...,-'Y-........

..-.....----........,.""V'-II'-J-.......-'-""-'---'¥\",....,..-

P3-01 w-"""""""'~""""'--""'-..._""V''''_''..rv...........Nv'1'''''"''''--"''''''''''..,.,...y...,..,..WI,/ Fp1-F7 F7-T3~---'I'

T3-T5 T5-01~

1 sec

300 IJ-V [

FIGURE ]-10 III Frontal intermittent rhythmic delta activity recorded in the EEG of a 14-year-old boy with obstructive hydrocephalus.

it probably relates to dysfunction of subcortical centers influencing the activity of cortical neurons. Its significance is considered on page 51.

cance by their response to eye-closure or reduction in the level of background illumination.

Triphasic Waves Breach Rhythm Breach rhythm is a mu-like rhythm found in patients with skull defects after surgical operations. It has a frequency of between 6 and 11 Hz, usually with faster components, and the waves often have spike-like negative phases. The rhythm recorded parasagittally is often responsive to fist-clenching and other stimuli, whereas that recorded more laterally (at the T3 or T4 electrode) is unresponsive to any stimuli.'? The presence of a breach rhythm has no predictive value for the development of seizures or the recurrence of the intracranial pathology that necessitated the original surgery.

Lambda Waves Lambda waves are electropositive sharp waves that may occur in the occipital region in normal subjects who are looking- at and scanning something (e.g., reading) in a well-illuminated field, particularly if their attention and in terest are aroused. Morphologically similar activity is sometimes seen during non-REM sleep. The nature of these potentials is unclear, and they have no known diagnostic significance at present. They are sometimes asymmetric over the two hemispheres but can be distinguished from sharp transients of pathologic signifi-

Triphasic waves typically consist of a major positive potential preceded and followed by smaller negative waves. They are found most characteristically in metabolic encephalopathies when they usually are generalized, bilaterally synchronous, and frontally predominant. They are sometimes reactive to external (painful) stimulation. Triphasic waves were originally thought to be specific for hepatic encephalopathy, but in fact are found in a variety of metabolic encephalopathies'P-" and suggest a poor prognosis for survival. 20 In one study of 50 patients with triphasic waves in their EEGs, the etiology was found to be hepatic dysfunction in 28, azotemia in 10, anoxia in 9, hypoglycemia in 1, and hyperosmolality in 2. 22 They have also been described in hypothermia's; myxedema comav'; neuroleptic malignant syndrome's; and a variety of other neurologic disorders, including dementia.i? Triphasic waves may occur in patients receiving pentobarbital, especially as the drug is being tapered after treatment of status epilepticus.P during which time they should not be mistaken for epileptiform activity. Similarly, they have been noted in patients with primary generalized epilepsy at a time when they had a postictal depression in level of consciousness.F The periodic complexes found in certain conditions (p, 52), especially Creutzfeldt-jakob disease, may take the form of triphasic waves.

Electroencephalography: General Principles and Clinical Applications

Spike Discharges One of the major uses of electroencephalography is in the investigation of patients with suspected epilepsy. In this regard, the presence in the EEG of interictal spike discharges or sharp waves is often held to be suggestive of an epileptic disturbance. Epileptiform activity is defined as abnormal paroxysmal activity consisting, at least in part, of spikes or sharp waves resembling those found in many patients with epilepsy. It is not synonymous with an electrographic seizure and should be clearly distinguished from the latter. A spike is defined arbitrarily as a potential having a sharp outline and a duration of 20 to 70 msec, whereas a sharp wave has a duration of between 70 and 200 msec. Because such activity may occur in nonepileptic subjects, its presence must be interpreted with caution. The distinction between epileptiform and nonepileptiform sharp transients is usually made intuitively but bearing in mind certain published guidelines." Epileptiform sharp transients are usually asymmetric in appearance, are commonly followed by a slow wave, have a duration that differs from that of the ongoing background activity, may be biphasic or triphasic, and often occur on a back-

Fp2-F4 F4-C4

ground containing irregular slow elements (Fig. 3-11). These criteria distinguish between epileptiform activity and background activity that is sharp and variable in amplitude (e.g., a spiky alpha rhythm). Pathologic spike discharges have different clinical implications depending on their characteristics and location. Focal epileptiform spike discharges arise from a localized cerebral region. The likelihood of spikes arising from a particular area depends on the age of the patient, type of underlying lesion, and epileptogenicity of the involved region. Slowly progressive lesions are more likely to be associated with such activity than are rapidly progressive ones, and the frontal and temporal lobes are more epileptogenic than the parietal and occipital lobes. The benign epileptiform discharges that occur in drowsy subjects (p, 53) have a different significance from that of the anterior or mesial temporal spike discharges found interictally in patients with complex partial seizures. Similarly, 3-Hz, 6-Hz, and 1- to 2-Hz spike-wave discharges differ in their clinical and prognostic relevance. Moreover, the significance of frankly epileptiform discharges depends on the clinical circumstances in which they are found.

-.;o.....,................. \.flJ¥'~~JJ'r'0('vfJ1"J\{I\fv~

.......,"Wo./.."...............~......~'lIIt"'......."..,.,~........,,...,..""""""" ......._..,,.,...............

~\f~

C4-P4 P4-02 Fp2-F8 F8-T4 T4-T6 T6-02 Fp1-F3 F3-C3 C3-P3 P3-01 Fp1-F7

---~---~-------------"'---"'---

F7-T3 yvT3-T5 T5-01

L...-...J 1 sec

1000 /LV [

III Interictal spike discharges arising independently in the central region of either hemisphere in the EEG of a patient with seizures since infancy.

FIGURE 3-11

49

50

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Epileptiform spike discharges arising focally in the temporal region are associated most commonly with complex partial seizures and are activated during drowsiness or sleep. Intermittent delta activity may also be present.F' Frontal epileptiform spike discharges have been associated with seizures of different types, including complex partial and focal motor seizures, posturing, and drop attacks.P? Occipital discharges of very short duration may occur in congenitalIy blind children without seizures. In general, occipital spikes tend to be benign, although they sometimes are related to structural abnormalities or occipital seizures, especially in older persons. One form of benign epilepsy of childhood is associated with spike-wave discharges that occur in one or both occipital regions after eye-closure.P' Focal epileptiform spike discharges relate to paroxysmal depolarizing shifts occurring synchronously (as a result of disinhibition and excitation) in a population of neurons. Such shifts consist of a 20- to 30-mV depolarization of the cell membrane for up to about 100 msec, with a superimposed train of action potentials and followed by an afterhyperpolarization that may last for approximately 2 seconds. The slow wave that is recorded from the scalp after many epileptiform spikes ("spike-wave discharge") has been attributed to this afterhyperpolarization. Inhibitory activity is also triggered in the surrounding cortex as well as contralaterally,32 and this activity may be responsible for the focal slow activity that often occurs intermittently in association with an active spike focus. Generalized spike-wave discharges are discussed in detail in the next section.

Paroxysmal Adivity Paroxysmal activity has an abrupt onset and termination, and it can be distinguished clearly from the background activity by its frequency and amplitude. It may occur as a normal phenomenon during hyperventilation in young adults and in response to arousal or sensory stimuli during sleep. The epileptiform transients just described are one example of paroxysmal activity. Abnormal paroxysmal activity that consists, at least in part, of epileptiform sharp transients (epileptiform activity) has a high correlation with the occurrence of epileptic seizures. It does not necessarily represent seizure (ictal) activity, however, and scalp-recorded electrographic seizures may not contain any epileptiform activity. SPIKE-WAVE ACTIVITY

Generalized, bilaterally symmetric and bisynchronous 2.5- to 3-Hz spike-wave activity is the expected finding in patients with primary or idiopathic generalized epilepsy; it is enhanced by hyperventilation or hypo-

glycemia. More is known about the fundamental mechanisms underlying the absence seizures that occur in primary generalized epilepsy than about other types of generalized seizure. Gloor and Fariello related the generalized spike-wave activity associated with absence seizures to abnormal oscillatory discharges between cortical and thalamic neurons. 3:1,34 This oscillation involves the regular alternation of brief periods of neuronal excitation (associated with a markedly increased firing probability) during the spike phase with longer periods of neuronal silence during the slow-wave phase. Large populations of cortical and thalamic neurons are affected in near synchrony, and the brainstem reticular formation also participates. Generalized 2.5- to 3-Hz spike-wave activity may also be found in secondary generalized epilepsy (i.e., in patients with generalized seizures secondary to known pathology), but it is then superimposed on a diffusely abnormal background and is often accompanied by other EEG abnormalities. In the secondary generalized epilepsies, however, slow (p. 61) or atypical (p, 62) spike-wave activity is found more commonly, Generalized, bisynchronous spike-wave activity rarely may arise from a unilateral cortical focus, particularly on the medial surface of the hemisphere; this phenomenon is referred to as secondary bilateral synchrony. In such circumstances, the paroxysmal activity usually has a faster or slower frequency than that in primary generalized epilepsy, and the form and relationship of the spike to the wave component of the complex is less regular. Further, a consistent asymmetry of amplitude and waveform may exist between the hemispheres, the activity being either more or less conspicuous on the affected side. Recognition of the cortical origin of such activity is facilitated when isolated focal discharges arise from one side, particularly if they consistently precede the bursts of bilaterally synchronous activity; otherwise, recognition can be difficult unless the paroxysmal discharges have a focal or lateralized onset. The mechanisms generating secondary bilateral synchrony of spike-wave discharges are not fully established but may involve either the propagation of discharges from one hemisphere to the other along the forebrain commissures or the activation by a cortical focus of diencephalic or other midline structures, which then elicit a synchronous discharge from both hemispheres. Studies have shown diminished but persistent bisynchronous epileptiform discharges after corpus callosotomy in patients with seizure disorders, suggesting that both mechanisms may be important." It should be noted that bilaterally synchronous spikewave activity may also be seen in rare instances in patients with structural subtentorial or midline lesions.!" as well as in unselected nonepileptic patients:17,3H and in the clinically unaffected siblings of patients with primary generalized epilepsy.

Electroencephalography: General Principles and Clinical Applications

Gloor and co-workers related EEG changes to the distribution of the pathologic process in 32 cases of diffuse encephalopathy.l" They found generalized paroxysmal disturbances, consisting of either slow activity, spike-wave activity, or sharp transients, in patients with a diffuse encephalopathic process involving predominantly the cortical and subcortical gray matter, but not when the white matter alone was involved. If both gray and white matter were affected, generalized paroxysmal activity occurred on a background of continuous polymorphic slow activity. Bilaterally synchronous spike-wave activity was seen in diffuse gray matter encephalopathies and was usually more slow and irregular than that in patients with primary epilepsy. INTERMITTENT RHYTHMIC DELTA ACTIVITY

Intermittent rhythmic delta activity (see Fig. 3-10) with a fron tal predominance (often designated FIRDA) in adults and an occipital emphasis in children was referred to on page 47. It may result from a destructive lesion or from pressure and concomitant distortion affecting midline subcortical structures, in particular the diencephalon and rostral midbrain; it also occurs with deep frontal lesions. However, in a study of the EEG in 154 patients with well-defined diencephalic, midbrain, or posterior fossa lesions, Schaul and colleagues found such activity nonspecific and of no particular diagnostic significance, occurring in only 26 percent of patients.t" There was no means of clearly distinguishing

L...J 1 sec

51

the EEG pattern in deep midline lesions from that in diffuse cortical or subcortical encephalopathies or in metabolic encephalopathies. In a further study, Schaul and co-workers compared the findings in 42 patients with FIRDA with those of a control group with normal EEGs; they found that in most of these patients there was a diffuse encephalopathic process rather than a lesion limited to deep midline structures.t" Similarly, other authors have found FIRDA in a variety of neurologic disorders and in metabolic encephalopathies." BURST-SUPPRESSION PATTERN

The so-called burst-suppression pattern is characterized by bursts of high-voltage, mixed-frequency activity separated by intervals of marked quiescence or apparent inactivity that may last for no more than a few seconds or as long as several minutes. It occurs with a generalized distribution during the deeper stages of anesthesia; in patients who are comatose following overdosage with central nervous system (eNS) depressant drugs; and in any severe diffuse encephalopathy, such as that following anoxia (Fig. 3-12). The bursts may be asymmetric or bisynchronous. The prognostic significance of a burst-suppression pattern depends on the circumstances in which it is found. When it follows cerebral anoxia, it is associated with a poor outcome. Spontaneous eye-opening,42,43 nystagmoid movements.r' pupillary changes, facial movements." myoclonus.t" and limb movements have occasionally been associated with

100 IloV [

FIGURE :J-12 • Burst-suppression pattern recorded in the EEG of a 70-year-old man

after a cardiac arrest from which he was resuscitated.

52

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

the EEG bursts," and these sometimes mimic volitional activity.t" In other instances, movements occur exclusively between EEG bursts,"? Experimental studies in

animals indicate that approximately 95 percent of cortical neurons become hyperpolarized and then electrically silent during periods of EEG suppression; this results from increased inhibition at cortical synapses, which also leads to functional disconnection of the cortex from its thalamic input.t" PERIODIC COMPLEXES

Repetitive paroxysmal slow- or sharp-wave discharges, or both, may occur with a regular periodicity in a number of conditions. Such periodic complexes are seen most conspicuously and with a generalized distribution in subacute sclerosing panencephalitis and Creutzfeldt-jakob disease and are sometimes found in patients with liver failure. Periodic complexes that exhibit, to a greater or lesser extent, a regular rhythmicity in their occurrence may also be found in patients with certain lipidoses, progressive myoclonus epilepsy, drug toxicity, anoxic encephalopathy, head injury, subdural hematoma, occasionally after tonic-clonic seizures, and in rare instances in other circumstances. Their diagnostic value therefore depends on the clinical circumstances in which they are found. PERIODIC LATERALIZED EPILEPTIFORM DISCHARGES

Repetitive epileptiform discharges sometimes occur periodically as a lateralized phenomenon (designated periodic lateralized epileptiform discharges, or PLEDs), although they are often reflected to some extent over the homologous region of the opposite hemisphere as well. Their amplitude has ranged between 50 and

Fp1-F3

...,......_'-""

F3-C3-~---"---

300 IlV, and their periodicity between 0.3 and 4 seconds in different series.t? They typically are seen in patients with hemispheric lesions caused, most commonly, by cerebral infarction (Fig. 3-13), hemorrhage, or tumors. In patients with acute hemispheric stroke, PLEDs are more likely when there are associated metabolic derangements, especially hyperglycemia and fever.t" Patients with herpes simplex encephalitis may have PLEDs, especially in the first 2 weeks of the illness, but their absence in no way excludes the diagnosis. Other infections occasionally associated with PLEDs include neurosyphilis, cysticercosis, bacterial meningitis, mononucleosis encephalitis," and early Creutzfeldt-Jakob disease. 52 PLEDs may also be found in patients with chronic seizure disorders or long-standing static lesions (e.g., old infarcts), sometimes as a persistent phenomenon 53 but more often for a brief period, especially when seizures or toxic metabolic disturbances have recently occurred. They are found occasionally with metabolic disorders that more typically produce diffuse EEG disturbances (e.g., anoxia, hepatic encephalopathy, and abnormal blood levels of glucose or calcium) and have also been reported with head injury, subdural hematoma, cerebral abscess or other cystic lesions, and sickle cell disease. 53•54 In many cases, no specific cause can be found. During the acute stage of illness, patients with PLEDs are usually obtunded and commonly have seizures (especially focal seizures) and a focal neurologic deficit. The PLEDs themselves are usually an interictal pattern, being replaced or obscured by other activity as a seizure develops. They may also occur ictally, however, as in patients with epilepsia partialis continua. Moreover, patients have been described with recurrent or prolonged confusional episodes during which PLEDs were present in the EEG; with clinical improvement, the EEG normalized. 55 PLEDs typically

--..I'--.. . . . .

~-------'--

--{'..---·~

C3-P3........-..r---JI'-----""f---.,....--......,'~ P3-01 _ . . . ,.,..._-"\r---v,,,...,.--....,,....,..--'l.r---""v'-----'1t--""'v'\,.r ----"'-

Fp2-F4---....-...~---.r...------.~------_......--A..

..........---------........,.---~-~-_--r"-,~

F4-C4-""""'-~

C4-P4~...-·---------.......-·------------------

P4-02 L..J 1 sec FIGURE 3-13 • Periodic lateralized epileptiform discharges (PLEDs) over the left

hemisphere in the EEG of a 78-year-old woman with a recent stroke.

Electroencephalography: General Principles and Clinical Applications

occur every 1 to 2 seconds; vary in morphology in different patients; and usually (but not always53) disappear over the course of a few days or weeks, to be replaced by a focal or lateralized polymorphic slow-wave disturbance or by isolated spike discharges. The underlying disease determines the prognosis of patients with PLEDs. In one study, 41 percent of patients with PLEDs died during their hospitalization or within 2 months of discharge." The pathophysiologic basis of PLEDs is not understood. Some have considered PLEDs to be equivalent to the terminal phase of status epilepticusr': others have reported their association with rhythmic discharges that have a stereotyped distribution, frequency, configuration, and amplitude for individual patients, and that may be obscured by the development of frank electrographic seizure discharges.F PLEDs may occur independently over both hemispheres; they are then designated BIPLEDs. The complexes over the two hemispheres differ in their morphology and repetition rate. BIPLEDs are most commonly caused by anoxic encephalopathy, multiple vascular lesions, and eNS infection with either herpes simplex or other agents, but they may also be found in patients with chronic seizure disorders or with recent onset of seizures. BIPLEDs are usually found in comatose patients and are associated with a much higher mortality than are PLEDs. 58

Low-Voltage Records A number of normal subjects have generally low-voltage EEGs, consisting of an irregular mixture of activity with a frequency ranging between 2 and 30 Hz and an amplitude of less than 20 1lV. A little alpha activity may, however, be present at rest or during hyperventilation, and it is sometimes possible to enhance the amplitude of background rhythms by simple or pharmacologic activating procedures.t? This low-voltage EEG pattern may occur on a hereditary basis; an autosomal dominant mode of inheritance has been suggested.I" Similar lowvoltage records are occasionally encountered in patients with Huntington's disease or myxedema, but they are of no diagnostic value. Such records should be distinguished from those consisting primarily of lowvoltage delta activity.

other EEG Patterns Over the years, special pathologic significance has been attributed (without adequate justification) to a number ofEEG patterns that are now known to occur as a normal phenomenon in some healthy subjects, particularly during drowsiness or sleep. These patterns are therefore of dubious clinical relevance, but

53

they merit brief comment to prevent their misinterpretation.

14- AND 6-Hz

POSITIVE SPIKES

During drowsiness and light sleep, especially in adolescents, runs of either 14- or 60Hz positive spikes, or both, may occur, superimposed on slower waves; generally they last for less than about 1 second. They are found especially in the posterior temporal or parietal regions on one or both sides and are best seen as surface-positive waveforms on referential recordings. They are of no pathologic relevance, although bursts of such activity have been described in comatose patients with Reye's syndrome and in adults with hepatic or renal disease. SMALL SHARP SPIKES OR BENIGN EPILEPTIFORM TRANSIENTS OF SLEEP

Small sharp spikes or benign epileptiform transients of sleep are found during drowsiness or light sleep in as many as one-quarter of normal adults. Generally, they consist of monophasic or biphasic spikes; they are sometimes followed by a slow wave but are unaccompanied by sharp waves or rhythmic focal slowing of the background. They usually occur independently over the two hemispheres with sporadic shifting localization, but are best seen in the anteromesial temporal regions. Their appearance varies in different patients. They are distinguished from transients of pathologic significance by their bilateral occurrence, failure to occur in trains, and disappearance as the depth of sleep increases, and by the absence of abnormal background activity. Although commonly less than 50 IlV, they are sometimes larger, so that size is not a reliable distinguishing feature. Such discharges are best regarded as normal and are of no diagnostic help in the evaluation of patients with suspected epilepsy?'

6-Hz SPIKE-WAVE ACTIVITY Brief bursts of 60Hz spike-wave activity, usually lasting for less than 1 second, may occasionally be seen in normal adolescents or young adults, especially during drowsiness, and are sometimes referred to as phantom spike-waves. They disappear during deeper levels of sleep, unlike pathologically significant spike-wave discharges. In some instances, the discharges are bilaterally symmetric and synchronous, but in others they are asymmetric. The spike is usually small compared with the slow wave and may be hard to recognize; when the spike is large, the discharges are more likely to be of pathologic significance. Discharges that are frontally predominant are also more likely to be associated with epilepsy than are discharges that are accentuated occipitally.

54

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

WICKET SPIKES

RHYTHMIC THETA DISCHARGES

A wicket spike pattern usually is found in adults, most commonly in drowsiness or light sleep but also during wakefulness. It may consist of intermittent trains of sharp activity resembling a mu rhythm or of sporadic single spikes that are surface-negative and essentially monophasic. Its frequency is usually between 6 and 12 Hz. Wicket spikes are generally best seen over the temporal regions, either bilaterally or independently over the two sides, and sometimes have a shifting lateralized emphasis. The spikes are not associated with a subsequent slow wave or with any background abnormality. Wicket spikes have no diagnostic significance and, in particular, do not correlate with any particular symptom complex, including epilepsy. When these spikes occur singly, however, they may be mistaken for interictal temporal spike discharges.F

In patients older than about 40 years, bursts of rhythmic sharpened theta activity at about 5 to 7 Hz may occur, often with an abrupt onset and termination. In contrast to seizures, they typically show no evolution of their frequency, distribution, or configuration, and they are not followed by focal or diffuse slow activity. These subclinical rhythmic electrographic discharges in adults (known by the acronym SREDA) are usually distributed bilaterally and diffusely but occasionally are focal or lateralized. When diffuse, the discharges are often most conspicuous over the parietal and posterior temporal regions. They can occur at rest, during hyperventilation, or with drowsiness. The bursts may last up to 1 or 2 minutes, or even longer, and may be mistaken for a subclinical seizure. Digital analysis has revealed that the bursts are maximal in the parietal or parietocentrotemporal region and that they consist of a complex mixture of rapidly shifting frequencies that show little spatial or temporal correlation.P Such discharges are probably of no diagnostic relevance and do not correlate with epilepsy or any specific clinical complaints.P' Unusual variants of SREDA include predominantly delta frequencies, notched waveforms, or discharges having a frontal or more focal distribution; bursts that are more prolonged in duration; and discharges that occur during sleep.65

RHYTHMIC TEMPORAL THETA BURSTS OF DROWSINESS (PSYCHOMOTOR VARIANT)

During drowsiness or light sleep, especially in young adults, bursts of rhythmic sharpened theta waves with a notched appearance sometimes occur, predominantly in the midtemporal regions either unilaterally or bilaterally (Fig. 3-14). If bilateral, they may occur synchronously or independently, with a shifting emphasis from one side to the other. Individual bursts commonly last for at least 10 seconds. Unlike electrographic seizure discharges, the bursts of activity do not show an evolution in frequency, amplitude, or configuration. They are of no pathologic significance.

Hereditary Fadors and the EEG Studies on twins have suggested a high correlation of EEG patterns within families. Moreover, Vogel has suggested that in family members other than twins, the

L--J 1 sec

[50~V

FIGURE 3·14 • Rhythmic sharpened theta activity seen bilaterally during drowsiness

in the EEG of a 16-year-old boy. Bursts of such activity are a normal variant (sometimes called psychomotor variant) without pathologic significance.

Electroencephalography: General Principles and Clinical Applications

low-voltage EEG pattern, fast alpha variant, and highvoltage monorhythmic alpha pattern all are inherited as autosomal dominant traits.v" Metrakos and Metrakos suggested that "centrencephalic" spike-wave EEG abnormalities were inherited as all autosomal dominant trait with age-dependent expressivity.'" They found that 37 percent of siblings of subjects with such activity had similar EEG abnormalities, as opposed to 9 percent among the siblings of control subjects. Others have found a lower frequency of spike-wave discharges, recording them in an average of 7 percent of siblings of subjects with this EEG abnormality and in 1.8 percent of siblings of control children."? The age distribution of positive findings in the siblings of patients with spike-wave discharges was clearly bimodal, being 12 to 13 percent in siblings between 3 and 6 years old, 1.4 percent in siblings 9 to 10 years old, and 10 percent in siblings 15 years old. These findings suggested that genetic influences were important for the development of spike-wave activity but that these influences were age-dependent and multifactorial. Photoparoxysmal responses occur with increased incidence in family members of individuals known to have such responses. In one study, an incidence of nearly 6 percent was found in a control population whereas a 23 percent incidence was found in siblings of children with photoparoxysmal responses.r" Their occurrence was age-dependent, and the age distribution showed a bimodal distribution among the female siblings, with the lowest incidence in 9- to 10-year-old children. The mode of inheritance is unclear. Rolandic spikes occur in patients with benign focal epilepsy of childhood (see Chapter 4) but are also found with increased frequency in the siblings or parents of these patients. Expression of this pattern is also age-dependent, with rolandic spikes found with the greatest incidence in subjects between 8 and 16 years of age.li~1 Other EEG patterns that have a familial incidel ICe are certain parietal theta and occipital delta rhythms, both of which occur as age-dependent phenomena in children.T:" Molecular studies to define further the basis of these findings remain to be undertaken.

EEG Changes with Aging A number of EEG changes occur with senescence, but the extent to which they occur varies widely in different subjects. It is widely believed that the mean alpha frequency slows in elderly subjects compared with young adults, but whether such slowing occurs in completely healthy and cognitively intact elderly subjects is unclear; if it docs, it is to a minimal degree. 72- 74 In subjects older than 50 years, alpha-like activity is sometimes

55

seen in one or both temporal regions and may be more conspicuous than the occipital alpha rhythm.?" The effect of aging on beta activity is less clear, and review of the published accounts is confounded by differences in the methods of analysis and recording circumstances, differences in the types of patients studied, and a failure to consider medical and drug histories and the state of intellectual function. No consistent effects on beta rhythms have been found in elderly subjects, although various alterations have been described by different authors." Diffuse theta and delta activity is significantly increased in elderly persons and is clearly related to intellectual deterioration and life expectancy." Polymorphic or rhythmic focal slow activity, encountered fairly commonly in the elderly, usually is localized to the left anterior temporal region. Such slow activity may be enhanced or brought out during drowsiness or hyperventilation. It is of little pathologic significance, showing no obvious correlation with life expectancy, neurologic disease, or intellectual changes. 73 . 76 It should not be regarded as abnormal if it occurs infrequently in bursts on an otherwise normal background; however, the bounds of normality are unclear. EEG sleep patterns also change with age, with a reduction in total sleep time and especially in the duration of stage 3 and 4 sleep.

EEG RESPONSES TO SIMPLE ACTIVATING PROCEDURES Hyperventilation The response to hyperventilation (p. 40) varies considerably in different subjects and is enhanced by hypoglycemia. Typically, there is a buildup of diffuse slow activity, first in the theta- and then in the delta-frequency ranges, with the activity settling in the 30 to 60 seconds after the conclusion of overbreathing. As indicated earlier, these changes are generally attributed to a fall in arterial PC0 2, which leads to cerebral vasoconstriction and thus to reduced cerebral blood flow; however, other mechanisms have also been suggested." The EEG response depends very much on the age of the subject, In normal children or in young adults it may be quite striking, with continuous or paroxysmal rhythmic, high-voltage delta activity coming to dominate the EEG record. By contrast, many elderly persons show little or no response. This lack of response has been attributed to an inability of these persons to alter their arterial PC0 2 and has been related to reduced cerebrovascular reactivity." Persistence of any response for an excessive period after hyperventilation has ceased is generally regarded as abnormal, as is an asymmetric response. However, individual variability in the response to hyperventilation can

56

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

make it difficult to evaluate the findings provoked by the maneuver. In some patients, hyperventilation leads to the reproduction of symptoms. The occurrence of symptoms should be noted by the technologist so that electroclinical correlations can be made. Spike-wave discharges can be provoked in patients with generalized epilepsy, especially in patients with absence seizures. Focal abnormalities may be enhanced in patients with partial epilepsy. Delta rhythm notched by faster activity (beta activity or muscle artifact) is an occasional nonspecific finding that may simulate spikewave discharges; however, it should not be taken as evidence of epilepsy. Review of the background activity usually permits the nature of this activity to be recognized. In patients with cerebrovascular disease, irregular theta and delta activity appears earlier and more conspicuously on the affected side than on the other, and in some instances repetitive paroxysmal discharges may be seen in a restricted distribution. In patients with tumors, hyperventilation may enhance the abnormalities seen on the resting record or provoke changes that are otherwise not apparent.

Photic Stimulation In response to photic stimulation (p. 40), it is usual to observe the so-called driving response over the posterior regions of the head. This response consists of rhythmic activity that is time-locked to the stimulus and has a frequency identical or harmonically related to that of the flickering light. Not ail normal subjects

exhibit a driving response, and many show it at only some stimulus frequencies. Moreover, a mild amplitude asymmetry is sometimes seen in normal subjects. Even when the amplitude difference between the two hemispheres is greater than 50 percent, an underlying structural lesion is unlikely unless other EEG abnormalities are present (e.g., a gross amplitude asymmetry of the alpha rhythm or the presence of focal slow waves or sharp transients). Focal hemispheric lesions may lead to an ipsilateral reduction in amplitude or, less commonly, to enhanced responses.F An asymmetry in the development of a driving response is more likely to be associated with focal slowing and other EEG abnormalities and with the presence of a structural lesion, although the site of the latter may vary markedly in differen t subjects. 78 Paroxysmal activity is sometimes found during photic stimulation, even when the EEG is otherwise normal. Polyspike discharges following each flash of light may be seen, particularly during stimulation while the eyes are closed. The discharges are muscular in origin; are most conspicuous frontally; stop when stimulation is discontinued; and may occur without clinical accompaniments, although in other cases a fluttering of the eyelids is seen. Activity of this sort is generally referred to as a photomyogenic (or photomyoclonic) response and is regarded as a normal response to highintensity light stimulation, being not uncommon in healthy persons. It must be distinguished from a photoparoxysmal response (Fig. 3-15). A photoparoxysmal response occurs in some (less than 5 percent) epileptic patients, particularly in those with generalized (myoclonic, tonic-clonic, or absence)

Fp1-F3------"...--~---"""""_t1\~Vl\

F3-C3 _~_~""'-"

........__J.i'_"'lv

Fp2-F4~-~~-------"1In\Jl\ F4-C4--~~""--~~~-..."

Ilhl/ill/I/IJUlII/lWI L--J 1 sec

300 fLV [

FIGURE J-15 IIlI Photoparoxysmal response recorded in a patient with epilepsy. The lowest trace records the light flashes. which were occurring at 12 Hz.

Electroencephalography: General Principles and Clinical Applications

seizures. Juvenile myoclonic epilepsy is closely associated with photosensitivity. An abnormal response occurs occasionally in patients with diverse CNS and metabolic disorders. Early reports of photoparoxysmal responses in alcohol and drug withdrawal syndromes are not supported by the study of Fisch and colleagues, who studied 49 subjects during acute alcohol withdrawal without finding a single instance of such a response.?? It may also occur on a familial basis in subjects without neurologic or metabolic disorders (p, 55). The response consists of bursts of slow-wave and spike or polyspike activity that has a discharge frequency unrelated to that of the flashing light and may outlast the stimulus. The activity is cerebral in origin and is usually generalized, bilaterally symmetric, and bisynchronous, although it may have a frontocentral emphasis. It is sometimes associated with clinical phenomena such as speech arrest, transient absence, or deviation of the eyes or head; if photic stimulation is continued, a generalized tonic-clonic seizure may result. In one study, responses outlasting the duration of photic stimulation were found to correlate much more closely with a clinical diagnosis of seizure disorder than when responses ended either as soon as stimulation was stopped or spontaneously despite continued stimulation. so However, another study failed to confirm any difference in the incidence of seizures in patients with photoparoxysmal responses outlasting or limited to the time of stimulation or stopping spontaneously.W'' Photoparoxysmal activity sometimes occurs with a more

57

restricted distribution in the occipital region, particularly in patients with an epileptogenic lesion in that area; such activity may be unilateral or bilateral.

Natural Sleep The electrophysiologic changes that occur during sleep are discussed in Chapter 32; only brief mention of them is made here. As the patient becomes drowsy (stage I), the alpha rhythm becomes attenuated and the EEG is characterized mainly by theta and beta rhythms. During light sleep (stage 2), theta activity becomes more conspicuous, and single or repetitive vertex sharp transient" occur spontaneously or in response to sensory stimuli. Bursts of high-voltage biphasic slow waves (K-complexes) also occur spontaneously or with arousal and are often associated with bursts of diffuse 12- to 14-Hz activity (sleep spindles). The spindles are usually most prominent in the central regions and may occur independently of the K-complexes. Positive occipital sharp transients (POSTs) may also occur spontaneously, either singly or (more commonly) repetitively in runs during this stage of sleep; they are bilaterally synchronous, but may be markedly asymmetric (Fig. 3-16). As sleep deepens, the EEG slows further until up to 50 to 60 percent (stage 3) or more (stage 4) of the record consists of irregular delta activity at 2 Hz or less; vertex sharp waves and spindles may also be found in stage 3 sleep. Stage 3 and 4 sleep is generally referred to as slow-wave sleep.

Fp1-F3 F3-C3 C3-P3 P3-01 Fp1-F7 F7-T3 T3-T5 T5-01

FIGURE ]·16 11II Runs of positive occipital sharp transients (POSTs) recorded during drowsiness in a subject with a normal EEG.

58

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

During REM sleep, the heart and respiratory rates become irregular, irregular eye movements occur, and the EEG comes to resemble that of stage 1 sleep. Focal cerebral lesions may lead to abnormalities of normal sleep activity, most often causing a voltage asymmetry of vertex sharp waves or sleep spindles. In some instances, such activity may be absent unilaterally. In rare instances, sleep spindles may show a frequency asymmetry, being slower on one side (usually that of the lesion) than the other." The EEG is sometimes recorded during sleep or following a period of sleep deprivation in the hope of activating abnormalities, especially when patients with suspected epilepsy are being evaluated. The incidence of epileptiform discharges is increased when the EEG is recorded during sleep as well as wakefulness, and the additional yield is greatest when patients with suspected complex partial seizures are being investigated. However, both the extent of this additional yield and the manner in which sleep recordings are best obtained in patients with suspected epilepsy are controversial. In particular, there is disagreement as to whether activation is best achieved with natural or drug-induced sleep. Nevertheless, in all patients with a suspected seizure disorder. the EEG probably should be recorded during both wakefulness and natural sleep because this may be helpful and involves no additional cost or risk. Slow or atypical spike-wave discharges are usually (but not always) more conspicuous during sleep." Blume and colleagues, for example, found that sleep augmented slow spike-wave activity in 12 cases, had no effect on it in 7, and diminished it in 2.85 Focal spike discharges can be enhanced or precipitated during any stage of non-REM sleep; but most interictal epileptiform discharges, whether focal or generalized, are usually seen less commonly during REM sleep. Classic 3-Hz spike-wave activity is seen more often but is less well organized during non-REM sleep than during wakefulness, whereas it occurs less frequently during REM sleep but is similar in form, rhythmicity, and regularity to its appearance in the waking record. Thus, recording during sleep can be useful, especially in patients with suspected complex partial seizures when the EEG recorded during wakefulness does not show focal or lateralized epileptiform activity. However, the interpretation of EEGs obtained in such circumstances can be complicated by difficulty in determining whether particular findings represent the normal electrical accompaniments of sleep or are of pathologic significance.

Sleep Deprivation Sleep deprivation is a harmless activating procedure and is therefore undertaken in many laboratories. The

early literature was reviewed by Ellingson and colleagues, who concluded that the preponderance of evidence suggested that sleep deprivation is effective as an activating procedure in patients with epilepsy.'" In one study of patients with possible epilepsy in whom the routine EEG during wakefulness was normal or only mildly abnormal, the yields of EEGs obtained during sedated sleep and following sleep deprivation were compared.f? The records of 44 percent of sleepdeprived patients were found to provide useful information (precipitation or enhancement of epileptiform discharges, presence of a new independent focus, or new type of epileptiform discharge), compared with 14 percent of records from patients in sedated sleep. Moreover, sleep deprivation was found to be significantly superior to sedated sleep in differentiating patients with a final clinical diagnosis of epilepsy from those in whom this diagnosis of epilepsy was rejected or very doubtful." Other, more recent studies have supported this view.88

EEG FINDINGS IN PATIENTS WITH NEUROLOGIC DISORDERS

Epilepsy Although the EEG is recorded from a very restricted portion of the brain and for only a limited time, it is an invaluable adjunct to the management of patients with epilepsy. The recording of electrocerebral activity during one of the patient's clinical attacks may be particularly helpful in determining whether the attacks are indeed epileptic in nature and whether they have a focal or lateralized origin. Because attacks usually occur unpredictably, however, the chances of a recording actually being in progress during an attack are not particularly good unless prolonged recordings are made or attacks are deliberately provoked. Moreover, even if an attack can be recorded, the EEG may be so obscured by muscle and movement artifact that little useful information can be gained from it. The interictal EEG is also abnormal in many epileptic patients and may exhibit features that help to establish the diagnosis. In this connection, the presence of paroxysmal activity consisting of spike, polyspike, or sharp-wave discharges, either alone or in association with slow waves, is of prime importance. Such epileptiform activity may be focal, multifocal, or diffuse and may appear unilaterally or bilaterally; if bilateral, it may be synchronous or asynchronous and symmetric or asymmetric. When multiple foci are present, one of them may be the primary one, generating a mirror focus in the homologous region of the contralateral hemisphere, although the development of such mirror

Electroencephalography: General Principles and Clinical Applications

foci in humans is not well documented. Alternatively, a single deep focus may discharge to homologous regions of the cortex so that the EEG reveals either a focus that shifts from side to side or bilaterally synchronous discharges. In many patients with multiple foci, however, the foci are distinct from each other, with discharges arising from them asynchronously. In one study of patients with such independent multifocal spike discharges in the EEG, most had extensive bilateral cerebral lesions and many experienced clinical seizures of more than one type. R!! Epileptiform activity is occasionally found in subjects who have never experienced a seizure. Zivin and Marsan'? found it in the initial EEG of 2.2 percent of 6,497 nonepileptic patients, and Goodin and Aminoff found it in 4 percent of 948 patients without seizure disorders" Accordingly, the presence of epileptiform activity, in itself, does not establish the diagnosis of epilepsy. Among patients with epilepsy, the initial EEG contains epileptiform activity in about 50 to 55 percent of cases. Because the incidence of epilepsy in the general population is only 0.5 percent, however, epileptiform activity is actually more likely to be encountered in EEGs from nonepileptics than in those from epileptics if recordings are made from an unselected group of subjects. By contrast, among patients with episodic behavioral or cerebral disturbances that could well be epileptic in nature, the presence of epileptiform activity in the EEG markedly increases the likelihood that epilepsy is the correct diagnosis.I" The presence of epileptiform activity can therefore be very helpful in establishing the diagnosis of epilepsy beyond reasonable doubt, depending on the clinical context in which the EEG is obtained.l" The absence of such activity cannot be taken to exclude this diagnosis. Several factors (e.g., the age of the patient and the type and frequency of seizures) bear on the presence or absence of epileptiform activity in the EEG of patients with undoubted epilepsy on clinical grounds. Some of these factors should influence the manner in which the EH; is obtained. First, epileptiform activity is more likely to be found if repeated records are obtained. Second, the diagnostic yield is increased by the routine use of activation procedures (e.g., hyperventilation, photic stimulation, sleep, and sleep deprivation for 24 hours; pp. 55-58) to precipitate epileptiform discharges. Third, the timing of the examination may influence the yield. Epileptiform activity is found more commonly if the EEG is recorded soon after (particularlv within 24 hours of) a clinical seizure.P''?' If seizure frequency is influenced by external or situational factors (e.g., the menstrual period), the EEG should be scheduled with these factors in mind. In patients with one of the reflex epilepsies, it may be possible to reproduce the provocative stimulus while the EEG is recorded.

59

King and associates examined the possibility of diagnosing specific epilepsy syndromes by clinical evaluation, EEG, and MRI in 300 patients who presented with a first unexplained seizure.?" They were able to diagnose a generalized or partial epilepsy syndrome on clinical grounds in 141 (47 percent) of patients, and subsequent analysis showed that only 3 of these diagnoses were incorrect. When the EEG data were also utilized, an epilepsy syndrome was diagnosed in 232 patients (77 percent). Neuroimaging revealed 38 epileptogenic lesions; there were no lesions in patients with generalized epilepsy confirmed by EEG. The EEG recorded within the first 24 hours was more useful in detecting epileptiform abnormalities than a later EEG (51 percent vs. 34 percent). This study emphasizes the utility of the EEG in evaluating patients with a single seizure for determining the precise diagnosis and thereby the prognosis and need for treatment. In reviewing the EEG findings in patients with seizure disorders, attention has been confined to the more common types of seizures encountered in clinical practice. PRIMARY (IDIOPATHIC) GENERALIZED EPILEPSY

The background activity of the interictal record is usually relatively normal, although some posterior slow (theta and delta) activity may be present in patients with absence (petit mal) seizures. Generalized, bilaterally symmetric, and bisynchronous paroxysmal epileptiform activity is often seen, especially during activation procedures such as hyperventilation or photic stimulation. Absence (petit mal) seizures. In patients with absence seizures, epileptiform activity consists of well-organized 2.5- to 3-Hz spike-wave discharges (Fig. 3-17) that may be seen both interictally (especially during hyperventilation, which increases the discharges in about 75 percent of patients, or with hypoglycemia) and ictally. The presence of observed clinical accompaniments depends on the duration of the discharges and the manner in which the patient's clinical status is evaluated. With sufficiently sensitive techniques, it may well be found that this type of epileptiform activity is generally associated with behavioral disturbances (i.e., is usually an ictal phenomenon). For example, auditory reaction time is commonly delayed during the first 2 seconds of a generalized spike-wave discharge even when there is no clinically obvious impairment of external awareness. The characteristics of the spike-wave complexes are best defined in referential derivations. They are often maximal in amplitude in the frontocentral regions, and careful analysis has shown that the spike has several components: it is not a simple, monophasic phenomenon." The frequency of the complexes is often a little faster than 3 Hz at onset and tends to slow to about 2 Hz before terminating. Sleep may influence the mor-

60

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp1-F3 ~

F3-C3

C3-P3~JVIN'lrYYYi('('''''''''''''''''''''~~-~--

P3-01 ~hN*",..".",..~~~\hI\f.J{'\I\-.Mt""v/W~'f-'W-_~Fp2-F4 F4-C4-----...."..-I\l\r'Il\I\JVV1lrur"Wmfilll/W'lvllN'lIY\ai''{\iW.......-------~­ C4-P4-~__-""'t"'....yr~"'""'"~~""""'.A+A1'--~~~-"\/"-------

P4-02 --,.,.,..."r-t4vtI~~\MN+!'M'Y'Y~~M.rtI'f«f\NN"'tJ"I'I'Ir<"~...,.....-~~-..,.,. Fp1-F7 "---v"'\ .---tV1I11J'\1\,II.I\I\N'v'V\r1...v"l,Mr'\r'\f\l\fll \t"J"INlI'II F7-T3 -----...-~WiV'hr-N"m#Mrfffll,H\Jr-N'dY.N\f'v----~_T3-T5 _ ........-_...-.O/\N1I11Jli\MI"Vor"fl'V'''''''''¥V''''V~MI.,,/''''''\rV'V'.J-..--..----­ T5-01 - - -_ _'II"'"1''V\Ar'lt1r\Nvryv..'"'''J''V''vV-(>..r>,'''If''f/W\.II;yv'''''''I(V..,....,...,.,...,~-.-,..,..,.,...._~ Fp2-F8 F8-T4 -----4~fflrri'"V'N"'N'"Mrff'''f'''t",N1\.f''t-('f.,M,....,...,,-------

T4-T6

---~'-"I'\/1¥I\MrtN~VVI!V"'~VV'..,....,...,~..."..".,~,,-------------

T6-02

--_...,.,...'V""I'fwoII\v..rv"II'VI,Mt"'lV"II'V>.~"""""'IV''''''''''''-.r"'''''''''.."."".....-.......------_ ............

1~

450 J.LV[

FIGURE J-17 III Paroxysmal, generalized, bilaterally synchronous and symmetric 2.5- to 3-Hz spike-wave activity recorded interictally in the EEG of a patient with absence (petit mal) seizures.

phology of the complexes (p, 58). In some instances, the discharges cease with eye-opening or in response to alerting. There may be inconstantly lateralized amplitude asymmetries or inconsistent asynchronous onsets over the two hemispheres when the recordings on different occasions are compared, but such findings are of no clinical consequence. The background between the generalized bursts of activity occasionally contains isolated spike discharges that may be focal (usually, frontal) in distribution'"; this finding is also without special significance. Similarly, high-voltage intermittent rhythmic delta activity may be found occipitally, especially in children during the first decade of life. EEG findings that are repeatedly normal in a child with clinical attacks resembling absence seizures suggest that a diagnosis of primary generalized epilepsy is incorrect. Absence epilepsy of childhood onset usually presents before the age of 12 years and commonly remits in late adolescence. By contrast, juvenile absence epilepsy presents in the second decade and is associated with less frequent absences but a greater likelihood that seizures will continue into adulthood and that generalized tonic-clonic convulsions will occur. Further details of these disorders are provided in Chapter 4. The absences that occur in juvenile myoclonic epilepsy are associated with spike- or polyspike-wave activity that has a frequency of 2.5 to 4 Hz (with a range of 2 to 7 Hz) and that is enhanced by hyperventilation; there is much intra- and interdischarge variation, and discharges may be interrupted for brief periods.P'

When absence attacks continue to occur in adults, they are often difficult to recognize because of their short duration. 95,96 Such patients usually also have generalized tonic-clonic seizures. Michelucci and colleagues found in adults that the interictal EEG background is normal but may be interrupted by generalized bursts of spike-wave or polyspike-wave discharges and, occasionally, by independent focal spikes; runs of polyspikes sometimes occur during non-REM sleep." Ictally, 3-Hz spike-wave discharges are characteristic, sometimes preceded by polyspikes with or without intermixed spikewave or polyspike-wave discharges. Similar findings have been reported by others.P" Myoclonic seizures. Bursts of generalized spike-wave or polyspike-wave activity are found both ictally and interictally, often with a frequency of 3 to 5 Hz and an anterior emphasis; they may be enhanced especially by photic stimulation or sleep. In juvenile myoclonic epilepsy, discussed in detail in Chapter 4, the interictal EEG usually has a normal background; however, this is interrupted by bursts of generalized, frontally predominant polyspikes or of polyspike-wave or spike-wave discharges having a variable frequency that is often between 3 and 5 Hz but may be higher. Photic stimulation provokes or enhances such abnormalities in 25 to 30 percent of cases, and hyperventilation also activates the epileptiform abnormalities. Focal or lateralizing features or shifting asymmetries are sometimes found. Ictally, the myoclonic seizures are associated with anteriorly dominant polyspike or

Electroencephalography: General Principles and Clinical Applications

polyspike-wave bursts followed by rhythmic delta activity. Absence seizures in this syndrome are accompanied by irregular spike-wave or polyspikewave discharges of variable frequency.P" Tonic-clonic (grand mal) seizures. Generalized, bilaterally synchronous spike discharges, or bursts of spikewave or polyspike-wave activity, or both, may be seen interictally (Fig. 3-18), the latter sometimes being identical to that found in absence attacks. The earliest change during a tonic-clonic convulsion is often the appearance of generalized, low-voltage fast activity. This activity then becomes slower, more conspicuous, and more extensive in distribution and, depending on the recording technique, may take the form of multiple spike or repetitive sharp-wave discharges that have a frequency of about 10 Hz and are seen during the tonic phase of the attack. In other instances, seizure activity may be initiated by a flattening (desynchronization) of electrocerebral activity or by paroxysmal activity such as that which occurs in the interictal period. In any event, as the seizure continues into the clonic phase, a buildup of slow waves occurs and the EEG comes to be characterized by slow activity with associated spike or polyspike discharges. Following the attack, the EEG may revert to its preictal state, although there is usually a transient attenuation of electrocerebral activity, followed by the appearance of irregular polymorphic slow activity that may persist for several hours or even longer. In a few cases, the postictal EEG is characterized by periodic complexes.

F~:=:

P4-A2

SECONDARY (SYMPTOMATIC) GENERALIZED EPILEPSY

In patients with secondary generalized epilepsy (i.e., with generalized seizures relating to known pathology, such as Lennox-Gastaut syndrome), the background activity of the interictal EEG is usually abnormal, containing an excess of diffuse theta or delta activity that is poorly responsive to eye-opening and may show a focal or lateralized emphasis. Single or independent multiple spike discharges may also be found.I" However, interictal paroxysmal activity characteristically consists of slow spike-wave or polyspike-wave discharges that are usually less well organized than in the primary generalized epilepsies, with a frequency that varies between 1.5 and 2.5 Hz (Fig. 3-19). This activity exhibits a characteristic variability in frequency, amplitude, morphology, and distribution on different occasions, even during the same examination." In some instances it is generalized and exhibits bilateral symmetry and synchrony; in others it is markedly asymmetric, showing a clear emphasis over one hemisphere or even over a discrete portion of that hemisphere. Hyperventilation and flicker stimulation are generally less effective as activating agents than in the primary generalized epilepsies, but non-REM sleep can increase the frequency of the paroxysmal discharges.v' Such paroxysmal activity may also occur as an ictal phenomenon accompanying absence and myoclonic attacks. The most common EEG changes associated

,

01-A1 -

.........-

........N¥.~~l\-(\""""'MN'""""N....,....;-y~--"V\i~""""'--_.....mA

F7-A1 ----(''\I¥I''''''ItVlN''''fV''''J'.1 "\I" T3-A1 T5-A1 .......,.."""...,.,.*""'f\J\~,...."-If'-o,....,...........,...,,-v-'f/'_..~_-'V'J'...,...,-w...--.,,.--'W'M L.-...J

1

sec

600 IJ.V [

III Paroxysmal, generalized, bilaterallysynchronous spike-wave and polyspike-wave discharges seen interictally in the EEG of a 62-year-old woman with tonic-clonic seizures caused by primary generalized epilepsy.

FIGURE 3-18

61

62

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp1-F3~

F3-C3 C3-P3 "-''--J'--'''''''....JI,...J-..,J'vV.......... P3-01

"""~_..N'-.I
..... Ar...r-"'""'_-JV'-l\r-..Jv...~.....A _ _

r-'"...,......,-.~...ANIr-Ni\"'.jV",...,.....-"'V'"-II'-J\I"'-v.f-...NlN.r-N-~--"---N'--./\Hir"'v-"--J\.,-...~

Fp2-F4 .,....J.........~-'lrV\I"\N\NV F4-C4 ..-"-" '-'>JV'\J\f'\~r_.III(ll" C4-P4 "-"'.-../'o_.....,..,...JIrV....."f-w-.....I......" """.....-.Y'V'-JIrv.f.....'"'fIr-JIf'vJ-""-'"-J'-"-------f\..../V'-..~~""'_""'---'''''__ P4-02 .--.,.~--v-.../I.r--.r_JV-llr--JV-"'"'-""'"''''''''
1~

600 fL V [

FIGURE J-19 • Interictal 2-Hz spike-wave activity in the EEG of 17-year-old boy with Lennox-Gastaut syndrome.

with tonicseizures are relative or total desynchronization (flattening), with sudden disappearance of any ongoing spike-wave activity, or the development of rhythmic activity (usually at about 10 Hz, or faster) that usually increases in amplitude, has a generalized distribution, and is followed by irregular slow waves (Fig. 3-20). In some instances, an initial desynchronization of the traces is followed by this pattern of generalized rhythmic fast activity. The ictal EEG features of tonic-clonic seizures are as described for patients with primary generalized epilepsy. Atonic (astatic) seizures are associated with polyspike-wave discharges, an attenuation of the background, or a brief burst of generalized fast activity or spikes. Generalized atypical spike-wave discharges have a more irregular appearance than that of 3-Hz or slow spikewave activity, and range in frequency between 2 and 5 Hz. They occur ictally or interictally, may be activated by sleep and sometimes by photic stimulation, and have been associated with various types of generalized seizures. In patients with myoclonus epilepsy the background is diffusely slowed and bursts of spikes, polyspikes or sharp waves, or spike-wave discharges are seen either

unilaterally or bilaterally. Flicker stimulation may lead to bursts of spike- or polyspike-wave activity. The spike activity is sometimes time-locked to the myoclonic jerking that the patient exhibits. PARTIAL EPILEPSY

Several varieties of so-called benign partial epilepsy have been described in children. These include benign rolandic, occipital, or frontal epilepsy; benign partial epilepsy of adolescence; and benign epilepsy associated with multiple spike foci. These are considered in Chapter 4. The interictal EEG findings may vary considerably at different times in patients with partial (focal) epilepsy, especially in those with complex symptomatology. The interictal record (see Fig. 3-11) may exhibit intermittent focal sharp waves or spikes, continuous or paroxysmal focal slow-wave activity, localized paroxysmal spikewave discharges, or any combination of these features. There is some correlation between the site of epileptiform discharges and the clinical manifestations of seizure phenomena. For example, among patients with epilepsy, complex partial seizures are more likely in

F3-C3 -~-----~~~~N~~~~~'\.I C3-P3 .........~~-~.~~~~ • •/II.~~~ P3-01 F4-C4 C4-P4 ...-...-..----'-----.......-J'*~II._i~IfI\I.,JII~~~~~~ P4-02 --...J--~...wIIt'4l~I'VtI\.IfIy)"'i\.MrWltIf\,fN"''''''ft./\

T3-Cz Cz-T4

1;;;

400 fL V[

FIGURE J-20 • Tonic seizure recorded in the EEG of an l l-year-old girl with Lennox-Gastaut

syndrome.

Electroencephalography: General Principles and Clinical Applications

those with interictal anterior temporal spike discharges, visual hallucinations are more common in those with occipital spikes, and hemifacial seizures occur more often in those with centrotemporal (rolandic) spikes. III patients with temporal lobe abnormalities. spike discharges may be confined to one hemisphere, may be found either independently or synchronously over both hemispheres, or may occur in the homologous region of hoth hemispheres with a consistent temporal relationship between the spikes. In the latter case it is presumed that the discharge originates from one side and spreads to the other via commissural connections (e.g., the corpus callosum) to produce a so-called mirror focus that eventually becomes independent of the original focus, although the evidence to support this concept is incomplete in humans. Focal epileptiform discharges may be restricted to the vertex electrodes, in which case they can be hard to distinguish from the normal vertex sharp activity seen during sleep. The occurrence of such discharges during- wakefulness, the presence of associated focal slowing. or a tendency for the discharges to lateralize to one side of the midline would suggest that they are abnormal, as also would fields of distribution that differ from those of normal vertex sharp waves." Bilaterally synchronous spike-wave activity is an occasional finding in patients with partial seizures. Unless this activity is markedly asymmetric, has a focal onset, or is preceded by focal or lateralized discharges, it can be difficult to distinguish from the similar activity seen in patients with primary generalized epilepsy. Interictal abnormalities may be provoked by such harmless activation procedures as hyperventilation, sleep, or sleep deprivation (pp. 55-58), but flicker stimulation is not helpful in most patients with partial epilepsy. When these abnormalities arise in the anterior temporal region they may be recorded only from nasopharyngeal, sphenoidal, or other accessory electrodes (p. 39). The site of focal interictal abnormalities does not necessarily correspond to the region in which

63

seizures usually originate, however, and this must be kept in mind when patients are being evaluated (see Chapter 7). Moreover, the possibility of an underlying structural abnormality (e.g., a tumor) must be remembered in patients with partial seizures. The presence of a focal polymorphic slow-wave disturbance may reflect such a lesion, and the significance of this finding should be clearly indicated to referring physicians. A repeat examination after an appropriate interval will help to exclude the common alternative possibility that the abnormality is postictal in nature. The presence of independent bilateral spike discharges makes a unilateral structural lesion unlikely. In some patients with partial seizures, and especially those with elementary symptomatology, the scalprecorded EEG shows no change during the ictal event."? More commonly, however, the EEG shows localized discharges (Figs. 3-21 and 3-22) or more diffuse changes during the ictal period. In some patients, the ictal EEG shows a transient initial desynchronization, which is either localized or diffuse and is followed by synchronous fast activity; in others, the initial flattening of the traces does not occur. In still other cases, rhythmic activity of variable frequency with or without associated sharp transients is seen during a seizure. The ictal discharge is often succeeded by a transient flattening of the traces and then by slow activity that may have a localized distribution. If focal slow activity is present, the EEG examination may have to be repeated after several days to distinguish postictal slowing from that caused by a structural lesion. For practical purposes, temporal lobe epilepsy is divided into medial and neocortical (lateral) temporal lobe epilepsy, and different ictal EEG patterns have been reported, depending on the site at which seizures arise. In seizures arising medially, lateralized rhythmic activity is typically maximal over the ipsilateral temporallobe, whereas for seizures arising laterally such ictal activity often has a hemispheric distribution and a lower maximal or irregular frequency.'Y" In some

F3-C3 An.11J,.i..fo,1'vv."""""-...........---~/oJrItIII~~_ C3-P3~~~ ......-""""""'...........M~W• •~~ P3-01 J F4-C4 ~~.,.J\,.,.,--"~'Ir--"\~rvy"'Yl'1'J/1.."""'(VV'(I C4-P4 ~~~~.--~...,.."....-.-. .........NN"'-'yrI'v. P4-02 '-----'

1 sec

200/l.V[

FIGURE ]-JI • Partial seizure recorded in the EEGof a 12-year-old girl. The seizure commences with rhythmic left-sided fast activity Ih~H increases in amplitude and becomes intermixed with slower elements, with some spread to the contralateral hemisphere. Slow waves with some associated spike discharges occur in the later part of the seizure. This activity is followed by lower voltage, diffuse slow activity with a left-sided emphasis. Clinically. the seizure was characterized by clonic jerking of the side of the face and the limbs on the right side, followed by unresponsiveness as slow activity became conspicuous in the later part of the trace.

64

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

Fp1-F7

----"',""""-....-.......---.-.,..-....

F7-T3

~~....".,........~""""I/f+N

.....

,.".,.,~~....., ,r..."'~~~._...,.., ..~~~~

T3-T5 "'+v~~,~!"W1o.""..tA.1vI~AM~....-~"""',~~ T5-01 ~ 1 A•.-I""oMo.....,..."""""~~...,..,~Iffl..J1~~'-~

F8-T4~y~\)~~,~~I1~~~~

T4-T6~~~~JII\~N/~.4~~~~"'*" T6-02~

/"It--A~~~~~~~~~~~

L....J

200 fJ-V[ 1 sec FIGURE 3·22 • EEG of a 13-year-old boy with frequent attacks of "dimming out." A burst of repetitive spikes is seen to occur in the right temporal region; during this time he experienced an attack.

(13 percent) patients with medial temporal lobe epilepsy, the electrographic seizure recorded over the scalp is contralateral to the side of seizure origin as determined by depth recordings and curative surgery; lateralized postictal slowing, however, is a reliable lateralizing finding. roo When a partial seizure becomes generalized as a secondary phenomenon, patients usually go on to experience a tonic-clonic convulsion. Focal EEG abnormalities are then replaced by the more diffuse changes occurring during such seizures. ANTICONVULSANT DRUGS, SEIZURE CONTROL,

AND THE

EEG

The EEG findings are influenced by antiseizure medication. An analysis of 23 controlled trials with therapeutic drug monitoring and serial EEG observations was made by Schmidt.l'" Suppression of "paroxysmal" discharges correlated well with an increase in the plasma concentrations of diazepam, phenobarbital, or phenytoin, either alone or in combination with phenobarbital or primidone. A more limited correlation was reported for carbamazepine and in patients with focal discharges receiving sodium valproate. Epileptiform discharges sometimes became more profuse with carbamazepine despite clinical improvement. An increase in beta activity correlated with increased plasma concentrations of clonazepam, phenytoin, and phenobarbital. Background slowing occurred with high plasma concentrations of diazepam or phenytoin, either alone or in combination with phenobarbital or primidone. Slowing of background activity was usually associated with clinical evidence of toxicity to carbamazepine, phenytoin, phenobarbital, or primidone.

The effects of newer anticonvulsant agents on the EEG are less clear. Lamotrigine reduces the frequency and duration of electrographic seizures and diminishes interictal epileptiform discharges without affecting background activity.102 Ictal and interictal EEG abnormalities are reduced or unaffected by vigabatrin independently of seizure control, which is improved by the drug in most patients, and the EEG background may be slowed. 103 The extent to which the suppression of interictal epileptiform discharges reflects the degree of seizure control is variable. In patients with clinical absences associated with generalized 3-Hz spike-wave activity, a good correlation usually exists between seizure control by pharmacologic agents and the amount of interictal epileptiform activity, but the relationship is otherwise more complex. Intravenous phenytoin and benzodiazepines acutely suppress both clinical seizures and interictal epileptiform activity, but in other contexts any correlation of the clinical and EEG effects of anticonvulsant drugs is hard to define.l'"

STATUS EPILEPTICUS

Tonic-Clonic Status

The EEG findings during the seizures may consist of diffuse repetitive spikes, sharp waves, or spike-wave complexes. Alternatively, the findings may resemble those seen during a single seizure except that the electrographic seizures occur repetitively and there is usually no transient postictal attenuation of the EEG. If seizures have a focal onset, the EEG may show focal spikes, which then spread diffusely. The interictal background is usually, but not always, abnormal, with an excess of irregular, asynchronous, diffuse slow activity, often with intermixed

Electroencephalography: General Principles and Clinical Applications

spikes and spike-wave complexes. The main value of recording the EEG during convulsive status epilepticus is to determine whether seizures are continuing when any motor manifestations are subtle because of associated brain damage,44,105 the duration of the epileptic state, or pharmacologic neuromuscular blockade. The EEG is also useful in monitoring treatment that involves induction of anesthesia, being important as a means of ensuring that adequate doses are given, and helps in determining whether involuntary motor activity that emerges when medication is being withdrawn relates to recurrence of seizures or is nonepileptic in nature. Nonconvulsive Status Epilepticus

This is suggested clinically by the occurrence of a fluctuating abnormal mental status with confusion, reduced responsiveness, lethargy, somnolence, automatisms, inappropriate behavior, or some combination of these and other symptoms. In patients without a history of prior seizures the epileptic basis of symptoms is often not recognized initially, and a primary psychiatric disorder is suspected until an EEG is obtained. EEG monitoring has shown that nonconvulsive status epilepticus occurs more commonly than previously appreciated in the neurologic intensive care unit, and that it is also a common sequela to convulsive status epileptic us but will generally not be recognized in the absence of an EEG. 106-1l~1 Nonconvulsive status epilepticus has been subdivided into absence status and complex partial status. These may resemble each other clinically, but their EEG features are quite distinct. The EEG is therefore important in establishing the diagnosis, distinguishing between the different types of nonconvulsive status, and monitoring the response to treatment. In absence status, which may occur in either children or adults, the EEG shows diffuse, bilaterally synchronous, continuous or discontinuous spike-wave and polyspike-wave discharges occurring at about 2 or 3 Hz. Less commonly, irregular arrhythmic spike-wave or polyspike-wave complexes may occur on a diffusely slowed background. By contrast, in complex partial status epilepticus, the EEG generally shows focal or lateralized electrographic seizures, although they are sometimes more generalized. Thus, paroxysmal activity (e.g., repetitive spikes, spike-wave discharges, slow waves, or beta rhythm) may be found unilaterally, bilaterally, or on alternating sides. Myoclonic Status Epilepticus

The most common cause of this disorder is anoxic encephalopathy. The EEG may show generalized periodic or pseudoperiodic complexes (usually consisting of spikes or sharp waves), often with an attenuated background between the complexes or a burst-suppression pattern. 46 ,110 In some patients, no epileptiform activity is seen.l!"

65

Status Epilepticus and Sleep

In rare instances, an EEG appearance suggesting status epilepticus may occur during sleep. Such an EEG consists of almost continuous, generalized spike-wave activity that is present primarily during non-REM sleep, which may make it impossible to recognize the normal sleep architecture. Epilepsia Partialis Continua

This may be regarded as a type of partial status epilepticus and is characterized clinically by rhythmic clonic movements of a group of muscles (Fig. 3-23). Thomas and co-workers reported the EEG findings in 32 patients with this disorder, in 28 of whom at least one recording was made during the ictal activity. I II In two patients, no abnormalities were detected. Other researchers have also reported that the EEG may be normal. Epileptiform abnormalities may have been obscured, however, by the background rhythms or by muscle or movement artifact; alternatively, these abnormalities may not have been detected because they arose from infolded regions of the cortex and were therefore not picked up by the recording scalp electrodes. Focal EEG abnormalities, which were sometimes enhanced by hyperventilation, were present in 21 patients and consisted of abnormal slow waves in 19 and/or sharp transients in 20. It was sometimes possible to identify a relationship between the EEG and muscle activity when the frequency of the jerking was slow. In 25 patients, the background activity was abnormally slow or asymmetric, and in 4 patients the EEG showed PLEDs.

LONG-TERM EEG MONITORING FOR EPILEPSY

Long-term EEG recordings are generally performed to distinguish between attacks that are epileptic or psychogenic in nature, to determine the frequency of seizure activity, and to localize an epileptogenic source. Monitoring procedures are considered in detail in Chapter 5. Ambulatory cassette recording is a useful long-term monitoring procedure that is relatively inexpensive because it does not require hospitalization; this type of recording is considered separately in Chapter 6. VALUE OF THE EEG IN THE MANAGEMENT OF PATIENTS WITH EPILEPSY

In addition to the aforementioned help that the EEG findings provide in supporting a clinical diagnosis of epilepsy, in permitting a precise epilepsy syndrome to be identified, and in excluding an underlying structural cause of the seizures, the findings are useful in a number of other ways with regard to the management of patients with epilepsy. They are an important aid to the clinician who is attempting to classify the seizure disor-

66

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp1-F3 F3-C3

L---J

1 sec

[150 IlV

FIGURE J-U • EEG of 46-year-old woman with epilepsia partialis continua. Sharp waves in the EEG can be seen arising from the left central region and were accompanied by repetitive jerking of the right arm. The seizures were found to result from neurosarcoidosis.

del' of individual patients and may therefore influence the choice of anticonvulsant drugs that are prescribed. For example, unless the EEG findings are taken into account it may be difficult to distinguish between certain types of partial seizures, absence seizures, and the so-called atypical absences that occur in patients with Lennox-Gastaut syndrome. The EEG may enable a focal or lateralized epileptogenic source to be identified, even when it is not apparent on clinical grounds: This identification may be of prime importance when the etiology or surgical treatment of the disorder is under consideration. Further comment on this aspect is made in Chapters 5 and 7. The EEG is particularly helpful in the diagnosis of absence and complex partial status epilepticus; indeed, it is often the only means that permits these diagnoses to be made with any confidence. In tonic-clonic (grand mal) status epilepticus, EEG monitoring is invaluable for monitoring the level of induced anesthesia and for determining whether seizure activity is continuing, as discussed earlier. The EEG can provide a limited guide to the prognosis of patients with epilepsy. Thus, an EEG that is normal and remains so following the standard activation procedures, discussed earlier, generally implies a more favorable prognosis than otherwise, although this prognosis is not always supported by the outcome in individual cases. In patients with absence attacks, 1- to 2-Hz spikewave discharges suggest a poorer prognosis than bilaterally symmetric and bisynchronous 3-Hz spike-wave activity. The presence of abnormal background activity also implies a poor prognosis.U'' The EEG findings have, however, proved to be disappointingly unreliable as a means of determining the prognosis for the subsequent development of seizures in children experiencing their first febrile convulsion or in patients with head injuries.

Emphasis is sometimes placed on the EEG findings when the feasibility of withdrawing anticonvulsant drugs is under consideration after epileptic patients have been free from seizures for some years. However, the EEG provides no more than a guide to the outlook in such circumstances, and patients can certainly have further attacks after withdrawal of medication despite a normal EEG or, conversely, can remain seizure-free despite a continuing EEG disturbance. The difficulty in obtaining reliable, definitive data concerning the utility of the EEG in this circumstance was underscored in the metaanalysis of Berg and Shinnar.U'' In patients with partial epilepsy, a more recent study suggests that the risk of relapse after drug withdrawal is not predicted by the findings in EEGs recorded before the beginning of drug withdrawal; by contrast, the changes that occur in the EEG recorded during drug withdrawal or in the 3 years thereafter do have prognostic value.!" Therefore, decisions concerning withdrawal of medication must be based on clinical grounds, with the context of individual cases as well as the EEG findings taken into account.

Syncope Loss of consciousness may occur because of diffuse cerebral hypoxia or ischemia. The EEG during a syncopal episode is generally said to show a diffusely slowed background initially and then high-voltage delta activity, usually after an interval of about 10 seconds or so, and followed sometimes by transient electrocerebral silence if the hypoperfusion persists. Recovery occurs in the reverse order. In fact, the EEG changes are more variable; in some instances there is no attenuation of electrocerebral activity, whereas in other instances electrocerebral quiescence develops with little or no preceding change

Electroencephalography: General Principles and Clinical Applications

in the frequency of background rhythms. I 1~,116 The precise temporal relationship of the EEG changes to the loss of consciousness is also variable, II~ depending in part on the cause of the syncope.!'?

Infedions KEG changes are generally more marked in patients with encephalitis than in patients with uncomplicated meningitis. In most of the acute meningitides and viral encephalitides, the EEG is characterized by diffuse rhythmic or arrhythmic slow activity, although focal abnormalities are sometimes found as well. The degree of slowing reflects, at least in part, the severity and extent of disease, the level of consciousness, and any metabolic and systemic changes. These EEG findings are not really helpful for diagnostic purposes, although they may be useful in following the course of the disordel; especially if the clinical features are relatively inconspicuous. Persistence or progression of the EEG abnormalities suggests advancing disease, complications, or residual brain damage, but clinical and EEG improvement do not necessarily have the same time course. The presence of focal abnormalities raises the possibility that an abscess is developing, although localized electrical abnormalities may also arise for other reasons (e.g., secondary vascular changes or scarring). Residual neurologic deficits, or complications such as a seizure disorder, may develop even if the EEG returns to normal after the acute phase of the illness. In the chronic meningitides, the EEG may show little change, although in other instances it may show diffuse' slowing. Similarly, the findings in aseptic meningitis are generally normal or consist of mild diffuse slow activity, usually in the theta-frequency range. Such findings are sometimes helpful in following the course of the disorder. The findings in patients with the encephalopathy associated with infection by human immunodeficiency virus (HIV) are discussed on page 74. An encephalopathy may occur in patients with systemic sepsis but no evidence of intracranial infection or other identifiable cause for its occurrence; it is probably multifactorial in etiology. The EEG is more sensitive than clinical evaluation for detecting such sepsis-associated encephalopathy. It becomes increasingly slowed, may show triphasic waves, and-in some cases-exhibits a hurst-suppression pattern or becomes increasingly suppressed. I IS Recovery may occur, however, even when a burst-suppression pattern is present. lIS HERPES SIMPLEX ENCEPHALITIS

The EEG findings in herpes simplex encephalitis are sometimes characteristic, especially if serial recordings

67

are made. Initially, diffuse slow activity is found, often with a lateralized or focal emphasis, especially over the temporal lobe on the affected side. Subsequently, stereotyped repetitive sharp- or slow-wave complexes, with a duration of up to about 1 second, may come to be superimposed on the slow background, usually developing between the 2nd and 15th days of the illness but occasionally not arising until even later (Fig. 3-24). These complexes may be found over one or both hemispheres, particularly in the temporal regions, and occur with a regular repetition rate, the interval between successive complexes usually being between 1 and 4 seconds in different patients. When the complexes are bilateral, they may occur synchronously or independently on the two sides. Such bilateral involvement generally implies a more serious outlook than otherwise. The occurrence of focal electrographic seizure activity may temporarily obliterate the periodic activity on the involved side. In time, the amplitude of the repetitive complexes becomes less conspicuous until they can no longer be recognized, the EEG showing instead a focal slow-wave disturbance; attenuation of activity over the affected region; or, in cases that have a fatal outcome, a progressive reduction in frequency and amplitude of background activity. These changes, when found during the course of an acute cerebral illness, may have considerable diagnostic value in suggesting the possibility of herpes simplex encephalitis.'!" Periodic complexes are not, however, an invariable finding in herpes simplex encephalitis, and their absence does not exclude the possibility of this disorder. Moreover, similar activity has been reported in rare cases in infectious mononucleosis encephali tis.5 1 SUBACUTE SCLEROSING PANENCEPHALITIS

Recurrent slow-wave complexes, sometimes with associated sharp transients, occur with a regular periodicity in this disease (Fig. 3-25). The complexes usually last for up to 3 seconds, but their form may show considerable variation in different patients or in the same patients at different times, as may the interval (usually 4 to 14 seconds) between successive complexes. In most instances, they have a generalized distribution and occur simultaneously over the two hemispheres, but in the early stages of the illness they are sometimes more conspicuous over one side. Sleep may have a variable influence on the complexes: In some patients it has no effect, but in others the complexes are either brought out, enhanced, or lost during sleep.F" Occasionally, they seem to disappear for a time, but this is unusual except in the terminal stages of the illness.R' When myoclonic jerking is a clinical feature of the patient's illness, the jerks are usually time-related to the periodic complexes, occur-

68

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp2-F4 F4-C4 C4_P4 P4-02

- -...........- . A J - V ----....J"'II.""""'........-""-.......r-"'N'-"""A .,.",.......--.,..-.....-II/'--

-..-.r--.."........------.".....'\rv-~

Fp2-F8 F8-T4 T4-T6 T6-02----.!'-'---Fp1-F3 __--v/V---v\ 1"'----''1 F3-C3 ---,-,-",,C3-P3------.,..,-"'V'--J-----"""'-"-------...--..........-..----' ,----r"""",-,,"", P3-01 ~--~~------~...........'""-~---......,.~Fp1-F7 F7- T3

~

"'--.--..,,' -----"'"--..

'__"__'_""'-...__--r~"""""...J-

T3-TS-------------------------..-..------""'""'-TS-01---------------------L-.J 1 sec

200 ....V[

FIGURE 3·14 • Repetitive complexes seen in the right temporal region of a child with herpes simplex encephalitis.

ring just before or just after them. Although the background EEG activity is sometimes relatively normal, it is generally characterized by a reduction or loss of alpha rhythm and the presence of diffuse slow activity. Randomly occurring spikes or sharp waves may be

found, especially in the frontal regions, as may bilaterally synchronous spike-wave activity or rhythmic frontal delta activity.'!' Transient, relative quiescence of the background following a complex is an occasional finding.

Fp1-F3 F3-C3 ~

C3-P3 P3-01 Fp2-F4 F4-C4

1J'-"


.-vv-~

C4-P4 P4-02

'\.."y-

F7-Fp1 Fp1-Fp2 - Fp2-F8

-/"'-

.....

--V\

JI.J'v

V--

v

wv-

~


~

"vv-~

_'\0/"'-.

-----

TS-01 0 1 - 0 2 - - -..........""" EMG right forearm L.....J

1 sec

--"

f" [ 300 ....V

FIGURE 3-15 • EEG of a 16-year-old boy with subacute sclerosing panencephalitis of 8 months' duration. Generalized slow-wave complexes are seen to occur approximately every 4 seconds and last for up to 1 second. Each complex is accompanied by a brief EMG discharge from the right forearm extensor muscles.

Electroencephalography: General Principles and Clinical Applications

Periodic complexes similar in form and repetition rate to those just described have been reported in patients with phencyclidine (PCP) intoxication.P" CREUTZFELDT-JAKOB DISEASE

In Creutzfeldt-Jakob disease, the EEG is also characterized by periodic complexes occurring on a diffusely abnormal, slowed background (Fig. 3-26). The complexes differ from those occurring in subacute sclerosing panencephalitis. They consist of brief waves (generally less than 0.5 second in duration) of variable (often triphasic) form and sharpened outline; these waves recur with an interval of 0.5 to 2 seconds between successive complexes and may show a temporal relationship to the myoclonic jerking that patients commonly exhibit. This periodic activity is usually present diffusely and is then bilaterally synchronous, but it may have a more restricted focal or lateralized distribution in the early stages of the disorder. Such a restricted distribution correlates with clinical cerebral dysfunction and with the findings on MR!. J23 Asymmetries occasionally persist, even at advanced stages. In Heidenhain's variant of the disease, periodic complexes may remain confined to the occipital regions and initially may be unilateral.l'" The typical periodic discharges may not be found in the early and terminal stages of the disease. When the disease is sufficiently advanced that it is strongly suspected on clinical grounds, typical EEG abnormalities have been reported in more than 90 percent of patients.l'" and in pathologically verified cases a rate greater than 85 percent has been obtained. 126 The EEG is less likely to show periodic complexes in patients with an unusually long clinical course.F'' The periodic discharges are not present in certain familial forms of Creutzfeldt-Jakob disease "? or in the new-variant disease, in which the EEG is often abnormal in a nonspecific manner. 128 ,129 They are

also absent in Gerstmann-Straussler-Scheinker syndrorne.P" Both the clinical features of a rapidly progressive encephalopathy and the EEG appearance of periodic complexes suggesting Creutzfeldt-Jakob disease have been reported in patients with lithium or baclofen toxicity, indicating the need to take a careful drug history in such circumstances.P'-P'' Bismuth subsalicylate toxicity may lead to a prolonged encephalopathy that is mistaken for Creutzfeldt-Jakob disease, but the EEG is usually diffusely slowed, without more characteristic features. 133 ABSCESS

In patients with acute supratentorial cerebral abscess, the EEG is characterized by low-frequency, high-amplitude, polymorphic focal slow activity such as that shown in Figure 3-9. PLEDs have also been reported over the involved hemisphere in occasional cases. The EEG changes are often much more dramatic than those encountered with other focal cerebral lesions. Depending on the degree to which consciousness is depressed, however, focal changes are sometimes obscured by a more generalized slow-wave disturbance or by frontal intermittent rhythmic delta activity. Infratentorial abscesses generally produce less marked EEG abnormalities characterized by intermittent rhythmic delta rhythms or diffuse slow activity. The EEG changes produced by chronic supratentorial abscesses may be indistinguishable from those produced by focal lesions such as tumors (p. 72).

Vascular Lesions In considering the changes that occur in patients with vascular lesions of the nervous system, it must be borne

Fp1-F3-......---v--"""'"--------V\r-.l\r-'""\'-JV---v---vv---vF3-C3 -v--_oJ'---'""'''----'''I.J'v--""'''''"-¥\r-..... L,.-4 ,-"'\r---'\r-..-"VI..o-"""v-""-C3-P3 - . r_ _' ...........-J""'"'--.-.J·v---'v_J'v--.l\...,.--.J\,.-__-.-JiIv_--""I.r--.J'--v'\.__ P3-01 -V"'--I\-.."""V'--oJ"/'v-IV"--./'---.(V'-_"---,j'---"",,,,"-""V--..r-..../"Fp2-F4 _""---'"""""_-.r-''''''---rr---"'I\r-.......''\/~\r-"''''-r--'V--~-v...--.._ F4-C4 _ _ A _--",,,,--.,.IIIr---'1,r.......\.r--...,..,..."'----~

~.---...,--....,JI.r---.fIv

C4-P4 -'"''''---.!''-.....-J'"""'-'""'-J'v''--'f\---I\.--..../I,,----''v-"'''I/I,.--''IIr-''''--'II'..-.''IIIP4-02 ---'\,t"--"--""'V'--..../'v--.f'v---J'"--.r----..v--.!'----r---,/'--r'-/'-L---l 1 sec

300

69

J.l.v[

FIGURE J·26 III Repetitive complexes, often triphasic in form, occurring about once every second on a rather featureless background in the EEG of a 66-year-old patient with Creutzfeldt-Jakob disease.

70

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

in mind that the EEG findings fail to provide a reliable means of distinguishing such lesions from other types of structural pathology (e.g., cerebral tumors). Similarly, the EEG findings do not provide a way of distinguishing the underlying pathologic basis of the vascular disturbance. and this drawback limits the utility of the technique. Only a brief description is therefore provided in this chapter. In patients with cerebral ischemia caused by occlusive disease of major vessels, the normal background activity is often depressed in the affected portion of the hemisphere, and a slow-wave (theta or delta) disturbance may be found, sometimes associated with sharp transients or PLEDs. Changes are usually most conspic1I0US in the ipsilateral midtemporal and centroparietal reg-ions in patients who have involvement of the middle cerebral or internal carotid artery, in the frontal region when the anterior cerebral artery is affected, and in the occipital region when the posterior cerebral artery is occluded. The topographic extent of these changes varies in individual patients, however, depending on factors such as the site of occlusion, rapidity of its development. and adequacy of the collateral circulation. The EEG may revert to normal with time despite a persisting clinical deficit. but sharp transients become more conSpiUlOUS in some cases. In patients with ischemia restricted to the internal capsule, the EEG is usually normal or shows only minor changes. Similarly, discrete subcortical vascular lesions (e.g., lacunar infarcts) generally do not alter the EEG. In patients with vertebrobasilar ischemia, the EEG is often normal, although it may show minor equivocal changes. In some cases of infarction of the lower brainstem (Fig. 3-27), however, widespread, poorly reactive alpha-frequency activity is found. With involvement of more rostral brainstem regions and of the diencephalon, the EEG is characterized by predominant slow activity, which is often

Fp1-A1

organized into discrete, bilaterally synchronous runs without constant lateralization. Following acute nonhemorrhagic stroke, the EEG may show focal abnormalities (e.g., a polymorphic slowwave disturbance) at a time when CT scans are normal, as in 19 percent of the patients studied by Gilmore and Brenner," Thus, in this context, the EEG may reflect dysfunction not yet apparent morphologically. Such changes may be helpful in distinguishing between a hemispheric and brainstem lesion. They have also been used to distinguish between a cortical and a lacunar infarct when such distinction is clinically difficult, although a recent study indicates that lateralized minor abnormalities are more common than previously appreciated after lacunar infarction. 134 In patients with mild atherosclerotic cerebrovascular disease, hyperventilation may induce focal or lateralized EEG abnormalities that are not apparent otherwise. Carotid compression is sometimes used to provoke EEG changes caused by inadequacies of the collateral circulation. This maneuver is not always innocuous, however, and may give misleading information (e.g., positive responses in patients without vascular disease or a lack of response in patients with well-established disease). In patients with intracerebral hematomas, the extent of the EEG changes depends on the site and size of the hematoma and on the rapidity of its development. Background activity is depressed over part or all of the affected side, and focal polymorphic delta activity is seen. Especially in the elderly, this slow activity is sometimes localized preferentially to the temporal region regardless of the site of the lesion. Sharp transients are seen more commonly than in patients with nonhemorrhagic vascular lesions, especially in the temporal region, and intermittent bilateral rhythmic delta activity also may be found, especially where secondary displacement of brainstem structures has occurred. EEG

'w/vooJwv..""""~,fIWo"v-MIvJvJ"~VWJo;",",,",,~oJ'vwo'ONoM~M.M!Wv"'I"WMfo/W"''{''IAvwN

F7-A1 1~~~'vf'/"VIf/JWoWM~~~~~~

T5-A1

'wW'MJw.""-.).fflI.NW'\~/VJM"~"""""'fVW-o,,,,,,,,,"~,,,,,,,,,,,-.;vv..~II/VotoJ""""">NwvoNWl''vIworwJlf'o

Fp2-A2 ~I\MtoIvWVwvo/f'AJ.Jo."""""'~MvJ.IwW'rl'Mto/lr.r-ttWvJww..w. F8-A2

WovNV'vJIV'I~tJ""'V"""""'.wNo~WoMJ~ofN..No~WIv.I'Ivo/VIV'w.J~.,pwI/w.tloI.lv'tA/'MNW

T6-A2 ~IWvJVvvW/lJAIWIW~~AtNwwJ.M"'NIIVwN-"WM/.IWM~""""""~~IoWwi~WMiWJ I 300 ILV[ 1 sec FIGURE 3·17 • Generalized, rhythmic 9- to lO-Hz activity. unresponsive to sensory

stimuli, in the EEG of a 68-year-old woman following cardiopulmonary arrest.

Electroencephalography: General Principles and Clinical Applications

changes are often inconspicuous in patients with hemorrhage into the lower brainstem, but widespread, poorly reactive alpha-like activity may be noted. Cerebellar hemorrhage is associated with little, if any, change unless tonsillar herniation complicates the clinical picture, in which case diffuse slowing results. The EEG findings in patients with cerebral venous or venous sinus thrombosis are similar to, but often more extensive than, those in patients with occlusive arterial disease. When the superior sagittal sinus is involved, the changes are usually bilateral, often variable, and commonly asymmetric. I t is unclear whether the EEG has any role in indicating the prognosis following stroke, and no definite conclusions can be reached on the basis of the few published studies on this point. The EEG has been used to monitor patients undergoing carotid endarterectomy, and this is discussed further in Chapter 9. A reversible posterior leukoencephalopathy has been related to hypertensive encephalopathy. It is associated with characteristic abnormalities on neuroimaging that suggest subcortical edema without infarction.!" Occipital lobe seizures may occur.P" and variable EEG abnormalities may be encountered. The diagnosis, however, is based 011 the clinical and neuroimaging abnormalities. SU.ARACHNOID HEMORRHAGE

Although the EEG may be normal following subarachnoid hemorrhage, diffuse slowing is a common finding, especially in patients with clouding of consciousness. Focal abnormalities may also be observed and can be related to the presence of a local hematoma; to the source of hemorrhage, especially when it is an angioma; or to secondary ischemia by arterial spasm. Continuous EEG monitoring in the intensive care unit is heing used increasingly to detect focal changes

caused by vasospasm while they are still subclinical or at least at a time when the vasospasm may still be reversible. 137 SUBDURAL HEMATOMA

The background activity is sometimes reduced in amplitude or virtually abolished over the affected hemisphere in patients with a chronic subdural hematoma (Fig. 3-28). In other instances, however, a focal ipsilateral slow-wave disturbance is the most conspicuous abnormality, and little suppression of background rhythms is found. In either case, generalized changes (e.g., frontal intermittent rhythmic delta activity) may be present as well, and repetitive periodic complexes have been described in rare instances. Because the EEG is sometimes normal in patients with a chronic subdural hematoma, the possibility of such a lesion cannot be excluded just because the electrophysiologic findings are unremarkable. Indeed, even when abnormalities are found, it can sometimes be difficult, ifnot impossible, to localize the lesion with any certainty owing to the interplay of the various changes described earlier. Moreover, the EEG findings do not reliably distinguish a chronic subdural hematoma from other space-occupying lesions. In all instances, therefore, detailed neuroradiologic assessment is necessary if cases are to be managed correctly. INTRACRANIAL ANEURYSMS AND ARTERIOVENOUS MALFORMATIONS

The findings in subarachnoid hemorrhage have already been described. Focal slow activity is an occasional finding in patients with aneurysms that have not bled, but in uncomplicated cases there is often little to find in the EEG. Following subarachnoid hemorrhage, focal or

Fp1-F7 ----------------------~--

p-T3

~_v___

_~

_

T3-T5 N""~v__._....w"""Ww-_.....w.JV"'--""-"'"""""~\N'M/WW..........._"""""__-_.....,..,....,..".". T5-01

-""'--"_ _""'"""_ _..,....J""'-".".,...""-vYwW""_....-._ _....."..,_---___.fW'--

Fp2-F8 ~---------------------­ F8-T4 - - - - - - - - - - - - - - - - - - - - - - - - - T4-T6 ---'"'------.-.... "....r"'"-----........- -__......,.-_ _- _ ~ _

T6-02 L-...J

1 sec

71

300/-lV[

FIGURE 3-28 III! Focal slowing in the right posterior quadrant and suppression of normal background rhythms over this side in the EEG of a patient with a suspected right subdural hematoma.

72

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

lateralized abnormalities are sometimes present and may be a guide to the site of bleeding. Focal slow activity or epileptiform discharges may be found in the EEG of patients with intracranial arteriovenous malformations. SrURGE-WEBER SYNDROME

Together with depression of normal background activity and the responses to hyperventilation and photic stimulation over the affected hemisphere, irregular slow activity and sharp transients are often seen.P" The reduction in background activity is not necessarily related to the presence and degree of intracranial calcification. Abnormalities sometimes have a more generalized distribution that can be confusing, but in such circumstances they are commonly more conspicuous over the involved side.

Headache Electroencephalography has little relevance to the diagnosis of migraine. In uncomplicated cases, the EEG is usually normal, or shows only minor nonspecific changes between and during migrainous attacks. Focal or unilateral slow-wave disturbances are common, however, particularly in patients developing lateralized aura or neurologic deficits in association with their attacks. Such localized abnormalities usually settle rapidly once the clinical disturbance has resolved, unless infarction has occurred or there is an underlying structural abnormality. Paroxysmal activity is sometimes found, but the proportion of cases with such a disturbance varies greatly in different series. In one study of the EEG in 100 children with migraine, 9 (without a history of epilepsy) were found to have benign focal epileptiform discharges at a time when they were free from headaches. I:19 This incidence of 9 percent is higher than in the general population and is of uncertain significance. It is possible that migraine and such discharges are genetically linked or that the vascular abnormality of migraine leads to cerebral ischemic damage and thus to the EEG abnormality.P" The EEG disturbance, in itself, is insufficient evidence to permit the headaches to be regarded as the manifestation of an epileptic disorder. There is always some concern that an underlying structural lesion may have been missed in patients with chronic headache syndromes that fail to respond to medical treatment. In this regard, a possible role has been suggested for the EEG as a screening procedure for patients requiring further investigation. The EEG is g-enerally less expensive than cranial CT scanning or MR!, but it is also less sensitive to intracranial masses than are these imaging procedures. Whether a costbenefit analysis would justify such a role for the EEG is

uncertain, however, and there is insufficient information in the literature to provide guidance on this point. 140 Moreover, given the implications of delay in the diagnosis of an intracranial mass lesion, it is hard to justify use of the EEG in place of imaging studies when these modalities are readily available and intracranial pathology is suspected clinically.U" There is little if any evidence that the EEG has a useful role in the routine evaluation of patients presenting with headache, and it is not helpful for predicting prognosis or in selecting therapy.r'?

Tumors Tumors may affect the EEG by causing compression, displacement, or destruction of nervous tissue; by interfering with local blood supply; or by leading to obstructive hydrocephalus. Considerable variation exists in the presence, nature, and extent of abnormalities in different subjects, however, and this variability seems to depend, at least in part, on the size and rate of growth of the tumor and the age of the individual patient. An abnormal record is more likely to be found in patients with a supratentorial tumor than an infratentorial one, and in patients with a rapidly expanding tumor rather than a slowly growing lesion. Superficial supratentorial lesions generally produce more localized EEG changes than do deep hemispheric lesions; the latter may lead to abnormalities over the entire side involved or to even more diffuse changes. EEG abnormalities are more common with rostral than with caudal infratentorial lesions." In general, abnormalities are more conspicuous with tumors in children than with those in adults. Even if the EEG is abnormal, however, the changes may be generalized rather than focal. Thus, they would not be particularly helpful in the diagnosis or localization of the underlying neoplasm, although they would provide information about the extent of cerebral dysfunction produced by it. Diffuse abnormalities are common in all patients with cerebral tumors when the level of consciousness is depressed; therefore, in this context, these abnormalities are particularly probable in patients with infratentorial lesions. In addition to diffuse slowing, intermittent bilateral rhythmic delta activity may be seen with a frontal emphasis in adults, or more posteriorly in children. In these circumstances, localization of the tumor by EEG is often less feasible than at an earlier period in the natural history, although a gross asymmetry of such activity between the two hemispheres raises the possibility of a lateralized hemispheric lesion. Earlier records sometimes permit more definite localization, but they may show only subtle changes, which are easily missed. Depression of electrical activity over a discrete region of the brain is a reliable local sign of an underlying

Electroencephalography: General Principles andClinical Applications

cerebral lesion, but this sign may be masked by volume conduction of activity from adjacent areas. The presence of a focal polymorphic slow-wave disturbance (Fig. 3-29) is also important, although such an abnormality has less localizing significance when it is found over the temporal lobe. Focal ictal or interictal sharp activity is of some, but lesser, localizing value unless it is associated with an underlying focal slow-wave disturbance. However, epileptiform activity may precede the appearance of more reliable focal EEG abnormalities by several months or even longer. It usually occurs at the margins of the lesion and is more likely with slowly growing tumors than with rapidly expanding ones and more likely with hemispheric tumors than with brainstem lesions. Some correlation exists between the presence of epileptiform activity in patients with brain tumors and the development of seizures, but in general this correlation is not close enough to be of prognostic relevance. Many patients with such EEG discharges do not have seizures; conversely, many patients with seizures from brain tumors do not have spikes and sharp waves in their EEGs. Epileptiform discharges may be found bilaterally, without obvious focality, in patients with parasagiual or mesial hemispheric lesions, but are often then asymmetric in distribution over the two sides or confined to the vertex.P' A number of other abnormalities have been described in patients with discrete cerebral lesions, including an asymmetry of druginduced fast activity, a local increase in beta activity, and the presence of a mu rhythm; but these findings are of much less significance and can lead to difficulty in determining which of the two sides is the abnormal one. In assessing the findings in patients with suspected brain tumors, it must be borne in mind that the main value of the EEG is in indicating which patients require

more detailed investigation, especially when facilities for CT scanning or MRI are not readily available. Its value even in this regard is limited, however, because a normal or equivocal EEG does not exclude the possibility of a tumor. Deep-seated supratentorial tumors may produce no abnormalities whatsoever at an early stage, and the EEG is likely to be normal in patients with pituitary tumors unless the lesion has extended beyond the pituitary fossa or has caused hormonal changes. Furthermore, there are no abnormalities that will allow the differentiation of neoplastic lesions from other localized structural disorders, such as an abscess or infarct. The EEG can provide no information about the nature of the tumor in individual cases; the findings in patients with tumors of different histologic types or in different locations are therefore not discussed in this chapter. Although its place as a noninvasive screening procedure has largely been taken over by CT scanning or MRI in the developed countries, the EEG still has an important complementary role in the evaluation of patients with known or suspected brain tumors. In particular, the EEG is helpful in the evaluation of episodic symptoms that might either be epileptic in nature or have some other basis, and it provides information about the extent of cerebral dysfunction.

Tuberous Sclerosis There are no pathognomonic EEG features of tuberous sclerosis, but epileptiform activity or changes suggestive of space-occupying lesions may be found. The EEG is often normal in mild cases, but in others focal slow- or sharp-wave disturbances are seen. In more advanced cases, a hypsarrhythmic pattern may be seen during infancy, whereas in older patients the record may contain generalized spike-wave activity or independent

Fp1-F3 F3-C3 C3-P3 ,J,r..J'-.mJ~tWv-_oNM"""""'W"""''V'I'NiI"'''Ir~'NwW'YI'''''''''('IW''''''''Jl,.yv-....JIII'''''''''('IV'M. P3-01

--v--.-"''''''''''''''''''-''''''''''''''''''-

Af{V.'WNoMN"\-"",-~_"""""--""",,,,-.,.,-..........

------""""'1-----/'iYW""-""'---""""'_.,.-_

Fp2-F4 ~..... F4-C4

,.".,..,._~_.....,.,._-

C4-P4 -"I""""""-....-......."."'W\(\_""""' P4-02

.........

-""\IW'-..,.;..,.oIWo....-~'""""'_

........."""""_.,.

_ _I""_J""""""_"'_"""_.-.If'-o.,..."..oJ\MJ_""""".,.".".._...,.,...'V"""""'-

..,......""""~,..,..,..

L....J 1 sec

300 fLV [

II Polymorphic slow-waveactivity in the left frontal region in the EEG of a 62-year-old man with a glioma.

FIGURE ]·29

73

74

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

multifocal spike discharges. Thus, focal or generalized changes may be found in this disorder.

Pseudotumor Cerebri In pseudotumor cerebri, or benign intracranial hypertension, the EEG is often normal, but abnormalities, consisting of a diffusely slowed background and bursts of activity in the alpha-, theta-, or delta-frequency ranges, are sometimes found.

Dementia The EEG changes in most patients with dementia are nonspecific and do not discriminate between the different types of dementing processes. In early cases, the findings may be normal. As the dementia advances, however, the amount and frequency of the alpha rhythm declines, and irregular theta and slower activity appear, sometimes with a focal or unilateral emphasis, especially in the temporal region. In patients with Alzheimer's disease, the alpha rhythm is often lost at a relatively early stage compared with patients with other disorders that cause dementia. As a general rule, abnormalities occur earlier and are much more conspicuous in Alzheimer's disease than in Pick's disease or frontotemporal dementia. In the latter disorder, the EEG is often norrnal'V; even if diffuse slow-wave activity is present, the alpha rhythm is commonly preserved. In some cases, the cause of dementia may be suggested by the EEG findings. Both Creutzfeldt-Jakob disease and subacute sclerosing panencephalitis are commonly associated with characteristic EEG changes, as described earlier. When a focal structural lesion (e.g., a tumor or chronic subdural hematoma) is responsible for the intellectual decline, there may be localized depression of electrocerebral activity or a focal polymorphic slow-wave disturbance, sometimes with relative preservation of the alpha rhythm. In multi-infarct dementia, significant asymmetries in background activity often occur over the two hemispheres, and focal slowwave disturbances are common. The findings in patients with parkinsonism, Huntington's disease, hepatolenticular degeneration, and Steele-Richardson-Olszewski syndrome are discussed on page 79 but are of no help in distinguishing the intellectual decline occurring in those disorders from other diseases associated with dementia. In patients with the dementia sometimes associated with HIV infection, the EEG may be normal or diffusely slowed. Focal slow-wave disturbances in these patients raise the possibility of neoplastic involvement or opportunistic infection. A high incidence of EEG ab normalities has been reported in asymptomatic Hlv-sero-

pOSItIVe homosexual men by sorne.l'" but not other, authors,144-146 raising the possibility that electrophysiologic studies may be useful as a means of indicating subclinical neurologic involvement. Among patients with acquired immunodeficiency syndrome (AIDS), 67 percent have been reported to have EEG abnormalities, usually in relation to dementia or opportunistic infection but sometimes without apparent neurologic disease.r'? Considerable overlap exists between the EEG findings in patients with dementia of the Alzheimer type and mentally normal elderly subjects; thus, the EEG cannot be used to indicate reliably whether an elderly patient has dementia. However, the presence of diffuse slow-wave activity supports the diagnosis of dementia rather than pseudodementia (depression). Again, the EEG findings cannot be used to distinguish with any confidence between an acute and a chronic disturbance of cognitive function, although in the former circumstance abnormalities are more likely to reverse with time. However, a markedly abnormal EEG in patients with clinically mild cognitive disturbances should suggest an acute process, such as a toxic or metabolic encephalopathy.

Metabolic Disorders The EEG has been used to detect and monitor cerebral dysfunction in patients with a variety of metabolic disorders, and changes may certainly precede any alteration in clinical status. The EEG findings may also be helpful in suggesting that nonspecific symptoms have an organic basis. Diffuse (rather than focal) changes characteristically occur in metabolic encephalopathies unless there is a pre-existing or concomitant structural lesion. Typically, desynchronization and slowing of the alpha rhythm occur, with subsequent appearance of theta, and then of delta, activity. These slower rhythms initially may be episodic or paroxysmal and are enhanced by hyperventilation. Triphasic waves (p. 48), which consist ofa major positive potential preceded and followed by smaller negative waves, may also be found and usually are generalized, bilaterally synchronous, and frontally predominant. They are sometimes reactive to external (painful) stimulation. Although triphasic waves were originally thought to be specific for hepatic encephalopathy, they are in fact found in a variety of metabolic encephalopathies'Pr" and have been held to suggest a poor prognosis for survival. 20 Diffuse slowing of the EEG has been described in hyperglycemia or hypoglycemia, Addison's disease, hypopituitarism, pulmonary failure, and hyperparathyroidism. In hypoparathyroidism, similar changes are found in advanced cases; but spikes, sharp waves, and slow spike-wave discharges may also be seen. A lowvoltage record in which the alpha activity is classically

Electroencephalography: General Principles and Clinical Applications

preserved but slowed characterizes myxedema, and there may be some intermixed theta or slower elements, These changes may pass unrecognized, however, unless premorbid records are available for comparison. In hyperthyroidism, the alpha rhythm is usually diminished in amount but increased in frequency, beta activity is often conspicuous, and scattercd theta elements may be present. The findings in Cushing's disease and pheochromocytoma are usually unremarkable. Water intoxication or hyponatremia may cause diffuse slowing of the EEG, and bursts of rhythmic delta activity are also commonly present. Hypernatremia and abnormalities of serum potassium concentration usually have little effect on the EEG. HEPATIC ENCEPHALOPATHY

The EEG shows progressive changes in patients with advancing hepatic encephalopathy, and in general a good correlation exists between the clinical and electrical tindings. 148 In early stages the alpha rhythm slows, gradually being replaced by theta and delta activity, but in some instances it may coexist with this slow activity.l48 As the disorder progresses, triphasic complexes usually (but not invariably) occur symmetrically and synchronously over the two hemispheres, with a frontal emphasis (Fig. 3-30). With further clinical deterioration, the EEG in patients with hepatic encephalopathy comes to

Fp2-F4 F4-C4 C4-P4 f'--..-.I'_r"'V'--r--r~";'.... I"""" v.." """, .... , P4-02 .A--V--v-"I.r--<'\/'oJV'-J',-.J\...-I'--.J-............v"-..r--J'V"'""--A Fp2-FB F8-T4 T4-T6 T6-02 ....-..~f'.,,--..r-'oo...........ro-r--.l'...,....""'""V'-..,......,/'-""'"'''--'' '-,/ '-'1J Fp1-F3 \foo- ~,r'\rv\jrvv F3-C3 C3-P3 P3--01 Fp1-F7 F7-T3 T3-T5 -""- r-«: J '---./ .......,1'-,;/"'-J~

-t-, I\,~,..-.JV'..,1

T5-01

~

300 ....V[

FIGUJlE J-JO • Triphasic and slow-wave activity in the EEG of a 79-year-old woman with hepatic encephalopathy.

75

consist of continuous triphasic and slow-wave activity, which shows a marked anterior emphasis and may be interrupted by periods of relative quiescence. The amplitude of the slow-wave activity gradually decreases as death approaches. After orthotopic liver transplantation, epileptiform activity may be encountered in the EEG, especially in those who ultimately fail to survive. 149 Other EEG abnormalities in liver transplantation recipients include diffuse slowing, local or generalized suppression of electrocerebral activity, the presence of triphasic waves, electrographic seizures that may be subclinical, and PLEDs.149 Focal or generalized abnormalities may be related to cerebrovascular complications or to infective processes to which patients are prone because of their immunosuppressed status. Generalized abnormalities may also be related to other metabolic disturbances, central pontine myelinolysis, and medication effects. RENAL INSUFFICIENCY

In patients with renal insufficiency, the EEG may be normal initially but the background eventually slows; theta and delta activity develop in increasing amounts and are sometimes paroxysmal; and, ultimately, the record is dominated by irregular slow activity that does not respond to external stimuli. Triphasic complexes are an occasional finding but usually are not as wellformed as those in hepatic encephalopathy. Spike and sharp-wave discharges may also be seen in some patients, as may photoparoxysmal responses. Hughes looked for correlations between the EEG findings and 10 different chemical indices in 23 uremic patients tested repeatedly for up to 18 months. l',O An abnormal EEG was recorded at least once in 70 percent of these patients. Slow-wave abnormalities, mainly mild and frontal, were found in 97 percent of abnormal records, and epileptiform discharges, consisting of bilateral spike-wave complexes, were seen in 14 percent. The main conclusion from the record was that the blood urea nitrogen (BUN) correlated best with the EEG findings. especially if the record was deteriorating, and variations in the amount of epileptiform activity were most closely related to changes in the BUN. During hemodialysis, the EEG findings are often dramatic, even in patients who previously had a relatively normal record, consisting of bursts of generalized, high-voltage rhythmic delta activity occurring on either a relatively normal or generally slowed background. In patients developing the progressive encephalopathy that sometimes occurs during chronic hemodialysis, the EEG contains diffuse slow-wave activity interrupted by bilaterally synchronous complexes of slow, sharp, triphasic, and spike waves. 15 J.152 Attempts have been made to predict the presence or probable development of this encephalopathy by the EEG findings and, in

76

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

particular, by the presence of bilateral spike-wave complexes.P'' Such complexes were found in 20 of26 patients with the disorder in one study, and enabled the presence or absence of dialysis encephalopathy to be predicted correctly in 91 percent of a total of 77 patients undergoing dialysis.152 In some instances the EEG shows such changes several months before the encephalopathy becomes manifest clinically,151,152 but in other cases spike-wave discharges are not found before clinical onset of the dlsorder.l'" ANOXIC ENCEPHALOPATHY

The EEG changes occurring during acute pancerebral hypoperfusion or hypoxia were discussed earlier in the section on syncope (p. 66). If cerebral hypoxia or anoxia is prolonged, it leads to an encephalopathy that may be irreversible. Diffuse slowing of variable degree is found in the EEG of some comatose patients, and short runs of fast activity may also be present. The responsiveness of the background to external stimulation (Fig. 3-31) provides important prognostic information in such circumstances (p. 77). In patients with more severe brain damage, the EEG may show continuous or intermittent epileptiform discharges, diffuse unresponsive alpha-frequency activity (p. 78), a burstsuppression pattern (see Fig. 3-12 and p. 51), or electrocerebral inactivity. A detailed discussion of the EEG findings in brain death and neocortical death is provided in Chapter 34.

Effects of Drugs and Alcohol on the EECi The effects on the EEG of various anticonvulsant drugs are considered on page 64. Neuroleptic and tricyclic

antidepressant drugs may cause diffuse slowing of the EEG and, in addition, provoke paroxysmal slow activity or spike-wave discharges. Clozapine treatment may lead to background slowing in the theta or delta ranges, and spike, spike-wave, and photoparoxysmal discharges have also been described.!" The slowing relates to plasma drug levels. 154 Barbiturates and benzodiazepines lead especially to an increase in beta activity, whereas acute withdrawal from them after chronic usage may be associated with generalized epileptiform discharges and photoparoxysmal responses. Lithium causes slowing of the alpha rhythm and paroxysmal slow-wave activity that may have a focal or lateralized emphasis. Periodic complexes resembling those occurring in Creutzfeldt-Jakob disease have been described with lithium 131 or baclofen toxicity.l'" Alcohol causes mild slowing during chronic intoxication. Paroxysmal activity is occasionally found during alcohol withdrawal, following which the EEG reverts to normal unless there is any co-existing pathology. Photomyogenic or photoparoxysmal responses may occur, but these responses were rarely encountered in one study of 49 subjects during acute alcohol withdrawal.?" In most patients with alcohol-withdrawal seizures, the EEG is either normal or mildly slowed; epileptiform discharges are uncommon. Focal abnormalities suggest the possibility of a structural lesion or may be the postictal sequelae of partial seizures. Recreational drugs also affect the EEG. Amphetamines are said to increase the amount of beta activity. Similarly, cocaine increases beta activity, as does its withdrawal in cocaine-dependent subjects. 11,155,156 Narcotic drugs may lead to a reduction of alpha frequency and duration and, with chronic administration, to slowing of the record.

Loud Clap

t

F3-C3

-w\/'t'~fJ'.J.-fll'l'J"',!

C3-P3~\ P3-01 F4-C4

!\,J"J\/V'w

-r-J:vi\rf"'A,JV\r/'v,~/~-~J\)\\.r"/\"""""""'/\~oI ~

C4-P4 '(\IJ'(""v\J'vllr/Vo0.J~\

P4-02 ",v'~,)V-:'~V"J'--...j"A./'~ T3-Cz

Vv../Ij

CZ-T4~ L-.J 1 sec

200 11V [

FIGURE J·J1 • EEG of a comatose child showing attenuation of the diffuse background slow activity in response to a sudden loud noise.

Electroencephalography: General Principles and Clinical Applications

Multiple Sclerosis Patients with multiple sclerosis commonly have abnormal EEGs. Focal or generalized activity in the theta and delta ranges may be found, as may diffuse or localized spike discharges. Focal changes are often evanescent and are probably related to foci of acute demyelination. The findings are of no help in the diagnosis of the disorder, however, and usually bear little relationship to the clinical signs.

Trauma Electroencephalography is generally undertaken to provide some guide as to the nature and severity of head injuries, the prognosis for recovery, and the likelihood of developing posttraumatic seizures. In addition, it is often requested when patients with posttraumatic syndromes are being evaluated at a later stage in the hope that it will provide an indication of whether nonspecific symptoms have an organic etiology; however, its use in this regard has no rational basis. No specific EEG abnormalities develop after head injury. In evaluating the findings at electroencephalography, one must bear in mind that abnormalities may have existed before the injury and that the significance of any findings will depend, in part, on the time at which the study was undertaken in relation to the trauma. This is important because the correlation between clinical and electrophysiologic findings is often poor when the EEG is recorded 3 or more months after the injury was sustained. The presence of a localized abnormality is often of particular concern because it may point to the existence of an intracerebral, subdural, or extradural hematoma. Unfortunately, there is no reliable way of distinguishing electroencephalographically between surgically remediable intracranial lesions, such as a subdural hematoma, and pathology that does not necessitate operative treatment. In such circumstances, the judicious use of imaging procedures clearly facilitates the rational management of individual patients. The EEG findings in individual cases depend, in greater part, on the patient's level of consciousness. The abnormalities that may be seen in comatose subjects are discussed separately, and there is further comment on this aspect in Chapter 34. Other abnormalities that may be found following head trauma include local or generalized depression of normal activity; slowing of the alpha rhythm; focal or diffuse slow-wave disturbances, especially in the temporal region in adults; and spike or paroxysmal discharges. These changes may develop progressively with time, if serial recordings are undertaken. Focal abnormal-

77

ities may be related directly to local injury but also may occur as a sequela to complications of cerebral trauma such as ischemia or edema. They may be obscured initially by more generalized changes in the EEG, becoming conspicuous only as the latter diminish. Before pathologic implications can be attributed to a localized depression of electrocerebral activity, however, extracerebral factors (e.g., subcutaneous edema or hemorrhage) must be excluded by careful scrutiny of the patient. Localized or lateralized paroxysmal discharges are sometimes the sole evidence of a structural abnormality, such as a hematoma, and indicate the need for further neuroradiologic investigations. Jennett and van de Sande found that although abnormal records were more common in patients who developed seizures following head injury, the EEG findings did not improve the accuracy of predicting which patients were likely to go on to develop posttraumatic epilepsy.P? Patients with EEG abnormalities could remain free from seizures, whereas 20 percent of those who went on to develop posttraumatic epilepsy had at least one normal record in the 3 months following injury. A recent development has been the use of continuous EEG monitoring in the intensive care unit of patients with traumatic brain injury to detect seizures that might be clinically inapparent and to guide early prognostication. Vespa and colleagues found in a study of94 patients monitored during the initiall4 days after injury that 22 percent had seizures; more than half of these were nonconvulsive and were diagnosed solely on the basis of the EEG.l08 In another study, monitoring during the first 3 days after injury, with particular attention to the percentage of alpha variability, was found to be sensitive and specific in this regard, improving prognostic ability independently of traditional clinical indicators of outcorne.l'" Clearly, further studies are necessary to replicate and extend these findings, which suggest an important role for EEG monitoring of such patien ts,

Coma Altered states of consciousness may result from many causes, and it is therefore not surprising that the EEG findings in comatose patients are variable. A few general points require emphasis. First, although the EEG changes are never specific to any particular disorder, they may direct attention to diagnostic possibilities that might otherwise be overlooked. Second, in evaluating patients with depressed levels of consciousness, it is particularly important, both for prognostic purposes and for following the course of the disorder, that serial records are obtained so that the direction of any

78

ElECTRODlAGNOSIS IN CLINICAL NEUROLOGY

change can be determined. Automated "trending" of the EEG may be particularly helpful in the context of the intensive care unit. Finally. any spontaneous variation in the EEG and the responses evoked by external stimuli must be considered when the comatose patient is being evaluated or followed electrophysiologically (see Fig. 3-31). A change in electrocerebral activity can be expected to occur following stimulation of a patient with a mildly depressed level of consciousness; this reactivity becomes inconstant, delayed, or lost as the depth of coma increases. More specifically, stimulation during light coma results in an attenuation of the background rhythms or, especially as coma deepens, a paradoxical arousal response in which slow-wave activity briefly becomes more profuse and conspicuous. With further progression, repeated stimulation may be needed to produce any EEG change, and finally the EEG becomes unresponsive. When consciousness is impaired, the EEG becomes slowed, the degree of slowing often (but not always) corresponding to the extent to which consciousness is depressed. The slow-wave activity may be episodic or continuous and, in the former instance, often shows a frontal emphasis and bilateral synchronicity. As the depth of coma increases, the EEG becomes unresponsive to afferent stimuli, and its amplitude diminishes until eventually it becomes flat and featureless, sometimes preceded by a burst-suppression pattern. Such a record should not be taken to indicate that irreversible brain death has occurred; similar changes may be found in severe hypothermia or in coma caused by intoxication with CNS depressant drugs (see Chapter ~H). Records characterized by rhythmic generalized fast activity intermixed with slower rhythms are found in patients who have taken excessive quantities of certain drugs, particularly benzodiazepines or barbiturates. In any severe diffuse encephalopathy, the EEG may show a burst-suppression pattern; its prognostic significance varies with the cause (p. 51). In some comatose patients, the EEG resembles that found during normal wakefulness in that it consists predominantly of activity in the alpha-frequency range (see Fig. 3-27). Such activity is distributed more widely than normal alpha rhythm, however, and unlike the latter is often unresponsive to sensory stimuli. This alpha-pattern coma has been reported particularly in patients with brainstem strokes or following hypoxia caused by cardiopulmonary arrestl59-161 and with drug intoxication.16~ It has also been described after trauma and in association with encephalitis, Reye's syndrome, thalamic tumor, or hyperglycemia with hyperosmolality.l'" Certain differences can be seen in the EEG depending on whether hypoxia or a brainstem lesion is responsible. In particular, the alpha-frequency activity is often maximal posteriorly and may retain some reactivity to

sensory stimulation (e.g., passive eye-opening or pain) with brainstem pathology, whereas it is generalized or more prominent anteriorly and usually unreactive in hypoxic coma. The designation theta-pattern coma is used when widespread, persistent, unreactive thetafrequency activity is present in the EEG of a comatose patient. It has the same significance as alpha-pattern coma. Indeed, in some recordings, the two patterns coexist (alpha-theta coma).163 Alpha-pattern coma has usually been associated with a poor prognosis for survival, although recovery sometimes occurs. J59.16\,16~ In fact, prognosis appears to depend more on the cause of the coma than on the presence of the alpha pattern in itself. Thus, Iragui and McCutchen reviewed the outcome in 94 patients with posthypoxic alpha-pattern coma and found that only 10 of the 86 patients who became comatose following cardiopulmonary arrest survived, whereas 7 of the 8 with a respiratory arrest survived.P? They also found that only 2 of approximately 40 patients with alpha-pattern coma from brainstem lesions survived, whereas all patients in whom the etiology was drug toxicity survived without sequelae. Similar findings have been reported by others. II\~ Kaplan and co-workers found that EEG reactivity in alpha-pattern coma predicted survival in that most patients with reactivity recovered consciousness, whereas most without such reactivity died.16~ Others have noted that alpha- or theta-pattern coma, or alphatheta coma, is a transient clinical phenomenon that usually changes to a more definitive pattern by 5 days after coma-onset. EEG reactivity then implies a more favorable prognosis, and a burst-suppression pattern implies a gloomy outcome.I'" The term spindle coma is used when the EEG shows activity resembling sleep spindles in patients who are comatose. Such spindles are, however, much more diffuse in distribution than normal sleep spindles. It has been postulated that they relate to functional derangement of the midbrain reticular formation. Hansotia and co-workers found that among 370 comatose patients, 22 (5.7 percent) showed spindle activity in the EEG.164 The etiology of coma among these patients was trauma in 8, non traumatic hemorrhage in 4, anoxia in 3, and miscellaneous causes in the remainder. There was no relationship between the occurrence of spindle activity and clinical deficit, depth of coma, or outcome. EEG reactivity seems predictive of outcome, however, as does the cause of the coma.l'" In general, patients with brainstem or cerebral infarction have the poorest outcome, and those with cerebral hypoxia or a cardiopulmonary arrest also have a high likelihood of death or of remaining in a persistent vegetative state.I'" The EEG may be diagnostically helpful in the evaluation of comatose patients. Sometimes e1ectrographic seizure activity is found; in some instances, this may

Electroencephalography: General Principles and Clinical Applications

have been unsuspected clinically, as when nonconvulsive status epilepticus is found in obtunded patients in the intensive care unit or in those recovering from convulsive status epilepticus. In other instances, localized changes in frequency and amplitude suggest the presence of a structural supratentorial lesion. Repetitive complexes are seen in the EEGs of some comatose patients. The character and distribution of the complexes, the degree to which they exhibit a regular periodicity, and the interval between successive complexes may be helpful in suggesting the cause of the coma if the clinical circumstances surrounding the case are obscure. The clinical significance ofPLEDs and BIPLEDs is discussed on page 52; either of these two patterns may be found in obtunded or comatose patients. In general, PLEDs occur most commonly with acute hemispheric lesions (e.g., infarcts or tumors), but occasionally occur with more diffuse disturbances (e.g., hypoxia). In contrast, the most common cause of BIPLEDs is diffuse pathologic involvement of the brain (e.g., anoxic encephalopathy or CNS infections), and this EEG pattern is generally associated with a higher mortality rate than when the epileptiform discharges are restricted to one side. The findings in subacute sclerosing panencephalitis, Creutzfeldt-Jakob disease, herpes simplex encephalitis, and hepatic encephalopathy have been described earlier. Periodic complexes may be found in a number of other disorders, including postanoxic coma. Despite the different EEG changes that may occur in comatose patients, the limited information that can be gleaned from some of these changes, and the advent of more sophisticated investigative techniques, the EEG remains useful for prognostic purposes. In one study, for example, it was found that the outcome of postanoxic coma could be predicted most accurately by combining the results of the Glasgow Coma Scale at 48 hours with the findings obtained by recording somatosensory evoked potentials (see Chapter 25) and, if these data were inconclusive, with the EEG.166 In another study involving a systematic review of the literature, the prognostic value of early neurologic and neurophysiologic findings in anoxic-ischemic coma was examined. The recording of somatosensory evoked potentials was again found to be the most useful method to predict poor outcome, but the EEG was also helpful, with an isoelectric or burst-suppression pattern having a specificity of 100 percent in five of six relevant studies. 167 Young has reviewed the topic and found that a single EEG indicates no possibility of recovery of consciousness with 100 percent specificity when there is complete generalized suppression (less than 10 IlV) after the first day following a cardiac arrest.l'" Less marked suppression, a burst-suppression pattern, periodic complexes, and an alpha-theta pattern usually but not invariably indicate a grim outlook.

79

De-Efferented State (Locked-In Syndrome) Patients with locked-in syndrome may be mistakenly regarded as comatose. Although alert, these patients are mute and quadriplegic and have a supranuclear paralysis of facial and bulbar muscles owing to the pathologic involvement of either major portions of the basis pontis with relative sparing of the tegmentum, or to bilateral midbrain lesions. The EEG is usually normal or shows minor nonspecific abnormalities. In one study of 23 EEGs recorded in 8 patients, 4 were normal, 15 showed scattered theta elements, and 4 showed scattered delta activity without consistent focal or lateralizing features.l'" The EEG was initially normal in 3 patients but later became minimally slowed in 2 of them. In all patients, there was clinical and EEG reactivity to pain, sound, and flicker stimulation. Prolonged polygraphic recording in 7 patients showed that both REM and nonREM sleep states were present in 2, whereas in the remaining 5 there was no REM stage and non-REM sleep showed variable disorganization. Further comment on this disorder is made in Chapter 34.

Miscellaneous Disorders The EEG abnormalities in a number of disorders are of little diagnostic help and are considered here only briefly. SPINOCEREBELLAR DEGENERATION

In spinocerebellar degeneration, the EEG is usually normal, but focal or generalized slowing sometimes occurs. In patients with an associated seizure disorder, epileptiform activity may also be found. PARKINSON'S DISEASE

Mild nonspecific changes, consisting usually of background slowing that mayor may not be lateralized, are occasionally found in patients with Parkinson's disease, but in most patients the EEG is normal. PAROXYSMAL CHOREOATHETOSIS

Epileptiform activity is not a feature of the EEG obtained in patients with paroxysmal choreoathetosis. Records usually are normal or show only an excess of generalized slow activity. HUNTINGTON'S DISEASE

Patients with Huntington's disease characteristically have a low-voltage, featureless record in which the

80

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

predominant background activity is sometimes in the beta range. Irregular slow activity, however, may be present in some cases, whereas in others the record is normal. Paroxysmal disturbances are occasionally found. Despite early hopes to the contrary, the EEG findings have no value in determining which of the apparently unaffected offspring of patients with this inherited disorder will go on to develop the disease. SYDENHAM'S CHOREA

EEG abnormalities are common but nonspecific in Sydenham's chorea, consisting usually of diffuse slow activity in the theta- or delta-frequency ranges. Focal or lateralized slow-wave disturbances may also be found, and epileptiform discharges are seen occasionally. Improvement occurs with time but does not necessarily parallel the time course of clinical changes. HEPATOLENTICULAR DEGENERATION

The EEG is normal in most patients with hepatolenticular degeneration, but in others either generalized or paroxysmal slowing is found, sometimes together with spike discharges or sharp waves. In uncomplicated cases, the presence ofEEG abnormalities shows no clear correlation with the clinical or biochemical state, and patients with predominantly hepatic or neural involvement show similar changes. Abnormalities are found most often when there are complications of the disease; these improve with clinical im provemen t. PROGRESSIVE SUPRANUCLEAR PALSY

In one study of 22 patients with this disorder, the EEG was normal in 12; among the remainder, there was an excess of theta activity in 8, minor lateralizing asymmetries in 1, and frontal intermittent rhythmic delta activity in I. There were no characteristic EEG features of the disorder. 170 TRANSIENT GLOBAL AMNESIA

Most EEGs recorded during or shortly after an episode of transient global amnesia are completely normal. In one study, EEGs were obtained during 13 episodes of transient global amnesia and were entirely normal in 8; none contained epileptiform activity or electrographic seizures.!"' In this same study, EEGs were also obtained after 103 amnestic episodes in 96 patients with isolated transient global amnesia: Most of the records were normal, but mild or moderate and nonfocal abnormalities were found in some patients. Epileptiform activity was not found unless patients also had unequivocal seizure disorders.

Psychiatric Disorders Although a vast literature has accumulated on the EEG findings in patients with psychiatric disorders, there is little evidence that the EEG is of any use in the diagnosis and management of such patients, apart from when it suggests the possibility of an organic disturbance.

REFERENCES 1. Nuwer MR, Comi G, Emerson Ret al: IFCN standards for digital recording of clinical EEG. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl, 52:11, 1999 2. Cooper R, Osselton jW, Shaw JC: EEG Technology. Butterworths, London, 1974 3. Goodin DS, Aminoff MJ, Laxer KD: Detection of epileptiform activity by different noninvasive EEG methods in complex partial epilepsy. Ann Neurol, 27:330, 1990 4. Sadler RM, GoodwinJ: Multiple electrodes for detecting spikes in partial complex seizures. Can J Neural Sci, 16:326, 1989 5. Patel VM, Maulsby RL: How hyperventilation alters the electroencephalogram: a review of controversial viewpoints emphasizing neurophysiological mechanisms. J Clin Neurophysiol, 4:101, 1987 6. Buzsaki G, Traub RD, Pedley TA: The cellular basis of EEG activity. P: 1. In Ebersole JS, Pedley TA (eds): Current Practice of Clinical Electroencephalography. 3rd Ed. Lippincott Williams & Wilkins, Philadelphia, 2003 7. Ebersole JS: Cortical generators and EEG voltage fields. p. 12.1n EbersoleJS, Pedley TA (eds): Current Practice of Clinical Electroencephalography. 3rd Ed. Lippincott Williams & Wilkins, Philadelphia, 2003 8. Alarcon G, Guy CN, Binnie CD et al: Intracerebral propagation of interictal activity in partial epilepsy: implications for source localisation. J Neurol Neurosurg Psychiatry, 57:435, 1994 9. Gloor P: Neuronal generators and the problem of localization in electroencephalography: application of volume conductor theory to electroencephalography. J Clin Neurophysiol, 2:327, 1985 10. Westmoreland BF, Klass DW: Unusual EEG patterns. J Clin Neurophysiol, 7:209, 1990 11. Kojelka jW, Pedley TA: Beta and mu rhythms. J Clin Neurophysiol, 7:191, 1990 12. Brenner RP, Atkinson R: Generalized paroxysmal fast activity: electroencephalographic and clinical features. Ann Neurol, 11:386, 1982 13. Rhee RS, Goldensohn ES, Kim RC: EEG characteristics of solitary intracranial lesions in relationship to anatomical location. Electroencephalogr Clin Neurophysiol, 38:553, 1975 14. Gloor P, Ball G, Schaul N: Brain lesions that produce delta waves in the EEG. Neurology, 27:326, 1977 15. Gilmore PC, Brenner RP: Correlation of EEG, computerized tomography, and clinical findings: study of 100 patients with focal delta activity. Arch Neural, 38:371, 1981

Electroencephalography: General Principles and Clinical Applications

16. Marshall DW, Brey RL, Morse MW: Focal and/or lateralized polymorphic delta activity. Association with either "normal" or "nonfocal" computed tomographic scans. Arch Neurol, 45:33, 1988 17. Schaul N. Green L, Peyster R et al: Structural determinants of electroencephalographic findings in acute hemispheric lesions. Ann Neurol, 20:703, 1986 18. Gloor p. Kalabay 0, Giard N: The electroencephalogram in diffuse encephalopathies: electroencephalographic correlates of grey and white matter lesions. Brain, 91:779,1968 19. Cobb WA, Guiloff RJ, CastJ: Breach rhythm: the EEG related to skull defects. Electroencephalogr Clin Neurophysiol, 47:251,1979 20. Bahamon-Dussan JE, Celesia GG, Grigg-Damberger MM: Prognostic significance of EEG triphasic waves in patients with altered state of consciousness. J Clin Neurophysiol, 6:313,1989 21. Fisch BJ, Klass DW: The diagnostic specificity of triphasic wave patterns. Electroencephalogr Clin Neurophysiol, 70:1,1988 22. Karnaze DS, Bickford RG: Triphasic waves: a reassessment of their significance. Electroencephalogr Clin Neurophysiol, 57:193, 1984 23. Reutens DC, Dunne lW, Gubbay S8: Triphasic waves in accidental hypothermia. Electroencephalogr Clin Neurophysiol, 76:370, 1990 24. River Y, Zelig 0: Triphasic waves in myxedema coma. Clin Electroencephalogr, 24: 146, 1993 25. Blatt I, Brenner RP: Triphasic waves in a psychiatric population: a retrospective study. J Clin Neurophysiol, 13:324, 1996 26. Lancman ME, Marks S, Mahmood K et al: Atypical triphasic waves associated with the use of pentobarbital. Electroencephalogr Clin Neurophysiol, 102:175, 1997 27. Ogunyemi A: Triphasic waves during post-ictal stupor. Can J Neurol Sci, 23:208, 1996 28. Maulsby RL: Some guidelines for assessment of spikes and sharp waves in EEG tracings. Am J EEG Technol, II :3, 1971 29. Reiher J, Beaudry M, Leduc CP: Temporal intermittent rhythmic delta activity (TIRDA) in the diagnosis of complex partial epilepsy: sensitivity, specificity and predictive value. CanJ Neurol Sci, 16:398, 1989 30. Westmoreland BF: Epileptiform electroencephalographic patterns. Mayo Clin Proc, 71:501, 1996 31. Gastaut H: Benign epilepsy of childhood with occipital paroxysms. Curr Probl Epilepsy, 2:159, 1985 32. Prince DA: Neuronal correlates of epileptiform discharges and cortical DC potentials. p. 56. In Creutzfeldt 0 (ed): Handbook of Electroencephalography and Clinical Neurophysiology. Vol 2. Elsevier, Amsterdam, 1974 33. Gloor P, Fariello RG: Generalized epilepsy: some of its cellular mechanisms differ from those of focal epilepsy. Trends Neurosci, 11:63, 1988 34. Gloor P: Epilepsy: relationship between electrophysiology and intracellular mechanisms involving second messengers and gene expression. Can J Neurol Sci, 16:8, 1989 35. Spencer SS, Spencer DD, Williamson PD et al: Effects of corpus callosum section on secondary bilaterally syn-

81

chronous interictal EEG discharges. Neurology, :~5: 1689, 1985 36. Marsan CA, Lewis WR: Pathologic findings in patients with "centrencephalic" electroencephalographic patterns. Neurology, 10:922, 1960 37. Zivin L, Marsan C.A: Incidence and prognostic significance of "epileptiform" activity in the EEG of nonepileptic subjects. Brain, 91:75 I, 1968 38. Goodin DS, Aminoff MJ: Does the interictal EEG have a role in the diagnosis of epilepsy? Lancet, 1:837, 1984 39. Schaul N, Gloor P, GotmanJ: The EEG in deep midline lesions. Neurology, 31:157, 1981 40. Schaul N, Lueders H, Sachdev K: Generalized, bilaterally synchronous bursts of slow waves in the EEG. Arch Neurol, 38:690, 1981 41. Fariello RG, Orrison W, Blanco G et al: Neuroradiological correlates offrontally predominant intermittent rhythmic delta activity (FIRDA). Electroencephalogr Clin Neurophysiol, 54:194,1982 42. McCarty GE, Marshall DW: Transient eyelid opening associated with postanoxic EEG suppression-burst pattern. Arch Neurol, 38:754, 1981 43. Jordan JE, Parrish DC, Cliett JB et al: Suppression burst associated with eye opening. Arch Neurol, 39:602, 1982 44. Simon RP, Aminoff MJ: Electrographic status epilepticus in fatal anoxic coma. Ann Neurol, 20:351, 1986 45. Reeves AL, Westmoreland BF, Klass DW: Clinical accompaniments of the burst-suppression EEG pattern. J Clin Neurophysiol, 14:150, 1997 46. Jumao-as A, Brenner RP: Myoclonic status epilepticus: a clinical and electroencephalographic study. Neurology, 40:1199, 1990 47. Pourmand R: Burst-suppression pattern with unusual correlates. Clin Electroencephalogr, 25: 160, 1994 48. Steriade M, Amzica F, Contreras D: Cortical and thalamic cellular correlates of electroencephalographic burstsuppression. Electroencephalogr Clin Neurophysiol, 90:1, 1994 49. Pohlmann-Eden B, Hoch DB, Cochius JI et al: Periodic lateralized epileptiform discharges: a critical review. J Clin Neurophysiol, 13:519, 1996 50. Neufeld MY, Vishnevskaya S, Treves TA et al: Periodic lateralized epileptiform discharges (PLEDs) following stroke are associated with metabolic abnormalities. Electroencephalogr Clin Neurophysiol, 102:295, 1997 51. Greenberg DA, Weinkle DJ, Aminoff ~J: Periodic EEG complexes in infectious mononucleosis encephalitis. J Neurol Neurosurg Psychiatry, 45:648, 1982 52. Au \o\J, Gabor AJ, Vijayan N et al: Periodic lateralized epileptiform complexes (PLEDs) in Creutzfeldt-jakob disease. Neurology, 30:611,1980 53. Westmoreland BF, Klass DW, Sharbrough FW: Chronic periodic lateralized epileptiform discharges. Arch Neural. 43:494,1986 54. Snodgrass SM, Tsuburaya K, Ajrnone-Marsan C: Clinical significance of periodic lateralized epileptiform discharges: relationship with status epilepticus. .J Clin Neurophysiol, 6:159, 1989 55. Terzano MG, Parrino L, Mazzucchi A et al: Confusional states with periodic lateralized epileptiform discharges

82

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

(PLEDs): a peculiar epileptic syndrome in the elderly. Epilepsia, 27:446,1986 56. Walsh .1M, Brenner RP: Periodic lateralized epileptiform discharges: long-term outcome in adults. Epilepsia, 28:533, 1987 57. Reiher], Rivest .I, Grand'Maison F et al: Periodic lateralized epileptiform discharges with transitional rhythmic discharges: association with seizures. Electroencephalogr Clin Neurophysiol, 78:12, 1991 58. de la Paz D, Brenner RP: Bilateral independent periodic lateralized epileptiform discharges: clinical significance. Arch Neurol, 38:713, 1981 59. Synek VM: The low-voltage electroencephalogram. Clin Eleetroencephalogr, 14:102, 1983 60. Vogel F: The genetic basis of the normal human electroencephalogram (EEG). Hum Genet, 10:91, 1970 61. White jC, Langston.JW, Pedley TA: Benign epileptiform transients of sleep. Neurology, 27:1061, 1977 62. Reiher j, Lebel M: Wicket spikes: clinical correlates of a previously undescribed EEG pattern. Can j Neurol Sci, 4:39,1977 63. O'Brien 1], Sharbrough FW, Westmoreland BF et al: Subclinical rhythmic e1ectrographic discharges of adults (SREDA) revisited: a study using digital EEG analysis. j Clin Neurophysiol, 15:493, 1998 64. Westmoreland BF,Klass DW: A distinctive rhythmic EEG discharge of adults. Electroencephalogr Clin Neurophysiol, 51:186, 1981 65. Westmoreland BF, Klass DW: Unusual variants of subclinical rhythmic electrographic discharge of adults (SREDA). Electroencephalogr Clin Neurophysiol, 102:1,1997 66. Metrakos K, Metrakos jD: Genetics of convulsive disorders. II. Genetic and electroencephalographic studies in centrencephalic epilepsy. Neurology, 11:474, 1961 67. Gerken H, Doose H: On the genetics of EEG-anomalies in childhood. III. Spikes and waves. Neuropadiatrie, 4:88,1973 68. Doose H, Gerken H: On the genetics ofEEG-anomalies in childhood. lV. Photoconvulsive reaction. Neuropadiatrie, 4:162, 1973 69. Heijbel .I, Blom S, Rasmuson M: Benign epilepsy of childhood with centrotemporal EEG foci: a genetic study. Epilepsia, 16:285, 1975 70. Gerken H, Doose H: On the genetics ofEEG-anomalies in childhood. n. Occipital 2-4/s rhythms. Neuropadiatrie, H37,1972 71. Doose H, Gerken H, Volzke E: On the genetics of EEGanomalies in childhood. I. Abnormal theta rhythms. Neuropadiatrie, 3:386, 1972 72. Torres F, Faoro A, Loewenson R et al: The electroencephalogram of elderly subjects revisited. Electroencephalogr Clin Neurophysiol, 56:391, 1983 73. Arenas AM, Brenner RP, Reynolds CF: Temporal slowing in the elderly revisited. Amj EEG Technol, 26:105,1986 74. Klass DW, Brenner RP: Electroencephalography of the elderly.j Clin Neurophysiol, 12:116, 1995 75. Obrist WD: Problems of aging. p. 275. In Chatrian GE, Lairy GC (eds): Handbook of Electroencephalography and Clinical Neurophysiology. Vol 6A. Elsevier, Amsterdam, 1976

76. Oken BS, Kaye jA: Electrophysiologic function in the healthy, extreme old. Neurology, 42:519,1992 77. Cantor FK, I1bag F: Facilitation of photic response by focal lesions. Electroencephalogr Clin Neurophysiol, 34:77, 1973 78. Coull BM, Pedley TA: Intermittent photic stimulation: clinical usefulness of non-convulsive responses. Electroencephalogr Clin Neurophysiol, 44:35, 1978 79. Fisch Bj, Hauser WA, BrustjCM et al: The EEG response to diffuse and patterned photic stimulation during acute untreated alcohol withdrawal. Neurology, 39:434, 1989 80. Reilly EL, Peters jF: Relationship of some varieties of electroencephalographic photosensitivity to clinical convulsive disorders. Neurology, 23:1050, 1973 81. jayakar P, Chiappa KH: Clinical correlations of photo paroxysmal responses. Electroencephalogr Clin Neurophysiol, 75:251,1990 82. So EL, Ruggles KH, Ahmann PAet al: Prognosis of photoparoxysmal response in nonepileptic patients. Neurology, 43:1719, 1993 83. Reeves AL, Klass DW: Frequency asymmetry of sleep spindles associated with focal pathology. Electroencephalogr Clin Neurophysiol, 106:84, 1998 84. Markand ON: Slow spike-wave activity in EEG and associated clinical features: often called "Lennox" or "Lennox-Gastaut" syndrome. Neurology, 27:746, 1977 85. Blume wr, David RB, Gomez MR: Generalised sharp and slow wave complexes: associated clinical features and long-term follow-up. Brain, 96:289, 1973 86. Ellingson R], Wilken K, Bennett DR: Efficacy of sleep deprivation as an activation procedure in epilepsy patients. j Clin Neurophysiol, 1:83, 1984 87. Rowan Aj, Veldhuisen R], Nagelkerke NJD: Comparative evaluation of sleep deprivation and sedated sleep EEGs as diagnostic aids in epilepsy. Electroencephalogr Clin Neurophysiol, 54:357,1982 88. Fountain NB, KimjS, Lee SI: Sleep deprivation activates epileptiform discharges independent of the activating effects of sleep. j Clin Neurophysiol, 15:69, 1998 89. Noriega-Sanchez A, Markand ON: Clinical and electroencephalographic correlation of independent multifocal spike discharges. Neurology, 26:667, 1976 90. King MA, Newton MR,jackson GD et al: Epileptology of the first-seizure presentation: a clinical, electroencephalographic, and magnetic resonance imaging study of 300 consecutive patients. Lancet, 352: 1007, 1998 91. Gotman j, Koffler Dj: Interictal spiking increases after seizures but does not after decrease in medication. Electroencephalogr Clin Neurophysiol, 72:7,1989 92. Weir B: The morphology of the spike-wave complex. Electroencephalogr Clin Neurophysiol, 19:284, 1965 93. Pedley TA: Interictal epileptiform discharges: discriminating characteristics and clinical correlations. Am .I EEG Technol, 20:101, 1980 94. Panayiotopoulos CP, Obeid T, Waheed G: Absences in juvenile myoclonic epilepsy: a clinical and video-electroencephalographic study. Ann Neurol, 25:39 I, 1989 95. Michelucci R, Rubboli G, Passarelli D et al: Electroclinical features of idiopathic generalised epilepsy with persisting absences in adult life. j Neural Neurosurg Psychiatry, 61:471, 1996

Electroencephalography: General Principles andClinical Applications 96. Panayiotopoulos CP, Chroni E, Daskalopoulos C et al: Typical absence seizures in adults: clinical, EEG, videoEEG findings and diagnostic/syndromic considerations. Neurol Neurosurg Psychiatry, 55:1002,1992 97. Devinsky 0, Sato S, Kufta CV et al: Electroencephalographic studies of simple partial seizures with subdural electrode recordings. Neurology, 39:527, 1989 9H. Foldvary N, Lee N, Thwaites G et al: Clinical and electrographic manifestations of lesional neocortical temporal lobe epilepsy. Neurology, 49:757,1997 99. f:bersole JS, Pacia SV: Localization of temporal lobe foci by ictal EEG patterns. Epilepsia, 37:386, 1996 100. Williamson PD, FrenchJA, Thadani VM et al: Characteristics of medial temporal lobe epilepsy: II. Interictal and ictal scalp electroencephalography, neuropsychological testing, neuroimaging, surgical results, and pathology. Ann Neurol, 34:781, 1993 101. Schmidt D: The influence of antiepileptic drugs on the electroencephalogram: a review of controlled clinical studies. p. 453. In Buser PA, Cobb WA, Okuma T (eds): Kyoto Symposia (Electroencephalogr Clin Neurophysiol Suppl 36). Amsterdam, Elsevier, 1982 102. Marciani MG, Spanedda F, Bassetti MA et al: Effect of lamotrigine on EEG paroxysmal abnormalities and background activity: a computerized analysis. Br J Clin Pharrnacol, 42:621, 1996 103. Marciani MG, Stanzione 1', Maschio M et al: EEG changes induced by vigabatrin monotherapy in focal epilepsy, Acta Neurol Scand, 95:115,1997 104. Duncan JS: Antiepileptic drugs and the electroencephalogram. Epilepsia, 28:259, 1987 105. Lowenstein DH, Aminoff M]: Clinical and EEG features of status epilepticus in comatose patients. Neurology, 42:100, 1992 106. DeLorenzo RJ, Waterhouse EJ, Towne AR et al: Persistent nonconvulsive status epilepticus after the control of convulsive status epilepticus. Epilepsia, 39:833, 1998 107. Towne AR, Waterhouse EJ, Boggs JG et al: Prevalence of nonconvulsive status epilepticus in comatose patients. Neurology, 54:340, 2000 lOR. Vespa I'M, Nuwer MR, Nenov Vet al: Increased incidence and impact of nonconvulsive and convulsive seizures after traumatic brain injury as detected by continuous electroencephalographic monitoring. J Neurosurg, 91:750,1999 109. Young GB, Jordan KG, Doig GS: An assessment of non-convulsive seizures in the intensive care unit using continuous EEG monitoring: an investigation of variables associated with mortality. Neurology, 47:83, 1996 110. Celesia GG, Grigg MM, Ross E: Generalized status myoclonic us in acute anoxic and toxic-metabolic encephalopathies. Arch Neurol, 45:781, 1988 III. Thomas JE, Reagan TJ, Klass DW: Epilepsia partialis continua: a review of 32 cases. Arch Neurol, 34:266, 1977 112. Sato S, Dreifuss FE, Penry JI{; Prognostic factors in absence seizures. Neurology, 26:788, 1976 113. Berg AT, Shinnar S: Relapse following discontinuation of antiepileptic drugs: a meta-analysis. Neurology, 44:601,1994

.r

83

114. Tinuper P, Avoni P, Riva R et al: The prognostic value of the electroencephalogram in antiepileptic drug withdrawal in partial epilepsies. Neurology, 47:76, 1996 115. Aminoff MJ, Scheinman MM, Griffin JC et al: Electrocerebral accompaniments of syncope associated with malignant ventricular arrhythmias. Ann Intern Med, 108:791, 1988 116. Clute HL, Levy \\j: Electroencephalographic changes during brief cardiac arrest in humans. Anesthesiology, 73:821, 1990 117. Brenner RP: Electroencephalography in syncope. J Clin Neurophysiol, 14:197, 1997 118. Young GB, Bolton CF, Archibald YM et al: The electroencephalogram in sepsis-associated encephalopathy. J Clin Neurophysiol, 9:145, 1992 119. Ch'ien LT, Boehm RM, Robinson H et al: Characteristic early electroencephalographic changes in herpes simplex encephalitis. Arch Neurol, 34:361, 1977 120. Westmoreland BF, Gomez MR, Blume WT: Activation of periodic complexes of subacute sclerosing panencephalitis by sleep. Ann Neurol, I: 185, 1977 121. Markand ON, PansziJG: The electroencephalogram in subacute sclerosing panencephalitis. Arch Neurol, 32:719, 1975 122. Fariello R, Black]: Pseudoperiodic bilateral EEG paroxysms in a case of phencyclidine intoxication. J Clin Psychiatry, 39:579, 1978 123. Cambier DM, Kantarci K, Worrell GA et al: Lateralized and focal clinical, EEG, and FLAIR MRI abnormalities in Creutzfeldt-Jakob disease. Clin Neurophysiol, 114:1729, 2003 124. Furlan AJ, Henry CE, Sweeney PJ et al: Focal EEG abnormalities in Heidenhain's variant of Jakob-Creutzfeldt disease. Arch Neurol, 38:312, 1981 125. Bortone E, Bettoni L, Giorgi C et al: Reliability of EEG in the diagnosis of Creutzfeldt-Jakob disease. Electroencephalogr Clin Neurophysiol, 90:323, 1994 126. Levy SR, Chiappa KH, Burke C.J et al: Early evolution and incidence of electroencephalographic abnormalities in Creutzfeldt-Jakob disease. J Clin Neurophysiol, 3:1,1986 127. Tietjen GE, Drury I: Familial Creutzfeldt-Jakob disease without periodic EEG activity. Ann Neurol, 28:585, 1990 128. Zeidler M, Stewart GE, Barraclough CR et al: New variant Creutzfeldt-Jakob disease: neurological features and diagnostic tests. Lancet, 350:903, 1997 129. Will RG, Zeidler M, Stewart GE et al: Diagnosis of new variant Creutzfeldt-Jakob disease. Ann Neurol, 47:575, 2000 130. Kuzuhara S, Kanazawa I, Sasaki H et al: GerstmannStraussler-Scheinker's disease. Ann Neurol, 14:216, 1983 131. Smith SJM, Kocen RS: A Creutzfeldt-Jakob like syndrome due to lithium toxicity. J Neural. Neurosurg Psychiatry, 51:120, 1988 132. HormesJT, Benarroch EE, Rodriguez M et al: Periodic sharp waves in baclofen-induced encephalopathy. Arch Neurol, 45:814, 1988 133. Gordon MF, Abrams RI, Rubin DB et al: Bismuth subsalicylate toxicity as a cause of prolonged encephalopathy with myoclonus. Mov Disord, 10:220, 1995

84

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

134. Petty GW, Labar DR, Fisch B] et al: Electroencephalography in lacunar infarction. ] Neurol Sci, 134:47, 1995 135. Hinchey], Chaves C, Appignani B et al: A reversible posterior leukoencephalopathy syndrome. N Engl] Med, 334:494, 1996 136. Bakshi R, Bates YE, Mechtler LL et al: Ocipital lobe seizures as the major clinical manifestation of reversible posterior leukoencephalopathy syndrome: magnetic resonance imaging findings. Epilepsia, 39:1381, 1998 137. Vespa PM, Nuwer M, Juhasz C et al: Early detection of vasospasm after acute subarachnoid hemorrhage using continuous EEG ICU monitoring. Electroencephalogr Clin Neurophysiol, 103:607, 1997 138. Brenner RP. Sharbrough FW: Electroencephalographic evaluation in Sturge-Weber syndrome. Neurology, 26:629, 1976 139. Kinast M, Lueders H, Rothner AD et al: Benign focal epileptiform discharges in childhood migraine (BFEDC). Neurology, 32:1309,1982 140. Gronseth GS, Greenberg MK: The utility of the electroencephalogram in the evaluation of patients presenting with headache: a review of the literature. Neurology, 45:1263, 1995 141. Pedley TA, Tharp BR, Herman K: Clinical and electroencephalographic characteristics of midline parasagittal foci. Ann Neurol, 9:142, 1981 142. Pasquier F, Petit H: Frontotemporal dementia: its rediscovery. Eur Neurol, 38:1, 1997 143. Koralnik lj, Beaumanoir A, Hausler R et al: A controlled study of early neurologic abnormalities in men with asymptomatic human immunodeficiency virus infection. N Engl] Med, 323:864, 1990 144. Goodin DS, Aminoff M], Chernoff DN et al: Long latency event-related potentials in patients infected with human immunodeficiency virus. Ann Neurol, 27:414, 1990 145. Newton TF, Leuchter AF, Miller EN et al: Quantitative EEG in patients with AIDS and symptomatic HIV infection. Clin Electroencephalogr, 25:18, 1994 146. Harrison M], Newman SP, Hall-Craggs MA et al: Evidence of CNS impairment in HIV infection: clinical, neuropsychological, EEG, and MRI/MRS study.] Neurol Neurosurg Psychiatry, 65:301, 1998 147. Gabuzda DH, Levy SR, Chiappa KH: Electroencephalography in AIDS and AIDS-related complex. Clin Electroencephalogr, 19:1, 1988 148. MacGillivrayBB:The EEG in liver disease. P: 26. In Glaser GH (ed): Handbook of Electroencephalography and Clinical Neurophysiology. Vol 15C. Elsevier, Amsterdam, 1976 149. Steg R, Wszolek ZK: Electroencephalographic abnormalities in liver transplant recipients: practical considerations and review.] Clin Neurophysiol, 13:60, 1996 150. Hughes ]R: Correlations between EEG and chemical changes in uremia. Electroencephalogr Clin Neurophysiol, 48:583, 1980 151. Lederman R], Henry CE: Progressive dialysisencephalopathy. Ann Neurol, 4:199,1978 152. Hughes]R, Schreeder MT: EEG in dialysis encephalopathy. Neurology, 30:1148, 1980

153. Malow BA, Reese KB,Sato S et al: Spectrum ofEEG abnormalities during clozapine treatment. Electroencephalogr Clin Neurophysiol, 91:205,1994 154. Freudenreich 0, Weiner RD, McEvoy ]P: Clozapineinduced electroencephalogram changes as a function of clozapine serum levels. Bioi Psychiatry, 42:132,1997 155. Heming RI, Jones RT, Hooker WD et al: Cocaine increases beta activity: a replication and extension of Hans Berger's historic experiments. Electroencephalogr Clin Neurophysiol, 60:470, 1985 156. Heming RI, Guo X, Better WE et al: Neurophysiological signs of cocaine dependence: increased electroencephalogram beta during withdrawal. Biol Psychiatry. 41:1087, 1997 157.]ennett B, van de Sande ]: EEG prediction of posttraumatic epilepsy. Epilepsia, 16:251, 1975 158. Vespa PM, Boscardin \\j, Hovda DA et al: Early and persistent impaired percent alpha variability on continuous electroencephalography monitoring as predictive of poor outcome after traumatic brain injury.] Neurosurg. 97:84,2002 159. Iragui \J, McCutchen CB: Physiologic and prognostic significance of "alpha coma." ] Neurol Neurosurg Psychiatry, 46:632, 1983 160. Westmoreland B, Klass DW, Sharbrough FW et al: Alphacoma: electroencephalographic, clinical, pathologic. and etiologic correlations. Arch Neurol, 32:713,1975 161. Austin E], Wilkus R], Longstreth WT: Etiology and prognosis of alpha coma. Neurology, 38:773, 1988 162. Kaplan PW, Genoud D, Ho TW et al: Etiology, neurologic correlations, and prognosis in alpha coma. Clin Neurophysiol, 110:205, 1999 163. Young GB, Blume WT, Campbell VM et al: Alpha, theta and alpha-theta coma: a clinical outcome study utilizing serial recordings. Electroencephalogr Clin Neurophysiol, 91:93,1994 164. Hansotia P, Gottschalk P, Green P et al: Spindle coma: incidence, clinicopathologic correlates, and prognostic value. Neurology, 31:83, 1981 165. Kaplan PW, Genoud D, Ho TW et al: Clinical correlates and prognosis in early spindle coma. Clin Neurophysiol, III :584, 2000 166. Bassetti C, Bomio F, Mathis] et al: Early prognosis in coma after cardiac arrest: a prospective clinical, electrophysiological, and biochemical study of 60 patients. ] Neurol Neurosurg Psychiatry, 61:610, 1996 167. Zandbergen EG, de Haan R]. Stoutenbeek CP et al: Systematic review of early prediction of poor outcome in anoxic-ischaemic coma. Lancet, 352:1808, 1998 168. Young GB: The EEG in coma.] Clin Neurophysiol, 17:473, 2000 169. Markand 0: Electroencephalogram in "locked-in" syndrome. Electroencephalogr Clin Neurophysiol, 40:529, 1976 170. Fowler CJ, Harrison M]G: EEG changes in subcortical dementia: a study of 22 patients with Steele-RichardsonOlszewski (SRO) syndrome. Electroencephalogr Clin Neurophysiol, 64:301, 1986 171. Miller ]W, Yanagihara T, Petersen RC, Klass DW: Transient global amnesia and epilepsy: electroencephalographic distinction. Arch Neural, 44:629, 1987

CHAPTER

4

Neonatal and Pediatric Electroencephalography JIN S. HAHN and BARRY R. THARP

NEWBORN INFANTS Technical Considerations Normal EEG CNS Maturation and Ontogenic Scheduling Behavioral States Distinctive Patterns in Neonatal EEGs Age-Specific Background Patterns Pitfalls in EEG Interpretation Abnormal EEG Abnormal Background Patterns Dysmature Patterns Value of SerialEEGs in Newborn Infants Abnormal EEG Transients Neonatal Seizures Neonatal Neurologic Disorders with Characteristic EEGs Patterns

The electroencephalogram (EEG) is an important tool in the evaluation of an infant or child with symptoms referable to the central nervous system (CNS). EEGs are used to assess seizure disorders, monitor the progression of a disease, and determine the prognosis for recovery or the development of long-term sequelae. Many EEGs are obtained to "rule out neurologic disorder" or simply to verify a diagnosis that has been well established clinically without much thought from the clinician as to the potential value or yield of such a test. However, the EEG may be overinterpreted by an electroencephalographer with little pediatric experience or training. The clinician, faced with a report of an abnormal EEG, may then feel obliged to undertake additional diagnostic tests, such as computed tomography (CT) or magnetic resonance imaging (MRI), or even to advise unnecessary therapy. This chapter is an overview of pediatric EEG with particular emphasis on newborns and young infants; it illustrates the value of the EEG in the era of CT and MRI, and provides a source of references for those interested in a particular topic. Not all facets of pediatric EEG are reviewed. Rather, the normal patterns seen during infancy and childhood are discussed in detail, with emphasis on those that are often misinterpreted or

OLDER INFANTS AND CHILDREN Normal Patterns During the First 2 Years The EEG During Wakefulness The EEG During Drowsiness The EEG During Sleep Normal Patterns after the First 2 Years EEG Patterns Of Dubious Significance Abnormal EEGs Epilepsy Infectious Diseases Genetic Syndromes with Specific EEG Patterns Progressive Neurologic Syndromes Brain Death

"over-read." Most EEG abnormalities are nonspecific and, in the absence of a clinical history, of little diagnostic value except to indicate the possibility of a pathologic process involving the CNS. Therefore, neurologic disorders that are often associated with specific and, in some cases, pathognomonic EEG patterns are emphasized.

NEWBORN INFANTS The EEG is an important adjunct to the neurologic evaluation of the sick newborn infant. It provides an excellent, noninvasive method of assessing at-risk newborns and of formulating a prognosis for long-term neurologic outcome.' EEGs are commonly utilized in the United States for the evaluation of premature and full-term infants. Improvements in the obstetric management of high-risk pregnancies and major advances in neonatal medicine have significantly decreased the mortality and lessened the morbidity of small, premature infants. Although neonatal mortality has declined since the advent of neonatal intensive care units (NICUs), morbidity statistics have changed less dramatically. The neurologic assessment of critically ill newborns is an important aspect of their initial care 85

86

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

because preservation of cerebral function is the goal of the intensive supportive care. The major long-term sequelae of surviving newborn infants are neurologic in nature: cerebral palsy, mental retardation, and other more subtle motor and cognitive deficits. The major areas in which the EEG can provide unique information in the assessment of newborn fullterm and preterm infants are: I. Diagnosis and treatment of seizures. 2. Evaluation of infants with compromised cerebral function caused by primary neurologic disorders (e.g., hypoxic-ischemic encephalopathy and cerebral infarction) and those with significant systemic disease (e.g., severe respiratory distress syndrome or sepsis) who are at risk for secondary encephalopathies. 3. Determination of conceptional age (CA), which is defined as the estimated gestational age (EGA) plus the legal age. The EGA is the age in weeks of the infant at birth calculated from the date of the mother's last menstrual period or by standardized examination. Legal age is that of the infant since birth. 4. Identification of specific neurologic entities (e.g., intraventricular hemorrhage, periventricular leukomalacia, congenital brain malformations, viral encephalitis, and metabolic encephalopathies). 5. Determination of prognosis and long-term neurologic outcome. The EEG is also used in neonatal research concerned with the study of infant behavior and the assessment of physiologic functions that are dependent on the sleep state. For example, studies of neonatal apnea have demonstrated marked differences in the quantity and quality of apneas in active and quiet sleep." The clinical relevance of these studies is still under investigation. The neurologic disorders of newborn infants are often unique to this period of life. Asphyxia is a common cause of neurologic compromise in the newborn and leads to hypoxic-ischemic encephalopathy. Placental dysfunction, particularly infection, can cause periventricular leukomalacia in premature infants. Metabolic encephalopathies (e.g., hypoglycemia) may cause transient EEG abnormalities yet permanently impair cerebral function. Transient and more benign metabolic abnormalities (e.g., hypocalcemia) cause seizures but are becoming less common with improvements in neonatal intensive care. Meningitis, traumatic lesions (including subarachnoid hemorrhage), congenital malformations of the CNS, drug withdrawal, and inherited disorders of CNS metabolism (e.g., amino or organic acidurias) constitute the bulk of the remaining neurologic syndromes of relevance to the electroencephalographer.

Technical Considerations Many technical aspects of EEG recording are unique to the newborn infant. It is beyond the scope of this chapter to detail the nuances of EEG recording in the newborn nursery, which is reviewed elsewhere." Table 4-1 is a summary of some major technical areas in which EEG recording in neonates differs from that in older children and adults. This recording method is used for preterm and full-term infants until approximately I to 2 months after term. The American Clinical Neurophysiology Society (formerly the American Electroencephalographic Society) has published guidelines for the recording of EEGs in infants." The technologist should have additional training in a laboratory that is already recording from neonates, should become familiar with nursery procedures, and should develop a good rapport with the nursery staff. The technologist should annotate normal and abnormal body and facial movements of newborn infants. In conjunction with the respirogram, electro-oculogram, and EEG, the infant's movements provide the necessary data to determine the behavioral or sleep state. An EEG lacking such clinical observations and polygraphic (non-EEG) variables is extremely difficult, if not impossible, to interpret unless it is grossly abnormal.

Normal EEG There is extensive literature on the normal cerebral electrical activity of the premature and full-term newborn. 5 ,6 Only a summary of the salient characteristics of the normal EEG at each conceptional age is given here. eNS MATURATION AND ONTOGENIC SCHEDULING

Spontaneous electrocerebral activity in premature newborn infants evolves more rapidly than at any other time during human life. A close and consistent relationship exists between the changes in the EEG and the maturational changes of the nervous system. To understand and interpret neonatal EEGs, it is important to recognize the rapid maturational changes that take place in the brain during the last trimester. Such development takes place in an extrauterine environment in the premature infant. There is a rapid enlargement of the brain, with the weight increasing four-fold from 28 weeks to 40 weeks CA. Its appearance changes from a relatively primitive-looking structure before 24 weeks, when the surfaces of the cerebral hemispheres are smooth, to 28 weeks, when the major sulci make their appearance. The sulci and gyri continue to develop, until ultimately the complex brain of the full-term infant is achieved. Rapid maturational changes are also apparent in the neurochemical milieu; in interneuronal

Neonatal and Pediatric Electroencephalography

87

TABLE 4-1 • EEG R.orellq inthe Neonatal PerIod Recording Techniques

Comments

Electrodes and placement Paste or collodion attachment

Electrodes and placement Needle electrodes are never used; collodion may notbe allowed in some nurseries. Small head limits electrodes to frontal, central, occipital, midtemporal, and Cz in premature infants. Frontal sharp waves, delta, and other activity are of higher amplitude in prefrontal than in frontopolar region. Many polygraphic parameters must be measured; allbrain areas can be monitored with one montage. These variables are essential in the determination of the behavior state (awake, active sleep, or quiet sleep) and the recognition of artifacts.

Minimum of 9 scalp electrodes applied in premature infants and entire 10-20 array interm infants Fp1and Fp2 placements replaced byFp3 and Fp4 (halfway between 10-20 placements Fp and F) Atleast 16-channel recording ispreferred Polygraphic (non-EEG) variables recorded routinely: respiration (thoracic, with or without nasal thermistor); extraocular movements (primarily in infants older than 34 weeks CA); and ECG Single montage used for entire recording, particularly with 16-channel recording Paper speed of 15 rnm/sec forentire record: long time constant-between 0.25 and 0.60 sec Frequent annotations bythe technologist of baby's body movement and, in small premature infants, eye movement Technologist attempts to record active and quiet sleep, particularly in older premature and term infants; duration of record may exceed 60 minutes Accurate notation of EGA. CA, recent drug administration, recent changes in blood gases

connectivity, with the development of dendritic trees and synaptogenesis; and in the myelination ofaxons. Despite the apparent simplicity of the anatomy of the premature infant's brain, its repertoire of function is rather extensive. The premature infant of 28 weeks CA is capable of complex, spontaneous motor activity; vigorous crying; and response to stimuli. Behavioral stales are also relatively well developed at 28 weeks CA. In parallel with the developmental maturation of the brain, there is maturation of the bioelectric cerebral activity. According to the rule of ontogenic scheduling, the bioelectric maturation of cerebral activity of the healthy infant occurs in a predictable, time-linked manner. This maturational process is dependent primarily on the age of the brain (i.e., CA) and is independent of the number of weeks of extrauterine life. Therefore, the EEG of a premature infant born at 30 weeks EGA whose legal age is 10 weeks (CA 40 weeks) is similar to that of a 38-week EGA infant who is 2 weeks old. Some minor differences exist in the EEGs of premature infants who mature to term and those of fullterm newborns, but these differences appear to be of little clinical significance. Consequently, an accurate estimate of the CA is essential for the correct interpretation of neonatal EEGs. The EEG patterns of the newborn infant are dependent not only on the CA but also on the behavioral states of the infant during the recording. Distinct EEG

Generalized changes in background activity and state-related changes are more important than exact localization of focal abnormalities. It iseasier to recognize interhemispheric synchrony and slow background activity. Important information for the determination of behavioral state and possible artifact. Presence or absence of well-developed sleep states isimportant for interpretation; some pathologic patterns are seen primarily in quiet sleep. Interpretation isdependent onknowledge of CA of infant. The EEG isvery sensitive to abrupt changes in blood gases and certain medications.

patterns of active (rapid eye movement or REM) and quiet (non-REM) sleep can be identified easily in normal infants by 35 weeks CA and in many infants as early as 27 to 28 weeks CA (Fig. 4-1). BEHAVIORAL STATES

The full-term newborn has easily recognizable waking and sleeping behavioral states that are very similar to those of older children and adults. These states are generally classified as waking, active sleep (REM sleep), quiet sleep (non-REM sleep), and indeterminate or transitional sleep (a sleeping state that cannot be definitely classified as either active or quiet sleep). Active sleep is the most common sleep-onset state in newborn infants and remains so for the first 2 to 4 months of postterm life'; it constitutes approximately 50 percent of the sleeping time in term infants and a slightly higher proportion in premature infants. The duration of active sleep at onset is usually 10 to 20 minutes but may exceed 40 minutes in some newborns. A normal infant is considered awake if the eyes are open. Behavior may vary from quiet wakefulness to crying with vigorous motor activity. Transient eye-closures may accompany crying and also occur during quiet wakefulness. If the eyes remain closed for an extended period of time (usually for more than 1 minute), the infant is considered asleep.

88

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp1-T3

~

~~~~.--.

T3-01

Fp1-T3 T3-01

Fp2-T4

Fp2-T4

T4-02 Fp1-C3

T4-o2 Fp1-C3

C3-01 Fp2-C4

C3-01 Fp2-C4

C4-02A~~ Rasp. Movement

~.~'w!"~------"-""-""""-~'"

C4-02

~VvVvwvvv til-

Movement-----------------

ECG A FIGURE 4·1 II A, Recording of quiet sleep of healthy premature infant (29 weeks CA). Discontinuous background is associated with infrequent limb movements. Note the burst of high-amplitude theta in the left frontotemporal region. Resp., abdominal respiration. Upper movement channel, movement monitor located under baby's body. Lower movement channel, surface EMG on right leg. Calibration: 50 flV; 2 seconds. B, Active sleep recorded in same infant. Continuous background is associated with frequent body and limb movements. Bilateral brushes in the central regions are present in the right half of the tracing. Respirations are slightly less regular than in A. Calibration: 50 flV; 2 seconds.

Active sleep in infants older than 29 weeks CA is characterized by eye-closure, bursts of rapid horizontal and vertical eye movements, irregular respirations, and frequent limb and body movements ranging from brief twitches of a limb to gross movements of one limb or the entire body, grimacing, smiling, frowning, and bursts of sucking. Quiet sleep in infants older than 29 weeks CA is characterized by the infant lying quietly with eyes closed, regular respiration, and the absence of rapid eye movements. There is a paucity of body and limb movements, although occasional gross body movements, characterized by brief stiffening of the trunk and limbs, may occur and are sometimes associated with brief clonic jerks of the lower extremities. DISTINCTIVE PATTERNS IN NEONATAL

EEGs

Certain distinct EEG patterns are common in the newborn, especially in the premature infant. These specific patterns serve as useful findings for determining the CA of the infant's nervous system. Figure 4-2 depicts some of these waveforms.

NORMAL PATIERNS

Fp1-C3 ~AA.J\..(

A.

y~ C3-01 C3-01~~ ••

Fp2-C4

-

:

-:

~ y

C4 02"

..~

~...c~ •. .•• rv

A

""-"""d"l

...

"'\f\J 'V V"" ~ L-..

C4-02

"''"-IVoo,IV~~

Frontal sharp wave

Frontal delta

B

T3-C3

C3-Cz Cz-C4

C4-T4

Theta Bursts (Sawtooth Waves) Occipital theta bursts consist of rhythmic, medium- to high-amplitude, 4- to 7-Hz activity located in the occipital regions. These bursts are present as a physiologic pattern in very premature infants (maximal between 24 and 20 weeks CA) and disappear by 32 weeks CA.8.9 This pattern serves as a useful hallmark in determining CA.

rel-.9~ "f'.. r-; ~ fi. '" -'''\1'V'' V V v V vv

Brushes

C FIGURE 4·1 II A, Bilateral frontal sharp waves (encoches frontales) in healthy full-term infant. B, Frontal rhythmic delta activity (anterior slow dysrhythmia) in healthy preterm infant. C, Central delta brushes in healthy premature infant (35 weeks CA). Calibration: 50 flV; 1 second.

Neonatal and Pediatric Electroencephalography

Temporal theta burst activity is commonly seen in the EEGs of slightly older premature infants. This activity appears at approximately 26 weeks CA, with the maximal incidence between 27 and 31 weeks, and is rarely seen beyond 35 weeks.l'v'! The bursts consist of sharply contoured, high-amplitude (50 to 150 /lV, occasionally 200 to 300 /lV), 4- to 6-Hz activity; they are maximal in the temporal areas, but often become diffuse and, commonly, bilateral and synchronous (see Fig. 4-1), 'These regular theta rhythms originate in the occipital regions in very premature infants and with increasing CA migrate toward temporal regions." The gradient of occipitotemporal maturation of this pattern appears to coincide with the timing of gyral development in these regions.

Delta Brushes Delta brushes are a hallmark of prematurity and are most abundant between 32 and 35 weeks CA (see Fig. 4-2). They consist of short bursts of 8- to 20-Hz rhythmic activity, often with a spindle morphology, superimposed on high-amplitude slow waves (0.8 to 1.5 Hz). The amplitude of the fast activity may range from 10 to 100 /l V (usually 20 to 50 /l V), and that of the slow waves from 25 to 200 /lv. Typically, delta brushes occur asynchronously in the homologous areas of the two hemispheres. Delta brushes predominate in the central (rolandic) region in very young premature infants (younger than 32 weeks CA) and extend to the occipital and temporal regions in older premature infants. Brushes are more abundant in active sleep in younger infants (up to 32 weeks CA) and in quiet sleep in infants older than 33 weeks CA.12,13 Lombroso counted the number of brushes in the parasagittal regions during 5-minute epochs of well-established quiet and active sleep in a group of healthy infants between 31 and 43 weeks CA.13 In infants between 33 and 34 weeks CA, he found an average of 29 brushes per epoch in quiet sleep and 22 per epoch in active sleep. The brushes decreased in number as the infant approached term, were infrequent during the quiet sleep of the term baby (0.8 per epoch), and were virtually absent in active sleep (0.1 per epoch). They disappear during the first few weeks of postterrn life.

89

from one side to the other during the course of the recording. Typical frontal sharp transients appear at 35 weeks CA and persist until several weeks after term (see Fig. 4-2) .14

Rhythmic Frontal Delta Activity (Anterior Slow Dysrhythmia) Rhythmic frontal delta activity consists of bursts and short runs of 1.5- to 4-Hz, 5()- to 200-/lV delta activity that is often monomorphic and of maximal amplitude in the frontal regions (see Fig. 4-2). Like frontal sharp transients, rhythmic frontal delta activity is most prominent during transitional sleep. This activity appears at approximately 37 weeks CA and lasts until approximately 6 weeks after term. Rhythmic frontal delta activity is often intermixed with frontal sharp transients. This activity should be distinguished from prolonged runs of monorhythmic bifrontal delta activity that persists in all sleep states, which is an abnormal finding. Trace Discontinu

In very young premature infants, the electrical activity of the brain is interrupted by long periods of quiescence. This pattern, called trace discontinu, may be present in all states of sleep in very premature infants and persists to some degree in infants of 34 to 36 weeks. During periods of quiescence, the EEG is very low in amplitude (less than 30 uv) and may even be flat at standard amplification. The active periods increase in duration as the CA increases (Fig. 4-3). They are composed of bursts containing monomorphic delta, theta and faster activity, delta brushes, and temporal theta bursts. 60 0

C-

I

0

0

o

0

g

50

"C 0 .;::: Q)

c.

40

CIl :::l

0

:::l

30

.5

"E 0

-

20

c:

10

t,)

CIl Q)

Cl

Frontal Sharp Transients (Encoches Frontales) Frontal sharp transients are biphasic sharp waves (usually surface negative, then positive), often followed by a slow wave; they are of maximal amplitude in the prefrontal regions (Fp3, Fp4) and occur primarily during sleep, particularly during the transition from active to quiet sleep. The typical amplitude is 50 to 150 /lV and the duration of the initial surface-negative componen t is 200 msec. They usually appear bilaterally and synchronously, although they may be asymmetric in amplitude. Sometimes they appear unilaterally, shifting

0

...J

9 0 0 o+--.....,....----..,,..--...,.--.....,....--,..--...,.--~

26-2728-2930-31 32-3334-3536-3738-40 CA (weeks) FIGURE 4-] III Longest continuous period of cerebral activity at various conceptional ages (CAs). The regression line was determined by the least-squares method. (From Hahn JS, Monyer H, Tharp BR: Interburst interval measurements in the EEG of premature infants with normal neurologic outcome. Electroencephalogr Clin Neurophysiol, 73:410, 1989, with permission.)

90

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

The mean duration of interburst intervals in the healthy premature infant decreases with increasing gestational age: The mean interval at 27 to 29 weeks CA is approximately 6 seconds, and at 30 to 34 weeks is approximately 4 to 5 seconds. The interburst intervals should not exceed approximately 30 seconds at any age beyond 27 weeks CA (Fig. 4-4).15,16 In healthy, very preterm infants, the maximum interburst interval duration can be much longer: 126 seconds at 21 to 22 weeks CA, 87 seconds at 23 to 24 weeks CA, and 44 seconds at 25 to 26 weeks CAP

Trace Alternant As the CA increases, the periods of relative electrocerebral inactivity become shorter in duration and the interburst intervals display generalized amplitude attenuation rather than quiescence. This pattern, called trace alternant; is the discontinuous pattern that characterizes the quiet sleep of the full-term newborn (Fig. 4-5). Trace alternant consists of 3- to 6-second bursts of high-amplitude delta and theta activity (l to 6 Hz, 50 to 100 IlV) admixed with lower-amplitude beta and theta activity, which occur at intervals of 3 to 6 seconds. The bursts may also contain scattered isolated sharp transients. The interburst intervals contain diffuse moderate-amplitude (25- to 50-1lV) , mixed-frequency (usually 4- to 12-Hz) activity similar to that occurring during wakefulness and active sleep following quiet sleep.

() Ql

!E.c: 0

~

...

::J

"C

!B E E

::J

.~

~

32 30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 0

0 0

0 0 0

0 0

~

0

0 0 0

8

e

~

9

I e

0

0

6

term, although even in the healthy full-term infan t, occasional short, very-low-amplitude interburst intervals are seen. Trace discontinu, therefore, differs from trace alternant primarily on the basis of the amplitude of the activity during the interburst interval.

Interhemispheric Synchrony Interhemispheric synchrony is defined as the relatively simultaneous appearance in both hemispheres of bursts of cerebral activity during discontinuous portions of the record. In the very young premature infant (younger than 29 weeks CA) the EEG activity during trace discontinu is hypersynchronous, with the degree of synchrony approaching 90 to 100 percent. The degree of interhemispheric synchrony decreases to 60 to 80 percent for the next 4 to 5 weeks (30 to 35 weeks CA) and gradually approaches 100 percent as the infant reaches term. AGE-SPECIFIC BACKGROUND PATTERNS

Normal Patterns at 23 to 26 Weeks CA The EEG in a very preterm infant is consistently discontinuous (trace discontinu) , although some variability in the organization of discontinuous patterns can be observed. 32 30 28 26 () 24 Ql !E.- 22 c: 20 0 ~::J 18 "C 16 !B 14 E ::J 12 E .~ 10 ~ 8 6

0

0

A gradual transition from trace discontinu to trace alternant occurs as the premature infant approaches

4 2 0

§

26-2728-2930-31 32-3334-3536-3738-40

JJ.

JJ. JJ. JJ.

A JJ.

A

JJ.

t

i

A JJ. JJ.

A

t JJ. JJ.

6 JJ.

JJ.

I

A

A

t

JJ. JJ.

i JJ.

JJ. JJ. JJ. JJ.

JJ.

26-2728-29 30-31 32-33 34-3536-37 38-40

CA (weeks)

A

JJ.

CA (weeks)

B

FIGURE 4-4 l1li The maximum interburst interval (IBI) versus CA for (A) type 1 IBIs (interval of at least 2 seconds containing no cerebral activity greater than 151lVin any channel); and B, type 2 IBIs (intervals that are similar to type 1 IBIs except for a brief transient lasting less than 2 seconds in one or two electrodes or continuous cerebral activity greater than 15 IlV in one electrode). (Modified from Hahn j5, Monyer H, Tharp BR: Interburst interval measurements in the EEG of premature infants with normal neurologic outcome. Electroencephalogr Clin Neurophysiol, 73:410, 1989, with permission.)

Neonatal and Pediatric Electroencephalography

T4-02 -""1\.1"'" "'ll\...~-------~--

FP1-C3~----------

C3-01~ :~ Fp2-C4 ~y----------­ L.. C4-02

A Fp1-T3 T3-01 Fp2-T4 T4-02 Fp1-C3 C3-01 Fp2-C4 C4-02

B FICURE 4·5 • A, Burst-suppression (paroxysmal) pattern during quiet state in stuporous, full-term infant with severe hypoxicischemic encephalopathy. B, Trace alternant pattern in quiet sleep of healthy full-term infant. Calibration: 50 IlV; 1 second.

In some phases bursts can be longer and intervals shorter, whereas at other times the discontinuity is more marked." In a study of five subjects in this age group, the maximum interburst interval duration ranged between 20 and 62 seconds and the minimum burst length was 1 to ~ seconds." Another study found that the maximum inierburst interval duration was 87 seconds at 23 to 24 weeks CA and 44 seconds at 25 to 26 weeks CA. 17 Occipital theta bursts are prominent (maximal at 25 weeks CA). Temporal theta bursts can be seen but are less common than in older preterm infants. Some delta brushes are observable, but the incidence is generally low (between 9 and 21 percent of bursts).

Normal Patterns at 27 to 29 Weeks CA The EEG at 27 to 29 weeks CA is characterized by a discontinuous background, trace discontinu, that consists of mixed-frequency activity, primarily in the delta range. Bursts contain runs of high-amplitude, very slow, rhythmic occipital delta activity; and somewhat lower-amplitude, more arrhythmic central and temporal delta

91

activity. 16 The maximum interburst intervals are generally less than 30 seconds.!" Beginning as early as 27 weeks CA, there may be behavioral and EEG differentiation of active and quiet sleep states." During active sleep, REMs are observed with bursts of slow waves admixed with theta activity. During quiet sleep, REMs are absent and the background becomes more discontinuous. High-amplitude temporal theta bursts (see Fig. 4-1) become prominent during this period. Isolated frontopolar slow waves with superimposed faster activity (socalled "delta crests") are also common in this age group;" Delta brushes are present primarily in the central (rolandic) and occipital regions. 16

Normal Patterns at 30 to 32 Weeks CA Two distinct types of background predominate at 30 to 32 weeks CA. A discontinuous background (tmci discontinu) composed of an admixture of occipital delta (1 to 2 Hz, 25 to 100 IlV) and centrotemporal delta-, theta-, and alpha-range activity occurs when the child is in quiet sleep. More continuous background occurs with eye movements and more active body movements, consistent with active sleep. Synchrony during trace discontinu reaches a nadir during this period (approximately 60 to 75 percent, depending on the study). Delta brushes are very abundant at this age; they are located in the occipital, central, and temporal regions with a higher incidence during the continuous portions of the record (active sleep). There are fewer temporal theta bursts than at earlier ages, and the delta crests and occipital theta bursts are less common.

Normal Patterns at 33 to 35 Weeks CA Sleep states at 33 to 35 weeks CA are becoming more clearly defined. Discontinuous activity (trace discontinu) is associated with quiet sleep. During wakefulness and active sleep, continuous background slow activity (primarily 1 to 2 Hz, 25 to 100 IlV) that is maximal in the temporal, central, and occipital regions is more abundant. In general, an increasing amount of lower-amplitude theta and faster activity is also present in all states. Synchrony approaches 80 percent at 35 weeks CA. Delta brushes are still abundant but fewer than at an earlier age, and they are more frequent during quiet sleep. Typical frontal sharp waves iencoches lrontales) appear at 34 to 35 weeks CA. Temporal theta activity should be much less common. I I

Normal Patterns at 36 to 38 Weeks CA Three different EEG patterns predominate at 36 to 38 weeks CA. (1) Continuous, diffuse, low-amplitude,

92

ELEaRODIAGNOSIS IN CLINICAL NEUROLOGY

4- to 6-Hz activity (usually less than 50 flV) characterizes quiet wakefulness (activite moyenne). This activity is often admixed with 50- to 100-flV delta (2- to 4-Hz) activity and is the predominant rhythm of sleep-onset active sleep. Delta brushes are rare in active sleep. (2) A discontinuous pattern (trace alternant) with bursts of mixed-frequency slow activity and occasional delta brushes, separated by low- to medium-amplitude background, typifies quiet sleep. During this period there is a transition from trace discontinu to trace alternant. The trace alternant pattern persists until 4 to 8 weeks after term. (3) The active sleep that follows quiet sleep is characterized by low- to moderate-amplitude theta activity with occasional delta waves at lower amplitude than that occurring during sleep-onset active sleep. Synchrony approximates 90 percent during the discontinuous portions of the EEG. Delta brushes are less abundant during quiet sleep and are virtually absent during active sleep. Frontal sharp transients and rhythmic frontal delta activity are prominent, particularly during the transition from active sleep to quiet sleep.

Normal Patterns at 38 to 42 Weeks CA (Full-Term Newborn) Sleep and waking cycles are well established at 38 to 42 weeks CA. Four distinctive patterns occur during sleep. (1) Low- to medium-amplitude theta activity with superimposed delta activity (2 to 4 Hz, less than 100 flY), the latter appearing as continuous activity or in short runs or bursts, characterizes active sleep, particularly at sleep onset. (2) Diffuse continuous delta activity (0.5 to 2 Hz,

25 to 100 flV) is found at the beginning and end of quiet sleep periods and is occasionally found during long portions of such sleep (slow-wave quiet sleep). (3) Trace alternant is characteristic of well-established quiet sleep. (4) Activite moyenne is a continuous low-amplitude activity (25 to 50 flV) at 1 to 10 Hz (predominantly 4 to 7 Hz) that characterizes wakefulness and active sleep, particularly following a period of quiet sleep. Interhemispheric synchrony during trace alternant approaches 100 percent. Delta brushes are rare in the EEGs of term infants; they occur primarily during quiet sleep. Frontal sharp transients and rhythmic frontal delta activity are abundant, particularly during the transitional period between active sleep and quiet sleep. The amplitude of the background activity is relatively symmetric over the two hemispheres, although transient asymmetries are common, particularly in the temporal regions. Rhythmic theta-alpha bursts, which consist of bursts or short runs (1 to 3 seconds) of sharply contoured, primarily surface-negative, rhythmic theta- and alpharange activity, are common in the central and midline frontocentral areas, particularly during quiet sleep.'? Scattered isolated sharp waves are common in the temporal regions and are less common in the central and midline regions. These sharp transients are usually incorporated within the bursts of the trace alternant. Figure 4-6 summarizes the various EEG patterns that are seen in the neonatal EEG and how these patterns change as the premature infant matures to term. By knowing which maturational patterns become prominent at a particular CA, an estimation of the CA can be determined.

FIGURE 4-6 .. A summary depicting the various maturational patterns seen in the EEGs of preterm and term infants as a function of the conceptional age (CA). These patterns include the number of temporal theta bursts and delta brushes, the durations of interburst intervals and bursts, and the degree of interhemispheric synchrony.

Neonatal and Pediatric Electroencephalography PITFALLS IN EEG INTERPRETATION

Inexperienced electroencephalographers often mistakenly characterize the frontal sharp transients and rhythmic frontal delta activity of the older premature infant and the temporal theta bursts of the very young infant as "epileptiform" or "paroxysmal." The abundant brushes in the EEGs of younger premature infants, particularly if slow paper speed is used, may also give the background a "spiky" or "paroxysmal" appearance. Multifocal sharp waves are also noted, particularly in the temporal and central regions of healthy infants. The various discontinuous background patterns of healthy premature infants must be distinguished from the abnormal paroxysmal patterns discussed in the next section. Transient interhemispheric amplitude asymmetries are common in all age groups, as is asynchrony of the bursts during the discontinuous portions of the tracings in premature infants (approximately 31 to 35 weeks CA). There may be significant asymmetries in the onset of the trace alternant pattern in normal infants, with trace alternant appearing in one hemisphere 1 to 5 minutes before it emerges in the other. 2o •2 1

Abnormal EECi It is beyond the scope of this chapter to discuss all the deviations from normal that are encountered in the EEGs of newborn infants. Rather, emphasis is placed on the patterns most commonly seen, those associated with certain specific neurologic disorders, and those that appear to be of prognostic value. ABNORMAL BACKGROUND PATTERNS

Only the most common EEG abnormalities that occur in the neonatal EEG are discussed in this section. Additional discussions of the prognostic value of the EEG in full-term newborns can be found elsewhere. 1,22- 24 One of the more common causes for a severely abnormal EEG background is hypoxic-ischemic encephalopathy. An EEG performed early can be valuable in predicting the severity of the encephalopathy and prognosis. 25- 27 The following section discusses the various abnormal EEG patterns and their prognostic significance (Table 4-2). Before interpreting the EEG, it is important to consider any acute changes in metabolic status, administration of neuroactive medications, or changes in respiratory status (e.g., oxygenation or pH) that may acutely affect the background EEG. For example, the EEG may become transiently abnormal and even isoelectric during acute hypoxemia. Many drugs used in nurseries (e.g., intravenous diazepam, lorazepam, midazolarn, and opioids) may cause an acute but

93

TABLE 4·2 • Prognostic Significance of Neonatal Background EEG Patterns EEG Background Patterns

Isoelectric Burst-suppression Low-voltage undifferentiated Excessively discontinuous Diffuse slow activity Gross asynchrony

Percent with Favorable Outcome*

0-5 0-15 15-30 40-50 15-20 10-20

*Favorable outcome includes those with normal and mildly abnormal outcomes. See text on the importance ofthe timing ofthe EEG with respect toprognosis. Data from Tharp et al.," Takeuchi and Watanabe,26 and Ortibus etal. 69

reversible depression of the EEG. The synthetic opioid sufentanil increases the discontinuity of the background in premature infants.i" Intravenous morphine infusions may cause excessive discontinuity, sharp waves, and even burst-suppression patterns.P Therefore if these drugs are being administered at the time of the EEG, the record should be interpreted with caution and another record should be obtained after their discontinuance.

Electrocerebral Inactivity (Isoelectric) Electrocerebral inactivity implies that there is no discernible cerebral electrical activity even at high sensitivities. Such an EEG pattern is seen in various clinical settings, most commonly following severe asphyxia, circulatory collapse, and massive intracerebral hemorrhage. It can also be seen in bacterial meningitis, encephalitis, severe malformations of the brain (e.g., hydranencephaly or massive hydrocephalus), and inborn errors of metabolism. The same technical requirements for recording isoelectric records in adults are applied to infants. High sensitivities (equal to or exceeding 21lVImm), long time constants (0.1 seconds or more, or low-frequency filter of 1 Hz), and long interelectrode distances should be used. The technologist should also perform various auditory and nociceptive stimulations to confirm lack of reactivity. In the absence of drug intoxication, acute hypoxemia, hypothermia, and postictal state, this EEG pattern carries a grave prognosis in neonates; however, it does not necessarily indicate "brain death." The vast majority of neonates with inactive EEGs either die in the neonatal period or survive with severe neurologic deficits. In a study of 80 full-term newborns who had EEGs within 24 hours after birth, Pezzani and colleagues found that 19 had "isoelectric" EEGs, and 17 of the 19 (90 percent) died.V Two neonates survived; one had epilepsy and the other was normal at 6 years of age. In neuropathologic studies of the brains of neonates who have inactive EEGs, widespread

94

ELEaRODIAGNOSIS IN CLINICAL NEUROLOGY

encephalomalacia and ischemic neuronal necrosis are found involving the cerebral cortex, corpus striatum, thalamus, midbrain, pons, and medulla. In newborns, the return of EEG activity may be seen after an isoelectric EEG recorded in the first 24 hours of life or immediately after an acute hypoxic-ischemic event. If recovery occurs, the isoelectric EEG is usually followed by other abnormal patterns (particularly, diffusely slow backgrounds), but occasionally a flat EEG persists for many months following the acute neurologic insult. Inactive EEG and Brain Death

There are no universally agreed criteria for brain death determination in newborn infants. The Task Force for the Determination of Brain Death in Children established the criteria for determining brain death in children but did not provide guidelines for diagnosing brain death in infants under the age of 7 days.'? For infants between the ages of 7 days and 2 months, the Task Force recommended that two physical examinations and EEGs separated by 48 hours be performed routinely. During the first week after birth, an isoelectric EEG is not a reliable test for determining brain death. However, Ashwal and Schneider found no reports of any neonates who recovered EEG activity if the baby remained clinically brain-dead." They concluded that if the initial EEG in the newborn shows electrocerebral inactivity in the absence of barbiturates (phenobarbital concentration less than 25 ug/rnl), hypothermia, or cerebral malformations (e.g., hydranencephaly or hydrocephalus), and if the infant's neurologic examination findings remain unchanged after 24 hours, electrocerebral inactivity is confirmatory of brain death. Most neurologists believe that the EEG is a useful adjunct to the determination of brain death if it is performed by an experienced technologist using the standards of the American Clinical Neurophysiology Society (formerly the American Electroencephalographic Society) 32 and is interpreted by an experienced electroencephalographer.

Burst-Suppression (Paroxysmal) Pattern The burst-suppression pattern (see Fig. 4-5, A) is characterized by an isoelectric background interrupted by nonperiodic bursts of abnormal activity: delta and theta with admixed spikes, beta activity, or both; less commonly, bursts or short runs of diffuse or focal alpha or theta activity that is occasionally rhythmic. The bursts, which are usually highly synchronous between hemispheres, contain no age-appropriate activity. In the most severe form, this pattern is invariant and minimally altered by stimuli and persists throughout waking and sleeping states. This abnormal pattern must be differentiated from the discontinuous

patterns seen in the quiet sleep of normal premature infants. Burst-suppression patterns are seen following a varietyofsevere brain insults (e.g., asphyxia, severe metabolic disorders, CNS infections, and cerebral malformation). A burst-suppression pattern can be induced pharmacologically with high doses of barbiturates and other neuroactive medications. In the absence of significant concentrations of neuroactive medications, a burstsuppression pattern is usually associated with a very poor prognosis (85 to 100 percent unfavorable prognosis depending on the study). Infants who have a burstsuppression pattern that changes with stimulation have a somewhat better prognosis." A burst-suppression pattern in the first 24 hours of life that is rapidly replaced by a less severely abnormal EEG is occasionally followed by a normal neurologic outcome. 27,33 Variants of Burst-Suppression Pattern

In a small group of full-term newborns with neonatal hypoxic-ischemic encephalopathy, Sinclair and coworkers compared different types of burst-suppression patterns with outcome." Burst-suppression pattern was defined as a background with bursts lasting 1 to 10 seconds alternating with periods of marked background attenuation (amplitude consistently less than 5 IlV). A modified burst-suppression pattern was defined as a burst-suppression pattern that was not constantly discontinuous throughout the recording, had periods of attenuation that contained activity higher than 5 IlV, or both. Those newborns with a burst-suppression pattern had poor outcome: death in 6 of 15, severe disabilities in 4, and normal outcome in 2 of 9 survivors. The outcome in those with a modified burst-suppression pattern was more favorable: neonatal death in 1 of 8, severe disabilities in 1, and normal outcome in 3 of 7 survivors.f The EEGs were performed within the first week of life, but the timing of the EEG in regards to the timing of the insult was not provided. The study did not assess whether the burst-suppression pattern could be modified with stimulation. Instead of rigidly distinguishing between burstsuppression and other constantly discont.inuous patterns based on the amplitude of activity during the suppressions, Biagioni and colleagues examined EEG parameters of discontinuity and correlated these scores with the outcome." In 32 full-term infants with hypoxicischemic encephalopathy who had an EEG with a constantly discontinuous pattern, they noted the minimum burst duration (activity greater than 45 IlV), maximum interburst interval duration (interburst. interval was defined as a period of activity less than 45 IlV), and mean interburst interval amplitude. The best indicators predictive of a normal outcome were maximum interburst interval duration shorter than 10 seconds, mean interburst interval amplitude greater than 25 IlV, and

Neonatal and Pediatric Electroencephalography

minimum burst duration longer than 2 seconds.P The maximum interburst interval duration correlated with the severity of the hypoxic-ischemic encephalopathy. It was also significantly lengthened (more than double) in children who received phenobarbital. The timing of the EEG was also an important factor in determining pn >!{nosis; of the nine subjects who had a constan tly discoutinuous EEG after the eighth day from birth, none had a normal outcome. All of the 32 EEGs in this study would be compatible with the definition of modified burst-suppression pattern, as defined by Sinclair and colleagues.i" because there were no records with interburst interval amplitudes less than 5 llV. Menache and co-workers also examined several EEG parameters of discontinuity in 43 term or near-term infants who had a variety of neurologic disorders." They included EEGs with constantly or transiently discontinuous patterns. Only 7 had a burst-suppression pattern as defined ahove. A predominant interburst interval duration (defined as the duration of more than 50 percent: of all interburst intervals with amplitudes less than 211 llV) of longer than 30 seconds correlated with the occurrence of both unfavorable neurologic outcome and subsequent epilepsy.'" Infants with this finding had 100 percent probability of severe neurologic disabilities or death (all had hypoxic-ischemic encephalopathy) and an 86 percent chance of developing subsequent epilepsy.

Excessively Discontinuous Background The background EEG is normally discontinuous in very premature infants, with periods of total absence of cerebral activity lasting many seconds. With the maturation of the brain, the duration of these flat periods, or interburst intervals, decreases as the preterrn infant approaches termy),li,37 Conversely, the duration of the longest period of continuous EEG increases with CA (see Figs. 4-3 and 4-4) .1:\17,'\i In healthy, very preterm infants the maximum interburst interval duration was 126 seconds at 21 to 22 weeks CA, 87 seconds at 23 to 24 weeks CA, and 44 seconds at 25 to 26 weeks CA.'? In premature infants at 26 to 30 weeks CA, the maximum interburst interval duration was usually less than 15 seconds, but in a few infants it reached 20 to 30 seconds. I!) An excessively discontinuous background is, therefore, defined as a pattern that exceeds these age-dependent norms. Unlike the burst-suppression pattern, there is some reactivity to tactile stimulation and often preservation of sleep-state transitions in an EEG with an excessively discontinuous background. Furthermore, in the premature infant, periods of EEG activity separated by excessively long interburst intervals may contain many of the normal transients that are abundant at this age (c.g., delta brushes and temporal theta bursts), whereas the bursts within the burst-suppression pat-

95

tern of term infants are composed of abnormal EEG activity. Normal discontinuous patterns may be affected by mild encephalopathies. The trace alternant may manifest excessively long or low-amplitude interburst intervals. Excessively long interburst intervals are indicative of encephalopathy at all CAs, and generally have unfavorable prognostic implications.r' Benda and co-workers reported the prognostic significance of excessively discontinuous background activity in 44 premature infants (mean EGA, 27.1 weeks) who were followed for approximately 2 years." Using the longest inactive phase as the benchmark, three groups of interburst intervals were identified: (l) those less than 20 seconds; (2) those between 20 and 29 seconds; and (3) those of 30 seconds or more. An inactive period of30 seconds or longer in an infant of 25 to 35 weeks CA was more common in those who died. However, the EEGs of normal and disabled survivors showed no differences with respect to inactive periods lasting for 30 seconds or longer. This study supports earlier reports on the clinical significance of prolonged interburst intervals in premature infants.l-" but the findings were not striking in the survivors. There are some differences among studies in regard to the duration of normal interburst intervals, owing in part to the various criteria used to define them. Nevertheless, a conservative statement is that the maximum interburst intervals duration should not exceed 40 seconds in infants younger than 30 weeks CA, 20 seconds in infants 33 to 36 weeks CA, and 6 seconds in full-term infants. 24 However, in very premature infants (younger than 30 weeks CA), there is a remarkably wide range of interburst intervals and more variability in outcome.'?

Low-Voltage Undifferentiated Pattern A low-voltage pattern is usually defined as activity between 5 to 15 uv during all states.t" Faster frequencies tend to be depressed or obliterated. Differentiation of sleep state in low-voltage records may be difficult, although some amplitude differences may exist he tween sleep states, with amplitudes being slightly higher in quiet sleep than in active sleep. Low-voltage records are seen in a variety of severe CNS disorders, including hypoxic-ischemic encephalopathy; toxic-metabolic disturbances; congenital hydrocephalus; and severe intracranial hemorrhage, including large subdural hematomas. Tharp and colleagues found that a low-voltage record that lacks reactivity or characteristic EEG patterns for CA carries a poor prognosis (four of five infants having severe sequelae). 22 The prognostic value of the low-voltage EEGs depend strongly on the timing of recording after a presumptive brain injury. The pattern is ominous, especially when it persists beyond the first week after the insult. Therefore, EEGs obtained shortly after the event

96

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

should be interpreted with caution, and a follow-up study should be performed. Several caveats about the interpretation oflow-voltage EEGs are required. It is important that the recording is long enough to include periods of quiet sleep (which are generally higher in amplitude than active sleep). Severe scalp edema, subgaleal hematomas, subdural hematomas, and extra-axial fluid collections may also artifactually attenuate the EEG amplitude. Transient generalized attenuation of the background (usually lasting less than a few minutes) may occur before the onset of quiet sleep" and after surfactant therapy.

Fp1-C3

~~"-wf""''''''''''-'v<_~'''''''''""",_",-""¢""",,,_

C3-01 Fp2-C4 -.._.................._---..__-...'--..I-....-..."~'-"'-'"V"V C4-02

C3-Cz

_~'V

Cz-C4 C4-T4

Low-Voltage Background with Theta Rhythms The low-voltage background with theta rhythms is a variant of the low-voltage pattern. Continuous low-voltage background (5 to 25 I!V) is accompanied by lowvoltage (5 to 15 I!V) theta activity that occurs in bursts of varying lengths or sometimes continuously. The theta activity may be diffuse but is more often focal or multifocal. The EEG shows no reactivity to stimulation and, usually, no discernible sleep states. This pattern is often the precursor of an isoelectric record and is associated with highly unfavorable outcomes. 22,27 Diffuse Slow Activity The pattern of diffuse slow activity consists of widespread amorphous delta activity that persists throughout the recording and is not significantly altered by sensory stimuli (Fig. 4-7). The faster patterns in the theta and beta range, which are normally abundant in the EEGs of term infants, are absent. The background is devoid of the normal patterns seen in the premature infant, such as delta brushes, temporal theta bursts, and occipital theta bursts. This type of abnormal background must be distinguished from the high-amplitude, slow-wave pattern that occurs during a portion of the normal quiet sleep of preterm and term infants, and that gradually replaces the trace alternant pattern during the first 4 to 6 weeks after term."! This pattern should also be distinguished from frontal rhythmic delta activity (anterior slow dysrhythmia), which is a normal pattern seen in transitional sleep that appears at 37 weeks CA and persists for several weeks after term. When a diffuse slowactivity pattern persists past the second week in full-term neonates, the prognosis appears to be poor.

...-r '_"_ _"'-_"--____' '--

v-...""'--"

L

II Diffusely slow background with absence of normal patterns for the conceptional age (CA). Premature infan t (34 weeks CA) with hypoxic encephalopathy, severe respiratory distress syndrome, and bloody CSF. The infant died on the 6th day of life, and autopsy revealed subarachnoid hemorrhage, cerebral edema, and widespread anoxic encephalopathy. Calibration: 50 IlV; I second.

FIGURE 4-7

cent synchrony for infants older than 30 weeks CA) .22 The majority of records with gross interhemispheric asynchrony have paroxysmal backgrounds (Fig. 4-8). Records with markedly asynchronous burst-suppression patterns are often seen in infants with severe hypoxicischemic encephalopathy and congenital malformations such as Aicardi syndrome. Neuropathologic

Fp1-T3 T3-01

Fp2-T4 T4-02

Fp1-G3 C3-01

Fp2-C4 C4-02

Grossly Asynchronous Records The degree of interhemispheric synchrony is dependent on the CA. A record is considered to be grossly asynchronous if, during the discontinuous state, all background rhythms are persistently asynchronous between the hemispheres (estimated less than 25 per-

4-8 III Excessive interhemispheric asynchrony. Premature infant with hypoxic encephalopathy and severe bronchopulmonary dysplasia (EGA 28 weeks; legal age, 11 weeks). The child exhibited severe developmental delay and seizures by the age of 16 months. Calibration: 50 IlV; 1 second.

FIGURE

Neonatal and Pediatric Electroencephalography

correlates of this pattern include abnormalities in the corpus callosum and periventricular leukomalacia.V A grossly asynchronous pattern is usually associated with an unfavorable outcome.F

Amplitude Asymmetry Pattern A persistent amplitude asymmetry in the background activity between the hemispheres (Fig. 4-9) that exceeds 50 percent and is present in all states is thought to be significant. This pattern commonly correlates with lateralized structural pathologies (e.g., intraparenchymal hemorrhages, strokes, or congenital malformations). It is important to exclude the presence of asymmetric scalp edema or cephalhematomas and technical pitfalls (e.g., electrode "salt bridges" or asymmetric electrode placement) . Transient amplitude asymmetries have nonspecific prognostic significance. Transient unilateral attenuation of background EEG activity, usually lasting about I minute, may rarely occur during slow-wave quiet sleep in normal newborns. These episodes usually occur within minutes of the time the infant first enters slowwave quiet sleep and are accompanied by normal background activity. Transient asymmetries may also occur after subclinical electrographic seizures. Focal Abnormalities

Focal abnormalities usually consist of localized amplitude- attenuation of background activity, with or without

Fp1-T3 T3-01 Fp2-T4

T4-02~

/V" Iv., ''''I It -J\~ 11 '1 (, I. (Vi I, (1 V 'vi :1""/\,.) v vv VA~/;lIvJ \1 ("""t/'V"

Fp1-C3/ C3-o1

\rt.',~)~~.,F·'VM-,V~V~

FP2-C4~~~ L C4-o2 FIGURE 4·9 III Attenuation of the amplitude of the background activity over the right hemisphere, maximal in the right temporal region, in a full-term 3-day-old child who had occasional seizures but was alert and without abnormality on neurologic examination. Computed tomography showed localized hemorrhage in the right temporal lobe. Carotid arteriogram suggested the presence of an arteriovenous malformation. Calibration: 50 IlV; 1 second.

97

spikes (see Fig. 4-9). Again, it is important to exclude electrode placement errors, salt-bridge formation, localized scalp edema, and cephalhematoma, any of which can cause localized attenuation of amplitudes. Less commonly seen is focal high-amplitude slowing, often accompanied by spikes and sharp waves; it is seen more commonly in older premature and term infants. This pattern may correlate with focal abnormalities on the neuroimaging and neuropathologic studies (e.g., hemorrhage, cerebral infarction, and periventricular Ieukomalaciar.w" However, the sensitivity and specificity of the EEG for diagnosing focal cerebral abnormalities seem to be poor. Some infants with gross focal lesions at autopsy did not have a significant asymmetry on the EEG. In a small study of 12 infants with subdural hematomas, focal or unilateral EEG abnormalities were seen only in two, and generalized abnormalities were present in the remainder." Disturbance of Sleep States Sleep state differentiation should be readily apparent after 34 weeks CA. EEGs that lack distinct sleep states despite a long period of recording (1 hour) usually occur in infants with encephalopathies from a variety of causes. The background is usually persistently low in amplitude or is excessively discontinuous. It is important to ensure that this lack of change in sleep state is not caused by excessive environmental stimulation in the NICU, hypothermia, toxic factors, or administration of neuroactive medications. If these causes can be eliminated, an EEG that contains no recognizable states is generally associated with a poor prognosis. If the EEG is performed within the first 24 hours, it should be repeated after several days. Several types of abnormalities of sleep states can be encountered in an EEG. They include the lack of welldeveloped active and quiet sleep; poor correlation (concordance) between the behaviors characterizing a particular state and the EEG; excessive transient or "indeterminate" sleep; excessive lability of sleep states; and deviations from the normal percentages of specific behavioral states. Much variability exists in the interpretation of these patterns, and few normative data are available. These patterns occur in babies with mild hypoxic or metabolic encephalopathies or subarachnoid hemorrhage, or following complicated pregnancies or deliveries. As the abnormal clinical state resolves, the EEG rapidly returns to normal. These abnormalities are etiologically nonspecific and apparently have little predictive value. Other Nonspecific Background Disturbances Many abnormalities are of little prognostic value in isolation but often accompany the severe abnormalities just discussed. These disturbances include excessive amounts of anterior slow dysrhythmia; increased incidence of encoches Jrontales (frontal sharp transients);

98

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

excessive amounts of fast background rhythms, particularly beta activity; and a transiently dysmature background. These abnormalities are etiologically nonspecific and have little predictive value. DYSMATURE PATTERNS

EEGs are considered dysmature if they contain patterns that are at least 2 weeks immature for the CA. EEG patterns that have been used as benchmarks for determining maturity include the degree of interhemispheric synchrony, number of delta brushes during active and quiet sleep, number of temporal theta bursts, and duration and morphology of the interburst intervals during discontinuous sleep. Normative scales for these patterns have been developed by several investigators 11,15,45; however, the determination of dysmaturity is usually performed subjectively. The determination of dysmaturity is more easily made when the infant's CA is near term. Dysmature patterns that resolve on serial EEGs (transient dysmaturity) are not significant with respect to outcome; however, persistently dysmature patterns on serial EEGs may indicate a poor prognosis.1,6,46,47 A variety of exogenous and endogenous factors may disturb the maturational schedule of the EEG. Seizures may produce a temporary dysmaturity or regression. When dysmaturity results from seizures, it does not consistently correlate with a poor outcome. Dysmature EEG patterns may be caused by prolonged physiologic disturbances, such as chronic lung disease or patent ductus arteriosus, that cause an arrest or delay in brain maturation.i" Dysmature EEG patterns are often observed in infants with severe bronchopulmonary dysplasia.l" This pattern appears after several weeks of ventilatory support and correlates with an unfavorable outcome (e.g., neurologic sequelae or early death). Biagioni and co-workers obtained EEGs within the first 2 weeks in 63 preterm infants (28 to 34 weeks EGA) and scored the degree of dysmaturity using precise maturational criteria." They found that a normal EEG was associated with a favorable prognosis (in 25 of 26 infants), but a highly dysmature EEG was not necessarily associated with a poor prognosis (only 2 of 9 infants had severe neurologic sequelae). The fact that some infants had a normal evolution despite very dysmature EEGs may reflect transient effects on the EEG of metabolic or circulatory disturbance early in the course of the preterm infant. VALUE OF SERIAL

EEGs

IN NEWBORN INFANTS

Serial EEG recordings beginning shortly after birth are useful to assess the timing and mode of brain injuries and may also assist in elucidating their pathogenesis in pre term infants.v' Serial EEGs may help to distinguish between acute and chronic pathologic processes. The

former are characterized by EEG findings of acute depression (e.g., increased discontinuity, decreased faster frequencies, and low amplitudes); the latter may consist of dysmature and disorganized EEG patterns. The timing of brain insults may be assessed by considering the stage of EEG abnormalities in relation to the time of birth. For example, serial EEGs in preterm infants may be useful for determining the timing of injury. Hayakawa and colleagues performed serial EEGs and categorized background EEG abnormalities into acute- or chronic-stage abnormalities in infants with cystic periventricular Ieukomalacia/''' The timing of injury was judged to be postnatal if the EEG was normal during the early neonatal course but afterward developed acute-stage followed by chronic-stage abnormalities. Insults just before or around birth would result in acute-stage abnormalities during the early neonatal period, whereas antenatal insults would result in chronic-stage abnormalities during this period. In infants whose initial EEG displayed chronic-stage abnormalities, the mean age of cystic degeneration on ultrasonography was 4 days earlier than those with acute-stage abnormalities, presumably because the injury occurred several days before birth.P?

Serial EEGs in Preterm Infants The prognostic value of serial EEGs in premature infants has been shown in several studies. 1,22,3Y In a study of 81 infants with an EGA of less than 36 weeks, Tharp and colleagues found a number of patterns that occurred primarily in infants who died in the neonatal period or who developed long-term sequelae.F The patterns included an isoelectric EEG; positive rolandic sharp waves; a burst-suppression (paroxysmal) disturbance distinct from the discontinuous pattern seen in normal, small premature infants (see Fig. 4-1); excessive interhemispheric asynchrony (see Fig. 4-8); major (more than 50 percent) and persistent hemispheric amplitude asymmetries (but excluding records with localized areas of amplitude attenuation) ; and an excessively slow background of variable amplitude (10 to 100 J..lV) that was unresponsive to stimulation and was characterized by a paucity or absence of physiologic rhythms such as delta brushes (see Fig. 4-7). Other less severe EEG abnormalities were also encountered in this age group but were of no prognostic value. These included mild asymmetries and asynchrony for the conceptional age, excessively discontinuous backgrounds, and alterations in the amount of background faster rhythms. In this series of 81 premature infants, a worsening of the EEG (e.g., a normal EEG progressing to an abnormal record or a moderately abnormal EEG evolving to a markedly abnormal EEG) was seen in 41 percent of the children with major neurologic sequelae, in contrast to

Neonatal and Pediatric Electroencephalography

3 percent of children with normal development and 15 percent of those with minor sequelae. ~~ This study reinforces the importance of serial EEGs in premature infants. Tharp and colleagues prospectively studied all premature infants (birth weight less than 1,200 g) admitted to an NICU and reported that neurologic sequelae occurred in all infants whose neonatal EEGs were markedly abnormal and in the majority of those with two or more moderately abnormal tracings (recorded at weekly intervals).' The EEG was more sensitive than the neurologic examination in predicting poor outcome (72 percent vs. 39 percent), whereas both were equally effective in predicting normal outcome. The EEG also proved more sensitive than the cranial CT and ultrasonography in establishing the severity of the encephalopathy. The combination of EEG and ultrasonography may be particularly useful in detecting brain injury in preterm infants. In a recent study, EEG abnormalities (within 72 hours) in conjunction with abnormal ultrasonography detected periventricular leukomalacia with a sensitivity of 94 percent and a specificity of 64 percent." Maruyama and colleagues evaluated acute EEG background abnormalities in 295 preterrn infants (EGA 27 to 32 weeks) on the basis of continuity, frequ<:ncy spectrum, and amplitude, and graded them on a 5-point scale. 52 The EEGs were performed within the first week of life. The maximal grade of the acute background abnormalities correlated with the subsequent development of cerebral palsy (mostly because of periventricular leukomalacia) and it" severity, but the presence of significant acute background abnormalities also had a high false-positive rate. The recording of serial EEGs is a very important aspect of the evaluation of premature infants. The EEG should be obtained at the time of acute neurologic insult and repeated 1 to 2 weeks later, particularly if the initial EEG is normal or only moderately abnormal. The authors have often been impressed by the development of a grossly abnormal EEG at a time when the child was stable or even improving clinically.

99

infants with a major depression of the background at any time (burst-suppression, or nearly isoelectric background); with moderate depression (abnormal trace alternant, a discontinuous background, or a very-lowvoltage irregular pattern) lasting longer than 4 days; or with mild depression present after 9 days. Similarly, in a study of nine term infants with hypoxic-ischemic encephalopathy, Pressler and colleagues found that an early EEG was an excellent prognostic indicat.or for a favorable out.come if normal within the first 8 hours after birth.F' The outcome was unfavorable if major background abnormalities persisted beyond 8 to 12 hours. However, an inactive or very depressed EEG within the first 8 hours correlated with a good outcome if the EEG activity recovered within 12 hours. Zeinstra and colleagues confirmed that serial EEGs are better than a single study performed early. They performed two EEGs, the first 12 to 36 hours after birth and the second in 7 to 9 days in 36 term infants with acute neonatal asphyxia. 54 Several infants with a burstsuppression pattern on the initial EEG showed a significant improvement on the second EEG and had a favorable outcome. If the first EEG was normal or mildly abnormal, the second EEG did not add substantially to the prognostic value. ABNORMAL

EEG

TRANSIENTS

Spikes and sharp waves are commonly seen in neonatal EEGs and must be interpreted conservatively. In older infants and children, interictal spikes and sharp waves are signatures of an epileptogenic disturbance. In term and premature newborns, however, such sharp or fast transients may be seen in asymptomatic newborns with a normal outcome.P'' Even when seen in "excessive" amounts, they tend to be relatively nonspecific and do not necessarily imply an epileptogenic abnormality.'" Unless they are repetitive, periodic, or confined to specific head regions, pathologic significance should be assigned with caution. This section focuses on some EEG transients that may have pathologic si~nificance. The relationship between sharp transients and neonatal seizures is discussed later.

Serial EEGs in Term Infants Serial EEGs have also been utilized in the assessment of full-term newborns. Takeuchi and Watanabe assessed I n high-risk, full-term infants with hypoxic-ischemic encephalopathy (defined as an episode of fetal distress or an Apgar score of 5 or less at I or 5 minutes after del;very).26 The severity of the depression of EEG background activity and its persistence correlated with the neurologic outcome. Infants with normal EEGs and with only minimal or mild background depression that disappeared during the first few days of life had good outcomes, By contrast, neurologic sequelae occurred in

Positive Rolandic Sharp Waves Positive rolandic sharp waves are moderate- to highamplitude (50 to 200 IlV), surface-positive transients lasting 100 to 250 msec (Fig. 4-10). They may occur in the central regions (C3, C4), either unilaterally or bilaterally, or in the central vertex (Cz) region. Positive rolandic sharp waves correlate with deep white matter lesions, particularly periventricular leukomalacia. 23.57-1iU They may be an early marker of white matter injury, preceding the uItrasonographic detection of cystic changes.f" They may be associated with intraventricular

100

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

FIGURE 4-10 • Positive rolandic sharp waves in premature

infant with intraventricular hemorrhage on CT scan. +: posilive sharp waves at Cz. Calibration: 50 )J.V; 1 second.

decrease rapidly on subsequent EEGs, whereas they persisted in the pathologic groups. Scher and co-workers found that positive temporal sharp waves were present in a population of healthy, asymptomatic infants, with the peak at 33 to 36 weeks CA.62 They were uncommon in term infants but persisted in premature infants who had matured to term. These authors postulated that the persistence of these sharp waves in the latter group may represent an electrographic pattern of dysmaturity.62 Others have found a correlation with hypoxicischemic insult in the newborn period. Full-term infants who have positive temporal sharp waves appear to have a high incidence of focal or diffuse structural lesions on neuroimaging studies and a high incidence (80 percent) of other EEG background abnorrnalities." From this study, the prognostic implication of positive temporal sharp waves is unclear. There is no consistent association with neonatal seizures.

Excessive Frontal Sharp Transients hemorrhage if there is a component of white matter injury. Positive rolandic sharp waves are usually associated with an unfavorable outcome. However, infants with these sharp waves often have other EEG background abnormalities that may confound determination of their prognostic significance. In one study of premature infants, the occurrence of positive rolandic sharp waves with a frequency exceeding 2 per minute was highly associated with motor disabilities or early mortality.'?

Temporal Sharp Transients Abnormal temporal sharp transients must be distinguished from normal, usually sharply contoured theta bursts of activity that are seen in the temporal areas during the discontinuous (trace discontinu) background in normal premature infants (26 to 32 weeks CA), with highest incidence between 29 and 31 weeks CA. These temporal theta bursts or "sawtooth" waves are normal patterns and are not associated with seizures.

Positive Temporal Sharp Waves Positive temporal sharp waves have a morphology and polarity similar to those of positive rolandic sharp waves hut occur over the midtemporal regions (T3 and T4). In a study of premature infants (31 to 32 weeks CA) they were seen in approximately half of asymptomatic infants and in three-quarters of children with various disorders (asphyxia, metabolic disorders, and cystic pcriventricular leukomalaciaj.v' Their incidence was higher in the asphyxia group when compared with the asymptomatic group but not in the other disorders. In the asymptomatic group, their frequency tended to

Frontal sharp transients are physiologic patterns seen in infants between 35 and 45 weeks CA. These sharp waves are seen more abundantly in mild encephalopathies and are absent in severe encephalopathies. Nunes and colleagues found increased density and asynchrony of frontal sharp transien ts in symptomatic hypoglycemic neonates compared with normal controls.v" However, it is not clear whether hypoglycemia was responsible because serum glucose concentrations were not recorded around the time of the EEG and the infants had encephalopathies of various etiologies including asphyxia, hydrocephalus, and sepsis.

Periodic Discharges Focal periodic discharges and periodic lateralized epileptiform discharges (PLEDs) are pathologic patterns that occur in various encephalopathic conditions but do not necessarily imply ictal events. The distinction between these two patterns is sometimes difficult to define but may relate to their duration and persistence. Focal periodic discharges consist of broad-based, often biphasic discharges that may occur focally or independently at various locations (Fig. 4-11). They may last from a few seconds to several minutes and sometimes become faster in frequency. The usual lack of evolution in the morphology, frequency, or field of the discharge differentiates them from electrographic seizures. However, focal periodic discharges may sometimes represent a "slow" ictal pattern, often associated with focal clonic seizures. Focal periodic discharges may occur in a variety of CNS disorders, such as cerebrovascular insults (strokes), hypotension, bacterial meningitis, viral encephalitis, and

Neonatal and Pediatric Electroencephalography

10 1

Fp1-T3 T3-01 ----..r------"--'\i'-J'-.'-'\r"----.-.A.J--..I---..J\-, I ~""'""\r'--....,..,if' ....-A....... r-""""-'''-'\ r-J'v........ 'AJo'"""-"lf"-J Fp2-T4 . - - - - - - - - - . - - - - - - - - - - ~_T4-02

~

Fp1-e3

...

~

---~

_~---.r­

---r,rv-------..f'v.---'''---...-J'I.. _ , ........~.~

C3-01

Fp2-C4 . -----------------------------~----­ C4-02-~--------------------~~-----

T3-C3 C3-ez Cz-C4

--~-----------------------------------

C4-T4 --~-FpZ-Fz _ - - - -

-------------------------~-...........---__ ~

-Fz-Cz - - - - - - - - - - - - - ----------------------_ Cz-Pz ---"'-------~--"____..__--~---------_ LOC-A1

----~----------------~----~-----------------

ROC-A1 ~:r-----..."---""---.,-"---.,,.,.....--.....,....------~--..,.....-

....

"""'"-_/"~~.~

FIGUIE 4·11 • Periodic sharp discharges located in the left hemisphere (maximal in the left temporal region) in a 33-week EGA infant with bacterial meningitis and multicystic abscesses, resulting in seizures. Calibration: 50 IlV; 1 second

brain malformation, and in the course of metabolic disorders. Scher and Beggarly studied 34 infants with focal periodic discharges in EEGs recorded in the neonatal period. 6!' Cerebral infarction was the most common underlying pathology. Most (75 percent) of these infants had an unfavorable prognosis (death or neurodevelopmental sequelae). Focal periodic discharges may occur in neonates without clinical seizures. However, they sometimes follow a high-frequency EEG discharge (with focal clinical seizures occurring during the high-frequency discharge). In preterm infants, focal periodic discharges are less commonly associated with electrographic seizures and a demonstrable cerebral lesion. Mizrahi and Tharp described a unique multifocal periodic or pseudoperiodic pattern in six infants with severe encephalitis/" The periodic activity, which appeared in the temporal, frontal, and central regions, consisted of slow waves or sharp slow waves that recurred at 1- to +second intervals and were usually invariant throughout the tracing, interrupted only by focal seizures. PLEDs are defined as stereotyped, repetitive, paroxysmal complexes occurring with a regular periodicity (around 1 second) and lasting at least 10 minutes. They

often last much longer. PLEDs exhibit no evolution in morphology, frequency, or field, and they are not associated with ictal manifestations. PLEDs have been seen in patients with focal pathology (e.g., cerebral infarcts) or more diffuse encephalopathies (e.g., perinatal asphyxia) . Focal periodic discharges and PLEDs may represent a continuum of a similar pathophysiologic process, appearing in brain-injured infants who sometimes exhibit clinically detectable seizures and who usually display other EEG background abnormalities and disorganized states. The prognosis appears poor for either pattern, although it may depend more on the underlying etiology.

Rhythmic Theta-Alpha Activity The EEG of preterm and term infants sometimes shows excessive amounts of rhythmic theta- or alpha-frequency activity. In a study of term infants (37 weeks CA or older) by Hrachovy and O'Donnell, such patterns were found in a variety of conditions including congenital heart diseases, congenital brain abnormalities, and hypoxemia, as well as in infants receiving neuroac-

102

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

tive medications.f? They concluded that this pattern was diagnostically nonspecific and may occasionally be seen in infants without overt CNS disease. Rhythmic theta-alpha bursts, which consist of bursts or short runs (1 to 3 seconds) of sharply contoured, primarily surface-negative, rhythmic theta- and alpharange activity, are common in the central and frontocentral midline areas, particularly during quiet sleep (Fig. 4-12).19,68 These are called wickets or rhythmic sharp theta bursts by some authors. Such bursts occur in healthy infants, although some authors believe that they are often seen in infants who have suffered various eNS insults, particularly when clearly defined sharp waves or spikes are intermixed with these bursts. However, in a study of newborns with electrographic seizures, the presence of this activity seemed to have a favorable prognostic value. 69 Biagioni and co-workers investigated the significance of rhythmic theta-alpha bursts and other "abnormal" EEG transients in preterm and full-term neonates.?? They scored the EEGs based on the abundance of isolated sharp transients (surface positive or negative), rhythmic sharp theta bursts, rhythmic delta activities, and rhythmic low-amplitude alpha discharges. The overall incidence of abnormal transients was low, with 21 percent of the EEGs having frequent abnormal transients. In the EEGs of full-term newborns or preterm infants reaching term, the score correlated with the outcome. During the preterm age, the EEG score did not correlate significantly with the outcome. Surprisingly, the authors considered these transients to be abnormal at any CA and attributed the lack of correlation with outcome in premature infants to the fact that the immature brains have a limited ability to generate them. However, based on their study, it is not clear whether patterns such as rhythmic theta-alpha

T3-C3 C3-Cz Cz-C4 C4-T4

FIGURE 4·12 • Rhythmic theta-alpha burst at the vertex (Cz)

region. Calibration: 50 IlV; 1 second. (Modified from Ortibus EL, Sum JM, Hahn JS: Predictive value of EEG for outcome and epilepsyfollowing neonatal seizures. Electroencephalogr Clin Neurophysiol, 98:175,1996, with perrnission.)

bursts represent clearly abnormal activity, because such transients do not consistently correlate with outcome. NEONATAL SEIZURES

Neonatal seizures are a common problem in the newborn nursery; they can occur in infants at any CA and have diverse etiologies." The clinical spectrum of neonatal seizures is quite different from that of older children. Generalized tonic-clonic seizures are exceedingly rare in the newborn, especially in the premature infant. The most common seizure types are focal or multifocal clonic, tonic, and myoclonic phenomena; and motor automatisms, such as oral-buccal-lingual movements and progression movements (e.g., pedaling, stepping, and rotatory arm movementsj.F The EEG is of particular diagnostic value in those infants whose seizures are subclinical, subtle, or easily confused with nonepileptic motor behavior. Studies using video-EEG monitoring of infants with eNS dysfunction have shown that several types of abnormal paroxysmal motor activity previously considered to be seizures are not associated with ictal EEG pattems.I''?" Included are generalized tonic posturing, motor automatisms, and some myoclonic jerks. The investigators contend that these behaviors may be elaborated by "brainstern release" or other nonepileptic mechanisms. Connell and colleagues recorded continuous (24 hours or more), two-channel EEGs from 275 fullterm and preterm infants, 55 of whom were found to have seizures." Clinical signs occurred simultaneously with EEG ictal patterns in 12, were minimal in 20, and were absent in 23. However, because only two channels ofEEG (F4-P4 and F3-P3) were recorded, it is possible that seizures in other regions were missed. Therefore, EEG monitoring is crucial in determining the epileptic nature of all but the most obvious "seizure-like" behavior (e.g., clonic and migrating clonic activity and tonic eye deviation). The EEG can also be useful in the evaluation ofjittery babies and or infants who have abnormal motor activity that is easily confused with seizures. Electrographic seizures and status epilepticus often occur without recognizable clinical accompaniments, particularly in depressed newborns. Possible reasons include the use of pharmacologic paralyzing agents (for management of neonates on mechanical ventilators);" the administration of anticonvulsants that decouple the clinical behavior and the EEG, and the inability of some immature neonatal cortex to generate the clinical expression of epileptic activity. In other cases, subclinical electrographic seizures occur in infants who are obtunded or comatose and who have suffered a significant hypoxic-ischemic insult; the seizures are usually associated with long-term neurologic sequelae. EEGs are essential in monitoring CNS function in such babies

Neonatal and Pediatric Electroencephalography

and in determining the effectiveness of anticonvulsant therapy.

Ictal EEG The- criteria of an electrographic seizure in neonates are the subject of debate." The electrographic patterns of neonatal seizures are highly variable, with complex and varied morphology and frequencies (Fig. 4-13). Most groups operationally define electrographic seizures as clear ictal events characterized by the appearance of sudden, repetitive, evolving stereotyped waveforms that have a definite beginning, middle, and end, and last a minimum of 10 seconds. 69 ,76 During a seizure, there is a progressive buildup of rhythmic activity at almost any frequency or repetitive sharp waves and spikes. Neonatal seizures are usually electrographically focal and have variable spread over the ipsilateral and often the contralateral hemispheres. Two or more focal seizures may appear concomitantly in the same hemisphere (Fig. 4-14); or, more commonly, appear in both hemispheres and progress independently at different frequencies. Multifocal seizures are common in severe encephalopathies. Multifocal ictal activity is more often associated with neurologic sequelae than unifocal seizure discharges, particularly if there are two or more independent foci."? Status epilepticus, defined as total seizure duration lasting more than 30 minutes or seizures occupying more than 50 percent of the

103

EEG, is also highly associated with severe neurologic sequelae or death. 69 The electroclinical correlation is rather poor, with a limited number of ictal EEG patterns accompanying a variety of clinical manifestations. Certain types of EEG patterns are reported to be commonly associated with certain clinical seizures (e.g., rhythmic alpha-frequency activity with apneic seizures and rhythmic delta waves with tonic seizures?"), but the relationships are rather tenuous. In benign familial neonatal convulsions there appears to be a characteristic pattern of electroclinical seizures: the seizures begin with a diffuse flattening of the background, accompanied by apnea and tonic motor activity, and are followed by rhythmic theta or delta activity that evolves into bilateral spike and sharpwave discharges during bilateral clonic activity. 79,80 Others have found that the seizures in benign familial neonatal convulsions are focal in onset with secondary generalization. 81,82 Figure 4-15 illustrates some of the artifacts that may be confused with seizure discharges.

Interictal EEG Interictal background patterns are extremely variable and can range from normal background activity to isoelectric records. The interictal background reflects the severity of the underlying encephalopathy responsible for the seizures. The background activity may be tran-

C3-Cz T4-o2

Cz-C4 251-lV~

1 sec Fp2-C4 C3-Gz C4-02

Cz-C

Fp2-T4

~"""""'''V'V'V''''''''''''''''''''',"""",'VVVv'..rv.rv..."."..........,.--.......,.....

T4-02 .,..".""""'"".....,..,...,.........,..,.---...,......-....."""''''''''----...__VV''

T4-o2

Fp1-T3

Fp2-T4 ('1/1 T4-o2.-n......

ftoJ, • . . . .

FIGURE 4-13

~

Fp2-T4

II

calibration: 50

T3-01

Typical electrographic seizure patterns in premature and full-term infants. Except where otherwise indicated, uv. 1 second,

104

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

T3-01 Fp4-T4 T4-02 Fp3-C3 C3-Q1 Fp4-C4 C4-02 TS-GS C3 C3-Cz Cz-C4

Cz-Pz .-..~-~---~..

L...

FIGURE 4-14 .. Multifocal electrographic seizures in a I-day-old infant (EGA 38 weeks; birth weight, 3100 g) with severe respiratory distress syndrome and hypoxemia. Independent ictal discharges are seen in the left frontal (Fp3) and central (C3) regions. In the midline frontopolar region, there are focal periodic discharges (arrowheads), followed by a brief seizure discharge. Calibration: 25 uv, 1 second.

siently worsened by altered cerebral metabolism and perfusion associated with a seizure, by intravenously administered anticonvulsant drugs, or by systemic postictal changes. Therefore the final interpretation of the EEG should be made on the basis of the entire record, particularly the pre-ictal background. If a seizure occurs in the early portions of the EEG, a prolonged recording session may be necessary to allow sufficient recovery from the postictal state before judgment of the interictal background is made. Regardless of the presence of electrographic seizures, the interictal EEG patterns (which reflect the severity of the underlying encephalopathy) can be most helpful in the prediction of neurologic sequelae. 69 ,83 Severely abnormal background patterns are highly associated with early neonatal mortality or severe neurologic outcome.P''?? In one study the presence of an isoelectric pattern was always associated with unfavorable outcome, whereas a burst-suppression or a low-voltage undifferentiated pattern was associated with 70 to 80 percent unfavorable outcome.P? The outcome in neonatal seizures appears to be particularly unfavorable if the patient population is

limited to those neonates with EEG-documented seizures. 71 •83

Spikes, Sharp Waves, and Seizures In older infants and children, interictal spikes and sharp waves are signatures of an epileptogenic disturbance. In the neonate, however, such sharp transients are relatively nonspecific and do not necessarily imply an epileptogenic abnormality. Most consider sharp transients occurring over the frontal, rolandic, and temporal areas to be abnormal if they are excessively frequent for the CA, appear in short runs, or are consistently unilateral or polyphasic. They are particularly abnormal if they occur during the attenuated phase of trace alternant or persist during the more continuous pattern of REM sleep or wakefulness. Multifocal sharp waves are also considered to be abnormal, although they do not show a statistically significant association with neonates having seizures. In newborns, sharp waves occur in a variety of pathologic situations (e.g., infarcts, hemorrhages, and leukomalacia). In the authors' experience, sporadic

Neonatal and Pediatric Electroencephalography

105

ECG

Fz-Cz Hiccoughs

C3-Q1

~/0NW~NVVVwv...jvWvl'"

Electrode artifact

Fp2-C4

c4;;;f

C3-Cz

~~I!Wf'w~

Cz-C4

Patting baby Pulse FIGURE 4-15

III

Rhythmic artifacts, which can be confused with ictal patterns. Calibration: 50

isolated negative or biphasic spikes and sharp waves occurring in an otherwise normal record should be interpreted conservatively. A record may be interpreted as abnormal solely on the basis of an excessive numbers of spikes or repetitive focal spikes or sharp waves, but its value as a predictor of long-term neurologic outcome remains uncertain unless other abnormali ties exist. Whether spikes and sharp waves represent interictal epileptiform discharges in a neonatal EEG is still debated. Clancy quantified the number and location of sharp transients (central and temporal only) in infants with electrographically proven seizures and found more focal and multifocal sharp transients than in a control population." The sharp transients also tended to occur in bursts (runs) in the group with seizures. However, there was a significant overlap in the frequency of such discharges between the two populations, and there was no comparison group of encephalopathic infants without seizures. In a study of 81 neonates with EEG-proven seizures, the abundance of negative temporal sharp waves in the interictal EEG was significantly correlated with the development of postneonatal seizures." Furthermore, a direct relation existed between the severity of the postneonatal seizure disorder and the number of temporal sharp waves. However, until further studies are completed in neonatal seizures with a control group,

~V;

1 second.

spikes and sharp waves in a neonatal EEG cannot be construed to indicate an "epileptogenic" disturbance. The diagnosis of neonatal seizures must be made in the clinical context, ideally with the recording of ictal discharges. Nevertheless, the authors believe that the presence of frequent (three or more per minute) negative temporal sharp waves is abnormal, and in their population was highly correlated with an unfavorable outcome and frequent postneonatal seizures." In the same study, the presence of rhythmic thetaalpha bursts in both the ictal and the interictal EEGs strongly correlated with a favorable outcome and a decreased likelihood of developing postneonatal seizures.P'' The authors believe that these patterns are a normal phenomenon of healthy newborns and may be a predictor of favorable outcome. They are depressed in neonates with severe encephalopathies who are destined to develop postneonatal seizures and neurologic sequelae.

Rhythmic Discharges The EEG ictal pattern of neonatal seizures has been arbitrarily defined as one that lasts at least 10 seconds. The significance of briefer rhythmic discharges was investigated in a study of 340 neonates (30 to 40 weeks CA) .85 Rhythmic discharges were defined as runs of repetitive, rhythmic, monomorphic, stereotyped, sinu-

106

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

soidal or sharply contoured waveforms occurring within a frequency of 0.5 to 20 Hz. The cohort was subdivided into three groups: (l) 67 with rhythmic discharges with maximal duration less than 10 seconds (brief rhythmic discharges); (2) 63 with long rhythmic discharges (exceeding 10 seconds); and (3) 210 without rhythmic discharges. In 40 percent of the subjects with rhythmic discharges accompanied by a clinical seizure, the electrographic discharge lasted less than 10 seconds. The authors hypothesized that brief rhythmic discharges occurring without clinical accompaniments may represent brief epileptic seizures or an interictal pattern. The incidence of postneonatal epilepsy was similar in all three groups. The presence of brief rhythmic discharges was associated with periventricular leukomalacia, and long rhythmic discharges with hypoxic-ischemic encephalopathy. The presence of any rhythmic discharges (brief or long) correlated with an abnormal neurodevelopmental outcome. NEONATAL NEUROLOGIC DISORDERS WITH CHARACTERISTIC fEG PATTERNS

Most of the neurologic syndromes encountered in neonates are associated with nonspecific EEG abnormalities. The EEG changes are usually generalized and reflect the severity of the cerebral insult. A burstsuppression pattern, for example, can be caused by severe meningitis, hypoxic-ischemic encephalopathy, or subarachnoid hemorrhage. Usually the EEG pattern, per se, does not suggest the etiology of the neurologic problem. A few specific neurologic disorders, however, may be associated with characteristic EEG abnormalities that suggest certain etiologic considerations to the clinician. In holoprosencephaly there are characteristic EEG abnormalities attributable to the distinct neuroanatomic features. When a large fluid-filled dorsal cyst is present, the EEG is low in amplitude or isoelectric over the affected region. The EEG recorded from scalp regions overlying the abnormal telencephalon is often grossly abnormal and manifests a variety of bizarre patterns. These include multifocal spikes and polyspikes (Fig. 4-16); prolonged runs of rhythmic alpha, theta, or delta activity; asynchrony; and a fast beta activity that probably represents subclinical seizures. In patients without seizures, the background often shows excessive hypersynchronous theta activity in all states and hypersynchronous beta activity during sleep." This hypersynchronous activity correlates with the severity of the thalamic and hemispheric abnormalities in the various types of holoprosencephaly. The EEG in lissencephaly (agyria-pachygyria) is said to show a characteristic EEG pattern'" of "major fast dysrhythmia" characterized by rapid, very-high-amplitude, alpha-beta activity in infants and children; however,

Fp1-T3 T3-01 Fp2-T4 T4-02

C3-Cz

~( ~~

~\ .1"oJI.--'i/--'''\II.J-''\,...~V III Multifocal sharp waves, absence of normal background rhythms, and interhemispheric asynchrony in a full-term infant with seizures, a large head, and ambiguous genitalia. At time ofEEG, baby was alert and vigorous and had a normal neurologic examination. Autopsy revealed absence of corpus callosum and septum pellucidum, and agenesis of the olfactory bulbs and olfactory tracts with a V-shaped bilobar cerebrum caused by agenesis of posterior temporal and occipital lobes. Calibration: 50 JlV; 1 second.

FIGURE 4-16

there is little information on the EEG findings in neonates. High-amplitude beta activity predominates as the child grows." Another rare anomaly, Aicardi syndrome, is associated with rather typical EEG abnormalities. This syndrome occurs in girls and is characterized by the association of infantile spasms appearing during the first few months of life and multiple congenital anomalies. The major eNS anomalies consist of agenesis of the corpus callosum, cortical heterotopias, porencephalic cysts, and ocular lesions. The characteristic EEG feature is a burstsuppression pattern that is completely asynchronous between the two hemispheres. Hypsarrhythmia may also develop in patients with Aicardi syndrome, and is quite asynchronous and discontinuous. A patient may have hypsarrhythmia in one hemisphere and a completely independent burst-suppression pattern in the other. Inborn errors of metabolism of the newborn are associated with a variety of nonspecific EEG abnormalities. One of these disorders may be suspected when a grossly abnormal EEG background develops in a normal infant who deteriorates after the initiation of feedings. The

Neonatal and Pediatric Electroencephalography

absence of a history of intrauterine distress, neonatal hypoxia, signs of infection, or extracerebral anomalies would suggest that another etiology is responsible for the marked disturbance of cerebral electrical activity. The authors have studied a full-term infant whose EEGs con tained, among other abnormalities, positive rolandic sharp waves. A CT scan documented abnormalities in the white matter, and subsequent metabolic studies were consistent with propionic acidemia. In the neonatal form of maple syrup urine disease, a characteristic pattern is seen in the first few weeks of life. sR Runs and bursts of 5- to 7-Hz, primarily monophasic negative (rnu-Iike or comb-like) activity in the central and midline central regions are present during all states, with the most abundant bursts occurring in non-REM sleep (Fig. 4-17). This pattern gradually disappears after the institution of dietary therapy. Similar patterns have been noted in propionic acidemia.

Herpes Simplex Encephalitis Herpes simplex encephalitis is one of the common fatal viral encephalitides in the neonatal period. The neonatal form is most often caused by the type 2 herpes simplex virus, which the newborn acquires during delivery from maternal genital lesions. Herpes simplex encephalitis in older children and adults is most often caused by type I herpes simplex virus and is often associated with characteristic periodic EEG patterns that typically appear in the temporal regions. In the newborn the encephalitis involves the brain more diffusely and is often preceded by characteristic skin lesions.

107

Mizrahi and Tharp described a characteristic multifocal periodic or pseudoperiodic pattern in six newborns with severe, isolated encephalitis/" The periodic activity, which appeared in the temporal regions as well as the frontal and central regions (Fig. 4-18), consisted of slow waves or sharp slow waves that recurred at 1- to 4second intervals and were usually invariant throughout the tracing, being interrupted only by focal seizures. This activity disappeared after a few weeks and was followed by a low-voltage or isoelectric EEG. All infants either had significant neurologic sequelae or died. Since the introduction of acyclovir therapy, this severe encephalopathy has virtually disappeared, and more nonspecific EEG abnormalities are usually recorded.r"

Other EEG Analysis Techniques In recent years, automated techniques for analyzing neonatal EEGs have been utilized. These include 2-channel EEGs, spectral edge frequency, and amplitude-integrated EEG. Recently amplitude-integrated EEG has been used to monitor neonates with hypoxicischemic encephalopathy. This method uses signal processing for rectifying and integrating the EEG amplitudes with a compressed time scale. It has been used as a bedside measure of the EEG background and found to be useful for identifying infants with more severe encephalopathy and worse outcomes.?" particularly when combined with the neurologic examination."! The technique relies on a limited number of channels, usually single bipolar derivation from the parietal regions. In a comparative study of amplitude-integrated

Fp1-A1 Fp3-C3 C3-01 Fp4-C4 C4-02 Jlrl"C··W\.N,';"'I1.) T3-C3 C3-Cz "WU'" W·", JII'

F3-A1 C3-A1 P3-A1 01-A1 Fp2-A2

Cz-C4

F4-A2

C4-T4

C4-A2

Fpz-Fz Fz-Cz Cz-Pz FIGURE 4·17 .. Rhythmic runs and bursts of 5- to 8-Hz, primarily monophasic, negative (rnu-like) activity in the central and central parasagittal regions during sleep in a 3-week-old, full-term infant with the neonatal form of maple syrup urine disease. Calibration: 50 ~V; 1 second.

P4-A2 02-A2 FIGURE 4·18 • Herpes simplex encephalitis in full-term infant. Moderate-amplitude periodic sharp waves in right central region and independent delta activity in left frontal region, which persisted throughout entire recording. Child died at 3 months of age with severe cortical atrophy. Calibration: 25 ~V; 1 second.

108

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

EEG and standard neonatal EEG, the former was very good at detecting severely abnormal patterns (e.g., a burst-suppression pattern, low-voltage undifferentiated pattern, and inactivityj.i" The ability to detect seizures was fairly good (sensitivity 80%, specificity 100%), although brief seizure discharges could be missed. Spectral edge frequency has been used in studying preterm infants who are at risk for cerebral white matter injury. Those with white matter injury had a marked decrease in the spectral edge frequency, representing loss of faster background frequencies.P" This method, which tracks frequencies without regard to amplitude, may be more suitable for premature infants. These methods have several limitations. Because they utilize a limited array of electrodes, focal transients and ictal discharges distant from the electrodes may be missed. The interpretation of various patterns (e.g., seizures and sharp waves) requires experience, as well as frequent correlation with routine EEGs. Management decisions about hypoxic-ischemic encephalopathy or treatment of neonatal seizures should not be made solely on the basis of these methods.

OLDER INFANTS AND CHILDREN Normal Patterns During the First 2 Years

attenuate with eye-opening and may represent the precursor of the mu rhythm. In addition, rhythmic theta activity may appear over the parieto-occipital and temporal regions during periods of crying. Runs of 4-Hz activity (50 to 200 IlV) are also present in the temporal regions of many normal infants. Lambda activity may occur in healthy babies as early as the first few months oflife (Fig. 4-19). These infants usually have a prominent driving response at the slower rates of photic stimulation. In some healthy children, high-amplitude (often reaching 100 to 200 IlV), biphasic slow transients that are sharply contoured may be seen in the posterior head regions following eye-blinks. These transients are present between 6 months and 10 years of age, with the peak incidence between 2 and 3 years. Both lambda activity and blink-related transients may be high in amplitude and occur asymmetrically; they should not be confused with epileptiform discharges. THE

EEG

DURING DROWSINESS

Drowsiness produces many interpretative problems for most electroencephalographers. The EEG patterns of drowsiness are often seen paradoxically when the child appears awake with eyes open. These patterns are noted particularly after arousal from a long period of sleep, when sedating drugs are administered, or following

The brain grows rapidly during the first 2 years of life, and significant maturational changes also occur in the cerebral electrical patterns, although they are less dramatic than those encountered during the neonatal period. A brief summary of the more important maturational features is given here, and several normal patterns are discussed. More comprehensive reviews are provided elsewhere.lt/" THE

EEG

DURING WAKEFULNESS

The predominant occipital rhythm, destined to become the alpha rhythm at 24 to 36 months of age, appears in normal infants at 3 to 6 months of age. This activity has a frequency of 3.5 to 4.5 Hz with maximal amplitude in the midline occipital region (Oz), particularly in younger infants, and appears when the child's eyes are gently and passively closed. The frequency of this activity remains in the 4- to 7-Hz range until 24 to 30 months of age when it reaches 8 to 9 Hz in most normal children." Compared with adults, children show higher amplitude of the occipital rhythm. A central theta rhythm (4 to 6 Hz), which is usually 1 to 2 Hz faster than the occipital rhythm, is present at 3 months of age; it gradually increases in frequency and reaches 8 to 9 Hz at 2 to 3 years of age. This activity does not

[ FIGURE 4·111 • Lambda activity in healthy full-term infant, 2 months old. There was respiratory distress during the neonatal period. The child's eyes were open and scanning the room during the EEG examination. Lambda waves are indicated by arrows. High-amplitude evoked responses to low-frequency photic stimulation were present later in record. Calibration: 50 IlV; 1 second.

Neonatal and Pediatric Electroencephalography

a prolonged period of vigorous crying. Most healthy full-term infants have definite EEG changes during drowsiness; these are usually characterized by an increase in the amplitude and a slowing of the frequenc'}' of the background rhythms. The EEG patterns of drowsiness are sometimes difficult to identify during the first few months of life, however, perhaps because they are admixed with sleep activity. At 4 to 6 months of age a distinctive pattern of hypnagogic hypersynchrony emerges during drowsiness. This pattern consists of runs of high-amplitude (lOa to 200 JlV), rhythmic 3- to 5-Hz activity lasting seconds to many minutes that appear in the parieto-occipital and temporal regions during the first year of life. During the second year, the drowsy state is often ushered in by a gradual decrease in the amplitude of the background activity and a slowing of the predominant waking occipital rhythm. In older infants and children, monorhythmic theta activity appears over the anterior scalp derivations, particularly in the frontocentral or central regions, and is most commonly encountered between the ages of 3 months and 2 years." Hypnagogic hypersynchrony often appears as bursts of high-amplitude, 4- to 5-Hz activity in the central and fron tal regions. Spikes and sharp waves are often admixed with the slow waves, resulting in an "epileptiform" appearance (Fig. 4-20). This pseudoepileptiform pattern has been described in approximately 8 percent of normal children aged 1 to 16 years." and was reported in a high percentage of children with a history of febrile seizures." The authors are conservative in the interpretation of all paroxysmal activity that is confined to the transition between wakefulness and sleep. Such activity is considered normal if it occurs only in drowsiness; if it disappears when typical sleep complexes such as vertex waves and spindles appear; and when the spikes or sharp waves have a random temporal relationship to the theta waves (compared with epileptogenic spike-wave activity, in which there is a more fixed temporal relationship between the spikes and the waves). THE EEG DURING SLEEP

Active (REM) sleep characterizes sleep onset in newborns. Between 2 and 3 months of age, quiet (nonREM) sleep replaces active sleep as the initial stage of sleep in the EEG laboratory setting.'! REM sleep is often recorded during the first year of life, but it is rarely seen thereafter in routine clinical EEGs. The lowto moderate-amplitude, mixed-frequency, slow background is usually easily distinguishable from non-REM sleep. Rhythmic, sharp 3- to 6-Hz activity occurring in bursts, or runs, may be seen in the posterior head regions. These sharp waves are confined to the midline

109

FIGURE 4-10 • Hypnagogic hypersynchrony during drowsiness in a healthy 3-year-old girl. Time constant: 0.035 seconds. Calibration: 50 IlV; 1 second.

occipital electrode (Oz) until 4 months of age, after which they can be recorded from more widespread areas posteriorly. This activity is often associated temporally with rapid eye movements. It should be considered an entirely normal pattern. Ellingson and Peters demonstrated the rapid maturational changes that occur during the first 3 months after term. 14 The trace alternant pattern gradually disappears (between 20 and 47 days) and is replaced by diffuse slow-wave quiet sleep. Frontal sharp waves and delta brushes also disappear, and 14-Hz sleep spindles appear between 27 and 61 days. The onset of sleep shifts from active to quiet sleep with a decrease in the percentage of time spent in active sleep. The 12- to 14-Hz sleep spindles that characterize stage 2 and 3 sleep in older children and adults can be seen in the quiet sleep of some infants as early as the first month oflife and are present throughout quiet sleep in most infants aged 3 to 4 months.':' Slower spindles (10 to 12 Hz) are more common in younger infants. Approximately 86 percent of spindles are 12- to 14-Hz by 6 months of age. 97 Spindles are usually present symmetrically over the central regions but are often asynchronous during the first year. The percentage of asynchronous spindles is 36 percent at 1.5 months and falls to 25 percent thereafter." This asynchrony may persist until 2 years of age and is often seen in older children during the early stages of non-REM sleep.

110

ELECTROOIAGNOSIS IN CLINICAL NEUROLOGY

In the first year of life, spindles have a monophasic and often a mu or "comb-like" morphology and may exceed 10 seconds in duration. The longest spindles appear at 3 to 4 months of age. As the child enters the second year of life the duration of sleep spindles decreases, the interval between spindles increases, and their incidence declines. The incidence during nonREM sleep remains low until 4 to 5 years of age when there may be a moderate increase. In the authors' experience, this apparent decrease in incidence is partially explained by the tendency for spindles to occur in deeper stages of sleep than those occurring during the first year. Therefore the absence of sleep spindles in infants between 5 and 12 months of age is unusual and considered abnormal, whereas in older infants in very light stages of sleep, spindles tend to be sparse or absent. Hypothyroid infants have a strikingly reduced number of spindles." In some normal infants, sleep spindles may be more abundant over one hemisphere during the entire tracing. Only major and persistent amplitude asymmetries or the absence of spindles during quiet sleep in the second half of the first year should be considered abnormal. Asymmetry of the background activity will probably accompany the spindle asymmetry in infants with cerebral pathology. Slower spindles (10 to 12 Hz) can also occur in older children, are usually more frontal in location than the 12- to 14-Hz spindles, and tend to appear in somewhat deeper stages of sleep. Vertex sharp waves can usually be identified during non-REM sleep after 3 to 4 months of life.94 Their voltage field is more diffuse in infants than adults, and they often blend into the background activity, making identification difficult. In newborns during the first month of life, repetitive sharp waves confined to the midline central region (Cz) may represent precursors of vertex waves. Vertex waves consist primarily of monophasic (negative) or biphasic (negative-positive) sharp waves during the first 2 years oflife. They are higher in amplitude and sharper in children than in adults. In older children, vertex activity may occur in bursts and runs of repetitive. high-amplitude sharp waves, which are often admixed with sleep spindles. The repetition may occur three to nine times within 1 to 3 seconds." Such highamplitude, short-duration, and repetitive vertex waves should not be classified as epileptiform. Vertex waves are usually symmetric and synchronous, but transient amplitude asymmetries may occur, and often a series of vertex waves will appear unilaterally for several seconds over one hemisphere. In the authors' experience, typical K-complexes do not appear until 2 to 3 years of age. These are high-amplitude, biphasic slow waves (surfacepositive transient followed by a slower, surface-negative component) that occur spontaneously over the vertex

region, occasionally as a response to sound or other environmental stimuli during sleep. The background activity during sleep consists of an admixture of delta and theta rhythms, with the former appearing at highest amplitude in the occipital regions and the latter appearing frontally at lower amplitudes. Slater and Torres defined this so-called frequencyamplitude gradient as a "progressive decrement in voltage from occipital to frontal areas, with an accompanying decrease in slow frequencies in the same posterioranterior direction.t'l'" They described it in the EEGs of normal children from birth to 10 years, appearing most commonly after 3 to 4 months of age. The pattern was poorly developed or absent in a group of infants and children with a variety of acute and chronic illnesses. In children, particularly 1 to 2 years of age, the transition from light to deep stages of sleep may be associated with the appearance of bilateral, moderate- to highamplitude biphasic slow transients (cone waves) over the occipital regions."

Normal Paltems after the First 2 Years The normal EEG patterns of childhood and adolescence are described in several published reviews94.95.101,I02 and are briefly summarized here. Posterior slotu activity in the delta and theta range is common in children of all ages during wakefulness. There are no definite guidelines as to what constitutes excessive posterior slowing: each electroencephalographer must develop his or her own normative range, A conservative approach is suggested. Many EEGs are overread because of an exaggerated concern for slightly excessive theta activity during wakefulness. In many instances the slowing is caused by drowsiness. For this reason, technologists should stimulate all cooperative children for at least a portion of the record by engaging the patient in conversation, and should attempt to record the EEG before sleep. "Excessive slowing" is very common after arousal from sleep. The child should be given several minutes to awaken fully before the EEG can be considered to represent a truly waking state. Occipital slow waves (slow-fused transients, posterior slow waves of youth) consist of single, high-amplitude delta waves that are admixed with the ongoing rhythm." The alpha activity immediately preceding the slow transient often has a higher amplitude than that of preceding alpha waves and, in conjunction with the slow wave, creates a sharp and slow-wave complex that may be misinterpreted as epileptiform by an inexperienced reader. Rhythmic 2.5- to 405-Hz activity in the temporooccipital regions is common in children (25 percent,

Neonatal and Pediatric Electroencephalography

according to one study'?') and reaches maximal expression at 5 to 7 years of age. This activity may have a sharp morphology, particularly when it is admixed with the alpha activity, and usually appears in short runs or bursts. Doose and colleagues studied the genetics of two varieties of posterior slowing: parietal theta (4 to 6 Hz) and occipital delta (2 to 4 HZ).IO~.104 The EEGs of children with generalized epileptiform activity were more likely to contain these patterns. The normal siblings of epileptic children with parietal theta had a higher incidence of this pattern (13 percent) than did siblings of epileptic children without parietal theta (3.2 percent) or a control group of normal children (5.6 percent). The authors found this pattern to be extremely rare in full wakefulness in normal children. Bursts and runs of occipital delta are commonly seen in children with nonconvulsive generalized (petit mal) epilepsy. The delta activity in epileptic children usually occurs in high-amplitude bursts and is often admixed with spikes, forming a typical 3-Hz spike-wave complex. A unique posterior delta rhythm provoked by eyeclosure may occasionally be seen in normal children between 6 and 16 years (phi rhythm) .105.106 This rhythm consists of rhythmic high-amplitude (100 to 250 IlV) slow waves of 3 to 4 Hz lasting for 1 to 5 seconds after eye-closure, The phi rhythm is not associated with epilepsy, and it often lacks the sinusoidal morphology and accentuation with hyperventilation that are characteristic of the pathologic occipital delta patterns. The phi rhythm occurs only after eye-closure, usually after a period of concentrated visual attention, and lasts less than 3 seconds. 106 A mu rhythm is seen occasionally in young children; the incidence increases progressively with age. A typical 7- to II-Hz central rhythm with a comb-like form was found by Petersen and Eeg-Olofsson in 7.1 percent of normal children.'?' If one accepts all centrally located 7- lO II-Hz rhythms that are not attenuated by eyeopening, the incidence of mu exceeds 7 percent. Monophasic sharply contoured mu often occurs unilaterally in the central regions, where it may be misinterpreted as an abnormal pattern. I~yperventilation usually produces a buildup of highamplitude mixed delta and theta activity in normal children at all ages. This response most commonly consists of an increase in the pre-existing posterior slow activity (e.g., posterior slow waves of youth) or a buildup ofposteriorly dominant or diffuse slow-wave activity. In children younger than 7 years, the slow-wave buildup is maximal posteriorly. After 7 or 8 years, it is maximal over the anterior head regions. A buildup with hyperventilation is abnormal only if persistent focal slow waves or epileptiform potentials appear, if generalized spike-wave activity is activated, or if the buildup is

111

markedly asymmetric. If the buildup is asymmetric, the hemisphere with the lower amplitudes is the one that is often more involved pathologically. A generalized buildup of delta activity, even if it has a paroxysmal appearance, should be considered normal unless definite spikes are admixed with the slow waves. The admixture of high-amplitude alpha waves and a prominent buildup of occipital delta waves often leads to the appearance of sharp and slow-wave complexes that can easily be misinterpreted as epileptiform activity. Some children may overbreathe so vigorously that they transiently lose contact with the environment and fail to respond to the technologist's commands. This behavior, which may resemble an absence seizure, is accompanied by diffuse high-amplitude delta activity without spikes. North and colleagues described 18 children with this unusual response, which they called "pseudoseizures caused by hyperventilation."!"? Intermittent photic stimulation is performed routinely in most EEG laboratories. The main value of this technique is to provoke epileptiform activity (photoparoxysmal response). Photoparoxysmal responses may be provoked in normal children and have been reported in 7 to 9 percent of the normal population.t'v'!" In approximately one-half of these children, the response was characterized by bursts of high-amplitude slow activity, often with admixed spikes. In the remainder, the response consisted of bitemporo-occipital sharp and slow-wave or spike and slow-wave activity!" Photoparoxysmal response appears to be a genetically determined trait that occurs more often in the siblings of epileptic children with the response than in the control population.l'" Occipital spikes may be provoked in a very small number of nonepileptic children in the absence of a photoparoxysmal response by using a photostimulator with a fine grid pattern. 109 These spikes, which are confined to the occipitoparietal regions, gradually increase in amplitude as the stimulus is continued. They occur in a variety of nonepileptic disorders and are of doubtful pathologic significance. Photic stimulation is not routinely performed on newborns and young infants. Anderson and Torres reported that 66 percent of premature infants (27 to 32 weeks EGA) manifested photic driving at flash frequencies of 2 to 10 HZ.110 Colon and colleagues described an "off-response" in infants with a variety of neonatal conditions, particularly the respiratory distress syndrorne.!!' Further studies are needed to assess the clinical utility of these findings. Drowsiness and sleep present as many interpretative problems in older children as they do in those 2 years of age or younger. High-amplitude paroxysmal hypnagogic hypersynchrony, which in the authors' experience is maximally expressed between 3 and 5 years of

112

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

age, is still encountered in the EEGs of children as old as 12 years of age (see Fig. 4-20). Drowsiness in children is usually characterized by the gradual disappearance of the alpha rhythm, an increase in the amount offrontocentral beta activity, and the appearance of rhythmic posterior 2.5- to 6-Hz activity. Artifacts from slow horizontal eye movements may be noted in the lateral frontal leads. More slow rhythms are noted anteriorly in children older than 10 to 12 years of age. During light sleep, runs of high-amplitude, repetitive, sharp vertex waves are common and are frequently admixed with central sleep spindles, leading to complexes often classified as epileptiform. Sleep spindles may appear to have a spiky appearance when recorded with bipolar montages (Fig. 4-21). A referential montage reveals the benign morphology of this paroxysmal sleep activity. Familial and twin studies have shown the individual characteristics of the EEG are largely genetically determined. A low-amplitude background, for example, is known to have an autosomal dominant mode of inheritance, and the gene has been localized to the distal part of chromosome 20q. m Doose and colleagues have reported that alpha activity extending into the frontal regions is more common in the parents of children with primary generalized epilepsy (18 percent) than in those parents whose children have focal epilepsy (8 percent) or in control subjects (9 percent) .113 The authors concluded that their study "reveals a clear correlation between the type ofEEG background activity in parents and the EEG characteristics in their children."

EEG '.ltems of Dubious Significance Certain EEG patterns that present difficulties in interpretation are reviewed here. These patterns often have an epileptiform appearance, although their relationship to pathologic processes has not been clearly established.

Positive spike bursts oj 14 and 6 Hz are commonly recorded in the EEGs of adolescents during drowsiness and sleep and are considered a benign pattern by most electroencephalographers. They consist of brief rhythmic arciform discharges that are maximal in the posterior temporal regions. They may be seen in children as young as 3 to 4 years of age, but are most common at 13 to 14 years of age. 114 They are found in 14 to 55 percent of asymptomatic children depending on age and recording method. A high incidence of positive spike bursts has been reported in hepatic coma, including Reye syndrome. I Ei This activity appeared during deep stages of coma, in some instances only following auditory or tactile stimulation. The bursts disappeared or were less abundant as the children recovered. Furthermore, 14- and 6-Hz positive spikes have been reported in adults during hepatic coma.!" The significance of these observations is still unclear. Drury found 14- and 6-Hz positive spikes in 6 percent of 111 children with diverse encephalopathies (patients with either normal EEGs or with severe abnormalities such as burst-suppression, alpha-pattern coma, and isoelectric backgrounds were excluded).'!" The positive bursts in the abnormal children occurred in EEGs with moderate-to-marked background slowing. The incidence of 14- and 6-Hz positive spikes was not significantly different from that in a control population of 184 psychiatric inpatients aged 2 to 21 years. The morphology and topography were the same in the two study groups, and none occurred in patients with hepatic coma. Psychomotor variant discharge or rhythmic temporal theta bursts consist of 5- to 7-Hz sharply contoured activity that is often notched in appearance. This pattern is most prevalent in young adults 114 but can also be seen in adolescents. Originally this pattern was considered to be abnormal and was associated with vegetative states, psychiatric disorders, and temporal lobe epilepsy; however, this pattern has been found subsequently in a variety of benign conditions and is now regarded as a nonspecific finding. It has no significance with regard to epilepsy. Benign sharp transients oj sleep or small sharp spikes are commonly seen in the normal population, particularly if nasopharyngeal electrodes are used. These sleepactivated spikes have a high incidence in adults but can occur in teenagers. In a study of 120 normal subjects between the ages of 10 and 80 years, they were recorded in 24 percent.!" They were present in 10 percent of individuals between 10 and 20 years of age, and are exceedingly rare in children younger than 10 years of age.

Abnormal EEGs FIGURE 4-11 II Sleep spindle (sigma) and vertex waves in a healthy 3-year-old girl (same child as in Figure 4-20). Calibration: 50 IlV; 1 second.

The EEG can be helpful in evaluating children with a variety of proven or suspected neurologic disorders. In

Neonatal and Pediatric Electroencephalography

most instances the EEG is used to verify the organic nature of a particular neurologic disturbance, confirm a clinical suspicion as to the location of a particular lesion, follow the progression of a disease and the response to therapy, investigate the possibility that episodic behavioral or motor phenomena are epileptic in nature, or provide data that will help the clinician classify a seizure disorder and develop rational therapy. This section discusses some of the EEG patterns seen in children after the age of 1 month that are relatively specific for certain diseases, contribute significantly to the therapy of the patient, assist in reaching a diagnosis, or are of prognostic value. EPILEPSY

The EEG is important in the evaluation of children with episodic disorders and in defining childhood epileptic syndromes; however, the episodic nature of the problem results in a significant number of nondiagnostic EEGs. To increase the yield of abnormal records, the clinician should be familiar with the factors that may promote the appearance of epileptiform activity. These factors include obtaining adequate recordings during wakefulness and sleep at a time of day when the child is more likely to be having seizures; activating procedures, such as hyperventilation, photic stimulation, and sleep deprivation (in older children); and the recording of an EEG in close proximity to a seizure, particularly if a previous EEG was normal or nonspecifically abnormal.l!'' Occasionally, antiepileptic drugs should be discontinued, particularly if the diagnosis of epilepsy is uncertain or if the patient is being evaluated for the surgical excision of a focal epileptogenic area. During and immediately after antiepileptic drug withdrawal, bursts of generalized spike-wave activity may transiently appear in the EEGs of patients with focal epileptogenic lesions as well as in nonepileptic patients who are withdrawing from short-acting barbiturate drugs. Studies in adult patients with partial seizures have not found significant EEG changes in the period of drug withdrawal and have suggested that the increase in seizure frequency and the appearance of new epileptiform patterns represents the unmasking of pre-existing foci. 120 The type of epileptiform activity in an individual child with a single seizure type may change over serial records. Camfield and colleagues reported 159 children with epilepsy who had at least two EEGs (excluding those with absences, akinetic-atonic, and myoclonic seizuresr.!" A 40 to 70 percent discordance for the type of abnormality was found on the second EEG. For example, of the 42 children with major focal abnormalities on the first EEG, 7 had only generalized spikewave on the second. A total of 7 of the 17 with only generalized abnormalities on the initial recording showed only focal abnormalities on the second EEG.

113

Lombroso followed 58 neurologically and cognitively normal patients with idiopathic generalized epilepsy for 9 to 20 years. 122 Ultimately. 56 percent developed focal EEG features, particularly in the temporal or frontal regions. For example, focal EEG spikes and slowing emerged over time in 60 percent of the children with idiopathic generalized nonconvulsive (absence) seizures only. These reports emphasize that caution should be used when trying to establish the child's seizure type from the EEG alone. Certainly. generalized 3-Hz spikewave activity with a clinical absence is diagnostic of generalized absence epilepsy. If the same EEG or subsequent recordings contain focal spikes or slowing, the clinical diagnosis should not necessarily be changed to localization-related epilepsy. It should also be emphasized that epileptiform activity. including focal and generalized discharges, may persist in children whose seizures have stopped and who are no longer in need of chronic anticonvulsant drugs. In the long-term study of Shinnar and colleagues, epileptiform abnormalities were not associated with an increased risk of seizure recurrence in children withdrawn from medication after a mean seizure-free interval of 2.9 years. 123 Video-EEG monitoring. reliable 8-channel ambulatory EEG monitoring, and telemetry have remarkably expanded our ability to investigate episodic disorders of behavior. These techniques are providing electroclinical correlative data that were previously unavailable. Further discussion of this aspect is provided in Chapters 5 and 6. The EEGs of patients with various types of epileptic syndromes. defined according to the International Classification of Epilepsies and Epileptic Syndromes, 124 are detailed in Chapter 3. Certain additional aspects that relate particularly to children are discussed here.

Childhood Absence Epilepsy The EEG is only rarely normal in children with untreated childhood absence (petit mal) epilepsy (generalized nonconvulsive seizures). Repeatedly normal EEGs in a child with episodes of impaired external awareness are therefore more consistent with a partial seizure disorder or with nonepileptic phenomena. Long-term EEG monitoring of children with untreated absences has shown occasional periods during the day when a child may be free from clinical seizures and the EEG relatively devoid of epileptiform discharges.!" It is therefore important to determine before the EEG whether such a seizure-free period is recognized by the family so that the EEG can be arranged for some other time. The morphology of the 3-Hz spike-wave discharge is modified by sleep. The bursts during non-REM sleep

114

ElECTRODIAGNOSIS INClINICAL NEUROLOGY

are shorter in duration and more irregular in morphology, with generalized polyspike and polyspike-wave discharges replacing the typical 3-Hz spike-wave bursts. This polyspike activity does not appear to predict the concurrence or future appearance of generalized convulsions. The complexes seen during sleep must not be confused with the slow spike-wave activity of the Lennox-Gastaut syndrome. The EEG is of limited value in formulating a longterm prognosis for a child with childhood absence epilepsy. Sato and colleagues evaluated the long-term outcome of 89 patients and found that a normal IQ, absence of hyperventilation-induced spike-waves, male sex, and normal neurologic findings were predictive of cessation of seizures. 126 Abnormality of the EEG background activity was not related to outcome. Other studies have not confirmed the prognostic value of the EEG.m

Other Epileptic Syndromes with Absence Seizures Absence seizures are seen in four childhood epileptic syndromes: (1) childhood absence epilepsy; (2) juvenile absence epilepsy; (3) epilepsy with myoclonic absences; and (4) juvenile myoclonic epilepsy. Childhood absence epilepsy often begins between 4 and 6 years of age and, on rare occasions, under the age of 1 year, ]28 whereas the absence seizures in the other syndromes occur later in the first decade or in the teenage years. Panayiotopoulos and colleagues analyzed the clinical and electrographic features of 20 patients with absence syndromes.P" They found that in childhood absence epilepsy there was often a cessation of hyperventilation with an absence seizure because of a more profound loss of consciousness. They also found that the ictal discharge in juvenile absence epilepsy was longer than in childhood absence epilepsy or juvenile myoclonic epilepsy. Janz has pointed out that the paroxysmal discharges in the childhood form of absence epilepsy occur more commonly at 2.5 to 3.5 Hz, whereas in other forms that begin later in childhood or adolescence the frequency is faster. 130 He also noted that photosensitivity is equally common among male and female patients with the childhood form of absence epilepsy, but is significantly more common in females with other syndromes, particularly those with juvenile myoclonic epilepsy. Seizure frequency is much higher in childhood absence epilepsy, as are complex absences, including those with retropulsive movements. In juvenile myoclonic epilepsy, the characteristic EEG pattern consists of generalized bursts of spike-wave activity, polyphasic spike-wave activity (often accompanied by a myoclonic jerk), fragmentation of the discharges, and multiple spike complexes with a compressed capital W appearance.l'" The repetition

rate of the individual spike-wave complexes is usually 3 to 5 Hz, but is occasionally up to 10 Hz. In some individuals the epileptiform activity is attenuated or eliminated on eye-opening. 132 Approximately 25 percent of patients are photosensitive, and hyperventilation commonly activates the epileptiform activity. The EEG may occasionally be normal, particularly in individuals receiving antiepileptic medication. Focal and asymmetric EEGs and clinical features can be observed injuvenile myoclonic epilepsy. Aliberti and colleagues found focal slow waves, spikes and sharp waves, and focal onset of the generalized discharge in 37 percent of 22 patients with this disorder. 133 They also noted that transient discontinuation of the discharges, called "fragmentation," is characteristic, as are multiple spikes preceding the slow waves resulting in a compressed W shape. Clinical and EEG asymmetries often delay the diagnosis because of the concern that the patient has partial seizures.P" In Lombroso's long-term follow-up study of patients with idiopathic generalized epilepsy, focal abnormalities ultimately appeared in 56 percent; they were particularly common in individuals with both absence and generalized tonic-clonic seizures and were least common in patients with juvenile myoclonic epilepsy (11 percent) .122 The myoclonic jerks are temporally related to the generalized spike- or polyspike-wave discharges. These rapid shock-like contractions may involve either the proximal or distal muscles of the limbs, the face, and occasionally the trunk. There may be a transient postmyoclonic inhibition of the involved musculature leading to a loss of tone and, in occasional patients, to a flapping movement. This inhibition is coincident with the aftergoing slow wave of the generalized paroxysm. 135 An understanding of these syndromes allows prognostic statements about the duration of the epilepsy and the risk for the appearance of other seizure types. In many cases, however, the patient does not fit easily into one of the classic syndromes. Berkovic and colleagues have proposed that absence epilepsy is a "biologic continuum" from the healthy child with typical absence seizures and 3-Hz spike-wave activity to the retarded, motor-handicapped individual with atypical absences and the slow (less than 205-Hz) spike-wave pattern seen in the Lennox-Gastaut syndrome. 1:\6

Generalized Epilepsies Associated with Encephalopathy Three major and characteristic EEG syndromes occur in children with static or progressive encephalopathies: (I) hypsarrhythmia; (2) Lennox-Gastaut syndrome; and (:~) independent multifocal spike foci pattern. These syndromes are characterized primarily on the basis of specific EEG abnormalities and are associated with a

Neonatal and Pediatric Electroencephalography

heterogeneous group of diseases. They are the electrographic expression of a diffuse encephalopathy. Any neurologic disorder that is associated with a diffuse involvement of the neocortex may manifest one or all of these EEG patterns. HVfJ,mrrhythmia is the electrophysiologic expression of an encephalopathy that occurs primarily in the first year of life and is caused by any diffuse insult to the developing brain. The EEG is characterized by a diffusely abnormal background composed of high-amplitude delta activity admixed with multifocal spikes and sharp waves. The typical pattern is best expressed in non-REM sleep. The syndrome is expressed clinically by infantile spasms, delay of developmental milestones, and a very poor prognosis for ultimate intellectual development. Infantile spasms can be the sequelae of a specific severe neonatal insult such as hypoxic-ischemic encephalopathy and viral encephalitis (symptomatic infantile spasms), or it may be caused by a genetic disorder such as tuberous sclerosis complex. In approximately 10 percent of infants, no etiology is discovered (cryptogenic infantile spasms) despite examinations, neuroimaging studies, and metabolic and genetic testing. Neurodevelopmental outcome is more favorable in the cryptogenic group. It has been recommended that the word infantile be discarded because spasms can occur at any age. Spasms designate a specific seizure involving the axial musculature that tends to occur in clusters. In older children (e.g., those with the Lennox-Gastaut syndrome), typical infan tile spasms may gradually evolve into more prolonged tonic seizures and yet maintain the typical electrodecrernental ictal EEG pattern. Infants in the symptomatic group are often neurologically abnormal in the neonatal period and have abnormal EEGs. In many cases, however, the EEG and the neurologic examination normalize during the first few months of life. Subsequently, minor abnormalities such as focal sharp waves gradually appear in the EEG and evolve into a pattern offull-blown hypsarrhythmia. Partial seizures may evolve into infantile spasms.l'? Approximately 25 percent of children with hypsarrhythmia do not have infantile spasms but rather have other types of seizures. These children tend to have a poorer prognosis than those with spasms, because of more serious underlying etiologies. Children with the clinical syndrome of infantile spasms may have EEG abnormalities other than typical hypsarrhythmia. Hrachovy and colleagues have stressed the highly variable and dynamic EEG patterns in this syndrome, even in the same infant, over a 24-hour monitoring period.P" Modified hypsarrhythmia is any pattern that differs from the classic pattern and includes disorganized patterns with preserved interhemispheric synchronization, patterns with little or no spike and

115

sharp-wave activity, unilateral hypsarrhythmia, and those with focal abnormalities. Focal abnormalities and asymmetric hypsarrhythmia have assumed a more important role with the advent of surgical treatment of infantile spasms. 139. 140 A subgroup of children with refractory spasms and partial seizures have been identified whose EEGs or seizure symptomatology suggests focal cortical pathology and who benefit by removal of the involved cortex. The major abnormalities identified pathologically include cysticgliotic encephalomalacia and a variety of dysgenic or dysplastic lesions. These focal lesions may not be seen on early conventional MRI studies and may be found only by positron emission tomography or by single photon emission computed tomography. 141 ,142 The electrodecremental response is the common ictal pattern associated with a typical symmetric spasm; it consists of a high-amplitude generalized slow wave or a midcentral medium- to high-amplitude positive slow wave, followed by a diffuse attenuation of amplitude. Fast activity in the beta range, often with a spindle morphology, may occur alone or may precede or be superimposed on the slow wave. 14 3 Ictal patterns that accompany clinically asymmetric and asynchronous spasms, which are more common in individuals with focal or regional pathology, include focal spikes and sharp waves and unilateral or asynchronous paroxysmal fast activity in the hemisphere contralateral to the major clinical component of the spasm.J"? Asymmetric spasms involve only one side of the body or are stronger on one side than the other l 40 and are usually associated with an asymmetric hypsarrhythmia pattern. In a study by Gaily and colleagues of 60 infants, many of whom were referred for possible surgical intervention, 52 percent had asymmetric or asynchronous spasms and only 4 patients had hypsarrhythmia at the time of their evaluations; approximately half had had at least one EEG in the past that showed hypsarrhythmia.l'? Infants whose spasms were predominantly asymmetric or asynchronous never had cortical pathology that was exclusively posterior to the central area or in the temporal lobes, implicating the sensorimotor cortex in the generation of these particular spasms. Others have implicated the brainstem and basal ganglia as the primary site of spasm generation. 144 The EEG is of little value in formulating a prognosis for ultimate intellectual development. There is a tendency for full recovery in those infants whose initial EEGs are less severely abnormal and in those with typical, symmetric hypsarrhythmia. Two-thirds of these infants with cryptogenic infantile spasms may have a favorable neurodevelopmental outcome.U'' The prognosis is clearly worse in those with symptomatic spasms (e.g., patients with tuberous sclerosis, static encephalopathies from perinatal events, or with known metabolic or chromosomal syndromes).

116

ELEaRODIAGNOSIS IN CLINICAL NEUROLOGY

The Lennox-Gastaut syndrome includes a heterogeneous group of static and progressive encephalopathies that are grouped together on the basis of the presence of slow (2.5-Hz or less) spike-wave discharges on the EEG. The clinical spectrum that accompanies this particular EEG pattern includes mental retardation; high incidence of abnormalities on neurologic examination; and frequent occurrence of tonic, akinetic, atypical absence, and generalized tonic-clonic seizures, which are characteristically refractory to anticonvulsant drugs. The EEG pattern is commonly recorded during the first two decades of life, but seizure onset is most frequent between 6 months and 3 years of age. Occasional patients have had the onset of seizures in the teenage years. The interictal EEG may also show slowing of the background activity, focal or multifocal spikes and sharp waves, and diffuse polyspike-wave activity. The EEG abnormalities are enhanced during sleep. During sleep, bursts of activity at approximately 10 Hz ("recruiting rhythm") may be associated with subtle tonic seizures. The pathology underlying the syndrome varies widely, ranging from encephalopathies dating to the prenatal or neonatal period to progressive neurodegenerative diseases. Approximately one-third of patients have a prior history of infantile spasms and hypsarrhythmia. In a study of 15 children with drop attacks ("epileptic fall"), Ikeno and associates described 7 with seizures characterized by flexor spasms resembling infantile spasms, with the typical slow-wave and electrodecremental ictal pattern.l'" All had spasms in infancy. Tassinari and colleagues have described a related syndrome characterized by the onset of virtually continuous spike-wave activity during sleep.!" This striking activation of atypical spike-wave activity with sleep has also been seen in children with Landau-Kleffner syndrome':" and so-called atypical benign partial epilepsy of childhood.U" The multiple focal spike pattern is seen often in children with a prior history of hypsarrhythmia or the Lennox-Gastaut syndrome. This pattern is defined as "epileptiform discharges (spikes, sharp waves, or both) which, on any single recording, arise from at least three noncontiguous electrode positions with at least one focus in each hemisphere.l'P'' This EEG pattern is seen most often in patients during the first two decades of life. Among children with this EEG, most (84 percent) have seizures, usually of more than one type, but most commonly of the generalized tonic-clonic variety.l'" Many patients are intellectually subnormal, particularly if seizures started before the age of 2 years, and many have abnormal neurologic examinations. Multifocal spikes on a normal EEG background can sometimes be seen in children with benign partial epilepsies who are neurologically and cognitively normal. The normal neurodevelopmental history and the benign form of

their epilepsy differentiate these patients from those with the multifocal spike pattern.

Early-Onset Epilepsies Associated with Encephalopathy Early myoclonic encephalopathy is characterized by erratic, fragmentary myoclonic jerks that begin in the first month of life. These jerks are replaced by partial seizures; massive myoclonus; infantile spasms; and, infrequently, tonic seizures. The EEG shows a characteristic burst-suppression pattern; bursts are composed of spikes, sharp waves, and slow activity lasting 5 to 6 seconds and alternate with 4- to 12- second periods of attenuation. 151 The burst-suppression pattern often evolves into hypsarrhythmia or multifocal spikes and sharp waves. Etiology is often not determined, although some cases are familial in nature. The infants have marked developmental delay, abnormal tone, and microcephaly. Known etiologies include inborn errors of metabolism (particularly nonketotic hyperglycinemia), pyridoxine dependency, and brain malformations such as hydranencephaly. Ohtahara syndrome (early infantile epileptic encephalopathy with suppression-burst) is characterized by frequent tonic spasms with onset in the neonatal period or early infancy and a burst-suppression pattern.P'' The burstsuppression pattern is composed of bursts lasting 1 to 3 seconds, and the nearly flat suppression phase lasts 2 to 5 seconds (Fig. 4-22). The tonic spasms are associated with the bursts or with an abrupt attenuation of cerebral activity (desynchronization). Partial seizures are seen in one-third of the infants, but myoclonic seizures are rare. The seizure types may evolve from tonic spasms to infantile spasms with hypsarrhythmia (West syndrome), Lennox-Gastaut syndrome, or both. The infants have severe encephalopathy. Etiology is varied but often related to structural abnormalities including cerebral dysgenesis, porencephaly, hemimegalencephaly, Aicardi syndrome, and diffuse or focal cortical dysplasias. Less commonly, metabolic disorders are associated with Ohtahara syndrome. Ohtahara syndrome and early myoclonic encephalopathy have overlapping etiologies and EEG patterns, which may lead to confusion in their nosology. The initial seizure types in the two syndromes are quite different. The burst-suppression pattern in early myoclonic encephalopathy may be seen only during sleep, whereas it is observed in waking and sleeping states in Ohtahara syndrome.

Partial Seizures Partial (localization-related) seizures are very common in infants and children. Many occur in normal children and have a benign prognosis. Although it is reasonably

Neonatal and Pediatric Electroencephalography

117

AF3-T3

-==-~~-­ T~: AF3-C3

=

C3-01 AF~ C~2

T3-C3

C3-Cz

B

A

FIGURE 4-22 II Otahara syndrome in a 2-week-old term infant with uncomplicated pregnancy and delivery. The infant had experienced tonic seizures every few minutes since birth. A, EEG shows a burst-suppression pattern, with bursts composed of spikes and sharp waves admixed with slow waves. B, EEG during a brief tonic seizure shows a highamplitude sharp wave (arrow) followed by a complex slow wave and low-amplitude fast activity, much of the latter representing artifact. (Modified from Tharp BR: Neonatal seizures and syndromes. Epilepsia, 43:2, Sup pi 3, 2002, with permission. )

easy in older children and adults to clinically distinguish partial and generalized seizures, it may be quite problematic in younger children, particularly infants. The International Classification of Epileptic Seizures in current use was developed primarily for use in adults and is difficult to use in children, particularly those under 2 years of age. This has led to proposals for a new classification specifically for infants (Table 4-3) .153,154 It is difficult to classify seizures in infants because the clinical manifestations are quite different from those of older children and adults. In a video-EEG study,154 two skilled epileptologists were asked to determine whether

seizures were partial or generalized in a group of children under 26 months of age. There was very poor agreement between the observers when they viewed only video recordings of ictal events. The interrater agreement on seizure onset when they used only the EEG was better for generalized-onset seizures (0.79) than for partial seizures (0.54). These and other authors'P" found that it was virtually impossible to determine the level of consciousness or awareness in a seizing infant, thus making it impossible to classify partial seizures as simple or complex. Seizures in infants may be so subtle that even a skilled observer is unable

Seizure Type

Seizure Semiology

Astatic (atonic)

Sudden 1055 of tone involving one or more muscle groups or entire body Sudden change of behavior (usually an arrest or marked reduction of motor activity-hypomotor seizure); automatisms usually simple and consist primarily of oral-buccal movements Clonic jerking of one or more limbs, or generalized and occasionally involving face and/or eyelids only; may alternate from side to side or be asynchronous between each half-body (resembling a generalized motor seizure) Symmetric or asymmetric tonic posturing, occasionally asymmetric tonic neck-like posturing, sometimes with eye and head version; may be followed byclonic phase In older children, it may be difficult to distinguish a spasm from true myoclonus or a short tonic seizure

Behavioral Clonic Tonic

InflllltlJe spasm,tlnyodonlc epI'-kspa-

Veme Unclassified seizures

Turning of the eyes only or eyes and head; may be associated with behavioral changes or tonic involvement of limb(s) Seizures that cannot be categorized in above classification

Adapted from Acharya IN, Wyllie E, Luders HO et al: Seizure symptomatology ininfants with localization-related epilepsy. Neurology, 48:189, 1997; and Nordli DR, Brazil CW, Scheuer ML etal: Recognition and classification ofseizures ininfants. Epilepsia, 38:553, 1997, with permission.

118

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

determine if it is partial or generalized. Many tonic motor events may be bilateral in a partial seizure, whereas asymmetric tonic posturing may be associated with a generalized ictal EEG discharge. Well-developed complex automatisms are much less common than in older children, whereas simple oral-buccal movement (e.g., sucking) is more common. Focal dystonic posturing, a helpful lateralizing feature of partial seizures in older children and adults, is seen less commonly in infants and has less lateralizing value. Versive movements, particularly of the eyes, have little lateralizing value. The EEG may be a helpful adjunct in determining the etiology of partial seizures. If the background activity is normal, and the child is neurologically and cognitively normal, it is probable that the seizures will fall in to one of the benign syndromes. In these disorders, the frequency of the epileptiform activity may be very high and is often enhanced by sleep. The morphology of the abnormal activity is also rather distinctive, particularly in benign rolandic epilepsy and childhood occipital epilepsies. to

Benign Childhood Epilepsy with Centrotemporal Spikes (Benign Rolandic Epilepsy) Benign rolandic epilepsy is one of the most common partial epilepsy syndromes of childhood. It is characterized clinically by the occurrence of seizures in an otherwise normal child between the ages of 5 and 12 years. The seizures are typically partial (usually hemifacial, often associated with somatosensory symptoms) when they occur during waking hours and tend to generalize during sleep. The interictal EEG features biphasic, high-amplitude (l00 to 300 ~V) sharp waves in the centrotemporal (rolandic) regions and a normal background activity. The epileptiform discharges are often bilateral and asynchronous and may markedly increase in frequency during sleep. A characteristic EEG feature of the spike in this syndrome is the "horizontal dipole": simultaneous surface negative spikes in the central region and positive spikes in the frontal region. The central negative spikes can be of maximum amplitude anywhere along the central sulcus; these have been divided into "high central foci" and "low central foci." Legarda and Jaykar l55 found that the horizontal dipole, which occurred in 50 percent of their patients, occurred with equal frequency in patients with high and low foci and was as common in children with seizures as in those without seizures. Gregory and Wong found that children without the dipole discharges were more likely to have more frequent seizures, developmental delay, school difficulties, and an abnormal neurologic examination compared with a group having rolandic

discharges with a horizontal dipole.l'" A more detailed EEG analysis found that a topographic pattern of nonstationary fields represented by a "double" spike-wave complex was highly correlated with seizure occurrence'F: by contrast, children without seizures were more likely to have a topographic pattern of stationary potential fields (i.e., a single spike-wave complex). The complex spike consisted of a sequence of dipoles with a small initial spike that was negative in the frontal region and positive in the centrotemporal regions, followed by a more obvious higher-amplitude spike that was negative centrotemporally and, when present, positive frontally. There were no significant morphologic (as differentiated from topographic) differences of the spike that provided a clue to its "epileptogenicity" or to the presence or absence of an organic lesion. LoB Other authors, by contrast, were able to distinguish children with uncontrolled benign focal epilepsy from those with focal spikes and other syndromic categories (e.g., symptomatic, cryptogenic, and Landau-Kleffner) on the basis of spike morphology using a single morphologic index, the composite spike parameter, which was derived from the amplitude, duration, and sharpness of the spike. I:;\) Magnetoencephalographic analysis of rolandic discharges has shown that the prominent negative sharp wave appears as a tangential dipole in the rolandic region, with the positive pole situated anteriorly, and is presumably generated through a mechanism similar to that for the middle-latency components of the somatosensory evoked responses from stimulation of the lower lip.160 The EEGs of a significant percen tage of children with benign rolandic epilepsy also contain generalized spikewave discharges typical of an idiopathic generalized epilepsy (approximately 10 to 15 percent); bilateral or shifting foci (34 percent) 161; multiple independent foci (9.8 percent) 162; and, occasionally, a photoparoxysmal response to intermittent photic stimulation. Long-term follow-up studies reveal that most children with benign rolandic epilepsy become seizure-free in their midteens. Rarely is any focal pathology demonstrable, and neuroimaging studies are not usually indicated if the seizure symptomatology and EEG are characteristic. An autosomal dominant gene with agedependent penetrance is presumably responsible for the EEG trait. Focal centrotemporal spikes are therefore quite common in siblings of children with this syndrome and can also occur sporadically in normal children (3.7 percent of 1,057 normal children studied by Okubo and colleagues'P"). Although benign rolandic epilepsy has an excellent prognosis with regard to seizure remission, behavior problems and cognitive dysfunctions may sometimes develop in its course. Comprehensive neuropsychologic studies have shown that children with rolandic epilepsy

Neonatal and Pediatric Electroencephalography

have a high incidence of learning disabilities, developmental language disorders, and attention deficits.F" Children with significant learning or behavioral problems were more likely to have long spike-wave clusters, generalized 3- to 4-Hz spike-wave discharges, asynchronous bilateral spike-wave foci, and focal slow waves. 165

Other Benign Epileptic Syndromes Occipital spikes are seen in two childhood epilepsy syndromes, early-onset benign occipitalseizuresusceptibility !>yndrome (EBOSS) and childhood epilepsy with occipital paroxysms (CEOP). Both syndromes have similar EEG features of unilateral or bilateral and synchronous occipital spikes and spike-wave discharges that are often very abundant and increase during sleep. During wakefulness the spikes are promptly suppressed or attenuated in the majority of patients by eye-opening (71 percent) or central fixation (88 percent). The seizures in EBOSS begin at a young age (between 3 and 7 years of age) and are characterized by vomiting, tonic eye and head deviation, progressive alteration of consciousness, and autonomic symptoms. They often progress to hemiconvulsions or generalized tonic-clonic seizures. Prolonged partial seizures may develop into status epilepticus. Seizures in this syndrome are predominantly nocturnal and infrequent (with half of the children having only one or two seizures). The epilepsy course is shorter, and in most children the seizures subside within 1 to 2 years.l'" In one recorded seizure, the ictal EEG showed the entire seizure dominated by high-amplitude, 2-Hz slow waves intermixed with spikes or polyspikes localized in the occipital region, spreading to the temporal region. 167 The onset of seizures in CEOP occurs in slightly older children (usually 7 to 10 years of age). Seizures are characterized by initial visual symptoms (usually in older children), which may be followed by a tonic deviation of the eyes, vomiting, and hemiclonic or generalized tonic-clonic seizures. The visual symptomatology includes amaurosis, simple or complex hallucinations, and illusions. Often the seizures are followed by headaches that have a migrainous quality. Seizures in CEOP are diurnal and more likely to be multiple. The EEC demonstrates unilateral or bilateral paroxysms of high-amplitude, rhythmic spikes or sharp waves in the occipital regions.l'" These epileptiform discharges atteu nate with eye-opening and are activated by the elimination of central vision and fixation (fixation-off sensitivity). During seizures, the occipital discharges spread to the centrotemporal regions. The seizures are relatively infrequent and reportedly usually remit spontaneously, although the occipital paroxysms may outlast the clinical remission for several years.l'" Cooper and Lee, however, have presented a somewhat more pes-

119

sirnistic long-term outlook, although their study included patients with symptomatic epilepsy.l'" From a retrospective analysis of 33 patients with reactive occipital epileptiform activity, they found that only 9.1 percent of patients were able ultimately to discontinue anticonvulsant drugs and that 63.6 percent continued to have uncontrolled seizures at follow-up of 6 months to 8 years. Occipital spikes may also occur in normal nonepileptic children, in children with the benign and malignant multifocal spike syndrome, and commonly in children with visual impairment that is either congenital or acquired.l"" Spikes associated with visual impairment initially have a short duration, but they increase in duration and amplitude as the child grows, becoming associated with a slow wave. They disappear by the teenage years or in young adulthood. Benign syndromes of partial epilepsy have also been described in children whose EEGs contain foci in the temporal, parietal, and frontal regions. These disorders are less common than the rolandic and occipital syndromes, and the interictal epileptiform discharges have no characteristics that distinguish them from those occurring in more malignant partial seizure syndromes.

Febrile Seizures The EEG is of little value in the evaluation of children with typical (simple) febrile seizures. The majority of children with febrile seizures have normal EEGs during the following week, 17l although some may show nonspecific disturbances such as excessive posterior slowing. Epileptiform activity rarely occurs in the immediate postseizure period. Another etiology must be considered if focal spikes or generalized epileptiform discharges are present. Paroxysmal activity subsequently develops in a significant number of children with a past history of febrile seizures (29 percent according to Lennox-Buchthal'J'' and less often in the authors' experience). This activity consists primarily of generalized spike-wave activity and focal spikes, usually centrotemporal or occipital in location. Children who develop these paroxysmal abnormalities do not have a higher risk for the development of nonfebrile seizures when compared with children whose EEGs lack paroxysmal activity.172 SchiottzChristensen and Hammerberg confirmed these findings in a follow-up study of 59 pairs of twins of the same sex. 173 The incidence of paroxysmal EEG abnormalities was significantly higher in the twin who had had a febrile seizure than in the normal twin. The EEG does not appear to be of value in predicting which child will have future febrile seizures or develop epilepsy. Risk factors for epilepsy after febrile seizures are clinical: family history of nonfebrile seizures, preexisting neurologic abnormality, and a complex febrile

120

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

seizure (greater than 15 minutes duration, more than one seizure in 24 hours, focal features). J74

Acquired Epileptic Aphasia (Landau-Kleffner Syndrome) Acquired epileptic aphasia, a disorder first described by Landau and Kleffner, is characterized by the appearance of a progressive aphasia in a previously healthy child. The degree of aphasia appears roughly to parallel the EEG abnormalities. Seizures commonly occur after the onset of the aphasia but are easily treated with antlepileptic medication. Standard antiepileptic treatment usually has no effect on the aphasia, which may improve over years but infrequently disappears completely. The etiology is unknown. The EEG is characterized by slow spike-wave discharges that are of higher amplitude over the temporal regions and spread diffusely through both hemispheres, giving the appearance of a generalized discharge. A striking EEG abnormality often occurs during sleep, also called "continuous spikes and waves during slow sleep" or "electrical status epilepticus during slow-wave sleep." This latter syndrome can occur independently of the Landau-Kleffner syndrome and is often associated with progressive cognitive and behavioral deterioration. 17.'i- 177 With neurophysiologic techniques, including invasive electrical recording and evoked potential studies as well as radiologic procedures including positron emission tomography, the focal epileptiform disturbance can be localized to the language cortex in one temporal lobe, with evidence of spread to homologous regions of the contralateral hemisphere. INFECTIOUS DISEASES

The EEG abnormalities in children with CNS infections are usually nonspecific. Localized slowing or epileptiform activity may indicate focal pathology (e.g., abscesses or subdural empyemas) and a unilateral attenuation of the background activity may suggest a subdural effusion. As a rule, however, the EEG has not been of particular value in the detection of subdural effusions and has been replaced by CT scans or MR!. This discussion is therefore limited to two infectious illnesses associated with EEG patterns that are relatively specific and may suggest the proper diagnosis.

Herpes Simplex Encephalitis Herpes simplex encephalitis is often associated with a characteristic periodic EEG pattern. Although this periodic pattern is seen more commonly in adults, it also occurs in children even during the first year of life. The periodic activity consists of high-amplitude, monophasic or biphasic sharp slow waves that recur every 2 to

4 seconds. These complexes appear focally, primarily in the frontotemporal area, or diffusely over one hemisphere with variable spread to the contralateral side. In neonatal herpes simplex encephalitis, a distinctive EEG pattern of multifocal periodic or quasiperiodic pattern (each focus with independent morphology and periodicity) has been found." Periodic EEG patterns may occasionally occur in other viral encephalitides of childhood (e.g., Epstein-Barr virus infection) .178

Subacute Sclerosing Panencephalitis (Dawson's Encephalitis) The periodic EEG complexes that characterize subacute sclerosing panencephalitis (SSPE) consist of relatively stereotyped, high-amplitude, bilaterally simultaneous bursts of slow activity and sharp waves with a periodicity approximating 3 to 10 seconds (occasionally as long as 90 seconds). They are usually accompanied clinically by myoclonic jerks. They occur early in the illness (stage II) and may be associated with normal waking and sleep activity. They may be asymmetric and, even less commonly, asynchronous. Occasionally, the periodic activity is focal, in which case it usually consists of repetitive, sharp slow-waves that have a less complex morphology than the more generalized complexes. The complexes occasionally appear only during sleep, with normal or nonspecifically abnormal tracings during wakefulness.'?" Prominent epileptiform activity in the form of spikes, sharp waves, or spike-wave discharges may accompany the bursts. This activity may assume a pattern resembling that seen in Lennox-Gastaut syndrome.P? Patients have been described with long runs of generalized epileptiform activity and a clinical state resembling absence (petit mal) status. The typical periodic pattern may appear after the epileptiform activity is suppressed by intravenous diazepam.P" Lombroso described one child presenting with the clinical and electrographic features of epilepsia partialis continua who never developed a periodic EEG pattem.l'" On rare occasions, repetitive myoclonic jerks occur during the waking state without the accompanying EEG complexes. More commonly, the EEG complexes persist during sleep after disappearance of clinical myoclonus. Periodic complexes resembling those of SSPE have been reported in other forms of encephalitis. Chronic progressive rubella encephalitis, which may resemble SSPE clinically, does not appear to manifest periodic EEG complexes. Townsend and co-workers described the EEG findings in one case as "generalized periodic low amplitude polyspike complexes that were closely related to myoclonic movements. "181 A comparison with the periodic complexes of SSPE was not made. Unfortunately, detailed EEG studies of this rare syndrome have not been published.

Neonatal and Pediatric Electroencephalography

Periodic patterns resembling those of SSPE may also occur in other neurologic disorders such as the postictal state in a child with seizures arising from the orbitofrontal region,182 phencyclidine (PCP) intoxication,183 and Rett syndrome.l'" GENETIC SYNDROMES WITH SPECIFIC EEG PATTERNS

121

that resembles facial or upper-limb myoclonus. IHn After 4 years of age, the typical EEG findings are replaced by focal epileptiform activity. Occasionally, girls with Angelman syndrome clinically resemble those wit.h Rett syndrome. The strikingly different EEG findings in these two disorders, particularly during t.he second year of life, are helpful in arriving at the proper diagnosis.!'"

Angelman Syndrome

Rett Syndrome

Angelman syndrome is characterized by severe mental retardation, inappropriate laughter, jerky movements, ataxic gait, severe speech impairment, and epilepsy. It is usually caused by a de novo deletion of maternally inherited chromosome 15 in the 15ql1-q13 critical region. The EEG may have many of the features of the LennoxGastaut syndrome, as well as other abnormalities that suggest the diagnosis. The typical EEG findings in young children consist of (1) high-amplitude persistent rhythmic theta waves; (2) runs of high-amplitude, usually anterior delta waves (2 to 3 Hz), often admixed with spikes; and (3) spikes mixed with high-amplitude, 3- to 4-Hz component" posteriorly that are facilitated or seen only with eyeclosure. 185 The epileptiform activity is often quite striking and at times may be associated with jerky motor behavior

Rett syndrome, a progressive encephalopathy in girls, is characterized by severe mental retardation with autistic features, stereotyped hand-wringing movements, episodic hyperventilation, seizures, and acquired microcephaly. The diagnosis is made by well-established clinical criteria. Approximately 80 percent of females with classic Rett syndrome have a mutation in the MECP2 gene (chromosomal locus Xq28). The EEG may also be helpful in confirming the diagnosis, particularly in later stages of the disorder. The EEG is usually normal or nonspecifically abnormal during stage I and early stage II of the disease.IH4.IHH Epileptiform activity (consist.ing primarily of spikes and sharp waves in the centrotemporal region) and slowing of the background appear in stage II (Fig. 4-23). The

Fp1-F7~.·"'--'\./V'--.--.·;~rv~ ~;v~\r--J~~

F7-T31~~1N1~

T3-T5 ............~~;I'--".\-r-

TS-'01~~\f\P~~r0N

:~~~~

Fp1-F3

F3-C3~{~lr,~Y

C3-P3....i,v--':Ir-~rl~

0~ /'v"',""'~J'v'~~-v---,r-r---..,'-'v-"~"'i •. 1 I

p3-01.-.yvoJlr0 ~rJrJ/ Fp2-F4

F4_C4-~·f\. -oJ~,{"'vJ'---yftV"'....,r'lV'4f'4\r~v-'0r\JJ"~~~

~~

C4-P4~~~~~~~'-'#f'~~~'~

P4-0~~\r''--'.fl~~

~

Fp2-F8 -~--"----'-'Ir------'\...'I,.../'-.r'--~r--,I\ Fa-T4~.~-I"oI'-'"·~-"~~"""'v"~J-·-·1\

~.v.---v
T4-T6 -.".."""-J".........I"'\-~""-"J'vvo~-A-.."...rJ'\f"V'-...JV\r-./--r-.--J\-~f'o---'''''''';V-\(''wN

L-.....

T6-o2~'~~\('-"'...........-"'J",J'_i\rv_".vv-A/\~\~v".", ....·,_!v--.-tv--I\r~ FIGURE 4·23 • Centrotemporal spikes during sleep in a 6-year-old girl with stage III Rett syndrome. The spikes are repetitive and predominate in the left hemisphere. Calibration: 50 flV; I second.

122

ElECTRODIAGNOSIS IN CLINICAL NEUROLOGY

spikes may be repetitive, enhanced by sleep, and occasionally triggered by tactile stimulation. 188,l 89 In stage III (the "pseudostationary stage," characterized by increasing seizures, mental retardation, gait difficulties, and pyramidal signs), the EEG background continues to deteriorate and multifocal spikes and slow spike-wave discharges resembling those of Lennox-Gastaut syndrome are common. The epileptiform activity often wanes during stage IV, which is associated with severe motor and mental handicaps, the necessity for the patient to use a wheelchair, and a gradual decline in the frequency of the seizures. Hagne and colleagues have pointed out that the EEG may be helpful in confirming the diagnosis ofRett syndrome in a patient who fits the clinical syndrome, if any of the following patterns are present: abundant centrotemporal spikes (which resemble those of benign rolandic epilepsy); unusual pseudoperiodic delta bursts; monorhythmic 4- to 5-Hz background slowing; and continuous, rhythmic, bilaterally synchronous, and generalized I-Hz spikes (that resemble those seen in Creutzfeldt-jakob disease) .184 Other syndromes that may resemble Rett syndrome, particularly during the early years of life, include autism, biotin-dependency disorders, and Angelman syndrome.l''? PROGRESSIVE NEUROLOGIC SYNDROMES

The EEG abnormalities accompanying most of the progressive neurologic syndromes of childhood are nonspecific and consist of variably slowed background activity and focal or generalized epileptiform discharges. In the early phases of the diffuse encephalopathies, a distinction can sometimes be made between a primary gray or white matter disorder on the basis of the EEG pattern. 190 Continuous non paroxysmal, polymorphic delta activity is more commonly associated with leukoencephalopathies, whereas multifocal cortical or generalized epileptiform activity in association with bilateral, paroxysmal slow activity is more consistent with neuronal disease, including cortical and subcortical gray matter. As the disease progresses, this distinction becomes less prominent. Several neurodegenerative disorders have EEG abnormalities that are sufficiently characteristic to warrant discussion (Table 4-4).

Neuronal Ceroid Lipofuscinosis Neuronal ceroid lipofuscinosis is a group of inherited, progressive, lysosomal-storage disorders characterized by progressive mental and motor deterioration, visual loss, seizures, and early death. Among the many clinical and pathologic SUbgroups in this disease, three have

TABLE 4-4 • EEG Patterns and Associated Disorders EEG Pattern

Disorder

Comb-like rhythm

Maple syrup urine disease Propionic acidemia Fast central spikes Tay-Sachs disease Rhythmic vertex positive spikes Sialidosis (type I) Vanishing EEG Infantile neuronal ceroid lipofuscinosis High-amplitude 16- to 24-Hz activity Infantile neuroaxonal dystrophy Diminished spikes during sleep Progressive myoclonic epilepsy Giant somatosensory evoked Progressive myoclonic epilepsy potentials Marked photosensitivity Progressive myoclonic epilepsy and neuronal ceroid Iipofuscinosis, particularly late infantile type Burst-suppression Neonatal citrulinemia Nonketotic hyperglycinemia Propionic acidemia Leigh disease o-Glyceric acidemia Molybdenum cofactor deficiency Menkes syndrome Holocarboxylase synthetase deficiency Neonatal adrenoleukodystrophy Hypsarrhythmia Zellweger syndrome Neonatal adrenoleukodystrophy Neuroaxonal dystrophy Nonketotic hyperglycinemia Phenylketonuria Carbohydrate-deficient glycoprotein syndrome (type III) Adapted from Nordli DR, De VIVO DC: Classification of infantile seizures: implications foridentification and treatment ofinborn errors of metabolism. J Child Neurol, 17: Suppl 3, 353-358, 2002, with permission.

been reported with peculiar EEG abnormalities (see Table 4-4). Infantile neuronal ceroid lipofuscinosis (SantavuoriHaltia type) has onset in the first two years of life and is characterized clinically by a regression of developmental milestones followed by myoclonus. Serial EEGs in this form reveals a progressive diminution in amplitude, culminating in isoelectricity.l'" The rapid development of an isoelectric EEG in conjunction with gradual loss of the electroretinogram (ERG) and visual evoked potentials (VEPs) is unique to this forms of neuronal ceroid lipofuscinosis. In the classic late infantile form (Jansky-Bielschowsky type), the EEG shows pseudoperiodic epileptiform discharges and a characteristic high-amplitude response to photic stimulation. 192 A high-amplitude VEP and lowamplitude or absent ERG are noted. This disease begins at 2 to 4 years of age, usually with seizures.

Neonatal and Pediatric Electroencephalographv

lnjuvenile neuronal ceroid lipofuscinosis (SpielmeyerVogt-Sjogren type), progressive visual loss begins at 4 to 7 years of age. The EEGs contain distinctive runs of high-amplitude spike-wave complexes. The ERG disappears at an early stage of the illness and the YEP gradually becomes of low amplitude.

123

missible agent. 199 The EEG shows, in addition to diffuse background abnormalities, unusual, high-amplitude (200 to 1,000 IlV) slow activity (I Hz or less) with superimposed lower-amplitude polyspikes that are often asymmetric and are more prominent over the posterior head regions, where the most significant neuronal loss is found at autopsy.200

Cherry-Red Spot-Myoclonus Syndrome Cherry-red spot-myoclonus syndrome (sialidosis I), a rare metabolic disorder occurring in the first few decades of life, is caused by an isolated deficiency of a lysosomal neuraminidase or sialidase. Patients suffer from a severe myoclonus, cerebellar ataxia, visual impairment, and generalized convulsions, and are noted to have cherry-red spots on the macula. In a young woman with this syndrome, who also had a peripheral neuropathy, EEGs contained bursts of primarily positive, centroparietal and vertex spikes that often correlated with the myoclonic jerks. 193 Engel and colleagues described two patients with this syndrome whose EEGs contained unusual low-amplitude fast activity with 10- to 20-Hz vertex positive spikes that correlated with the myoclonus.l'" The EEG in sialidosis type II (deficiency of betagalactosidase and sialidase) is less specific and consists of bursts of generalized spike- or polyspike-wave discharges (often coincident with myoclonic jerks) and high-amplitude somatosensory evoked potentials, common findings in many of the myoclonic epilepsies.l'" Mitochondrial Encephalopathy An increasing number of progressive and often familial neurologic syndromes have been associated with an abnormality of mitochondrial metabolism, including a variety of defects in respiratory chain and phosphorylation-system enzymes. Two common phenotypes are myoclonic epilepsy and ragged-red fibers (MERRF); and mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes (MElAS). Other disorders now considered to fall into this category include the Ramsay Hunt syndrome and other progressive myoclonic epilepsy syndromes, Alpers' disease, Leigh disease, and some of the "system degenerations." So and colleagues described the electrophysiologic features of] 3 patients with MERRF syndrome.P" These included generalized, bilaterally synchronous, atypical spike-wave discharges; multiple spike-wave bursts; and focal, often occipital, spikes with a slowed background. Only three patients had a photoparoxysmal response. Alpers' disease is characterized by a progressive encephalopathy with intractable seizures and liver disease, the latter consisting of fatty changes, cell loss, fibrosis, and often cirrhosis.I'" The etiology of this disorder is still unknown though some investigators have incriminated mitochondrial dysfunction 198 and a trans-

Infantile Neuroaxonal Dystrophy Infantile neuroaxonal dystrophy is a progressive neurodegenerative disorder of infancy characterized pathologically by accumulations of large eosinophilic axonal swellings or spheroids. A characteristic EEG pattern has been described, consisting of high-amplitude, unreactive, 16- to 24-Hz rhythms and an absence of responses to a variety of sensory stimuli.i'" However, in a juvenile form of this disorder, two brothers had EEGs that lacked this striking beta activity but were characteristic of a myoclonic epilepsy.202 The EEG and evoked potential alterations were similar to those described in other progressive myoclonic encephalopathies. BRAIN DEATH

The Task Force for the Determination of Brain Death in Children has set forth the criteria for determining brain death in children.I" The Task Force recommended that EEGs be obtained routinely in children between the ages of 7 days and 1 year but not in older children if an irreversible cause exists. Most neurologists believe that the EEG is a useful adjunct to the determination of brain death if it is performed by an experienced technologist and interpreted by an experienced electroencephalographer, as is discussed in Chapter 34. The criteria for irreversible brain death are not clearly established in infants less than 3 months of age.

ACKNOWLEDGMENT The authors thank Dr. Seju Kim for assistance with the updates and review of the chapter.

REFERENCES 1. Tharp BR, Scher MS, Clancy RR: Serial EEGs in normal and abnormal infants with birth weights less than 1200 grams: a prospective study with long-term follow-up. Neuropediatrics, 20:64, 1989 2. Gaultier C: Apnea and sleep state in newborns and infants. BioI Neonate, 65:231, 1994 3. Clancy RR, Bergqvist AG, Dlugos DJ: Neonatal electroencephalography. p. 161. In Ebersole JS, Pedley TA (eds): Current Practice of Clinical Electroencephalography. 3rd Ed. Lippincott Williams & Wilkins, Philadelphia, 2003

124

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

4. American Electroencephalographic Society: Guideline two: minimum technical standards for pediatric electroencephalography.J Clin Neurophysiol, 3: Suppll, 7,1986 5. Stockard-Pope JE, Werner SS, Bickford RG: Atlas of Neonatal Electroencephalography. 2nd Ed: Raven Press, New York, 1992 6. Lombroso CT: Neonatal Electroencephalography. p. 725. In Niedermeyer E, Lopes da Silva F (eds): Electroencephalography, Basic Principles, Clinical Applications and Related Fields. 2nd Ed. Urban & Schwarzenberg, Baltimore, 1987 7. Curzi-Dascalova L, Peirano P,Morel-Kahn F: Development of sleep states in normal premature and full-term newborns. Dev Psychobiol, 21:431,1988 8. Hughes JR, Miller JK, Fino lJ et al: The sharp theta rhythm on the occipital areas of prematures (STOP): a newly described waveform. Clin Electroencephalogr, 21:77, 1990 9. Biagioni E, Frisone MF, Laroche S et al: Occipital sawtooth: a physiological EEG pattern in very premature infants. Clin Neurophysiol, 111:2145, 2000 10. HughesJR, FinolJ, Hart LA: Premature temporal theta. Electoencephalogr Clin Neurophysiol, 67:7, 1987 11. Biagioni E, Bartalena L, Boldrini A et al: Background EEG activity in preterm infants: correlation of outcome with selected maturational features. Electroencephalogr Clin Neurophysiol, 91:154,1994 12. Watanabe K, Iwase K: Spindle-like fast rhythms in the EEGs oflow birthweight infants. Dev Med Child Neurol, 14:373, 1972 13. Lombroso C: Quantified electrographic scales on 10 preterm healthy newborns followed up to 40-43 weeks of conceptional age by serial polygraphic recordings. Electroencephalogr Clin Neurophysiol, 46:460,1979 14. Ellingson R, Peters ]: Development of the EEG and daytime sleep patterns in normal full-term infants during the first 3 months of life: longitudinal observations. Electroencephalogr Clin Neurophysiol, 49:112, 1980 15. HahnJS, Monyer H, Tharp BR: Interburst interval measurements in the EEG of premature infants with normal neurological outcome. Electroencephalogr Clin Neurophysiol, 73:410, 1989 16. Selton D, Andre M, HascoetJM: Normal EEG in very premature infants: reference criteria. Clin Neurophysiol, 111:2116,2000 17. Hayakawa M, Okumura A, Hayakawa F et al: Background electroencephalographic (EEG) activities of very preterm infants born at less than 27 weeks gestation: a study on the degree of continuity. Arch Dis Child Fetal Neonatal Ed, 84:F163, 2001 18. Anderson CM, Torres F, Faoro A: The EEG of the early premature. Electroencephalogr Clin Neurophysiol, 60:95,1985 19. Hayakawa F, Watanabe K, Hakamada S et al: Fz theta/alpha bursts: a transient EEG pattern in healthy newborns. Electroencephalogr Clin Neurophysiol, 67:27, 1987 20. Challamel MJ, Isnard H, Brunon AM et al: Asymetrie EEG transitoire 11 l'entree dans le sommeil calme chez Ie nouveau-ne: etude sur 75 observations. Rev Electroencephalogr Neurophysiol Clin, 14:17, 1984

21. O'Brien MJ, Lems YL, Prechtl HF: Transient flattenings in the EEG of newborns: a benign variation. Electroencephalogr Clin Neurophysiol, 67:16, 1987 22. Tharp BR, Cukier F, Monod M: The prognostic value of the electroencephalogram in premature infants. Electroencephalogr Clin Neurophysiol, 51:219,1981 23. Clancy RR, Tharp BR, Enzmann D: EEG in premature infants with intraventricular hemorrhage. Neurology, 34:583, 1984 24. Holmes GL, Lombroso CT: Prognostic value of background patterns in the neonatal EEG.J Clin Neurophysiol, 10:323, 1993 25. Van Lieshout HBM,JacobsJWFM, RotteveellJ et al: The prognostic value of the EEG in asphyxiated newborns. Acta Neurol Scand, 91:203,1995 26. Takeuchi T, Watanabe K: The EEG evolution and neurological prognosis of perinatal hypoxia neonates. Brain Dev, 11:115, 1989 27. Pezzani C, Radvanyi-Bouvet MF, Relier JP et al: Neonatal electroencephalography during the first twenty-four hours of life in full-term newborn infants. Neuropediatrics, 17:11, 1986 28. Nguyen The Tich S, Vecchierini MF, Debillon T et al: Effects of sufentanil on electroencephalogram in very and extremely preterm neonates. Pediatrics, Ill: 123, 2003 29. Young GB, da Silva OP: Effects of morphine on the electroencephalograms of neonates: a prospective, observational study. Clin Neurophysiol, Ill: 1955, 2000 30. Task Force for the Determination of Brain Death in Children: Guidelines for the determination of brain death in children. Ann Neurol, 21:616, 1987 31. Ashwal S, Schneider S: Brain death in the newborn. Pediatrics, 84:429, 1989 32. American Electroencephalographic Society: Guideline three: minimum technical standards for EEG recordings in suspected cerebral death. J Clin Neurophysiol, 3: Suppll, 12, 1986 33. Douglass LM, WujY, Rosman NP et al: Burst suppression electroencephalogram pattern in the newborn: predicting the outcome. J Child Neurol, 17:403, 2002 34. Sinclair DB, Campbell M, Byrne P et al: EEG and longterm outcome of term infants with neonatal hypoxicischemic encephalopathy. Clin Neurophysiol, 110:655, 1999 35. Biagioni E, Bartalena L, Boldrini A et al: Constantly discontinuous EEG patterns in full-term neonates with hypoxic-ischaemic encephalopathy. Clin Neurophysiol, 110:1510, 1999 36. Menache CC, Bourgeois BF, VolpelJ: Prognostic value of neonatal discontinuous EEG. Pediatr Neurol, 27:93, 2002 37. ConnellJA, Oozeer R, Dubowitz V: Continuous 4-channel EEG monitoring: a guide to interpretation, with normal values, in preterm infants. Neuropediatrics, 18:138, 1987 38. Benda G, Engel R, Zhang Y: Prolonged inactive periods during the discontinuous pattern of prematurity in the electroencephalogram of very-low-birth-weight infants. Electroencephalogr Clin Neurophysiol, 72:189, 1989

Neonatal and Pediatric Electroencephalography

39. Hughes], Fino], Gagnon L: Periods of activity and quiescence in the premature EEG. Neuropediatrics, 14:66, 1983 40. Holmes G, Rowe], Hafford] et al: Prognostic value of the electroencephalogram in neonatal asphyxia. Electroencephalogr Clin Neurophysiol, 53:60, 1982 41. Watanabe K, Iwase K, Hara K: Development of slow-wave sleep in low birthweight infants. Dev Med Child Neurol, 16:23, 1974 42. Aso K, Scher MS, Barmada MA: Neonatal electroencephalography and neuropathology.] Clin Neurophysiol, 6:103,1989 4~~. Scher M, Tharp B, Sylvestri L: Significance of focal abnormalities in neonatal electroencephalograms: radiological correlation and outcome. Ann Neurol, 12:217, 1982 44. Allemand F, Monod N, Larroche ]C: L'electroencephalogramme dans les hernorragies sous-durales du nouveau-ne. Rev Electroencephalogr Neurophysiol Clin, 7:365,1977 45. Nunes ML, Da Costa]C, Moura-Ribeiro MY: Polysomnographic quantification of bioelectrical maturation in preterm and full-term newborns at matched conceptional ages. Electroencephalogr Clin Neurophysiol, 102:186,1997 46. Hahn ]S, Tharp BR: The dysmature EEG in infants with bronchopulmonary dysplasia and its prognostic implications. Electroencephalogr Clin Neurophysiol, 76: I 06, 1990 47. Hayakawa F, Okumura A, Kato T et al: Dysmature EEG pattern in EEGs of preterm infants with cognitive impairment: maturation arrest caused by prolonged mild CNS depression. Brain Dev, 19:122, 1997 48. Biagioni E, Bartalena L, Biver P et al: Electroencephalographic dysmaturity in preterm infants: a prognostic tool in the early postnatal period. Neuropediatrics, 27:311,1996 4!:l. Watanabe K, Hayakawa F, Okumura A: Neonatal EEG: a powerful tool in the assessment of brain damage in pretenn infants. Brain Dev, 21:361,1999 50. Hayakawa F, Okumura A, Kato T et al: Determination of timing of brain injury in preterm infants with periventricular leukomalacia with serial neonatal electroencephalography. Pediatrics, 104:1077, 1999 51. Kubota T, Okumura A, Hayakawa F et al: Combination of neonatal electroencephalography and ultrasonography: sensitive means of early diagnosis of periventricular leukomalacia. Brain Dev, 24:698, 2002 52. Maruyama K, Okumura A, Hayakawa F et al: Prognostic value of EEG depression in preterm infants for later development of cerebral palsy. Neuropediatrics, 33:133, 2002 53. Pressler RM, Boylan GB, Morton M et al: Early serial EEG in hypoxic ischaemic encephalopathy. Clin Neurophysiol, 112:31, .2001 54. Zeinstra E, Fock ]M, Begeer ]H et al: The prognostic value of serial EEG recordings following acute neonatal asphyxia in full-term infants. Eur ] Paediatr Neurol, 5:155, 2001 55. Scher MS, Bova ]M, Dokianakis SG et al: Physiological significance of sharp wave transients on EEG recordings

125

of healthy pre-term and full-term neonates. Electroencephalogr Clin Neurophysiol, 90: 179, 1994 56. Clancy RR, Spitzer AR: Cerebral cortical function in infants at risk for sudden infant death syndrome. Ann Neurol, 18:41, 1985 57. Clancy R, Tharp B: Positive rolandic sharp waves in the electroencephalograms of premature neonates with intraventricular hemorrhage. Electroencephalogr Clin Neurophysiol,57:395, 1984 58. Novotny E], Tharp BR, Enzmann DR et al: Positive rolandic sharp waves in the EEG of the premature infant. Neurology, 37:1481,1987 59. Marret S, Parain D, Menard] et al: Prognostic value of neonatal electroencephalography premature newborns less than 33 weeks of gestational age. Electroencephalogr Clin Neurophysiol, 102:178, 1997 60. Baud 0, d'Allest AM, Lacaze-Masmonteil T et al: The early diagnosis of periventricular leukomalacia in premature infants with positive rolandic sharp waves on serial electroencephalography. J Pediatr, 132:813, 1998 61. Vecchierini-Blineau MF, Nogues B, Louvet S et al: Positive temporal sharp waves in electroencephalograms of the premature newborn. Neurophysiol Clin, 26:350, 1996 62. Scher MS, BovaJM, Dokianakis SG et al: Positive temporal sharp waves on EEG recordings of healthy neonates: a benign pattern of dysmaturity in pre-term infants at post-conceptional term ages. Electroencephalogr Clin Neurophysiol, 90:173, 1994 63. Chung H], Clancy RR: Significance of positive temporal sharp waves in the neonatal electroencephalogram. Electroencephalogr Clin Neurophysiol, 79:256, 1991 64. Nunes ML, Penela MM, da Costa ]C: Differences in the dynamics of frontal sharp transients in normal and hypoglycemic newborns. Clin Neurophysiol, 111:305, 2000 65. Scher MS, Beggarly M: Clinical significance of focal periodic discharges in neonates.] Child Neurol, 4:175, 1989 66. Mizrahi E, Tharp B: A characteristic EEG pattern in neonatal herpes-simplex encephalitis. Neurology, 32:1215,1982 67. Hrachovy RA, O'Donnell DM: The significance of excessive rhythmic alpha and/or theta frequency activity in the EEG of the neonate. Clin Neurophysiol, 110:438, 1999 68. Zaret BS, Guterman B, Weig S: Circumscribed midline EEG activity in neurologically normal neonates. Clin Electroencephalogr, 22:13, 1991 69. Ortibus EL, Sum ]M, Hahn ]S: Predictive value of EEG for outcome and epilepsy following neonatal seizures. Electroencephalogr Clin Neurophysiol, 98:175,1996 70. Biagioni E, Boldrini A, Bottone U et al: Prognostic value of abnormal EEG transients in preterm and full-term neonates. Electroencephalogr Clin Neurophysiol, 99: I, 1996 71. Scher MS, Painter M], Bergman I et al: EEG diagnosis of neonatal seizures: clinical correlations and outcome. Pediatr Neurol, 5:17,1989 72. Mizrahi EM, Kellaway P: Characterization and classification of neonatal seizures. Neurology, 37: 1837, 1987

126

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

73. Connell], Oozeer R, de Vries L et al: Continuous EEG monitoring of neonatal seizures: diagnostic and prognostic considerations. Arch Dis Child, 64:452, 1989 74. Tharp BR, Laboyrie P: The incidence of EEG abnormalities and outcome of infants paralyzed with neuromuscular blocking agents. Crit Care Med, 11:926, 1983 75. Shewmon DA: What is a neonatal seizure? Problems in definition and quantification for investigative and clinical purposes.] Clin Neurophysiol, 7:315,1990 76. Clancy RR, Legido A: The exact ictal and interictal duration of electroencephalographic neonatal seizures. Epilepsia, 28:537, 1987 77. Bye AME, Cunningham CA, Chee KYet al: Outcome of neonates with e1ectrographically identified seizures, or at risk of seizures. Pediatr Neurol, 16:225, 1997 78. Watanabe K, Hara K, Myazaki et al: Apneic seizures in the newborn. Am] Dis Child, 136:980, 1982 79. Hirsch E, Velez A, SelIaI F et al: Electroclinical signs of benign neonatal familial convulsions. Ann Neurol, 34:835, 1993 BO. Ronnen GM, Roslaes TO, Connolly M et al: Seizure characteristics in chromosome 20 benign familial neonatal convulsions. Neurology, 43:1993 81. Aso K, Watanabe K: Benign familial neonatal convulsions: generalized epilepsy? Pediatr Neurol, 8:226, 1992 82. Bye AME: Neonate with benign familial neonatal convulsions: recorded generalized and focal seizures. Pediatr Neurol, 10:164, 1994 83. Legido A, Clancy RR, Berman PH: Neurologic outcome after electroencephalographically proven neonatal seizures. Pediatrics, 88:583, 1991 84. Clancy RR: Interictal sharp EEG transients in neonatal seizures.] Child Neurol, 4:30, 1989 85. Oliveira A], Nunes ML, Haertel LM et al: Duration of rhythmic EEG patterns in neonates: new evidence for clinical and prognostic significance of brief rhythmic discharges. Clin Neurophysiol, 111:1646, 2000 86. Hahn ]S, Delgado MR, Clegg N] et al: Electroencephalography in holoprosencephaly: findings in children without epilepsy. Clin Neurophysiol, 114:1908, 2003 87. Gastaut H, Pinsard N, Raybaud CH et al: Lissencephaly (agyria-pachygyria): clinical findings and serial EEG studies. Dev Med Child Neurol, 29:167,1987 88. Tharp BR: Unique EEG pattern (comb-like rhythm) in neonatal maple syrup urine disease. Pediatr Neurol, 8:65, 1992 89. Sainio K, Ganstrom ML, Pettay 0 et al: EEG in neonatal herpes simplex encephalitis. Electroencephalogr Clin Neurophysiol, 56:556, 1983 90. Toet MC, Hellstrom-Westas L, Groenendaal F et al: Amplitude integrated EEG 3 and 6 hours after birth in full-term neonates with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed, 81:FI9, 1999 91. Shalak LF, Laptook AR, Velaphi SC et al: Amplitude integrated electroencephalography coupled with an early neurologic examination enhances prediction of term infants at risk for persistent encephalopathy. Pediatrics, 111:351,2003 92. Toet MC, van der Meij W, de Vries LS et al: Comparison between simultaneously recorded amplitude integrated

electroencephalogram (cerebral function monitor) and standard electroencephalogram in neonates. Pediatrics. 109:772, 2002 93. Inder TE, Buckland L, Williams CE et al: Lowered electroencephalographic spectral edge frequency predicts the presence of cerebral white matter injury in premature infants. Pediatrics, 111:27,2003 94. Kellaway P: An orderly approach to visual analysis: elements of the normal EEG and their characteristics in children and adults. p. 100. In Ebersole]S, PedleyTA (eds): Current Practice of Clinical Electroencephalography. 3rd ed. Lippincott Williams & Wilkins, Philadelphia, 2003 95. Eeg-Olofsson 0, Petersen I, Sellden U: The development of the electroencephalogram in normal children from the age of 1 through 15 years: paroxysmal activity. Neuropadiatrie, 2:375, 1971 96. Alvarez N, Lombroso C, Medina C et al: Paroxysmal spike and wave activity in drowsiness in young children: its relationship to febrile convulsions. Electroencephalogr Clin Neurophysiol, 56:406,1983 97. Louis], Zhang]X, Revol M et al: Ontogenesis ofnocturnal organization of sleep spindles: a longitudinal study during the first 6 months of life. Electroencephalogr Clin Neurophysiol, 83:289, 1992 98. Schultz M, Schulte F, Akiyama Yet al: Development of electroencephalographic sleep phenomena in hypothyroid infants. Electroencephalogr Clin Neurophysiol, 25:351, 1968 99. Metcalf D, Mondale ], Butler F: Ontogenesis of spolltaneous K-complexes. Psychophysiology, 8:340, 1971 100. Slater G, Torres F: Frequency-amplitude gradient: new parameter for interpreting pediatric sleep EEGs. Arch Neurol, 36:465, 1979 101. Petersen I, Eeg-Olofsson 0: The development of the electroencephalogram in normal children from the age of 1 through 15 years: non-paroxysmal activity. Neuropadiatrie, 2:247,1971 102. Petersen I, Sellden V, Eeg-Olofsson 0: The evolution of the EEG in normal children and adolescents from ] to 21 years. p. 31. In Lairy G (ed): Handbook of Electroencephalography and Clinical Neurophysiology. Elsevier, Amsterdam, 1975 103. Doose H, Gerken H, Volzke E: On the genetics of EEGanomalies in childhood. I. Abnormal theta rhythms. Neuropadiatrie, 3:386, 1972 104. Gerken H, Doose H: On the genetics ofEEG-anomalies in childhood. II. Occipital 2-4/s rhythms, Neuropadiatrie, 3:437,1972 105. Belsh ], Chokroverty S, Barabas G: Posterior rhythmic slow activity in EEG after eye closure. Electroencephalogr Clin Neurophysiol, 56:562,1983 106. Silbert PL, Radhakrishnan K,]ohnson] et al: The significance of the phi rhythm. Electroencephalogr Clin Neurophysiol, 95:71, 1995 107. North K, Ouvrier R, Nugent M: Pseudoseizures caused by hyperventilation resembling absence epilepsy.] Child Neurol, 5:288, 1990 108. Doose H, Waltz S: Photosensitivity: genetics and clinical significance. Neuropediatrics, 24:249, 1993 109. Harding GF, Fylan F: Two visual mechanisms of photosensitivity. Epilepsia, 40:1446, 1999

Neonatal and Pediatric Electroencephalography

110. Anderson C, Torres F: Photic driving in the early premature infant. Electroencephalogr Clin Neurophysiol, 58:302, 1984 11 I. Colon E, Vingerhoets H, Noterrnans S et al: orr response: its clinical incidence in very young children. Electroencephalogr Clin Neurophysiol, 46:601, 1979 112. Steinlein 0, Anokhin A, Yping Metal: Localization of a gene for the human low-voltage EEG on 20q and genetic heterogeneity. Genomics, 12:69, 1992 II:t Doose H, Castiglione E, Waltz S: Parental generalized EEG alpha activity predisposes to spike wave discharges in offspring, Hum Genet, 96:695, 1995 114. Klass OW, Westmoreland BF: Nonepileptogenic epileptiform electroencephalographic activity. Ann Neurol, 18:627,1985 II f,. Yamada T, Young S, Kimura J: Significance of positive spike bursts in Reye syndrome. Arch Neurol, 34:376,1977 1I (i. Silverman 0: Fourteen- and six-per-second positive spike pattern in a patient with hepatic coma. Electroencephalogr Clin Neurophysiol, 16:395, 1964 117. Drury I: 14 and 6 Hz positive bursts in childhood encephalopathies. Electroencephalogr Clin Neurophysiol, 72:479,1989 118. White j, Langston W, Pedley T: Benign epileptiform transients of sleep. Neurology, 27: 1061, 1977 II ~l. Cotman ], Koffler OJ: Interictal spiking increases after seizures but does not after decrease in medication. Electroencephalogr Clin Neurophysiol, 72:7, 1988 120. Marks DA, Katz A, Scheyer R et al: Clinical and electrographic effects of acute anticonvulsant withdrawal in epileptic patients. Neurology, 41:508, 1991 121. Camfield P, Gordon K, Camfield C et al: EEG results are rarely the same if repeated within six months in childhood epilepsy. Can ] Neurol Sci, 22:297, 1995 122. Lombroso C: Consistent EEG focalities detected in subjects with primary generalized epilepsies monitored for two decades. Epilepsia, 38:797, 1997 12::1. Shinnar S, Berg AT, Moshe SL et al: Discontinuing antiepileptic drugs in children with epilepsy: a prospective study. Ann Neurol, :~.1):534, 1994 124. Commission on Classification and Terminology of the International League Against Epilepsy: Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia, 30:389, 1989 12;;. Kellaway P, FrostJ, Crawley J: Time modulation of spikeand-wave activity in generalized epilepsy. Ann Neurol, 8:491. 1980 12(;. Sato S, Dreifuss F, Penry K et al: Long-term follow-up of absence seizures. Neurology, 33: 1590, 1983 127. Bouma PA, Westendorp RG, Van Dijk.lG et al: The outcome of absence epilepsy: a meta-analysis. Neurology, 47:802,1996 128. Cavazzuti GB, Ferrari F, Galli V et al: Epilepsy with typical absence seizures with onset during the first year of life. Epilepsia, 30:802, 1989 129. Panayiotopoulos CP, Obeid T, Waheed G: Differentiation of typical absence seizures in epileptic syndromes. Brain, 112:1039,1989 130. janz D: The idiopathic generalized epilepsies of adolescence with childhood and juvenile age of onset. Epilepsia, :~8:4, 1997

127

131. Panayiotopoulos CP, Obeid T, Tahan AR: Juvenile myoclonic epilepsy: a 5-year prospective study. Epilepsia, 35:285, 1994 132. Panayiotopoulos CP: Effectiveness of photic stimulation on various eye-states in photosensitive epilepsy. J Neurol Sci, 24:95, 1974 133. Aliberti V, Grunewald RA, Panayiotopoulos CP et al: Focal electroencephalographic abnormalities in juvenile myoclonic epilepsy. Epilepsia, 35:297, 1994 134. Lancman ME, Asconape .U, Penry.lK: Clinical and EEG asymmetries in juvenile myoclonic epilepsy. Epilepsia, 35:302, 1994 135. Oguni H, Mukahira K, Oguni M et al: Video-polygraphic analysis of myoclonic seizures in juvenile myoclonic epilepsy. Epilepsia, 35:307,1994 136. Berkovic SF, Andermann F, Andermann E et al: Concepts of absence epilepsies: discrete syndromes or biological continuum? Neurology, 37:993, 1987 137. Yamamoto N, Watanabe K, Negoro T er al: Partial seizures evolving to infantile spasms. Epilepsia, 29:34, 1988 138. Hrachovy R, FrostJ, Kellaway P: Hypsarrhythrnia: variations on the theme. Epilepsia, 25:317, 1984 139. Vinters HV, DeRosa MJ, Farrell MA: Neuropathological study of resected tissue from patients with infantile spasms. Epilepsia, 34:772, 1993 140. Gaily EK, Shewmon DA, Chugani HT et al: Asymmetric and asynchronous infantile spasms. Epilepsia, 36:873, 1995 141. Miyazaki M, Hashimoto T, FIUii E et al: Infan tile spasms: localized cerebral lesions on SPECT. Epilepsia, 35:988, 1994 142. Chugani HT, Conti JR: Etiologic classification of infantile spasms in 140 cases: role of positron emission tomography. J Child Neurol, 11:44, 1996 143. Fusco L, Vigevano F: Ictal clinical electroencephalographic findings of spasms in West syndrome. Epilepsia, 34:671, 1993 144. Juhasz C, Chugani HT, Muzik 0 et al: Neuroradiological assessment of brain structure and function and its implication in the pathogenesis of West syndrome. Brain Dev, 23:488,2001 145. Dulac 0, Plouin P,.Iambaque 1: Predicting favorable outcome in idiopathic West syndrome. Epilepsia, 34:747, 1993 146. Ikeno T, Shigematsu H, Miyakoshi M et al: An analytic study of epileptic falls. Epilepsia, 26:612, 1995 147. Tassinari C, Bureau M, Dravet C et al: Epilepsy with continuous spikes and waves during sleep. p. 194. In Roger .I, Dravet C, Mureau M et al (eds): Epileptic Syndromes in Infancy, Childhood and Adolescence. Libbey Eurotext, London, 1985 148. Beaumanoir A: The Landau-Kleffner syndrome. p. 181. In Roger .I, Dravet C, Mureau M et al (eds): Epileptic Syndromes in Infancy, Childhood and Adolescence. Libbey Eurotext, London, 1985 149. Aicardi J, Chevrie JJ: Atypical benign partial epilepsy of childhood. Dev Med Child Neurol, 24:281,1982 150. Blume W: Clinical and electroencephalographic correlates of the multiple independent spike foci pattern in children. Arch Neurol, 4:541, 1978

128

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

151. Lombroso CT: Early myoclonic encephalopathy, early infantile epileptic encephalopathy, and benign and severe infantile myoclonic epilepsies: a critical review and personal contributions.J Clin Neurophysiol, 7:380, 1990 152. Yamatogi Y, Ohtahara S: Early-infantile epileptic encephalopathy with suppression-bursts, Ohtahara syndrome; its overview referring to our 16 cases. Brain Dev, 24:13,2002 153. AcharyaJN, Wyllie E, Luders HO et al: Seizure symptomatology in infants with localization-related epilepsy. Neurology, 48:189, 1997 154. Nordli DR, Bazil CW, Scheuer ML et al: Recognition and classification of seizures in infants. Epilepsia, 38:553, 1997 155. Legarda S, Jaykar P: Electroclinical significance of rolandic spikes and dipoles in neurodevelopmentally normal children. Electroencephalogr Clin Neurophysiol, 95:257, 1995 156. Gregory DL, Wong PK: Clinical relevance of a dipole field in rolandic spikes. Epilepsia, 33:36, 1992 157. Van der Meij W, Van Huffelen AC, Wienecke GH et al: Sequential EEG mapping may differentiate "epileptic" from "non-epileptic" rolandic spikes. Electroencephalogr Clin Neurophysiol, 82:408, 1992 158. Van der Meij W, Wieneke GH, Van Huffelen AC et al: Identical morphology of the rolandic spike-and-wave complex in different clinical entities. Epilepsia, 34:540, 1993 159. FrostJD, Hrachovy RA, Glaze DG: Spike morphology in childhood focal epilepsy: relationship to syndromic classification. Epilepsia, 33:531, 1992 160. Minami T, Gondo K, Yamamoto T et al: Magnetoencephalographic analysis of rolandic discharges in benign childhood epilepsy. Ann Neurol, 39:326, 1996 161. Aicardi J: Benign rolandic epilepsy. Int Pediatr, 2:176, 1987 162. Beydoun A, Garofalo EA, Drury I: Generalized spikewaves, multiple loci, and clinical course in children with EEG features of benign epilepsy of childhood with centrotemporal spikes. Epilepsia, 33:1091, 1992 163. Okubo Y, Matsuura M, Asai T et al: Epileptiform EEG discharges in healthy children: prevalence, emotional and behavioral correlates, and genetic influences. Epilepsia, 35:832, 1994 164. Doose H, Neubauer B, Carlsson G: Children with benign focal sharp waves in the EEG:developmental disorders and epilepsy. Neuropediatrics, 27:227, 1996 165. Massa R, de Saint-Martin A, Carcangiu R et al: EEG criteria predictive of complicated evolution in idiopathic rolandic epilepsy. Neurology, 57:1071, 2001 166. Ferrie CD, Beaumanoir A, Guerrini R et al: Early-onset benign occipital seizure susceptibility syndrome. Epilepsia, 38:285, 1997 167. Vigevano F, Lispi ML, Ricci S: Early onset benign occipital susceptibility syndrome: video-EEG documentation of an illustrative case. Clin Neurophysiol, Ill: Suppl 2, S81,2000 168. Panayiotopoulos CP: Benign childhood epilepsies with occipital paroxysms: a 15-year prospective study. Ann Neurol, 26:51, 1989

169. Cooper GW, Lee SI: Reactive occipital epileptiform activity: is it benign? Epilepsia, 32:63, 1991 170. Biagioni E, Cioni G, Cowan F et al: Visual function and EEG reactivity in infants with perinatal brain lesions at 1 year. Dev Med Child Neurol, 44:171, 2002 171. Maytal ], Steele R, Eviatar L et al: The value of early postictal EEG in children with complex febrile seizures. Epilepsia, 41:219, 2000 172. Lennox-Buchthal M: Febrile convulsions: a reappraisal. Electroencephalogr Clin Neurophysiol Suppl 32: I, 1973 173. Schiottz-Christensen E, Hammerberg P: EEG in twins with febrile convulsions: follow-up on 59 pairs of the same sex. Neuropadiatrie, 6:142,1975 174. Nelson K, EllenbergJ: Predictors of epilepsy in children who have experienced febrile seizures. N Engl J Med, 295:1029, 1976 175. Tassinari CA, Bureau M, Dravet C et al: Epilepsy with continuous spikes and waves during slow sleep-otherwise described as ESES (epilepsy with electrical status epilepticus during slow sleep). p. 245. In Roger J, Bureau M, Dravet C et al (eds): Epileptic Syndromes in Infancy, Childhood and Adolescence. 2nd Ed. John Libbey, London, 1992 176. Perez ER, Davidoff V, Despland P et al: Mental and behavioral deterioration of children with epilepsy and CSWS: acquired epileptic frontal syndrome. Dev Med Child Neurol, 35:661, 1993 177. Morrell F, Whisler WW,Smith MC et al: Landau-Kleffner syndrome: treatment with subpial intracortical transection. Brain, 118:1529, 1995 178. Greenberg DA, Weinkle DJ, Aminoff MJ: Periodic EEG complexes in infectious mononucleosis. J Neurol Neurosurg Psychiatry, 45:648, 1982 179. Westmoreland BF, Gomez MR, Blume WT: Activation of periodic complexes of subacute sclerosing panencephalitis by sleep. Ann Neurol, 1:185, 1977 180. Lombroso C: Remarks on the EEG and movement disorder in SSPE. Neurology, 18:60, 1968 181. Townsend.IJ, Baringer JR, WolinskyJS et al: Progressive rubella panencephalitis: late onset after congenital rubella. N EnglJ Med, 292:990, 1975 182. Tharp B: Orbital frontal seizures: an unique electroencephalographic and clinical syndrome. Epilepsia, 13:627, 1972 183. Fariello R, BlackJ: Pseudoperiodic bilateral EEG paroxysms in a case of phencyclidine intoxication. J CUn Psychiatry, 39:579, 1978 184. Hagne I, Witt-Engerstrom I, Hagberg B: EEG development in Rett's syndrome: a study 01'30 cases. Electroencephalogr Clin Neurophysiol, 72:1, 1989 185. Buoni S, Grosso S, Pucci Let al: Diagnosis of Angelman syndrome: clinical and EEG criteria. Brain Dev, 21:296, 1999 186. Guerrini R, DeLorey TM, Bonanni P et al: Cortical myoclonus in Angelman syndrome. Ann Neurol, 40:39, 1996 187. Laan LA, Brouwer OF, Begeer CH et al: The diagnostic value of the EEG in Angelman and Rett syndrome at a young age. Electroencephalogr Clin Neurophysiol, 106:404, 1998

Neonatal and Pediatric Electroencephalography

18H. Robb SA, Harden A, Boyd SG: Rett syndrome: an EEG study in 52 girls. Neuropediatrics, 20:192, 1989 189. Robertson R, Langill L, Wong PKH et al: Rett syndrome: EEG presentation. Electroencephalogr Clin Neurophysiol, 70:388, 1988 190. Gloor P, Kalabay 0, Giard N: The electroencephalogram in diffuse encephalopathies: electroencephalographic correlates of grey and white matter diseases. Brain, 91:779, 1968 191. Vanhanen SL, Sainio K, Lappi M et al: EEG and evoked potentials in infantile neuronal ceroid-lipofuscinosis. Dev Med Child Neurol, 39:456, 1997 192. Veneselli E, Biancheri R, Buoni S et al: Clinical and EEG findings in 18 cases of late infantile neuronal ceroid lipofuscinosis. Brain Dev, 23:306, 2001 193. Steinman L, Tharp BR, Dorfman L] et al: Peripheral neuropathy in the cherrry-red spot-myoclonus syndrome (sialidosis type I). Ann Neurol, 7:540, 1980 194. Engel], Rapin I, Giblin D: Electrophysiological studies in two patients with cherry red spot-myoclonus syndrome. Epilepsia, 18:73, 1977 195. Tobimatsu S, Fukui R, Shibasaki H et al: Electrophysiological studies of myoclonus in sialidosis type 2. Electroencephalogr Clin Neurophysiol, 60:16, 1985

129

196. So N, Berkovic S, Andermann F et al: Myoclonus epilepsy and ragged-red fibers (MERRF). 2. Electrophysiological studies and comparisons with other progressive myoclonus epilepsies. Brain, II2:1261, 1989 197. Harding BN: Progressive neuronal degeneration of childhood with liver disease (Alpers-Huttenlocker syndrome): a personal review.] Child Neurol, 5:273,1990 198. Prick M], Gabreels F], Treibels ]M et al: Progressive poliodystrophy (Alpers' disease) with a defect in cytochrome aa3 in muscle: a report of two unrelated patients. Clin Neurol Neurosurg, 85:57,1983 199. Manuelidis E, Rorke L: Transmission of Alpers' disease (chronic progressive encephalopathy) produces experimental Creutzfeldt-]akob disease in hamsters. Neurology, 39:615,1989 200. Boyd SG, Harden A, Egger] et al: Progressive neuronal degeneration of childhood with liver disease ("Alpers' disease"): characteristic neurophysiological features. Neuropediatrics, 17:75, 1986 201. Ferriss GS, Happel LT, Duncan MC: Cerebral cortical isolation in infantile neuroaxonal dystrophy. Electroencephalogr Clin Neurophysiol, 43:168,1977 202. Dorfman L], Pedley TA, Tharp BR et al: Juvenile neuroaxonal dystrophy: clinical, electrophysiological and neuropathological features. Ann Neurol, 3:419, 1978

CHAPTER

5 Long-Term Monitoring for Epilepsy JEROME ENGEL, Jr., and JASON R. SOSS

INDICATIONS Differential Diagnosis Characterization and Classification Determination of Frequency and Temporal Pattern of Seizures Localization of the Epileptogenic Region TECHNICAL CONSIDERATIONS Equipment for EEG Recording Electrodes Amplifiers Transmission Recording, Storage, Retrieval, and Review Equipment for Monitoring Clinical Behavior Methods of Procedure Electrodes Recording System Recording Techniques LTME System Configurations

Long-term monitoring for epilepsy (LTME), also referred to as long-term neurodiagnostic monitoring and as long-term video electroencephalographic monitoring, includes a variety of techniques used to simultaneously record electrical and clinical characteristics of paroxysmal disturbances in cerebral function over extended periods. 1.2 This approach is useful for describing events that are difficult to record during routine electroencephalography (EEG) and is employed for four purposes: (I) to distinguish between epileptic seizures and other intermittent behaviors; (2) to characterize the electroclinical features of habitual ictal event" in order to diagnose seizure type and, when possible, a specific epileptic syndrome; (3) to determine the frequency and temporal pattern of ictal events to identify precipitating factors and assess the effectiveness of therapeutic interventions; and (4) to localize the site of seizure origin in patients with medically refractory seizures who are candidates for surgical treatment."? CIME is designed for the diagnosis of intermittent abnormalities that occur infrequently and unpredictably. Consequently, the length of time required for capturing a number of event" sufficient to answer the necessary clinical questions cannot be readily predicted. Sometimes

INTERPRETATION OF RESULTS Artifacts Epileptic Activity Nonepileptic Abnormal Activity RECOMMENDATIONS Presurgical Evaluation Diagnosis of Nonepileptic Seizures Classification and Characterization of Epileptic Events Quantification of Electrographic Abnormalities APPENDIX 5-1 Behavioral Signs and Symptoms Associated with Electrographic Ictal Discharges Recorded from Specific Cerebral Areas

weeks of recording are involved, but at other times only several hours are required. Thus the term long-term monitoringrefers more to the capability for recording over long periods than to the actual duration of each recording. 1.2 LTME became readily available as a clinical diagnostic tool in the 1960s as a result of two independent technologic developments. First, the use of stereotactically implanted chronic depth electrodes provided a means of easily recording ictal EEG discharges during complex partial and secondarily generalized epileptic seizures without contamination by muscle artifact.v" Second, EEG telemetry was devised by scientists working with the National Aeronautics and Space Administration (NASA) in order to record electrophysiologic changes occurring in animals (and later, humans) put into earth's orbit. Combining telemetry technology with depth electrodes chronically implanted into brains of epileptic patients allowed artifact-free ictal EEGs to be recorded during spontaneous seizures occurring unpredictably over prolonged periods.!" Since then, many advances in LTME have occurred, and a variety of approaches are currently available. No specific approach is generally accepted for standard use; rather, each of the various technologies and methods 131

132

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

TABLE 5·1 • TermlnolOlY for the Common Subcategories ofLTME Long-term EEC monitoring: scalp/sphenoidal electrodes, direct cable, no

video Long-term intracranial EEC monitoring: depth, subdural, epidural. orfora-

men ovale electrodes; direct cable, no video Long-term EEC recording with video monitoring: scalp/sphenoidal elec-

trodes; direct cable, video Long-term intracranial EEC recording with video monitoring: depth, sub-

dural. epidural. orforamen ovale electrodes; direct cable, video Long-term EEC telemetry: scalp/sphenoidal electrodes, cable orradio

telemetry, no video Long-term intracranial EEC telemetry: depth, subdural, epidural. orforamen

ovale electrodes; cable orradio telemetry, no video Long-term EEC telemetry with video monitoring: scalp/sphenoidal elec-

trodes, cable orradio telemetry, video Long-term intracranial EEC telemetry with video monitoring: depth, sub-

dural. epidural. orforamen ovale electrodes; cable orradio telemetry, video EEC ambulatory recording: scalp/sphenoidal electrodes, ambulatory recording

has relative strengths and weaknesses that allow specific equipment and configurations to be chosen, depending on the needs and limitations of individual clinical facilities. The American Clinical Neurophysiology Society and the International Federation of Clinical Neurophysiology have issued "Guidelines for Long-Term Monitoring for Epilepsy.'"> which reviews available technologies and methods and makes recommendations concerning indications for their use. The technology for LTME has advanced considerably since these guidelines were published, however, and technologic updates are now needed. Nevertheless, the suggestions for a standard terminology for the most common subcategories of LTME, shown in Table 5-1, are still appropriate. This chapter is concerned with the indications for LTME, the available monitoring equipment and methods of procedure, the interpretation of data, and recommendations for specific diagnostic purposes. The equipment and procedures discussed here are appropriate for LTME of adults and children; more details on adaptations necessary for application to infants and neonates can be found elsewhere. I 1,12

INDICATIONS LTME can be expensive and labor intensive, but it is cost effective in many circumstances. Its use should be limited to diagnostic problems that cannot easily be resolved in the routine EEG laboratory.

Differential Diagnosis LTME is often used as a last resort in the differential diagnosis between epilepsy and other disorders associated with intermittent or paroxysmal disturbances that may resemble epileptic seizures. 13 A definitive diagnosis is easily made when habitual events are shown to consist of clinical behaviors characteristic of epilepsy and are associated with well-defined ictal EEG discharges; or when other etiologies can be clearly demonstrated (e.g., cardiac arrhythmias or sleep disturbances). Often, however, the events in question occur without obvious EEG or other electrophysiologic changes. In this case, a diagnosis is usually reached with reasonable confidence based on features of the ictal behavior in association with other clinical and laboratory information. Simple partial seizures usually have no EEG correlates that can be recorded with extracranial electrodes.!" Consequently, seizures without impaired consciousness are most often diagnosed by characteristic behavioral features and, at times, elevated serum prolactin levels.P rather than by the occurrence of ictal EEG discharges. Myoclonic jerks may also have no EEG correlates but usually can be diagnosed on the basis of characteristic motor symptorns.l" Many other nonepileptic disorders can be recognized readily by clinical examination during the habitual event, by review of videotapes, or both. In most of these situations, however, the results of LTME merely confirm the clinical impression derived from historical and other information and are not diagnostic in themselves. The most difficult, and most important, differential diagnosis for which LTME is used is the distinction between epileptic seizures and nonepileptic psychogenic seizures. Because, by definition, nonepileptic psychogenic seizures have no EEG correlates, this diagnosis is usually one of exclusion. Certain features of nonepileptic psychogenic seizures may distinguish them from generalized convulsions; these include uncoordinated nonsynchronous thrashing of the limbs, quivering, pelvic thrusting, side-to-side head movements, opisthotonic posturing, screaming and talking throughout the ictal episode, prolongation for many minutes or even hours, abrupt termination without postictal confusion, evidence of some recall during the ictal event, and features that are not stereotyped but differ from one episode to another. 13 ,17,18 Any of these symptoms, however, can be epileptic phenomena, and differential diagnosis between nonepileptic psychogenic seizures and complex partial seizures is particularly dimcult.'? Furthermore, nonepileptic psychogenic seizures may be associated with autonomic changes (e.g., pupillary dilatation, depressed corneal reflexes, Babinski responses, cardiorespiratory changes, and urinary and fecal incontinence) as well as self-injury induced by falling or biting the lips and tongue. 13 ,18 Consequently,

long-Term Monitoring forEpilepsy

a definitive diagnosis cannot be made solely on the basis of ictal clinical features observed during LTME; other positive evidence (e.g., secondary gain) is necessary. A conclusion that an ictal event captured by LTME is nonepileptic and psychogenic does not, in itself, rule out the existence of an epileptic condition, because some patients with epilepsy also have nonepileptic psychogenic seizures. 13 . IS When the two phenomena coexist, it should be possible to obtain a history of more than one seizure type and to use LTME to record examples of each type to determine which are nonepileptic and which are epileptic. Based on this information, the patient and family can be instructed to record these events separately to determine the effects of antiepileptic and psychiatric interventions on each independently. Nonepileptic psychogenic seizures, which are involuntary and as disabling as epileptic events, need to be distinguished from malingering, which is voluntary simulation, and from factitious disorder, which is selfinduction of epileptic attacks for the purpose of gaining patient status." This differential diagnosis can sometimes be accomplished by data obtained during LTME, but usually it depends on historical information.

Charaderization and Classification Patients with known epileptic seizures that do not respond to antiepileptic medication may be undergoing treatment for the wrong type of epilepsy. In this situation, LTME can provide crucial information that characterizes the epileptic events so that the physician can make a seizure diagnosis" and, when possible, a diagnosis ofa specific epileptic syndrome." Seizure type usually determines the most appropriate antiepileptic drugs and whether a patient might be a candidate for surgical intervention. A specific epileptic syndrome is often associated with a known prognosis, which is helpful information for the patient and physician.!' LTME is particularly useful for distinguishing partial seizures from generalized epileptic events on the basis of characteristic ictal EEG features. It is important to distinguish the different epileptic causes of brief loss of consciousness, which can be a typical absence seizure, an atypical absence seizure, or a complex partial seizure characterized by impaired consciousness only. It may be impossible to distinguish between atypical absences resulting from bilateral or diffuse brain damage and atypical complex partial seizures originating primarily from frontal lobe lesions with secondary bilateral synchrony.22 LTME is also useful for distinguishing simple from complex partial seizures, particularly when trained personnel are available to examine the patient during the ictal event. The hallmark of a complex partial seizure is

133

amnesia for the ictal period. Although this information usually has no direct therapeutic relevance, because simple and complex partial seizures are treated with the same medications, a definitive diagnosis can at times have medicolegal implications (e.g., a patient with only simple partial seizures might be allowed to drive), and it is important for counseling patients regarding activities of daily living. Documentation of the degree of disability during and after the seizure can be important in the decision about whether surgical treatment is warranted. More detailed tests (e.g., reaction time tasks) during ictal events can be used to identify subtle disturbances of function for the same purposes.

Determination of Frequency and Temporal Pattern of Seizures When patients are known to have epileptic seizures of a specific type, but it is unclear how often these seizures are occurring, LTME can be used to determine the frequency of ictal events. Patients with seizures that involve relatively brieflapses of consciousness may not be aware of each ictal event; therefore they depend on observers to know whether therapeutic interventions have resulted in benefit. In these situations, LTME is a more accurate way of documenting seizure frequency before and after treatment. Unexplained deterioration in mental function can be caused by unrecognized brief daytime seizures or more severe nocturnal events. Knowledge about the occurrence and frequency of such ictal events is important for medicolegal reasons, for counseling patients regarding the activities of daily living, and for deciding whether surgical intervention is warranted. LTME can reveal when seizures are most likely to occur. At times, this information is useful for identifying specific precipitating factors that might be avoided.F Combining LTME with serum drug level assessments can help to determine whether seizures occur because serum levels are subtherapeutic at specific times of the day, and it can aid in suggesting more effective drug dosing schedules.F'

Localization of the Epileptogenic Region The most common use of LTME is for localization of a discrete epileptogenic region in patients with medically refractory epilepsy who are candidates for surgical therapy. Consideration for surgical therapy is essentially the only situation in which detailed information about localization is of clinical value. This localization includes not only identification of an area that can be removed when localized resection is contemplated, but

134

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

also demonstration that no such well-defined epileptogenic region exists in patients who are candidates for nonlocalized therapeutic surgical procedures such as hemispherectomy and corpus callosum section." LTME is capable of revealing clinical signs and symptoms of habitual ictal events that have localizing value (see Appendix 5-1); however, these clinical behaviors may result from propagation to distant cortical areas and can never be considered definitive evidence of an epileptogenic region." Consequently, identification of the area to be surgically resected usually requires clear demonstration of the site of electrographic ictal onset. Reliable localization of an epileptogenic region can often be determined with scalp and sphenoidal EEG recordings, in association with a variety of other confirmatory tests; but at times LTME with intracerebral, epidural, or subdural recordings is necessary.7.25,26 A variety of electrode types are used for this purpose, including depth, strip, grid, and foramen ovale electrodes. The performance of long-term intracranial EEG monitoring requires specialized technical and clinical expertise to place electrodes, to guarantee patient safety in the recording unit, and to interpret the EEG data obtained. In large part as a result of technical advances in LTME, approximately twice as many patients underwent surgical treatment for medically refractory epilepsy in 1991 as did in 1985. 27 For some surgical procedures (e.g., standard anterior temporal lobectomy), 70 to 90 percent of patients with medically refractory complex partial seizures can expect to become seizurefree, whereas almost all of the remainder experience worthwhile irnprovement.P The results of LTME techniques are also increasingly used to guide extratemporal cortical resections with beneficial results,26,28 and patients with secondary generalized epilepsies who would not have been considered surgical candidates only a few years ago are now benefiting from large multilobar resections and, to a much lesser extent, corpus callosum sections. 29,30 Safe and effective treatment of epileptogenic zones in primary cortical areas can now he accomplished with multiple subpial transection. 31,32

Electrodes should have a hole in the top to permit periodic regelling, and should be applied with collodion and gauze. Sphenoidal electrodes are often used to record from the mesial or anterior aspects of the temporal lobe in the region of the foramen ovale. 22 These electrodes are constructed of fine, flexible, braided stainless steel wire, insulated except at the tip, and are inserted through a needle guide, as illustrated in Figure 5-1. Sphenoidal electrodes may be left in place from days to several weeks. Evidence suggests that earlobe or Tl/T2 placements are better than nasopharyngeal electrodes,34.35 and they are preferred when sphenoidals are not used. Nasoethmoidal, supraoptic, and auditory canal electrodes have also been used but are difficult to place; indications for their use are not defined, and their routine use is discouraged.F Some investigators prefer to lise a

TECHNICAL CONSIDERATIONS Equipment, methods of procedure, and typical system configurations used for LTME are considered in this section, which is adapted, with permission, from guidelines published elsewhere. 1,2.33

Equipment for EEG Recording ELECTRODES

Disk electrodes are used for scalp EEG recordings; needles and electrode caps are not recommended.

FIGURE 5-1 • The placement of sphenoidal electrodes. The needle is inserted approximately 1 inch anterior to the tragus immediately under the zygomatic bone (black dot on lateral view). The tip of the electrode should lie close to the foramen ovale (basilar view). Inset shows how multistranded Tefloncoated wire protrudes from the tip of the insertion needle and is bent backward on the Teflon coating to prevent breakage of wire strands. Inner lip of the needle can also be beveled to further ensure against breakage of the sphenoidal wire. (From Engel J Jr: Seizures and Epilepsy. FA Davis, Philadelphia, 1989, with permission.)

long-Term Monitoring for Epilepsy

variety of basal electrodes (e.g., sphenoidal plus Tl/T2, and other placements on the face and ear) to better define the fields of basal epileptiform transients. % A variety of rigid and flexible depth electrodes are used for intracerebral recording.s" Most are multicontact. constructed of either stainless steel or metals compatible with magnetic resonance imaging (MRI) , such as platinum and nickel-chromium alloy. They are inserted stereotactically by several techniques that enter the skull from the side, back, or top of the head. 26 Depth electrodes are best suited for recording from structures deep within the brain (e.g., the hippocampus and amygdala), orbital frontal cortex, and cortex in the interhemispheric fissure (e.g., supplementary motor area and anterior cingulate). Consequently, depth electrode evaluations are usually preferred in patients with complex partial seizures of limbic origin. 26 Recordings from the surface of the brain are made with strip electrodes or grid electrodes, which can be placed epidurally or subdurally. Strip electrodes are inserted through burr holes, whereas grid electrodes require a craniotomy for placement." Electrode strips consist of a row of stainless steel or MRI-compatible platinum disks embedded in Silastic, or a bundle of fine wires with recording contacts at the tips. Electrode grids consist of 4 to 64 small platinum or stainless steel disks arranged in two to eight rows and embedded in soft Silastic. The disks in strip and grid electrodes are typically spaced so that there is 1 cm from disk center to disk center. Strip and grid electrodes are preferred in patients with partial seizures whose epileptogenic region is likely to be in the lateral neocortex. Strips are easier to insert than grids and can be used bilaterally. Grids are usually used only unilaterally because bilateral craniotomy is rarely justified, but their extensive coverage allows not only accurate topographic mapping of interictal and ictal epileptic events but also detailed functional mapping of normal essential cortex.'17 Intermediately invasive electrodes are used at some centers. Foramen ovale electrodes are constructed of flexible stainless steel or MRI-eompatible metals and contain one to four recording contacts.P They are placed in the ambient cistern through a needle inserted into the foramen ovale and record from hippocampal gyrus in a manner similar to the most mesial contacts of subtemporally inserted strip or grid electrodes. Although foramen ovale electrodes cannot record directly from hippocampus and amygdala, as do depth electrodes, and do not record as broad a field as strip and grid electrodes, they do have a definite advantage over sphenoidal and other extracranial basal electrodes. Epidural peg electrodes are inserted through twist: drill holes in the skull; they can record from selected areas of the lateral cortical surface as an alternative to strip or grid electrodes.P

135

Foramen ovale electrodes may be used in association with grid electrodes for more comprehensive recordings of mesial temporal structures, including those on the contralateral side. Also, strip electrodes and epidural pegs may be used contralateral to grid electrodes as sentinel electrodes to identify the occurrence of distant ictal discharge, even though the spatial distribution of these epileptic events cannot be mapped. Similarly, strip or epidural peg electrodes can be used in association with depth electrodes to provide additional information about ictal discharges at the cortical surface. AMPLIFIERS

Amplifiers used for LTME are smaller and lower in power than amplifiers used for standard EEG but should have the following performance specifications: low-frequency response of at least 0.16 Hz; highfrequency response of at least 70 Hz; noise level less than 0.5 J.1V RMS; input impedance of at least 1 Mohm; common mode rejection of at least 110 dB; and dynamic range of at least 40 dB.! ,:l3 If a preamplifier is used, preamplification input impedance must be greater than 100 Mohm.P Amplifiers should be able to capture a minimum of 24 channels, and more preferably, 32 or greater, at a minimum of 200 samples per second. The analog-todigital converter should have a minimum of 12-bit resolution with the ability to discriminate the EEG at 0.5 J.1V steps or less.33 To prevent aliasing artifacts, a high frequency antialiasing filter with a minimum rolloff of 12 dB/octave must be used before digitization." The maximum cutoff frequency for the antialiasing filter is determined by the sampling rate. For example, at a sampling rate of 200 Hz, the amplifier must have an antialiasing high-frequency filter no greater than 70 Hz. For higher sampling equipment, proportionately higher-frequency cutoffs can be used. For recording purposes, the low-frequency filter should be set at 0.16 Hz or less. The use of notch filters for acquisition is discouraged. Because LTME systems allow data to be modified when reviewed, frequency filters and gain of the recording system should be set initially to obtain maximum information rather than clean tracings. Amplifiers may be mounted on the head, carried on the body, or remote from the patient. The closer the amplifier is to the signal source, the shorter are the electrode leads and the less artifact that results from movement or interference. The amplifier is most commonly carried on the body because of size and weight. In scenarios where the patient has violent or continuous movement, the amplifier may be mounted on the head to minimize artifact. Small, lightweight amplifiers and preamplifiers are commercially available for this purpose, although they are limited in the number of

136

ElECTRODIAGNOSIS IN ClINICAL NEUROLOGY

channels that can be recorded. Movement and interference artifact is worst with remote amplifiers, and this arrangement is least satisfactory for recording seizures. TRANSMISSION

Standard cable used for routine EEG is the simplest and most widely available means of transmission, but it greatly impairs patient mobility and is associated with the greatest amount of artifact. Telemetry refers to a system in which the EEG signal is amplified close to the patient, multiplexed, and then transmitted to a remote recording device, where it is decoded.l" With cable telemetry, the multiplexed signal is transmitted over a lightweight cable that consists of a single wire that is long enough to allow the patient to move around the room. Cable telemetry is inexpensive, is associated with low interference, and is the most common form ofEEG telemetry used; however, patient mobility is relatively limited. More mobility is possible with radio or infrared telemetry, in which signals are transmitted without a wire, but the range of these devices is still restricted, and there is increased interference with the EEG signal from outside sources. To overcome this patient restriction, some recently developed systems can store the multiplexed signal directly onto flash memory within the amplifier/headbox and retransmit it to the system at a later time. This allows the patient to be unhooked from the system for moderate periods (up to 4 hours with most systems) and move about freely without loss of data. Although this does not offer the same degree of freedom as radio telemetry, it has the advantage of allowing the patient limited periods in which they are not restricted by cable lengths without subjecting the data to radio frequency interference artifacts. Ambulatory recording for longer periods (discussed in Chapter 6) is a specific form of LTME in which the patient wears the recording device and no remote transmission is necessary. However, because of limitations of storage, the number of channels are limited, and often these devices are event recorders and do not provide continuous long-term records. Amplification close to the body and multiplexing of the resultant strong signal accounts for the major advantage of telemetry. The low-voltage signal travels only a short distance and so is less prone to movement artifact and interference from outside sources. Recordings of epileptic seizures, which are usually associated with considerable movement, are therefore relatively artifact-free compared with ictal recordings obtained in a routine EEG laboratory. These advances have made feasible LTME with scalp and sphenoidal electrodes. Multiplexers typically combine 16 to 64 channels into a single channel of information, which is then demultiplexed at some later point. If the multiplexed signal is recorded and stored, the recording

apparatus must have a higher frequency response than if the signal is demultiplexed before recording and storage. RECORDING, STORAGE, RETRIEVAL, AND REVIEW

Many methods of EEG recording and storage are currently available. 1- 4,38-40 The simplest, a continuous paper printout, is an expensive and labor-intensive approach that requires a dedicated EEG machine and the attention of a qualified technologist. The inability to manipulate the waveforms in addition to the bulky nature of paper printouts have made this technique obsolete for LTME. Current systems use analog-to-digital conversion methods, and the data is stored on a computer hard drive, server, or compact disc/DVD media. Previous systems used videocassette recorders, or audiotape recorders in the case of ambulatory cassette recording; this has given way to digital systems using large-storage hard drives or flash memory, Digital video has replaced analog video and videotape. This approach allows continuous unattended recording from 6 hours to several days, depending on the system. Digital storage allows data to be easily modified on review, and it is amenable to computer reduction and analysis. Compact discs and DVDs are also less bulky and cheaper to store than videotapes. The storage capacity of the system depends on the sampling rate and the number of channels acquired. For 32 channels of EEG sampled at 200 Hz, a compact disc can hold approximately 20 hours of continuous data, whereas DVD media can store approximately 90 hours. Higher sampling rates and greater numbers of channels decrease this amount proportionately. Systems that permit selected EEG storage employ a built-in time delay so that the data stored include EEG activity recorded both before and after the device is triggered. 1-4 ,40 Computer-recognized interictal or ictal events are used to activate the system, so only the events of interest are stored. Because computerized detection programs are not completely accurate, false-positive detections and failure to detect genuine events occur. However, this approach greatly reduces the amount of data that needs to be stored. Storage can also be activated by a pushbutton that the patient or an observer uses to identify ictal events. Such an event-recording approach requires that the patient or an observer be able to recognize when a seizure is occurring, and seizures that occur without warning or subclinically may be missed. A portable system is now available that allows computerized detection of ictal and interictal events for use with an ambulatory recorder.v'r" The advantages of digital video, including the ability to zoom in on the image after recording and the ease of EEG and video time-locking on review, have made it

Long-Term Monitoring forEpilepsy

the current standard. In addition, the ability to store the video on the same medium as the EEG information has eliminated the need to keep separate libraries of videotapes and EEG. However, the digital video image is the most limiting factor for storage. The minimum video resolution for LTME is 640 X 480 pixels. At this low resolution, even with compression techniques, these image files are very large. The higher the image resolution, the larger are the image files. Thus there is a trade-off between the quality of the image and the amount of video that can be stored. To overcome the difficulty with storage, most systems ofIeI' the ability to edit and delete unwanted video on line before permanent storage, keeping only the clinically necessary video segments. The most common method for review is to display both the EEG and video on a high-resolution monitor. This has several advantages including the ability to reformat the EEG montage, filter settings, and gain. In addition, the digital video image can be time-locked to the EEG via a cursor to easily correlate clinical behavior with the EEG. In order to display the EEG waveform accurately, the display must have certain minimal characteristics. The display should have a minimum scaling ability for each channel such that 1 second of EEG occupies approximately 30 mm, with a resolution of 120 data points per second.P Vertical scaling depends on the number of channels displayed, but a minimum of four pixels per vertical millimeter is required for accurate reproduction of the waveforms.V The system should also offer the ability to digitally mark the EEG for technologist comments, pushbutton events, and automatic detections. The major advance in this area of LTME is the capability to record digitally onto disk or flash memory so that the continuous attention of an EEG technologist is not necessary. This has made 24-hour monitoring relatively inexpensive. The added refinements of computer-detected and patient-triggered selective identification and/ or storage have reduced the amount of data that need to be reviewed. Techniques for rapid review of EEGs have also facilitated data analysis, particularly for ambulatory recording, in which the addition of an audio channel has enhanced recognition of meaningful events." The development of digital recording systems makes it possible to reproduce data on a high-resolution computer monitor in any montage, gain, or filter setting desired. 38.4o Movable time markers, spike maps, and other programs for analysis permit facilitated interpretation, and networking between workstations at multiple locations is feasible.

Equipment for Monitoring Clinical Behavior Methods of monitoring behavior include self-reporting, observer-reporting, video and/or audio recording, and

137

detailed monitoring of specific physical or cognitive functions. 1-4.38.39 Self-reporting requires the patient to make notes in a daily diary or log, indicating the occurrence of the specific events in question. This self-reporting can be accompanied by the use of a pushbutton that marks the occurrence of an event on the EEG tracing. Self-reporting is the common method of recording clinical behavior during ambulatory cassette monitoring. Observer-reporting uses the same methods but someone else, such as a parent, maintains a log and activates the event marker. For inpatient LTME, specially trained nurses or technologists can be made available to examine the patient during a seizure in order to elucidate the clinical behavior and to better elicit mental and neurologic deficits. A useful approach to the clinical examination during spontaneous epileptic seizures is shown in Table 5-2. Video and/or audio recording is ideally used in association with self- and observer-reporting and is the usual practice for inpatient LTME. This approach requires one or more video cameras (which are continuously focused on the patient) and a mechanism for synchronizing the video-recorded behavior with simultaneously recorded EEG activity. In selected situations it may also be helpful to use polygraphic instrumentation, continuous reaction time monitoring, or other automated techniques to identify specific physiologic or cognitive disturbances during ictal events. These data are usually recorded along with EEG activity.

TABLE 5-2 • Clinical Examination During Epileptic Seizures A. Ictal phase 1. Mental status a. Determine responsivity to commands, orientation, language function. b. Present a nonsense phrase for later recall, to determine amnesia. 2. Motor a. Note site of initiation and pattern of motor symptoms, clonic and/or postural. b. Note focal or lateralizing motor deficits during spontaneous movements and, if possible, provoke movements to confirm deficits. 3. Sensory In special situations it might be useful to demonstrate a general analgesia to pinprick or to document a specific sensory deficit, such as ictal blindness. B. Postictal phase 1. Observe spontaneous abnormal behavior (e.g., automatisms, combativeness, or unresponsiveness); determine time course of resolution. 2. Examine for specific focal or lateralizing neurologic deficits, including cognitive deficits. 3. Test for recall of nonsense phrase given in A-Ib,to determine amnesia for ictal event. 4. Elicit description, if possible, of aura, behavioral seizure, and postictal symptoms.

From Engel JJr: Seizures and Epilepsy. FA Davis, Philadelphia, 1989,with permission.

138

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

A variety of black-and-white and color video cameras are available for use in LTME, including low light-level cameras and infrared cameras that can record at night. A common approach is to use two cameras, providing a split-screen image with one camera focused on the entire body and the other on a close-up of the face. Lens irises are available that automatically adjust to changing light conditions. Remote zoom and pan-and-tilt devices allow monitoring personnel to maintain the patient in the desired place on the video screen. For large units with several monitoring beds, a full-time monitoring technologist can be employed for this purpose. For smaller units, nurses or EEG technologists can watch the monitor and make adjustments when they have a chance; however, information may be lost. An alternative approach is to allow the patient to maintain the appropriate adjustments of the video equipment. Four approaches to auto tracking are currently available for video monitoring of patients who are moving around a large room. One depends on a radio frequency transmitter worn by the patient. Another uses a coded pulse of infrared light emitted from the camera, which is reflected off a strip worn by the patient and picked up by a detector on the camera. For the third, multiple cameras surround the room and the recording device determines which camera is recording from the patient by use of a computer that senses contrast. The fourth technique uses image processing to track specific defined features of the patient. In previous systems, analog video and/or audio recordings were synchronized with EEG recordings by a specialized time code generator that displays the date and time on the video image digitally and on the EEG in binary code or alphanumeric.l ? With modern systems, EEG and behavior are automatically synchronized when the data is stored digitally, and simultaneous display of EEG and behavior on the video image occurs with the reformatting technique. Clinical events are digitally marked on the EEG record at the appropriate point in time, and then re-reviewed by reading the EEG with the clinical notes. It is also possible to review the video time-locked to the EEG, simultaneously displayed on a high-resolution computer monitor. In this way discrete EEG events can be correlated with behavior directly, or vice versa. Unidirectional and omnidirectional audio microphones can be used during LTME. It is important to have a microphone that eliminates unwanted extraneous noise but does record sounds corning from observers who are describing ictal events, as well as sounds corning from the patient. The most bothersome extraneous noise on LTME units is the television in the patient's room. Consequently, it can be useful to wire the pushbutton that the patient uses to signal the occurrence of an event so that it also automatically turns off the television set.

Digital video is preferred over cassette videotape recorders. Cassette videotape recorders can only record 6 hours of video per cassette, requiring four recorders for 24 hours of continuous, unattended recording. Because digital video records to the computer hard drive or server, the continuous duration of recording is limited only by the storage capacity of the system. Editing and mastering of the digital image is much easier than with the systems used for cassette videotape, and image quality does not degrade with copying. With the networking capabilities of most systems, digital video also adds the ability to remotely view the patient from any location. In the future, both higher data storage systems and new data compression techniques will allow higher resolution video for longer continuous lengths of time.

Methods of Procedure ELECTRODES

Disk and sphenoidal electrodes should be checked periodically for breakage. Regelling of disk electrodes should be performed as necessary to maintain low impedance. Scalp irritation can usually be avoided if the patient uses a strong antidandruff shampoo before the electrodes are applied. Impedance is measured at the beginning, periodically during, and with ambulatory EEG at the end of recording. Impedance should be maintained at less than 5,000 ohms. Malfunction of intracranial electrodes cannot be corrected after they are inserted; however, problems with the special connectors can be resolved, and these must be inspected periodically. Impedance measurements of intracranial electrodes can be performed safely with currents in the range of 10 nA. For patients with intracranial electrodes, the guidelines for indwelling devices should be followed (in the United States, UL type A patient). RECORDING SYSTEM

The integrity of the entire recording system from electrode to storage medium should be checked before beginning LTME, and periodically during the monitoring. This is most easily done by observing the ongoing EEG and by having the patient generate physiologic artifacts. In addition, most systems have automatic impedance checks, can identity loose or bad electrodeto-amplifier connections, and have self-diagnostics to ensure proper recording parameters. The technician should also periodically review the record files to ensure that they are being stored properly. All instruments should also be calibrated as suggested by the manufacturer. More specific calibration systems are recommended for ambulatory EEG recorders.

long-Term Monitoring for Epilepsy RECORDING TECHNIQUES

Although a minimum of 24 channels is recommended for LTME, 32 or more are routinely used and are particularly necessary when basal electrodes are employed. 1•2. 33 Thirty-two channels are standard for scalp and sphenoidal LTME to localize epileptogenic regions for surgery; and 64, 96, or more channels are being used increasingly with intracranial electrodes. Sixteen-channel and 24-channel ambulatory recordings are now also possible with selective storage triggert'd by computerized spike and seizure detection and with an event marker operated by the patient or an observer.i'r" With most recording systems it is possible to record extracranially using a common reference montage, and then to review the data in any desired montage.t" The most useful montage for extracranial recordings with basal electrodes (e.g., sphenoidal, earlobe, or Tl/T2) employs standard lateral temporal bipolar chains in association with independent bipolar chains involving the basal electrode sites (Table 5-3), with additional derivations as equipment allows. Normal variants such as small sharp spikes and 14- and 6-Hz positive spikes are more easily distinguished from epileptic transients

TABLE 5-3 • Basal Electrode Montages forLTME Basal Electrode (PG) Montages Independent from Lateral Temporal Derivations 16-channel 12-channel Fp1-F7 Fpl-F7 F7-13 F7-T3 T3-TS T3-T5 T5-01 T5-01 Fp2-FB Fp2-FB FB-T4 FB-T4 T4-T6 T4-T6 T6-02 T6-02 C3-PGl 0-13 T3-PGl or T3-PGl PG1-PG2 PG1-PG2 PG1-Nz PG2-T4 PG2-C4 Nz-PG2 ECG ECG PG2-T4 T4-C4 C4-Cz ECG Basal Electrode (PG) Montage Not Independent from Lateral Temporal Derivations Fp l-F7 Fp2-FB F7-PGl FB-PG2 PG1-T3 PG2-T4 T3-T5 T4-T6 T5-01 T6-02 From Engel JJr, Burchfiel J, Ebersole Jetal: long-term monitoring for epilepsy: report of anIFCN committee. Electroencephalogr Clin Neurophysiol, 87:437, 1993, with permission.

139

in this manner. Inclusion of C3/C4 electrodes on the basal electrode chain is useful when cancellation occurs between basal electrodes and T3 or T4. The same montages are appropriate for 16-channe1 ambulatory recording. When fewer channels are availahle, montages are usually designed to display data symmetrically from both hemispheres, preferentially sampling frontal and temporal regions as discussed in Chapter 6.1.2 Montages used with intracranial electrodes vary greatly depending on the electrodes used and the cerebral areas of suspected involvement.

LYME System Configurations Table 5-4 lists the common basic system configurations recommended by the American Electroencephalographic Society- and the International Federation of Clinical Neurophysiology.' Although newer systems all use digitally acquired EEG and video, older systems are still available and in use for LTME, particularly in developing countries. Their configurations are summarized in this section. Monitoring with paper printout only is appropriate for documentation, characterization, and quantification of clinical and subclinical ictal events and interictal EEG features over a period of hours, as is assessment of the relationship of these EEG events to behavior, performance tasks, naturally occurring events or cycles, or therapeutic intervention. It is not appropriate fill' evaluations that require the patient to move freely or when continuous monitoring is necessary fill' days. This form of long-term monitoring is relatively labor-intensive and becomes progressively more cumbersome with duration. At least 16 channels of EEG data and synchronized video monitoring are required for pr-esurgical localizetion of epileptogenic regions with this and all subsequently described configurations. Because of the limitations of this method, it is no longer considered the standard of care. Monitoring with continuous storage is appropriate for the same purposes as monitoring with paper printout only, but for days and weeks rather than for hours. It is not appropriate for the quantitative analysis of subclinical ictal or interictal features or for evaluations benefiting from complete freedom of movement. Radio telemetry provides more mobility than does cable telemetry; however, video monitoring becomes difficult or impossible when this degree of mobility is required. Computer-assisted selective monitoring is most appropriate for the same purposes as the previous two configurations, but it also records subclinical as well as clinical ictal EEG events. It is not appropriate for evaluations benefiting from complete freedom of movement. Computerized recognition programs generate false-negative and false-positive errors, so important data may be missed with this technique.

140

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

TABLE 5-4 • Commonly Used LTME System Conflaurations Monltorlnl with p.per printout only EEC transmission: "hard wire" (standard cable) ortelemetry (cable orradio) EEC recording/storage: continuous paper printout EEC review/analysis: complete manual review Behavior monitoring: self, observer, and video Monltorlnl with continuous stor'le EEC transmission: cable orradio telemetry EEC recording/storage: analog orvideo tape, digital media (hard disk, server, CD/DVD) EEC review/analysis: selective review ofclinical ictal events, random sampling review for subclinical ictal and interictal events Behavior monitoring: self, observer, and video Computer·asslsted selective monltorlnl EEC transmission: cable orradio telemetry EEC recording/storage: digital tape/disk, computer-assisted selective storage EEC review/analysis: selective review ofclinical and computer-recognized ictal and interictal events on high-resolution monitor Behavior monitoring: self, observer, and video Ambulatory casseUe-seledlve event recordlnl!epoch sampllnl EEC transmission: ambulatory flash memory orhard disk (16- or 24-channel) EEC recording/storage: flash memory, hard disk, ordigital media (CD/DVD); event recording/epoch sampling (periodic orafter trigger) EEC review/analysis: epoch review Behavior monitoring: self, observer

Continuous storage recording with computerassisted event detection is preferred; it is used at most epilepsy centers. This technique reduces the amount of data that need to be reviewed while maintaining the continuous data if a more detailed review is desired. Typically, the data are reviewed, and the technologist saves selected segments. Sixteen- and 24-channel ambulatory monitoring is most appropriate for documentation and quantification of ictal (clinical and subclinical) and interictal EEG features and assessment of their relationship to reported behavior, naturally occurring events or cycles, or therapeutic intervention, when the data are most likely to be obtained outside the hospital or laboratory environment. With the addition of computerized spike and seizure detection, it is also appropriate for detailed characterization of EEG features for classification of seizure types and presurgical evaluation.:" This configuration, as well as the other ambulatory configurations, can be used in an inpatient setting and is particularly useful when mobility is of benefit. The absence of EEG changes during a reported ictal event does not rule out an epileptic condition.

INTERPRETATION OF RESULTS

Artifads EEG data obtained using LTME techniques are likely to contain unusual artifacts that are uncommon, or never encountered, in the standard EEG laboratory (Fig. 5-2). These are often peculiar to the systems in use, and the electroencephalographer working on an LTME unit must become familiar with their appearance. Scalp and sphenoidal electrodes show the usual biologic artifacts seen with standard EEG, but, in addition, chewing, talking, and teeth-brushing can produce dectromyographic (EMG), glossokinetic, and reflex extraocular movement potentials that are more difficult to identify. Intracranial electrode recordings are usually free from biologic artifacts except for pulsation and respiration. Altered electrode-scalp contact and intermittent lead-wire disconnection caused by body movement are the most common sources of mechanical artifacts during LTME. Artifacts induced by movement of a directconnection standard cable are avoided when telemetry systems are used. Rhythmic mechanical artifacts produced by rubbing or scratching the scalp, or movements of the head or extremity (particularly when associated with accompanying biologic artifacts) can be mistaken for ictal discharges. Interference from nearby electrical equipment, electrostatic potentials from movements of persons with dry clothing, or telephone ringing can also produce bothersome EEG transients. Mechanical artifacts caused by body movement and electrical interference are usually much less with intracranial than with extracranial recording. Electrodes, wires, amplifiers, receivers, switches, reformatters, tape recorders, oscillographs, or any other part of the LTME system can potentially cause artifact. Electrode pops, faulty switches or connectors, and contact between dissimilar metals are common causes of spurious transients, whereas chipped silver-silver chloride coating, instability of the electrode-scalp interface, and electrode wire movement are the most common causes of rhythmic slow waves. As with routine EEG, diagnosis of an epileptic ictal EEG event is made by recognition of well-formed spikewave or other characteristic patterns with a believable field and typical ictal progression, followed by postictal slowing and appropriate interictal abnormalities in other portions of the record. Confounding artifacts caused by biologic and mechanical disturbances can usually be identified when simultaneous videorecorded behavior is available. For ambulatory EEG monitoring, when video is not available, it is standard to have the patient and technologist produce common biologic and mechanical artifacts at the beginning or at the end of the recording, where they can be used as a

141

Long-Term Monitoring for Epilepsy

J~ ~~V'

--+-I

~-,A,V

~fA. j

J"M. VJ\f"-

II ,A

FN

r= ~~

~

I

~~

U

v

,;

".

y ,

•

v-

\,i

"

'v

,cAr\,JV ''-"

~y~ -J'-Jv,J;f'JV",,~,/'

,

,

1

t~:~r,Nj\-/

NV\yV~ ~A)'V:VV\\.I)/'

"

~·V~;-,-/·,tV·

~

"

Aft.

I

v

"'V·,

,,', t,

. v ,. '{

. . A

1

..

',,"

v YVV v v

'

-e

,t-.;J..,A "V",

vy'

h,,1; " v Av 'Yi,V

ft.

V v

~

elM)

.AM;

JlI\r.. v

"

yY

'.--;---'-.

" ' v

,

!\!\

1/

.,

'~""l"·

B

A ''''1

~"

~.~ ~-

"

--....

~

~I( rt

<,--

~

....... _101

-

-~ ,""L

"'~~

.,. C

IJ

.,

,

i

:-

~VI

c

VV_Y

\-I

'-J\J"iN'Ir/"" I

I ~.

--,__

. '"

, v ~~_J_vv~ " ~------

C4-T4A/'<~~

'-' ~

v YV

'VVVV V

-

i

T4-S2~-V~

A "'-.A

~.~~

IX "V~VX~

"""

A....,..,-..-"..--.

n,y f!YrWt fMYR~ ~@;:~

rl'~

~~ ~A

~

F3-FI

I"

A'~~

v

,

I

F&-F~

r-.

~I'. ]~J1

~"'vV-~~J\'

~' "

,

~ L

ci

.

.

~

j'h'

'""

L

'-"I I~i

.,

;]I

r.' w

D

E

F

FIGURE 5-1 • Artifacts commonly encountered during EEG telemetry recordings. A, Tapping on T3 produces an artifact in SI as well because the sphenoidal wire is curled under the head-dressing. This gives the false impression of a field; however, careful inspection reveals a double phase-reversal. B, Scratching over the right posterior head region demonstrates how this apparent field effect can result in rhythms that resemble an ictal discharge. Again, the double phase-reversal provides the clue that this is not a cerebral event but a mechanical artifact. C, Rhythmic muscle activity is often seen during chewing. D, Many patients will rock their head wrapping back and forth in order to relieve itching and discomfort, producing regular or irregular slowwave artifacts that have a wide distribution but display multiple phase reversals. E, Intermittent sudden loss of signal usually indicates a poor connection. When the artifact is generalized, but some signal remains, as shown here, the most likely site of disconnection is the montage card, where the end of the card remains connected. F, Current telemetry systems are so good that movement artifact other than direct mechanical interference rarely distorts the signal. However, movement artifact, as shown here, can occur as the only consequence of a disconnected ground lead. Calibrations: I second and 100 1lV.

reference for confusing transients on that tape. Conservative interpretation of unusual or equivocal EEG events is recommended.

Epileptic Adivity Interictal spike-wave discharges are identified on LTME, as on standard EEG, using artifact-free segments of the recording for analysis. Interictal spikewave discharges recorded from depth electrodes are commonly much more widespread than those seen during extracranial recording, and they do not correlate as well with the location of the epileptogenic region. 45 ,46

Ictal discharges recorded from depth electrodes can originate focally, from one or two electrode sites, or regionally, from several electrodes simultaneously. Focal onsets are believed to be more localizing, particularly when mesial temporal structures are involved, because regional onsets may represent propagation from a distant epileptogenic area that is not detected by available recording electrodes. Ictal onsets that originate simultaneously from both sides of the brain, or simultaneously from multiple intracranial and extracranial electrodes, are generally not useful for identifying an epileptogenic region. 45 ,46 Examples of focal and regional ictal onsets recorded with depth electrodes are shown in Figure 5-3.

142

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

13 sec

LAMYG~~

LA H"""'-....-........ ...............

•

i

..... .........,..."....

..."' ..... tr-

"i..ItW,fft,'~

~ .............

LPG~~

LMG;~~

A

o

L AMYG-~-~~~-----....--~-~~""",

;1 ~~

L AMYGM~,~r~~~·~rti~:;.J'fl.Ji ~ ~

L A H----~--------~-----............-.r

~~-v/------

I~ 'I~

G~'''¥('Iv a~v''''WI'.t-(vJ\v.h/~~

LM •• L P G~~~J'~'''--''''"··'''''-/''"'-J'v''i'I'''/'Ii\\1

~---/'-~--..~~~

•

I

E

B RAMYG~~'-"/II~~

R A H - R M H~M~~~!lJr~

R A H~VVW"'~/J",/'N';(t RM

c

I,:

L M H~------'.-c-~o~""",~v,
L P G~~"---"-~------"""""""'----"--"""""''''''-V'''''' L0 F

~'I\,~'

I

H~MMfvVv~W~w~~,M

RMH-RPH~~J\"I\.i\.J~ R AMYG - R A G~~~~V\J'{'\J\!\~I\.tI\WI!t~

R M GJ\V"'VI,fi~~N~IifVI!'vfV"

RAG - R P G~!v(\~rvv-v~l);jj)~~~

F

FIGURE 5-:1 • Segments of depth electrode EEG telemetry. Ictal onsets from 6 patients are shown. These tracings represent

examples of progressively decreasing localizingvalue. A, Very-low-voltage fast activity seen initially only as thickening of the trace (arrow) isolated to the left anterior hippocampal pes (LAH); this continues for 17 seconds before it is barely seen at other areas. B, Low-voltage fast activity of much lower frequency than that seen in part A begins (arrow) at the left posterior hippocampal gyrus (LPG) and appears in all depth leads on the left after 5 seconds. C, 4- to 5-Hz sharp activity begins in the right middle hippocampal pes (RMH) (arrow) and 1 second later is slightlyreflected in all depth electrodes on the right. D, Sharp activity begins with phase reversal in the left posterior hippocampal pes (LPH) (arrow) and remains most prominent there, although it is reflected in all the other depth electrodes. E, Ictal rhythmic activityfirst appears in the left middle hippocampal gyrus (LMG) and later spreads to other depth electrodes; this is preceded by a regional suppression (arrow) involving all left temporal depth electrodes. F, Ictal discharges begin with irregular, regionally synchronous spike, polyspike, and slow-wave bursts followed by a buildup oflow-voltagefast activity. which is also synchronous in both hippocampal pes and gyrus. Calibration: I second, For each sample, channels not shown recorded from homologous contralateral depth sites, extratemporal, skull, and sphenoidal derivations. (From Engel J Jr, Crandall PH, Rausch R: The partial epilepsies. p. 1349. In Rosenberg RN [ed]: The Clinical Neurosciences. Vol 2. Churchill Livingstone, New York, 1983,with permission.) Two distinctive types of intracranially recorded focal ictal onset-patterns are recognized, One consists of lowvoltage fast activity that builds up over several seconds and resembles a localized version of the recruiting rhythm of generalized tonic-clonic convulsions (Fig. 5-4, A). A variation of this pattern is suppression of ongoing EEG activity. The second type, typical of hippocampal onset. consists of localized high-amplitude spike or spike-wave discharges, which may resemble interictal spike-waves and can occur at various frequencies (Fig. 5-4. B and Fig, 5-5). Some may begin very slowly and increase in frequency as the seizure progresses, whereas others may begin with relatively rapid 3- or 4-Hz discharges. These latter patterns are more often subclinical or associated with auras, and commonly evolve into the low-voltage fast discharges that then propagate and give rise to complex partial seizures (Fig. 5-4, B) ,47 Ictal EEG events that have no clinical cor-

relate are most often seen when computerized seizure detection is used; these electrographic seizures are helpful in identifying the epileptogenic region and may also have prognostic significance.f Ictal discharges, as seen with intracranial electrodes, may not be reflected in the scalp or sphenoidal EEG.14 Particularly with mesial temporal ictal patterns, the initial discharge in the depth will continue for tens of seconds before propagating contralaterally, and only at that time do ictal EEG rhythms appear in the sphenoidal and, later, scalp derivations. On most (85 to 90 percent) occasions, the ictal onset will appear first in the sphenoidal electrode on the side of seizure origin. even though activity in the depth of the brain is usually bilateral at that time.t? The early sphenoidal ictal pattern consists most often of characteristic 5- to 7-Hz rhythmic activity, which may occur at the outset or may begin after a more diffuse initial ictal EEG change.

long-Term Monitoring forEpilepsy

143

136

"/··J·~·j,,j/\;,,(lj\p')IV\/\I\rJ·.t~W'/'~~~~'J/,~~

r\r~lJc ,Jv-JlrJ~lrJlrJlrJLJ~ ~l,;

FIGURE 5·4 • Segments of telemetry recordings from two patients showing stereoelectroencephalographic activity at selected depth electrode bipolar tips during the onset of complex partial seizures. A, The classic depth electrode recorded ictal onset consisting of a buildup oflow-voltage fast discharge, here beginning in a single channel (arrow). B, Three continuous segments show a more common ictal onset-pattern, beginning with rhythmic high-amplitude sharp and slow transients (arrow), eventually giving way to a low-voltage fast discharge, which then evolves into higher-amplitude repetitive spikes or spike-waves. L, left; R, right; A, amygdala; AH, anterior hippocampus; MH, mid-hippocampus; PS, presubiculum; PC, posterior hippocampal gyrus. Calibration: 1 second. (From Engel J Jr: Brain metabolism and pathophysiology of human epilepsy. p. 1. In Dichter MA [ed]: Mechanisms of Epileptogenesis: The Transition to Seizure. Plenum Press, New York, 1988, with permission.)

A delayed focal sphenoidal onset, defined as one in which the characteristic 5- to 7-Hz rhythmic activity develops within 30 seconds after any initial ictal EEG change that is not predominantly contralateral, has been found to be as accurate as an initial focal pattern in localizing the epileptogenic region (Fig. 5-6) .49 The ictal discharge is known to persist for long periods in one mesial temporal structure before propagating to the scalp EEG, and so the temporal relationship of the sphenoidal-recorded ictal onset and behavioral ictal signs or symptoms was not considered to be important in this classification.

Ictal onsets recorded from both intracranial and extracranial electrodes can be extremely subtle and difficult to recognize. At times, they appear only as a change in the frequency of ongoing activity that is restricted to a few channels and is not consistent with a state change (see Fig. 5-3, A and B). Usually, the new activity is more rhythmic than the ongoing baseline tracing and may contain faster frequencies that are not seen in the interictal recording. Such changes are much more likely to occur with intracranial recordings than with extracranial ones, although they do occasionally appear with scalp and sphenoidal recordings as well.

'~I~ RAH-RP5

r-~

RPS-RMH RMH-RPG RA-MMTG RM-P MTG

LA-M MTG LM-P MTG T4-52 52-51

"I'm having an aura"

~f\~~ ~!f'I'M',,",I~"viJW""'W'/'~'i\f'M,-JWIf'J'-'-.P~~"I\t-' 'v.t"v"""""'w"W~""""I!.M~IIV/t(I~~I/ftI,o\\~~

FIGURE 5·5 • Forty continuous seconds of an EEG recorded from depth, sphenoidal, and scalp electrodes during a simple partial seizure involving the right temporal lobe. The ictal event begins in the left portion of the upper panel as an increase in interictal spike discharges that are maximal at the right anterior hippocampal electrode (RAH). After 8 to 9 seconds, these spikes become regular, eventually developing into a 3-Hz spike-wave pattern involving all derivations from the right mesial temporal lobe before ending abruptly in the right portion of the lower panel. Note that no low-voltage fast ictal activity is seen, either initially or at any part of the ictal episode. Videotape analysis indicated that the patient reached for the call button at the arrow, at which point regular rhythmic slow activity is also seen in the left anterior hippocampus (LAH) and the right sphenoidal (S2) electrode. The patient then indicated that she was having an aura, which consisted of a sensation of fear in her stomach. Calibration: 1 second. RA, right amygdala; RAH, right anterior hippocampal pes; RPS, right presubiculum; RMH, right middle hippocampal pes; RPG, right posterior hippocampal gyrus; RA-M MTG, right anterior-to-middle middle temporal gyrus; RM-P MTG, right middle-to-posterior middle temporal gyrus; LA, left amygdala; LAH, left anterior hippocampal pes; LPS, left presubiculum; LMH, left middle hippocampal pes; LA-M MTG, left anterior-to-middle middle temporal gyrus; LM-P MTG, left middle-to-posterior middle temporal gyrus. (From Engel] Jr: Brain metabolism and pathophysiology of human epilepsy. p. 1. In Dichter MA [ed]: Mechanisms of Epileptogenesis: The Transition to Seizure. Plenum Press, New York, 1988, with permission.)

145

long-Term Monitoring forEpilepsy

ALr..~---"'I""'1I'-~---~'-"""'J~ ~VV\[~~ 1~~"'-"-

'"""-__""-"-r-"""""""""'A'v"""'t4N'~4 ' i \ , . ~.......,

----'.....,..,---........".....~--.r-\__.r.-.-'V""V

'AA"....---.......""""'_---."...

~~"""'lil ......W4IIIr-r'-'.......,...........IIiiNIIIMfoI. ..~~_ .,;-.1.I'kli11tW----~~~~-----~~--'

.............

,t
,~

............".~

FIGURE 5-6 .. Examples ofEEG telemetry-recorded ictal onsets from four patients with complex partial seizures. A, Low-voltage 6- to 7-Hz rhythmic activity appears at the right sphenoidal electrode (arrow) 5 seconds before it is seen over the right temporal convexity (initial focal onset). B, Following a diffuse burst of muscle and eye movement artifact, low-voltage fast activity is recorded by the right sphenoidal electrode (arrow; initial focal onset). This becomes progressively slower, and the amplitude incre-ases; 5 seconds later it is seen diffusely over the right hemisphere. C, Irregular, sharply contoured slow waves demonstrate phase reversal at the right sphenoidal electrode (arrow) and are reflected as low-amplitude delta, without phase reversal, over the right hemisphere (of undetermined localizing value). D, In this lateralized but not localized ictal onset, voltage suppression and low-voltage fast activity occur over the right frontotemporal area and are best seen at the right sphenoidal electrode (arrow). This precedes by 3 seconds the appearance of diffuse 3-Hz spike-wave discharges, which are also more prominent from the right frontotemporal and sphenoidal derivations. After 10 seconds, this latter activity evolves into high-voltage 7-Hz sharp-waves, which show phase reversal at the right sphenoidal electrode and laterally at the right anterior to midtemporal region (delayed focal onset). Calibrations: I second and 100 1lV. Note that sensitivity is the same for parts A, B, and C, and the first half of part D, but decreased to half in part D at the first calibration mark. (From Engel J Jr, Crandall PH, Rausch R: The partial epilepsies. p. 1349. In Rosenberg RN red]: The Clinical Neurosciences. Vol 2. Churchill Livingstone, New York, 1983, with permission.)

The most common extracranial correlate of this ictal onset-pattern is a cessation of interictal spikes, because interictal and ictal events do not co-exist. Consequently, the disappearance of frequent interictal spiking may be the only EEG evidence that a simple partial seizure has occurred, and ongoing interictal spikes often appear to stop several seconds before a complex partial seizure is recorded from sphenoidal electrodes.

Nonepileptlc Abnormal Adivity The presence of continuous focal slowing or focal attenuation of normal fast activity during the interictal

state has important localizing value. Such interictal asymmetries recorded with depth electrodes implanted into mesial temporal structures are indicative of hippocampal dysfunction and usually correlate with the presence of hippocampal sclerosis. It is therefore important to make use of LTME to identity nonepileptic abnormalities when the purpose of this evaluation is to localize an epileptogenic region for surgical therapy. Functional mapping is an essential part of the presurgical evaluation when the epileptogenic region to be removed is adjacent to primary motor, sensory, or language cortical areas that cannot be damaged. Evoked potentials and cortical stimulation can be used to identity these areas extraoperatively, during LTME

146

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

with subdural grid electrodes.F Maps of the epileptogenic region, derived from recording interictal and ictal epileptic abnormalities, can then be combined with the results offunctional mapping to determine the boundaries of cortical resection before the patient enters the operating room (Fig. 5-7).

RECOMMEN DATIONS The American Electroencephalographic Society' and International Federation of Clinical Neurophysiology'

have indicated the following rmmmum standards of LTME practice for specific indications. Since these standards were published, however, the use of digital EEG technology with paperless review that permits post hoc data manipulation has become standard in most laboratories. Also, the American Academy of Neurology has published a practice parameter on epilepsy surgery'" that impacts on the use ofLTME for this purpose. When inpatient LTME is performed, an EEG technologist, monitoring technologist, epilepsy staff nurse, or other qualified personnel must be available to observe the patient, record events, and maintain recording integrity.

o

Seizure onset

• Speech cortex (Broca's area) • Motor cortex

1A-CZ ~~~.rvJ0""~"""' 1B -

. ,. . . .

"""1IIlII4......

"""VV"--",;/,"""''IWI(V-"",,",!'f'I.''''NI'''h'lW\~ffII.,-.J~...-.........~- ..... ~.,

,....

*",.,1'.

......

~

. .f

If

I". fI\e",'" ..

",,~ ~..,.. ...",.,#."'.{'fIftW~1'vy­ ,,";JII.*,,/,I

~It-N~

1C -

101E -

1F1H 1J -

2A2B-

2C20 -

v¥t~~~\rW{Whl~'~~~""""""'''''''''''''''''~'''''H+'''''''''''~~~ r~~o/'Ji"I1'--"'-'''''''''~~-~~'''''''''~~--''''''''''''''''''~~~V''W' ............. ,.......r-rww--,~w...,N>M,............,''r,,",,,,,,'''''''''"''

""~,~~rl-o"""''''''''''~~''''''''I'
2E2F 2H -

Stops talking,

Lifts head

reaches for buzzer FIGURE 5-7 II Extraoperative functional mapping and ictal recording with subdural grid electrodes. Lateral skull radiographic outline shows placement of an anterior and a posterior grid of electrodes over the left lateral convexity. Electrode numbers are shown for the posterior grid, and the EEG below is derived from the bottom two rows of electrodes. Electrodes are identified where motor and language responses to electrical stimulation were obtained and where a typical seizure originated. This was a simple partial seizure of forced thought and speech arrest followed by a complex partial seizure with ictal vocalization. The elertrographic seizure recording shows focal onset of low-amplitude fast activity and sharp waves (arrows) followed by sustained highamplitude fast activity restricted to electrodes lA, IB, lC, 2A, 2B, and 2C. Note that lC and 2C are also over Broca's area. Left frontal resection included most of the epileptogenic zone and extended up to Broca's area. Calibrations: 1 second and 100 /-lV. (Adapted from Sutherling WW, Risinger MW, Crandall PH et al: Focal functional anatomy of dorsolateral frontocentral seizures. Neurology, 40:87, 1990, with permission.)

Long-Term Monitoring for Epilepsy

Presurgical Evaluation The most exacting evaluation in LTME is the attempt to localize, by means of extracranial and/or intracranial electrodes, a region of epileptogenic brain tissue that is the site of origin of recurrent seizures and that is amenable to surgical removal. The following are minimum acceptable standards. EEG Transmission. Standard cable ("hard wire") or EEG telemetry with at least 16 channels of EEG data. Cable telemetry is the most common methodology. Ambulatory EEG is not acceptable for final evaluation but may serve a useful triage function. EEG Recording and Storage. Continuous paper printout with time code is no longer the standard of care. Continuous or computer-assisted selective tape (analog instrumentation, video cassette, or digital) or disk (digital) recording of EEG with a synchronized time code and subsequent selective playback on paper or other high-quality display is recommended. EEG Review and Analysis. Detailed visual analysis of all seizures and representative interictal abnormalities from a paper printout or other high-quality display is adequate. Repeat analyses of seizures recorded on tape or disk with variations in playback parameters are preferable. Additional computer analyses of EEG abnormalities (temporal and distribution characteristics) may be beneficial. Behavior Monitoring. Continuous video recording synchronized to EEG. Observer- or self-reported behavior is not sufficient. Time-lapse video recording is discouraged.

147

(preferably 16 or more) channels of EEG monitoring are needed for detailed analyses. The classification of epileptic seizures usually requires video documentation of the behavioral features, synchronized with EEG data.

Quantification of Eledrographic Abnormalities Four-channel ambulatory EEG is adequate for quantifying the frequency of occurrence of recognizable electrographic interictal or ictal events. In order to differentiate clinical seizures from subclinical ones, however, video monitoring is essential. The identification of the relationship of epileptic events to specific precipitating factors may require video recording or ambulatory capability, depending on features of the provocative stimuli or circumstances.

ACKNOWLEDCMENTS Drs. James Burchfiel, John Ebersole, John Gates, Jean Gotman, Richard Homan, John Ives, Donald King, Jeffrey Lieb, Susumo Sato, and Robert Wilkus participated in the preparation of the "Guidelines for LongTerm Monitoring for Epilepsy," from which large parts of this chapter were adapted. Original research reported by the author was supported in part by Grants NS-02808, NS-15654, and GM-24839 from the National Institutes of Health, and Contract DE-AC03-76-SFOOOI2 from the Department of Energy.

Diagnosis of Nonepileptic Seizures REFERENCES Minimum standards of practice in the differentiation of nonepileptic seizures from epileptic seizures are the same as discussed earlier, although 8 channels of EEG data are adequate for identifying most epileptic events. Regardless of the number of channels, however, the absence of clear ictal EEG abnormalities during a behavioral event must be interpreted with reference to the complete clinical evaluation before a diagnosis of non epileptic seizures can be made.

Classification and Charaderization of Epileptic Events Although 4-channel ambulatory EEG is adequate for documentation of certain interictal and ictal electrographic abnormalities, only systems with 8 or more channels can provide basic characterization of epileptic EEG events. Classification and characterization of epileptiform EEG features are enhanced by an increased number of EEG channels, and at least 12

1. Engel J Jr, Burchfiel J, Ebersole J et al: Long-term monitoring for epilepsy: report of an IFCN committee. Electroencephalogr Clin Neurophysiol, 87:437, 1993 2. American Electroencephalographic Society: Guideline 12: guidelines for long-term monitoring for epilepsy. J Clin Neurophysiol, 11:111, 1994 3. Gumnit RJ (ed): Intensive Neurodiagnostic Monitoring. Raven Press, New York, 1987 4. GotmanJ, IvesJR, Gloor P (eds): Long-term monitoring in epilepsy. Electroencephalogr Clin Neurophysiol Suppl, 37, 1985 5. Commission on Classification and Terminology of the International League Against Epilepsy: Proposal for revised clinical and electroencephalographic classification of epileptic seizures. Epilepsia, 22:489, 1981 6. Commission on Classification and Terminology of the International League Against Epilepsy: Proposal for revised classification of epilepsies and epileptic syndromes. Epilepsia, 30:389, 1989 7. Liiders HO, Comair YG: Epilepsy Surgery. 2nd Ed. Lippincott Williams & Wilkins, Philadelphia, 2001

148

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

8. Crandall PH, Walter RD, Rand RW: Clinical applications of studies of stereotactically implanted electrodes in ternporallobe epilepsy.J Neurosurg, 20:827, 1963 9. Talairach J, Bancaud J, Szikla G et al: Approche nouvelle de la neurochirurgie de l'epilepsie: me thodologie stereotaxique et resultats therapeutiques. Neurochirurgie, 20:Suppl 1,240, 1974 10. Dymond AM, Sweizig JR, Crandall PH et al: Clinical application of an EEG radio-telemetry system. p. 16. In Proceedings, Rocky Mountain Bioengineering Symposium, 1971 11. Burgess RC: Neurophysiologic monitoring devices for infants and chiidren.J Clin Neurophysiol, 7:442, 1990 12. Gotman J, Flanagan D, Zhang Jet al: Automatic seizure detection in a newborn: methods and initial evaluation. Electroencephalogr Clin Neurophysiol, 103:356, 1997 13. Rowan AJ, Gates JR (eds): Non-epileptic Seizures. Butterworth-Heinemann, Boston, 1993 14. Devinsky 0, Kelly K, Porter RJ et al: Clinical and electroencephalographic features of simple partial seizures. Neurology, 38:1347,1988 15. Sperling MR, Pritchard PB III, EngelJJr et al: Prolactin in partial epilepsy: an indicator of limbic seizures. Ann Neurol, 20:716,1986 16. Fahn S, Marsden CD, Van Woert MH: Definition and classification of myoclonus. Adv Neurol, 43:1, 1986 17. Charcot J-M: Lecons sur les maladies du systerne nerveux, recueillies et publiees par Bourneville. Vol 1. Paris, 1886 18. Lesser RP: Psychogenic seizures. p. 273. In Pedley TA, Meldrum BS (eds): Recent Advances in Epilepsy, 2. Churchill Livingstone, Edinburgh, 1985 19. King DW, Gallagher BB, Murvin AI et al: Pseudoseizures: diagnostic evaluation. Neurology, 32:18, 1982 20. American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders. 4th Ed. (DSM-IVTR). American Psychiatric Association, Washington, DC, 2000 21. Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 22. Engel J Jr: Seizures and Epilepsy. FA Davis, Philadelphia, 1989 23. Porter RJ, Penry JK, LacyJR: Diagnostic and therapeutic re-evaluation of patients with intractable epilepsy. Neurology, 27:1006, 1977 24. Liiders HO, Engel J Jr, Munari C: General principles. p. 137. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. 2nd Ed. Raven Press, New York, 1993 25. Wieser HG, Morris H III: Foramen ovale and peg electrodes. p. 1707. In Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 26. Spencer SS, Sperling MR, Shewmon DA: Intracranial electrodes. p. 1719. In Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 27. Engel J Jr, Shewmon DA: Overview: who should be considered a surgical candidate? p. 23. In Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998

28. Engel J Jr, Van Ness P, Rasmussen TB, Ojemann LM: Outcome with respect to epileptic seizures. p. 609. In Engel JJr (ed): Surgical Treatment of the Epilepsies. 2nd Ed. Raven Press, New York, 1993 29. Chugani HT, Shields WD, Shewmon DA et al: Infantile spasms. I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol, 27:406, 1990 30. Gates JR, Wada JA, Reeves A et al: Re-evaluation of corpus callosotomy. p. 637. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. 2nd Ed. Raven Press, New York, 1993 31. Morrell F, Whisler WW, Bleck TP: Multiple subpial transection, a new approach to the surgical treatmen t of focal epilepsy.J Neurosurg, 70:231, 1989 32. Morrell F, Whisler WW, Smith MC: Landau-Kleffner syndrome: treatment with subpial intracortical transection. Brain, 118:1529, 1995 33. Nuwer MR, Comi G, Emerson Ret al: IFCN standards for digital recording of clinical EEG. The International Federation of Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl, 52:11, 1999 34. Sperling MR, EngelJJr: Electroencephalographic recording from the temporal lobes: a comparison of ear, anterior temporal, and nasopharyngeal electrodes. Ann Neurol, 17:510, 1985 35. Sperling MR, Mendius JR, Engel J Jr: Mesial temporal spikes: a simultaneous comparison of sphenoidal, nasopharyngeal, and ear electrodes. Epilepsia, 27:81, 1986 36. Lesser RP, Luders H, Morris HH et al: Commentary: extracranial EEG evaluation. p. 173. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. Raven Press, New York,1987 37. Jayakar P, Lesser RP: Extraoperative methods. p. 1785. In Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 38. Legatt AD, Ebersole JS: Options for long-term monitoring. p. 1001. In EngelJ Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 39. Binnie CD, Mizrahi EM: The epilepsy monitoring unit. p. 1011. In Engel J Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 40. Gottman J, Ives J: Computer-assisted data collection and analysis. p. 1029. In EngelJ Jr, Pedley TA (eds): Epilepsy: A Comprehensive Textbook. Lippincott-Raven, Philadelphia, 1998 41. Ives JR, Mainwaring NR, Schomer DL: A solid-state EEG event recorder for ambulatory monitoring of epileptic patients. Epilepsia, 31:661, 1990 42. Gilliam F, Kuzniecky R, Faught E: Ambulatory EEG monitoring.J Clin Neurophysiol, 16:111, 1999 43. Shewmon DA, Krentler KA: Off-line montage reformatting. Electroencephalogr Clin Neurophysiol, 57:591, 1984 44. Chang BS, Ives JR, Schomer DL et al: Outpatient EEG monitoring in the presurgical evaluation of patients with refractory temporal lobe epilepsy. J Clin Neurophysiol, 19:152,2002

Long-Term Monitoring for Epilepsy

45. Lieb JP, Engel J Jr, Gevins A et al: Surface and deep EEG correlates of surgical outcome in temporal lobe epilepsy. Epilepsia, 22:515, 1981 46. Lieb JP, Engel J Jr, Brown \\J et al: Neuropathological findings following temporal lobectomy related to surface and deep EEG patterns. Epilepsia, 22:539, 1981 47. Velasco AL, Wilson CL, Babb TL et al: Functional and anatomic correlates of two frequently observed temporal lobe seizure-onset patterns. Neural Plasticity, 7:49, 2000

149

48. Sperling MR, O'Connor MJ: Auras and subclinical seizures: characteristics and prognostic significance. Ann Neurol, 28:320, 1990 49. Risinger MW, Engel J Jr, Van Ness PC et al: Ictal localization of temporal lobe seizures with scalp/sphenoidal recordings. Neurology, 39:1288,1989 50. Engel J Jr, Wiebe S, French J et al: Practice parameter: Temporal lobe and localized neocortical resections for epilepsy. Neurology, 60:538, 2003

APPENDIX

5-1 Behavioral Signs and Symptoms Associated with Electrographic Ictal Discharges Recorded from Specific Cerebral Areas*t

SEIZURES ARISING FROM THE TEMPORAL LOBE Simple partial seizures have autonomic signs or symptoms, psychic symptoms, and/or certain sensory symptoms. Seizures of hippocampal-amygdalar (mesiobasal limbic or primary rhinencephalic psychomotor) origin often begin with an indescribable strange sensation, rising epigastric discomfort, or nausea. Other common initial signs and symptoms include fear, panic, and/or marked autonomic phenomena such as borborygmi; belching; pallor; fullness of the face; flushing of the face; arrest of respiration; and pupillary dilatation. Seizures of lateral temporal origin often begin with auditory or visual perceptual hallucinations, illusions, dreamy states, and/or vertiginous symptoms. Language disorders indicate involvement of the language-dominant hemisphere. Gustatory hallucinations may indicate involvement of parietal or rolandic operculum, and olfactory hallucinations may indicate involvement of orbital

*Si!l,"IlS and symptoms may be determined by functions of cerebral structures indicated, or their preferential propagation patterns. The information contained in this appendix was adapted in part from deliberations of the Commission on Classification and Terminology of the International League Against Epilepsy, chaired by Dr. J. Roger, and in part from the experiences of Dr. J. Bancaud and his group of Hopital Sainte-Anne in Paris. This appendix is reproduced with the permission of Drs. Roget and Bancaud, with the understanding that it often reflects an oversimplification of complicated anatomic interactions that are as yet incompletely identified. Although ictal behavioral signs and symptoms are useful for postulating possible anatomic substrates of epileptogenic dysfunction, such clinical information alone never definitively indicates the site of seizure origin. [From Engel] ]1': Seizures and Epilepsy. FA Davis, Philadelphia, 1989, with permission.

frontal cortex. Complex partial seizures often, but not always, begin with motor arrest followed by oroalimentary automatisms. There must be amnesia for the ictal event. Common features include reactive automatisms, duration of more than 1 minute, postictal confusion, and gradual recovery. Secondary generalization occurs occasionally.

SEIZURES ARISING FROM THE FRONTAL LOBE Simple or complex partial seizures often have prominent motor manifestations, can include drop attacks, and may be mistaken for psychogenic seizures. Some seizure types are commonly associated with rapid secondary generalization or status epilepticus. Complex partial seizures often are brief and frequent and have minimal or no postictal confusion, and can be associated with urinary incontinence. Seizures involving supplementary motor cortex may have postural (including fencing postures) or focal tonic motor signs, vocalization, or speech arrest. Seizures involving cingulate cortex may be associated with changes in mood and affect, vegetative signs, and elaborate motor gestural automatisms at onset. Seizures involving orbital frontal cortex may be associated with olfactory hallucinations and illusions; early motor signs, including gestural automatisms; and autonomic signs and symptoms. Dorsolateral frontal lobe involvement may give rise to simple partial seizures with tonic or, less commonly, clonic signs and versive eye and head movements. The signs and symptoms of seizures involving frontal operculum and prerolandic cortex are described in the following with respect to seizures arising from multilobar regions.

150

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

SEIZURES ARISING FROM THE PARIETAL LOBE Simple partial seizures consist of positive somatosensory signs and symptoms, as described for seizures arising from the perirolandic area or an intra-abdominal sensation of sinking, choking, or nausea (particularly with inferior and lateral parietal lobe involvement); rarely pain (either as a superficial burning dysesthesia or as a vague but severe episodic painful sensation); negative somatosensory symptoms, including numbness, feeling as if a body part were absent, or loss of awareness of part of half of the body (asomatognosia, particularly seen with nondominant hemisphere involvement); severe vertigo or disorientation in space (suggesting inferior parietal lobe involvement); receptive or conductive language disturbances (suggesting dominant parietal lobe involvement); rotary or postural movements; and/or visual symptoms as described for seizures arising from the temporo-parieto-occipital junction.

SEIZURES ARISING FROM THE OCCIPITAL LOBE Simple partial seizures usually, but not exclusively, include visual phenomena. Visual symptoms consist of fleeting visual perceptions, which may be either negative (scotoma, hemianopia, amaurosis) or, more commonly, positive (sparks or flashes, phosphenes), originating in the visual field contralateral to occipital cortical involvement, or visual perceptual illusions or hallucinations described for seizures originating at the temporo-parieto-occipital junction. Motor signs include clonic and/or tonic contraversion (or occasionally ipsiversion) of eyes and head, or eyes only (oculoclonic or oculogyric deviation), palpebral jerks, or forced closure of the eyelids. Nonvisual sensory symptoms include sensations of ocular oscillation, whole body oscillation, or headache (including migraine). Ictal discharges may spread to produce seizure manifestations of temporal lobe, parietal lobe, or frontal lobe involvement. There is an occasional tendency for seizures to become secondarily generalized.

SEIZURES ARISING FROM MULTILOBAR REGIONS Some seizure patterns are characteristic of more than one anatomically defined lobe of the brain. Features suggestive of seizures arising from the perirolandic (sensory motor) area can originate in either the precentral (frontal) or the postcentral (parietal) gyrus. Simple

partial seizures with motor signs and/or sensory symptoms involve body parts in proportion to their representation on the precentral and postcentral gyrus. Thus involvement of face, tongue, hand, and arm occur most often. Common signs and symptoms, which may occasionally spread in jacksonian manner, include tonic or clonic movements, tingling, a feeling of electricity, a desire to move a body part, a sensation of a part being moved, and/or loss of muscle tone. Lower perirolandic involvement may be associated with speech arrest, vocalization, or dysphasia; movements of the face on the contralateral side; swallowing; tongue sensations of crawling, stiffness, or coldness; and/or facial sensory phenomena, which can occur bilaterally. Movements and sensory symptoms of the contralateral upper extremities occur with involvement of the middle and upper perirolandic area. Sensory symptoms and/or motor signs of the contralateral lower extremity, welllateralized genital sensations, and/or tonic movements of the ipsilateral foot occur with involvement of the pericentral lobule. Postictal Todd's paralysis and secondary generalization occur commonly with seizures of perirolandic origin. Features suggestive of seizures arising from the opercular (perisylvian, insular) area can originate in the frontal, parietal, or temporal operculum. Mastication, salivation, swallowing, laryngeal symptoms, epigastric sensations with fear, and/or vegetative phenomena are characteristic of opercular involvement. Simple partial seizures, particularly with clonic facial movements, are common. Secondary sensory symptoms include numbness, particularly in the hands. Bilateral movement of the upper extremities may be seen. Features suggestive of seizures arising from the temporo-parieto-occipital junction commonly derive from epileptic discharges involving cortex of more than one lobe. Simple partial seizures often consist of visual perceptual illusions or formed hallucinations. Visual illusions include a change in size (macropsia or micropsia), a change in distance, an inclination of objects in a given plane of space, distortion of objects, or a sudden change of shape (metamorphopsia, more common with nondominant hemisphere involvement). Formed visual hallucinations may include complex visual perceptions including colorful scenes in varying complexity; in some cases, the scene is distorted or made smaller, or in rare instances the subject sees his or her own image (autoscopy). Multimodality formed hallucinations may include auditory and occasionally olfactory or gustatory symptoms, and autonomic signs and symptoms appropriate to the visual perceptions. Vertiginous symptoms also arise from this region. Language deficits suggest dominant hemisphere involvement. Complex partial seizures often ensue, presumably owing to mesial temporal spread.

CHAPTER

Ambulatory Electroencephalographic Monitoring

6

JOHN S. EBERSOLE

EQUIPMENT TECHNOLOGY Electrodes Combined A/EEG and Videotape Recording REVIEW TECHNIQUES Computer-Assisted Spike and Seizure Detection Printout Review Artifact and Normal Transient Recognition

When recording systems for ambulatory electroencephalography (A/EEG) were first introduced, it was quickly recognized that this technology had the potential to fill the diagnostic gap that existed between routine electroencephalography (EEG) and intensive inpatient monitoring in the evaluation of paroxysmal disorders. The brief EEG provided by routine laboratory studies is not well suited to the identification of abnormalities that are infrequent. For this reason epilepsy centers evolved to provide long-term monitoring, as discussed in Chapter 5. Despite widespread acclaim for this form of evaluation, there remained the inherent disadvantages of necessary hospitalization, restricted patient mobility, insufficient availability, and expense. A more convenient, mobile, available, and less expensive means of obtaining long-term EEG data was and continues to be needed.

EQUIPMENT The concept of prolonged monitoring of physiologic data on mobile patients by means of a portable tape recorder was first introduced by Holter in the electrocardiographic (ECG) evaluation of arrhythmias.' This scheme was not immediately applicable to EEG because the early recorders were limited to one channel, the EEG signals required considerable additional amplification, and there was no efficient method for analyzing the data once recorded. When a -l-channel miniature

INDICATIONS Epilepsy Sleep Disorders Syncope and Dizziness Psychiatric Disorders and Pseudoseizures Presurgical Evaluation OVERALL YIELD IN CLINICAL PRACTICE

cassette recorder became available," Ives and Woods showed that recording EEG on it was feasible." Development of a solid-state, on-head preamplifier chip solved the second problem," and the introduction ofa rapid video/audio playback device solved the last." Complete 4-channel ambulatory cassette systems were commercially available by 1979. Four-channel cassette recorders were standard analog devices utilizing four recording heads. Tape speed was reduced to approximately 2 mm per sec so that a standard C120 cassette could record at least 24 hours of continuous EEG. Recorders weighed approximately 1.5 pounds and were easily worn by belt or strap on most patients, including small children. Paginated rapid video playback was the conceptual breakthrough that made efficient analysis of long-term ambulatory cassette tape recordings feasible. The first playback units incorporated a video display of data at selectable page lengths and speeds of replay, plus a simultaneous audio reproduction of one data channel. At the fastest replay speed (usually 60 times real time), 24 hours could be reviewed in 24 minutes. In 1983, a cassette system capable of recording 8 channels of continuous EEG, as well as digital real time and event markers, came on the market.r? A new recording method called "blocked analog" was developed to record 8 channels of physiologic data plus real digital time and events on lI8-inch tape. The size of the early 8-ehannel recorders was only slightly larger than the 4-channel version. The 8-channel playback units provided not only a means of displaying additional

151

152

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

physiologic data, but a number of improved operational features were incorporated into them. These included digital real time as a separate channel, automatic search to a specific time, up to 64 seconds of memory so that approximately 30 seconds before and after the present screen could be viewed without tape movement, gain and filter acljustments without tape movement, alpha-numeric registry of channel gain and filtering status, continuous printout of data as well as epoch printout, and stereo audio reproduction utilizing any number of component channels. In 1988, a computer-based replay system made its debut, and 16-channel continuous EEG recordings became possible by electronically linking two cassette recorders. In 1996, a 17-channel continuous recorder was developed. This was made possible by the introduction of large-capacity removable hard disks, originally designed for use in notebook computers. With this new recorder, 16 channels of EEG and I channel of ECG could be recorded for 24 hours (at a 200-Hz sampling rate) onto a miniature hard drive within the device. Montage reformatting is possible during replay. EEG analysis can be performed by rapid video review; by audio transformation of signals from selected channels (up to 120 x real time); and by offline, computerassisted spike and seizure detection. In addition to onscreen interpretation, sections of EEG containing features of interest may be printed on individual sheets via laser-jet or on continuous paper via thermal printer. Since then, 24- to 32-channel continuous recorders that use flash memory chips for data storage have been introduced. Various schemes have been devised for analyzing the data. Most systems use offline spike and seizure detection software, including neural network detectors that can rank the probability of a detection being genuine. This has the advantage of allowing the reviewer to verify the most likely detections preferentially rather than having to screen all detections. New continuous recording systems offer additional operational features. Originally it was impossible to archive any EEG data other than the entire tape because the complex tape recording scheme did not allow the data to be copied. Digital storage of EEG epochs by the new computer-based replay units has eliminated this problem. Stored with these files may be patient information; review comments; and the final report, if desired. Direct, online patient monitoring is now also possible with some systems. Isolation electronics built into the replay unit allow the ongoing EEG of patients to be displayed on the replay screen as scrolling waveforms. Alternatively, an optional laptop computer has also been configured to serve as a display of ongoing EEG. In either case, the electronic EEG display can be used in lieu of a polygraph to test the qual-

ity of an ambulatory recording. Several systems also have the capability of displaying, editing, and analyzing (either manually or automatically) polysomnographic data, including oximetry. Historically, ambulatory EEG has also developed along another line, namely the discontinuous or epoch recorder. The original concept in this evolution was that more channels could be recorded at a higher sampling rate if only discrete epochs of EEG were recorded rather than continuous data. Ives introduced the first such recorder in 1982. 8 The recorder was like a commercial "Walkman" and used standard tape speed in order to achieve the frequency response necessary to faithfully record 16 multiplexed channels. The recording was done in selectable periodic epochs (e.g., 15 seconds every 10 minutes over 24 hours); the recorder could also be turned on by a pushbutton, which the patient activated when a spell was experienced. An electronic buffer memory allowed recording of EEG before the button push. Approximately 45 minutes of EEG could be recorded on a tape. Both the amplifiers and multiplexing device were incorporated into one small box that was usually worn on the patient's head and secured by a gauze turban. The 16-channel epoch recorder did not use a video playback device; instead, the recorded epochs were transcribed onto paper in real time. Analysis was like that of standard EEG. Many of the deficiencies of intermittent and pushbutton EEG sampling were overcome by linking the epoch recorder to a portable computer." This device monitored the ongoing EEG and used spike and seizure detection programs to identify segments of abnormal EEG that were then stored on its hard disk. Although these computers were portable, they were not truly ambulatory because they required mains power. They were appropriate for use in a setting where the patient moved only a limited distance (e.g., from bed to chair). Presently, the most common discontinuous system records 16 channels of EEG and 2 channels of other physiologic data such as ECG, electromyogram (EMG), and electro-oculogram (EOG). Up to 15 hours of data can be recorded on the attached portable computer. This includes pushbutton actuation (with 2 minutes before and after the push), periodic sampling, and spike and seizure detection. The EEG is recorded in one of three bipolar montages, including a standard "double-banana" montage. Data are routinely printed out on a laser-jet and reviewed like standard EEG or copied to a CD and reviewed digitally on any computer. These systems can also be configured with more polygraphic channels, in lieu of EEG channels, to record polysomnographic data. An enhanced, waist-worn version of this 18-channel recorder has recently been developed that contains

Ambulatory Electroencephalographic Monitoring

sufficient computing power to perform the spike and seizure detection. It is no longer necessary to attach the recorder to a portable computer in order to obtain online EEG analysis. Discontinuous recorders of27- and 32-channels have also been developed; they offer the possibility of remontaging the output using referential reconstruction from bipolar recordings.!" The 27- and 32-channel systems include all of the standard electrodes in the 10-20 system plus two basal temporal electrodes. Built-in, online spike and seizure detection is currently being implemented in these recorders also. Most recently, a digital video recording system with a wide-angle lens has been added to this line of discontinuous recorders. It is synchronized to the recorder to provide online monitoring of behavior just as with inpatient monitoring. It is clear that these two lines of recording technology are converging. As digital storage devices increase in capacity and decrease in size and power consumption, continuous recording of 24 or more hours of 24 to 32 EEG channels has become reality. A similar evolution of central processing units will allow simultaneous online spike and seizure detection within the confines of a small, truly ambulatory recorder. With such a device, both detections and the continuous EEG record will be available at the end of a recording session. Furthermore, the distinction between ambulatory and inpatient long-term scalp EEG monitoring is now, for the most part, only in the location of the recording. No longer is there a major difference in the number of recording channels, the data sampling rate, or how the data are analyzed. Montages or analysis strategies for ambulatory EEG no longer differ from those of traditional long-term monitoring. Inpatient monitoring continues to have the advantage, however, of convenient audio/video recording of behavior and the safety of antiepileptic drug reduction in a hospital environment.

TECHNOLOGY Eledrodes Application of disk electrodes by collodion technique is currently the only method that will ensure stable longterm recordings. Use of techniques involving electrode application with paste, needles, or headgear is not recornrnended.'! For emergency recordings of several hours' duration on nonambulatory inpatients, selfadhering "stick-on" electrodes commonly used for nerve conduction studies or pediatric ECG recordings have been shown to be useful when employed in a subhairline montage,12.13 but these are not secure enough for ambulatory outpatients.

153

Combined A/EEG and Videotape Recording In hospitals, the conveniences of ambulatory EEG recording can be combined with portable video recording to create a mobile intensive monitoring system. Combined ambulatory EEG and video recording of outpatients can similarly be accomplished in physician's offices or at home so that the inconvenience and expense of hospitalization can be avoided. The only technical necessity is that the EEG and video recordings must be synchronized so that temporal correlations can be made. This is most easily accomplished by adding the same time code to each. Both continuous and discontinuous ambulatory recording systems offer this option.

REVIEW TECHNIQUES The relationships of artifacts, confusing normal transients, and epileptiform abnormalities to normal sleepwakefulness cycles provide the basis for a useful protocol for reviewing continuous A/EEG.14 Analysis of A/EEG during active wakefulness should be aimed principally at the identification of seizures (Fig. 6-1). Active wakefulness is filled with eye movement, muscle, and electrode artifacts. Individual epileptiform transients, even if present, are difficult to recognize or differentiate during these periods. The exception may be prominent generalized or bilateral interictal discharges. The rhythmic character and longer duration of a seizure pattern contrasts sufficiently with the irregular background of wakefulness, however, to permit relatively easy detection. Certain artifacts are also rhythmic, but these can be distinguished by their distribution to channels rather than scalp regions and by their usual nonvarying frequency and interrupted nature. Eye movement, muscle, and electrode artifacts diminish during quiet wakefulness and essentially cease during stages of slow-wave sleep. To the contrary, epileptiform abnormalities increase in frequency or may be apparent only during stages I through 3 of sleep. These two opposing relationships make light sleep the most reliable and productive period to identify interictal epileptiform abnormalities. If no interictal abnormalities are apparent during the first two or three sleep stage cycles, the likelihood of detecting any interictal features decreases significantly.

Computer-Assisted Spike and Seizure Detedion Discontinuous and continuous A/EEG recording systems have become increasingly dependent upon spike

154

ElECTRODIAGNOSIS INCLINiCAl NEUROLOGY •

•

..

•

II

..

..

..

•

..

..

•

.... •

----"--"-'---~"-----.

..

..

'---

-

..

......

T5-T3 --.....,.....,.I_~rNfNIt T3-F7 F7-F3 F3-C3 - - - _ - -............... ........_,~ _ 0'"' C4-F4 _ _-...\_ _._.\_ __ ¥ _ •

••

0

".".

. . . . .

__ ,. . --"""""""'---r--""--+

• ...,.. .,."..

~

F4-F8 -~--------~\oIIoWlI

, v...

dlilllII

...,

"".,,~~­

• • , ...

F8-T4 T4-T6 _ _.....-..-._ _~-_.........-• . . . It • ......... sa '" $

B

A ..

...

. . . . . . . . . .. . .

•

-~-~~~~~~~~~~~~-

'''~~.~\ty..:

F3-C3 """ 4 i T5-T3 .-... #..,f'\.wu Ii. . . . . ..

~'JIIWl~:I'

'./lA'"

T3-F7 ~"~W~\~~l~ F7-F3- U -.."y- • ... jJ~~

F4-F8 \ ."u ~ F8-T4 ~~,.."... ,

~~~~ C

=

;1'\\(:;'",'·

••cnl~J,!!,I-IYlJ~rtI.·¥!.,.i(

"'"~u,,,,J(h~~ '~IY"~~\\ • ~.,~tr·l~' '7"1 '.

!\:::::::J:V:n~~X~,~~(:{Y~:: ,

I

",*rl'''V~.' ~I'\IJ

I

..

..

..

..

..

..

..

•

II

•

..

..

..

It

•

.. ... ........., .... ,.. ............. 4.

T5-T3 • •

.... -

..

--~"""'--------~------'-'-~---.----

au .........

,

F7-F3r = f " ~ _ F4-F8

,W.·.J..I~"'M'fJ.J~\Y/~'v-'.r'tlv-

F8-T4,.".,•

~~~~ i.I:'/,'t·.~~ift'¥Vt"'W·it"l'~'-''\~ 0

~~~~

...,..-v-

FIGURE 1-• • A/EEG segments (l6-sec), photographed from an 8-channel video playback unit. A, Generalized rhythmic spikewave discharge. B, Onset of a left temporal partial seizure. C, Onset of a generalized seizure. D, End of a right temporal partial seizure. Time marker: I second.

and seizure detection software, developed originally for long-term monitoring, to screen the massive amount of accumulated EEG data for epileptiform abnormalities. Most systems employ variations of the algorithms developed by Gotman and colleagues.P'<"

Printout Review In many laboratories, EEG technologists may be responsible for a video prescreening of the ambulatory EEG record to define segments for later review and clinical interpretation by the electroencephalographer. At a minimum, pertinent segments ofthe continuous A/EEG should be analyzed by the responsible physician. Considerable additional information is gained by seeing the context in which a particular paroxysm or transient was recorded. Often what precedes or follows a suspicious event clarifies its nature as real or artifact. When clinical interpretation of continuous ambulatory EEG is dependent upon review of printouts, they should contain more than just a single example of the supposed abnormal feature. In ambulatory EEG, even more so than in routine EEG, artifact can mimic anything. Segments of background rhythms; multiple

examples of suspicious events, if present; and ongoing activity before and after any discharge thought to be a seizure should be included to reassure the interpreter of the significance of a supposed abnormality. Typically, data from discontinuous recorders include pushbutton events, spike and seizure detections, and periodic samples (Fig. 6-2). These are printed out with the time of the events. In recordings with few false-positive artifact detections, the time necessary to read the data may be substantially reduced. However, the same caveats apply to the interpretation of epoch printouts; in fact, the disconnected nature of the tracing may make it more difficult to distinguish between the artifactual and epileptiform nature of isolated paroxysms. Given the difficulty in distinguishing artifact from epileptiform abnormalities in an ambulatory recording, more than one example of an interictal abnormality should be identified, and seizure rhythms should show an appropriate progression of frequency and amplitude.

Artlfad and Normal Transient Recognition The differentiation of artifacts and normal physiologic transients from true EEG abnormalities can be very

i.

Gl-TJ

~ ~

lJv."

~ -r

ra-sot 4

sot-sea Sp2-T4

"} , V

.. ~

~

r •.1

,

:iI iiJ: num 'H"

ce

~ J!'''II. JA JA J. J. I.

ecc

V

Fpl F7

'V

I

I

I

I

j

,)

u J! J! JA jJ jl II I. jl]! JI

'0'

I

~~7 - T3

l'3-T5

.-

l

12

Y

I

15

!'"

!'4-T6

16

II?

usia

'I'

/.1 ~ J~.L1.I r ...'lIIVf' 11'1.... I...J.IL T

,

V

,"

'V

OfO~

It,

II.

r f\.

T

.,.

...

I

I'V

13:50:21

jll,{ IT

IT III T

.1.

' ' ' 'I

"no

\"

'1

T

.~

I ,,,

1

13:50:26

13:50:31

nl,li~ 'J'''I "\

I",

1

m-02

sec

~

.11

1

....

-

'l'

~

,,,I.,

,Ll IA •.•.A '1

y

,...

IV

.,,,,

J..

v

....

13 F-iJ2-FB

,

.'

1";

.. 1 .'" U· - "'

'f "

T~-Ol

14

J.', '1"

V

10 11

.:::.~

JA JA JA I. 1.1, I. 11.1.1.1.111'.11 II.L II

r

T

'V

"" "::

A

11"1P'

c-

~·"'r,.

1" '~,

r

I 13:50:36

13:50:41

13:50:46

13:50:51

13:50:56

13:51:01

A

Fpl-F3

F3-C3

C3-P3 3 P3-Ql

5 Fp2-F4

F4-e4

7 C4-P4

8 P4-02

Fp1-F7

10 F7-T3

11 T3-T5

12 T5-01 13 Fp2-FB

14 F8-T4

15 T4-T6

18 T6-02

17 AuxA

18 AuxB

~O()IJ-V 1 sec

00:27:33

B FIGURE 1-2 • A, Printout of an epoch of EEG from a discontinuous recorder, which portrays a seizure originating in the right temporal lobe. Note that EEG data before the pushbutton activation was also stored, which allowed an analysis of the seizure onset. B, Printout of short, 2-second segments of EEG stored in the portable computer attached to a discontinuous recorder. Selections were made by spike and seizure detection software residing in the computer that screens ongoing EEG. The four center segments show right frontocentral spikes. (Courtesy of DigiTrace Care Services.)

156

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

difficult on ambulatory EEG recordings. 14,21,22 By recording during normal activity, an abundance of artifact, principally related to eye movement and muscle tension, is introduced. Quiet wakefulness, as is achieved in the standard EEG laboratory, constitutes only a small portion of the record. In addition, a reviewer faces this plethora of artifacts without benefit of a technologist's observations concerning the patient's behavior. Electroencephalographers or technologists reviewing ambulatory recordings should accordingly familiarize themselves with the common artifacts and transients of active wakefulness and sleep. A conservative interpretation of unusual or equivocal events is always mandatory. Several aids make the task of A/EEG interpretation easier. A diary in which the patient or an observer records the time of suspicious episodes, sleep, and activities that may produce rhythmic artifact (e.g., eating, scratching, teeth brushing, or face washing) is invaluable. Eye movements and other activity artifacts disappear with sleep, Interictal abnormalities are most readily identified during this period because they are both more frequent and less likely to be obscured. Sleep transien ts, however, appear and become the major problem in differentiation. Pediatric sleep complexes may appear very sharp because of their high amplitude. These can usually be identified by a lack of spike-wave morphology and their absence in any but appropriate sleep stages. Vertex waves also may be asymmetric, but seldom are they truly lateralized. Although in teric tal spikes commonly precede a spindle burst, caution is necessary in the interpretation of any sharp transients seen only in association with sleep complexes and not as isolated events.

INDICATIONS Epilepsy A/EEG has been most useful in the evaluation of seizure disorders. 23-26 Many of the early investigative efforts in A/EEG were directed at the generalized epilepsies because the electrographic patterns of the abnormalities were easily recognized even with a reduced number of channels. Furthermore, in the case of absence seizures, A/EEG provided a very convenient way of identifying and quantifying the discharges over a long period. Numerous clinical reports began appearing in the literature soon after ambulatory EEG equipment became available. Most attested to the usefulness of A/EEG in the differential diagnosis of epilepsy,27-35 particularly 3-Hz spike-wave discharges or generalized tonic-clonic ictal episodes.P" Objective measurement of drug efficacy in reducing the frequency of interictal and ictal discharges was also shown to be feasible. 37.38 In

nearly all the above studies, improved diagnostic yield over routine EEG recordings was reported. Ives and Woods and later Ebersole and colleagues demonstrated with only 4 EEG channels the feasibility of cassette monitoring for the lateralized ictal discharges of partial seizures. 39-42 The greatest advantage of intensive inpatient monitoring over ambulatory EEG was not its electroencephalographic superiority but the additional information obtained through video monitoring of behavior and the ability to withdraw antiepileptic medications under medical observation. The evaluation of partial epilepsies benefited from the introduction of 8-channel A/EEG systems. Compared with simultaneous 16-channel cable telemetry records (the standard of the day), both 3- and 8-channel ambulatory EEG reviews correctly identified 93 percent of the records as either normal or epileptiforrn.t" Lateralization of abnormalities was equally good with either cassette system, but more detailed characterization was achieved with 8-channel ambulatory EEG. Although 100 percent of seizures were detected on both systems, there were more false-positive errors when only three data channels were available. Better ability to differentiate real abnormalities from artifacts was the most significant advantage of 8-channel over 3- to 4-channel ambulatory EEG. A/EEG is particularly useful in pediatric practice because epilepsy is a common neurologic affliction of childhood, and these patients do not tolerate long-term monitoring well.v' The small size of A/EEG recorders and the outpatient setting make this technique well suited to the pediatric patient. The utility of A/EEG in children, particularly those with absence and other generalized epilepsies, was demonstrated early in the development of the technique. 34,45-47 Its usefulness in documenting the nature of "spells" of uncertain etiology has also been shown in children. 48,49 The utility of cassette EEG for seizure detection in neonates has been assessed in several studies.50-54 Bridgers and colleagues documented that extended cassette EEG recording, even when restricted to as few as 3 EEG channels, can result in substantially increased detection of seizures over routine recordlngs/"

Sleep Disorders The capability of recording multiple channels of electrophysiologic data over long periods also lends itself to the evaluation of sleep-associated disturbances. As an alternative to hand scoring, sleep staging is easily accomplished using combined video/audio analysis of ambulatory taped data.55 Computer programs have been developed that produce sleep statistics and hypnograms from data entered by the reader during such rapid tape review.56,57 Automated sleep analysis for

Ambulatory Electroencephalographic Monitoring

cassette tapes has also been introduced and is being evaluated and progressively refined. 58-63 The physiologic parameters other than EEG most commonly used for sleep studies with A/EEG recorders are ECG, EOG, and EMG. The EMG and movement data from extremities can be used to identify nocturnal periodic movements and myoclonus. 64•65 Measures of respiration by means of strain gauges, impedance pneumonography, or thermistors may be added in order to investigate paroxysmal disorders of breathing (e.g., sleep apnea,64.66 sudden infant death syndrome, and neonatal apnea). Monitoring all these sleep parameters in ambulatory recording has continued to evolve, particularly with the newer 16- to 32-channel systems. 67- 72 Fry and colleagues evaluated the 18-ehannel recorder for in-home sleep studies.?" They were able to record 4 channels of EEG; 2 channels of EOG; 2 channels of EMG; and ECG, nasal airflow, respiratory effort, microphone signals of snoring, EMG from the anterior tibial muscles bilaterally, and body movement. Of the 177 recordings, none were lost for technical reasons, and 95 percent of epochs were scorable. In a subgroup of 16 patients who had simultaneous A/EEG and sleep recordings performed in the laboratory, there was very high correlation for the scoring of all parameters.

Syncope and Dizziness A/EEG provides a sufficiently long recording of both ECG and EEG to capture syncopal events, if they are frequent. Several studies have shown that this form of combined monitoring can be useful in clarifying the etiology of the patient's complaint, particularly if it is a cardiac arrhythmia. 74--79 Predominantly cardiac abnormalities should be expected, given the relative infrequency with which epilepsy is uncovered in patients presenting with syncope.f" One review showed that only 1.5 percent of ambulatory recordings from patients complaining of syncope or dizziness contained epileptiform abnormalities.flO.RI Despite concerns that an increased incidence of sudden death among people with seizure disorders may be related to cardiac arrhythmias, a combined A/EEG and ECG study showed no increase in cardiac rhythm disturbances among known epileptics when compared with nonepileptics. 7R,79,82 ECG abnormalities may accompany seizures 79,83; in most instances these consist of ictal tachycardia and, at times, abrupt changes in rate rather than dangerous rhythm disturbances.

Psychiatric Disorders and Pseudoseizures The differential diagnosis of behavioral episodes has also been investigated using A/EEG, particularly

157

documenting those attacks that were seizures rather than of psychiatric origin. 84--88 Observing no electrographic changes on A/EEG during episodes of apparent total loss of consciousness or major motor convulsions can provide support for a diagnosis of pseudoseizures, In the evaluation of pseudoseizures, observation of behavior is essential. A/EEG monitoring can be combined with videotape recording to provide the objective documentation necessary for these differential diagnoses." In the case of psychiatric disorders that have no paroxysmal features, the yield for detecting underlying epilepsy has been very low, and in one investigation was zero. 80 A/EEG monitoring may, however, be useful in identifying and quantitating disturbances in sleep architecture, particularly REM-onset latency, that purportedlyare observed in patients with depression and that resolve with drug treatment.P"

Presurgical Evaluation Detailed EEG characterization and localization of spikes and seizures are essential in a presurgical evaluation. At least 16 and preferably 24 to 32 scalp EEG channels sampled at a standard 200 Hz are required. Offline data manipulations such as remontaging and filtering, as well as more sophisticated analyses such as voltage mapping or source modeling, are often very useful. Video recording of ictal behavior is also considered mandatory at most epilepsy surgery centers. Fortunately, the newest generation of ambulatory recording systems has all these capabilities. In fact, at several epilepsy centers the same recording equipment is used for both inpatient and outpatient monitoring." In the hospital setting the "ambulatory" recorders are linked to fixed video recording systems, whereas in the outpatient setting they are linked to portable digital video recorders. With responsible patients, sphenoidal electrodes can also be used in home recordings. Withdrawal of antiepileptic medication is perhaps the only part of the normal inpatient scalp EEG evaluation that cannot be performed as an outpatient procedure. For patients with frequent seizures this is not a major consideration, however.

OVERALL YIELD IN CLINICAL PRACTICE The utility of ambulatory EEG in general practice is dependent upon the appropriateness of the question asked and on the likelihood of answering the question, even if appropriate. As would be expected, rates for recording attacks or spells vary dramatically with the clinical frequency of these episodes. A success rate of 77 percent was achieved in patients with one or more

158

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

seizures daily,29 whereas only 16 percent of unselected patients had spells recorded." A capture rate of 50 percent could be attained in patients who had only one attack per week by allowing at least 3 days for monitoring each of them. 92 Similarly, the yield for documenting epileptiform abnormalities is quite variable in the literature and is most likely caused by different patient populations. Reports of continuous 4-channel A/EEG use in unselected patients have suggested a positive yield of evidence to support a diagnosis in the range of 10 to 15 percent. 27,33 Of attacks recorded in several series, the proportion that were identifiable as seizures have ranged from 23 to 73 percent. 28,30 The author previously reviewed experience with 500 consecutive patients, aged between 2 months and 82 years, undergoing 8-channel A/EEG for the first time. 80 Seizures, interictal epileptiform abnormalities, or both were detected in 87 of the tapes, or 17.4 percent. Among these, there was a 64 percent increase in the yield of interictal epileptiform abnormalities and a 2l-fold increase in seizure recording with A/EEG, as compared with laboratory EEGs. These overall yields included patients referred for many reasons. When analyzed by specific clinical problems or by the perception of the physician, definite trends emerged. A particularly high yield (34.9 percent) was found in those patients whose requests included an affirmative statement regarding epilepsy as the diagnosis. Positive yield was 15.3 percent in patients referred with a wide range of episodic alterations of behavior, perception, sensation, or motor function thought possibly to represent seizures. To the contrary, EEG yield was very low in patients with syncope, dizziness, and particularly nonepisodic behavioral alterations without a history consistent with seizure. 80.81 These results underscore the need for reasoned clinical judgment in the application of ambulatory EEG, as in any diagnostic procedure. Morris and colleagues completed in 1994 a study of the clinical usefulness of If-channel discontinuous, but computer-assisted, EEG recording." A total of 344 patients were recorded for an average of 1.4 days using a 16-channel bipolar montage. Pushbutton events plus spike and seizure detections and periodic samples were recorded on a portable computer attached to the recorder. Epileptiform abnormalities (seizures or spikes) were identified in 38 percent of patients (26 percent by computer only) and in 25 percent of patients with previously normal EEGs. Seizures were identified in 12 percent of patients. The higher overall yield in this study compared with previous 8-ehannel studies may be related either to improved detection of EEG abnormalities with the additional channels or to differences in patient population. It is noteworthy that computer assistance is very important in identifying interictal abnormalities with a discontinuous recording system. A full two-thirds of patients had only computer-

detected EEG abnormalities. Morris later examined the manner in which 16-channel discontinuous A/EEG altered patient care in 149 patients.P" Monitoring resulted in subsequent nonneurologic referrals and further evaluation in 40 percent of patients, in additional inpatient video EEG monitoring in 14 percent of patients, in a change in medications in 23 percent of patients, and in no changes of care in only 19 percent of patients. Liporace and colleagues compared A/EEG to sleepdeprived EEG in 46 patients." Improved detection of interictal abnormalities over routine EEG was similar in both. There was 24 percent improvement for sleepdeprived EEG compared with 33 percent for A/EEG; however, A/EEG detected seizures in 15 percent of patients, whereas none were detected during sleepdeprived recordings. In 3 patients the seizures were recognized only by the computer and not by the patient. Tatum and associates used A/EEG to assess the frequency of unreported seizures." They reviewed 552 records from 502 patients with suspected epilepsy; 47, or 8.5 percent of records contained partial seizures. Twenty-nine of the 47 seizures (62 percent) were identified by patient pushbutton, but 18 (38 percent) of seizures were unrecognized by the patient. Foley and colleagues have studied the usefulness of A/EEG in children and adolescents by comparing it with routine EEG.97 They obtained 18-channel discontinuous records from 84 patients. The mean recording time was 1.4 days, and useful results were obtained in 87 percent of patients. In those with known epilepsy, spells were recorded in 73 percent and electrographic seizures in 45 percent. In patients with suspected but undiagnosed epilepsy, spells were recorded in 86 percent, but of these only 17 percent were seizures. Seizure diagnosis and classification were concordant with routine interictal EEG findings in only 19 percent. Olson used A/EEG to discriminate between epileptic and nonepileptic events in children who had at least three events per week. 98 Of the 167 children, 140 or 89 percent had their typical event during the recording, and 76 percent were nonepileptic. He concluded that A/EG was useful in identifying nonepileptic events on an outpatient basis.

REFERENCES 1. Holter NJ: New method for heart studies. Science. 134:1214, 1961 2. Marson GB, McKinnon]8: A miniature tape recorder for many applications. Control and Instrumentation, 4:46. 1972 3. IvesJR, Woods JF: 4-channel 24-hour cassette recorder for long-term EEG monitoring of ambulatory patients. Electroencephalogr Clin Neurophysiol, 39:88, 1975 4. Quy RJ: A miniature preamplifier for ambulatory monitoring of the electroencephalogram. J Physiol, 284:23. 1978

Ambulatory Electroencephalographic Monitoring 5. Stores G, Hennion T, Quy RJ: EEG ambulatory monitoring system with visual play-back display. p. 89. In Wada.JA, Penry .JK (eds): Advances in Epileptology. The 10th Epilepsy International Symposium. Raven Press, New \brk, 1980 6. Ebersole jS, Bridgers SL: Performance evaluation of an Schannel ambulatory cassette EEG system. p. 476. 15th Epilepsy International Symposium Abstracts, Washington, 198c~

7. Sarns MW: Recording and playback instrumentation for ambulatory monitoring. p. 13. In Ebersole .IS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 8, Ives jR: A completely ambulatory ISchannel cassette recording system. p. 205. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium, Bonn, May 1982. Fischer, New York, 1982 9. Ives JR: Evolution of ambulatory cassette EEG. p. 1. In Ebersole .IS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 10. lves JR, Ichihashi K, Gruber L.J et al: New topographic mapping of temporal lobe seizures. Epilepsia, 34:890, 1993 11. Clenney SL: Techniques of cassette EEG recording. p. 27. In EbersoleJS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 12. Bridgers SL, Ebersole JS: EEG outside the hairline: detection of epileptiform abnormalities. Neurology, 38:146, 1988 l:t Ebersole jS, Bridgers SL: Cassette EEG monitoring in the emergency room and intensive care unit. p. 231. In Ebersole .IS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 14. Ebersole JS: Audio-video analysis of cassette EEG. p. 69. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 1:1. Gotman ], Gloor P: Automatic recognition and quantification ofinterictal epileptic activity in the human scalp EEG. Electroencephalogr Clin Neurophysiol, 41:513, 1976 Hi. Gotman .I, Ives JR, Gloor P: Automatic recognition of inter-ictal epileptic activity in prolonged EEG recordings. Electroencephalogr Clin Neurophysiol, 46: 510, 1979 17. Gotman J: Automatic recognition of epileptic seizures in the EEG. Electroencephalogr Clin Neurophysiol, 54:530, .1982 18. Gotman J: Automatic seizure detection: improvements and evaluation. Electroencephalogr Clin Neurophysiol, 76:317, 1990 19. Koffler OJ, Cotman J: Automatic detection of spike-andwave bursts in ambulatory EEG recordings. Electroencephalogr Clin Neurophysiol, 61:165, 1985 20. GotmanJ: Automated analysis of ambulatory EEG recordings. p. 97. In Ebersole jS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 21. Blumhardt LO, Oozeer R: Problems encountered in the interpretation of ambulatory EEG recordings. p. 37. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 22. Ebersole .IS, Bridgers SL, Silva CG: Differentiation of epileptiform abnormalities from normal transients and artifacts on ambulatory cassette EEG. Amj EEG Technol, 23:1I3, 1983

159

23. EbersoleJS: Ambulatory cassette EEG.J Clin Neurophysiol, 2:397• .1985 24. Ebersole JS: Ambulatory EEG: telernetered and cassette recorded. p. .139. In Gumnit R (ed): Intensive Neurodiagnostic Monitoring. Raven Press, New York, 1986 25. Ebersole JS, Bridgers SL: Ambulatory EEG monitoring. p. 111.ln Pedley TA. Meldrum BS (eds): Recent Advances in Epilepsy. Vol 3. Churchill Livingstone, Edinburgh, 1986 26. Ebersole JS: Clinical utility of cassette EEG in adult seizure disorders. p. Ill. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, .1988 27. Green ], Scales D, Nealis] et al: Clinical utility of ambulatory EEG monitoring. Clin Electroencephalogr, 11:173, .1980 28. Callaghan N, McCarthy N: Twenty-four hour EEG monitoring in patients with normal, routine EEG findings. p. 357. In Dam M. Gram L, Penry JK (eds): Advances in Epileptology: 12th Epilepsy International Symposium. Raven Press, New York, 1981 29. Davidson DLW, Fleming AMM, Kettles A: Use of ambulatory EEG monitoring in a neurological service. p. 319. In Darn M, Gram L, Penry JK (eds): Advances in Epileptology: 12th Epilepsy International Symposium. Raven Press, New York, 1981 30. Ives JR, Hausser C, Woods JF et al: Contributions of 4channel cassette EEG monitoring to differential diagnosis of paroxysmal attacks. p. 329. In Dam M, Gram L, Penry JK (eds): Advances in Epileptology: 12th Epilepsy International Symposium. Raven Press, New York, 1981 31. Oxley J, Roberts M, Dana-Haeri ], Trimble M: Evaluation of prolonged 4-channel EEG-taped recordings and serum prolactin levels in the diagnosis of epileptic and nonepileptic seizures. p. 343. In Dam M, Gram L. Penry .JK (eds): Advances in Epileptology: 12th Epilepsy International Symposium. Raven Press, New York, 1981 32. Ramsay RE, Herskowitz A: 24-hour ambulatory EEG: a clinical appraisal. Electroencephalogr Clin Neurophysiol, 5.1:20, 1981 33. Ramsay RE: Clinical usefulness of ambulatory EEG monitoring of the neurological patient. p. 234. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 34. Stores G, Brankin P, Crawford C: Aspects of differential diagnosis using ambulatory EEG monitoring. p. 55. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 35. Zschoske ST, Hunger J, Alexopoulos T: Gain of information using mobile EEG long-term monitoring. p. 19. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May .1982. Fischer, New York, 1982 36. Quy RJ, Fitch P,Willison RG et al: Electroencephalographic monitoring in patients with absence seizures. p. 69. In Wada jA, Penry JK (eds): Advances in Epileptology: The .IOth Epilepsy International Symposium. Raven Press, New York, 1980 37. Milligan N, Richens A: Ambulatory monitoring of the EEG in the assessment of anti-epileptic drugs. p. 224. In Stott FO, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings

160

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 38. Stores G: Patterns of occurrence of seizure discharge. p. 115. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 39. Ives JR, Woods JF: A study of 100 patients with focal epilepsy using a 4-channel ambulatory cassette recorder. p. 383. In Stott FD, Raftery EB, Goulding L (eds): ISAM 1979: Proceedings of the Third International Symposium of Ambulatory Monitoring. Academic Press, London, 1980 40. Ebersole JS, Leroy RF: An evaluation of ambulatory, cassette EEG monitoring: II. Detection of interictal abnormalities. Neurology, 33:8, 1983 41. Ebersole JS, Leroy RF: Evaluation of ambulatory cassette EEG monitoring: III. Diagnostic accuracy compared to intensive inpatient EEG monitoring. Neurology, 33:853, 1983 42. Bridgers SL, Ebersole JS: The clinical utility of ambulatory cassette EEG. Neurology, 35:166, 1985 43. Ebersole JS, Bridgers SL: Direct comparison of 3- and 8channel ambulatory cassette EEG with intensive inpatient monitoring. Neurology, 35:846, 1985 44. Stores G, Bergel N: Clinical utility of cassette EEG in childhood seizure disorders. p. 129. In EbersoleJS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 45. Horwitz SJ, Burgess RC, Kijewski KN: Twenty-four hour, four-channel EEG recording in children using a miniature tape recorder and computer analysis. Am J EEG Technol, 18:133, 1978 46. Hall DMB: Experience with ambulatory monitoring in children. p. 151. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 47. Stores G, Lwin R: Precipitating factors and seizure activity. p. 183. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 48. Bachman DS: 24-hour ambulatory electroencephalographic monitoring in pediatrics. Clin Electroencephalogr, 15:164,1984 49. Aminoff MJ, Goodin DS, Berg BO et al: Ambulatory EEG recordings in epileptic and nonepileptic children. Neurology, 38:558, 1988 50. Eyre J, Crawford C: Prolonged electroencephalographic recording in neonates. p. 143. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 51. Eyre.lA, Oozeer RC, Wilkinson AR: Diagnosis of neonatal seizures by continuous recording and rapid analysis of the electroencephalogram. Arch Dis Child, 58:785, 1983 52. Fenichel GM, Fitzpatrick.JE: Difficulty in clinical identification of neonatal seizures: an EEG monitor study. Electroencephalogr Clin Neurophysiol, 58:33P, 1984 53. Bridgers SL, Ebersole .IS, Ment LR et al: Cassette EEG in the evaluation of neonatal seizures. Arch Neurol, 43:49, 1986

54. Eyre J: Clinical utility of cassette EEG in neonatal seizure disorders. p. 141. In Ebersole.JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 55. Reitman M: Techniques of cassette polysomnography. p. 243. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 56. Erwin CW, Ebersole JS, Marsh GR: Combined auditoryvisual scoring of polysomnographic data at 60 times real time. J Clin Neurophysiol, 4:214, 1987 57. Erwin CW, Ebersole JS: Data reduction of cassetterecorded polysomnographic measures. p. 257. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York,1988 58. Koerner E, Ladurner G, Flooh E et al: Basic criteria for automatic analysis of mobile long-term EEG. p. 227. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 59. Martens WLJ, Declerck AC, Kums GJTM et al: Considerations on a computerized analysis of long-term polygraphic recordings. p. 265. In Stefan H, Burr W (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 60. Crawford C: Evaluation of the Oxford Medilog sleep stager. p. 697. In Palu C, Pessina AC (eds): Proceedings of the Fifth International Symposium on Ambulatory Monitoring. Padua. Cleup, Padua, 1986 61. Erwin CW, Hartwell]W: Sleep staging in ambulatory taperecorded polysomnographic data: what a difference an epoch makes.J Clin Neurophysiol, 4:215,1987 62. Marsh G, Erwin CW: The Oxford sleep stager: assessment ofvariability.J Clin Neurophysiol, 4:291,1987 63. Broughton RJ: Ambulatory sleep-wake monitoring in the hypersomnias. p. 277. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 64. Ancoli-Israel S, Kripke DF, Mason W, Kaplan DJ: Sleep apnea and periodic movements in an aging sample . .I Gerontol, 40:419,1985 65. Radtke RA, Hoelscher TJ, Bragdon AC: Ambulatory evaluation of periodic movements of sleep. p. 317. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 66. Ancoli-Israel S: Ambulatory cassette recording of sleep apnea. p. 299. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 67. Wilkinson RT, Mullaney D: Electroencephalogram recording of sleep in the home. Postgrad MedJ, 52:Suppl 7,92,1976 68. Ancoli-Israel S, Kripke DF, Mason W et al: Comparisons of home sleep recordings and polysomnograms in older adults with sleep disorders. Sleep, 4:283,1981 69. Mounaimne MW, Riley TL: Twenty-four-hour ambulatory recording for diagnosis of narcolepsy. Electroencephalogr Clin Neurophysiol, 53:22P, 1982 70. Riley TL, Peterson H, Mounaimne M: Sleep studies in the subject's home. Electroencephalogr Clin Neurophysiol, 53:37P, 1982 71. Mason ~, Kripke DF, Messin S et al: The application and utilization of an ambulatory recording system for the screening of sleep disorders. Am J EEG Technol, 26:145, 1986

Ambulatory Electroencephalographic Monitoring

72. McCall WV, Edinger JD, Erwin CW: Clinical utility of cassette polysomnography in sleep and sleep-related disorders. p. 267. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 73. Fry JM, DiPhilLipo MA, Curran K et al: Full polysomnography in the home. Sleep, 21:635,1998 74. Lai C-W, Ziegler DK: Syncope problem solved by continuous ambulatory simultaneous EEGjECG recording. Neurology, 31:1152, 1981 75. Hlumhardt LD, Oozeer R: Simultaneous ambulatory monitoring of the EEG and ECG in patients with unexplained transient disturbances of consciousness. p. 171. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 76. Callaghan N, McCarthy N: Ambulatory EEG monitoring in fainting attacks with normal routine and sleep EEG records. p. 61. In Stefan H, Burr W (eds): Mobile LongTerm EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 77. Graf M, Brunner G, Weber H et al: Simultaneous longterm recording of EEG and ECG in "syncope" patients. p. 67. In Stefan H, BurrW (eds): Mobile Long-Term EEG Monitoring: Proceedings of the MLE Symposium. Bonn, May 1982. Fischer, New York, 1982 78. KeiLson MJ, MagriIL JP, Hauser WA et al: Electrocardiographic abnormalities in patients with epilepsy. Epilepsia, 25:645, 1984 79. Keilson MJ, MagriILJP: Simultaneous ambulatory cassette EEGjECG monitoring. p. 171. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 80. Bridgers SL, EbersoleJS: Ambulatory cassette EEG in clinical practice: experience with 500 patients. Neurology, 35:1767,1985 81. Bridgers SL: Evaluation of episodes of altered awareness or behavior. p. 217. In EbersoleJS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 82. Keilson MJ, Hauser WA, MagriILJP et al: ECG abnormalities in patients with epilepsy. Neurology, 37:1624,1987 83. Blurnhardt LD, Smith PEM, Owen L: Electroencephalographic accompaniments of temporal lobe epileptic seizures. Lancet, 1:1051, 1986 84. Stores G: Ambulatory EEG monitoring in neuropsychiatric patients using the Oxford mediLog 4-24 recorder with visual play-back display. p. 399. In Stott FD, Raftery ER, Goulding L (eds): ISAM 1979: Proceedings of the Third International Symposium on Ambulatory Monitoring. Academic Press, London, 1980 85. Stores G: Differential diagnosis of seizures: psychiatric aspects. p. 259. In Dam M, Gram L, Penry JK (eds):

86.

87.

88.

89.

90.

91.

92.

93.

94. 95.

96.

97.

98.

161

Advances in EpiLeptology: 12th Epilepsy International Symposium. Raven Press, New York, 1981 Forrest GC, Crawford C: Ambulatory monitoring and child psychiatry. p. 157. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 Oxley J, Roberts M: The role of prolonged ambulatory monitoring in the diagnosis of nonepileptic fits in a population of patients with epilepsy. p. 195. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 Smith EBO: The value of prolonged EEG monitoring to the clinician in a psychiatric liaison service. p. 162. In Stott FD, Wright SL, Raftery EB et al (eds): ISAM 1981: Proceedings of the Fourth International Symposium on Ambulatory Monitoring. Academic Press, London, 1982 Leroy RF, Rao KK, Voth BJ: Intensive neurodiagnostic monitoring in epilepsy using ambulatory cassette EEG with simultaneous video recording. p. 157. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 Marsh GR, McCall WV: Sleep disturbances in psychiatric disease. p. 331. In Ebersole JS (ed): Ambulatory EEG Monitoring. Raven Press, New York, 1988 Schomer DL, Ives JR, Schachter SC: The role of ambulatory EEG in the evaluation of patients for epilepsy surgery.J Clin Neurophysiol, 16:116, 1999 Powell TE, Harding GFA, Jeavons PM: Ambulatory EEG monitoring: a preliminary follow-up study. p. 1c~ 1. In Ross E, Chadwick D, Crawford R (eds): Epilepsy in Young People. John Wiley & Sons, London, 1987 Morris GL III, GalezowskaJ, Leroy R et al: The results of computer-assisted ambulatory 16-channel EEG. Electroencephalogr Clin Neurophysiol, 91 :229, 1994 Morris GL: The clinical utility of computer-assisted ambulatory 16-channel EEG.J Med Eng Technol, 21:47,1997 Liporace J, Tatum WO, Morris GL et al: Clinical utility of sleep-deprived versus computer-assisted 16-channel EEG in epilepsy patients: a multi-center study. Epilepsy Res, 32:357, 1998 Tatum WO, Winters L, Gieron M et al: Outpatient seizure identification: results of 502 patients using computerassisted ambulatory EEG. J Clin Neurophysiol, 18:14, 2001 Foley CM, Legido A, Miles DK et al: Long-term computerassisted outpatient electroencephalogram monitoring in children and adolescents. J Child Neurol, 15:49, 2000 Olson DM: Success of ambulatory EEG in children . .J Clin Neurophysiol, 18:158,2001

CHAPTER

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

7

WILLIAM J. MARKS, Jr., and KENNETH D. LAXER

INVASIVE TECHNIQUES IN EPILEPSY Epilepsy Surgery and Localization of the Epileptogenic Zone Role and Limitations of the Surface EEG in the Evaluation of Patients with Epilepsy Role of Invasive Recordings in the Evaluation of Patients with Epilepsy DEPTH ELECTROENCEPHALOGRAPHY Definition and Indications Techniques Findings Advantages, Limitations, and Complications SUBDURAL STRIP ELECTRODE RECORDINGS Definition and Indications Techniques Findings Advantages, Limitations, and Complications RECORDING AND STIMULATION STUDIES WITH SUBDURAL GRIDS Definition and Indications Techniques

Invasive recording techniques, in which electrodes are placed on or in the brain to record its activity, are important electrophysiologic tools in the presurgical evaluation and surgical treatment of patients with epilepsy. Similarly, intraoperative microelectrode recordings obtained from deep brain structures are valuable in guiding the surgical treatment of patients with movement disorders, particularly Parkinson's disease. essential tremor, and dystonia.

INVASIVE TECHNIQUES IN EPILEPSY Epilepsy Surgery and Localization of the Epileptogenic Zone The pioneering work in the 1930s of Penfield and colleagues at the Montreal Neurological Institute established the effectiveness of surgical treatment in

Findings Advantages, Limitations, and Complications EPIDURAL RECORDING TECHNIQUES Definition, Indications, Techniques, and Findings Advantages, Limitations, and Complications ELECTROCORTICOGRAPHY Definition and Indications Techniques Findings Advantages, Limitations, and Complications MICROELEGRODE RECORDINGS IN THE SURGICAL TREATMENT OF MOVEMENT DISORDERS Definition and Indications Techniques Findings

Globus Pollidus Thalamus Subthalamic Nucleus Advantages, Limitations, and Complications

controlling seizures in some epileptic patients who do not adequately respond to medical therapy. Despite the wide range of medical treatments for epilepsy currently available, approximately 20 to 25 percent of persons with epilepsy continue to experience seizures.' For these patients with medically refractory epilepsy, surgical removal of the epileptogenic brain tissue is often an effective treatment.' Surgical resection is predicated on the ability to identify the seizure focus or epileptogenic zone. This area of the brain is responsible for the generation of seizures, and its removal or disconnection results in the cessation of seizures. Techniques for identifying the seizure focus are continually evolving and being refined. Although a number of components typically constitute the contemporary presurgical evaluation of patients with medically refractory epilepsy, ictal electrophysiology remains the "gold standard" in this endeavor. When the scalp-recorded EEG fails to provide adequate electrophysiologic localization of the

163

164

ElECTRODIAGNOSIS INClINICAL NEUROLOGY

epileptogenic region, or when it suggests localization that conflicts with the findings from other elements of the preoperative evaluation (e.g., neuroimaging), invasive recordings are necessary to identify clearly the brain region from which seizures arise.

Role and Limitations of the Surface EEG in the Evaluation of Patients with Epilepsy The ability to record the electrophysiologic activity of the brain with the use of surface electrodes applied to the scalp is rather remarkable when one considers the relatively low amplitude of the signal compared with other biologic signals and with environmental noise, the attenuating effects of tissues intervening between brain and scalp, and the distance between surface electrodes and cerebral generators. Without question, the surface EEG is indispensable in the evaluation of epileptic patients. Its utility includes detection of interictal epileptiform activity, strengthening a suspected diagnosis of epilepsy; identification of focal abnormalities, suggesting a focal structural brain lesion as a possible basis for seizures; and documentation of specific epileptiform patterns associated with particular epilepsy syndromes. The ictal EEG (i.e., the EEG recorded during a seizure) is especially important in defining the epileptogenic focus in patients with localization-related (partial) epilepsy. Long-term recordings of the scalp EEG combined with simultaneous video recordings of clinical behaviors during seizures (video-EEG monitoring) are essen tial in the evaluation of patients being considered for surgical treatment. For patients to be good candidates for focal resective surgery (e.g., temporal lobectomy or frontal topectomy), their seizures must be localized electrophysiologically to one discrete brain region. With current EEG technology, electrophysiologic seizure localization can often be accomplished with the scalp-recorded EEG; however, in approximately 30 percent of patients, the surface-recorded ictal EEG is inadequate for this purpose. At times, the surface EEG is simply not helpful in providing information of localizing value, whereas on other occasions it is misleading. For example, the EEG recordings may provide localizing but not lateralizing information, as when simultaneous bitemporal seizure onsets are found (Fig. 7-1). Similarly, the EEG may lateralize the seizure focus to the right or left hemisphere, but may not localize the epileptogenic zone to a particular region. Furthermore, the EEG can be misleading, reflecting activity predominantly in areas to which the seizure has spread and not the area from which it has arisen.

Role of Invasive Recordings in the Evaluation of Patients with Epilepsy In patients with medically refractory epilepsy who are being considered for surgical treatment, invasive recording techniques are used in situations in which the surface EEG (recorded with scalp and sphenoidal electrodes) does not provide adequate localization of the epileptogenic focus or when it provides discordant localization in relation to other studies. These invasive techniques involve the surgical placement of intracranial electrodes in the subdural or epidural spaces or within the brain parenchyma (Fig. 7-2, A to C). No consensus exists among experienced epileptologists concerning the specific indications for one technique rather than another.F Regardless of the technique, the rationale is to place recording electrodes close to brain regions thought to be generating seizures in order to identify the epileptogenic region with certainty. In addition to their role in recording the electrical activity of the brain, arrays of electrodes contained on a grid can be located atop the cortical surface and used for mapping studies. These studies use cortical stimulation to identify areas of functional importance that are to be avoided in the surgical resection. Invasive electrodes, when properly located, provide the obvious advantage of recording seizure activity directly from, or close to, its source. Nevertheless, for practical reasons the extent of brain area sampled by invasive electrodes is limited, and therefore the ability to survey widespread areas of the brain (accomplished well by the scalp-recorded EEG) is sacrificed to varying degrees in invasive recordings. If invasive electrodes are not near the site of seizure origin, the recordings obtained may be misleading. Thus, the type of electrodes used and the areas in which they are placed in a particular patient are of critical importance. The placement of invasive electrodes is guided by the scalp ictal and interictal EEG, as well as by other studies, such as magnetic resonance imaging (MRI) , single photon emission computed tomography (SPECT), and positron emission tomography (PET). Abnormalities identified on high-resolution MRI, especially the features that suggest the structural substrate of the epileptogenic region, provide important information for guiding the placement of invasive electrodes (Fig. 7-3). Other components of the evaluation that may be helpful in directing intracranial electrodes to brain regions likely to yield reliable localizing information include abnormalities identified on neurologic examination, the ictal clinical features (clinical behaviors and signs during the seizure, as discussed in Chapter 5), and abnormalities revealed on functional imaging studies, including PET, SPECT, and magnetic resonance spectroscopy (MRS). These factors must also be considered when

165

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

Fp1-F7

F7-T3

~"AA0..I\

r. MVI\\/.~it~~vv' rJ\ ~vJIf\lv. . V

rJlPJ\M,(1V Iv~)V~VfI ~ A}V~\

.• .... 1\. 'WT'N"'V-'" V V

J

V

V

y

v~~

T3-T5

F8-T4

~j\fVvv\lVI"A

v..._.r'''NV''\fV V"'I/V

T4-T6

T6-02~' 1

sec

Example of nonlateralizing ictal EEG recorded with scalp electrodes. Repetitive, sharply contoured deltafrequency activity at seizure onset is seen simultaneously over both temporal regions. FIGURE 7-)

III

one interprets the recordings provided by invasive electrodes. Because invasive recording techniques, though generally safe, are associated with potential morbidity, they are reserved for patients who desire operative treatment and in whom surgery is considered appropriate. These approaches are necessary when adequate localizing information cannot be obtained by noninvasive methods. Specific intracranial recording techniques are considered in the following sections.

DEPTH ELECTROENCEPHALOGRAPHY

Definition and Indications Depth electrodes allow the direct recording of cerebral activity from the brain parenchyma into which they are implanted. The indications for depth electrode recordings are not universally agreed, but the method is particularly well suited to investigating seizures suspected of arising from deep structures such as the hippocampus, amygdala, and medial frontal lobe. Seizures arising from these areas may be difficult to localize with surface recordings because of the closed electric fields and

attenuation of the activity by the time it arrives, if at all, at superficial electrode sites. Furthermore, rapid transit of seizures to homologous contralateral regions sometimes precludes the ability to lateralize the area of onset of seizures, even with subdural electrode strips. For example, seizures originating from one hippocampus often spread quickly to the opposite hippocampus.f with scalp-sphenoidal or subdural recordings commonly reflecting an apparent synchronous onset of the seizure from both temporal regions. Depth electrodes, because of their unique ability to record seizure activity directly from the hippocampus, are often useful in establishing the hippocampus from which the seizures arise. In addition, the presence of bilateral, independent epileptogenic foci-not always obvious from surface recordings-can often be demonstrated with depth recordings."

Techniques Rigid and flexible electrodes constructed of a variety of metals and containing a variable number of contacts have been used for chronic depth recordings; the recent availability of electrodes constructed of

166

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

FIGURE 7-2 • Various electrodes used for

invasive recordings in the evaluation of epilepsy and the representative brain regions from which they record. A, Depth electrode implanted with the orthogonal approach to record from the medial and lateral portions of the temporal lobe. B, Subdural strip electrode inserted to cover the subtemporal region. C, Subdural grid covering the frontoparietal and superior temporal regions.

nonmagnetic materials has allowed postimplantation imaging with MRI to document the anatomic location of the electrodes. Modern placement of depth electrodes utilizing MRI- or computed tomography (CT)-guided stereotactic techniques allows accurate and safe implantation of the electrodes through cranial burr holes with the patient under local or general anesthesia. The trajectory of electrode implantation depends on the location of the suspected epileptogenic focus and on the customary practices at the center where the study is to be undertaken. Two common depth electrode arrangements are employed to study temporal lobe epilepsy: (1) an orthogonal approach in the horizontal plane, in which three electrodes are implanted on each side through the middle temporal gyrus, with the most distal contacts inserted into the amygdala, anterior hippocampus, and posterior hippocampus''; and (2) a longitudinal approach, in which a single electrode is inserted through an occipital

burr hole on each side and traverses the long axis of the hippocampus.? The orthogonal technique offers the advantage of recording from both the medial and the lateral aspects along the temporal lobe but, compared with the longitudinal approach, requires a greater number of electrodes to do so.

Findings The interpretation of depth recordings, as with all intracranial recordings, differs from that of surface EEG recordings. First, the electrode is in close proximity to the source of activity, and so the amplitude of the signal is relatively high. Various components of the signal (especially high-frequency activity) are therefore more robust than those recorded with surface electrodes. Second, depth electrodes record activity from a spatially restricted field.

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

FIGURE 7·] • Coronal Tl-weighted magnetic resonance image from a patient with medically refractory temporal lobe epilepsy demonstrating left hippocampal atrophy (right side of image).

LT1-LT2

~

LT4-LT5

,j

RT3-RT4

167

The relevance of background abnormalities recorded by depth electrodes is not clear. Slow-wave disturbances and attenuation of background rhythms may be seen. Such findings do not necessarily correspond to the epileptogenic zone and may represent injury potentials secondary to electrode insertion." In addition, normal background activity is commonly quite sharp in configuration and may assume a rhythmic appearance. Interictal epileptiform abnormalities recorded by depth electrodes are typically of steeper slope, higher amplitude, and briefer duration than those recorded with surface electrodes.f The frequency of bitemporal spikes and sharp waves in patients with unilateral ternporallobe epilepsy (defined by ictal depth recordings) is substantially higher in depth recordings than that found in scalp EEG recordings." Ictal patterns recorded at seizure onset with depth electrodes in temporal lobe epilepsy include attenuation of background activity; repetitive, sharp slow waves; rhythmic discharges of beta, alpha, or theta frequency; and irregular slow or sharp waves (Fig. 7-4).10 In temporal lobe seizures, intraparenchymal recordings commonly demonstrate electrographic onset of the ictus earlier than do subdural recordings. Auras (simple partial seizures),

~4

1000 fLV

1 sec

RT5-RT6

~~~~-~~-~~-~~~~-~

-r-v-:

---.A

~--"""",..r---~~-~"'-""~~-

FIGURE 7·4 • Ictal pattern commonly recorded with depth electrodes in medial temporal lobe epilepsy. Rhythmic sharp activity of abrupt onset, phase-reversing at RT2 (the contact within the hippocampus) evolves in frequency before slowing and then terminating. Note that the activityremains confined to the right medial temporal region. LT and RT refer to the left and right temporal depth electrodes, respectively, inserted via a posterior approach; each electrode has six contacts, with contact 1 the most anterior and contact 6 the most posterior. Recordings are formatted into a sequential bipolar montage.

168

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

although typically not associated with ictal patterns on scalp recordings, often correlate with electrocerebral changes that are detected by implanted electrodes. Similarly, subclinical seizures are more commonly detected by depth recordings. I I Focal seizure onsets, in which electrographic changes begin at one or two adjacent contacts of the depth electrode, most commonly are recorded from the hippocampus, but regional onsets may also be seen. Less often, seizures originate in a focal manner from the amygdala. In the absence of a lesion, seizures rarely are seen to arise from the lateral temporal neocortex.F

Advantages, Limitations, and Complications Depth electrode recordings from brain parenchyma provide a sensitive means of recording, with negligible artifact, activity occurring in a limited volume of brain in the vicinity of the electrode. Although the sensitivity and clarity of the signal is heightened with depth recordings, the constricted "field of view" inherent to all intracranial techniques results in the risk that the recorded activity originated from afar and propagated to a brain region sampled by the implanted electrode. Thus interpretation of depth recordings must be performed within the larger context of data provided from other aspects of the evaluation. Morbidity complicating the use of depth electrodes includes hemorrhage and infection. The risk of hemorrhage is estimated to be less than 2 percent but may result in permanent neurologic sequelae or even death." Most infections respond to antibiotic treatment.

SUBDURAL STRIP ELECTRODE RECORDINGS

Definition and Indications Subdural strip electrodes consist of a matrix of flexible material in which electrode contacts, arranged in a single row at fixed interelectrode distances (typically 10 mm), are embedded. Each strip is inserted under the dura through a cranial burr hole and directed to overlie the cortical surface of interest. This method is most commonly used to ascertain the localization and lateralization of an epileptogenic focus in a patient whose surface recordings fail to provide this information. Although subdural electrode recordings have been used in the evaluation of the epileptic brain since the 1930s, the technique gained more routine use in the United States in the early 1980s, when reports indicated that the relatively less invasive subdural approach often supplanted the need for the more invasive depth electrode techniques.J!

Placement of multiple subdural electrode strips is typically undertaken to survey several cortical regions bilaterally. For example, in patients suspected of having a temporal lobe focus, subdural strips can be inserted through a temporal burr hole and directed laterally and subtemporally on either side to record from neocortical and mediobasal structures, respectively. Additional electrode strips are often placed through a frontal burr hole to record from the dorsolateral or opercular frontal regions; seizures may arise from these areas and spread to temporal regions, with clinical seizure symptomatology and the surface EEG erroneously suggesting temporal lobe onset. Subdural strip electrodes are also useful in the clarification of whether the epileptogenic region is associated with a lesion evident on MRI when scalp recordings are discordant with the imaging findings. For example, the scalp-recorded EEG tracings depicted in Figure 7-5, A are from a patient with medically refractory complex partial seizures whose MRI is shown in Figure 7-5, B. These surface recordings suggested that seizures were arising from the left temporal lobe, whereas the MRI revealed a lesion in the right anterior temporal lobe consistent with a cavernous angioma. Because cavernous angiomas are often implicated in an epileptogenic region, the surface EEG localization was questioned and subdural strip electrodes were therefore implanted in both subtemporal regions. Ictal recordings with these electrodes demonstrated that, indeed, the patient's seizures did originate from the area of the lesion (Fig. 7-5, C). Resection of the lesion and the immediately adjacent cortex resulted in cessation of the patient's seizures. The presence of a lesion on MRI does not necessarily imply that it is the site of seizure origination, however, because lesions iden tified on neuroimaging may be incidental and not related to the epileptogenic region. Further complicating the presurgical evaluation, some patients demonstrate multiple areas of abnormality on imaging (e.g., multiple areas of posttraumatic encephalomalacia, vascular malformations, and areas of cortical malformation); one, some, or none of these areas may be epileptogenic.

Techniques Strip electrodes are available in a variety of lengths and widths and with various numbers of contacts and interelectrode spacings. Six-contact strips are often used routinely, although longer or shorter arrays may be employed depending on the locations in which they are to be placed. Strip electrodes are surgically implanted through cranial burr holes, usually with the patient under general anesthesia; multiple strips, directed at different regions, may be inserted through a single

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

burr hole or through multiple burr holes bilaterally. The electrodes are typically placed under fluoroscopic guidance to ensure delivery to the region of interest. Commonly, burr holes are made superior to the zygoma and slightly anterior to the ear for the placement of electrodes in the subtemporal and lateral temporal regions. Strips also can be directed posteriorly to record from the occipital regions, if indicated. Frontal burr holes, located 4 to 6 ern anterior to the coronal suture and slightly lateral to the midline, allow subdural placement of strips along the interhemispheric fissure, to record from medial frontal areas; and across the lateral frontal region, to record from dorsolateral frontal and orbitofrontal regions. With the use of methods similar to those for depth electrodes, the cables from each electrode exit through stab wounds separate from those associated with the burr hole; such an arrangement helps to anchor each strip and reduces cerebrospinal fluid leaks and infection. Skull x-rays, CT, or MRI allows verification of electrode location in relation to the anatomic target. Both referential and bipolar montages are employed to interpret subdural strip recordings. A scalp electrode can serve as a convenient reference; because of the voltage difference between the scalp and the brain surface, the scalp reference is rarely active but can introduce electromyographic artifact into the recording.

Findings The interpretation of subdural strip recordings, as with depth electrode recordings, must take into consideration the somewhat restricted sampling of brain activity provided by the technique. Recording from homologous contacts of bilateral strips assists in the assessment of background activity, aids in differentiating localized or lateralized changes in the EEG (or more precisely, the electrocorticogram [ECoG]), and enables a better understanding of seizure propagation patterns. A~ with depth recordings, localized disturbances of background rhythms may be seen and do not necessarily correspond to the epileptogenic region. Interictal epileptiform discharges are detected more commonly with subdural than with surface recordings, whereas the frequencies of spikes recorded with subdural and depth recordings are similar. 15.16 Ictal patterns consist of a distinct alteration of background activity. Similar to depth recordings, electrographic seizure onsets recorded with subdural strips may assume several morphologies, including the abrupt onset of rhythmic spike or sharp-wave activity, low-voltage fast activity, or suppression of background activity (see Fig. 7-5, C). Several series comparing subdural and depth electrode recordings in localization-related (mainly temporal lobe) epilepsy have been published.2.17-2o These reports generally conclude that inferior temporal sub-

169

dural strip electrodes are somewhat less sensitive than depth electrodes (80 percent compared with 100 percent, respectively) in localizing seizures of medial temporal lobe onset. Importantly, though, the 20 percent of seizures not localized by subdural recordings were not falsely localized: Their recordings simply were not localized at all. Thus, a case can be made for subdural recordings as the next phase in evaluating patients with suspected temporal lobe epilepsy where scalp recordings are nonlocalizing. Such recordings are reasonably sensitive, can cover a larger region of brain than depth electrodes, are typically not misleading, and are less invasive and therefore less likely to cause morbidity or mortality compared with depth electrodes. For the minority of patients with temporal lobe epilepsy in whom subdural recordings are also non localizing, depth recordings may then be undertaken. Many epileptologists employ this graduated approach to invasive studies, especially for patients in whom other studies (e.g., MRI or PET) suggest localization. Subdural recording techniques are valuable in the evaluation of extratemporal lobe epilepsy. Although depth electrodes can be utilized, the large area of cortical surface to be assessed limits their utility. Even in patients with mediofrontal or orbitofrontal epilepsy, subdural electrodes may be superior to depth electrodes in that they cover larger cortical areas, do not need to traverse long distances of brain uninvolved in seizure generation (i.e., white matter), and provide the capacity to record from a greater cortical volume at each contact. In certain clinical situations, the data provided by subdural and depth electrode recordings are complementary, and their contemporaneous use may be considered."

Advantages, Limitations, and Complications The technique of recording with multiple subdural strip electrodes offers the advantage of acquiring ictal recordings directly from the cortical surface at many locations in a minimally invasive and relatively safe manner. Compared with depth electrodes, subdural strips are easier to implant as well, because stereotactic procedures are not required. Like other invasive techniques, subdural strip recordings provide a limited sample of cerebral activity. If the electrodes are not placed near the epileptogenic region, recordings may be nonlocalizing or even misleading, detecting propagated activity rather than activity representing the onset of the ictus. As discussed earlier, even when strips are placed subtemporally along the inferior temporal lobe, the subdural recordings may fail to localize hippocampal seizure onsets. Of all the invasive, extraoperative techniques used for evaluating epilepsy, subdural strip electrodes carry

170

ElECTRODIAGNOSIS INClINICAL NEUROLOGY

T3-T5 T5-01 Fp1-F3 F3-C3

P3-01 ...... ~,.f\n./V'\ Fp2-F4 F4-C4

~/-Jv-v/lr".if1J1'N(.""""-.,..M-r"l,""N~-y,...r......-.I" "......I'~Jv' .,..,jV'-'"V'-""'v'-...."....,.,...,.

P4-02 Fp2-F8 F8-T4 T4-T6 ""'.,."1. N'V' T6-02 A

FIGURE 7-5 • A, Scalp-recorded ictal EEG suggesting onset of seizure from the left temporal region in a patient with medically refractory epilepsy and right temporal lobe lesion. B, Axial T2weighted magnetic resonance image demonstrates the lesion. consistent with a cavernous vascular malformation, in the right temporal lobe (left side of image).

B

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

171

LF1 ('v";Ji'\,"iV/';Vvv'vVV\~V\/V",/jJ'J\)I/'J".JY/JV'v~'''''''-'''''/~~'/\.-J';/''/"""~JV'''-JVV'V'-.!'-"",·V''J''~VfV'.r..;V'./J', LF2 LF3

rv-',~./'vV'JVV"~'V\./V,/'V"-v''/~~~~~.;v--.~·~rv-,r·v/,rv--/,-,v'v'V'..

LF4

~~,rI"-ryr,·v'\~-"'.f"""""":·__"'·':·V

LF5

~~'V-',F.""""""'~""""~"';~_~J"'·~-·",--,--~·J'_'_\lVV-"'N~

......r __ A/'_""__

~

"_JV""'·'A..~J,",,J-~vr~~--,~V"'..__ "',jV"'~

V·_ _

.dytVV~~"""""

..

~,.,....~~·,.'vr,.J,'Ir"f'\..;·"",,,,,,,_J

RF1 ~V'~'~'~!V\/\;',/,/"/'I',/",·~'/"",/>I\v"./\-'\AJ"vj,,/.,_,.i"··'··,~,-·/,V···_V\/ "/'v' /,~.-\/V·\J,./\./--V·~..'./·.,/',././·./\~./"~ RF2

~0-f ...:J~VV~'v-v"\;J'

.....,r_t·J·-.r..-VvA.A,.A/\./'-./' V,/"""'..JV'~~~~

RF3 ~,~..,r/'vJ·"'·V'.J\"",,,,,..),,M;J\"'~.."'ol'.,A."~-'-"'-'~V\_,,,,,,,'_,"""'I'-'.~,,,~~J~,"~~ RF5 ............_~~ ..~~v~~-~~~.......~_'.....vv_-~-"'''''' .._~.......----v'~'_~.;.",...••••-..._-./_,_/'vy..v'v"_·~ RF6

~-"I"J''';'-"'N''-'/~V~-''--'f,~t-/I.'-A.--AAN'''VV''---''-''''''''''''''''./IrI","",,,,,,,,......../ V'.-.

.......

C FIGURE 7-5 • cont'd C, Subdural strip recordings from the left and right frontal (LF, RF) and left and right temporal (IT, RT) regions in the same patient. The ictal recordings demonstrate seizure activity arising from, and remaining confined to, the right temporal region. Repetitive sharp activity is most conspicuous in contacts 2 to 4 of the right temporal electrode in these referential recordings.

the lowest morbidity rate (less than I percent)." Infection accounts for the majority of complications and typically is readily eradicated with antibiotics. Other reported complications, which are rare with subdural strips, include unintended extraction of the electrodes by patients in the peri-ictal period (a potential complication of any invasive recording procedure), subdural empyema, and cortical contusion. 22•23 No mortality has been reported with the technique.

RECORDING AND STIMULATION STUDIES WITH SUBDURAL GRIDS

Definition and Indications Subdural grids are similar to subdural strips except that they contain multiple parallel rows of electrode contacts, usually spaced 10 mm apart, embedded in an inert material. Typical arrays are rectangular or square and contain 16 to 64 contacts, although their arrange-

ment can be customized for each patient by simply cutting the grid to conform to the particular need. The large number of contacts available on subdural grids allows coverage of a relatively widespread cortical area compared with other invasive techniques. Hence subdural grids enable ascertainment of the spatial distribution of epileptiform activity. In addition, grids provide a means of performing chronic extraoperative stimulation studies to identify the localization of various cortical functions in relationship to the epileptogenic zone. Subdural grids are used to localize more precisely the epileptogenic region when other studies have established lateralization but have failed to define with sufficient certainty the regional localization of the seizure focus. Subdural grids are also employed when previous recordings suggest localization of the epileptogenic zone to a cortical region likely to sub serve important neurologic function. Commonly, grids are used to define the relationship of frontal or parietal epileptogenic foci to motor and sensory areas or to establish the

172

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

proximity of frontal or lateral temporal epileptogenic foci to language areas. The ability to identify both epileptogenic cortex and "eloquent" functional cortex is essential in the planning and execution of resective surgery in these regions. Although some functional information can be obtained with the use of intraoperative mapping, the usual lack of ictal recordings in the operating room precludes definite identification of the region from which seizures arise. In addition, from a practical standpoint, young children and some adults do not tolerate intraoperative language mapping, which requires that the patient be awake during the initial portion of the operation. Finally, performing extraoperative recording and stimulation studies over a period of days in an epilepsy monitoring unit allows a more comprehensive evaluation of the epileptogenic region and a more thorough assessment of language function in a setting where the patient is awake, relaxed, and more cooperative.

Techniques Subdural electrode grids are placed through a craniotomy, with scalp and bone flaps tailored both to the area of suspected epileptogenicity and to functional areas that the grid is to cover (Fig. 7-6). An osteoplastic bone flap, in which the bone flap is attached to vascularized muscle flaps, reduces the likelihood of osteomyelitis. Typically, the cortical surface is photographed, the grid is inserted to cover the regions of interest, and additional photographs are made. The intraoperative photographs allow correlation of anatomic landmarks to particular grid contacts, and thus allow the creation of a "map" of the cortical sur-

FIGURE 7·6 • Intraoperative photograph showing placement of a 54--contactsubdural grid following craniotomy.

face that details epileptogenic and functional regions. The grid is sutured to the overlying dura to reduce any tendency to shift. Cerebrospinal fluid leakage is minimized with tunneling and a tight dural closure around the electrode cable. During the course of monitoring with subdural grids, prophylactic antibiotics are administered. For the first 1 to 2 postoperative days, fluid restriction and, sometimes, treatment with steroids are used to reduce the cerebral edema associated with the grid. Contemporary recording techniques involve the use of computerized equipment to amplify, digitize, analyze, and store data. These techniques are particularly valuable for grid recordings, where the capacity to review data (and to manipulate the manner in which it is presented) from a large number of channels is vital. The recordings are obtained in a monopolar fashion, with the reference being an electrode fastened to the skull at the time of craniotomy, a scalp electrode, or an "inactive" electrode on the grid. Cortical stimulation studies, in which electrical current is applied to the cortex through the contacts of the grid, provide valuable information concerning cortical function. Such studies are typically performed during the course of monitoring in the epilepsy monitoring unit. Techniques vary from one center to another, but the general approach is to stimulate sequentially, one contact at a time, the cortical surface underlying the grid to identify motor, sensory, language, and other functional areas. This approach has shown a certain amount of variability from subject to subject in the anatomic localization of sensory, motor, and language functions. 24•25 Mapping is accomplished by briefly passing current through each contact and observing the patient for motor responses; by inquiring about sensory and other phenomena induced by the stimulation; and by testing the patient's language function with the use of a variety of exercises, including counting and picture-naming. Typical stimulation parameters include bipolar 60-Hz square-wave pulses, 0.2 to 2.0 msec in duration. The intensity of stimulation is progressively increased until a clinical response is obtained or an afterdischarge (a repetitive discharge evoked by the stimulation) is recorded, usually at 4 to 12 V with constant-voltage stimulation or 2 to 6 rnA with constantcurrent stimulation. The EEG is monitored during stimulation to detect the presence of afterdischarges. Recognition of afterdischarges is important because the propagation of this activity to cortical areas other than those directly beneath the stimulated electrode contact may provide misleading functional information. Additionally, afterdischarges may evolve to become a clinical seizure. Findings from stimulation studies, along with ictal and interictal data, may then be depicted as a map of the cortical surface covered by the grid. This map allows appreciation of the relationship between

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

functionally important areas and epileptogenic areas. Correlation of the map with intraoperative photographs facilitates planning of the surgical resection. Once a sufficient number of localizing ictal recordings have been obtained and mapping studies are complete, the patient undergoes reopening of the craniotomy under general anesthesia. After the grid position is confirmed, the grid is removed and the resection is undertaken based on the map of epileptogenic and functional areas.

cate regions of potential epileptogenicity or may be geographically distinct from the seizure focus. Ictal recordings, with morphologic features as previously described, provide the primary basis for defining the epileptogenic zone, although no standardized approach to their interpretation exists (Fig. 7-7). Although some authors have found subdural grids rarely to be adequate for seizure localization, others have found the technique generally to be quite useful. 26 .27

Advantages. Limitations. and Complications

Findings Interpretation of cortical recordings made with grids is similar to that with strips (previously discussed). One difference is the absence of bilateral information in grid recordings, and this lack can make the assessment of background, interictal, and ictal patterns challenging. As with other invasive recordings, localized background disturbances may be present in the epileptogenic zone. Interictal epileptiform discharges may serve to demar-

17 .r-"v-'"'--''"""'18

173

......• r-.~"""", _..--.~..--.~

The number and density of contacts contained in a subdural grid allow the localization of epileptogenic and functional cortical regions and an appreciation of the relationship between the two. The anatomic correlation of these localizations is known precisely because grids are placed under direct visualization. The recordings obtained from grids are restricted to the tops of gyri, however, with little contribution from sulcal cortex. Although generally safe and well tolerated by patients,

___

~..,.....,.-------~"'-~............______

19 --~----...---~~ --20 ---~------.,..~~----/'21 _ _.-r--...-- ............._ _ ~

22

--~-..-r----_'"""_~v_"~--r-""--~

23 24 25

..-M"-"-,J.M-Jo....dvvNw~oJvJv-A.AAIJJ'J"~"-A!I,.!'

26 27

~.r"""'....,r'v--.,.,--,-"-_________ ~""---"~------'----,-"",,~-...-.,,,,-..r-../'-'---'---v-

,--",,-~

_.,.r-.~j ..r'"'~'----....

/

~

-~----~--..../

rv--.f'\~'"

,,-----..--J

28

------~~~~

29

~--------

____

~_~

30 ~------......--...........~...,.....---.........- .................,""'f..........,.;....J\-,v_""\/v'JW'.,.."""""wy.."""""''"Y'vNv'w:V'''''''''irvVV'''l'W_/Vw~11< 31 ~'"""''-'''-NV'"-....J\,./V''-W\ll 32 ~~,.r--."'Vv''''''''v-~..,;vN~'''''V'-v-/VvVV'''V\'Wv''' 33

~./"-~~

~..__

----~---------.......-r-----"""-----

_.A_~~v-"V'\./.........r__."-""'_'""'·......""'_.,,.;....r'\J'\\I{

34 __-----..-'""""'..---.,~~_..../-----............-..--.....J'"-~ . . . . . .----~.......,._,.~v--r-~___..".r__"....,..,...vN"'----.rv-r~..._.....Jh..-/I""\

35 36

----~~_~-_..,/"'--"~.r-~--.r'---....,_'""'v....r_-".....,........,.,..,r_A'v.._.!I~'''''"''w-',..rv·-'_'\.rJ\...rrvy,·'\'I

37

--~

~-~'-"'--...---y.

38 ---------.

. . .,.....--..,..r--....

....,_.,.r--.......,_~---.,.,.--~-"v___r--.-""'-"""""---V"'V'.J'I".,.~I'VV'."""

~-......,.....,...,."I'-'-"....,.. ..........._.............rv..r,J.M.."'""'""~-"..,J.,Iy_vrJ'vv'l"'/''\VIrv..,..'II'''''''JV\,I

39 ------_....--_.........,J'--"""-~~J'-ty,/'-'\~~.vf'v.v~"\.Iy.J""""""~~0.rJvI'.rv\~"J-v.'''''''\,)./VvI

40

..,.booo;-~~:;';"-"""'--"'-,-r"'-"-"--""'--~"V""'~'",""vIV''I''V'.,J'-..flfl'lv'-JV''''''''-II\IV'V''''\(\flJ..l''V''''v{·v

0.5 sec

v ....

FIGURE 7-7 II Ictal recording from an 8-contact x 8-contact subdural grid placed in the right frontal region in a boy with medically refractory epilepsy. Referential recording of selected channels (channels 17 to 40) demonstrates rhythmic sharp betafrequency activity beginning at electrode 23, followed by rapid spread to adjacent electrodes (24, 31, and 32).

174

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

subdural grids have the highest potential for morbidity (about 4 percent) of any of the invasive techniques because of the requirement for craniotomy and the size of the grid. 28 Infection is the most common complication but is minimized with meticulous technique, proper maintenance of the craniotomy site and electrode cables, and the use of prophylactic antibiotics. Brain edema can occur but usually responds to fluid restriction and diuresis. Less commonly, subdural hematomas may occur.

EPIDURAL RECORDING TECHNIQUES Definition, Indications, Techniques, and Findings Strip or grid electrodes may be placed on top of the dura through burr holes or following craniotomy (rather than in the subdural space) to record activity from the underlying cerebral cortex. Additionally, socalled epidural peg electrodes may be inserted through small skull holes to enable recording in a similar manner, although this technique seems to have fallen out of favor with many epileptologists. The indications for, technical aspects of, and findings with epidural techniques and subdural techniques are similar. The epidural approach may be preferable when electrodes must be placed in regions involved in a previous resection, because adhesions may prevent access to the subdural space.

Advantages, Limitations, and Complications Compared with subdurally placed electrodes, those placed epidurally may be less prone to inciting infection because the intact dura serves as a protective barrier. By the nature of their location, epidural electrodes are easier to place in patients with subdural adhesions from previous surgery. Unlike subdural strips and grids, however, epidural electrode arrays cannot be placed in interhemispheric, basiotemporal, or orbitofrontal regions. Thus, epidural electrodes are reserved mainly for recording from the lateral convexities. Epidural peg electrodes do provide the capacity to obtain relatively extensive coverage of cortex, and these electrodes may be closely spaced to obtain reasonably highdensity regional recordings (with the requirement of relatively extensive trephining in such a case). In distinction to subdural electrodes, stimulation studies cannot be undertaken with epidural electrodes unless the exposed dura is denervated before electrode placement to avoid discomfort to the patient during electrical stimulation.

ELECTROCORTICOGRAPHY Definition and Indications Electrocorticography (ECoC) consists of the intraoperative recording of electrical activity directly from the exposed cerebral cortex. Penfield andJasper developed the technique in the 1940s, and their monograph on the subject remains a masterful and instructive treatise even today.29 Electrocorticography is used to help delineate the epileptogenic cortical region in more detail than that established by the preoperative, noninvasive evaluation. It is employed, along with electrical stimulation, to identify functional cortical areas, much in the same manner as with grid stimulation studies. Finally, ECoC is performed following the initial resection to assess the completeness of the resection and to identify areas of residual epileptiform activity that may be considered for removal. The circumstances during which ECoC is employed vary among epileptologists. At some institutions, recordings are made during all cases of epilepsy surgery, and the intraoperative findings are used to tailor the surgical resection to each patient. At other centers, ECoC is used selectively (e.g., only in extraternporal procedures), or it may be universally performed for research purposes but the findings are then not used in determining the volume of brain tissue to be resected. When resection of central regions is contemplated, motor mapping (in a manner similar to that previously described) may be employed. Such mapping is performed with the patient under general anesthesia but without the use of paralytic agents. Operations undertaken on the language-dominant hemisphere may be performed with the patient under local anesthesia so that intraoperative mapping of language function may be performed, with the findings from this mapping then used to guide the resection. Other centers perform extraoperative language mapping (using subdural grids, as described earlier) for all resections involving the language-dominant hemisphere in which the resection might encroach on cortical areas suspected of being involved in language function. Some surgeons utilize intraoperative language and motor mapping to guide the resection of tumors and other lesions in patients who do not necessarily have epilepsy, with ECoC used primarily to monitor for afterdischarges and to protect the patient from clinical seizures triggered by electrical stimulation.P

Techniques A variety of electrode arrangements may be used in ECoC. Many centers use a crown, attached to the skull, containing a number of electrodes (often 16) that may

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

be directed in various positions atop the cortex following craniotomy. This crown provides a flexible means of covering the exposed cortex of interest while it maximizes the access to and the visibility of the cortex (Fig. 7-8). Electrode strips, such as those used for extraoperative subdural recordings, may also be used for intraoperative ECoC, but the fixed interelectrode distances reduce versatility. Particularly for temporal lobe resections, crown or strip electrodes are often supplemented with depth electrodes and individual flexible wire electrodes. In such cases the depth electrodes are typically inserted through the middle temporal gyrus to enable the distal contacts to record from medial structures (the amygdala and hippocampus), providing electrophysiologic information from structures often involved in the epileptogenic process but not readily sampled by means of the cortical surface electrodes. Flexible wire electrodes offer the advantage of insertion beneath the dura, allowing recordings from neocortical subternporal regions not otherwise accessed. In extraternporal cases, wire electrodes enable recording from the interhemispheric fissure or from other neocortical areas not directly exposed during the craniotomy. The post of the crown is typically used as the ground, and an electrode clipped to the exposed dura provides the reference. Referential and bipolar montages can be used. Despite the relatively hostile electrical environment of the operating room, high-quality recordings with little artifact can be obtained. Correct selection of filters can limit the interference and amplifier blocking produced during cortical stimulation. The stimulation parameters are similar to those described earlier for grid recordings, although monopolar or bipolar stimulation can be used.

FIGURE 7-8 • Intraoperative photograph demonstrating the placement of the crown electrode used for electrocorticographv (ECoG).

175

A close working relationship between the epileptologist, the anesthesiologist, and the neurosurgeon is mandatory for obtaining high-quality and useful electrophysiologic data. Although practices vary, many epileptologists find that inducing anesthesia with an opiate (e.g., fentanyl) and maintaining general anesthesia with the opiate and an inhalational agent (e.g., nitrous oxide) provides effective anesthesia and permits acquisition of a satisfactory ECoC. For procedures in which it is desirable for the patient to be awake (e.g., language mapping), local analgesia and sedation with opiates and other agents, including droperidol, are used."

Findings The activity recorded during ECoC usually is composed of a mixed-frequency background, often with a considerable amount of fast activity; but the frequency, amplitude, and morphology may vary substantially from case to case. The factors influencing the appearance of this activity include the level of consciousness of the patien t, the particular anesthetic and sedative agents employed during the operation and their concentrations, the region of the cortex from which the activity is recorded, and the nature of the underlying pathologic disturbance. Halogenated inhalational anesthetic agents (e.g., halothane, enflurane, and isoflurane) invariably increase the amount of fast activity present, and this effect often persists for some time after discontinuation of the drug. A profuse amount of background fast activity obscures the recognition of epileptiform discharges, significantly reducing the usefulness of the recording. In general, ECoC from the central, perirolandic cortex demonstrates predominantly fast activity that can be quite rhythmic and sharp in appearance, mimicking epileptiform activity. Recordings from other regions (e.g., the temporal neocortex) contain predominantly intermediate frequency activity with less fast activity. When ECoC is undertaken in cases of brain tumors or other structural lesions, suppression and slowing of the activity recorded in the vicinity of the lesion are commonly seen (Fig. 7-9). In the course of the usual, relatively brief intraoperative ECoC recording, one rarely records a seizure. Thus, in most cases, identification of interictal epileptiform discharges provides the basis for delineating the epileptogenic zone during epilepsy surgery. Epileptiform discharges recorded by ECoG can assume spike, polyspike, and sharp-wave morphologies and can be of varied amplitude (Figs. 7-10 and 7-11). Many authors maintain that the site of interictal epileptiform discharges is related to the site from which seizures themselves arise and, moreover, that removal of the tissue producing interictal epileptiform abnormalities is

176

ElEaROOIAGNOSIS INCLINICAL NEUROLOGY

1-2 2-3 3-4

5-6

9-10

II Intraoperative electrocorticogram from a patient with a left frontoparietal brain tumor. The electrodes were arranged in four rows of four atop the cortical surface, with electrodes 1 to 4 placed along the inferior central region in a posterior-to-anterior fashion and electrodes 5 to 8, 9 to 12, and 13 to 16 placed sequentially more superior. A bipolar montage is depicted and shows irregular slowing and spikes recorded most conspicuously from the inferior frontoparietal region (channels 1 to 3). Note the fast, sharply configured activity present in the bottom half of the recording, which typifies the morphology of activity recorded from the central region.

FIGURE 7-9

/\t\'>J"~rJ~~M'l~~'\~\IVIf'IVJ~~Jvl'vVvv\~ ~~

10-11 ~~"w:~~v"l'lM~~~\I.NW'\\,~~~WII..rvtWlW/"VJIr 11-12 ....V''''''''''''~~~~~M~l'f'/W!i~~~ 13-14 14-15 ~1;,

15-16

~3001LV

......._ - -.....--~ 1 sec

important in achieving seizure control.V Although some series have found that the absence of epileptiform discharges on postresection ECoG predicts a successful postoperative outcome, and the presence of residual discharges is usually associated with incomplete seizure control following surgery, other reports have concluded that ECoG is of little value in guiding the surgical resection.P:" Pharmacologic activation procedures using intravenous administration of methohexital, etomidate, or other medications may enhance the frequency with which epileptiform discharges are recorded (Fig. 7-12, A and B), although more significance is placed on the spontaneously appearing discharges than on those that are induced.

Advantages, Limitations, and Complications Intraoperative ECoG with mapping provides the ability to define epileptogenic and functionally important cortical areas relatively quickly and with more spatial precision than when this information is obtained extraoperatively with the use of implanted, fixed-array elec-

trodes. It allows acquisition of the information, at the time of a single craniotomy, for the resective surgery, and these proximate ECoG findings are then available for the planning of the operation. When intraoperative ECoG provides sufficient information to proceed directly with the resection, it spares the patient the risk (albeit small), discomfort, and cost associated with invasive monitoring. As previously mentioned, ECoG typically provides only interictal information about the epileptogenic zone, and the relevance of the distribution of interictal discharges to the zone of ictus onset and the requisite tissue that must be resected to abolish seizures is not always clear. At times, the intraoperative recordings contain no epileptiform abnormalities and are thus not helpful. In such cases, when better definition of the epileptogenic region in relation to functional areas is required, a subdural grid is usually placed and extraoperative monitoring carried out before resection. Performing ECoG prolongs the duration of the operation and the total time during which the patient receives anesthesia. Thus such an operation theoretically carries the risk of increased morbidity compared with an operation of shorter duration, but this concern seems to be of little actual clinical relevance.

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

1-2

177

~~

2-3 3-4 5--6 6-7 7-8

~

9-10 ~ 10-11 11-12 1~14

14-15 15-16 ~3001J.V 1 sec

AD 1-2 AD 2-3

~

PO 1-2

---J'v

FIGURE 7-10 • Intraoperative electrocorticogram from a patient with right temporal lobe epilepsy. The crown electrodes were arranged in four rows of four atop the cortical surface, with electrodes I to 4 placed along the middle temporal gyrus, electrodes 5 to 8 placed along the superior temporal gyrus, and electrodes 9 to 12 and 13 to 16 placed sequentially more superior in the suprasylvian region. The top 12 channels of the recording represent a bipolar montage from these electrodes. In addition, four-contact depth electrodes were inserted through the middle temporal gyrus anteriorly and posteriorly to record activity from the amygdala and hippocampus, respectively (AD and PO). Spikes are seen to arise from electrodes 2, 3, 6, 7, PD2, and PD3. In addition, spikes are recorded independently from POI, the distal contact of the posterior depth electrode.

PO 2-3 PO 3-4

MICROELEORODE RECORDINGS IN THE SURGICAL TREATMENT OF MOVEMENT DISORDERS Basal ganglia surgery for Parkinson's disease and other movement disorders dates back to the 1940s, after it was fortuitously discovered that lesions of the globus pallidus could improve parkinsonian symptoms. Various techniques, including pallidotomy and thalamotomy, gained popularity until the introduction of more effective pharmacologic therapies, particularly levodopa. As the limitations and adverse effects of these medications became clearer, however, surgical options for the treat-

ment of Parkinson's disease, essential tremor, dystonia, and other movement disorders provoked renewed interest. The development of deep brain stimulation (DBS) therapy heralded a new era in surgical intervention for movement disorders, and this treatment is now in common use. Several technologic advances have contributed to the revitalization of surgical therapy for movement disorders. With improvements in modern functional stereotactic neurosurgical techniques and neuroimaging, identification of the neuroanatomic structures of interest and placement oflesions and brain stimulators in these structures has become more precise. Additionally, advances in

178

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

1-2~~0~'v~~~ 2_~-v"'~

~

5-6 6-7""'v-.."....-..,.rv"-""~""'"""'~-v~r--.J""'.r~"""""'''''''~or-''--./'-..p....,....,r...AM~''\A....w,.""..,..~

11-12 ~~"" A~~

A-J""';-""'" rV--...-v.........WV"'v'v'-J""-.J'o-"h""";""V'''''''.,.r--'''-J''o...,......''/V'''.N'\f'..............

e.--J500J.lV

AD

1II Intraoperative electrocorticogram from a patient with medically refractory left temporal lobe epilepsy. Recording configuration is similar to that in Figure 7-10. Note isolated epileptiform discharges arising synchronously from the amygdala and hippocampal electrodes (ADI and POl).

FIGURE 7-11

10-11 """,, II \"I'j\,NwN"'I_""""~/Y'-h'ww"-,""""./""V'~"I'\J'o"i"I

1_2~~~""A.--""'"'Y\A_r-..rwv.J"".,-......1 s~e/\.c.....-"l"'~~ I

AD 3-4 W{\/\.

"""'~

'-"'v.r"'l\--.-/'"Or'-""...,.".....rA."M-''''"'''~~''Vv

I

PD1-24t~~~/~~,.;V1(~ PD2-3f'"..-J~~~'\,

the understanding of the neurophysiologic underpinnings of Parkinson's disease (and, to a lesser extent, other movement disorders) have provided the scientific rationale for a surgical approach to the treatment of these disabling conditions. The advent of deep brain stimulation therapy provided a nondestructive, adjustable, and reversible treatment with advantages over ablative surgical procedures. Of special interest for the present discussion, neurophysiologic recordings from the basal ganglia and the thalamus have enhanced the precision with which lesions are made or stimulator leads are placed. The following sections address the use of microelectrode recordings to guide implantation of deep brain stimulation leads, pallidotomy, and thalamotomy.

Definition and Indications Microelectrode recordings are obtained with the use of a fine recording electrode, typically having a tip diameter of 15 to 25 11m, through a burr hole or twist-drill

hole in the skull with a trajectory toward a stereotactically defined target. The technique enables the recording of extracellular, single-unit neuronal activity. The electrophysiologic properties of the activity so recorded provide an indication of the location of the electrode in relation to the various gray matter nuclei and white matter tracts encountered along the trajectory. Indeed, the characteristic neurophysiologic signatures of cells within the subthalamic nucleus (STN), globus pallidus internus (GPi) , and ventralis intermedius nucleus of the thalamus (Vim) enable identification of the areas targeted in surgical procedures aimed at providing relief from the symptoms of various movement disorders. Furthermore, the motor subterritories within these target nuclei (i.e., the specific regions in which DBS leads are placed or lesions are generated) can be defined, because activity in such areas is modulated by active or passive patient movement. The role of microelectrode recordings in the surgical treatment of movement disorders has not been fully

Invasive Clinical Neurophysiology in Epilepsv and Movement Disorders

1-2

I\~~

179

~~~ft'tl~~iV';~

2-3 ,"'I~~~~~~f"r'\tII..~ 3-4

~~~r\Ii\~~~~'

4-5

\':I'lIfJi'&A"v'JV'''''v'\,~

6-7

'ii:(-"'-l\~WI"r,Nf.fIr~~,N\!]tAJtJll~tI(~~l\J\~w/11jni.,./~.}~i~rIH~"{~Jr,if }

7-8 \ W#'V'rltt,,~Jj~1,~~~~'N(V1';~f\~'t'VN{~/'\~vr~\~ \

8-9 9-10

"'~'(.J'l~v¥\~J~~~ #~\~~~

A

~ ~500fJ.V 1 sec

1-2

--'---.~~---~

2-3 ..~-'~

3-4 -' 4-5

6-7

'-'---f\)~~vh~~r~~\L-{J~

FIGURE 7-12 • A, Intraoperative

electrocorticogram from the left frontoparietal region in a 9-month-old girl with tuberous sclerosis and intractable seizures. The background consists of mixed-frequency activi ty with a considerable amount of fast activity. No epileptiform discharges are obvious. B, Following the administration of methohexital, background activity is dramatically suppressed and frequent spike and polyspike epileptiform discharges emerge, mainly in electrode 4.

--~--iI/I'-.w"~"i'll~~W---+«--fY___yyvt'__~

-..J''--.~

''''-''''''''-..../'~_~

~

7-8---·'·--V-V-~~~~~

8-9 9-10

-----.....-J~~~~~--""'-..r___~--.. .

r~~

B

established, yet many clinicians believe that the current limitations of image-based targeting require the addition of physiologic information to optimize the accuracy of surgical targeting. Current image-guided stereotaxis relies upon historical (preoperative) imaging, coupled with fiducial markers provided by a rigid head frame affixed to the patient's head or by frameless markers embedded in the skull. These techniques have inherent in them several factors that diminish the precision of targeting, including mechanical inaccuracies, image distortion (for MRI-based techniques), brain shifts that occur after preoperative images have been obtained, and variability in the correlation between physiologic function and anatomy. Whether the employment of microelectrode mapping in surgery for movement disorders results in an

~300!J.V 1 sec

improved success rate, greater efficacy, or altered morbidity remains to be settled. One study found that microelectrode recordings significantly altered the ultimate placement of the stimulator lead (by 2 mm or more) relative to the initial site determined by neuroimagingguided stereotactic techniques in 25 percent of patients. 35

Techniques The neurosurgical procedures involved in microelectrode recordings have been well described in the neurosurgical literature and will only be briefly outlined here." The initial target for the microelectrode is determined with the use of CT- or MRI-guided stereotactic

180

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

techniques. This procedure usually entails attachment of a stereotactic frame to the patient's head, neuroimaging, and calculation of the coordinates for the desired target. Following local analgesia, a small burr hole or twist-drill hole is made in the skull anterior to the coronal suture and lateral to the midsagittal line. The microelectrode, typically constructed of tungsten or platinum-iridium and housed in a protective carrier tube, is inserted under stereotactic guidance and advanced by submillimeter increments with the aid of a microdrive. The neuronal activity is recorded with the use of conventional preamplifiers, AC-De amplifiers, filters, spike triggers, and processors. In addition to an oscilloscope-type visual monitor, an audio monitor provides an effective means of identifying and differentiating the activity recorded from various structures. Microelectrode localization of the target (STN for subthalamic DBS, GPi for pallidal DBS or pallidotomy, or Vim for thalamic DBS or thalamotomy) is based on the characteristic patterns of spontaneous neuronal discharges recorded from the various structures, and on the identification of cells whose activity is modulated by limb movement.

Findings GLOBUS PALLIDUS

Deep brain stimulation of the globus pallidus or pallidotomy for the treatment of Parkinson's disease, dystonia, and various other movement disorders relies on identification of the globus pallidus. Once the microelectrode is advanced to the calculated target for the globus pallidus, recordings enable distinction between globus pallidus externus (GPe) and globus pallidus internus (Gpi); the latter, and particularly its motor-related region, is the desired location in the globus pallidus in which to implant the stimulator lead or place the lesion. In patients with Parkinson's disease, GPe single-cell activity tends to have a lower frequency and a more irregular pattern of firing than does activity recorded from GPi (Fig. 7-13) .36 Two different discharge patterns are characteristic of GPe neurons: (1) a bursting pattern consisting of high-frequency bursts of activity interrupted by pauses; and (2) a low-frequency, irregular pattern.3~ Neuronal activity recorded from GPi, by contrast, is distinctly higher in frequency in the parkinsonian brain; it often assumes a sustained but irregular firing pattern.36-38,40 In patients with dystonia, GPi discharge rates tend to be lower," So-called border cells, located in the laminae surrounding GPe and GPi, are identified by their relatively lower frequency and more regular discharge patterns.36-38 Within both GPe and GPi, cells responsive to active or passive movement, or both, are found. 36•38.40 Although most of these units respond to contralateral limb movement, some respond in a similar manner with

'1 1· H...... rt

H-ttH

Striatum

Dorsal thalamus

STN

SNr

~

GPebursting cell

~

GPepausingcell

~

"Border"cell

_GPi

Motor thalamus "tremor cell" (ET) FIGURE 7-13 • Microelectrode recordings of spontaneous neuronal activity from the thalamus and basal ganglia. Each trace represents a l-second recording in a patient with Parkinson's disease (upper 8 traces) or essential tremor (bottom trace). STN: subthalamic nucleus; SNr: substantia nigra pars reticulata; GPe: globus pallidus externus; GPi: globus pallidus internus; ET: essential tremor. (From Starr P: Technical considerations in movement disorders surgery. p. 269. In Schulder M, Gandhi CD [eds]: Handbook of Stereotactic and Functional Neurosurgery. Marcel Dekker, Inc., New York, 2003, with permission.)

ipsilateral movement. In addition, some cells preferentially respond to one phase of movement more than another (e.g., flexion but not extension, or vice versaj.t" Identification of the optic tract, located inferior to the inferior margin of GPi, is important so that one can avoid placing the stimulator lead or lesion in this structure and thereby causing visual deficits. Such identification is accomplished by detecting light-evoked action potential discharges at the base of the pallidum. 36--:IH Microstimulation, in which 2 to 20 ~ of current passed through the microelectrode evokes visual phenomena in the patient's contralateral visual field, is also used for identifying the optic tract.36-38

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

To avoid motor deficits, the internal capsule is identified. As for any white matter region, recordings from this structure are characterized by a relative absence of neuronal action potentials and, occasionally, the presence of axonal spikes. Microstimulation of the internal capsule, resulting in contraction of the tongue, face, or hand, confirms localization. 36-38 THAUlMUS

For the surgical treatment of essential tremor or other forms of tremor, the target for stimulator lead implantation or lesion placement is the Vim. The trajectory of the microelectrode for this procedure usually enters the caudate nucleus first, where recordings demonstrate a slow rate (up to 10 Hz) of spontaneous cellular discharges.l" As the electrode enters the dorsal thalamus, the recordings are relatively quiescent except for occasional bursts of actlviry.'" On entry of the microelectrode into the ventral aspect of the thalamus, cells responsive to movement may be detected. So-called "tremor cells," cells discharging at the rate of the patient's tremor, may also be encountered (see Fig. 7_13).:}6.42,4:} Placement of the surgical lesion in the area populated by these tremor cells has been shown to correlate with clinical improvement in the patient's tremor. 44.45 The sensory thalamus can be identified, and thus avoided, by recording cells responsive to light cuta-

neous stimuli. Microstimulation, too, enables identification of the sensory thalamus and the internal capsule." SUBTHALAMIC NUCLEUS

Deep brain stimulation of the STN has emerged as a commonly used procedure to treat the motor symptoms of advanced Parkinson's disease when patients experience disabling symptoms despite optimized pharmacologic therapy." The STN, a major output nucleus of the basal ganglia, exhibits excessive and abnormally patterned neuronal activity in the parkinsonian brain.f Stimulating this structure with high-frequency (greater than 100 Hz) electrical pulses by means of a chronically implanted DBS lead can reversibly suppress the major motor symptoms and signs of Parkinson's disease and provide patients with more consistent and higher-quality motor function." The STN is a small nucleus relative to the other structures targeted in surgery for movement disorders, but its characteristic pattern of neurophysiologic activity enables intraoperative identification with relative ease. With a typical microelectrode trajectory, cells in the caudate nucleus of the striatum and then the dorsal thalamus are recorded en route to the STN. Caudate and thalamic neurons have a relatively low rate of discharge (see Fig. 7-13), and individual units are generally easy to isolate." As the microelectrode is advanced further and the STN is approached, back-

II Microelectrode recording of movement-related STN neuronal activity during passive movement of the contralateral upper extremity. The upper trace shows the microelectrode recording and the lower trace shows output of a wrist-mounted accelerometer. Arrows indicate a burst of neuronal activity evoked by limb movement. The trace represents a 4--second recording in a patient with Parkinson's disease. (Courtesy of Dr. Philip A. Starr.)

FIGURE 7-14

181

182

ElECTRODIAGNOSIS IN CLINICAL NEUROLOGY

when single STN units, discharging at rates of 20 to 50 Hz, are isolated. 36 While recording within the STN, particularly within its dorsal aspect, neurons can be identified that respond to active or passive movement of the patient's contralateral arm or leg (Fig. 7_14).48 This movement-induced modulation of activity is most easily perceived as a subtle increase in volume in the audio signal coincident with the movement; this confirms that the electrode is traversing the region of STN subserving motor function-the target for DBS lead implantation. With passage of the microelectrode through the inferior border of STN and into the substantia nigra pars reticulata, the pattern of activity changes abruptly. Higher-frequency activity (50 to 70 Hz), more regular in its pattern and with single units more easy to isolate, appears. In clinical practice, mapping with 1 to 3 microelectrode tracks usually allow confident identification of the relevant borders of STN and its motor subterritory (Fig. 7_15).49.50 From this information, the desired location of DBS lead implantation can be determined.P

Advantages, limitations, and Complications

FIGURE 1·15 • Microelectrode mapping from a case of subthalamic nucleus (STN) DBS surgery, illustrating somatotopy of the motor area. Four microelectrode penetrations were performed in two parasagittal planes. Arm-related cells (filled symbols) are in the anterior, posterior, and lateral aspects of the motor territory, and leg-related cells (open symbols) are in the mediocentral part of the motor territory. Squares represent the most proximal joint (hip, shoulder); circles the middle joint (knee, elbow); and triangles the most distal joint (ankle, wrist) for each extremity. A bold line at the end of a microelectrode trajectory indicates the start of substantia nigra pars reticulata. Microelectrode track reconstructions are superimposed on sagittal plane anatomy drawn from a standard brain atlas, with (A) 11.5 mm lateral to the midline and (B) 13.5 mm lateral to the midline. (From Starr PA, Theodosopoulos PV, Turner R: Surgery of the subthalamic nucleus: use of movement-related neuronal activity for surgical navigation. Neurosurgery, 53:1146, 2003, with permission.)

ground neuronal activity increases. Upon entering the STN, dense, multi-unit recordings are common; the activity is thus seemingly faster than that observed

Stereotactic neurosurgical procedures for the treatment of Parkinson's disease, essential tremor, dystonia, and other movement disorders can diminish the debilitating symptoms of these disorders for many patients. These procedures rely on the precise placement of a stimulator lead or lesion into deep brain structures; thus the exquisite spatial resolution offered by neurophysiologic techniques utilizing microelectrode recording appears to be useful in identifying the target structures of the basal ganglia and thalamus.

REFERENCES 1. NIH Consensus Conference: Surgery for Epilepsy. National Institutes of Health Consensus Conference. JAMA, 264:729,1990 2. Arroyo S, Lesser RP, Awad 1A et al: Subdural and epidural grids and strips. p. 377. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. 2nd Ed. Raven Press, New York, 1993 3. Spencer 58, Williamson PD, Spencer DD et al: Human hippocampal seizure spread studied by depth and subdural recording: the hippocampal commissure. Epilepsia, 28:479, 1987 4. Sirven JI, Malamut BL, Liporace JD et al: Outcome after temporal lobectomy in bilateral temporal lobe epilepsy. Ann Neurol, 42:873, 1997 5. Talairach J, Bancaud J, Bonis A et al: Functional stereotaxic exploration of epilepsy. Cantin Neural, 22:328, 1962

Invasive Clinical Neurophysiology in Epilepsy and Movement Disorders

6. Spencer DO: Depth electrode implantation at Yale University. p. 603. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. Raven Press, New York, 1987 7. EngelJ, Crandall PH: Intensive neurodiagnostic monitoring with intracranial electrodes. p. 85. In Gumnit RJ (ed): Intensive Neurodiagnostic Monitoring (Advances in Neurology, Vol 46). Raven Press, New York, 1987 8. Ajmone-Marsan C: Chronic intracranial recording and electrocorticography, p. 535. In Daly DO, Pedley TA (eds): Current Practice of Clinical Electroencephalography. Raven Press, New York, 1990 9. Hirsch LJ, Spencer SS, Williamson MD et al: Comparison of bitemporal and unitemporal epilepsy defined by depth electroencephalography. Ann Neurol, 30:340,1991 10. Risinger MW, Lee BI, EngelJ Jr et al: Electrographic morphology of ictal recordings from temporal depth electrodes. Neurology, 38:Suppll, 106, 1988 II. Sperling MR, O'Connor MJ: Comparison of depth and subdural electrodes in recording temporal lobe seizures. Neurology, 39:1497, 1989 12. Morris HH, Estes ML, Luders H et al: Electrophysiologic pathologic correlations in patients with complex partial seizures. Arch Neurol, 44:703, 1987 13. Spencer SS, So NK, EngelJJr et al: Depth electrodes. p. 359. In EngelJJr (ed): Surgical Treatment of the Epilepsies. 2nd Ed. Raven Press, New York, 1993 14. Laxer KD, Needleman R, Rosenbaum TJ: Subdural electrodes for seizure focus localization. Epilepsia, 25:651, 1984 15. luders H, Lesser RP, Dinner DS: Chronic intracranial recording and stimulation with subdural electrodes. p. 297. In Engel J Jr (ed): Surgical Treatment of the Epilepsies. Raven Press, New York, 1987 16. Privitera MD, QuinlanJG, Yeh H: Interictal spike detection comparing subdural and depth electrodes during electrocorticography. Electroencephalogr Clin Neurophysiol, 76:379, 1990 17. Sperling MR, O'Connor MJ: Comparison of depth and subdural electrodes in recording temporal lobe seizures. Neurology, 39:1497,1989 18. Spencer SS: Con troversies in epileptology: depth vs. subdural electrode studies for unlocalized epilepsy. J Epilepsy, 2:123,1989 19. Rosenbaum TJ, Laxer KD: Subdural electrode recordings for seizure focus localization. J Epilepsy, 2:129, 1989 20. Barry E. Bergey GK, Wolf AL: Simultaneous subdural gTid and depth electrode recordings of patients with refractory complex partial seizures. Epilepsia, 30:695, 1989 21. Wyler AR, Walker G, Somes G: The morbidity of longte-rm seizure monitoring using subdural strip electrodes. J Neurosurg, 74:734, 1991 22. Rosenbaum TJ, Laxer KD, Vessely M et al: Subdural electrodes for seizure focus localization. Neurosurgery, 19:73, 1986 23. Wyler AR, Ojemann GA, Lettich E et al: Subdural strip electrodes for localizing epileptogenic foci. J Neurosurg, f:iO: 1195, 1984 24. Uematsu S, Lesser R, Fisher R: Individual variations and broad representation of motor and sensory cortex in humans. Epilepsia, 30:643, 1989

183

25. Ojemann GA, Ojemannj, Lettich E et al: Cortical language localization in left dominant hemisphere. J Neurosurg, 71:316,1989 26. Masuoka LK, Spencer SS: Seizure localization using subdural grid electrodes. Epilepsia, 34:Suppl 6, 8, 1993 27. Gates JR, Maxwell RE, Fiol ME et al: Usefulness of subdural grid electrodes in resecting epileptic lesions adjacent to eloquent cortex. Epilepsia, 29:660, 1988 28. Spencer SS, Lamoureux 0: Invasive electroencephalography evaluation for epilepsy surgery. p. 562. In Shorvon SO, Dreifuss F, Fish 0 et al (eds): The Treatment of Epilepsy. Blackwell Science, Oxford, 1996 29. Penfield W, Jasper H: Functional Localization in the Cerebral Cortex. Little, Brown, Boston, 1954 30. Berger MS: Minimalism through intraoperative functional mapping. Clin Neurosurg, 43:324, 1996 31. Smith M: Anaesthesia in epilepsy surgery. p. 794. In Shorvon SD, Dreifuss F, Fish 0 et al (eds): The Treatment of Epilepsy. Blackwell Science, Oxford, 1996 32. Ojemann GA: Intraoperative electrocorticography and functional mapping. p. 189. In Wyler AR, Hermann BP (eds): The Surgical Management of Epilepsy. ButterworthHeinemann, Boston, 1994 33. Bengzon A, Rasmussen T, Gloor P et al: Prognostic factors in the surgical treatment of temporal lobe epileptics. Neurology, 18:717, 1968 34. McBride MC, Binnie CD,Janota I et al: Predictive value of intraoperative electrocorticograms in resective epilepsy surgery. Ann Neurol, 30:526, 1991 35. Starr PA, Christine CW, Theodosopoulos PV et al: Implantation of deep brain stimulators into the subthalamic nucleus: technical approach and magnetic resonance imaging-verified lead locations. J Neurosurg, 97:370, 2002 36. Starr P: Technical considerations in movement disorders surgery. p. 269. In Schulder M, Gandhi CD (eds): Handbook of Stereotactic and Functional Neurosurgery. Marcel Dekker, New York, 2003 37. Vitek JL, Baron M, Bakay RAE et al: Microelectrodeguided pallidotomy: technical approach and application for medically intractable Parkinson's disease.J Neurosurg, 88:1027, 1998 38. Lozano A, Hutchison W, Kiss Z et al: Methods for microelectrode-guided posteroventral pallidotomy.J Neurosurg, 84:194,1996 39. Lenz FA, Dostrovsky JO, Kwan HC et al: Methods for microstimulation and recording of single neurons and evoked potentials in the human central nervous system. J Neurosurg, 68:630, 1988 40. Sterio D, Beric A, Dogali M et al: Neurophysiological properties of pallidal neurons in Parkinson's disease. Ann Neurol, 35:586, 1994 41. VitekJL, Chockkan V, ZhangJYet al: Neuronal activity in the basal ganglia in patients with generalized dystonia and hemiballismus. Ann Neurol, 46:22, 1999 42. Lenz FA, Tasker RR, Kwan HC et al: Single unit analysis of the human ventral thalamic nuclear group: correlation of thalamic "tremor cells" with the 3-6 Hz component of parkinsonian tremor. J Neurosci, 8:754, 1988 43. Benabid AL, Pollak P, Gao D et al: Chronic electrical stimulation of the ventralis intermedius nucleus of the thala-

184

ElEaRODIAGNOSIS IN CLINICAL NEUROLOGY

mus as a treatment of movement disorders. .I Neurosurg, 84:203, 1996 44. Hirai T, Miyazaki M, Nakajima H et al: The correlation between tremor characteristics and the predicted volume of effective lesions in stereotaxic nucleus ventralis intermedius thalamotomy. Brain, 106:1001, 1983 45. Lenz FA, Normand SL, Kwan HC et al: Statistical prediction of the optimal site for thalamotomy in parkinsonian tremor. Mov Disord, 10:318, 1995 46. Limousin P, Krack P,Pollak P et al: Electrical stimulation of the subthalamic nucleus in advanced Parkinson's disease. N Engl] Med, 339:1105, 1998

47. DeLong MR: Primate models of movement disorders of basal ganglia origin. Trends Neurosci, 13:281, 1990 48. Theodosopoulos PV, Marks V\J Jr., Christine C et al: Locations of movement-related cells in the human subthalamic nucleus in Parkinson's disease. Mov Disord, 18:791, 2003 49. Starr PA: Placement of deep brain stimulators into the subthalamic nucleus or globus pallidus internus: technical approach. Stereotact Funct Neurosurg, 79:133, 2002 50. Starr PA, Theodosopoulos PV, Turner R: Surgery of the subthalamic nucleus: use of movement-related neuronal activityfor surgical navigation. Neurosurgery, 53:1146, 2003

CHAPTER

Intraoperative Eledroencephalographic Monitoring During Carotid Endarteredomy and Cardiac Surgery

9

GREGORY D. CASCINO and FRANK W. SHARBRQUGH, III

CAROTID ENDARTERECTOMY

EEG CHANGES RELATED TO CAROTID ARTERY CLAMPING

INTRAOPERATIVE EEG MONITORING: METHODOLOGY

CEREBRAL BLOOD FLOW

EEG AND ANESTHESIA Symmetric EEG Patterns at Subanesthetic or Minimal Anesthetic Concentrations Symmetric EEG Changes with Induction Symmetric EEG Patterns at Sub-MAC Anesthetic Concentrations

EEG AND CAROTID ENDARTERECTOMY EVOKED POTENTIALS AND CAROTID ARTERY SURGERY EEG MONITORING AND CARDIAC SURGERY

EEG CHANGES UNRELATED TO CAROTID ARTERY CLAMPING

The intraoperative recording of electroencephalographic (EEG) activity began soon after the development and introduction of the EEG as a neurodiagnostic technique.l" One of the initial clinical applications of the intraoperative recording of cerebral electrical activity was electrocorticography (ECoG) at the time of focal corticectomy for intractable epilepsy.l" This technique was used before the development of extraoperative EEG monitoring to localize the epileptogenic zone in patien ts with partial seizures.l" ECoG is still used at selected epilepsy centers before and after excision of the epileptic brain tissue. The use of ECoG during epilepsy surgery is addressed Chapter 7 in Other potential indications for intraoperative EEG recordings include the monitoring of cerebral function during carotid endarterectomy and cardiopulmonary bypass surgical procedures.r" The rationale for intraoperative EEG monitoring is the detection of electrophysiologic alterations that are intimately associated with cerebral ischemia or cerebral hypoperfusion before the development of cerebral infarction.e" Extracranial or scalp-recorded EEG monitoring during carotid endarterectomy has been shown to be a reliable indicator of cerebral ischemia.F''" The findings obtained with extracranial EEG monitoring at the time of an endarterectomy may lead to an alteration in operative strategy (i.e., may indicate the need for placement of a

carotid artery shunt) .12 Intraoperative EEG recordings have also been performed during cardiac bypass surgery.r"? The effect of hypothermia, however, significantly limits the potential utility of EEG monitoring during the latter procedure." Profound hypothermia produces a suppression ofEEG activity that reduces the diagnostic yield of intraoperative recordings in identifying alterations resulting from cerebral ischemia.v'" Reversible causes of ischemia during cardiac bypass surgery occur much less commonly than during a carotid endarterectomy for carotid artery stenosis." Intraoperative EEG recordings were the first monitoring technique used during carotid and cardiac surgical procedures to minimize the likelihood of a postoperative neurologic deficit that might significantly affect the individual's quality of life. s.12 The potential adverse effects of cardiovascular and cerebrovascular surgery include stroke, cognitive impairment, and a postoperative encephalopathy. 12 Importantly, additional neurodiagnostic studies may be used in the operating room to monitor cerebral perfusion and the effect of anesthesia. fi - 11,13-15 Transcranial Doppler, carotid ultrasound, and xenon blood flow studies are performed in some centers during carotid endarterectomy and cardiac surgery.6-II,l3-15 The use of transcranial Doppler in selected patients may be predictive of cerebral ischemia in the absence of appropriate EEG

207

208

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

changes." Intraoperative carotid ultrasound is increasingly being used to validate the patency of the internal carotid artery after endarterectomy. II Monitoring the level of anesthesia by the bispectral index may reduce the hemodynamic changes that occur in some patients undergoing cardiopulmonary bypass." This chapter provides an overview of some of the aspects of intraoperative EEG monitoring for the identification of changes related to cerebral ischemia during carotid endarterectomy and cardiac surgery. In particular, the clinical applications and potential limitations of such cerebral function monitoring are considered.

CAROTID ENDARTERECTOMY Carotid endarterectomy is the cerebrovascular surgical procedure most commonly performed to reduce the risk of stroke in patients with symptomatic carotid artery stenosis related to atherosclerosis.6.12-25 Even before the publication of rigorous scientific studies confirming the efficacy of carotid endarterectomy, the use of this operative technique to protect against ipsilateral stroke had become popular.F The increase in the number of operative procedures performed annually in the United States over several years reflects the interest in identifying a protective surgical approach to stroke.P Fewer than 15,000 carotid endarterectomies were performed in 1970, but by 1985 an estimated 100,000 carotid endarterectomies had been performed in the United States. 22 The rationale for this operative procedure is the removal of atherosclerotic thrombotic material that may come to restrict flow and lead to either an occlusion of the carotid artery or an artery-toartery embolus. 21-25 Carotid artery surgery is the direct result of observations in the 1950s that established a relationship between extracranial internal carotid artery disease and stroke. Studies have indicated that carotid endarterectomy is effective in reducing the risk of ipsilateral stroke in patients with high-grade stenosis (70 to 99 percent) of the internal carotid artery.21-25 Selected patients with completed strokes may also be candidates for a carotid endarterectomy, depending on their neurologic deficits and the presence of coexistent medical problems. The North American Symptomatic Carotid Endarterectomy Trial (NASCET) demonstrated a significant reduction in the risk of stroke in patients with carotid artery stenosis greater than 70 percent. 24 Ipsilateral stroke occurred at 24 months' follow-up in 26 percent of 328 nonsurgical patients and in 9 percent of 331 patients undergoing a carotid endarectomy.24 A direct correlation existed between surgical benefit and the degree of carotid artery stenosis. The benefit of carotid artery surgery in the NASCET for patients with 30 to 69 percent stenosis is still undeter-

mined." The Veterans Affairs Symptomatic Stenosis Trial (VASST) also showed a significant reduction in the risk of stroke in patients with carotid artery stenosis of greater than 50 percent who underwent a carotid endarterectomy.P The Asymptomatic Carotid Atherosclerosis Study (ACAS) reported in 1995 that there is an aggregate risk reduction of 53 percent with surgical treatment for asymptomatic carotid artery stenosis.P A total of 1,659 patients were entered into this study.23 All patients had greater than 60 percent carotid artery stenosis and were randomized to surgical or medical therapy." A significant difference was evident between the two treatment groups at 3 years. 23 The risks of carotid endarterectomy must be considered in any discussion of the putative beneficial effects of this operative procedure. The morbidity of this treatment depends on a number of factors including surgical expertise; degree of carotid artery stenosis; the presence of cerebral infarction; and coexistent medical problems, especially ischemic heart disease. The published morbidity of carotid endarterectomy has ranged from 0 to 20 percent.P In one multicenter study, the perioperative morbidity was 2.2 percent and the mortality was 3.3 percent (combined operative complication was 5.5 percent) in "relatively high-risk patients. "25 Importantly, angiography has a reported risk of approximately I percent.F The operative procedure is performed with the patient receiving general anesthesia for both comfort and safety. A combination of nitrous oxide and isoflurane is used at the Mayo Clinic. Induction is routinely performed with thiopental.P The common carotid artery, internal carotid artery, and carotid bifurcation are exposed at the time of surgery. Atherosclerotic changes are most prominent in the proximal internal carotid artery and the carotid bifurcation. Clamping of the internal carotid artery, common carotid artery, and external carotid artery is necessary for an arteriotomy and endarterectomy to be performed.v'< A shunt between the common carotid artery and the internal carotid artery may be placed if the surgical team is concerned that the period of clamping may be associated with cerebral ischemia and perioperative stroke. 5.12 The attitude of neurosurgical teams concerning the use of shunt placement is variable. At the Mayo Clinic, a shunt is placed only ifEEG monitoring of cerebral function or cerebral blood flow studies, or both, suggest a hemodynamic insult that may produce a significant reduction in cerebral blood floW. 5•12 The routine use ofshunts may be associated with an increased risk of cerebral infarction related to artery-to-artery embolus, and it prolongs the duration of the operarion." Importantly, a minority of patients exhibit a significant change in cerebral function monitoring indicative of a hemodynamic insult with a diminished cerebral blood flOW. 5•12 The risk of

Intraoperative Electroencephalographic Monitoring

shunt placement in one study was low, with embolism occurring in 2 of 511 patients (0.4 percent) .14 The introduction of carotid angioplasty and carotid artery sten ting for carotid artery stenosis may be an alternative to carotid endarterectomy.'v-f The methodology for these therapeutic interventions is similar to the techniques introduced for the treatment of cardiovascular disease. Carotid revascularization has also been used in selected patients with restenosis following a carotid endarrerectomy.-" EEG monitoring is not routinely performed during these procedures because the carotid artery is not occluded and there is no consideration of shunt placement. The role for these alternative techniques in the management of asymptomatic or symptomatic carotid artery disease remains under study. 26-28

INTRAOPERATIVE EEG MONITORING: METHODOLOGY Intraoperative EEG monitoring is commonly performed during carotid endarterectomy. There is a diversity of opinion regarding the clinical application of this neurodiagnostic technique in cerebrovascular and cardiac surgery.5.6.12,16 The "hostile" environment of the operating room often makes intraoperative EEG monitoring technically difficult. Potential problems include electrical interference, which produces significant artifacts; difficulty in ensuring the stable application of the electrodes; and the use of anesthesia and pharmacotherapy that may alter the EEG recording. The EEG technologist and electroencephalographer may also have difficulty in examining the patient and in negotiating their way through the operating room because of the surgical team and the necessary equipment. The technical factors that must be considered during intraoperative extracranial EEG monitoring include proper placement of the scalp electrodes (collodion is used at the Mayo Clinic) before anesthesia induction. Usually, 21 to 23 scalp electrodes are used for intraoperative EEG monitoring. 5,12-15 For appropriate intraoperative recordings, at least 8 channels of EEG should be available. In most instances, 16 or 21 channels for EEG monitoring is strongly preferred.4.5.12.13 Proper grounding is essential for patient safety. The use of a 60-Hz filter is required. The linear frequency is set between 1 and 15 or 30 Hz. 5 Sensitivities of 3 to 5 IlV/mm are often necessary for recording the extracranial EEG with the patient under general anesthesia. 5,12-15 A longitudinal bipolar anteroposterior montage is routinely used for intraoperative EEG monitoring.P-V The Laplacian montage may also be useful in identifying a focal alteration. The paper speed during the cerebral function monitoring can be "compressed" from the 30 mm/sec used in routine

209

EEG recordings to 5 mm/sec because of the large amount of data acquired.5.12-15 Subtle changes in amplitude and frequency related to cerebral ischemia can be visualized easily at the slower paper speeds. The diagnostic yield of the reduced paper speed for identifying reversible alterations associated with cerebral ischemia has been demonstrared.S'v" The paper speed can be restored to 30 mm/sec if a continuous or paroxysmal EEG pattern occurs that cannot be iden tified. The use of digital EEG recordings in the operating room has removed concerns regarding paper storage during intraoperative EEG monitoring. The introduction of digital EEG has improved off-line analysis and allowed individuals remote from the operating room to review the EEG as it is being acquired (i.e., "real-time" review). The current practice at the Mayo Clinic is to continue to use a paper speed of 5 mm/sec for digital EEG intraoperative recordings because of the amount of data generated. The technologists at our institution are also more familiar with the effects of cerebral ischemia displayed at a slower paper speed. Computer processing with a compressed spectral array can be used for data interpretation but may have no advantage over visual inspection alone. 5,12.18

EEG AND ANESTHESIA It is necessary to review the relationship between anesthesia and the EEG before considering the effects of cerebral ischemia on the intraoperatively recorded EEG. Anesthetic agents may significantly alter the normal background EEG (Fig. 9-1). To some extent, individual anesthetic drugs may have different effects, depending on their concentrations.5.12.29-31 The EEG patterns produced by different anesthetic agents, when used at concentrations below their MAC level (i.e., the minimal alveolar concentration necessary for preventing movement to a painful stimulus in about 50 percent of subjects), are quite similar. 5. 12,29 Selected anesthetic agents (e.g., thiopental, halothane, enflurane, isoflurane, and nitrous oxide) produce a similar subanesthetic or anesthetic effect associated with a symmetric EEG pattern. 5.12.29-31

Symmetric EEG Patterns at Subanesthetic or Minimal Anesthetic Concentrations Thiopental produces the characteristic drug-induced beta effect at subanesthetic concentrations, which tends to be maximal in the anterior midline distribution. 29 Halothane, enflurane, isoflurane, and 50 percent nitrous oxide, administered at subanesthetic concentrations, also produce a similar pattern. (Fig. 9-2)

210

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Fp1-F3M\'!'''*''''''I'''wo''-.'''IN''''''''''Ifr;fo.........,t''''If'-r'M~''v'''k''''''''''~~~~1'4I''-

Fp1-F3

F3-C3

F3-C3-c._ft"""'

C3-P3

C3-P3 ........,\-I'A-J'''tA}''>v...........''''''''''_

P3-01

P3-01 *-'oI'\N"'VI~v...~""""'"""""""'-'........., (cont'd)

~~~'N~,.MJ.---~ ..... .fM.tWV'....l~-_,;..,
(cont'd)

(cont'd)

(cont'd)

(cont'd)

~301.N

~30JlV 2 sec

2 sec

B

A

FIGURE 9-1 • EEG patterns related to A, anesthesia induction and B, decreasing levels of anesthesia. Note the frontal intermittent rhythmic delta activity (FIRDA) that occurs with induction, associated with generalized background slowing.

The drug-induced beta activity is less prominent with these other agents than with thiopenta1,5·12.29

Fp1-A2 j-'~~~~II~~~~~~

FP2-.~~~~~~~"~ F3-~

FzF4-

"M~

Symmetric EECi Changes with Indudion

~1~~~~\~~W

C3-

~~~

Cz-

~rfI~ ,.•rIl'iWri'r"W,~WrI'tJ~~~\V\it

C4-~~~~

P3-

~'i(IIIwi"'w.\'+I#'I~ 'J'\",,~v"i"~~"""'JJM!'wI.Jv'-t1vf1

pz-

~ \ J ~ ~ W'y1vl~I",!vNvI"lv"V'w'\,iJvNt':V

P4-

~ v,i'w/'l'lvv~~iJ#'I""f\/'

01-

~ .~.""""f'JW~~

02-

...,..~v~v."""'\""·"",\"';;"""",",' .pr.~~,~~,.~~~N'

~501Lv

2 sec

~501LV

1 sec

FIGURE 9-2 • Anesthesia-induced widespread, but anteriorly predominant, rhythmic fast activity.

Inductions for surgery are usually performed with the rapid administration of thiopental." A characteristic pattern of EEG changes occurs in relation to induction. The drug-induced beta activity noted with thiopental at subanesthetic concentrations becomes more widely distributed, gradually increases in amplitude, and slows in frequency.12.29 The background frequency ultimately slows from the beta to the alpha range. Paroxysmal bursts of high-amplitude, intermittent slowing occur; these are frontally predominant and resemble frontal intermittent rhythmic delta activity (FIRDA) (see Fig. 9-1; Fig. 9_3).12 Induction with other anesthetic agents may produce a similar

Intraoperative Electroencephalographic Monitoring

v+N_'..

1 .....~........"""-'_ _-~~~MI"It!-......,..1\

Fp1-F7

Fp1-F7

F7-T3

F7-T3

T3-T5

T3-T5 ..,........-................._ _"'"""'

T5-01~ Fp1-F3

o.L

211

''''''''b'''~'

~...,.",.. .........."'"""',.....,""""

T5-01 V"'---------""/'-~__.rcr-/\~~~~'Vv-~.rr-"'-....A._~ Fp1-F3~~.,,~_~

F3-C3~~ F3-C3~--V'~'-~~~ C3-P3~~v~~

C3-P3

P3-Q1"""--vv--"""'\../"'~~........._~"""""""",-.I\."""",".-"\-.A.--",,,,

P3-C1 Fp2-F4M<1v~

FP2-F4~~~

~ F4-C4v"'--.~~~~""'NI~_~""""""J"""""""'rvJ\,

F4-C4

C4-P4 .........
C4-P4

............,....,.....r-.._

P4-0~~ P4-02~v-v-"""'--"""v_-.r.....,-_.......-'-""'-J'vI\A..~--vv"'v.N""" Fp2-F8

,~i~

Fp2-FB

~~t#\/IWJ~.~okl\ll~~~~IjHIil..A..K'fIt

F8-T4 ~"~fI¥\\~~~~_~1!N(, T4-T6 """".....".."""'~_"v".~

F8-T4 T4-T6

..r-"'""""_,......

.....__

T6-02~~

T6-02

~oO fJ.V 1 sec

r.=

A

-=

1.~

J50 fJ.V 1 sec

B Fp1-F7 F7-T3 T3-T5 T5-01 Fp1-F3~

F3-C3'V C3-P3

...A,...r'LrJV-"""'_.rv...,.,.,.,~'-Af-"vVV',r"AJVV\~"-'"""...... /'v-..r--..,,~

P3-Q1 Fp2-F4 F4-C4 C4-P4 ....t'<~."""''''''-.'''''''..,..,...'-../'........,''"''''~~~'\J'-vv/'NI,.,.~M!'J'' P4-02 Fp2-F8 Fa-T4 ......r-; /"I.,..I-#'-" _"""-./'-'-'Vv\/'v...~.-v'''-''"f'-'''"''~

T4-T6 --"Vv-'\.r"-."""-.. r--.,-..,...r-"'-""",",", T6-02 .... .,...~.rA...A-. ._.r-.Ar-~,J-\",,--..r-/'~

-=

~50fJ.V

1 sec

C FIGURE 9-] • EEG tracings. A, Before induction; B, 35 seconds after induction; and C, 70 seconds after induction with fentanyl. Note the posterior alpha activity in the awake EEG before induction.

212

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

sequence of EEG changes, but with the FIRDA-like pattern being less prominent.P

a history of cerebral infarction may have a persistent or intermittent EEG abnormality (e.g., increased focal slow-wave activity) that should be recognized before carotid clamping. Even in the absence of a history of a cerebral infarction, the baseline EEG is Symmetric EEG Patterns at Sub-MAC important for comparative purposes. Subtle, lateralAnesthetic Concentrations ized, or localized EEG abnormalities may occur in The EEG pattern for all anesthetic drugs at a sub-MAC patients with transient cerebral ischemic attacks and anesthetic concentration consists of generalized backtransient monocular blindness that may reflect a sigground slowing with anteriorly predominant rhythmic nificant hemodynamic insult. It is presumed in these fast activity in the beta or alpha frequency range. 5,12.29-31 patients that diminished cerebral blood flow is suffiIncreasing the concentration of anesthetic drug is associent to produce an EEG change in the absence of a ciated with slowing of the frontal fast activity. The antepersistent neurologic deficit. 5.12,32 Widely distributed rior fast activity is nearly continuous in nature and is and generalized background alterations are less presumed to represent drug-induced beta activity, meaningful and may relate to preoperative medicawhich is concentration dependent. tions. The sensitivity of the EEG in patients with stroke depends on the size and pathophysiology of the infarct and the temporal relationship between EEG CHANGES UNRELATED TO EEG monitoring and the stroke. CAROTID ARTERY CLAMPING A focal abnormality may also be evident at the time of thiopental induction and usually correlates with that noted in the baseline EEG. 5 The focal abnormalAn intraoperative baseline, awake EEG should be ity may consist of a unilateral reduction of the anteobtained before induction to identity any background asymmetry or localization-related abnormality. 5,12 rior, rhythmic fast activity by more than 30 to 40 percent.v'? This reduction is commonly associated However, one study indicated that a preoperative with an increase in the wavelength and amplitude of EEG is not predictive of adverse effects associated with a carotid endarterectomy. I? The baseline, awake any persistent polymorphic slow activity. In most cases, any focal baseline EEG abnormality correlates with EEG was "highly predictive" of anesthesia-induced preoperative neurologic deficits." Focal anestheticEEG changes but not of alterations associated with related EEG abnormalities usually correlate with those carotid clamping. This study suggested that the prein the preoperative baseline EEG. Rarely, however, the operative EEG could be "eliminated."!" Patients with anesthetic effect may obscure an abnormality that was present in the EEG tracing obtained during wakefulness. More commonly, an EEG abnormality develops that was inapparent during the baseline EEG Fp1-A2 "'--~~'I.~~~ Fp2~".,.J.v,lj~~v--.~~~ (Fig. 9-5) .12 The EEG recorded during anesthesia may show a major reduction in anterior, rhythmic fast activF 3 - ~ ~ ~ ~ ~ ity and an increase in polymorphic slowing despite a Fz- ~~V"'~~!J',,-,,",--""'MlW'l"~~ normal baseline EEG with symmetric and reactive posterior alpha activity.5,12,29-31 Almost invariably in these F4instances, the drug-induced beta activity seen during C3the preanesthetic state is reduced on the side with Cz- ~ diminished anterior rhythmic fast activity. Commonly, the EEG during anesthesia shows enhanced abnorC4malities compared with the baseline recording, espeP3cially when the latter contains intermittent, rhythmic pzslowing in the temporal region on the side of the cerebral ischemia. This intermittent abnormality is often P4converted into a more obvious and persistent focal 01alteration that is associated with a reduction in the 02anesthetic pattern. ~""M"'~"""""",,lJJr"-----""""""~""~ Finally, it must be noted that a paroxysmal FIRDA<----J 50 fLv like discharge or continuous generalized background 2 sec slowing cannot be regarded as abnormal when it occurs FIGURE IH • Anterior rhythmic fast activity with superimduring anesthesia; these patterns occur in most patients posed anterior triangular slow waves during induction with pentothal. at some stage of anesthesia.l''

Intraoperative Electroencephalographic Monitoring

213

.....

Fp1-F3 "~""""""""""----""""""""""",,,,,,-.....J..J,ollol ~ F3-C3 .-w.----""""Wf-J'vkv""M1h""""'..-N~wo-"" C3-P3 .......,w-...,..-J'-~-...-"""'"""""""'f-.~~-"'"'wt ...... P3-01i11AMr1WYl!'JIJ.I,f'/V
Fp2-F4 _..r-v..,."--~.................. -",~-"-",--,-,,,,,~ F4-C4 ...-.r"""","",~.-.J'""""1V""fV'v'W""">l~,.,...,..,,,,,,,",,,~'W'-or C4-P4 ~"""",,-""'VvW'v..MI''''''''''''wvJi''''''''~'"''''MJ~'hoW'''-

P4-°4.fiMIJvJ.J.IvWVtJVjM,Nw1!WV'MIWNw~~NN'M ~ ~ ~501LV

1 sec

B

A

FIGURE g·s III EEG (A) during wakefulness and (B) after anesthesia with halothane and nitrous oxide in a patient with a right cerebral infarct. There is anesthesia-activated right hemisphere slowing. Note the absence of a focal abnormality during the awake EEG.

EEG CHANGES RELATED TO CAROTID ARTERY CLAMPING The rationale for intraoperative EEG monitoring is to assess the likelihood that the patient will acquire a neurologic deficit related to diminished cerebral blood flow at the time of carotid artery clamping. 5,12-19,32-36 The EEG should be recorded for at least 10 minutes with the patient under anesthesia and prior to carotid artery clamping.!" An alteration in the EEG that is transient or persistent occurs in approximately 25 percent of patients

Before clamping

with carotid artery stenosis on clamping of the internal carotid, clamping of the common carotid artery, or both. 5,12,13 These changes almost invariably occur within 20 to 30 seconds after clamping and are associated with a reduction of cerebral blood flow to below a critical level (Figs. 9-6 and 9_7).5,12 The critical level needed to affect the EEG is determined by the specific anesthetic drug. The severity and rapidity of onset of the EEG change vary in proportion to the degree to which blood flow is lowered (Fig. 9-8). Minor changes consist of a 25 to 50 percent reduction in the faster EEG activity and an

40 sec after clamping

2 min after shunting

Fp1-F3~

~

~

F3-C3~

~

~

C3-P3~

~

~

P3-01~

~

~

Fp2-F4~

~

~

F4-C4~~

~

~

C4-P4~

~

~

P4-02~

~

~

CBF = 8 ml/100 gm/min

CBF = 42 ml/100 gm/min

CBF = 25 ml/100 gm/min

~301LV 2 sec FIGURE 9'6 III EEG changes related to carotid artery clamping in a patient with high-grade stenosis of the left internal carotid artery, The reduction in cerebral blood flow (CBF) to 8 mlliOO gm/min correlates with a marked attenuation of EEG activity on the left, The use of a shunt results in improvement in cerebral blood flow to 42 mlliOO gm/min that parallels the improvement in the EEG,

214

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Before clamping

4 min after clamping

40 sec after clamping

FP1-F3~

~

~

F3-C3~~ C3-P~~

P3-0~~

CBF = 88 ml/100 gm/min

CBF = 15 ml/100 gm/min

CBF = 49 ml/100 gm/min

'--_---J] 40 Il-V 1 sec

c

B

A

FIGURE 1·7 II Compared with the EEG before clamping (A), the EEG during moderate reduction in cerebral blood flow (CBF) to 15 ml/lOO gm/min (B) reveals a reduction of ipsilateral fast activity with an increase in slowing. Recovery of the anesthetic-induced fast activity occurs with shunting and an improvement in cerebral blood flow (C).

increase in amplitude and wavelength of slower components (Fig. 9-7) .5.12-15 Changes seen with more severe reductions in blood flow (i.e., in the range of 6 to 7 ml/ I 00 gm/min or less) consist of a greater reduction of anesthetic-induced fast activity and a reduction in amplitude of the slower components, producing a low-amplitude, relatively featureless EEG on the side of clamping (Figs. 9-6, 9-8, 9-9, and 9_10).5.13 A major change is defined as an attenuation of at least 75 percent of all EEG activity or an increase in slow activity having a frequency of I Hz or less.12.13 Although approximately 25 percent of EEGs may show some change at the time of carotid artery clamping, only 1 to 3 percent of EEGs show such major alteration." Unilateral changes are more common than bilateral alterations (Figs. 9-6 and 9_9).6,7 The latter usually reflect a severe compromise in collateral circulation." Focal transient changes, occur-

Before clamping

c5 Age: 50 yr

(Left)

T3-C3~

(Right)

c5 Age: 53 yr

48

C4-T4~

(Left)

Q Age: 77 yr

CBF 72

35

T3-C3~

ring at times other than with clamping, can be seen in as many as 10 percent of patients.5.12.13 In the majority, this proves to be caused by transient, asymmetric effects of changing levels of anesthesia on a pre-existing focal abnormality and is of no consequence. IS Some are probably a transient consequence of embolization from the operative site. However, in about 1 percent of endarterectomies, a focal EEG change develops during the course of surgery, is unassociated with carotid clamping, persists throughout the procedure, and is ultimately associated with a new neurologic deficit in the immediate postoperative period. 13 The cause of this EEG change is usually cerebral embolization. An effective way of identifying embolization is to measure cerebral blood flow when a new, persistent focal EEG change develops. Embolization is associated with a change in the EEG in the absence of an alteration in cerebral blood flOW.5.12.13

After clamping Time in sec 150

4 min after shunting

CBF 17

~

50

CBF 50 ~

15

57

~

~

50 5 ~

L.--.J 50 Il-v 2 sec

FIGURE 1·8 II EEG findings in three patients with a significant reduction in cerebral blood flow (CBF) related to carotid clamping. Note that the severity of the EEG alteration correlates with cerebral blood flow measurements. The EEG alterations were reversible, as demonstrated by the recordings obtained after shunting.

215

Intraoperative Electroencephalographic Monitoring

Fp1-F7 'fNt',~~~~~~Ii'rMftNt'W~~~o/tIttI.lIH­ F7-T3 ~Wf'l~fIM~~~~~IIoNIt_~~tN'#I T3-T5 ~w.,~~~~~~~~~/fIAM_~ T5-01 ~1fIYV'Wf.,~f.IV'~~~~~wy,/lM-w..MWJ\ Fp1-F3 W'M~W~o/W"'~.w~~~~~W¥WfJ. F3-C3 ~~"""'~M4I//'f41~.~_~.~_ C3-P3 ~MI'fWtVtttVMW-~_WWJ~_~~4"t P3-01 ffl~iMoJ.N-W~~_~~WNo\~~~ Fp2~F4 WV-..AA_....-~M-............"..,Wi~IIMM~~~ F4-C4 """"""""""-"""""~--",,"'W¥Wr1'!\Wl/Wlf'll'fNi~ C4-P4 "'-"""""'",.......,_"""'"""""""""""'~~~~~~~ P4-02~~,."'(~~

Fp2-F8 ~'V'O>J\I'A.r-..lVV'-../\/'~~~""""JN,Jl/IItI'lIiW"""""I'W;ol"""" F8-T4 """"" ,.."",.""""'.........."Y"'\f'~~~Nll'Illo/lillljl\l~~ T4-T6 """'"' "........""V-<_-.~WtfIn~.,.,.,.""""',.....~1l'/Ml T6-02~~~~~ot-~\IOM~~~"""~'I'Nf' 'Shunting FIGURE 9·9 • Marked EEG changes related to clamping of the right carotid artery in a patient with left carotid stenosis. There is attenuation of the EEG activity on the right, with an increase in background slowing. Collateral flow was inadequate at the time of the right internal carotid artery occlusion.

The intraoperative EEG in patients undergoing local anesthesia has also been studied." Investigators at the University of Iowa compared the EEG changes in 96 patients receiving local anesthesia and in 121 patients receiving general anesthesia." The two patient groups

appeared similar preoperatively. EEG changes were more common in the general anesthesia group (15.7 percent vs. 6.3 percent). 37 The explanation for this difference was not obvious, but may suggest that the sensitivity and perhaps specificity of intraoperative EEG are altered by the type of anesthesia.F The intraoperative EEG during carotid endarterectomy performed in awake patients using a cervical block was evaluated at the University of Rochester.!" The EEG findings in 135 such patients were compared with those in 288 patients undergoing endarterectomy under general anesthesia. EEG changes were again more common in the group that was asleep during surgery (15.3 percent vs. 7.4 percent). is Global or bihemispheric EEG changes only occurred in the patients receiving general anesthesia. 16 The authors suggested that local anesthesia may be "cerebroprotective," and that variations in blood pressure associated with general anesthesia may be responsible for the global EEG changes.!"

CEREBRAL BLOOD FLOW Cerebral blood flow is measured in the operating room at the time of carotid endarterectomy by the xenon-I33 intracarotid injection technique.P Cerebral blood flow measurements are more sensitive than is determination of the "distal stump pressure" obtained by occluding the common carotid and external carotid arteries and measuring the residual pressure in the internal carotid artery.6.33 The cerebral blood flow studies are correlated with the EEG findings (cerebral function monitoring)

Fp1-F3 ~M~,m~~/I/IIH"""~"""l4W..II~""~~~~~fY!M.IFp1-F3 """'...--..-.....,........._-""""'......,........,. F3--e3 F3-C3 ---.r--_"'-"......,~ C3-P3 C3-P3 ..,oI.o, . . - -__-v_ _"""',.....__

_IV"

P3--01

P3-01 """"

Clamping

..._

......,,,.....----_ _

_1"""..---""""-......__

,.....,_ _..........-............."""........

Changing

Improving

Shunting (continued)

(continued)

-..-.. *

I

,,, •

.

..

.... Ie...... _

..........flt~

~·'1Il

_ _ j.,\

~Id_~~~"";~'~ ~J.~~~""'''~~ L-J 30 Jl.V CBF = 10 mll100 gmlmin

''t1

, ,

~30

"' . . . . ~

I '.....

,,'Io_ ~"'"'_

tI

'I*'""'~ ,",4' I .....

"'~..

• ....,..

t'

Alt, ."" ,... ...

.f.. . . . . . . . . . . ~

1t,Wr "

,wI.+.li\,J~"I.I."\

....IWl\\I'~

....

'''1111

Ill\,. *"'"..

.""

l.. ~ ...,.'1 '''~I"ilI • •

~j~~,~'*'''''1 'JIOl"I'i~~~.......

M'.., _'.. II."....,.,~

......w',"

2 sec

A

_-

"""_.............._ _.,.,..,._

Jl.V

CBF = 40 ml1100 gmlmin

2 sec

B

FIGURE 9-10 • A, Marked EEG changes related to carotid clamping during a left carotid endarterectomy. Note the rapid attenuation of fast activity and the emergence of increased delta activity at 1 Hz or less, correlating with a significant reduction in cerebral blood flow (CBF). B. Shunting increases cerebral blood flow and leads to a gradual increase in fast activity.

216

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

to determine the indication for carotid artery shunting at the time of carotid clamping. Xenon is injected into the internal carotid artery above a high-grade stenosis or into the common carotid artery (below the stenosis) with the external carotid artery clarnped.P A reduction in the cerebral blood flow to 18 ml/lOO gm/min or less is an indication for shunting.v" Blood flow is clearly influenced by the PaC0 2 level and requires that the xenon be delivered directly to the internal carotid artery while the external carotid artery is clamped. Further, blood flow techniques are intermittent and usually are done before clamping, immediately after clamping, immediately after placement of a shunt (when shunting is necessary), and at the end of the procedure. Xenon intracarotid blood flow studies and EEG monitoring are complementary procedures. 5,12,31-33 Focal EEG changes, caused by ischemia brought about by decreased perfusion pressure beyond a clamped carotid artery, are always associated with low blood flow, as measured by the xenon technique. Severe or major EEG changes occur with cerebral blood flow measures below 18 ml/lOO gm/min, and minor changes occur with values between 18 and 23 ml/l00 gm/min. 31-33 No change occurs when the cerebral blood flow is greater than 23 ml/lOO grrr/min." An EEG change that persists and is associated with a normal cerebral blood flow is indicative of embolization. The presence of normal blood flow following embolization is explained on the basis of the so-called "look-through phenomenon. "34 If the ischemia results from embolic occlusion of, for example, half the blood vessels to a region, with the other half remaining patent, injection of xenon results in a normal or, at times, increased flow and washout of xenon through the patent blood vessels, and the totally occluded vessels receive no xenon and therefore do not contribute to the overall measurement of flow,31,33

EEG AND CAROTID ENDARTERECTOMY The significance of EEG changes during a carotid endarterectomy has been evaluated.5,12-19,~H,33,35There is conflicting evidence regarding the prognostic importance of cerebral function monitoring.F The clinical practice at the Mayo Clinic has been to place a shunt whenever there is an EEG change related to carotid occlusion or when the cerebral blood flow with clamping is less than 18 ml/l00 gm/min. 12,13,31-33 A previous study at the University Hospital in London, Ontario, reported the postoperative outcome in 176 consecutive patients who underwent a carotid endarterectomy without the placement of a shunt. 12,36 No patient without EEG changes or with only minor EEG alterations had a neurologic deficit after surgery. Two of 22 patients (9 percent) with severe EEG changes related to carotid artery clamping had an

intraoperative stroke." Some, if not all of this disagreement may be resolved with an appropriate understanding of the concept of ischemic tolerance of neurologic tissue.P The length of time that neural tissue can tolerate ischemia without permanent infarction is inversely related to the severity of the reduction of blood flow during the time of ischemia. For instance, at zero flow (e.g., with cardiac arrest) it is approximately 4 minutes. At higher levels of flow (approximately 15 ml/ 100 gm/ min), laboratory studies suggest that the tolerance may be as long as 1 hour or more. The very low incidence of infarction reported in patients developing EEG changes during occlusion and in whom shunting was not performed is probably related to the fact that only a small percentage of patients (1 to 3 percent) have a severe reduction in cerebral blood flOW. 5,12,13 Cerebral infarction may not occur in patients with EEG changes because of the relatively short occlusion time (often less than 20 minutes) that is necessary for the carotid endarterectomy. 12 The efficacy of intraoperative EEG as a criterion for placement of a shunt was recently evaluated at one center during 1,661 operations.I" The overall number of patients with cerebral infarctions was 5. 38 One patient had a "minor" stroke without EEG change. Patients at risk for stroke had an abnormal EEG and contralateral internal carotid artery occlusion (3.3 percent had a cerebral infarction). It was concluded that intraoperative EEG is an excellent predictor for identifying patients who may require placement of a shunt during carotid endarterectomy.'" The value of intraoperative EEG was confirmed in another study that included 564 patients who underwent a carotid endarterectomy. Again, the EEG was highly predictive of the need for shunt placement, even in individuals with contralateral internal carotid artery occlusion.P Bydon and colleagues evaluated the stroke risk in 15 patients with contralateral occlusion without intraoperative shunting. 40 The mean clamp time of the common carotid artery was 18.5 minutes (range, 14 to 30 minutes). None of the patients experienced a cerebral infarct.t'' Therefore controversy remains regarding the primary indication for intraoperative EEG monitoring during carotid endarterectomy and the frequency of adverse effects subsequent to surgery. The divided opinions are the reason that some centers rely on the intraoperative EEG to determine the need for shunt placement, whereas other institutions either use other preoperative or intraoperative data or follow local guidelines.

EVOKED POTENTIALS AND CAROTID ARTERY SURGERY Somatosensory evoked potentials (SEPs) have been obtained during carotid endarterectomy to demon-

Intraoperative Electroencephalographic Monitoring

strate alterations related to cerebral ischemia. 4 1,42 The cortical waveforms are altered by a reduction in cerebral blood flow. Persistent SEP abnormalities in one study correlated with the presence of a postoperative neurologic deficit.f At present, SEP monitoring has "no advantages" over intraoperative EEG recordings."

EEG MONITORING AND CARDIAC SURGERY A number of patients undergoing cardiac surgery develop a transient or persistent neurologic deficit (e.g., a cognitive disorder). Several techniques to monitor for the development of such complications have been evaluated including EEG, transcranial Doppler, cerebral oximetry, and bispectral index monitoring.P !? A reduction in cerebral blood flow may occur during cardiopulmonary bypass for cardiac surgical procedures. EEG monitoring may be useful in detecting significant cerebral ischemia during cardiac surgery'v" EEG monitoring is limited by the use of profound hypothermia, which is associated with increased slowing as the temperature falls below 30°C and a burstsuppression pattern at 20°C. Other potential causes of global EEG changes in patients undergoing heart surgery include hypotension, hemodilution, anesthesia, and air or clot emboli.ti,43 Another cause of a significant EEG change is an obstruction of the pump lines or a low profusion flow and pressure.P Significant hypotension may produce generalized EEG changes, including increased background slowing or attenuation. The multifact.orial origin of EEG changes during cardiac surgery makes cerebral function monitoring less useful for open-heart surgery than for carotid endarterectomy."

REFERENCES 1. Walker AE, Marshall C, Beresford EM: Electrocorticographic characteristics of the cerebrum in posttraumatic epilepsy. Assoc Res Nerv Ment Dis, 26:502, 1947 2. Penfield W, Jasper HH: Epilepsy and the Functional Anatomy of the Human Brain. Little, Brown, Boston, 1954 3. Gloor P: Contributions of electroencephalography and electrocorticography to the neurosurgical treatment of the epilepsies. p. 59. In Purpura DP, Penry JK, Walter RD (cds): Neurosurgical Management of the Epilepsies. Raven Press, New York, 1975 4. Ajmone-Marsan C, Baldwin M: Electrocorticography. p. 368. In Baldwin M, Bailey P (eds): Temporal Lobe Epilepsy. Charles C Thomas, Springfield, 1958 5. So EL, Sharbrough FW: Cerebral function monitoring. p. 523. In Daube JR (ed): Clinical Neurophysiology. 2nd Ed. Oxford University Press, Oxford, 2002 6. Costin M, Rampersad A, Solomon RA et al: Cerebral injury predicted by transcranial Doppler ultrasonography but

217

not electroencephalography during carotid cndarterectomy, ] Neurosurg Anesthesiol, 14:287,2002 7. Edmonds HL, Rodriguez RA, Audenaert SM et al: The role of neuromonitoring in cardiovascular surgery. J Cardiothorac Vasc Anesth, 10:15, 1996 8. Puri GO, Murthy SS: Bispectral index monitoring in patients undergoing cardiac surgery under cardiopulmonary bypass. Eur J Anesthesiol, 20:451, 2003 9. Edmonds HL: Multi-modality neurophysiologic monitoring for cardiac surgery. Heart Surg Forum, 5:225, 2002 10. Lehmann A, Karzau J, Boldt J et al: Bispectral indexguided anesthesia in patients undergoing aortocoronary bypass grafting. Anesth Analg, 96:336, 2003 11. Ascher E, Markevich N, Kallakuri S et al: Intraoperative carotid artery duplex scanning in a modern series of 650 consecutive primary endarterectomy procedures. J Vase Surg, 39:416, 2004 12. Blume WT, Sharbrough FW: EEG monitoring during carotid endarterectomy and open heart surgery. p. 797. In Niedermeyer E, Lopes da Silva F (eds): Electroencephalography: Basic Principles, Clinical Applications and Related Fields. 4th ed. Williams & Wilkins, Baltimore, 1999 13. Sharbrough FW, MessickJM, Sundt TM,Jr: Correlation of continuous electroencephalograms with cerebral blood flow measurements during carotid endarterectomy. Stroke, 4:674, 1973 14. Sundt TM,Jr, Sharbrough FW, Piepgras DG et al: Correlation of cerebral blood flow and electroencephalographic changes during carotid endarterectomy. Mayo Clin Proc, 56:533, 1981 15. Sundt TM, Jr, Sharbrough FW, Anderson RE et al: Cerebral blood flow measurements and electroencephalograms during carotid endarterectomy. J Neurosurg, 41:310,1974 16. Illig KA, Sternbach Y, Zhang Ret al: EEG changes during awake carotid endarterectomy. Ann Vase Surg, 16:6,2002 17. Illig KA, BurchfielJL, Ouriel K et al: Value of preoperative EEG for carotid endarterectomy. Cardiovasc Surg, 6:490, 1998 18. Chiappa KH, Burke SR, Young RR: Results of electroencephalographic monitoring during 367 carotid endarterectomies: use of a dedicated minicomputer. Stroke, 10:38], 1979 19. Cucchiara RF, Sharbrough FW, Messick JM,.Jr et al: An electroencephalographic filter-processor as an indicator of cerebral ischemia during carotid endarterectomy. Anesthesiology, 5]:77, 1979 20. Quasha AL, Sharbrough FW, Schweller TA et al: Hypothermia plus thiopental: synergistic EEG suppression. Anesthesia, 51:S20, 1979 21. Mayberg MR: Carotid endarterectomy for symptomatic carotid stenosis. J Stroke Cerebrovasc Dis, 6:185, 1997 22. Cohen S: Carotid endarterectomy for asymptomatic disease . .J Stroke Cerebrovasc Dis, 6: 180, 1997 23. The Asymptomatic Carotid Atherosclerosis Study Group: Endarterectomy for asymptomatic carotid artery stenosis. JAMA, 273:1421, 1995 24. North American Symptomatic Carotid Endarterectomy Trial Collaborators: Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade stenosis. N EnglJ Med, 325:445,1991

218

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

25. Mayberg MR, Wilson SE, Yatsu F et al: Carotid endarterectomy and prevention of cerebral ischemia and symptomatic carotid stenosis. JAMA, 266:3289, 1991 26. Brooks WH, McClure RR,Jones MR et al: Carotid angieplasty and stenting versus carotid endarterectomy for treatment of asymptomatic carotid stenosis: a randomized trial in a community hospital. Neurosurgery, 54:318, 2004 27. CARESS Steering Committee. Carotid revascularization using endarterectomy or stenting systems (CARESS):phase I clinical trial.J Endovasc Ther, 10:1021,2003 28. McDonnell CO, Legge D, Twomey E et al: Carotid artery angioplasty for restenosis following endarterectomy. EurJ Vase Endovasc Surg, 27:163, 2004 29. Stockard lJ, Bickford RG: The neurophysiology of anaesthesia. p. 3. In Gordon E (ed): A Basis and Practice of Neuroanaesthesia. Excerpta Medica, Amsterdam, 1975 30. Hansotia PL, Sharbrough FW, Berendes J: Activation of focal delta abnormality with methohexital and other anesthetic agents. Electroencephalogr Clin Neurophysiol, 38:554, 1975 31. McKay RD, Sundt TM,Jr, Michenfelder JD et al: Internal carotid artery stump pressure and cerebral blood flow during carotid endarterectomy: modification by halothane, enflurane, and innovar. Anesthesiology, 45:390, 1976 32. Yanagihara T, Klass DW: Discrepancy between CT scan and EEG in hemodynamic stroke of the carotid system. Trans Am Neurol Assoc, 104:141, 1980 33. Sundt TM, Jr, Sharbrough FW, Trautmann JC et al: Monitoring techniques for carotid endarterectomy. Clin Neurosurg, 22:199,1975 34. Donley RF, Sundt TM, Jr, Anderson RE et al: Blood flow measurements and the "look through" artifact in focal cerebral ischemia. Stroke, 6:121, 1975

35. Baker WH, Dorner DB, Barnes RW: Carotid endarterectomy: is an indwelling shunt necessary? Surgery, 82:321, 1977 36. Blume WT, McNeill DK, Ferguson GG: EEG during carotid artery clamping at endarterectomy without shunting. Electroencephalogr Clin Neurophysiol, 56:27P, 1983 37. Wellman BJ, Loftus CM, Kresowik TF et al: The differences in electroencephalographic changes in patients undergoing carotid endarterectomies while under local versus general anesthesia. Neurosurgery, 43:769, 1998 38. PinkertonJA: EEG as a criterion for shunt need in carotid endarterectomy. Ann Vase Surg, 16:756,2002 39. Schneider JR, Droste JS, Schindler N et al: Carotid endarterectomy with routine electroencephalography and selective shunting: influence of contralateral internal carotid artery occlusion and utility in prevention of perioperative strokes.J Vase Surg, 35:1114, 2002 40. Bydon A, Thomas AJ, Seyfried D et al: Carotid endarterectomy in patients with contralataeral internal carotid artery occlusion without intraoperative shunting. Surg Neurol, 57:325, 2002 41. Markand ON, Dilley RS, Moorthy SS et al: Monitoring of somatosensory evoked responses during carotid endarterectomy. Arch Neurol, 41:375,1984 42. Horsch S, De Vleeschauwer P, Ktenidis K: Intraoperative assessment of cerebral ischemia during carotid surgery. J Cardiovasc Surg, 31:599,1990 43. Salerno TA, Lince DP,White DN et al: Monitoring of electroencephalogram during open heart surgery: a prospective analysis of 118 cases.] Thorac Cardiovasc Surg, 76:97, 1978

CHAPTER

11 Clinical Electromyography MICHAEL J. AMINOFF

PRACTICAL ASPECTS Procedure ELECTRICAL ACTIVITY OF NORMAL MUSCLE EMG Activity at Rest EMG Findings During Activity Motor Unit Action Potentials Motor Unit Recruitment Pattern EMG ACTIVITY IN PATHOLOGIC STATES EMG Activity at Rest Insertion Activity Fibrillation Potentials Positive Sharp Waves Fasciculation Potentials Myotonic Discharges Complex Repetitive Discharges Motor Unit Action Potentials EMG Findings DuringActivity Motor Unit Action Potentials Abnormalities of Recruitment Pattern EMG FINDINGS IN VARIOUS CLINICAL DISORDERS Myopathic Disorders Muscular Dystrophies and OtherFamilial Myopathies

The term electromyography refers to methods of studying the electrical activity of muscle. Over the years, such methods have come to be recognized as an invaluable aid to the diagnosis of neuromuscular disorders. As is discussed in this chapter, electromyography (EMG) has been used to detect and characterize disease processes affecting the motor units and to provide a guide to prognosis. Electromyographic examination is often particularly helpful when clinical evaluation is difficult or equivocal. The findings commonly permit the underlying lesion to be localized to the neural, muscular, or junctional component of the motor units in question. Indeed, when the neural component is involved, the nature and distribution of EMG abnormalities may permit the lesion to be localized to the level of the cell bodies of the lower motor neurons or to their axons as they traverse a spinal root, nerve

Inflammatory Disorders of Muscle Endocrine and Metabolic Myopathies Myopathies Caused by Drugs or Alcohol Critical Illness Myopathy Congenital Myopathies of Uncertain Etiology Myotonic Disorders Rippling Muscle Disease Neuropathic Disorders Spinal CordPathology RootLesions Plexus Lesions Peripheral Nerve Lesions Disorders of Neuromuscular Transmission Miscellaneous Disorders DIAPHRAGMATIC ELECTROMYOGRAPHY SPHINCTERIC ELECTROMYOGRAPHY LARYNGEAL ELECTROMYOGRAPHY QUANTITATIVE ASPECTS OF ELECTROMYOGRAPHY

plexus, or peripheral nerve. The EMG findings per se are never pathognomonic of specific diseases and cannot provide a definitive diagnosis, although they may justifiably be used to support or refute a diagnosis advanced on clinical or other grounds. Electromyography is also used in conjunction with nerve conduction studies to obtain information of prognostic significance in the management of patients with peripheral nerve lesions. For example, EMG evidence of denervation implies a less favorable prognosis than otherwise in patients with a compressive or entrapment neuropathy. Again, evidence that some motor units remain under voluntary control after a traumatic peripheral nerve lesion implies a more favorable outlook than otherwise for ultimate recovery, indicating as it does that the nerve remains in functional continuity, at least in part.

233

234

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

The clinical utility of electrodiagnostic testing in patients presenting with a chief complaint of weakness has been examined. 1 The referring diagnosis was compared with the diagnosis immediately after electrophysiologic evaluation, and then with the final diagnosis as recorded 9 months later. This revealed that the testing had resulted in a single correct diagnosis in 73 percent of patients; where it resulted in more than one possible diagnosis, one of them was ultimately confirmed as correct in another 18 percent of patients, to yield an overall diagnostic accuracy of 91 percent. The electrophysiologic diagnosis was unsuspected by the referring physician, regardless of his or her specialty, in approximately one-third of cases. I Over the years, the activity of individual muscles in the maintenance of posture and during normal or abnormal movement has also been studied by EMG. Such studies are of considerable academic interest, and their clinical relevance is considered further in Chapter 18. The interested reader is also referred to a glossary of terms commonly used in electromyography.''

PRACTICAL ASPECTS The electrical activity of muscle is studied for diagnostic purposes by inserting a recording electrode directly into the muscle to be examined. The bioelectric potentials that are picked up by this electrode are amplified, then displayed on a cathode ray oscilloscope for visual analysis and fed through a loudspeaker system so that they can be monitored acoustically. A permanent photographic record of the oscilloscope trace can be made, if desired, or the amplified bioelectric signals can be recorded for retrieval at a later date. Modern commercially available equipment includes analog-to-digital converters that permit the easy storage of data, advanced signal processing and analysis, and alteration in the display characteristics (e.g., time base and sensitivity) of stored potentials. The concentric needle electrode is a convenient recording electrode for clinical purposes. It consists of a pointed steel cannula within which runs a fine silver, steel, or platinum wire that is insulated except at its tip. The potential difference between the outer cannula and the inner wire is recorded, and the patient is grounded by a separate surface electrode. Alternatively, a monopolar needle electrode can be used. This consists of a solid needle (usually of stainless steel) that is insulated except at its tip. The potential difference is measured between the tip of the needle, which is inserted into the muscle to be studied, and a reference electrode (e.g., a conductive plate attached to the skin or a needle inserted subcutaneously). The pick-up area of the concentric needle electrode is smaller than that of the monopolar electrode, and it is asymmetric as opposed

to the more circular pick-up area of the monopolar electrode. Both concentric and monopolar electrodes are available in disposable or reusable forms. There is no evidence that one type of electrode is more painful to patients than the other' An electrode exhibits some opposition or impedance to the flow of an electric current, and it is therefore important that it is connected to an amplifier having a relatively high input impedance to prevent loss of the signal. The amplifier, and the recording system to which it is connected, should have a frequency response of 2 or 20 to 10,000 Hz so that signals within this frequency spectrum are amplified uniformly without distortion. When necessary, however, the frequency response of the amplifier can be altered by the use of filters. This allows the attenuation of noise or interference signals that have a frequency different from that of the potentials under study. Noise, which appears as a random fluctuation of the baseline, is generated within the amplifier and by movement of the recording electrodes or their leads. It can obscure the bioelectric signals to be studied, as can any unwanted interference signals that are picked up by the recording apparatus. Interference signals are usually generated by the AC power line, by appliances such as radios, or by paging systems; occasionally, however, they are biologic in origin. Technical and safety factors are important when the electrical activity of muscle is to be recorded. For a detailed account of this aspect and the various technical problems that those undertaking EMG might face, see Chapter 2 and the review by Gitter and Stolov,"

Procedure The patient is examined in a warm, quiet room. The time base of the oscilloscope is allowed to sweep freely from left to right with a speed of 10 msec / cm, or with a slower sweep speed when firing patterns are to be characterized. The gain is commonly set at 50 or 100 IlVIcm for examining insertion and spontaneous activity, and at 200 or 500 IlVIcm for studying motor unit action potentials. Appropriate filter settings were discussed earlier. Proper grounding is essential as discussed in Chapter 2. Muscles are selected for examination on the basis of the patient's symptoms and signs and the diagnostic problem that they raise. A ground lead is attached to the same limb as the muscles that are to be examined. The needle electrode is inserted into the muscle while it is relaxed so that the presence and extent of any insertion activity can be noted. The muscle is then explored systematically with the electrode for the presence of any spontaneous activity. Following this, the parameters of individual motor unit action potentials are defined in different sites during graded muscle contraction, attention being directed not only

Clinical Electromyography

to the shape and dimensions of the potentials but also to their initial firing rate and the rate at which they must fire before additional units are recruited. Finally, the interference pattern is compared with the strength of contraction during increasingly powerful contractions, until full voluntary power is being exerted. Needle EMG is an invasive procedure, and concern has increased about infective complications, involving both patients and electromyographers, especially involving human immunodeficiency virus (HIV) , hepatitis virus, or Creutzfeldt-Jakob disease. It is wise for the physician to wear latex or rubber gloves during the examination and then to discard them in an appropriate receptacle once the procedure is over. Disposable needle electrodes are preferable for patient safety; they must be handled with care and disposed in proper containers after use. Needles should not be inserted through an infected space. Remote electrode

235

Other complications are mild or rare. Pain at the site of needle insertion sometimes lasts for I or 2 days, and bruising may occur at the site of needle insertion. Pneumothorax is a rare complication of diaphragmatic EMG or of examination of the chest wall, supraspinatus, or cervical or thoracic paraspinal muscles. One patient is reported to have developed an acute compartment syndrome of the leg following needle EMG, probably because of puncture of a small blood vessel by the examining electrode; this necessitated a surgical release procedure:" The needle examination carries a small but definite risk of hemorrhagic complications in patients with acquired or inherited coagulopathies and in those patients taking anticoagulants or even antiplatelet agents. When it is required, nevertheless, small electrodes should be used; the advice of a hematologist should be obtained in severe cases. Certain over-the-counter herbal remedies such as ginkgo biloba Recorded potential changes

A

Direction of impulse •

FIGURE 11-1 • Schematic diagram of the passage of an action potential along a nerve or muscle fiber in a conducting medium. The active electrode is on the surface of the fiber, and the reference electrode is at a remote point in the conducting medium. (From Brazier MAB: Electrical Activity of the Nervous System. 4th ed. Pitman Medical, London, 1977, with permission.)

o

236

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

~--------------J100

fLV

L-..J 100 msec FIGURE 11-2 • Insertion activity evoked in normal muscle. In accordance with convention, an upward deflection in this and subsequent figures indicates relative negativity of the active electrode.

and ginseng may also increase the risk of bleeding." Rare reports of needle electrodes breaking off within a muscle and requiring minor surgery for retrieval have been published."

ELEORICAL AOIVITY OF NORMAL MUSCLE In clinical EMG, electrical activity is recorded extracellularly from muscle fibers embedded in tissue, which is itself a conducting medium. The action potentials that are recorded in this way have a tri- or biphasic configuration. The basis for this is illustrated schematically in Figure 11-1, where the active electrode is shown on the surface of a fiber and the reference electrode is placed at a remote point in the conducting medium. Because there is a flow of current into the fiber (i.e., a current "sink") at the point of excitation, and an outward flow of current in adjacent regions, the propagated impulse can be considered a moving sink of current, preceded and followed by current sources. Accordingly, when an impulse travels toward the active electrode, this electrode becomes relatively more positive as it comes to overlie the current source preceding the action potential (Fig. 11-1, A). A short time later, as the impulse itself arrives, the active electrode registers a negative potential in relation to the distant electrode (Fig. 11-1, B), and then a relatively positive potential as the impulse passes on and is followed by the current source behind it (Fig. 11-1, C). As this too passes on, the electrode comes again to be on a resting portion of the fiber and the potential between the two electrodes returns to the baseline (Fig. 11-1, D). Clearly, the recorded action potential will be biphasic with a negative onset (rather than triphasic with a positive onset) if the active electrode is placed over the region of the fiber where the impulse is initiated.

EMG Adivity at Rest EMG activity usually cannot be recorded outside of the endplate region of healthy muscle at rest, except imme-

diately after insertion or movement of the needle recording electrode. The activity related to electrode movement (i.e., insertion activity) is caused by mechanical stimulation or injury of the muscle fibers and usually stops within about 2 seconds of the movement (Fig. 11-2). After cessation of this activity, spontaneous activity may be found in the endplate region but not elsewhere. This endplate noise, as it is called, consists of monophasic negative potentials that have an irregular, high-frequency discharge pattern; a duration of between 0.5 and 2.0 msec; and an amplitude that is usually less than 100 flV. The potentials correspond to the miniature endplate potentials that can be recorded with microelectrodes in animals. Biphasic potentials with a negative onset are also a constituent of endplate noise and have a duration of 3 to 5 msec and an amplitude of 100 to 200 flV. They have been held to represent muscle fiber action potentials arising sporadically because of spontaneous activity at the neuromuscular junction or activity in intramuscular nerve fibers.

EMG Findings During Adivity MOTOR UNIT ACTION POTENTIALS

Excitation of a single lower motor neuron normally leads to the activation of all of the muscle fibers that it innervates (i.e., those that constitute the motor unit). The motor unit action potential is a compound potential representing the sum of the individual action potentials generated in the few muscle fibers of the unit that are within the pick-up range of the recording electrode. When recorded with a needle electrode, motor unit action potentials are usually bi- or triphasic in shape (Fig. 11-3), but in the limb muscles about 12 percent may have five or more phases and are then described as polyphasic. The number of phases is calculated by adding one to the number of times that the baseline is crossed. The total duration of the potentials (i.e., the time taken for the trace to return finally to the baseline after its initial deflection at the beginning of a potential) is normally between 2 and 15 msec depending on the muscle being examined. The precise value relates to the anatomic scatter of endplates of the muscle

Clinical Electromyography

I

L.....J

A

I

20 msec

20 msec

B

237

fibers that generate the spike and becomes shorter as this distance is reduced. A rise time of approximately 200 usee or less is necessary if the other characteristics of the potentials are to be evaluated usefully. The area of the negative spike depends on the number and diameter of muscle fibers closest to the recording electrode, and the temporal dispersion of their action potentials. The area of the entire potential provides similar information, but for all of the muscle fibers contributing to the potential. Physiologic Factors Influencing Motor Unit Action Potentials

500 fJ-V

[--~ ...- -

L.J

c

20 msec

L.J

20 msec

o

FIGURE 11-] II Motor unit action potentials. A, Normal potential. B, Low-amplitude, short-duration, polyphasic potential. C, Long-duration, polyphasic potential. D, Polyphasic potential with a late component.

fibers in the units under study that are within the pickup zone of the recording electrode. This is because of the different distances along which the muscle fiber action potentials will have to be conducted from the individual endplates to the recording zone of the electrode. The amplitude of the individual motor unit action potentials is measured between the greatest positive and the greatest negative deflections of the potentials. When recorded by a concentric needle electrode, it is usually between 200 ~V and 3 mY; this is determined largely by the distance between the recording electrode and the active fibers that are closest to it, and by the recruitment order. The number of active fibers lying close to the electrode and the temporal dispersion of their individual action potentials also affect the amplitude of the potentials but to a lesser extent. Computer simulations of motor unit action potentials have suggested that amplitude is determined by less than eight fibers situated within 0.5 mm of the electrode." Other features of individual motor unit action potentials that merit consideration are the rise time and area. The rise time of the negative spike is the interval between the onset and the peak of this component of the motor unit action potential. It reflects the distance between the recording electrode and the muscle

The configuration and dimensions of individual motor unit action potentials are normally constant provided that the recording electrode is not moved. They are, however, influenced by the characteristics of the recording electrode and apparatus in the electromyograph system; and by physiologic factors such as patient age, intramuscular temperature, the site of the recording electrode within the muscle, and the particular muscle under examination. The potentials recorded with concentric needle electrodes tend to have slightly lower amplitudes and shorter durations than those recorded with monopolar needle electrodes. With increasing age from infancy to adulthood there is an increase in the mean duration of motor unit action potentials in limb muscles, probably because of growth in width of the territory over which endplates are scattered. Later increases in duration relate to increasing fiber density within motor units as a result of the reduction in muscle volume that occurs in older subjects. Mean duration of motor unit potentials and the number of polyphasic potentials also increase as temperature declines.V' Abnormalities in the parameters of motor unit action potentials occur in neuromuscular diseases, as discussed later. MOTOR UNIT RECRUITMENT PATTERN

When a muscle is contracted weakly, a few of its motor units begin firing at a low rate. As the force of contraction increases, the firing rate of these active units increases until it reaches a certain frequency, when additional units are recruited. The frequency at which a particular unit must fire before another is recruited (i.e., the recruitment frequency) depends on the muscle and motor unit being studied and on the number of units capable of firing and the tension that they can generate, but it is usually between 5 and 20 times per second. The ratio of the number of active motor units to the firing frequency of individual units is generally less than 5 and is relatively constant for individual muscles. Thus, with a firing frequency of 20, the number of active units will be 4 or more. In general, the units recruited first are smaller in amplitude than those recruited later.

238

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

I

i

I

I

reduced or absent when the needle electrode is not situated in muscle, and in certain metabolic disorders (e.g., hypokalemic periodic paralysis). Insertion activity is prolonged in denervated muscle and in polymyositis, the myotonic disorders, and some of the other myopathies; in these instances it consists of repetitively firing fibrillation potentials and positive sharp waves. In myotonic disorders, waxing and waning myotonic discharges are typically found (p, 240) Insertion activity consisting of an irregular discharge of variable duration that "snaps, crackles, and pops" and follows normal insertion activity has been described by Wilbourn. III It has no particular pathologic significance. After cessation of all insertion activity,various types of spontaneous activity may be recorded from fully relaxed muscle in patients with a neuromuscular disorder.

A

0

B

FIBRILLATION POTENTIALS

iJ

200 !-LV

I,

c 400 msec FIGURE 11·4 II Motor unit action potentials recorded during (A) slight, (B) moderate, and (C) maximal voluntary contraction of muscle.

With increasing muscle contraction, so many units are eventually active that the baseline is interrupted continuously by the potentials, and individual potentials cannot be distinguished from each other. The resulting appearance of the oscilloscope trace is referred to as the interference pattern (Fig. 11-4). Abnormalities of recruitment pattern are discussed on page 242.

EMC ACTIVITY IN PATHOLOGIC STATES EMC Adivity at Rest INSERTION ACTIVITY

Insertion activity is found only when some muscle tissue remains viable and is therefore absent in the advanced stages of various neuromuscular disorders. It is also

Fibrillation potentials are action potentials that arise spontaneously from single muscle fibers. When they occur rhythmically, their genesis may relate to oscillations of the resting membrane potential of denervated skeletal muscle fibers. ll o12 Less commonly, they occur irregularly; this has been attributed to the occurrence of random, discrete, spontaneous depolarizations that originate in the transverse tubular system of the muscle fiber.P The initiating events of both types of fibrillation are suppressed by tetrodotoxin or removal of external sodium ions, suggesting that they are related to changes in sodium conductance.'! Fibrillation potentials usually have an amplitude of between 20 and 300 I1V, a duration of less than 5 msec, and a firing rate of between 2 and 20 Hz (Fig. 11-5). They have a bi- or triphasic shape, the first phase being positive except when the potentials are recorded in the endplate region. This positive onset facilitates their distinction from endplate noise. Over the loudspeaker, they give rise to a high-pitched repetitive click, which aids their detection. They are found in denervated muscle, provided that some tissue remains viable and the muscle is warm when examined; however, they may not appear for 3 to 5 weeks after an acute neuropathic lesion. Once present, they may persist for months or even years, until the muscle fibers have come to be reinnervated or have degenerated, but there is some evidence that their amplitude declines with time.!" They are not in themselves diagnostic of denervation, however, because they are also seen in primary muscle diseases such as polymyositis, inclusion body myositis, and muscular dystrophy, and in patients with botulism, trichinosis, muscle trauma, or metabolic disorders such as acid maltase deficiency or hyperkalemic periodic paralysis. The presence of fibrillation potentials in myopathic disorders probably relates to isolation of part of the muscle fibers from their endplates, so that they are

Clinical Electromyography

_I 9

~

I

"*'-

}=

~

I

~

I ........

A

V-

'1-

239

y-

..

'1-

J100/LV

r-

FIGURE 11-5 • Spontaneous activity recorded in relaxed, partially denervated muscle. A, Fibrillation poten tials. B, Positive sharp waves. J200 /LV

B 100 msec

functionally denervated.P'!" Scanty fibrillation potentials occasionally may be found in normal healthy muscle; pathologic significance should therefore not be attributed to them unless they are detected in at least three separate sites within the muscle being examined. PosmVE SHARP WAVES

Positive sharp waves usually are found in associanon with, and have the same clinical significance as, fibrillation potentials. They occur both in denervated muscle and in certain primary disorders of muscle. Indeed, Nandedkar and colleagues have shown the transformation of a fibrillation potential to a positive sharp wave, and vice versa, reinforcing the concept that these potentials are two manifestations of the same phenomenon.!? They are thought to arise from single muscle fibers that have been injured. 17•18 As viewed on the oscilloscope, they consist of an initial positive deflection, followed by a slow change of potential in a negative direction that may be extended into a small negative phase (see Fig. 11-5). Their amplitude is usually about the same as or slightly greater than that of fibrillation potentials, their duration is often 10 msec or more, and their discharge rate is similar to fibrillation potentials but can sometimes reach values of up to 100 Hz. They may fire regularly or, less commonly, irregularly. Positive sharp waves may occur diffusely in the absence of fibrillation.l'' sometimes on a familial basis.l" FASCICULATION POTENTIALS

Fasciculation potentials are similar to motor unit action potentials in their dimensions and have been attributed to spontaneous activation of the muscle fibers of individual motor units. Their detection is aided by the sudden dull "thump" that they produce over the loudspeaker. They may be found in normal muscle or in

patients with chronic partial denervation, particularly when this is caused by spinal cord pathology, and especially in amyotrophic lateral sclerosis. Pathologic significance should not be attributed to them unless they are accompanied by other evidence of denervation (e.g., fibrillation potentials or changes in motor unit action potentials) . Early investigators considered that fasciculation potentials arose at the neuromuscular junctions and then spread antidromically by axonal reflex to the other nerve terminals of the motor unit. It was then suggested that distal immature collateral sprouts in fasciculating muscle reacted to humoral substances such as acetylcholine to cause axon reflex discharges of surviving enlarged motor units; or that fasciculations had a central basis, arising because the motor neuron was unable to maintain a normal membrane potential, with resulting hyperirritability and abnormal synchronization of spontaneous discharges. Wettstein, using a collision technique, studied the origin of distal fasciculations in 15 patients with anterior horn cell involvement from various causes." He found that they could originate at multiple sites along motor axons or the somas of diseased motor neurons. In 2 of the 25 fasciculating motor units studied, there was an exclusively distal origin, in IS an exclusively proximal origin, and in 8 a mixed origin. These three categories of fasciculations had no other common characteristics and were found in similar proportions in patients with amyotrophic lateral sclerosis or other disorders involving the anterior horn cells, and with or without symptoms of upper motor neuron involvement. Using a similar technique, however, Roth determined the origin of 100 fasciculations in various lower motor neuron lesions: In 82 percent it was on the distal extremity of the axon, regardless of the type or duration of the lesion or severity of denervation.F The basis of fasciculation potentials remains uncertain and may vary in different circumstances. Although

240

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

widely regarded as representing the activity of muscle fibers in individual motor units, the evidence is conflicting in motor neuron disease and amyotrophic lateral sclerosis. On the one hand, certain investigators have noted that fasciculation potentials are not identical to the motor unit action potentials that can be recorded at the same site during mild voluntary contraction or passive stretch. 23 •24 On the other, some patients are able voluntarily to activate their fasciculating motor units," and electrical stimulation of the distal motor axons may elicit potentials that are identical to spontaneously occurring fasciculation potentials." By contrast, benign fasciculations may result from ephaptic activation of adjacent muscle fibers in irregularly fibrillating muscle fibers. 26•27 MYOTONIC DISCHARGES

Myotonic discharges consist of high-frequency trains of action potentials that are evoked by electrode movement or by percussion or contraction of the muscle; they are enhanced by cold and reduced by repeated contractions. Some of the potentials resemble fibrillation potentials or positive sharp waves; others resemble motor unit potentials. Their frequency (15 to 150 Hz) and amplitude (10 ,..tVto 1 mY) wax and wane (Fig. 11-6), and in consequence the trains of potentials produce a sound like that of a dive-bomber on the loudspeaker. They are found in patients with one of the various myotonic disorders. They may also be found in hyperkalemic periodic paralysis, and occasionally are found in patients with polymyositis or the myopathy of acid maltase deficiency. Myotonic discharges do not result simply from mechanical irritation because they can be recorded with surface electrodes, and they are not prevented by pharmacologic blockade of neuromuscular transmission. They apparently relate to a disorder of the muscle fiber membrane. Altered chloride channels have been incriminated in myotonia congenita, and altered regu-

lation of sodium channels has been implicated in myotonic dystrophy. Disease of the sodium channels appears to be responsible in hyperkalemic periodic paralysis and paramyotonia congenita.P COMPLEX REPETITIVE DISCHARGES

Trains of high-frequency action potentials are sometimes found in the muscles of patients with muscular dystrophy, polymyositis, or chronic partial denervation; they are also occasionally found in patients with metabolic disorders such as hyperkalemic periodic paralysis, hypothyroidism, or certain of the glycogen storage diseases. They are often believed to indicate chronicity of the causal disorder, but a recent study failed to support this concept, indicating that the likelihood of finding them was the same in acute as in chronic conditions.r" The discharges occur spontaneously and after electrode movement or voluntary contraction. They have an abrupt onset and termination, but unlike myotonic discharges, their amplitude and frequency remain constant (see Fig. 11-6). The individual action potentials are often polyphasic. Their origin is uncertain, but they appear to arise in the muscle itself. Experimental studies suggest that discharges are probably initiated by a fibrillating muscle fiber that, by ephaptic transmission, depolarizes one or more adjacent hyperexcitable muscle fibers.t" The configuration of the discharges depends on the synchrony of firing of the fibers that contribute to the discharges.t" MOTOR UNIT ACTION POTENTIALS

Motor unit activity may occur spontaneously and repetitively despite full voluntary relaxation of the muscle in patients with myokymia, muscle cramps, or tetany. Individual action potentials sometimes exhibit a rhythmic, grouped pattern of firing so that double, triple, or multiple discharges are seen (Fig. 11-7). Double discharges also sometimes occur at the beginning of a

A 11·6 II Spontaneous high-frequency activity. A, Myotonic discharge recorded in a patient with myotonia congenita. B, Complex repetitive discharges recorded in a partially denervated muscle.

FIGURE

~~~~~ ~ ~~~~J100"V

8

I

I

100 msec

Clinical Electromyography

FIGURE 11·7 II Spontaneous repetitive motor unit activity, showing grouped discharges.

241

-----'l\~~'+--_.~-J200

voluntary contraction in normal subjects and patients with disorders of the anterior horn cells, roots, or peripheral nerves. The term myokymic discharges is used to refer to spontaneously occurring grouped action potentials, each group being followed by a period of silence, with subsequent repetition of a grouped discharge of identical action potentials in a semirhythmic manner." The multiplets or grouped motor unit action potentials of tetany may resemble myokymic discharges but are under voluntary control." Myokymic discharges may occur in the limbs of patients with radiation-induced plexopathy or, less commonly, patients with radiation myelopathy, multiple sclerosis, acute inflammatory polyradiculoneuropathy, chronic radiculopathy or entrapment neuropathy.'! syringomyelia.F or gold intoxication. Myokymic discharges in facial muscles occur in some patients with multiple sclerosis, brainstem neoplasms, or polyradiculoneuropathy. The discharges are consistently unilateral when a brainstem neoplasm is responsible, may occur unilaterally or on both sides of the face at different times in multiple sclerosis.l" and may be bilateral in polyradiculoneuropathy.t! The pathophysiologic basis of myokymic discharges is unclear, but suggested mechanisms include abnormal excitability of lower motor neurons or peripheral nerves; ephaptic excitation; and the development of a rhythmic, oscillating, intra-axonal generator of action potentials." It has been suggested that membrane "bistability" of alpha motor neurons is sometimes responsible, the cell membrane having two equilibrium potentials, one at the resting potential and a higher one above threshold, generating maintained rhythmic firing. Further details may be found elsewhere.P Cramp discharges consist of the involuntary repetitive firing at high rates (up to 150 Hz) of motor unit action potentials in a large area of muscles, usually associated with painful contraction of the muscle. The number of motor unit action potentials and their discharge frequency increase gradually and subsequently decline as the muscle contraction ceases. Regularly and spontaneously discharging motor units at a frequency of 5 to 15 Hz have been described in relaxed muscles of patients with infantile spinal muscular atrophy, even during sleep." Neuromyotonic discharges consist of irregular bursts of high-frequency (up to 300 Hz) discharges of motor unit action potentials as doublets, triplets, or multiplets. The

fLV

L-.J 20 msec

discharges often have an abrupt onset and termination and usually last for a few seconds. Some of the action potentials are very short in duration. The amplitude of the discharges typically wanes. The discharges occur spontaneously or with needle movement, activity, ischemia, or percussion over the nerve, and they do not disappear during sleep. They are abolished by curare. In many instances, the discharges can also be abolished by proximal nerve block, whereas the extent to which more distal nerve block stops the discharges is variable, suggesting that the hyperexcitability sometimes arises in the distal nerve trunk." Neuromyotonic discharges may occur in patients with peripheral neuropathy of the axonal or demyelinative type (e.g., in chronic acquired demyelinating polyneuropathy" or multifocal motor neuropathy'") and on either a sporadic or a hereditary basis in patients with muscle stiffness and other neuromuscular symptoms (see p. 253) but no evidence of neuropathy (Isaacs' syndrome). They have also been described in patients with myasthenia gravis, raised titers of acetylcholine receptor antibodies, or thymoma, or in those receiving penicillamine.t" At least in some instances, they appear to relate to the presence of autoantibodies that reduce the number of functional voltage-gated potassium channels and thus lead to hyperexcitability of motor nerves and increased release of acetylcholine; other antigenic targets may also be involved.'?

EMCi Findings During Adivily MOTOR UNIT ACTION POTENTIALS

Any change in the number of functional muscle fibers contained in motor units will affect the parameters of motor unit action potentials. This is best exemplified by the character of the action potentials that are recorded from a muscle when the number of muscle fibers per unit is reduced. The mean duration of the action potentials is shortened because of the loss of some of the distant fibers that previously contributed to their initial and terminal portions, whereas their mean amplitude is reduced because of the loss of some of the fibers lying close to the electrode. If the spikes generated by the surviving muscle fibers of individual units are widely separated in time, there may also be an increased incidence of polyphasic potentials. As might be expected, therefore, the mean duration and amplitude of motor unit

242

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

action potentials are reduced in patients with myopathic disorders (see Fig. 11-3), and there is an increased incidence of polyphasic potentials. This temporal dispersion of spikes may relate to scatter of endplate regions within surviving muscle fibers, variation in diameter of the muscle fibers (so that the conduction velocity in the fibers is more varied), longitudinal fiber splitting, regeneration of muscle fibers, or some combination of these factors. Similar findings may also be encountered in patients with disorders of neuromuscular transmission (e.g., myasthenia gravis or Lambert-Eaton myasthenic syndrome) and during the reinnervation of muscle following severe peripheral nerve injury; however, the characteristic feature in such circumstances is the variability that the action potentials exhibit in their parameters, and especially in their amplitude, during continued activity (Fig. 11-8). A similar variability in amplitude and morphology of motor unit action potentials may be encountered in amyotrophic lateral sclerosis and progressive spinal muscular atrophy; it signifies active disease and thus a poor prognosis." In neuropathic disorders, it is the number of functional motor units that is reduced, and the average number of muscle fibers per unit actually may be increased if denervated muscle fibers are reinnervated by collateral branches from the nerve fibers of surviving units. The motor unit action potentials recorded in such circumstances are of longer duration than normal and may be polyphasic (see Fig. 11-3). This is because the activity recorded by the electrode is temporally more dispersed than normal, caused primarily by the greater anatomic scatter of endplates in the units but also by the reduced conduction velocity at which immature collateral branches conduct impulses. The action potentials may also have a greater amplitude than normal if the number of muscle fibers lying close to the electrode is increased; such potentials are often particularly conspicuous in patients with involvement of anterior horn cells in the cord. Polyphasic motor unit action potentials are found in limited numbers in normal muscle, and care must be exercised in attaching any pathologic significance to them unless they are present in excessive numbers and have abnormal dimensions. Low-amplitude, short-duration

polyphasic potentials are seen most characteristically in myopathies and myositis but are also found during early reinnervation after axonal loss and in disorders of neuromuscular transmission. Similarly, largeamplitude, long-duration polyphasic potentials are typical of disorders characterized by chronic partial denervation with reinnervation (e.g., motor neuron diseases) but may also occur in polymyositis and muscular dystrophies. Motor unit action potentials are sometimes followed by smaller potentials called satellite potentials. These can appear at any point after the termination of the motor unit action potential. Such potentials sometimes occur after an interval of about 15 to 25 msec or more (see Fig. 11-3). These late components of motor unit action potentials have been reported in both neurogenic disordersf and muscle diseases such as polymyositis and muscular dystrophy.f They can be explained by the presence of an ectopic endplate or by delayed conduction along unmyelinated collateral nerve sprouts innervating previously denervated muscle fibers; however, in the latter case a change in the latency of the late component would be expected to occur with sprout myelination. In muscular dystrophy, the denervated fibers probably arise by segmentation of existing muscle fibers or by muscle regeneration.t" ABNORMALITIES OF RECRUITMENT PATTERN

When the number of functional motor units is reduced, as in patients with neurogenic weakness, there may be a diminution in the density of electrical activity that can be recorded from affected muscles during a maximal voluntary contraction. In severe cases, it may be possible to recognize individual motor unit action potentials (Fig. 11-9). In such circumstances, there is an increase in the rate at which individual units begin to fire and also in the rate at which they must fire before additional units are recruited (i.e., in the recruitment frequency) which, as indicated earlier, is influenced by a number of variables but is usually between 5 and 20 Hz. Firing rates as high as 50 Hz are sometimes encountered. In patients with myopathic disorders, the number of functional units remains unchanged until an advanced

I I

.,

I L-...J 500 msec

III Variation ill amplitude of a motor unit action potential during continued weak voluntary activity of a reinnervated muscle.

FIGURE 11-8

] 200 fJ-V

Clinical Electromyography

FIGURE 11-9 III Activity recorded during maximal voluntary contraction of a partially denervated muscle.

stage of the disorder, and therefore the interference pattern remains full. Indeed, it may be more complete than normal for a given degree of voluntary activity because there is an increase in the number of units activated to compensate for the reduced tension that individual units (with their reduced fiber content) are able to generate. Motor unit firing rates tend toward normal unless the myopathy is severe, when the frequency of firing at onset and at recruitment of other units may be increased. When patients are unable to cooperate fully because of limited comprehension or pain, or for psychologic reasons, the interference pattern may be reduced during maximal voluntary effort because only a few units are activated and they fire at low rates (e.g., 10 Hz). The motor unit action potentials, however, are normal in configuration. The distinction of poor activation from the poor recruitment of neurogenic disorders is best made by the rate of firing of individual motor unit action potentials (slow in poor activation and rapid in poor recruitment) and by the ratio of the number of active motor units to the firing frequency of individual units (which is generally less than 5 and is relatively constant for individual muscles).

EMG FINDINGS IN VARIOUS CLINICAL DISORDERS Myopathic Disorders The EMG findings in myopathic disorders do not, in themselves, indicate the etiology of the underlying muscle disease. Indeed, the findings do not even establish with certainty that the underlying pathology is a primary disorder of muscle. During the reinnervation of muscle after a severe peripheral nerve lesion, for example, the EMG appearance may be similar to that in a myopathy. Nevertheless, EMG is of undoubted practical usefulness in patients with myopathies. 1. It is helpful in establishing the diagnosis, pro-

vided that the clinical context of the examination is borne in mind and the findings are integrated with the results of other laboratory procedures. 2. It is important in providing an indication about the extent and severity of the underlying disease. ~t It permits the course of the disorder and its response to treatment to be followed.

243

L-_----.J 200 msec

4. It may provide a guide to prognosis. 5. It is helpful in suggesting the best muscle to biopsy. The selected muscle should be homologous to one that is moderately affected electromyographically, but should not itself have been examined by a needle electrode because this may lead to histopathologic changes. The needle examination should involve the sampling of numerous muscles, especially proximal ones (e.g., the iliacus, glutei, spinati, and paraspinal musclesv'), when a myopathy is suspected. Certain midlimb muscles (e.g., the brachioradialis and tibialis anterior) also have a high yield. It is sensible to examine the muscles on only one side of the body, so that an appropriate muscle on the other side can be used for biopsy if required. An increased amount of insertion activity may be found in myopathic disorders, and abnormal spontaneous activity is sometimes present, particularly in patients with inflammatory diseases of muscle. Indeed, in the myotonic disorders, the characteristic EMG finding consists of spontaneous high-frequency discharges of action potentials that wax and wane in frequency and amplitude, thereby producing a sound like that of a dive-bomber on the loudspeaker. The features that are most helpful in the EMG recognition of myopathic disorders, however, relate to the character and recruitment pattern of motor unit action potentials. The number of short or small action potentials, or both, is characteristically increased, and many of the potentials are polyphasic. Such action potentials produce a characteristic crackle on the loudspeaker, and this aids in their detection. Because there is a reduction in the tension that individual motor units can generate, an excessive number of units are activated in weak contractions. During strong voluntary contractions a full interference pattern is seen except at advanced stages, when it may be reduced with loss of all of the muscle fibers in individual units. In assessing the EMG findings in patients with a suspected myopathic disorder, it must be appreciated that abnormalities may not be detected despite a meticulous search and quantitative analysis of the data so obtained. It may be, for example, that the pathologic process is patchy in distribution, that only certain types of muscle fibers are affected, or that the electrical properties of the muscle fibers are unaffected by the disease. By contrast, abnormalities are sometimes found when

244

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

histopathologic studies are normal, so that these two diagnostic approaches are best regarded as complementary procedures. Buchthal and Kamieniecka contrasted the electrophysiologic and histologic findings in 188 patients with a myopathy diagnosed on clinical grounds; they noted that the EMG findings accorded with the clinical classification in 87 percent of patients, whereas the histologic or histochemical findings (or both) did so in 79 percent of patients.P The findings in various myopathic disorders are discussed in the following sections, with attention directed to the manner in which they may differ from the changes just described. Further details are given elsewhere.w'? MUSCULAR DYSTROPHIES AND OTHER FAMILIAL MYOPATHIES

Electromyography has an important role in determining the distribution and severity of involvement in the muscular dystrophies, and in separating these disorders from others with which they may be mistaken. The usual EMG findings are as described earlier, although abnormalities are sometimes inconspicuous, particularly in the relatively benign forms of muscular dystrophy such as the limb-girdle or facioscapulohumeral varieties. Insertion activity is reduced in areas where muscle has been replaced by fatty tissue, and there are changes in the configuration, dimensions, and recruitment of motor units. Insertion activity is sometimes increased, however, and fibrillation potentials, positive sharp waves, and complex repetitive discharges may be found because of segmental necrosis of muscle fibers or the presence of regenerating fibers. The findings are generally more abnormal in Duchenne's than in Becker's dystrophy and may be especially mild in limb-girdle, facioscapulohumeral, scapuloperoneal, ocular or oculopharyngeal, and distal dystrophies. A review of the clinical and genetic aspects of distal myopathies can be found elsewhere.f In Emery-Dreifuss dystrophy, the findings sometimes suggest both myopathic and neurogenic processes. Singlefiber EMG (see Chapter 12) may also be abnormal, revealing increased fiber density and even increasedjittel' and blocking, especially in Duchenne's dystrophy but occasionally also in other varieties such as limbgirdle and facioscapulohumeral dystrophy." Attempts to identify female carriers of the gene of the sex-linked variety of muscular dystrophy by studying the parameters of motor unit action potentials, the refractory period of muscle fibers, or the pattern of electrical activity of muscle have met with only limited success. Advances in molecular biology now make these electrophysiologic approaches of little practical significance. In congenital muscular dystrophies, a heterogeneous group of autosomal recessive neuromuscular dis-

orders, EMG changes of a myopathy are typical and early findings in proximal muscles; an EMG performed at birth may, however, be normal. 50 An associated neuropathy may be evident in patients with a merosindeficient disorder and sometimes even when merosin expression is normal. INFLAMMATORY DISORDERS OF MUSCLE

Inflammatory myopathies are characterized clinically by weakness, tenderness, and-in severe cases-wasting of affected muscles. Proximal muscles tend to be affected more severely than distal muscles. Electromyography is helpful in establishing that weakness is myogenic and in distinguishing an underlying inflammatory process from other disorders (e.g., one of the muscular dystrophies) that have no specific treatment. In addition to changes in the configuration and dimensions of motor unit action potentials, abnormal spontaneous activity (fibrillation and positive sharp waves) is often profuse. The EMG abnormalities are patchy in distribution, however, so that the findings may vary in different muscles or in different parts of the same muscle. Abnormalities are more common in proximal muscles than in distal ones and are especially common in the paraspinal muscles. The EMG findings also vary with the stage of the disease and the activity of the inflammatory process. Determination of muscle-fiber conduction velocity may improve specificity of the electrodiagnostic evaluation, but it is not useful in discriminating inflammatory myopathies from other myopathic disorders and is not a standard technique at this time.f In polymyositis, the character and recruitment pattern of motor unit action potentials are similar to those seen in other myopathic disorders; however, late components are sometimes found, and some polyphasic potentials may be of long duration and large amplitude. 52 Insertion activity is commonly excessive, and spontaneous fibrillation potentials, positive sharp waves, and complex repetitive discharges are found more often and are usually much more conspicuous than in patients with muscular dystrophy. These findings of abnormal spontaneous activity are patchy in distribution but have a high prevalence in the paraspinal muscles and are usually prominent when the disease is in an active phase. The evolution ofEMG abnormalities with time is uncertain because virtually all patients receive treatment for the underlying disorder. As the disease progresses from acute to chronic stages, however, there is an increase in the proportion of longerduration large-amplitude motor unit action potentials" The long-duration polyphasic potentials in more chronic but less active disease are related to the presence of regenerating muscle fibers on muscle biopsy. Treatment is usually with steroids, but methotrexate,

Clinical Electromyography

azathioprine, or other immunosuppressive agents have also been used. The presence or absence of abnormal spontaneous activity in the EMG reflects the response to treatment: persisting fibrillation potentials imply a less favorable therapeutic response than otherwise. In patients with increasing weakness that develops during the course of steroid therapy, the EMG findings may help to distinguish between a steroid-induced myopathy and reactivation of the myositis. In the former circumstance, fibrillation potentials are likely to be absent in weak muscles, whereas in the latter context they are likely to be present profusely. Polymyositis may develop at either an early or a late stage of human immunodeficiency virus (HlV) infection. 54 •55 In addition, some HIV-infected patients may present with the clinical and electrophysiologic features of polymyositis but are found at muscle biopsy to have selective loss of thick filaments and rod-body formation, with only minor inflammatory changes; benefit may follow immunosuppressive therapy or plasmapheresis.""·5ti Treatment of HIV-associated disease with zidovudine may lead to a myopathic disorder that has the EMG features of an inflammatory process, but biopsy fails to reveal inflammatory exudatesY-59 Inclusion body myositis is a common inflammatory disorder,tiO,IH but it is without effective treatment. It is the most common chronic myopathy presenting after the age of 50 years and is slowly progressive. The quadriceps femoris and forearm muscles (long flexors of the fingers) are characteristically affected early and most severely, but there is some variation in this regard. ti2 Diagnosis is made by muscle biopsy. The EMG findings do not reliably distinguish it from the inflammatory myopathies discussed earlier, or from other inflammatory disorders such as trichinosis and toxoplasmosis.t" In some instances, motor unit action potentials are of increased amplitude, but the concomitant reduction in area or duration is compatible with myopathic remodeling of the motor unit. 63 In patients with polymyalgia rheumatica, the EMG findings are usually normal. ENDOCRINE AND METABOLIC MYOPATHIES

In patients with an endocrine myopathy, the electrophysiologic findings are similar to those of other myopathies, and abnormal spontaneous activity is usually not seen. Abnormalities are often inconspicuous in patients with a steroid-induced myopathy, however, presumably because it is the type 2 muscle fibers that are predominantly affected in this condition. In thyrotoxicosis, fasciculation potentials may be conspicuous. In hypothyroidism, insertion activity is sometimes increased, and spontaneous fibrillation and fasciculation potentials, together with trains of complex repetitive discharges, may also be found in rare

245

instances. 54 In interpreting findings in patients with endocrine disturbances, it must be remembered that a myopathy may coexist with other types of neuromuscular disorders, and that these may complicate the EMG findings. For example, patients with hypothyroidism are liable to develop peripheral nerve entrapment syndromes, whereas myasthenia gravis sometimes occurs in association with thyrotoxicosis. The EMG features of a myopathy may also be found in patients with osteomalacia, chronic renal failure, and a number of other less common metabolic disorders. In patients with hypokalemic periodic paralysis, the EMG findings between attacks may be normal or, less commonly, suggestive of a proximal myopathy. During the attacks, no abnormality of insertion or spontaneous activity is found, but motor unit action potentials are reduced in duration and number; the interference pattern during attempted voluntary contraction is diminished, and in severe cases there may be complete electrical silence. During attacks of hyperkalemic or normokalemic periodic paralysis, insertion activity is increased; spontaneous fibrillation potentials, myotonic discharges, and complex repetitive potentials may be found; and motor unit action potentials are reduced in duration and number." In the glycogen storage diseases caused by deficiencies of either phosphorylase (McArdle's disease) or phosphofructokinase, the EMG of the resting muscle is usually normal but occasionally is suggestive of a myopathy. No electrical activity can be recorded during the contractures that may develop during continued exercise. In patients with acid maltase deficiency (Pompe 's disease), EMG may reveal increased insertion activity; profuse, spontaneous fibrillation and positive sharp waves; myotonic discharges; and complex repetitive discharges. However, the motor unit action potentials are similar in character to those seen in other myopathies." With deficiency of debrancher enzyme (Cori's disease), the electrophysiologic findings may reveal "a mixed pattern," with abnormal spontaneous activity and short, polyphasic motor unit action potentials of normal or reduced amplitude in some patients, and large or long-duration polyphasic potentials in others.?" Myopathy is a rare manifestation of primary systemic amyloidosis and is usually then diagnosed by muscle biopsy. Needle EMG reveals findings similar to those of a chronic inflammatory myopathy. Fibrillation potentials are common, most often in the gluteus medius and paraspinal muscles.P? Motor unit action potentials may be of short duration and low amplitude, especially in proximal muscles; however, long-duration large potentials are sometimes found, and occasionally a mixed population of motor units is encountered. There may also be electrophysiologic evidence of an underlying peripheral neuropathy, which probably accounts for

246

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

the presence of large-amplitude, long-duration motor unit action potentials. MYOPATHIES CAUSED BY DRUGS OR ALCOHOL

Changes suggestive of a myopathy may be found in patients taking certain drugs. In many instances, however, the clinical or EMG features are more extensive than are those of a simple myopathy. Thus, the presence of abnormal spontaneous activity and the character of the motor unit action potentials suggest that a myositis is sometimes associated with use of cimetidine or Dpenicillarnine/" and the EMG features of a myopathy with increased muscle fiber irritability may also occur with epsilon-aminocaproic acid, emetine, hypokalemic agents, and certain beta-blockers. Colchicine may induce a myopathy, sometimes accompanied by a neuropathy,69-71 as may chloroquine. Disorders of neuromuscular transmission may occur with certain antibiotics, n-penicillamine, and chloroquine.F Among the hypocholesterolemic agents, myopathic changes may occur with clofibrate, the statin group of drugs, gemfibrozil, and niacin. In addition, acute rhabdomyolysis has been associated with the statin drugs and with gemfibrozil," and myotonic discharges are associated with diazocholesterol. 74 Chronic use of steroids sometimes leads to a myopathy, but the EMG findings may be normal. Among chronic alcoholics with either the acute reversible muscle necrosis that sometimes follows intoxication or the acute hypokalemic myopathy that develops during an alcoholic binge, EMG typically reveals spontaneous fibrillation, positive sharp waves, and motor unit action potentials that are small and of short duration. EMG features of myopathy may also be found in alcoholics developing a more slowly progressive proximal muscle weakness, and sometimes are found in alcoholics without any clinical evidence of myopathy.'" CRITICAL ILLNESS MYOPATHY

The occurrence of an acute weakness and muscle wasting in the context of fever, sepsis, and treatment with high-dose steroids, nondepolarizing neuromuscular blockers, or both, is now well-recognized,75,76 and it seems especially common in patients treated for status asthmaticus. Weakness tends to be diffuse in the limbs and may also affect the neck flexors, facial muscles, and diaphragm. The clinical features overlap those of critical illness polyneuropathy or prolonged impairment of neuromuscular transmission, emphasizing the importance of electrodiagnostic evaluation.?" The EMG findings include small, short-duration, polyphasic, motor unit action potentials and early recruitment of motor units; spontaneous fibrillation is sometimes conspicuous and cannot be used to distinguish critical illness

neuropathy from myopathy.76,78 In addition, compound muscle action potentials may be grossly attenuated, sensory nerve action potentials are sometimes reduced, and there may be a decremental response to repetitive nerve stimulation suggestive of a postsynaptic defect of neuromuscular transmission (see Chapter 15). The skeletal muscle may be inexcitable by direct electrical stimulation.Y'" CONGENITAL MYOPATHIES OF UNCERTAIN ETIOLOGY

A number of congenital myopathies have been described; these may be associated with specific structural changes that enable distinct entities to be recognized. Appropriate EMG changes may be found in some cases, but in others the findings are normal. In nemaline and centronuclear (myotubular) myopathy, however, abnormal spontaneous activity may be conspicuous. Nemaline myopathy may be manifest by the dropped-head syndrome, other causes of which include polymyositis, isolated neck extensor myopathy, amyotrophic lateral sclerosis, and myasthenia gravis.l'' MYOTONIC DISORDERS

The characteristic EMG feature in myotonic dystrophy type 1 (dystrophia myotonica, DM1) or type 2 (DM2, proximal myotonic myopathy, PROMM), the dominant and recessive forms of myotonia congenita, paramyotonia congenita, and Schwartz-Jampel syndrome.F is the occurrence of myotonic discharges that are evoked by electrode movement and by percussion or voluntary contraction of the muscle being examined. Motor unit potentials are normal in appearance except in DMI or DM2 and in the recessive type of myotonia congenita, in which an excess of small, short-duration or polyphasic potentials may be found, as in other myopathic disorders. Myotonic dystrophy is a dominantly inherited disorder, and EMG may reveal abnormalities in the clinically unaffected relatives of patients with the disease. Similarly, brief myotonic discharges may be found in at least one unaffected parent in approximately 67 percent of the families of patients with recessive myotonia congenita." RIPPLING MUSCLE DISEASE

An autosomal dominant myopathy characterized by muscle stiffness, hypertrophy, and rippling was described by Torbergsen in 1975,84,85 The selfpropagating, wave-like rippling of the muscles can often be induced by passive muscle stretch and is electrically silent, thus being a form of muscle contracture. Myalgia is sometimes conspicuous, and mild proximal weakness

Clinical Electromyography

may be present.t" Percussion of the muscle may induce a prolonged contraction resembling percussion myotonia and a localized mounding lasting for several seconds. A rare autosomal recessive form has been described and is associated with severe cardiac disease."? Sporadic cases of muscle rippling have also been reported and sometimes resemble familial cases." In other instances, muscle rippling may have a neurogenic basis and has been associated with myasthenia gravis. 88

Neuropathic Disorders Immediately after the development of an acute neuropathic lesion, EMG reveals no abnormality other than a reduction in the number of motor unit action potentials under voluntary control in affected muscles. A complete interference pattern is not seen during maximal effort despite an increase in the firing rate of individual units; in severe cases, there may be no surviving units, so that no electrical activity is recorded during attempted voluntary contraction. The subsequent changes depend on whether denervation has occurred. If it has, the amount of insertion activity increases after several days and abnormal spontaneous activity may subsequently be found, although its appearance may be delayed for up to 5 weeks, depending on the site of the lesion. In particular, fibrillation potentials are usually detected sooner when the lesion is close to the muscle than when a more distant lesion is present. As reinnervation occurs, spontaneous activity becomes less conspicuous and low-amplitude motor unit action potentials are seen. These potentials may exhibit a marked variability in their size and configuration, and some have a complex polyphasic configuration. Initially, the duration of these potentials is quite short, but it increases progressively as more muscle fibers come to be reinnervated; eventually, long-duration polyphasic potentials are found. With time, many potentials regain a normal appearance and the interference pattern becomes more complete, but the extent of any residual EMG abnormality depends on the completeness of recovery. In patients with chronic partial denervation, insertion activity is increased and spontaneous fibrillation, positive sharp waves, and complex repetitive discharges are found. Fasciculation potentials are often conspicuous in patients with diseases such as motor neuron disease or poliomyelitis, in which the lower motor neurons in the spinal cord are affected, but they may also be found with more peripheral lesions. The mean duration of motor unit action potentials is increased if reinnervation has occurred by collateral sprouting; and there may be an increased incidence of large units, especially in patients with involvement of anterior horn cells. In addition, an excessive number of polyphasic

247

motor unit potentials is usually encountered. There is an increase in the rate at which individual units begin firing and at which they must fire before additional units are recruited, and the interference pattern during maximal contractions is reduced. The EMG findings may have important medicolegal implications by suggesting the duration of a neuropathgic disorder. For example, if a patient claims to have developed a footdrop immediately after a medical or surgical intervention, and needle EMG at that time reveals abnormal spontaneous activity in the weak muscles, the lesion is likely to be at least 1 to 3 weeks old and may therefore have been present before the alleged time of injury. Similarly, the presence of longduration, large-amplitude polyphasic potentials points to a long-standing lesion because there has been time for reinnervation to occur. SPINAL CORD PATHOLOGY

Electromyography may help to define the level and severity of lower motor neuron involvement in patients with a cord lesion. Signs of chronic partial denervation are found in the affected muscles of patients with chronic myelopathies if the anterior horn cells are involved. Insertion activity is increased, and spontaneous fasciculations, fibrillation, positive sharp waves, and complex repetitive discharges may be found. The mean duration and amplitude of motor unit action potentials (and the number of polyphasic potentials) are increased, giant potentials may be encountered, and the firing rate and recruitment frequency of individual potentials is increased, whereas the interference pattern is reduced. Upper motor neuron involvement is indicated by poor activation of motor units: the number of units firing and their rate of firing are reduced during maximal volitional activity. By making it possible to define the precise segmental distribution of a lesion involving lower motor neurons, the EMG examination can aid in the localization of spinal cord lesions. Electromyography therefore has an important role in distinguishing motor neuron diseases from discrete, surgically remediable conditions affecting the spinal cord. The examination should be continued until it is clear whether the pattern of involved muscles can be accounted for by a restricted cord lesion; proximal and distal muscles supplied by different roots and peripheral nerves are examined, as are muscles in clinically uninvolved limbs. Fibrillation potentials are usually most conspicuous in weak, wasted muscles. Four patients with slowly progressive, asymmetric weakness and fasciculations of the lower limbs but insignificant sensory findings were recently described and exemplify such a circumstance.r" Their clinical presentation was suggestive of early motor neuron disease, with some accompanying findings of upper

248

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

motor neuron involvement. Needle EMG indicated chronic and often active denervation in the weak muscles of the legs with, in two instances, abnormalities also in the paraspinal muscles; examination of the upper limbs was normal. Motor and sensory nerve conduction studies were normal except for small compound muscle action potentials from weak muscles. Imaging studies indicated spondylotic lumbar myelopathy with disc herniation or osteophytic formation, and the purely motor presentation may have resulted from ischemic myelopathy secondary to compression of the anterior spinal arter y.'" The term poliomyelitis is used with regard to an acute flaccid paralysis associated with viral infection. It is not necessarily synonymous with poliovirus infection but may, for example, also embrace infection with West Nile virus as well as enteroviruses, echoviruses, and coxsackieviruses.?" Recent studies on patients with poliomyelitis have focused particularly on those with West Nile virus infection. There is asymmetric acute denervation of muscles and small or absent compound muscle action potentials of affected muscles. In other instances, patients complain oflocalized or generalized weakness and of a sense of muscle fatigue, but EMG and other electrodiagnostic studies reveal no abnormalityY°,\1I In amyotmphic lateral sclerosis, EMG has an important role in confirming lower motor neuron involvement in clinically affected areas, detecting such involvement in clinically unaffected areas, providing evidence for a slowly progressive disorder, and indicating that abnormalities are too widespread to be explained by a single focal lesion. The EMG abnormalities are widespread. Accordingly, there should be evidence of chronic partial denervation and reinnervation in at least three limbs, or in an upper limb, lower limb, and bulbar muscles before this diagnosis is made. The EI Escorial criteria of the World Federation of Neurology advise that electrophysiologic confirmation of the diagnosis requires evidence of acute and chronic partial denervation in at least two muscles supplied by different roots or spinal nerves and by different cranial or peripheral nerves in at least two distinct anatomic regions (bulbar, cervical, thoracic, lumbosacral) .92 Abnormalities of the facial, tongue, or sternocleidomastoid muscles provide evidence of bulbar involvement." Patients with involvement of the small muscles of the hand sometimes show preferential or greater involvement of the abductor pollicis brevis than the abductor digiti minimi muscle." Fasciculations are often conspicuous. Motor units are recruited abnormally, firing frequency may be increased (except when upper motor neuron involvement predominates), and repetitive or double discharges may occur. Occasionally, when motor units having a high threshold for activation are lost early, apparently normal recruitment may be seen with two or three

motor units firing at rates less than 10 Hz, but additional units will not be recruited with increasing effort/" An excess of large-amplitude, long-duration polyphasic potentials is characteristic and indicates reinnervation of denervated fibers. In late or rapidly progressive stages, however, low-amplitude, shortduration potentials may be present if reinnervation fails to keep pace with denervation. In advanced cases, a single motor unit may be found in severely affected muscle firing at rates as high as 50 Hz, or there may be no remaining motor units under voluntary control. Marked variation in the configuration of motor unit action potentials from moment to moment is sometimes seen because of impulse blocking or failure of neuromuscular transmission, and single-fiber EMG may correspondingly reveal increased jitter and impulse blocking.t? Such findings," or EMG evidence of denervation in the diaphragm, indicate a poor prognosis. Serial estimates of motor unit number (Chapter 12) may also have prognostic value. The findings on needle EMG may help to diagnose amyotrophic lateral sclerosis, but they do not permit distinction of this disorder from other motor neuron diseases (e.g., that occurring with lymphoma.f" hexosaminidase deficiency, antiganglioside antibodies, paraproteinemia, or progressive spinal muscular atrophy) or from disorders simulating an anterior horn cell disease (e.g., multifocal motor neuropathy or organophosphate neurotoxicity) . The spinal muscular atrophies may be hereditary or acquired, and EMG is important in localizing the pathology to the anterior horn cells and in guiding prognosis. The EMG findings are similar to those just described, with evidence of chronic partial denervation and reinnervation, but fasciculations are less common than in amyotrophic lateral sclerosis. In Sobue's disease or monomelic amyotrophy, clinical involvement is usually restricted to one of the upper limbs; the EMG findings are correspondingly restricted in distribution, but abnormalities are commonly also present on the clinically unaffected side, suggesting that bilateral involvement is not unusual." Bulbospinal neuronopathy is an X-linked, recessively inherited, spinal muscular atrophy in which a characteristic twitching of the chin may occur spontaneously or with voluntary activity such as pursing of the lips. The EMG findings in affected muscles are as described earlier, but the twitching of the chin is associated with grouped discharges of motor units resembling myokymic discharges or with 20- to 40-Hz repetitive discharges of individual motor units that may last for several seconds/" In addition, sensory nerve action potentials are often small or absent. In syringomyelia, the electrodiagnostic findings are usually nonspecific, but various types of spontaneous EMG activity may be encountered including continuous

Ciinical Electromyography

motor unit activity with the firing frequency of individual potentials ranging from 8 to 13 Hz, respiratory synkinesis (bursts of motor unit action potentials in one or more limb muscles during inspiration), and myokymic discharges in limb or paraspinal muscles.f" Respiratory synkinesis and myokymia seem to be present only at an advanced stage of the disorder. Spinal cord disorders usually can be distinguished from peripheral nerve or plexus lesions by the pattern of muscle involvement, but measurement of motor and sensory conduction velocity also may be helpful in this respect. Maximal motor conduction velocity is normal or only slightly slowed with pathology restricted to the cord, although compound muscle action potentials (M waves) may be small, especially when elicited from weak, wasted muscles; sensory conduction studies are normal. Repetitive motor nerve stimulation may yield a small decrement in size of the M wave in amyotrophic lateral sclerosis because of unstable neuromuscular transmission along collateral nerve terminal sprouts; this is held to indicate active disease and a poorer prognosis." It is sometimes hard to distinguish a discrete spinal cord lesion from a root lesion, because in both instances EMG abnormalities may have a segmental distribution and changes may be found in the paraspinal muscles. In the former, however, several segments may be involved and bilateral changes usually can be expected. Electromyography per se does not provide any direct information regarding the pathology of the underlying abnormality. ROOT LESIONS

Despite the advent of sophisticated imaging techniques, EMG has an important role in the evaluation of patients with suspected root lesions. It detects both compressive and noncompressive radiculopathies, indicates the clinical relevance of structural abnormalities that are a common and often incidental finding.l'" provides a prognostic guide, and can be used to follow the course of the disorder. It therefore complements, rather than substitutes for, imaging studies and has a comparable diagnostic yield. For example, in a study of 47 patients who underwent both EMG and magnetic resonance imaging of the spine within 2 months of each other,'?' Nardin and associates found that 55 percent had an EMG abnormality and 57 percent had an imaging abnormality that correlated with the clinical level of the lesion. The two studies agreed in 60 percent of instances, with both normal in 11 patients and both abnormal in 17; only one study was abnormal in 40 percent of patients, however, indicating the complementary nature of the studies. In patients with axonal degeneration, EMG signs of partial denervation may be found in muscles supplied by the affected segment. The most helpful sign is the

249

presence of fibrillation potentials. These tend to appear earlier in more proximal muscles, appearing initially in the paraspinal muscles, then in the proximal, and subsequently in the distal limb muscles. They disappear in the same sequence as reinnervation occurs. This generally accepted concept has been questioned by some authors, however, who found no relationship between the presence of abnormal spontaneous activity in the paraspinal muscles and duration of symptoms in patients with cervical radiculopathy.l''? Complex repetitive discharges in a myotomal distribution may also be found in chronic radiculopathies but are rarely the sole abnormality. Similarly, fasciculations are encountered occasionally; when found as the sole abnormality, however, their distribution is widespread and they usually reflect generalized benign fasciculations or motor neuron disease. Motor unit action potentials may be decreased in number and fire at an increased rate, and the incidence of large, long, polyphasic potentials may be increased if the lesion is long-standing. Reliance for diagnosing a radiculopathy must not be placed solely on increased polyphasicity, even when in a myotomal distribution. The paraspinal muscles should always be examined, and the number of limb muscles examined must be sufficient to distinguish a root lesion from more peripheral (i.e., plexus or peripheral nerve) involvement by the distribution of electrical abnormalities. Radiculopathy may be diagnosed on EMG grounds when abnormalities are present in at least two muscles supplied by the same nerve root but by different peripheral nerves, and abnormalities are not present in muscles supplied by normal roots adjacent to the involved one. It is not necessary for all of the muscles in the myotome to be affected in order to diagnose a radiculopathy. Muscles may be electrophysiologically spared in patients with suspected root involvement because the lesion is only partial, because the muscles are examined at the wrong time (so that changes of denervation have not yet developed or have disappeared with reinnervation), or because the initial diagnosis was incorrect. They may also be spared because of inaccuracies in myotornal charts depicting the segmental innervation of individual muscles. Tsao and colleagues examined the segmental innervation of muscles supplied by the L2-S1 segments by comparing the surgical, imaging, and electrodiagnostic findings in patients with unilevel radiculopathies.l'" They found that the tibialis anterior was predominantly innervated by the L5 segment, both heads of the gastrocnemius by 51, and the two heads of biceps femoris exclusively by SI. Their findings conflict with some earlier studies, as summarized in their report. In a similar study, the pattern of muscle involvement in cervical radiculopathies was examined.l'" With single-level root lesions of C5, C7, and C8, changes were relatively stereotyped: The spinati, deltoid, biceps, and brachioradialis were

250

ElECTRODIAGNOSIS IN CLINICAL NEUROLOGY

involved with C5; pronator teres, flexor carpi radialis, triceps, and anconeus with C7; and first dorsal interosseous, abductor digiti minimi, abductor pollicis brevis, flexor pollicis longus, and extensor indicis proprius with C8. The findings with C6 lesions were variable and resembled those in patients with lesions at either the C5 or C7 level. Several muscles supplied by the root in question should be examined, and muscles supplied by adjacent roots and by more peripheral structures must also be evaluated. The importance of this is shown most clearly by means of a simple example. Neurogenic weakness of the elbow, wrist, and finger extensors may result from a lesion in anyone of several sites, and the examination of a patient with such a deficit must be meticulous for correct localization of the lesion. EMG evidence that flexor carpi radialis (C6, C7) is involved but brachioradialis (C5, C6) is spared would favor a lesion of the C7 root or the middle trunk of the brachial plexus; conversely, involvement of brachioradialis but sparing of flexor carpi radialis would suggest a radial nerve lesion unless the deltoid (C5, C6) muscle is also affected, in which case the posterior cord of the brachial plexus may well be involved. It is helpful to examine the paraspinal muscles. These are supplied from the spinal roots by their posterior primary rami, whereas the limb plexuses and peripheral nerves are derived from the anterior primary rami. Abnormalities in the paraspinal muscles are therefore common in patients with a radiculopathy, and fibrillation potentials are generally found in them before the limb muscles; they would not be expected in patients with a plexus or peripheral nerve lesion. However, fibrillation potentials will disappear first from these muscles after an acute radiculopathy, as reinnervation occurs. Furthermore, no conclusion should be reached about the level of the lesion from the findings in these muscles because there is a marked overlap in the territory supplied by the posterior primary rami. Finally, paraspinal fibrillation potentials are not specific for compressive radiculopathy. They occur also, for example, with inflammatory or diabetic radiculopathies, myelopathies, arachnoiditis, amyotrophic lateral sclerosis, myopathies, myositis, malignant disease, and after back surgery, and they become increasingly common in asymptomatic subjects with advancing age. lOb Diabetic thoracoabdominal radiculopathy presents with gradual onset of severe burning pain that sometimes shows some relation to posture or activity and may be confined to the front of the trunk or involve the back as well. It often involves multiple, usually adjacent, dermatomes. In almost all cases, fibrillation potentials are found in the paraspinal muscles, as well as in the abdominal or intercostal muscles if these are examined. Evidence of a concomitant polyneuropathy may also be found. lOn , l ll7 In most instances, diabetic amyotrophy is

probably caused by a polyradiculopathy or radiculoplexopathy; painful, progressive, asymmetric wasting and weakness of the thigh muscles is characteristic and may be associated with other neuromuscular complications of diabetes such as a distal polyneuropathy. The EMG findings indicate that the deficit is more extensive than can be attributed to a femoral neuropathy, which was once held responsible for the syndrome. A thoracoabdominal polyradiculopathy has been described in sarcoidosis in rare instances. The EMG findings are of active denervation in a pattern consistent with multiple root involvement. 108 Herpes zoster also can present with dermatomal pain on the trunk and motor disturbances with fibrillation potentials in paraspinal muscles. Other electrophysiologic studies are generally not as helpful as needle EMG in patients with root lesions. \09,1 III Motor conduction studies are normal, and M-wave responses usually are not reduced in size unless several adjacent roots are involved (as in cauda equina syndrome following central disc protrusion) or when, as occurs occasionally, the muscle from which the recording is made has atrophied. Sensory conduction studies are normal because the pathology is proximal to the dorsal root ganglia (see Chapter 13). H reflexes (in some Sl radiculopathies) and F-wave responses may be abnormal (see Chapter 16) but provide no indication of the site of the lesion, which can be anywhere along the length of the pathway that is tested. Dermatomal somatosensory evoked potentials (see Chapter 25) are sometimes abnormal. PLEXUS LESIONS

The brachial plexus is formed from the anterior primary rami of the C5, C6, C7, C8, and T1 roots, with a variable contribution from C4 (prefixed) and T2 (postfixed plexus) (Fig. 11-10). The C5 and C6 roots combine to form the upper trunk, C7 continues as the middle trunk, and C8 and T1 join to form the lower trunk. The trunks traverse the supraclavicular fossa, and at the upper border of the clavicle each divides into anterior and posterior divisions. The three posterior divisions then combine to form the posterior cord of the plexus, whereas the anterior divisions of the upper two trunks unite to form the lateral cord and the anterior division of the lower trunk continues as the medial cord. The cords give rise to the main peripheral nerves of the arm. The lumbar portion of the lumbosacral plexus is derived from the anterior primary rami of the first four lumbar roots, whereas the sacral portion is formed from the lumbosacral trunk (L4 and L5) and the anterior primary rami of the first four sacral roots. Needle EMG may be helpful in assessing the functional integrity of a nerve plexus, in determining the extent and severity of a plexus lesion, and in

Clinical Electromyography

251

Suprascapular nerve-pure C5 to supraspinatus (abduction) and infraspinatus (ext. rotation) ~--......... ~",,----~ of shoulder Lateral anterior thoracic nervepectoralis major (abduction)-G7 Musculocutaneous nerve-to biceps, brachialis, and coracobrachialis (elbow f1exion)-C5 (C6 contribution variable) 1111 Anatomy of the brachial plexus. (From Patten JP: Neurological Differential Diagnosis. Springer-Verlag, New York. 1977, with permission.)

FIGURE 11-10

Median nerve C6 pronators in forearm C7 wrist flexors C8 long-finger flexors (1,2) 01 small hand muscles

Long thoracic nerve to serratus anterior (abduction of scapula) ? mainly C5

Medial anterior thoracic nervepectoralis major and minor (shoulder adduction)-C8 Subscapular nerves (2)subscapularis and teres major Axillary nerve(shoulder inversion)-C5 deltoid and teres minor (shoulder Thoracodorsal nervelatissimus dorsi abductio~ / Ulnar nerve (shoulder adduction)-G7 Radial nerve C6 supinators (brachioradialis, C8 ulnar wrist flexion! long-finger flexors (3,4) supinator) 01 ulnar innervated small hand C7 triceps (elbow extn) muscles and wrist extensors C8 long-finger extensors

distinguishing such a lesion from root or peripheral nerve pathology. The muscles examined should include those supplied by each of the main peripheral nerves and spinal roots in question so that the site of the lesion can be determined by the pattern of involved muscles. The importance of examining the paraspinal muscles when one attempts to distinguish between a root lesion and a plexus lesion has already been stressed. Proximal involvement is further indicated by EMG evidence of denervation in muscles whose nerve supply arises before the brachial plexus (e.g., the rhomboids or serratus anterior muscles). Such a distinction can sometimes also be made by stimulation techniques in which appropriate sensory nerve action potentials are recorded in the extremities (see Chapter 13). These potentials may be small or absent in patients with lesions that are distal to the dorsal root ganglia and have caused afferent fibers to degenerate, but they are preserved in patients with radiculopathy because the lesion is more proximal and peripheral sensory fibers therefore remain intact.

With mild plexopathies, the only electrophysiologic abnormality may be the presence of fibrillation potentials in affected muscles, but not all muscles in the territory supplied by the involved structure necessarily contain them. In more advanced cases, the amplitude of sensory nerve action potentials diminishes, and with increasing severity the amplitude of the M wave to affected muscles also declines. These principles are well exemplified by the findings in patients in whom the lower trunk of the brachial plexus, or the anterior primary rami of the C8 and Tl nerve roots, are compressed or angulated by a cervical rib or band. The muscle wasting that results may be restricted to, or especially conspicuous in, the lateral part of the thenar pad, but careful clinical and EMG examination shows that motor involvement is more extensive than can be accounted for by a median nerve lesion alone, conforming instead to the distribution of muscles supplied by the C8 and Tl segments. Moreover, sensory nerve action potentials may be small or absent when recordings are made from the ulnar nerve at the wrist after stimulation of its digital fibers in

252

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

the little finger, indicating that the lesion is distal to the dorsal root ganglia, III and the latency of the F wave recorded from the hypothenar muscles following ulnar nerve stimulation at the wrist may be increased. l1l ,ll2 The electrophysiologic findings are usually normal in patients with suspected thoracic outlet syndrome when there are no objective neurologic signs.!'! In traumatic brachial plexopathy, EMG is performed to localize the lesion, determine its severity, and guide prognosis. In mild lesions, motor unit recruitment is reduced and sparse fibrillation potentials may be present in affected muscles. With more severe trauma, motor unit recruitment is impaired, motor unit action potentials are lost, and fibrillation potentials are found in the territory supplied by the damaged structures. The axonal dysfunction also may lead to reduced amplitudes of sensory nerve action potentials and, depending on the severity of the lesion, to small M responses with stimulation of the motor nerves to weak muscles. Conduction velocity usually is not reduced unless only a few fibers remain intact or just a few poorly myelinated regenerating axons are functional. For prognostic purposes, it is important to distinguish a plexopathy from root avulsion. In this regard, EMG abnormalities in the paraspinal muscles and preserved sensory nerve action potentials suggest that the lesion involves the nerve roots, whereas paraspinal abnormalities combined with attenuated sensory nerve action potentials imply a combined lesion. Sufficient time (10 days or so after injury) must be allowed for paraspinal abnormalities to develop before it is concluded that the EMG examination is normal. In idiopathic brachial plexopathy (neuralgic amyotrophy or Parsonage-Turner syndrome), similar changes are found on needle EMG, and there may be abnormalities of sensory and motor nerve conduction studies indicating single or multiple lesions of peripheral nerves'!" that are not necessarily at the level of the brachial plexus. The musculocutaneous nerve is involved more often than the ulnar and median nerves. EMG abnormalities may be found bilaterally even though only one side is clinically involved. The pattern of EMG abnormalities sometimes suggests that an individual nerve (e.g., the axillary, radial, or suprascapular) is affected, with little or no involvement of muscles supplied from the same level of the plexus through other nerves. Phrenic neuropathy may occur, leading to unilateral or bilateral diaphragmatic paralysis; needle EMG of the diaphragm then shows abnormal spontaneous activity and reduced numbers of (or absent) motor unit action potentials on inspiration.U" In radiation plexopathy, needle EMG may reveal myokymic discharges, without clinical evidence of myokymia, in addition to signs of denervation. Electromyography is probably important as a means of determining the prognosis of obstetric lesions of the brachial plexus. Evidence of denervation found during

the first week of life does not necessarily indicate an antenatal lesion but may simply result from the short length of the affected axons. Examination at about 3 months of age (when surgical intervention to restore nerve continuity becomes a consideration) may reveal motor unit action potentials in clinically paralyzed muscles, and thus that functional connections already exist between the spinal cord and affected muscles. The cause of this apparent paradox is unclear, but several possible explanations have been proposed. I 15 Lumbar plexopathies often can be distinguished from radiculopathies by examination of the paraspinal muscles and from femoral neuropathy by sampling the thigh adductor muscles, which are supplied by the obturator nerve. Sacral plexopathies can similarly be distinguished from multiple lumbosacral radiculopathies by examination of the paraspinal muscles for the presence of fibrillation potentials. Sensory nerve action potentials in the legs are generally small or absent with sacral plexopathies but may also be small in elderly patients. PERIPHERAL NERVE LESIONS

The EMG distinction of peripheral nerve from other neuropathic lesions has been discussed. In addition to the detection of signs of denervation in muscle supplied by the individual peripheral nerves, motor and sensory conduction can be studied in these nerves; this is discussed in Chapter 13, where the evaluation of polyneuropathies is also considered. In the axonal neuropathies, there may be EMG signs of chronic partial denervation in distal muscles (with or without accompanying evidence of reinnervation), whereas in the demyelinative neuropathies the only or most conspicuous abnormality on the needle examination is reduced recruitment of motor units. The EMG examination should be part of the electrodiagnostic evaluation of patients with suspected entrapment neuropathies. In some patients with symptoms of carpal tunnel syndrome, for example, the needle examination may suggest another diagnosis (e.g., a proximal median neuropathy) when results of the nerve conduction studies are not definitive. I 16 In addition to its diagnostic relevance, EMG may be of prognostic importance in this context. Needle EMG has an important role in providing a guide to prognosis after peripheral nerve injuries and in following the course of recovery. After injuries in which the function of a nerve is temporarily deranged but its structure remains intact, needle EMG reveals only a reduction in the number of motor units under voluntary control. If the anatomic continuity of nerve fibers is interrupted, however, the amount of insertion activity increases after a few days and abnormal spontaneous activity is eventually found (as described on page

Clinical Electromyography

247, where the subsequent changes that occur with regeneration and reinnervation are also considered). In evaluating patients with peripheral nerve injuries, needle EMG may provide evidence that some motor units remain under voluntary control after an apparently complete nerve lesion; this implies a more favorable prognosis than otherwise, provided that the possibility of these units being innervated anomalously has been excluded. Needle EMG is an important method of determining whether denervation has occurred, and this too has prognostic significance. Finally, needle EMG may indicate at an early stage whether recovery is occurring after a complete palsy, because voluntary motor unit activity reappears long before any signs of clinical recovery.

Disorders of Neuromuscular Transmission The EMG findings in patients with disorders of neuromuscular transmission are discussed in Chapter 15. The only point that need be made here is that individual motor unit action potentials may show a marked variation in their dimensions and configuration because of blocking of impulse transmission to individual muscle fibers within the motor unit. Such variability is therefore a common finding in patients with myasthenia gravis, Lambert-Eaton syndrome, and botulism, and is also encountered as reinnervation occurs after a neuropathic lesion and following muscle trauma. There may also be an increased incidence of short-duration or polyphasic motor unit action potentials in patients with myasthenia gravis, caused by a reduction in the number of functional muscle fibers per unit or by the development of a secondary myopathy. Similar findings may be encountered in Lambert-Eaton syndrome or botulism because of the defect in neuromuscular transmission.

Miscellaneous Disorders Intermittent bursts of high-frequency, repetitive motor unit discharges are found in different parts of the muscle in patients with cramps. These bursts have an abrupt onset and termination. The number of active motor units and their firing frequency increase as the cramp evolves, declining again as the cramp eases off. In patients with myokymia, EMG may reveal prolonged bursts of repetitive motor unit action potentials, or repetitive multiplets containing from 2 to more than 200 action potentials with intervals of about 20 msec between component spikes. Single or double discharges of individual motor units occur spontaneously and regularly at intervals of 100 to 200 msec in patients with facial myokymia, or motor units may fire

253

intermittently for up to 900 msec at intervals of 2 to 4 seconds at rates of 30 to 40 Hz. The findings in patients with myotonia or cramps differ from the spontaneous, high-frequency (up to 300 Hz) motor unit action potential discharges (neuromyotonic discharges, p. 241) seen in patients with Isaacs' syndrome. This disorder is characterized clinically by muscle cramps and stiffness, difficulty in muscle relaxation, muscle twitching, and hyperhidrosis, and may develop spontaneously; on a hereditary basis; or in association with lymphoma, Il7 peripheral neuropathy, or certain immunologic disorders such as myasthenia gravis or a thymoma. 37•4o Electromyography reveals neuromyatonic discharges that may be continuous in severe cases, but there is a marked variability in the configuration and dimensions of the action potentials that are found, some being of particularly low amplitude and short duration. us The spontaneous discharges are increased temporarily by voluntary activity, but a brief period of electrical silence may follow strenuous, continuous activity; they commonly disappear after proximal nerve block but may persist despite distal block, disappear after intravenous curare or succinylcholine, and are reduced by phenytoin. Two types of abnormal movements have been described in patients with hemifacial spasm. The first consists of brief twitches that affect several different muscles simultaneously and is accompanied electromyographically by isolated bursts of repetitive, highfrequency motor unit discharges, each burst consisting of discharges from the same unit. The second consists of a prolonged, irregular, fluctuating contraction during which motor units fire irregularly at lower frequencies, although bursts of activity identical to those just described are also seen. The EMG abnormalities of hemimasticatory spasm are similar to those of hemifacial spasm, but they are found in the jaw-closing muscles supplied by the trigeminal nerve. I 19 During the spasms of tetany, motor unit action potentials may be seen to fire repetitively in doublets, triplets, or multiplets. The number of activated units increases as the spasms intensify, and a full interference pattern may eventually be obtained. The syndrome of painful legs and moving toes is uncommon but well described. The EMG accompaniments of the continuous, involuntary toe movements that typify the disorder are varied, but may include myokyrnic discharges, 120 normal firing of motor units,12I and the rhythmic firing of motor units at rates between 0.5 and 3 Hz.122 Motor unit discharges may alternate in agonist and antagonist muscles or occur synchronously in both groups.'!' The stiff-man syndrome (Moersch-Woltman syndrome) is characterized by muscle stiffness and spasms that begin in axial muscles and then become more extensive, involving especially the proximal muscles in the limbs.P"

254

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

It has an autoimmune basisl 24 and relates to impaired central presynaptic inhibition. Electromyography reveals continuous motor unit activity that disappears during sleep, with spinal or general anesthesia, and after peripheral nerve block. Nerve conduction studies are normal, but exaggerated and widespread muscle responses follow electrical stimulation. Contractures are characterized clinically by a painful, involuntary shortening of muscle, and electromyographically by electrical silence.

DIAPHRAGMATIC ELEaROMYOGRAPHY Needle examination of the diaphragm may be helpful in clarifying the nature of respiratory disturbances in patients with neuromuscular diseases, as discussed in Chapter 33. It is probably best avoided, however, in agitated or uncooperative patients, and in those in whom it may be more hazardous or difficult (e.g., patients with coagulopathies, local infections or neoplasms, hiccups, frequent coughing, abdominal distention, or gross obesity). The reader is referred to the original sources for technical details of the several approaches that have been described. The subcostal approach is safe in that no major complications have followed it, but the examination is technically unsatisfactory in about 10 percent of cases.125.126 The easier approach favored by Bolton's group involves insertion of a needle electrode through an intercostal space just above the costal margin, but pneumothorax has occurred in a few patients with chronic obstructive pulmonary disease.P? The substernal approach is an alternative, but experience with it is limited.F'' The EMG findings should indicate whether a neuropathic or myopathic process is responsible for diaphragmatic weakness, but interpretation may be difficult because the motor unit action potentials of the diaphragm are normally smaller and more abundant than those of the limb muscles. The presence of abnormal spontaneous activity may provide evidence of partial denervation, whereas a reduced interference pattern during inspiration may occur from a demyelinating neuropathy of the phrenic nerve. In amyotrophic lateral sclerosis, the EMG findings may provide a guide to prognosis and planning supportive care. 129

SPHINaERIC ELECTROMYOGRAPHY It is possible to examine the function of the sphincteric muscles by needle EMG, but a variety of other physiologic approaches are also available for this purpose. Further details are provided in Chapter 30.

LARYNGEAL ELEaROMYOGRAPHY The utility of laryngeal EMG in the diagnosis, prognosis, and treatment of certain laryngeal disorders has recently been evaluated by an evidence-based review of the voluminous literature that has accumulated. 130 Laryngeal EMG may be useful as an aid to the injection of botulinum toxin into the thyroarytenoid muscle for the treatment of adductor spasmodic dysphonia, but there is only anecdotal evidence about the possible utility oflaryngeal EMG for other purposes. The technique is usually undertaken by otolaryngologists and merits no further discussion here.

QUANTITATIVE ASPEaS OF ELECTROMYOGRAPHY In an endeavor to improve the accuracy, reliability, and speed with which the examination is performed, and to gain further insight into the pathophysiology of neuromuscular disorders, different workers have developed a number of quantitative techniques over the years. These include techniques for measuring the parameters of individual motor unit action potentials, a variety of which are incorporated into newer commercially available equipment. These techniques and others that have been developed for the measurement of motor unit territory, for frequency analysis of the EMG to facilitate the diagnosis of muscle disease, and for estimating the number of motor units in a muscle are discussed in Chapter 12. Single-fiber electromyography is a particularly important quantitative approach and merits some comment here, although it is described in detail in Chapter 12. In brief, muscle fiber action potentials are recorded from a needle electrode with an electrode surface that is 25 urn in diameter, mounted in its side; this is inserted into the muscle while the latter is under slight voluntary activation. The action potentials have a constant shape for consecutive discharges, provided that the time resolution of the display is 10 J.lsec. They are essentially biphasic and often are followed by a small, long-duration terminal phase. They are commonly 5 to 10 mV in amplitude and have a total duration of less than 2 msec; the rise time of the positive-negative deflection is less than 200 usee when its source is close to the recording site. With the electrode appropriately positioned, activity is usually recorded from only one muscle fiber in an individual motor unit. In a certain number of cases, however, activity can be recorded from two or more muscle fibers belonging to the same motor unit (Fig.ll-ll). In such circumstances, the time interval between the recorded action potentials depends on differences in conduction time along the nerve and muscle fibers and on the anatomic

Clinical Electromyography

255

FIGURE II-II III The jitter. Electrode E is recording activity from two muscle fibers belonging to the same motor unit. A, Recording from a normal muscle. Band C, Recording from a myasthenic. The oscilloscope sweep is triggered by the first action potential. Interval variability between the action potentials is seen as a variable position of the second potential. In the upper row, 10 to 15 action potentials are superimposed. In the lower row, the oscilloscope sweep is moved downward. A, Normal jitter. B, Increased jitter but no impulse blockings. C, Increased jitter and occasional blockings (arrows). Calibration: 500 usee, (From Stalberg E: Single fiber electromyography. Disa Electronics, 1974, with perrnission.)

localization of motor endplates. A temporal variability, the jitter, is found between the two action potentials at consecutive discharges; this is mainly caused by variation in the neuromuscular transmission time in the two motor endplates involved. The jitter can be expressed numerically as the mean value of consecutive differences (MeD) of 50 to 100 interpotential intervals; it is normally between 5 and 50 usee, depending on the muscle examined. Its value does not depend on the recording site, suggesting that variability in the conduction velocity of muscle or nerve fibers contributes only a minor proportion of the jitter. Measurement ofjitter is a sensitive means of evaluating neuromuscular transmission. When there is uncertain transmission in terminal nerve twigs or immature motor endplates (e.g., as occurs following reinnervation), jitter is increased and there may be an intermittent block of transmission. In patients with myasthenia gravis, increased jitter and impulse blocking are the expected findings in muscles affected clinically. In patients with myopathy, a slight increase in jitter may be seen in 10 to 15 percent of the recordings, and blockings may be found in 5 to 10 percent." Jitter is normal in myotonia congenita but is increased in myotonic dystrophy. It may be increased or, occasionally, reduced (because of split muscle fibers) in Duchenne's, limb-

girdle, or facioscapulohumeral muscular dystrophy; the findings in polymyositis vary with the stage of the disease, but jitter is typically increased. In neurogenic disorders, the jitter is markedly increased and frequent blocking is found. When recordings are made from three or more muscle fibers at the same time, it is sometimes apparent that two or more of the components will block only together, and they show a common large jitter in relation to the rest of the action potential complex. This is caused by impaired or blocked transmission in the distal axonal branch supplying the muscle fibers whose action potentials are affected in this way.?" Increased jitter and impulse blocking may also occur in certain normal "series-fibered" muscles because of doubly innervated muscle fibers. 131 Single-fiber electromyography can also provide an estimate of mean fiber density of motor units. The electrode is positioned so that a muscle fiber action potential is recorded at maximal amplitude. This action potential is then made to trigger the oscilloscope sweep, and the number of synchronous action potentials that are larger than 200 IlVand have a rise time of less than 300 usee is counted for this particular electrode position. To obtain a measure of fiber density in the motor units of the muscle under examination, the process is repeated for at least 20 different positions of

256

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

Normal

""

S

'&

""\t-.......

1 msec

- - ........

, .......

'>

2

Ii,

{ --.E- T' . ---~

-.~--

Reinnervation

•

-r----

E

,u

11-12 l1\li Single-fiber EMG recordings in normal and reinnervated muscle. The diagram illustrates the number of muscle fibers of one motor unit (blackened). The uptake area of the recording electrode is represented as the half-circle and E. In the normal (l and 2), only action potentials from one or two fibers are recorded. In reinnervation (3), many fiber action potentials are recorded because of increased fiber density in the motor unit (right part of figure). (From Hakelius L, Stalberg E: Electromyographical studies of free autogenous muscle transplants in man. Scand] Plast Reconstr Surg, 8:211,1974, with permisFIGURE

... ....

3 _J

.......

the recording electrode, and the mean number of responses is calculated. Seldom is it possible to record from three or more muscle fibers belonging to the same normal motor unit in anyone electrode position. After reinnervation, however, fiber density is increased because of collateral sprouting, and it may be possible to record potentials from up to about 10 fibers of the same motor unit with the electrode in one position (Fig. 11-12). Conventional EMG gives little information about the size of motor units (i.e., the total number and size of the component muscle fibers). Macro-EMG is a technique that may provide such information. A modified single-fiber electrode is used. Electrical activity recorded by the 15 mm of exposed electrode shaft during voluntary muscle activity is averaged after triggering from a single muscle fiber action potential, so that the contribution from one motor unit is extracted. Scanning EMG was developed to study motor units to obtain a better understanding of their topography. It indicates both the spatial and the temporal distribution of activity in individual units. Further information about both techniques is provided in Chapter 12.

REFERENCES I. Nardin RA, Rutkove SB, Raynor EM: Diagnostic accuracy of electrodiagnostic testing in the evaluation of weakness. Muscle Nerve, 26:201, 2002

sion.)

2. American Association of Electromyography and Electrodiagnosis: AAEE Glossary of Terms in Clinical Electromyography. Muscle Nerve, 10:Suppl, 1987 3. Walker WC, Keyser-Marcus LA,]ohns]S et al: Relation of electromyography-induced pain to type of recording electrode. Muscle Nerve, 24:417, 2001 4. Gitter A], Stolov WG: Instrumentation and measurement in e1ectrodiagnostic medicine. Muscle Nerve, 18:799 and 812, 1995 (two parts) 5. Farrell CM, Rubin DI, Haidukewych G]: Acute compartment syndrome of the leg following diagnostic electromyography. Muscle Nerve, 27:374, 2003 6. Al-Shekh1ee A, Shapiro BE, Preston DC: Iatrogenic complications and risks of nerve conduction studies and needle electromyography. Muscle Nerve, 27:517, 2003 7. Stalbert E, Andreassen S, Falck B et al: Quantitative analysis of individual motor unit potentials: a proposition for standardized terminology and criteria for measurement.] Clin Neurophysiol, 3:313,1986 8. Rutkove S: Effects of temperature on neuromuscular electrophysiology. Muscle Nerve, 24:867, 2001 9. Denys EH: The influence of temperature in clinical neurophysiology. Muscle Nerve, 14:795, 1991 10. Wilbourn A]: An unreported, distinctive type of increased insertional activity. Muscle Nerve, 5:S101, 1982 11. Purves D, Sakmann B: Membrane properties underlying spontaneous activity of denervated muscle fibres. J Physiol, 239:125, 1974 12. Thesleff S, Ward MR: Studies on the mechanism of fibrillation potentials in denervated muscle. ] Physiol, 244:313, 1975 13. SmithJW, ThesleffS: Spontaneous activity in denervated mouse diaphragm muscle.] Physiol, 257:171,1976

Clinical Electromyography

14. Kraft GH: Decay of fibrillation potential amplitude following nerve injury. Electroencephalogr Clin Neurophysiol, 60:105P, 1985 15. Partanen]V, Danner R: Fibrillation potentials after muscle injury in humans. Muscle Nerve, 5:S70, 1982 16. Desmedt]E: Muscular dystrophy contrasted with denervation: different mechanisms underlying spontaneous fibrillations. Electroencephalogr Clin Neurophysiol Suppl, 34:531, 1978 17. Nandedkar SD, Barkhaus PE, Sanders DB et al: Some observations on fibrillations and positive sharp waves. Muscle Nerve, 23:888, 2000 18. Dumitru D, King JC, Rogers WE et al: Positive sharp wave and fibrillation potential modeling. Muscle Nerve, 22:242, 1999 19. Kraft GH: Fibrillation potentials and positive sharp waves: are they the same? Electroencephalogr Clin Neurophysiol, 81:163,1991 20. Wiechers D, Johnson E: Syndrome of diffuse abnormal insertional activity. Arch Phys Med Rehabil, 63:538, 1982 21. Wettstein A: The origin of fasciculations in motoneuron disease. Ann Neurol, 5:295, 1979 22. Roth G: The origin of fasciculations. Ann Neurol, 12:542, 1982 23. Trojaborg W, Buchthal F: Malignant and benign fasciculations. Acta Physiol Scand, 41:SuppI13, 251, 1965 24. Janko M, Trontelj]v, Gersak K: Fasciculations in motor neuron disease: discharge rate reflects extent and recency of collateral sprouting. .I Neurol Neurosurg Psychiatry, 52: 1375, 1989 25. Conradi S, Grimby L, Lundemo G: Pathophysiology of fasciculations in ALS as studied by electromyography of single motor units. Muscle Nerve, 5:202, 1982 26. Trontelj]v, Stalberg E: Spontaneous activity within the motor unit: a single fibre EMG study. Electroencephalogr Clin Neurophysiol, 43:613,1977 27. Stalberg E, Trontelj]V: Abnormal discharges generated within the motor unit as observed with single-fiber electromyography. p. 443. In Culp V\J, Ochoa .I (eds): Abnormal Nerves and Muscles as Impulse Generators. Oxford University Press, New York, 1982 28. Ptacek L],]ohnson K], Griggs RC: Genetics and physiology of the myotonic muscle disorders. N Engl .I Med, 328:482, 1993 29. Fellows LK, Foster B], Chalk CH: Clinical significance of complex repetitive discharges: a case-control study. Muscle Nerve, 28:504, 2003 30. Daube JR: Electrodiagnosis of muscle disorders. p. 764. In Engel AG, Franzini-Armstrong C (eds): Myology. 2nd Ed. McGraw-Hill, New York, 1994 31. Albers .JW, Allen AA, Bastron jA et al: Limb myokymia. Muscle Nerve, 4:494, 1981 32. Nogues MA, Stalberg E: Electrodiagnostic findings in syringomyelia. Muscle Nerve, 22:1653,1999 33. Radu EW, Skorpil V, Kaeser HE: Facial myokymia. Eur Neurol, 13:499, 1975 34. Daube JR, Kelly Il. Martin RA: Facial myokymia with polyradiculoneuropathy. Neurology, 29:662, 1979 35. Baldissera F, Cavallari P, Dworzak F: Motor neuron "bistability": a pathogenetic mechanism for cramps and myokymia. Brain, 117:929, 1994

257

36. Buchthal F, Olsen PZ: Electromyography and muscle biopsy in infantile spinal muscular atrophy. Brain, 93:15, 1970 37. Vincent A: Understanding neuromyotonia. Muscle Nerve, 23:655, 2000 38. Meriggioli MN, Sanders DB: Conduction block and continuous motor unit activity in chronic acquired demyelinating polyneuropathy. Muscle Nerve, 22:532, 1999 39. Oleary CP, Mann AC, Lough] et al: Muscle hypertrophy in multifocal motor neuropathy is associated with continuous motor unit activity. Muscle Nerve, 20:479,1997 40. Newsorn-Davis ], Mills KR: Immunological associations of acquired neuromyotonia (Isaacs' syndrome). Report of five cases and literature review. Brain, 116:453, 1993 41. Daube jR: Electrophysiologic studies in the diagnosis and prognosis of motor neuron diseases. Neurol Clin, 3:473, 1985 42. Borenstein S, Desmedt.IE: Range of variations in motor unit potentials during reinnervation after traumatic nerve lesions in humans. Ann Neurol, 8:460, 1980 43. Desmedt .IE, Borenstein S: Regeneration in Duchenne muscular dystrophy: electromyographic evidence. Arch Neurol, 33:642,1976 44. Wilbourn A]: The electrodiagnostic examination with myopathies..I Clin Neurophysiol, 10:132, 1993 45. Buchthal F, Kamieniecka Z: The diagnostic yield of quantified electromyography and quantified muscle biopsy in neuromuscular disorders. Muscle Nerve, 5:265, 1982 46. Aminoff Mj: Electromyography in Clinical Practice: Electrodiagnostic Aspects of Neuromuscular Disease. 3rd Ed. Churchill Livingstone, New York, 1998 47. Kimura J: Electrodiagnosis in Diseases of Nerve and Muscle: Principles and Practice. 2nd Ed. FA Davis, Philadelphia, 1989 48. Saperstein DS, Amata AA, Barohn RJ: Clinical and genetic aspects of distal myopathies. Muscle Nerve, 24:1440,2001 49. Stalberg E, Trontelj]v: Single Fiber Electromyography. 2nd Ed. Raven Press, New York, 1994 50. Quijano-Roy S, Renault F, Romero N et al: EMG and nerve conduction studies in children with congenital muscular dystrophy. Muscle Nerve, 29:292, 2004 51. Blijham Pj, Hengstman GD], Ter Laak HJ et al: Musclefiber conduction velocity and electromyography as diagnostic tools in patients with suspected inflammatory myopathy: a prospective study. Muscle Nerve, 29:46, 2004 52. Uncini A, Lange DJ, Lovelace RE et al: Long-duration polyphasic motor unit potentials in myopathies: a quantitative study with pathological correlation. Muscle Nerve, 13:263, 1990 53. Gilchrist JM, Sachs GM: Electrodiagnostic studies in the management and prognosis of neuromuscular disorders. Muscle Nerve, 29:165, 2004 54. Dalakas MC, Pezeshkpour GH, Gravell M et al: Polymyositis associated with AIDS retrovirus. .lAMA, 256:2381, 1986 55. Simpson DM, Bender AN: Human immunodeficiency virus-associated myopathy: analysis of II patients. Ann Neurol, 24:79, 1988

258

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

56. Gonzales MF, Olney RK, So YT et al: Subacute structural myopathy associated with human immunodeficiency virus infection. Arch Neurol, 45:585, 1988 57. Bessen LJ, GreeneJB, Louie E et al: Severe polymyositislike syndrome associated with zidovudine therapy of AIDS and ARC. N EnglJ Med, 318:708,1988 58. Gertner E, Thurn JR, Williams DN et al: Zidovudineassociated myopathy. AmJ Med, 86:814, 1989 59. Gorard DA, Henry K, Guiloff RJ: Necrotising myopathy and zidovudine. Lancet, 1:1050, 1988 60. Griggs RC, Askanas V, Di Mauro S et al: Inclusion body myositis and myopathies. Ann Neurol, 38:705, 1995 61. Lotz BP, Engel AG, Nishino H et al: Inclusion body myositis: observations in 40 patients. Brain, 112:727, 1989 62. Phillips BA, Cala lA, Thickbroom GW et al: Patterns of muscle involvement in inclusion body myositis: clinical and magnetic resonance imaging study. Muscle Nerve, 24:1526,2001 63. Barkhaus PE, Periquet MI, Nandedkar S: Quantitative electrophysiological studies in sporadic inclusion body myositis. Muscle Nerve, 22:480, 1999 64. Rao SN, Katiyar BC, Nair KRP et al: Neuromuscular status in hypothyroidism. Acta Neurol Scand, 61:167, 1980 65. Engel AG: Acid maltase deficiency in adults: studies in four cases of a syndrome which may mimic muscular dystrophy or other myopathies. Brain, 93:599, 1970 66. DiMauro S, Hartwig GB, Hays A et al: Debrancher deficiency: a neuromuscular disorder in 5 adults. Ann Neurol, 5:422, 1979 67. Rubin DI, Hermann RC: Electrophysiologic findings in amyloid myopathy. Muscle Nerve, 22:355, 1999 68. HallaJT, Fallahi S, Koopman "'1: Penicillamine-induced myositis: observations and unique features in two patients and review of the literature. Am J Med, 77:719, 1984 69. Kuncl RW, Cornblath DR, Avila a et al: Electrodiagnosis of human colchicine myoneuropathy. Muscle Nerve, 12:360,1989 70. RiggsJE, Schochet SS, Gutmann L et al: Chronic human colchicine neuropathy and myopathy. Arch Neurol, 43:521, 1986 71. Younger DS, Mayer SA, Weimer LH et al: Colchicineinduced myopathy and neuropathy. Neurology, 41:943, 1991 72. Sghirlanzoni A, Mantegazza R, Mora M et al: Chloroquine myopathy and myasthenia-like syndrome. Muscle Nerve, 11:114, 1988 73. Pierce LR, Wysowski DK, Gross TP: Myopathy and rhabdomyolysis associated with lovastatin-gemfibrozil combination therapy. JAMA, 264:71, 1990 74. SomersJE, Winer N: Reversible myopathy and myotonia following administration of a hypocholesterolemic agent. Neurology, 16:761, 1966 75. Hirano M, Ott BR, Raps EC et al: Acute quadriplegic myopathy: a complication of treatment with steroids, nondepolarizing blocking agents, or both. Neurology, 42:2082, 1992 76. Zochodne DW, Ramsay DA, Saly V et al: Acute necrotizing myopathy of intensive care: electrophysiological studies. Muscle Nerve, 17:285, 1994

77. Lacomis D, Zochodne DW, Bird SJ: Critical illness myopathy. Muscle Nerve, 23:1785, 2000 78. Gutmann L, Blumenthal D, Gutmann Let al: Acute type II myofiber atrophy in critical illness. Neurology, 46:819, 1996 79. Rich MM, Teener lW, Raps EC et al: Muscle is electrically inexcitable in acute quadriplegic myopathy. Neurology, 46:731, 1996 80. Rich MM, Bird SJ, Raps EC et al: Direct muscle stimulation in acute quadriplegic myopathy. Muscle Nerve, 20:665, 1997 81. Lomen-Hoerth C, Simmons ML, DeArmond SJ et al: Adult-onset nemaline myopathy: another cause of dropped head. Muscle Nerve, 22:1146, 1999 82. Pascuzzi RM, Gratianne R, Azzarelli B et al: Schwartzjampel syndrome with dominant inheritance. Muscle Nerve, 13:1152, 1990 83. Deymeer F, Lehmann-Horn F, Serdaroglu P et al: Electrical myotonia in heterozygous carriers of recessive myotonia congenita. Muscle Nerve, 22:123,1999 84. Torbergsen T: A family with dominant hereditary myotonia, muscular hypertrophy, and increased muscular irritability, distinct from myotonia congenita Thomsen. Acta Neurol Scand, 51:225, 1975 85. Torbersen T: Rippling muscle disease: a review. Muscle Nerve Suppl11:S103, 2002 86. SoYT, Zu L, Barraza C et al: Rippling muscle disease: evidence for phenotypic and genetic heterogeneity. Muscle Nerve, 24:340, 2001 87. Koul RL, Chand RP, Chacko A et al: Severe autosomal recessive rippling muscle disease. Muscle Nerve, 24:1542, 2001 88. Vernino S, Auger RG, Emslie-Smith AM et al: Myasthenia, thymoma, presynaptic antibodies, and a continuum of neuromuscular hyperexcitability. Neurology, 53:1233, 1999 89. Kleopas KA, Zamba-Papanicolaou E, Kyriakides T: Compressive lumbar myelopathy presenting as segmental motor neuron disease. Muscle Nerve, 28:69, 2003 90. Leis AA, Stokic DS, Webb RM et al: Clinical spectrum of muscle weakness in human West Nile virus infection. Muscle Nerve, 28:302, 2003 91. Amer-Shekhlee A, Katirji B: Electrodiagnostic features of acute paralytic poliomyelitis associated with West Nile virus infection. Muscle Nerve, 29:376, 2004 92. World Federation of Neurology Research Group on Neuromuscular Diseases: El Escorial World Federation of Neurology criteria for the diagnosis of amyotrophic lateral sclerosis. J Neurol Sci, 124:Suppl, 96, 1994 93. Li J, Petajan J, Smith G et al: Electromyography of sternocleidomastoid muscle in ALS: a prospective study. Muscle Nerve, 25:725, 2002 94. Kuwabara S, Mizobuchi K, Ogawara K et al: Dissociated small hand muscle involvement in amyotrophic lateral sclerosis detected by motor unit number estimates. Muscle Nerve, 22:870, 1999 95. Daube JR: Electrodiagnostic studies in amyotrophic lateral sclerosis and other motor neuron disorders. Muscle Nerve, 23:1488, 2000 96. Younger DS: Motor neuron disease and malignancy. Muscle Nerve, 23:658, 2000

Clinical Electromyography 97. Sobue I, Saito N, lida M et al: Juvenile type of distal and segmental muscular atrophy of upper extremities. Ann Neurol, 3:429, 1978 911. Olney RK, Aminoff MJ, So IT: Clinical and electrodiagnostic features of X-linked recessive bulbospinal neuronopathy. Neurology, 41:823, 1991 99. Nogues MA, Stalberg E: Electrodiagnostic findings in syringomyelia. Muscle Nerve, 22:1653,1999 100. Jensen MC, Brant-Zawadski MN, Obuchowski N et al: Magnetic resonance imaging of the lumbar spine in people without back pain. N EnglJ Med, 14:331, 1994 101. Nardin RA, Patel MR, Gudas TF et al: Electromyography and magnetic resonance imaging in the evaluation of radiculopathy. Muscle Nerve, 22:151, 1999 102. Pezzin LE, Dillingham TR, Lauder TO et al: Cervical radiculopathies: relationship between symptom duration and spontaneous EMG activity. Muscle Nerve, 22:1412, 1999 I lB. Tsao BE, Levin KH, Bodner RA: Comparison of surgical and electrodiagnostic findings in single root lumbosacral radiculopathies. Muscle Nerve, 27:60, 2003 104. Levin KH, Maggiano HJ, Wilbourn A.J: Cervical radiculopathies: comparison of surgical and EMG localization of single-root lesions. Neurology, 46: I 022, 1996 10:>. Date ES, Mar EY, Bugola MR et al: The prevalence of lumbar paraspinal spontaneous activity in asymptomatic subjects. Muscle Nerve, 19:350, 1996 lOti. Sun SF, Streib EW: Diabetic thoracoabdominal neuropathy: clinical and electrodiagnostic features. Ann Neurol, 9:75, 1981 107. Kikta GO, Breuer AC, Wilbourn AJ: Thoracic root pain in diabetes: the spectrum of clinical and electromyographic findings. Ann Neurol, 11:80, 1982 10K. Koffman B,Junck L, Elias SB et al: Polyradiculopathy in sarcoidosis. Muscle Nerve, 22:608, 1999 1m!. Wilbourn AJ, AminoffM]: The electrophysiologic examination in patients with radiculopathies. Muscle Nerve, 21:1612,1998 110. AminotfMJ, Goodin OS, Parry GJ et al: Electrophysiologic evaluation of lumbosacral radiculopathies: electromyography, late potentials, and somatosensory evoked potentials. Neurology, 35:1514,1985 III. Aminoff MJ, Olney RK, Parry GJ et al: Relative utility of different electrophysiologic techniques in the evaluation of brachial plexopathies. Neurology, 38:546, 1988 112. Wulff CH, Gilliatt RW: F waves in patients with hand wasting caused by a cervical rib and band. Muscle Nerve, 2:452, 1979 II:t Cruz-Martinez A, Barrio M, Arpa.J: Neuralgic amyotrophy: variable expression in 40 patients. J Periph Nerv Syst, 7:198, 2002

259

114. Lahrmann H, Grisold W, Authier FJ et al: Neuralgic amyotrophy with phrenic nerve involvement. Muscle Nerve, 22:437, 1999 115. Van DijkJG, Pondaag W, Malessy MJA: Obstetric lesions of the brachial plexus. Muscle Nerve, 24:1451, 2001 116. Gnatz SM: The role of needle electromyography in the evaluation of patients with carpal tunnel syndrome: needle EMG is important. Muscle Nerve, 22:282,1999 117. Lahrmann H, Albrecht G, Drlicek M et al: Acquired neuromyotonia and peripheral neuropathy in a patient with Hodgkin's disease. Muscle Nerve, 24:834, 2001 118. Isaacs H: Continuous muscle fibre activity in an Indian male with additional evidence of terminal motor fibre abnormality.J Neurol Neurosurg Psychiatry, 30:126, 1967 119. Auger RG, Litchy \\J, Cascino TL et al: Hemimasticatory spasm: clinical and electrophysiologic observations. Neurology, 42:2263, 1992 120. Mitsumoto H, Levin KH, Wilbourn AJ et al: Hypertrophic mononeuritis clinically presenting with painful legs and moving toes. Muscle Nerve, 13:215, 1990 121. Schoenen J, Gonce M, Delwaide P.J: Painful legs and moving toes: a syndrome with different pathophysiologic mechanisms. Neurology, 34:1108, 1984 122. WulffCH: Painful legs and moving toes: a report of3 cases with neurophysiological studies. Acta Neurol Scand, 66:283, 1982 123. Meinck HM, Thompson PO: Stiff man syndrome and related conditions. Mov Disord, 17:853, 2002 124. Lohmann T, Londei M, Hawa M et al: Humoral and cellular autoimmune responses in stiff person syndrome. Ann NYAcad Sci, 998: 215, 2003 125. Saadeh PB, Cristafulli CF, Sosner J: Electrodiagnostic studies of the neuromuscular respiratory system. Phys Med Rehabil Clin North Am, 5:541, 1994 126. Saadeh PB, Cristafulli CF, Sosner J et al: Needle electromyography of the diaphragm: a new technique. Muscle Nerve, 16:15, 1993 127. Bolton CF, Grand'Maison F, Parkes A et al: Needle electromyography of the diaphragm. Muscle Nerve, 15:678, 1992 128. Lagueny A, Ellie E, Sain tarailles J et al: Unilateral diaphragmatic paralysis: an electrophysiological study. J Neurol Neurosurg Psychiatry, 55:316, 1992 129. Bolton CF: Clinical neurophysiology of the respiratory system. Muscle Nerve, 16:809, 1993 130. AAEM Laryngeal Task Force: Laryngeal electromyography: an evidence-based review. Muscle Nerve, 28:767, 2003 131. Lateva ZC, McGill KC, Johanson ME: Increased jitter and blocking in normal muscles due to doubly innervated muscle fibers. Muscle Nerve, 28:423, 2003

CHAPTER

16 H-Reflex and F-Response Studies MORRIS A. FISHER

H REFLEX Physiology Technique of Recording H Reflexes Uses of H-Reflex Studies Disorders of the Peripheral Nervous System Disorders of the Central Nervous System

H reflexes and F waves are commonly recorded electrophysiologic responses (Fig. 16-1, Fig. 16-2, and Fig. 16-3). H reflexes are reflexes that are produced by afferent conduction in large afferent fibers and by efferent conduction in alpha motor neurons. By contrast, F waves are produced by antidromic activation of the alpha motor neurons. Despite their differences (Table 16-1), they are commonly discussed together. H reflexes and F waves are commonly used for similar clinical problems, are found at comparable latencies, reflect conduction to and from the spinal cord, and involve motor neuron activation.

H REFLEX

Physiology In 1918, Hoffman described a reflex response in calf muscles that followed stimulation of the posterior tibial nerve and was comparable in latency to the Achilles' reflex.t-" In recognition of Hoffman's original contribution, the response was named the H reflex by Magladery and McDougal. 3 Because H reflexes involve conduction from the periphery to and from the spinal cord, they occur at latencies considerably longer than the latency of a direct motor response. A necessary condition for establishing an H reflex is that this "late" response must be larger than the preceding direct motor response. This condition can occur only with central amplification of the motor response caused by reflex activation of motor neurons. Careful observation of H reflexes as they increase in amplitude often reveals a decrease in latency.

F RESPONSE Physiology Technique of Recording F Waves Clinical Application of F-Wave Studies Disorders of the Peripheral Nervous System Disorders of the Central Nervous System

This decrease is consistent with the orderly activation of smaller and then larger motor neurons, with an associated increase in axonal conduction velocities, as would be expected in a reflex response. The arc of the H reflex includes conduction in large, fast-conducting Ia fibers. In that sense, the H reflex is similar to the phasic myotatic ("deep tendon") reflex produced by muscle stretch. Although experimental data support the possible monosynaptic nature of the H reflex, intraneural studies have questioned the exclusively monosynaptic nature of both H and phasic myotatic reflexes." Calf H and Achilles' reflexes are generally correlated.i-" and calf H-reflex amplitudes have been positively correlated with prominence of the Achilles' reflex." Unlike the phasic myotatic reflex, however, the H reflex does not involve muscle spindle activation. This difference may at times explain the presence of calf H reflexes in the absence of Achilles' reflexes, and vice versa. H reflexes are inhibited as the stimulus intensity is increased from submaximal to that required for eliciting a maximal direct (M) response. This relationship has been explained by "collision" of orthodromic impulses with impulses conducted antidromically in motor axons. In fact, this mechanism has little or no role in the inhibition of H reflexes that occurs with increasing stimulus intensity. To be complete, such collision must occur distal to the motor neurons. Allowing for rise-times for motor neuron excitatory postsynaptic potentials (EPSPs) of at least 3 msec, as well as for at least one synapse, the differences in afferent and efferent conduction in H reflexes necessary for collision inhibition have not been established. Large H reflexes are obtained from calf muscles even with supramaximal

357

358

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

stimulation if the stimuli are timed appropriately with phasic contractions of the muscles.f This indicates that H-reflex inhibition does not depend on peripheral collision of afferent and efferent impulses. Experimental studies in normal and spastic subjects are consistent with central inhibition of H reflexes occurring as stimulus intensities are increased," Inhibitory mechanisms present in the spinal cord can explain H-reflex inhibition. Inhibitory interneurons (Renshaw cells) are activated by antidromic stimulation.l? are distributed widely throughout the motor neuron pool, II and discharge more strongly and with shorter latency as stimulus intensity increases.V H reflexes may be monosynaptic, whereas H-reflex inhibition by Renshaw cells would involve two synapses. Any resulting difference in the onset of inhibitory and excitatory effects may be as brief as 0.3 msec!' and therefore well within a reasonable physiologic range given motor neuron EPSP rise times. Direct connections between motor neurons are also present'V" and could contribute to H-reflex inhibition. Single-fiber dectromyographic (EMG) studies of H reflexes support a process of active inhibition involving inhibitory synapses with stimuli of increasing intensity" H-reflex studies in turn have been used to demonstrate prominent central inhibition, probably from Renshaw cells, following supramaximal nerve stimulation. 16

FIGURE 16-1 • Drawings of H reflexes (upper) and F responses (lower) recorded (R) from calf muscles. Stimulation (S) of the tibial nerve in the popliteal fossa activates Ia fibers with a resultant reflex discharge of motor (m) axons. H reflexes are present at low levels of stimulation, when the M wave may be absent or lower in amplitude than the H reflex, and are inhibited with increasing stimulus intensity. With supramaximal stimulation, low-amplitude, variable F responses are produced by antidromic activation of motor neurons. G, ground. (From Fisher MA: H reflexes and F waves: physiology and clinical indications. Muscle Nerve, 15:1223, 1992,with permission.)

2rnV

10 rnsec

35.3 rnA rnV M-Arnp 40.8 rnA

44.7 rnA

8 6

4 2

rnV H-Arnp

x

x

4

x x

x

65.1 rnA

10 rnA

0

0

0 0 0

2 rnA

0405060 70 54.1 rnA

6

0

rnA 40 50 60 70 10 rnsec

FIGURE 16-2 • H reflexes with increasing stimulus intensity shown in a raster mode (left) and superimposed (right). The inserts show in graphic form that as the M-wave amplitude increases, the H-reflex amplitude initially increases and then decreases. At the highest level of stimulation (last response in the raster mode), the H reflex has disappeared and is replaced with an F-wave (calibration per division, 2 mVand 10 msec).

H-Reflex and F-Response Studies

5 msec

500 fLY

A

5 msec

5mV

B III A, F waves (right) with associated M waves (left) recorded from the abductor pollicis brevis muscle following supramaximal stimulation at the wrist. The inherent variability of F waves is emphasized in the superimposed recordings (B). The chronodispersion is the difference between the shortest and longest F-wave latencies. The persistence in this series of 10 recordings is 90 percent, because in one an F wave is absent. The two largest responses are repeater waves (calibration per division, 5 mV for M waves, 500 IlV; 5 msec). (From Fisher MA: Normative F-wave values and the number of recorded F waves. Muscle Nerve, 17:1185, 1994.with perrnission.)

FIGURE 16-3

TABLE 1&-1 • H Reflexes Ind FWIVes Feature

H Reflexes

F Waves

Response

Reflex

Afferent fibers Efferent fibers Distribution Stimulus intensity Increasing stimulus intensity Amplitude Morphology Motor units

la afferents Alpha motor fibers Restricted low Inhibits response

Antidromic firing of motor neurons Alpha motor fibers Alpha motor fibers Ubiquitous High Facilitates response

large Stable Different from M wave

Small Variable Same as in M wave

359

In infants younger than 2 years, H reflexes are widely distributed.!? Beyond infancy, H reflexes are regularly found in calf muscles (primarily the soleus) and homologous forearm flexors. They are also commonly present in the quadriceps and occasionally in plantar foot muscles. This restricted distribution of H reflexes with age is caused by changes associated with physiologic maturation of the central nervous system (CNS). The fraction of the soleus motor neuron pool activated in an H reflex is usually about 50 percent but can be as high as 100 percent. The ratio of the peak-to-peak maximum H-reflex to maximum M-wave amplitude (HIM ratio) provides a measure of motor neuron pool activation and therefore excitability. The HIM ratio for calfH reflexes is normally less than 0.7. 18 H reflexes involve the activation of a portion of the segmental motor neuron pool and are therefore enhanced by maneuvers that increase excitability of the motor neuron pool. H reflexes can be produced by facilitation maneuvers (e.g., contraction or post-tetanic potentiation) in muscles where they are not otherwise present, such as the small hand muscles. H reflexes have been used to explore the physiology of the CNS. A recent review has emphasized the problems with equating H-reflex responsiveness with motor neuron excitabiliry.l? Such equating of these two phenomena is limited, for example, by the effects of presynaptic inhibition and postactivation depression on H reflexes. With these limitations, H reflexes remain an important tool for investigating the central control of movement. H reflexes have been used to study patterns of central reflex organization. Studies have shown changes in H reflexes consistent with an influence on the reflex of both group Ia and group II afferents. 20 •2 1 H reflexes have been used to analyze patterns of reciprocal inhibition. The methodology for this in the forearm has been well described. 22 •23 Myotatic arcs between proximal and distal muscles also affect H reflexes. Activation of the quadriceps, for example, inhibits the soleus H reflex.v' and this reflex interaction has been employed in techniques for evaluating presynaptic inhibition." Presynaptic inhibition of Ia afferent fibers is decreased on fibers projecting on motor neurons of contracting muscles, but it is increased on afferents to motor neurons supplying muscles that are not involved in the contraction." Passive cyclic movements of the legs produce inhibition ofH reflexes; this is probably because of presynaptic inhibition that can be integrated at the spinal cord leve1. 27 •28 Characteristic changes in H-reflex amplitude can be defined if test stimuli are given at varying intervals after a conditioning stimulus. These excitability or recovery curves vary with the level of stimulation'? (Fig. 16-4) and are not clearly established until 1 year of age.

360

ElECTRODIAGNOSIS IN ClINICAL NEUROLOGY

50 49 48 47 46 45 44 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 28 27 26 25

40.14

80

39

75

38 37 36 35 34 33

70 65 60 55

32 31 30 29

50 45 40

28 27 26 25 24 22.64

35 30 25 20

Leg length Latency Age (ern) (msec) (yrs) FIGURE 16·4 • Nomogram of the regression of H-reflex

latency on leg length and age. (From Braddom RI,Johnson EW: Standardization of the H reflex and diagnostic use in 81 radiculopathy. Arch Phys Med Rehabil, 55:161,1974, with permission.)

H reflexes are inhibited by vibration as a result of vibration-induced activation of large afferent fibers with consequent peripheral "busy line" interference, presynaptic inhibition of afferent input.t" and activation of spindles in antagonist muscles.V

Technique of Recording H Reflexes H reflexes are readily obtained with percutaneous stimulation and surface recording techniques. Responses are stable, but only at similar conditions of stimulation and recording. The stimulating cathode should be proximal to avoid anodal block. Stimulus pulses of long duration (1 msec) are used to preferentially activate the large sensory fibers." Stimulus frequency should be 0.2 Hz or less to allow full recovery of the H reflex from a prior stimulus. A series of responses should be obtained. Starting with submaximal stimuli and increasing to supramaximal stimulation, one should verify whether the "late"

response can be larger than the preceding direct motor response, determine which H reflex has the largest amplitude, and demonstrate that inhibition of the H reflex occurs with increasing stimulus intensity. Latencies should be measured to the onset of the responses (either negative or positive), and amplitudes should be measured in a peak-to-peak fashion. For calf H reflexes, the posterior tibial nerve is stimulated in the popliteal fossa. Bipolar stimulation is usually adequate. Use of an anode with a large surface area at the patella can decrease stimulus artifact and may provide more discrete cathodal excitation of the nerve in the popliteal fossa. Recordings are made from the soleus muscle. Although techniques vary, a standard and convenient location for the active electrode is medial to the tibia at a point that is half the distance between the stimulation site and the medial malleolus, with the indifferent electrode placed on the Achilles' tendon. H reflexes in the forearm are recorded from the flexor carpi radialis muscle.F The recording electrode is placed at the junction of the upper one-third and lower two-thirds of the distance between the medial epicondyle and the radial styloid. The median nerve is stimulated percutaneously in the cubital fossa. H reflexes are routinely recorded with the muscle at rest. Contraction of the recording muscle will enhance H reflexes by facilitating the motor neuron pool. Such contraction can help to identify H reflexes in muscles in which H reflexes are normally present, as well as elicit H reflexes in muscles where they are not normally found." This facilitation of H reflexes by contraction is sometimes clinically useful and demonstrates that the normal distribution of H reflexes is based on physiologic, not structural, factors.

Uses of H-Reflex Studies The upper limit of normal latency for the H reflex of the soleus is 35 msec, and that of the flexor carpi radialis is 21 msec. H-reflex latencies are directly related to leg or arm length, height, and, to a lesser degree, age 34-36 (see Fig. 16-4). Normal values for infants and children are available. 37.3BWith careful technique, the upper limits of normal for side-to-side latency differences have been reported to be as low as 1.5 msec for calf muscles'" (mean, 0.09 ± 0.70 standard deviation [SD]36) and 1.0 msec for forearm flexor muscles 22 ,g9 (mean, 0.002 ± 0.42 SD). For routine clinical work and consistent with a criterion of 3 SD from the mean, 2 msec should be allowed for side-to-side differences when one records from the calf and 1.5 msec allowed when one records from the forearm. Side-to-side latency differences may be larger in the elderly" The upper limit of normal for the interside ratio of H-reflex

H-Reflex and F-Response Studies

amplitudes for calf muscles has been reported as 2,41 but a preferable figure for general clinical work is probably 3. This is similar to the interside ratio for H-reflex amplitudes in the forearm." Calf and forearm H reflexes are usually present in normal subjects. As with phasic myotatic reflexes, however, this statement is not always true, and symmetrically absent H reflexes are not necessarily abnormal. The percentage of absent responses increases in the elderly.f DISORDERS OF THE PERIPHERAL NERVOUS SYSTEM

H reflexes are a sensitive test for polyneuropathies and may be abnormal even in mild neuropathies. H reflexes involve conduction in proximal fibers. These studies can therefore define proximal nerve injury and may be abnormal even when distal studies are unremarkable. Absent H reflexes are characteristic of acute idiopathic polyneuropathy (Guillain-Barre syndrome). This loss of H reflexes occurs early after onset of the disorder, and absent H reflexes often may be an isolated finding in these patients if studied within several days after the onset of their illness. H reflexes can also be abnormal in plexopathies'" and in radiculopathies. H reflexes in the forearm flexor muscles may be abnormal with C6 or C7 root injury,39 and calf H reflexes may be abnormal with S1 radiculopathies.r'V H reflexes are affected by injury to both the dorsal and the ventral roots. H reflexes can therefore be important in the electrodiagnostic evaluation of radiculopathies by documenting nerve injury even when needle EMG is unrevealing owing to sparing of the ventral roots.

361

weeks to months following a cerebrovascular lesion associated with the appearance of features of the upper motor neuron syndrome (e.g., increased tone, brisk reflexes, and extensor plantar responses). In patients with chronic upper motor neuron lesions, vibratory inhibition of H reflexes is less than expected, possibly because of decreased presynaptic inhibition.P In contrast, vibratory inhibition of H reflexes may be enhanced in patients with acute cerebral lesions. H reflexes are relatively well preserved acutely after spinal shock at a time when both Achilles' reflexes and H-reflex recovery curves are depressed.'? Alterations in soleus H reflexes have been related to specific features of the upper motor neuron syndrome, in particular, decreased vibratory inhibition of H reflexes with hypertonia, increased HIM ratios with increased reflexes, and late facilitation of the recruitment curve with clonus." Within several months after a central injury, H-reflex excitability curves can show an abnormally rapid pattern of recovery (Fig. 16-5) associated with increased HIM ratios. 5,28 These patterns differ from those in patients with parkinsonian rigidity or cerebellar hypotonia."

-

Controls ... -. Patients with spasticity

100 Q)

~ 60 0.. E

<{

20

A 100

DISORDERS OF THE CENTRAL NERVOUS SYSTEM

H reflexes may be abnormally widespread in patients with CNS lesions and upper motor neuron signs.:" The presence in adults of H reflexes in muscles where they are not normally present (e.g., in the tibialis anterior or small hand muscles) may be clinically useful for documenting dysfunction of the central motor system. Patients with even mild hemiparesis have decreased potentiation of H reflexes with muscle contraction, and this finding is consistent with decreased background facilitation of motor neurons. At the same time, HIM ratios tend to be increased in patients with CNS lesions and upper motor neuron signs, and recruitment curves are altered" in a manner consistent with increased excitability of the central motor neuron pool. Conversely, H reflexes during cataplexy are depressed." Changes in H reflexes after CNS lesions are time dependent. H reflexes are depressed acutely after spinal cord injury. Increased HIM ratios develop during the

60

20

10 B

100

1000

5000

Intervals (msec)

FIGURE 16-5 .. H-reflex recovery curves recorded from calf muscles. Note the less pronounced inhibition in patients with spasticity in comparison to control subjects for stimuli near threshold (A), as well as at levels producing maximal H reflexes (B). Abscissa, time between conditioning and test stimuli (logarithmic scale); ordinate, test reflex amplitudes as percentages of conditioning reflex amplitudes. The vertical bars denote mean errors. (Modified from Olsen PZ, Diamantopoulos E: Excitability of spinal motor neurones in normal subjects and patients with spasticity, parkinsonian rigidity, and cerebellar hypotonia. J Neurol Neurosurg Psychiatry, 30:325, 1967, with permission.)

362

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Studies of recovery curves and patterns of reciprocal inhibition of forearm flexor H reflexes have revealed abnormalities in patients with various types of dystonia, even in clinically normal parts of the body.49.5o Altered patterns of reciprocal reflex activity have been found in hereditary hyperekplexia ("startle disease"). 51 In patients with chronic long-tract motor dysfunction, H reflexes have been used to analyze altered patterns of reflex activity between flexor and extensor muscles. 52.53 These findings have included increased reciprocal inhibition of flexor muscles by extensors. H-reflex studies have been helpful for understanding the specific pathophysiology of CNS dysfunction. H-reflex analysis of recurrent inhibition in patients with upper motor neuron lesions, for example, has indicated altered patterns of Renshaw cell activation but only with postural or voluntary contractions.54 These findings are consistent with disturbed Renshaw cell activity caused by supraspinal disruption of inhibitory control.

F RESPONSE

Physiology F waves are low-amplitude, ubiquitous responses inherently variable in amplitude, latency, and configuration. They are produced by antidromic activation ("backfiring") of motor neurons. F waves were originally recorded in small foot muscles,05 hence their name. The initial concept that these responses were reflexes was challenged particularly by the observation that motor units were present only in F waves if they were also present in the direct motor response.t" The antidromic origin of F waves has been confirmed by the presence of F waves in deafferented animals and humans'" and by single-fiber EMG analysis, which has indicated that an F-wave requires activation of the motor axon producing that particular response.t" For an individual motor unit, the size and shape of the motor unit in the direct motor (M) and F waves should be the same. Motor neurons are activated by depolarization at the low-threshold initial segment and subsequent invasion of the soma. 59 The available evidence would indicate that this is true regardless of whether the stimulus is orthodromic or antidromic. The shortest F-wave latencies are comparable to H-reflex latencies, but they may be 1 to 2 msec shorter. In contrast to H reflexes, F waves are most prominent when elicited by stimuli of high intensity.''? In addition, impulses must pass orthodromically through an axonal initial segment that has been discharged by a preceding antidromic impulse. Therefore, the effect of altered motor neuron excitability on F waves is variable. If a motor neuron is at a high level of excitability (i.e., already relatively depolarized),

neuronal activation will occur rapidly and the resultant orthodromic axonal discharge may present at the initial segment at a time when it is still refractory. Increased excitability of the motor neuron pool could then result in decreased prominence of F waves." As a result, the effects of agonist and antagonist contraction are less predictable with F waves than with H reflexes and are dependent on the physiologic organization of a particular muscle.P Individual motor neurons are activated infrequently with antidromic stimuladon.?' and F waves usually have no more than several motor unit action potentials.59.6o.63 As a result, F waves may not appear after each stimulus, are variable in configuration, and are low in amplitude (see Fig. 16-3). An important area of controversy is whether the range of F-wave latencies reflects the full range of conduction in motor axons.P' Several reports have suggested an absence of bias of motor unit discharge in F waves.65-67 Circumstantial and direct evidence, however, is consistent with a selective discharge of larger motor neurons in the generation of F waves.68-72 The indirect evidence includes the observation that changes in minimal F-wave latencies, even with relatively few F waves, are comparable over distance with changes in maximal evoked motor response latencies70.7&-75 despite the relatively large range of conductions that might be expected in F waves." All studies supporting an absence of selection in motor unit discharge in F waves have been performed using submaximal stimulation, whereas F waves clinically are recorded using supramaximal stimuli. Renshaw cells inhibit large motor neurons less than small ones. 77.7il Activation of these inhibitory interneurons by antidromic nerve discharges therefore provides a physiologic model for the selective firing of larger motor neurons in the generation of F waves. Other possible physiologic effects of an antidromic volley include field potential induction within the ventral horn sufficient to alter motor neuron excitability, as well as inhibition by direct recurrent collaterals from motor neuron to motor neuron. 79.80 Whatever the uncertainties, the variation in F-wave latencies is neither so random nor large as to preclude their clinical use. F waves often arise from an unstable baseline, and axon reflexes may be superimposed. In addition, the time between the start of antidromic activation and subsequent orthodromic discharge (i.e., the central "turnaround" time) is uncertain. Based on a statement by Eccles/" the turnaround time is commonly estimated to be 1 msec, but this figure has never been demonstrated directly. In invertebrates, these central delays can be orders of magnitude greater" In humans, identical F waves have been recorded from foot muscles with latency variations of 3 msec. 82 For these reasons and because of the inherent variability ofF waves, the number ofF waves used for analyses

H-Reflex and F-Response Studies

in clinical studies has varied. At times this number has been 50 or greaten?" whereas others have argued that examination of only 3 to 5 F waves is needed for an accurate determination ofF-wave latencies.P A consensus has developed in the literature that analysis of 10 to 20 F waves is a reasonable balance between feasibility (including patient tolerance) and adequate data. Studies of the abductor pollicis brevis muscle, however, have indicated that evaluation of 16 to 20 F waves (20 stimuli) is needed for accurate measurement of "true" latencies based on an analysis of responses following up to 100 stimuli. 7 ) ,84-8 6 Whatever the uncertainties, the normal variability of latencies in a series of F waves is about 10 percent, which is comparable to the range of error for measurements of other commonly used electrophysiologic responses. The clinical importance of defining "true" latency values varies. For example, it is important when clinical evaluation depends on small latency differences between sides (e.g., for radiculopathies) but not when latency prolongation may be profound (e.g., in demyelinating neuropathies) . Latencies are the most commonly reported parameters of F waves, F-wave latencies are directly related to height; limb length; and, to a lesser degree, age. 87-89 Considering these variables when defining normal F-wave latency values increases the clinical sensitivity of individual F-wave latency measurements. F-wave latencies are reported most often as minimal latencies, but recording F-wave latencies as mean values minimizes the errors inherent in a single latency measurement and provides results that are more reproducible. 86 ,87,9o Median latency values have been suggested as the most statistically correct way of defining F-wave Iatencies.F This approach, however, introduces an additional complexity that does not seem warranted because F wave latencies are normally distributed." Proximal conduction can be compared with distal conduction by contrasting the latencies of F waves and M waves. F-wave latencies have been used to estimate conduction in limited portions of proximal nerves.f as have F-wave conduction velocities. It has been argued that the latter is a sensitive method for detecting mild nerve dysfunction and has the further advantage of allowing comparison of conduction velocity in subjects with differing limb lengths.f" However, F-wave conduction velocities are less accurate than latency values alone because additional errors of measurement may be introduced, and F-wave latencies can readily be normalized to a particular arm or leg length. Analysis of F-wave parameters other than latency has clinical utility and may at times be more important than latency measurements. The difference between minimal and maximal latencies in a series of F waves (i.e., F chronodispersion) provides a measure of the range of conduction velocities in the axons contributing to the recorded F waves. F-wave duration and amplitude are

363

related to both the size and the number of motor units in a particular F-wave.63 F-wave persistence refers to the percentage of measurable F responses that follow a series of stimuli and is related to the antidromic excitability of a particular motor neuron pool. The recurrence of individual motor units in a series of F waves measures the selectivity ofF-wave discharge. The ratio of F-wave amplitudes to that of the associated M waves (i.e., F1M ratios) is a measure of the proportion of a motor neuron pool activated by antidromic stimulation. Given the F-wave variability, mean F-wave rather than maximum F-wave amplitudes are preferable for calculating F1M ratios, and these are used routinely.f F-wave recovery curves have been defined by measuring F-wave amplitude or persistence following conditioning stimuli that precede the test stimulus at varying intervals. As with H reflexes, there is an early depression of the response, followed by a later facilitation from about 80 to 300 msec.P" F waves have helped to define normal CNS physiology. F-wave studies, for example, have shown changes consistent with activation of group 11 afferents in leg muscles following stimulation of the sural nerve." In resting individuals, F-wave studies have shown a relatively increased central excitability of the antigravity calf muscle in comparison to the antagonist tibialis anterior muscles. With isometric contraction, the central excitability of the tibialis anterior muscle increases relative to that of calf muscles so that statistically significant differences in F-wave persistence are no longer present. 62 In contrast to the situation at rest, these results are consistent with a more balanced central excitability state between flexor and extensor muscles with activity. In general, F waves can be used as a "probe" for changes in spinal cord excitability"

Technique of Recording F Waves F waves are recorded in a manner similar to that used for direct motor responses except that the stimulating cathode should be proximal to the anode to avoid anodal block. Unlike H reflexes, a long stimulus duration is not required because there is no reason to activate preferentially the large afferent fibers. The author stimulates at rates less than 0.5 Hz to avoid the effects of an earlier stimulus on a subsequent response. In contrast to H reflexes, F waves are enhanced by stimulation at high intensity because of increased amplitude and persistence. High stimulus intensities also block axon reflexes that may be confused with F waves, as well as limiting contamination by H reflexes. The conventional stimulus intensity is 25 percent above maximal for eliciting a direct response. This provides a consistent physiologic environment for eliciting F waves.F waves are present at submaximal stimulation. This has the advantage of

364

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

increased patient comfort and may provide adequate information in situations in which the only question is whether F-wave latencies are markedly prolonged. A recent article has provided support for eliciting F waves by low-intensity stimulation." To date, however, all normative F-wave data are based on supramaximal stimulation. This should therefore be used until it is clearly established that submaximal stimulation can produce meaningful, reproducible information for all F-wave parameters. Adequate display of F waves usually requires an amplifier gain of 200 or 500 IlV per division and a sweep of 5 or 10 msec per division. As such, different recording parameters are usually required to evaluate fully the associated larger M wave. To be clearly identifiable, F waves should be at least 20 IlV in peak-to-peak amplitude. As discussed earlier, the number of F waves required to permit measurement of different F-wave parameters has been examined.71.84.86,97 Data from 8 to 10 identifiable, sequential F waves (10 stimuli) can provide a reasonable estimate of persistence. Accurate F-wave latency measurement requires 16 to 20 F waves (20 stimuli) and is best expressed as a mean, rather than a minimal value. The same number of F waves is adequate for measuring F1M ratios and the percentage of responses in a series of F waves that are the same (i.e., repeater waves). Even two F waves may define an abnormal chronodispersion if the two latencies are sufficiently dispersed, but determination of a truly representative value requires 45 to 55 F waves (50 to 60 stimuli). Accurate determination of the number of individual waves that repeat may require more than 90 F waves. When there is low persistence, more stimuli than usual may be needed to obtain adequate data. Muscle belly-tendon recordings are standard for hand and foot muscles, with the recording cathode placed over the motor point. For calf F responses, muscle belly recordings are often preferred because the effects of extraneous muscle activity are decreased when recording the low-amplitude F waves. The active electrode is placed as for H reflexes. F waves are routinely recorded with the muscle relaxed. Slight voluntary contraction will enhance F waves. This is sometimes helpful clinically, but the contraction will alter F-wave parameters such as amplitude and will increase the possibility of contamination by H reflexes. F waves are ubiquitous in distribution. Recording from proximal muscles, however, is difficult because the low-amplitude F waves are superimposed on the associated M wave. F waves are therefore routinely recorded only from muscles of the hand, foot, and leg with standard stimulation sites at the wrist, ankle, and knee, respectively. F waves can be recorded in hand muscles with stimulation in the axilla if collision tech-

niques are used so that the orthodromic M wave from the axilla is blocked by collision with antidromic impulses from the wrist.73 F waves are usually recorded in a raster fashion so that individual responses and the associated maximum M waves are available for analysis. Standard protocols include measurement of minimal and mean F-wave latency, F-wave chronodispersion, F-wave persistence, and the F1M amplitude ratio. The latter is calculated as the mean of the amplitudes of the F waves divided by the M-wave amplitude, with both measured from peak to peak.

Clinical Application of F-Wave Studies The upper limits of normal for minimal F-wave latencies in the author's laboratory are 33, 36, and 64 msec when recording from hand, calf, and foot muscles, respectively. Mean latencies are about 2 to 3 msec longer. Side-to-side differences exceeding 2 msec are regarded as abnormal for both minimal and mean values in recordings from the hand; 3 msec in recordings from the calf; and 4 msec in recordings from the foot. Detailed tables have been published of normal minimal and mean F-wave latencies related to height for hand and foot muscles. 9B--101 Regression equations relating minimal or mean latencies to age and height (or limb length) for the abductor pollicis brevis, abductor digiti minimi, calf, and abductor hallucis muscles are also available. In recordings from hand muscles, values that exceed by 3 msec or more the predicted value based on regression equations are abnormal, and those between 2 and 3 msec are borderline; comparable values for the soleus are 4 and 3 msec. The highest reported normal values for F-wave chronodispersion (mean ± SD) for the abductor pollicis brevis are 3.6 ± 1.2 102; for the abductor digiti minimi, 3.3 ± 1.189; for the soleus, 2.8 ± 1.1 103; for the extensor digitorum brevis, 4.9 ± 2.3. 101, and for the abductor hallucis, 4.5 ± 2.4. 100 Normal F1M ratios in the author's laboratory, based on mean F-wave amplitudes, are 2.2 ± 1.0 percent for the abductor pollicis brevis and 2.5 ± 1.2 percent for the soleus, equivalent to an upper limit of normal of about 5 percent. Mean values for F-wave persistence recording from the abductor pollicis brevis, abductor digiti minimi, soleus, and abductor hallucis are about 0.8 to 0.9, whereas in the antigravity antagonist tibialis anterior, extensor digitorum brevis, and extensor digitorum communis muscles these values are about 0.3 to 0.4, but the range of normal is high for individual measurements. When recording from the extensor digitorum brevis, for example, only 55 percent of normal subjects would be expected to have a persistence of 0.5 or greater;'?' whereas this might be

H-Reflex and F-Response Studies

expected in 96 percent of subjects when recording from the abductor hallucis.l'" The low persistence in antigravity antagonist muscles (e.g., the extensor digito rum brevis) argues against the routine recording of F waves from these muscles. The normal maximum frequency of an individual response in a series of F waves from hand muscles is about 10 percent,60,97but the percen tage of responses in which repeater waves are seen is higher because there may be more than one repeater wave and each individual repeater wave may be present more than twice. A frequency of an individual repeater wave as high as 58 percent has been noted in recordings from the extensor digitorum brevis of normal subjects (mean, 21.5 percent) .82 DISORDERS OF THE PERIPHERAL NERVOUS SYSTEM

Prolonged F-wave latencies are a sensitive indicator in polyneuropathies and may be found even when other measures of distal motor nerve conduction are unremarkable. 74.103--106 They may be more sensitive than standard motor conduction studies in axonal injury, 104 F-wave recordings are the most stable and reliable conduction study for monitoring patients with neuropathies.l'" and F-wave latencies are the most sensitive measure for defining nerve injury in patients with diabetes mellitus.l'" F waves are slowed in amyotrophic lateral sclerosis to a degree comparable to that of motor conduction velocity.log F-wave conduction studies have been normal in limb-girdle muscular dystrophy'!" but abnormal in myotonic dystrophy, III thus providing evidence of peripheral nerve dysfunction in the latter disorder. Prominent slowing of proximal F-wave conduction in comparison to distal motor nerve conduction studies has been found in patients with Guillain-Barre syndrome, thus confirming the importance of proximal nerve lesions in these patients. F-wave studies have been abnormal in 92 percent of nerves in patients with Guillain-Barre syndrome and in 95 percent of nerves in those patients with chronic inflammatory demyelinating polyneuropathy/" In about 20 percent of these nerves, motor conduction studies were unremarkable. F-wave abnormalities consisted mainly of loss of responses or prolonged latencies. Proximally predominant abnormalities have not been found in patients with uremia,IIO,112 diabetes mellitus.U'' or CharcotMarie-Tooth disease":!'! The value of F-wave latency studies in detecting injury to proximal nerve segments (e.g., as occurs with radiculopathies) is controversial.U" Concerns have included overlapping root innervation in muscles as well as the potentially limited information available from F-wave studies in a situation in which needle electromyography also evaluates injury to motor fibers. There has also been concern that latency abnormalities

365

may be masked because the effect of the lesion is "diluted" by the long course of the pathway being tested. Nevertheless, F-wave latencies may be prominently slowed in conditions such as Guillain-Barre syndrome, in which the nerve injury is proximal and may be limited to the roots. F-wave latencies have been prolonged in plexus'P as well as in root injury. I16 F-wave latencies from hand muscles may be slowed in syringomyelia, possibly owing to disturbed function of roots as well anterior horn cells.'!" Soleus F-wave latencies have been reported abnormal in a high percentage of patients with Sl radiculopathies. Studies of patients with focal, proximal nerve injury have usually been limited by a failure to examine a sufficient number (16 to 20) of F waves or to consider mean, rather than minimal, latency values. These studies have also generally failed to evaluate F-wave parameters other than latency, even though such analyses may add to the sensitivity of F-wave studies. Certain studies have indicated a sensitivity of F-wave studies in L5/S1 radiculopathies comparable to that of needle electromyography, with the F-wave abnormalities often providing unique rather than overlapping information. I Ii\--I2I All of these studies, however, have considered F-wave parameters other than latency (e.g., persistence and chronodispersion). F-wave chronodispersion in calf muscles may be increased in studies performed after standing, compared with prior studies performed at rest in patients with lumbosacral radiculopathies, consistent with dynamic changes in root function.P'' There is no evidence that F waves are clinically useful in the evaluation of cervical radiculopathies. The number of identical responses in a series of F waves may be increased in patients with neurogenic atrophy, which is consistent with a decreased number of motor neurons capable of responding to antidromic stimulation.t" Increased repeater waves have also been found in patients with amyotrophic lateral sclerosis and cervical myeloradiculopathies. 123 This increase in repeater waves may be related to increased discharge of responding motor neurons in patients with upper motor neuron syndromes. An increase in the percentage of repeater waves following 100 supramaximal stimuli has been noted in the carpal tunnel syndrome, I 13 but this observation is not clinically useful given the large number of stimuli required. F-wave chronodispersion may be prolonged in polyneuropathies. 89,102,104,124 Chronodispersion tends to be larger in nerves with demyelinating, rather than axonal, injury,104 and tends to be relatively decreased with conduction block. 124 DISORDERS OF THE CENTRAL NERVOUS SYSTEM

F-wave analyses to define increased central excitability states are more complicated than analyses of H reflexes.

366

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

Single-fiber studies in patients with upper motor neuron syndromes indicate that antidromically activated motor neurons fire more frequently than do those in normal individuals.P' At the same time, a frequently backfired motor neuron will discharge less frequently with activation by muscle contraction, whereas motor neurons that discharge infrequently will increase their firing rate. These observations are consistent with F-wave studies in deafferented animals. 57 At times, therefore, increased central excitability results in decreased discharge of larger motor neurons in an F-wave owing to blockage at the still refractory initial segment of the motor neuron. Despite this complexity, analyses of F waves are a valuable technique for monitoring central motor neuron exci tabili ty. In patients with CNS lesions, the normal, relatively increased prominence of F waves in resting extensor muscles compared with flexor muscles may be disrupted. F-wave amplitudes and persistence are decreased in clinically involved limbs, and this finding is compatible with decreased central excitability in patients studied early after unilateral cerebrovascular lesions, a period when decreased tone and reflexes are common findings.125.126 Similarly, F waves are absent in patients with spinal shock caused by injury and are decreased in prominence in similar patients without spinal shock.t? F-wave amplitudes and persistence can be decreased by cerebellar stimulation, which is consistent with increased cerebellar inhibitory ourflow.P? By contrast, F-wave persistence and average F-wave amplitudes, as well as F1M ratios, are increased in patients with "spasticity." Huge F waves (i.e., as large as 75 percent of Mwave amplitudes) have been found in chronic tetanus.P" These waves were associated with a shortened or absent silent period, which is compatible with the failure of inhibition of Renshaw cells and thereby indirectly supports a role for Renshaw cell activity in F-wave discharge. In patients with upper motor neuron syndromes, F-wave latencies may be prolonged, and durations and amplitudes are increased.P? These data are consistent with the discharge of a greater number of smaller, slower-conducting motor units owing to increased central excitability while larger motor neurons are blocked because of too rapid activation, as discussed earlier. Correlations between F-wave latencies, durations, and amplitudes are also disturbed in patients with motor disorders of central origin.P? These data suggest that analyses of F waves could be used to define different abnormal states of the motor system for clinical purposes. Knowledgeable use of F waves requires an understanding that these responses originate at the interface between the central and the peripheral nervous systems. F-wave studies can provide physiologic insight

into that interface. F1M amplitude ratios, for example, are increased in patients with polyneuropathy, as well as in those with spastic hyperreflexia.P' Log F1M values are normally directly correlated with neuromuscular efficiency as defined by twitch tension/M-wave amplitudes. This relationship is disturbed most prominently in patients with CNS abnormalities, but is also disturbed in patients with peripheral nerve dysfunction.P''

REFERENCES 1. Hoffman P: Uber die Beziehungen del' Schnenreflexe zur willkurlichen Bewegun zum Tonus. Z BioI, 68:351, 1918 2. Hoffman P: Untersuchungen libel' die Eigenreflexe [Sehnenreflexe] menschilicher Muskeln. Springer, Berlin, 1922 3. Magladery JW, McDougal DB: Electrophysiological studies of nerve and reflex in normal man. I: Identification of certain reflexes in the electromyogram and the conduction velocity of peripheral nerve fibers. Bull Johns Hopkins Hosp, 86:265, 1950 4. Burke D, Gandevia SC, McKeon B: The afferent volleys responsible for spinal proprioceptive reflexes in man. J Physiol, 339:535, 1983 5. Katirji MB, Weissman ]D: The tibial H reflex and the ankle jerk. Muscle Nerve, 11:971, 1988 6. Weintraub ]R, Madalin K, Wong M et al: Achilles tendon reflex and the H response. Muscle Nerve, 11:972, 1988 7. Katirji B, Weissman ]D: The ankle jerk and the tibial H-reflex: a clinical and e1ectrophysiological correlation. Electromyogr Clin Neurophysiol, 34:331, 1994 8. Gottlieb GL, Agarwal GC: Extinction of the Hoffman reflex by antidromic conduction. Electroencephalogr Clin Neurophysiol, 41:19, 1976 9. Hilgevoord AA, Rour L], Koehlman ]H et al: Soleus H reflex inhibition in controls and spastic patients: ordered occlusion or diffuse inhibition? Electroencephalogr Clin Neurophysiol, 97:402, 1995 10. Renshaw B: Influence of discharge of motor neurons upon excitation of neighboring motor neurons. J Neurophysiol, 14:167, 1941 11. Eccles ]C, Fatt P, Koketsu K; Cholinergic and inhibitory synapses in a pathway from motor-axon collaterals to motor neurons.] Physiol, 126:524, 1954 12. Renshaw B: Central effects of centripetal impulses in axons of spinal ventral roots.] Neurophysiol, 9:190, 1946 13. Cullheim S, KellerthJ-O, Conradi S: Evidence for direct synaptic interconnections between cat spinal alpha-motor neurons via the recurrent axon collaterals: a morphological study using intracellular injection of horseradish peroxidase. Brain Res, 132:1, 1977 14. Gogan P, Gueritaud]P, Horcholle-Bossavit G et al: Direct excitatory interactions between spinal motoneurones of the cat.] Physiol, 272:755, 1977 15. Trontelj]V: A study of the H-reflex by single fiber EMG. J Neurol Neurosurg Psychiatry, 36:951, 1973 16. Bussel B, Pierrot-Deseilligny E: Inhibition of human motoneurones, probably of Renshaw origin, elicited by an orthodromic motor discharge.] Physiol, 269:319, 1977

H-Reflex and F-Response Studies

17. Thomas.JE, Lambert EH: Ulnar nerve conduction velocity and H-reflex in infants and children. J Appl Physiol, 15:1, 1980 18. Delwaide P.J: Contribution of human reflex studies to the understanding of the pyramidal syndrome. p. 77. In Shahani BT (ed): Electromyography in CNS Disorders: Central EMG. Butterworth, Boston, 1984 19. Misiaszek.JE: The H-reflex as a tool in neurophysiology: its limitations and uses in understanding nervous system function. Muscle Nerve, 28:144, 2003 20. Day BL, Marsden CD, Obeso JA et al: Reciprocal inhibition between muscles of the human forearm. J Physiol, 349:519,1984 21. Gottlieb GL, Agarwal GC: Effects of initial conditions on the Hoffman reflex. J Neurol Neurosurg Psychiatry, 34:226, 1971 22. .Jabre JF: Surface recording of the H-reflex of the flexor carpi radialis. Muscle Nerve, 4:435, 1981 23. Fuhr P, Hallett M: Reciprocal inhibition of the H-reflex in the forearm: methodological aspects. Electroencephalogr Clin Neurophysiol, 93:202, 1994 24. Meunier S, Mogyoros I, Kiernan MC et al: Effects of femoral nerve stimulation on the electromyogram and reflex excitability of tibialis anterior and soleus. Muscle Nerve, 19:1110, 1996 25. Hultborn H, Meunier S, Morin AC et al: Assessing changes in presynaptic inhibition of Ia fibers: a study in man and the cat.J Physiol, 389:729,1987 26. Pierrot-Deseilligny E: Assesssing changes in presynaptic inhibition of Ia afferents during movement in humans. J Neurosci Methods, 74:189, 1997 27. Misiaszek .IE, Brooke JD, Lafferty KB et al: Long-lasting inhibition of the human soleus H reflex pathway after passive movement. Brain Res,.677:69, 1995 28. Brooke JD, McIlroy WE, Collins DF et al: Mechanisms within the human spinal cord suppress fast reflexes to control the movement of the legs. Brain Res, 679:255, 1995 29. Olsen PZ, Diamantopoulos E: Excitability of spinal motor neurones in normal subjects and patients with spasticity, parkinsonian rigidity, and cerebellar hypotonia ..I Neurol Neurosurg Psychiatry, 30:325, 1967 30. Hagbarth K-E: The effect of muscle vibration in normal man and in patients with motor disorders. p. 428. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Vol 3. Karger, Basel, 1973 31. Godaux E, DesmedtJE: Human masseter muscle: H- and tendon reflexes. Their paradoxical potentiation by muscle vibration. Arch Neurol, 32:229, 1975 32. Panizza M, NilssonJ, Hallett M: Optimal stimulus for the H reflex. Muscle Nerve, 12:576, 1989 33. Burke D, Adams RW, Skuse NF: The effect of voluntary contraction on the H reflex of human limb muscle. Brain, 112:417, 1989 34. Braddom RI, Johnson EW: Standardization of H reflex and diagnostic use in SI radiculopathy. Arch Phys Med Rehabil, 55:161, 1974 35. Schimsheimer R.I, Ongerboer de Visser BW, Kemp B: The flexor carpi radialis H-reflex in lesions of the sixth and seventh cervical nerve roots. J Neurol Neurosurg Psychiatry, 48:445, 1985

367

36. Schimsheimer R.J, Ongerboer de Visser BW, Kemp Bet al: The flexor carpi radialis H-reflex in polyneuropathy: relations to conduction velocities of the median nerve and the soleus H-reflex latency. J Neurol Neurosurg Psychiatry, 50:447, 1987 37. Miller RG, Kunz N: Nerve conduction studies in infants and children. .I Child Neurol, 1:19, 1986 38. Bryant PR, Eng GD: Normal values for the soleus Hreflex in newborn infants 31-45 weeks post conceptual age. Arch Phys Med Rehabil, 72:28, 1991 39. de Visser BW, Schimsheimer RJ, Hart AAM: The H reflex of the flexor carpi radialis muscle: a study in controls and radiation-induced brachial plexus lesions. J Neurol Neurosurg Psychiatry, 47:1098, 1984 40. Falco FJ, Hennessey \\]' Goldberg G et al: H reflex latency in the elderly. Muscle Nerve, 17:161, 1994 41. Nishida T, Kompoliti A, Janssen I et al: H reflex in 501 radiculopathy: latency versus amplitude controversy revisited. Muscle Nerve, 19:915, 1996 42. Weintraub JR, Madalin 1(, Wang M et al: Achilles tendon reflex and the H response. Muscle Nerve, 11:972,1988 43. Aiello I, Rosati G, Serra G et al: The diagnostic value of H-index in SI root compression . .I Neurol Neurosurg Psychiatry, 44:171,1981 44. Teasdall RD, Park AM, Languth HW et al: Electrophysiological studies of reflex activity in patients with lesions of the nervous system. II: Disclosure of normally suppressed monosynaptic reflex discharge of spinal motoneurones by lesions of lower brainstem and spinal cord. Bull Johns Hopkins Hosp, 91:245,1952 45. Hilgevoord AA.I, Koelman JHTM, Bour LJ et al: Normalization of H-reflex recruitment curves in controls and a population of spastic patients. Electroencephalogr Clin Neurophysiol, 93:202, 1994 46. Stahl SM, Layzer RB, Aminoff MJ et al: Continuous cataplexy in a patient with a midbrain tumor: the limp man syndrome. Neurology, 30:1115, 1980 47. Leis AA, Kronenberg MF, Stekafovii 1 et al: Spinal motoneuron excitability after acute spinal cord injury in humans. Neurology, 47:231,1996 48. Koelman .IHTM, Bour LJ, Hilgevoord AAJ et al: Soleus H-reflex test and clinical signs of the upper motor neuron syndrome.J Neurol Neurosurg Psychiatry, 56:776, 1993 49. Panizza M, Hallett M, Nilsson.J: Reciprocal inhibition in patients with hand cramps. Neurology, 39:85, 1989 50. Panizza M, Lelli S, Nilsson J et al: H-reflex recovery curve and reciprocal inhibition of H-reflex in different kinds of dystonia. Neurology, 40:824, 1990 51. Floeter MK, Andermann F, Andermann E et al: Physiological studies of spinal inhibitory pathways in patients with hereditary hyperekplexia. Neurology, 46:766, 1996 52. Arteida J, Quesada P, Obeso .IA: Reciprocal inhibition between forearm muscles in spastic hemiplegia. Neurology, 41:286, 1991 53. Yang]F, Fung J, Edamura M et al: H-reflex modulation during walking in spastic paretic subjects. Can J Neurol Sci, 18:443, 1991 54. Katz R, Pierrot-Deseilligny E: Recurrent inhibition of alpha-motor neurons in patients with upper motor neuron lesions. Brain, 105:103, 1982

368

ElEaRODIAGNOSIS INCLINICAL NEUROLOGY

55. Magladery jw McDougal DB, Stoll J: Electrophysiological studies of nerve and reflex activity in normal man. II: The effects of peripheral ischemia. Bull Johns Hopkins Hosp, 86:291, 1950 56. Thorne J: Central responses to electrical activation of the peripheral nerves supplying the intrinsic hand muscles.J Neurol Neurosurg Psychiatry, 28:482, 1965 57. Fox JE, Hitchcock ER: F wave size as a monitor of motoneuron excitability: the effect of deafferentation. J Neurol Neurosurg Psychiatry, 50:453, 1987 58. Trontelj ]V: A study of the F response by single fiber electromyography. p. 318. In Desmedt JE (ed): New Developments in Electromyography and Clinical Neurophysiology. Vol 3. Karger, Basel, 1973 59. Eccles JC: The central action of antidromic impulses in motor nerve fibers. Pflugers Arch, 260:385, 1955 60. Yates SK, Brown WF: Characteristics of the F response: a single motor unit study. J Neurol Neurosurg Psychiatry, 42:161,1979 61. Schiller HH, Stalberg E: F responses studied with single fibre EMG in normal subjects and spastic patients. J Neurol Neurosurg Psychiatry, 41:45, 1978 62. Fisher MA: Relative changes with contraction in the central excitability state of the tibialis anterior and calf muscles.J Neurol Neurosurg Psychiatry, 43:243,1989 63. Feasby TE, Brown WF: Variation of motor unit size in the human extensor digitorum brevis and thenar muscles. J Neurol Neurosurg Psychiatry, 37:916,1974 64. Fisher MA: F waves. Muscle Nerve, 8:71, 1985 65. Kimura J, Yanagisawa H, Yamada T et al: Is the F wave elicited in a select group of motor neurons? Muscle Nerve, 7:392, 1984 66. Dengler R, Kossev A, Wohlfahrt K et al: F waves and motor unit size. Muscle Nerve, 15:1138, 1992 67. Doherty TJ, Komori T, Shahsuk DW et al: Physiological properties of single thenar motor units in the F-response of younger and older single adults. Muscle Nerve, 17:860, 1994 68. Fisher MA: Evidence for selective activation of faster conducting motor fibers in F waves. Muscle Nerve, 11:983, 1988 69. Shahani BT, Macleod WE, Burdocks GM: Minimal F response latencies as a measure of conduction in largestdiameter alpha motor axons. Neurology, 37:Suppl I, 114, 1987 70. GuiloffRJ, Modarres-Sadeghi H: Preferential generation of recurrent responses by groups of motor neurons in man. Brain, 114:1771, 1991 71. Fisher MA, Hoffen B, Hultman C: Normative F waves and the number of recorded F waves. Muscle Nerve, 17:1185,1994 72. Vatine .II, Gonen B: Behavior of F-response and determination of actively involved motor neurons. Electromyogr Clin Neurophysiol, 36:349, 1996 73. KimuraJ: 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, 24:539, 1974 74. 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, 1975

75. Panayiotopoulos CR, Scarpalezus S, Nostas PE: F wave studies of the deep peroneal nerve.] Neurol Sci, 31:319, 1977 76. BurgJ, Grimby L, Honers S: Axonal conduction velocity and voluntary discharge properties of individual short toe extensor motor units in man. J Physiol, 277:143, 1978 77. EcclesJC, Eccles RM, Iggo A et al: Distribution of recurrent inhibition among motor neurons. J Physiol, 159:479, 1961 78. Granit R, Pascoe JE, Steg G: The behavior of tonic alpha and gamma motor neurons during stimulation of recurrent collaterals.J Physiol, 138:381, 1957 79. Cullheim S, Kellerth HO, Conrado S: Evidence for direct synaptic interconnections between cat spinal alpha-motoneurones via the recurrent axon collaterals: A morphological study using intracellular injection of horseradish peroxidase. Brain Res, 132:1, 1977 80. Cullheim S, Kellerth JO, Conrado S: A morphological study of the axons and recurrent collaterals of alpha motoneurones supplying different hind-limb muscles. J Physiol, 281:285, 1978 81. Tauc L: Identification of active membrane areas in the giant neuron of Aplysia.] Gen Physiol, 45:1099,1962 82. Petajan ]H: F-waves in neurogenic atrophy. Muscle Nerve, 18:690, 1985 83. Marra TR: F wave measurements: a comparison of various recording techniques in health and peripheral nerve disease. Electromyogr Clin Neurophysiol, 27:33, 1987 84. Chroni E, Taub N, Panayiotopoulos CP: The importance of sample size for the estimation of F wave latency parameters in the peroneal nerve. Electroencephalogr Clin Neurophysiol, 101:375, 1996 85. Payiotoupoulos CP, Chroni E: F-waves in clinical neurophysiology: a review, methodological issues and overall value in peripheral neuropathies. Electroencephalogr Clin Neurophysiol, 101:365, 1996 86. Raudino F: F-wave: sample size and normative values. Electromyogr Clin Neurophysiol, 37:107, 1997 87. Fisher MA: F response latency determination. Muscle Nerve, 5:730, 1982 88. Peioglou-Harmoussi S, Howel D, Fawcett PRW et al: F-response behavior in a control population. J Neurol Neurosurg Psychiatry, 48:1152, 1985 89. Kiers L, Clouston P, Zuniga G et al: Quantitative studies of F responses in Guillain-Barre syndrome and chronic inflammatory demyelinating polyneuropathy. Electroencephalogr Clin Neurophysiol, 93:255, 1994 90. NobregaJAM, Manzano GM, Novo NF et al: A comparison between different F wave parameters in F-wave studies. Clin Neurophysiol, 112:866, 2001 91. Hong C-Z,Joynt RL, LinJC et al: Axillary F-Ioop latency of ulnar nerve in normal young adults. Arch Phys Med Rehabil, 62:565, 1981 92. Fox JE, Hitchcock ER: Changes in F wave size during dentatotomy. J Neurol Neurosurg Psychiatry, 45:1165. 1982 93. Mastaglia FL, Carroll W: The effects of conditioning stimuli on the F-response. J Neurol Neurosurg Psychiatry, 48:182, 1985

H-Reflex and F-Response Studies

94. Delwaide P], Crenna P, Fleron MH: Cutaneous nerve stimulation and motoneuronal excitability: I. Soleus and tibialis anterior excitability after ipsilateral and contralateral sural nerve stimulation. ] Neurol Neurosurg Psychiatry, 44:699, 1981 95. Mercuri B, Wasserman EM, Manganotti P: Cortical modulation of spinal excitability: an F-wave study. Electroencephalogr Clin Neurophysiol, 101:16, 1996 96. DiBenedetto M, Gale SO, Adarrnes 0 et al: F-wave acquisition using low-current stimulation. Muscle Nerve, 28:82,2003 97. Peioglou-Harrnoussi S, Fawcett PRW, Howell D et al: Fresponses: a study of frequency, shape and amplitude characteristics in healthy control subjects. ] Neurol Neurosurg Psychiatry, 48:1159, 1985 98. Buschbacher RM: Median nerve F-wave latencies recorded from the abductor poIlicis brevis. Am J Phys Med Rehabil 78:Suppl, S32, 1999 99. Buschbacher RM: Ulnar nerve F-wave latencies recorded from the abductor digiti. Am J Phys Med Rehabil 78:Suppl, S38, 1999 100. Buschbacher RM: Tibial nerve F-wave latencies recorded from the abductor hallucis. Am ] Phys Med Rehabil 78:Suppl, S43, 1999 1OJ. Buschbacher RM: Peroneal nerve F-wave latencies recorded from the extensor digitorum brevis. Am] Phys Med Rehabil, 78:SuppI. S48, 1999 102. Panayiotopoulos CP: F chronodispersion: a new electrophysiologic method. Muscle Nerve, 2:68, 1979 103. Lachman T, Shahani BT, Young RR: Late responses as aids to diagnosis in peripheral neuropathy. ] Neurol Neurosurg Psychiatry, 43: 156, 1980 104. Fraser ]L, Olney RK: The relative diagnostic sensitivity of different F wave parameters in various neuropathies. Muscle Nerve, 14:912, 1991 105. Ackil AA, Shahani BT, Young RR: Sural nerve conduction studies and late responses in children undergoing hemodialysis. Arch Phys Med Rehabil, 62:487, 1981 106. Walsh ]C, Yiannikas C, McLeod ]G: Abnormalities of proximal conduction in acute idiopathic polyneuritis: comparison of short latency evoked potentials and Fwaves.] Neurol Neurosurg Psychiatry, 47:197,1984 107. Kihara N, Kimura], Kaji R et al: Multicenter analysis on intertrial variability of nerve conduction studies: healthy subjects and patients with diabetic neuropathy. p. 809. In Kimura], Shibasaki H (eds): Recent Advances in Clinical Neurophysiology. Elsevier, Amsterdam, 1996 108. Andersen A, Stalberg E, Falck B: F-wave latency, the most sensitive nerve conduction parameter in patients with diabetes mellitus. Muscle Nerve, 20:1296,1997 109. Argyropoulos Cj, Panayiotopoulos CP, Scarpalezos S: Fand M-wave conduction in amyotrophic lateral sclerosis. Muscle Nerve, 1:479, 1978 110. Panayiotopoulos CP, Scarpalezos S: F-wave studies on the deep peroneal nerve. Part 2. I. Chronic renal failure 2. Limb-girdle muscular dystrophy. J Neurol Sci, 31:331, 1977 Ill. Panayiotopoulos CP: F-wave conduction velocity in the deep peroneal nerve: Charcot-Marie-Tooth disease and dystrophia myotonica. Muscle Nerve, 1:37, 1978

369

112. Fierro B, Modica A, D'Arpa A et al: F-wave study in patients with chronic renal failure on regular haemodialysis.] Neurol Sci, 74:271,1986 113. Kimura], Yamada T, Stevland NP: Distal slowing of motor nerve conduction velocity in diabetic polyneuropathy.] Neurol Sci, 42:291, 1979 114. Fisher MA: F-wave studies: clinical utility. Muscle Nerve, 21:1098, 1998 115. Wulff CH, Gilliatt RW: F waves in patients with hand wasting caused by a cervical rib and band. Muscle Nerve, 2:452, 1979 116. Ongerboer de Visser BW, van del' Sande lJ, Kemp B: Ulnar F-wave conduction velocity in epidural metastatic root lesions. Ann Neurol, I I: 142, 1982 117. Peioglou-Harmoussi S, Fawcett PRW, Howel D et al: Fresponses in syringomyelia. J Neurol Sci, 75:293, 1986 118. Tullberg T, Svanborg E, Isacsson] et al: A preoperative and postoperative study of the accuracy and value of electrodiagnosis in patients with lumbosacral disc herniation. Spine, 7:837, 1993 Il9. SceIsa SN, Herskovitz S, Berger AR: The diagnostic utility of F waves in L5/S1 radiculopathy. Muscle Nerve, 18:1496, 1995 120. Toyojura M, Murakami K: F-wave study in patients with lumbosacral radiculopathies. Electromyogr Clin Neurophysiol, 37:19,1997 121. Berger AR, Sharma K, Lipton RB: Comparison of motor conduction abnormalities in lumbosacral radiculopathy and axonal polyneuropathy. Muscle Nerve, 22:1053,1999 122. Tang LM, Schwartz MS, Swash M: Postural effects on F wave parameters in lumbosacral root compression and canal stenosis. Brain, III :207, 1980 123. Peioglou-Harmoussi S, Fawcett PRW, Howel D et al: F-response frequency in motor neuron disease and cervical spondylosis.] Neurol Neurosurg Psychiatry, 50:593, 1987 124. Shivde AJ, Fisher MA: F chronodispersion in polyneuropathy. Muscle Nerve, 11:960, 1988 125. Fisher MA, Shahani BT, Young RR: Assessing segmental excitability after acute rostral lesions. I. The F response. Neurology, 28:1265, 1978 126. Dory VE, Neufeld MY, Korczyn AD: F-wave characteristics following acute and chronic upper motor neuron lesions. Electromyogr Clin Neurophysiol, 33:441, 1993 127. Fisher MA, Penn RD: Evidence for changes in segmental motoneurone pool by chronic cerebellar stimulation and its clinical significance. J Neurol Neurosurg Psychiatry, 41:630, 1978 128. Risk WS, Bosch EP, Kimura] et al: Chronic tetanus: clinical report and histochemistry of muscle. Muscle Nerve, 4:363, 1981 129. Fisher MA: F response latencies and durations in upper motor syndromes. Electromyogr Clin Neurophysiol, 26:327, 1986 130. Fisher MA: F response analysis of motor disorders of central origin.] Neurol Sci, 62:13, 1983 131. Fisher MA: F/M ratios in polyneuropathy and spastic hyperreflexia. Muscle Nerve, 1l:217, 1988 132. Fisher MA: F responses and neuromuscular efficiencythe relationship in normals and its disturbance with neurological dysfunction. Neurology, 38:Suppl 1,289, 1988

CHAPTER

19

Evaluation of the Autonomic Nervous System MICHAEL J. AMINOFF

ANATOMY Afferent Pathways and Central Structures Sympathetic Efferent Pathways Parasympathetic Efferent Pathways CLINICAL ASPECTS OF DYSAUTONOMIA TESTS OF AUTONOMIC FUNCTION Cardiovascular Tests Heart Rate Variation Response to Valsalva Maneuver Blood Pressure Variation Cutaneous Vasomotor Control Tests of Baroreflex Sensitivity at Rest Sweat Tests Thermoregulatory Sweat Test

Autonomic disturbances are a characteristic feature of certain neurologic disorders; may be a cause of death in others; and sometimes complicate general medical disorders, such as diabetes mellitus. For most clinical purposes, autonomic function is evaluated by a number of simple noninvasive tests. These tests have an important role in several clinical contexts. First, they help to confirm a clinical diagnosis of dysautonomia and to exclude other causes of symptoms. Second, they provide an indication of the extent and severity of autonomic involvement and indicate whether the sympathetic and parasympathetic divisions are affected equally or whether one is selectively involved. Third, they may permit the site of the lesion to be localized more precisely, although this sometimes requires more sophisticated or invasive studies. Finally, autonomic function tests may be helpful in the evaluation of small-fiber neuropathies. In this chapter, attention is directed primarily at investigations that can conveniently be undertaken in a clinical neurophysiology laboratory. The enteric component of the autonomic nervous system is also important, but it will not be discussed here because it is usually not the focus of clinical neurophysiologists.

Sympathetic Skin Response and Related Responses SudomotorAxon Reflex Testing Plasma Catecholamine Levels and Infusions Tests of Pupillary Function Intraneural Recordings Other Investigative Techniques SELECTION OF TESTS TEST FINDINGS IN SPECIFIC DISORDERS Central Nervous System Peripheral Nervous System Acute Autonomic Neuropathies Chronic Autonomic Neuropathies

ANATOMY Details of the anatomy of the autonomic nervous system are provided in standard textbooks; only a short summary of certain aspects of clinical relevance is provided here to facilitate understanding of clinical test procedures and their interpretation.

Afferent Pathways and Central 5trudures Autonomic afferent fibers pass along autonomic or somatic peripheral nerves to the central nervous system (eNS), but their precise pathways have not been well defined. Fibers from the retina pass along the optic nerve and tract to the pretectal nucleus and then to pupilloconstrictor neurons in the Edinger-Westphal nucleus. The trigeminal nerve carries afferent fibers from the cornea and the nasal and oropharyngeal mucosa to the trigeminal nuclei and the nucleus tractus solitarius; their activation causes lacrimation and nasal and oral secretions. The glossopharyngeal and vagus nerves carry afferent impulses from baroreceptors in the carotid sinus and aortic arch to the brainstem

407

408

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

(Fig. 19-1). Cardiac afferent fibers also pass in the vagus nerve and sympathetic nerves. Afferent impulses travel in the vagus nerve from the tracheobronchial tree and abdominal viscera to the nodose ganglion and nucleus tractus solitarius. Somatic afferent fibers also influence autonomic activity. Sensory neurons projecting to the sympathetic system reside in the dorsal root ganglia and relay information to the dorsal horns of the spinal cord. A number of regions within the CNS have an important role in modulating autonomic function. These include the frontal and parietal cortical regions, which may influence heart, blood pressure, and respiratory functions. The cingulate cortex is involved in controlling sphincter (bladder and bowel) functions, and bilateral lesions may therefore lead to sphincter disturbances. The temporal lobe and amygdala have autonomic functions, and autonomic features are well-known accom paniments of seizures arising in these regions. The hypothalamus, cerebellum, and various brainstem nuclei also have major roles.' Afferent fibers from arterial baroreceptors and chemoreceptors end in the nucleus tractus solitarius in the dorsomedial medulla, and this nucleus also receives input from neocortical, forebrain, diencephalic, and rostral brainstem structures (see Fig. 19-1). In turn, it projects to the nucleus ambiguus and dorsal nucleus of the vagus and to the lateral reticular formation; it thus influences the cardiovascular and gastrointestinal systems. Various pontine and medullary regions are involved in the regulation of ventilation.

Central Nervous System

Peripheral Nervous System

Sympathetic Efferent Pathways Descending fibers from the brainstem conduct impulses to the preganglionic sympathetic neurons, which are located in the intermediolateral columns of the spinal cord between about Tl and L2. The axons of these neurons pass to the sympathetic trunk in the white rami communicantes (Fig. 19-2). The sympathetic trunk, on each side of the vertebral column, consists of a chain of ganglia connected by longitudinally running fibers. Within the sympathetic trunk, the axons synapse in the paravertebral ganglia with secondorder neurons. Some of the preganglionic axons pass up or down in the sympathetic trunk to ganglia at other levels before synapsing. There are 3 paired sympathetic ganglia in the cervical region, 12 in the thoracic region, 4 in the lumbar region, 4 or 5 sacrally, and 1 unpaired ganglion in the coccygeal region. The inferior cervical and first thoracic ganglia may fuse to form the stellate ganglion. Unmyelinated postganglionic fibers pass back to the spinal nerves in the gray rami communicantes or form perivascular plexuses along major arteries as they pass to their final destinations. Some preganglionic sympathetic fibers pass through the sympathetic ganglia without a synapse to form the splanchnic nerves; these fibers synapse in the prevertebral (preaortic) ganglia (i.e., the celiac, superior mesenteric, and inferior mesenteric ganglia), from which postganglionic fibers pass to the viscera in the

and

Target Structures

Other Cerebral Regions

~

CIl

c:

~--------------Sympathetic vasomoter fibers in peripheral nerves

FIGURE 19·1 II Diagrammatic representation of the anatomic pathways involved in the baroreceptor reflex regulation of cardiovascular function. (From Aminofl MJ: Electromyography in Clinical Practice. 3rd Ed. Churchill Livingstone, New York, 1998, with perrnission.)

Evaluation ofthe Autonomic Nervous System

Target

409

>--------.. .

FIGURE 19·2 II Sympathetic efferent pathways from the spinal cord.

Paravertebral ganglion

Gray ramus communicans

Target Prevertebral ganglion

hypogastric, splanchnic, and mesenteric plexuses (Table 19-1). The head and neck are innervated by preganglionic neurons in the TI and T2 segments through the superior cervical ganglion and the upper four cervical TABLE 19·1 • The Sympathetic Outflow from the Thoracolumbar Cord

Target Structure

Outflow

spinal nerves. The upper limb is supplied by preganglionic neurons in the T2 to T8 segments and the stellate ganglion, with varying contributions from the middle cervical and upper thoracic ganglia. The legs are supplied from the TI0 to 1.2 segments through the paravertebral ganglia. Postganglionic sympathetic fibers differ in their properties depending on their target organ. Thus, cutaneous sympathetic fibers have differing conduction velocities depending on whether they are vasomotor or sudomotor in function."

SKOIId-order neuron Inparaverte"'al ga.,11a Eyes Salivary glands Face Head Upper limbs Heart Tracheobrochialtree Lungs Pulmonary vessels Trunk Lower limbs

Parasympathetic Efferent Pathways The parasympathetic cranial and sacral outflow is summarized in Table 19-2. The peripheral ganglia are located close to the target organs so that the postganglionic pathways are short.

CLINICAL ASPEOS OF DYSAUTONOMIA

SKOIId-onter neuron In pnvertebral gallliia Nerve Greater splanchnic

Ganglian Celiac

Plexus Hypogastric

Lesser splanchnic

Superior mesenteric

Splanchnic

Small intestine Colon

Lumbar splanchnic

Ihferior mesenteric

Mesenteric

Distal colon Bladder Genitalia

Liver. bile ducts. gallbladder Pancreas Spleen Stomach Small intestine Proximal colon

The clinical features of dysautonomia are described in standard textbooks, but a brief summary is provided here as they point to the need for investigation and are reflected by the manner in which patients are investigated. The most disabling feature is probably an impaired regulation of blood pressure, especially postural hypotension or paroxysmal hypertension. Other cardiovascular abnormalities include syncope, disturbances of cardiac rhythm, and facial flushing. Disturbances of sweating are also common and are manifest by impaired thermoregulatory sweating (anhidrosis or hypohidrosis) or hyperhidrosis. Symptoms of gastrointestinal dysfunction include dysphagia from esophageal peristalsis or impaired relaxation of the esophageal

410

ElEaRODIAGNOSIS INCLINICAL NEUROLOGY

TABLE 19-1 • Parasympathetic Efferent System

Broinstem outflow

Sacral outflow(S2-S4)

Nerve

Peripheral Ganglia

Target Structure

Oculomotor Facial

Ciliary Pterygopalatine Submandibular

Glossopharyngeal Vagus

Otic Ganglia orplexuses related to target organs

Pelvic

Pelvic

Eye Palatine glands Submaxillary glands Sublingual glands Parotid glands Heart Airways and lungs Abdominal viscera Distal colon and rectum Bladder Genitalia

From Aminoff MJ: Electromyography in Clinical Practice. 3rd Ed. Churchill Livingstone, New York, 1998, with permission.

sphincter; early satiety, postprandial discomfort, gastric fullness or distension, and vomiting from gastroparesis; and intestinal pseudo-obstruction, constipation, or diarrhea from altered intestinal motility. Urinary frequency, urgency, and incontinence are features of bladder involvement in some patients, whereas hesitancy, retention, or overflow incontinence occurs in other patients depending on the site of the lesion. Fecal incontinence may occur. Sexual disturbances are common. Impotence may have many causes including an underlying dysautonomia. Lesions of the sacral roots or pelvic nerves may be responsible and, in women, may lead to a failure of arousal. The effect of cord lesions depends on their completeness and segmental level. Neuro-ophthalmologic disturbances are other manifestations of a dysautonomia that are well described and beyond the scope of this chapter. In patients being evaluated for symptoms suggestive of dysautonomia, nonneurologic causes of their complaints require exclusion because they are often reversible. Clinical evaluation is directed with this in mind. The neurologic examination may suggest the cause of a dysautonomia. In particular, it may reveal evidence of a polyneuropathy or of a focal CNS lesion, or a combination of signs (parkinsonism with or without upper or lower motor neuron or cerebellar signs) indicative of a degenerative disorder (e.g., multisystem atrophy).

TESTS OF AUTONOMIC FUNCTION

Cardiovascular Tests Tests of cardiovascular function are important and provide information about both the parasympathetic and sympathetic divisions of the autonomic nervous system.

HEART RATE VARIATION

In normal resting subjects the heart rate is determined mainly by background vagal activity. The laboratory tests of heart rate variation are therefore mainly tests of parasympathetic function. Heart Rate Variation with Breathing An increase in heart rate occurs during inspiration because of decreased cardiac vagal activity; thus, it is blocked by atropine but not by propranolol.P The variation in heart rate that occurs depends on the rate and depth of breathing. The test must therefore be standardized if it is to be used for clinical purposes. Normal values are affected by age, which must also be taken into account. The difference between the maximum and minimum heart rate during breathing decreases with increasing age and is diminished or absent in diabetes and other disorders that affect central or peripheral autonomic pathways.r" as shown in Figure 19-3. For clinical purposes, the recumbent patient is asked to rest quietly for 5 minutes, and is then asked to take deep breaths at the rate of six per minute, for I minute. The timing of the breaths can be directed verbally or by any other convenient means. The heart rate can be measured with a rate monitor or by recording the RR intervals on an electrocardiogram (ECG). The ECG may be displayed on a chart recorder or on standard electromyographic equipment. If the heart rate is measured directly, the difference between the highest and lowest rate during the minute of deep breathing is determined. When the ECG is recorded, several different measurements of RR variation may be made. 5 .6 The measurements made most commonly are of (1) the difference between the longest and the shortest RR interval during the period

Evaluation ofthe Autonomic Nervous System

411

c: .2 Expiration

-l

[

as

Inspiration

a:

90 G)

'S 80

!I.e t'lI E a: ...

70

)8. 60

:I:

50

40

FIGURE 19·J • A, Effect of deep breathing on heart rate in a normal subject, showing a tachycardia during inspiration and a bradycardia with expiration. B, The heart rate response is markedly blunted in patients with dysautonomia, as shown in a patient with multisystem atrophy.

1 -

o

10

I

I

I

20

30

40

50

60

TIme (seconds) A c:

i

Expiration

-i

[

~ Inspiration

~

I'~

100

) en

80

a:s. :I: )

-

[ I I I I I I O~I--:-L=-----:'-::~-~-~~-~-~ 10 20 40 30 50 60

TIme (seconds)

B

of deep breathing; and (2) the expiratory/inspiratory (E/I) ratio, which is the ratio of the mean of the maximum RR intervals in expiration to the mean of the minimum RR intervals in inspiration." The normal range of heart rate variation is agedependent.t-' but normal subjects generally have differences in heart rate exceeding 15 beats per minute. Values of less than 10 beats per minute are abnormal. From a review of the existing literature, Freeman indicated a likely decline in heart rate variability of 3 to 5 beats per minute per decade in normal subjects." The use of a single normal value regardless of age therefore reduces the utility of the test, leading to false-negative results in younger patients and false-positive results in older subjects." The E/I ratio decreases with age, but up to the age of 40 years, ratios less than 1.2 may be regarded as abnormal."

Other factors affecting the results in normal subjects include the time of the test (with increased heart rate variability occurring at night), body weight, physical fitness, medication, and body position."

Heart Rate Response to Change in Posture Immediate Increase in Heart Rate with Standing Heart rate and blood pressure responses to postural change can be measured conveniently on a tilttable. The patient lies supine until consistent values are obtained for at least 5 (preferably 10) minutes, The patient is then tilted so that he or she is in a 60degree head-up position and the heart rate and blood pressure are monitored. The normal decline in systolic and diastolic pressures should not exceed 25 and 10 mmHg, respectively. The heart rate normally

412

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

increases by about 10 to 30 beats per minute, but the response declines with age. 4 ,9 30:15 Ratio

On changing from a recumbent to a standing position, a tachycardia normally occurs and is followed after about 20 seconds by a bradycardia that reaches a relatively stable rate at about the 30th beat after standing (Fig. 19-4). The ratio of the RR intervals that correspond to the 30th and 15th heartbeats has therefore been used as a measure of parasympathetic function.l" This 30:15 ratio decreases with age, but in young adults a ratio of less than 1.04 is regarded as abnormal. The ratio of the absolute maximum to minimum RR interval is sometimes preferred. Atropine blocks the heart rate response to standing, indicating that it depends on vagal innervation of the heart.!" The biphasic response is not present with passive tilting. Power Spectral Analysis of Heart Rate

Various modifications of heart rate testing have been described but are not in widespread use. Power spectral analysis has been described of the heart rate at rest and after postural change. Two major peaks of interest on the power spectrum occur at rest: a high-frequency peak (greater than 0.15 Hz) representing heart rate changes with respiration (parasympathetic activity), and a lowfrequency peak (at 0.05 to 0.15 Hz) that reflects sympathetic and parasympathetic activity.6.11.12 Another component, at very low frequency (less than 0.05 Hz), also

-

104

"e...

96

Co

92

occurs, but its physiologic origin is unclear." A shift in the power spectrum from high to low frequencies occurs with a change in posture to the upright position and may reflect sympathetic activatlon." Although the results of power spectral analysis correlate with the results of other tests of autonomic function, the clinical role of such an approach remains to be defined. RESPONSE TO THE VALSALVA MANEUVER

The Valsalva maneuver consists of a forced expiration against a closed glottis or mouthpiece with a calibrated air-leak. Characteristic changes in heart rate and blood pressure occur during and after performance of the maneuver and relate to changes in cardiac vagal efferent and sympathetic vasomotor activity as a result of stimulation of carotid sinus and aortic arch baroreceptors and other intrathoracic stretch receptors. For clinical purposes, it may be adequate simply to record the heart rate responses with a heart rate monitor (Fig. 19-5) or an ECG. For more detailed information of the changes in heart rate and blood pressure, however, it is necessary to use a servo-plethysmomanometer device (Finapres) or record from an intra-arterial needle (Fig. 19-6). The test is performed with the subject in a semirecumbent position with a rubber clip over the nose. The subject is then required to blow into a mouthpiece (with a calibrated air-leak) connected to a mercury manometer and to maintain an expiratory pressure of

~ 100 c:::

Q)

tn

as

88

Q)

84

a:

76

-as .8

t::

m

::I:

72

t

Standing up

I

-10

o

10

20

30

TIme (seconds) FIGURE 19·4 • Heart rate response to standing in a normal subject, showing an initial tachy-

cardia that is maximal at about the 15th beat after standing, and a subsequent bradycardia.

Evaluation ofthe Autonomic Nervous System

413

110

FIGURE 19-5 II Normal heart rate response to the Valsalva maneuver, recorded with a heart rate monitor. A tachycardia occurs during the forced expiration, and a bradycardia occurs on release of the maneuver.

50 ~

:::::l (I)

£F40[

-E ~ E

0

~-

11...

0

...

r:-l

-'-_ I _...!.-----''--

20

10

c

L-!.

30

-J.....

I

---L

40

50

I

I

---l

60

Time (seconds)

40 mmHg for 15 seconds while the heart rate is recorded. The normal response has four stages. Stages 1 and 3 are artifactual and are characterized by an increase (stage 1) or a decline (stage 3) in blood pressure because of the increase or decrease, respectively, in intrathoracic pressure at the beginning and end of the maneuver. In stage 2, the reduction in venous return

leads to a progressive decline in systolic, diastolic, and pulse pressure, accompanied by a tachycardia resulting from increased cardiac sympathetic activity. The decline in blood pressure is arrested after about 5 to 8 seconds by a reflex vasoconstriction. With release of the blow at the end of the maneuver, the artifactual decline in mean blood pressure as a result of the release

Intraarterial Pressure

(mmHg)

FIGURE 19-6 II Valsalva response as recorded intra-arterially. A, Normal response. B, Abnormal response in a patient with multisystem atrophy. (From Amiuoff MJ: Electromyography in Clinical Practice. 3rd Ed. Churchill Livingstone, New York, 1998, with permission.)

Intraoral Pressure

40

(mmHg)

0

[ L.J

5 sec

A Intraarterial Pressure

(mmHg)

160 80 0

Intraoral Pressure

(mmHg) B

40 0

~

[~

\ L--.J

5 sec

414

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

of intrathoracic pressure (stage 3) is followed by a rebound in blood pressure to above resting levels because of the persisting peripheral vasoconstriction and the increased cardiac output that follows the increased venous return to the heart. This overshoot in the blood pressure, which varies in extent depending upon age, is accompanied by a compensatory, vagally induced bradycardia. Abnormalities of the Valsalva response in dysautonomic patients may take the form of a loss of the tachycardia in stage 2 or of the bradycardia in stage 4; or a lower heart rate in stage 2 than in stage 4. Other abnormalities include a decline in mean blood pressure in stage 2 to less than 50 percent of the resting mean pressure or loss of the overshoot in systolic pressure in stage 4 (see Fig. 19-6). With isolated impairment of efferent sympathetic vasoconstriction, the blood pressure fails to show an overshoot in stage 4, and consequently there is no compensatory bradycardia despite otherwise intact baroreflex pathways. When the response to the Valsalva maneuver is studied simply by recording the ECG, the Valsalva ratio is calculated by dividing the longest interbeat interval occurring after the maneuver by the shortest interbeat interval during it. The highest ratio from three successive attempts, each separated by 2 minutes, is recorded." The ratio reflects both parasympathetic (vagal) and sympathetic function. The normal range of values depends on age, the duration of the forced expiration, and the extent to which intrathoracic pressure is increased. A value of 1.1 or less is commonly regarded as abnormal and a value greater than 1.2 as normal, but in normal subjects younger than 40 years the ratio usually exceeds 1.4. 16 Low values are sometimes recorded in patients with heart and lung disease. The Valsalva ratio is sometimes normal when the blood pressure response is abnormal. 15 BLOOD PRESSURE VARIATION

Change in Posture The effect of postural change on blood pressure is important. The blood pressure is recorded with the subject supine and at rest for at least 10 minutes. The patient than stands with the arm held horizontally to avoid the hydrostatic effect of the column of blood in the dependent ann leading to a falsely elevated blood pressure. The blood pressure is taken immediately on standing and then at l-minute intervals for 5 minutes. In normal subjects a slight decline in systolic pressure may occur, and diastolic pressure typically increases slightly. A decline that is greater than 20 mmHg in systolic pressure or 10 mmHg in diastolic pressure is generally regarded as abnormal. A tilt-table can also be used to evaluate postural changes in blood pressure, but the response may dif-

fer from that obtained by standing because there is less enhancement of the venous return to the heart by contraction of leg and abdominal muscles, and thus greater peripheral pooling of blood. After the patient has been supine for 10 minutes, the table is tilted to an angle of at least 60 degrees, and the patient remains in this upright position for 10 minutes. Blood pressure can be measured with a sphygmomanometer or by continuous recordings from digital arteries using the Finapres device mentioned earlier, which accurately records pressure changes. 17 Prolonged testing (for up to 60 minutes) on a tilttable at an angle of at least 60 degrees is being used increasingly to evaluate patients with suspected syncope. Many studies have shown that blood pressure changes with posture are unrelated to age, but there is no agreement on this point and asymptomatic postural hypotension is common in elderly patients. Postural hypotension occurs in a variety of medical contexts including cardiac disease; endocrine disorders; hypovolemia; and in patients taking medications such as antihypertensive drugs, dopaminergic medication, and CNS depressants. Thus, detailed investigation may be necessary to clarify its cause, including other tests of autonomic function such as the response to the Valsalva maneuver.

Isometric Exercise Sustained handgrip increases heart rate and blood pressure," partly by central command and partly by changes in contracting muscles that activate fibers subserving the afferent limb of the reflex arc." The semirecumbent subject is required to maintain a pressure of 30 percent of maximal handgrip pressure for 5 minutes. The change in diastolic blood pressure is defined as the difference between the last value recorded before release of handgrip pressure and the mean resting value over the 3 minutes of recording preceding the isometric exercise.P An increase in diastolic pressure of at least 15 mmHg is normal and is not dependent on age. The test is not widely used, however, because the results are not as reproducible as the response to postural change. CUTANEOUS VASOMOTOR CONTROL

Cutaneous blood flow is altered by stimuli such as a sudden inspiratory gasp, mental stress, startle, or alteration in temperature of another part of the body. Such changes in blood flow can be examined by plethysmography or laser Doppler velocimerry.?' as illustrated in Figure 19-7. In normal subjects, an inspiratory gasp leads to a digital vasoconstriction through a spinal or brainstem reflex. The response is

Evaluation ofthe Autonomic Nervous System

415

Sudden inspiratory

60

1

en 50 E:l::: .2 § 40 (I)

~(I)

"C

~ 30

0_

~ ~ 20 10

o

10

20

30

40

50

60

Time (seconds) FIGURE 19-7 II Effect of a sudden deep inspiration on blood volume of the index finger, recorded photoplethysmographically by an infrared emitter and detector on the pad of the finger. The upper trace shows the amplitude of each pulse in relative units. The lower trace is the sensor output after amplification and reflects absolute blood volume in the finger. Each wave represents a heartbeat, and the amplitude of the waves reflects blood volume in the region of the sensor. The apparent shift of the DC signal component results from the long time-constant used to prevent loss of signal information.

lost in the presence of a cord lesion or dysfunction of sympathetic efferent fibers to the digit under study, such as in patients with a polyneuropathy or entrapment neuropathy.22,23 A cold stimulus to the opposite hand, such as water at 4°C, similarly leads to reflex vasoconstriction. The digital vasoconstriction that occurs in response to mental stress (e.g., performing mental arithmetic despite distraction, or startle from a sudden loud noise) causes a transient increase in sympathetic vasomotor activity without involving a reflex arc and thus evaluates sympathetic efferent pathways directly. Normal subjects sometimes have no response, however, leading to false-positive results. TESTS OF BAROREFLEX SENSITIVITY AT REST

The relationship between heart rate and blood pressure can be analyzed on a beat-to-beat basis, quantifying the mathematical relationship between increase in systolic pressure and decline in heart rate, and vice versa. The slope of the regression between spontaneous increments or decrements in blood pressure and alterations in heart rate or RR interval will provide a measure of baroreflex sensitivity. Values in the order of 6 to 8 msec per mmHg have been calculated." and correlate with values obtained by phenylephrine infusion.P At present, the method is more of academic interest than practical relevance.

Sweat Tests Disturbances of sweating are common in dysautonomia and can be evaluated by several different approaches. Techniques that have not gained widespread acceptance (e.g., the sweat imprint technique/") will not be discussed. THERMOREGULATORY SWEAT TEST

The thermoregulatory sweat test evaluates central and efferent sympathetic sudomotor function. The patient is warmed with a radiant heat cradle so as to increase the body temperature by 1°C. The skin is covered with a powder that changes color when moist, the most commonly used being alizarin red. This permits the presence and distribution of sweating to be characterized. Disturbances may reflect preganglionic or postganglionic lesions. A symmetrical distal loss is suggestive of a polyneuropathy. Patchy or segmental anhidrosis may suggest a disorder such as leprosy'? or Ross (Ross-Adie) syndrome.P' The procedure is somewhat time-eonsuming, messy, and requires the patient to be unclothed. SYMPATHETIC SKIN RESPONSE AND RELATED RESPONSES

A simple test of sudomotor function is the so-called galvanic skin response. A reduction in the electrical

416

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

resistance of the skin occurs in response to a deep inspiration or noxious stimulus. It is related to increased sudomotor activity. This can be detected as a voltage or current change between the region under study and an indifferent area (Fig. 19-8). The sympathetic skin response (SSR) represents the change in voltage measured at the skin surface following a single electrical stimulus. It depends in part on the electrical activity arising from sweat glands and is conveniently and quickly recorded in a clinical neurophysiology Iaboratory-? Pairs of surface electrodes are placed on the hands and feet, with the active electrodes on the palmar or plantar surface and the remote electrodes placed dorsally; limb temperature is maintained at 32 to 34°e. The electrical stimulus has a duration of 0.1 msec and an intensity of 5 to 20 rnA and is delivered randomly to a mixed or cutaneous nerve at the contralateral wrist or ankle. I? The interstimulus interval should be at least 30 seconds. The response is filtered at a bandpass of 0.1 or 0.5 to 1,000 or 2,000 Hz. A response can be obtained in normal subjects, but at present only its absence and not the absolute values of latency or amplitude are held to be of pathologic significance for clinical purposes. The latency in the upper limb is in the order of 1.5 sec and in the lower limb is about 2 sec. This long latency reflects the slow conduction velocity of postganglionic e fibers (approximately 1 m/sec). Abnormalities in the SSR correlate reasonably well with other sweat tests. 30 Although useful and simple to perform, the SSR is variable and habituates rapidly and is nonquantitative in nature." The SSR is absent in some patients with axonal

neuropathies but is generally preserved m those with demyelinative neuropathies. SUDOMOTOR AXON REFLEX TESTING

The intradermal injection of acetylcholine (5 to 10 mg) causes local sweating and piloerection if the postganglionic sympathetic fibers are intact. This provides a simple, nonquantitative test of sudomotor and pilomotor functions. Sweat output in response to an axon reflex can be measured directly and accurately in a specially designed chamber placed on the skin. Iontophoresed acetylcholine activates axon terminals, generating impulses that pass antidromically to a branch point and then are conducted orthodromically down another axon to its terminals, where acetylcholine is released, generating a sweat response. Quantitative sudomotor axon reflex testing (QSART) is a sensitive means of assessing postganglionic sympathetic function'< and yields reproducible results. It requires sophisticated and expensive equipment, however, so that it is used only in specialized centers.

Plasma Catecholamine Levels and Infusions The resting plasma norepinephrine level provides a useful index of sympathetic function. The main utility of this measurement is to indicate whether a sympathetic

Inspiratory gasp

6.0

!

'0 5.5 0

.r:

E 5.0 4.5 ~ c as 4.0

,3, t) :::J "0

3.5

8

3.0

c

2.5 I

0

10

20

30

40

50

60

Time (seconds) FIGURE 1'-8 • Galvanic skin response recorded in a normal subject in response to a sudden deep inspiration. There is a change in skin conductance (which is the reciprocal of resistance and is measured in mhos) that is determined by applying a constant voltage across a pair of surface electrodes and measuring the resultant current.

Evaluation of theAutonomic Nervous System

lesion is preganglionic, postganglionic, or both, because this is of pharmacologic importance. The level is diminished with postganglionic but not with preganglionic lesions. The plasma norepinephrine level normally increases on changing from a supine to standing position, but this increment may be markedly attenuated or absent in patients with either pre- or postganglionic lesions. The infusion of pressor drugs also helps to localize sympathetic lesions by demonstrating the presence and severity of denervation supersensitivity. Exaggerated pressor responses and a lower response threshold to sympathomimetic drugs (e.g., phenylephrine, norepinephrine, or epinephrine) are characteristic of degeneration of postganglionic sympathetic fibers. The blood pressure and heart rate are recorded with the patient at rest and then after intravenous infusion of the pressor agent at different doses. For example, an increase in systolic pressure of 40 mmHg usually requires norepinephrine to be infused at a rate of 15 to 20 /lg per minute in normal subjects, 5 to 10 /lg per minute in patients with multisystem atrophy, and 2.5 /lg per minute in pure autonomic failure." With infusion of graded doses of pressor drugs having 110 direct effect on heart rate, baroreflex sensitivity may be measured by the relationship between changes in heart rate and blood pressure. As mentioned earlier, the baroreflex sensitivity is indicated by the slope of the line obtained by plotting the rate of each heartbeat against the systolic pressure of the preceding beat, and is a measure of vagal activity and baroreflex afferents.

Tests of Pupillary Fundion A sympathomimetic agent applied topically leads to dilatation of denervated pupils at concentrations that are ineffective in normal pupils because of denervation sensitivity. For example, 0.1 percent epinephrine, applied to the conjunctival sac, does not affect the normal pupil but leads to dilatation when postganglionic sympathetic fibers are affected. Cocaine hydrochloride (4 percent) instilled into the conjunctival sac normally causes pupillary dilatation, but the response is reduced or absent with lesions of the oculosympathetic pathways. Thus, in Horner's syndrome occurring from an interruption of peripheral sympathetic pathways, the response to cocaine is lost. The responses to 6-hydroxyamphetamine distinguish preganglionic from postganglionic oculosympathetic lesions because pupillary dilatation occurs only when postganglionic sympathetic fibers are intact. There is usually no response of normal pupils to the instillation of very weak solutions of parasympathomimetic agents (e.g., methacholine 2.5 percent or pilocarpine 0.125 percent), but pupillary constriction

417

occurs when parasympathetic innervation is impaired because of denervation supersensitivity.

Intraneural Recordings Intraneural recording of postganglionic sympathetic activity in humans has been important in adding to understanding about the operation of the autonomic nervous system in health and disease.:H-~ti The topic is considered further in Chapter 14. At present. however, such approaches do not have a major role in diagnosing dysautonomias for clinical purposes and are therefore not considered further.

Other Investigative Techniques Radiologic studies and urologic procedures such as uroflowmetry and measurement of urethral pressure profiles are not the province of clinical neurophysiologists and are not considered further. The postganglionic sympathetic innervation of the heart has been studied by recording the myocardial concentration of radioactivity after injection of the marker iodine-123-metaiodobenzylguanidine, which accumulates presynaptically in norepinephrine storage granules. Again, this is not the province of the clinical neurophysiologists and further me thodologie details will not be provided here, but the findings in certain parkinsonian syndromes are considered later in this chapter. Sphincteric electromyography,"? pudendal and perineal nerve conduction studies, and evaluation of certain pelvic reflexes have also been used as a means of evaluating autonomic function. Those techniques useful in evaluating sacral function are considered further in Chapter 30.

SELECTION OF TESTS No full comparative study of autonomic function tests has been reported, but certain general comments can be made. Adequate assessment of autonomic function requires tests of the control of heart rate, blood pressure, and sweating, with evaluation of both sympathetic and parasympathetic pathways." Five simple noninvasive tests of cardiovascular reflexes have been deemed adequate for assessment of diabetic autonomic neuropathy'": the heart rate responses to (1) the Valsalva maneuver, (2) standing (30:15 ratio), and (3) deep breathing; and the blood pressure responses to (4) standing and (5) sustained handgrip. Definite autonomic neuropathy is indicated by an abnormality in two or more tests.

418

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

After a study involving QSART and the heart rate responses to deep breathing and the Valsalva maneuver, Low and co-workers found that abnormalities of vagal function and postganglionic sympathetic function occurred with a similar frequency in patients with diabetic neuropathy.t" They therefore concluded that these tests of parasympathetic and sympathetic function are a sensitive index of autonomic dysfunction and are adequate for investigating patients with suspected dysautonomias. Abnormalities in two or more ofa range of tests (i.e., cardiovascular responses to postural change, isometric exercise, deep breathing, and the Valsalva maneuver; thermoregulatory sweat tests; and plasma norepinephrine levels) correlate well with abnormal responses to the Valsalva maneuver as recorded intra-arterially." This suggests that abnormalities in two or more of such tests are adequate to confirm the presence of a dysautonomia.

TEST FINDINGS IN SPECIFIC DISORDERS The pattern of autonomic abnormalities may help to suggest certain disorders, even though both sympathetic and parasympathetic functions eventually come to be affected.

Central Nervous System Disorders of the CNS associated with autonomic failure include multisystem atrophy (MSA) , pure autonomic failure (PAF) , and Parkinson's disease. Multisystem atrophy is a progressive neurodegenerative disorder characterized by autonomic dysfunction; parkinsonism; upper motor neuron deficits; ataxia; and, sometimes, by lower motor neuron signs. Depending on the predominant clinical feature, the disorder may be designated the Shy-Drager syndrome (dysautonomia), stnatonigral degeneration (parkinsonism), or olivopontocerebellar atrophy (ataxic syndrome). In pure autonomic failure, orthostatic hypotension, commonly associated with bladder dysfunction, impotence, and impaired sweating, is not accompanied by clinical evidence of CNS involvement. In MSA, pathologic changes in the autonomic nervous system involve predominantly the preganglionic neurons, whereas in PAF the sympathetic ganglia and postsynaptic neurons are affected more severely. In both MSA and PAF, postural hypotension is often dramatic. Impaired thermoregulatory sweating and abnormal heart rate responses to deep breathing, the Valsalva maneuver, and postural change are also found. Resting plasma norepinephrine levels are reduced in

PAFbut not in MSA; in both disorders there is no increment with head-up tilt, and standing levels are low.42 In Parkinson's disease, postural hypotension has often been reported and, when present, may be iatrogenic. The literature is conflicting, but in the author's experience the cardiovascular reflexes are usually preserved." However, several recent studies have suggested that many patients with Parkinson's disease have cardiac sympathetic denervation that accounts for postural hypotension.r'-f In such patients, responses to the Valsalva maneuver are also abnormal. Abnormalities of the sympathetic skin response 46,47 and of heart rate variation'" have been found, but others have reported normal heart rate variability." Other dysautonomic symptoms (e.g., disturbances of bladder or gastrointestinal function and excessive salivation) are relatively common. Minor abnormalities of autonomic function may occur in patients with progressive supranuclear palsy and Huntington's disease." In patients with the rare syndrome of congenital deficiency of dopamine-/3-hydroxylase, there is pure sympathetic failure but normal cholinergic function, including sweating. Postural hypotension is an early feature.

Peripheral Nervous System Autonomic disturbances are an expected accompaniment of many peripheral neuropathies even though they have received far less attention than the motor and sensory disturbances that occur. Such a dysautonomia may be an incidental finding that is more of academic interest than clinical relevance, as in certain entrapment neuropathies. It may be a minor feature of neuropathies such as chronic inflammatory demyelinating neuropathy (ClOP) or that associated with acquired immunodeficiency syndrome (AIDS); and a major factor of others, such as the Cuillain-Barre syndrome or diabetic neuropathy. The dysautonomia may be the presenting feature of the neuropathy or simply an inconsequential accompaniment, its most disabling (even lifethreatening) feature or an asymptomatic manifestation recognized only by laboratory investigations. It may develop acutely, reaching a maximum within 4 weeks, or evolve more insidiously. Clinically, the dysautonomia may involve sympathetic or parasympathetic fibers, or both (pandysautonomia); similarly, an adrenergic or cholinergic predominance may occur. With generalized sympathetic involvement, postural hypotension and anhidrosis are conspicuous, whereas with parasympathetic involvement there is a fixed heart rate, xerostomia, xerophthalmia, gastrointestinal dysfunction (e.g., colicky abdominal pain, nausea, vomiting, satiety, bloating, gastric fullness, and diarrhea, sometimes alternating with constipation), bladder dysfunction, and-in men-

Evaluation of theAutonomic Nervous System

sexual disturbances such as impotence. With localized involvement, the site of the pathologic process governs manifestations. ACUTE AUTONOMIC NEUROPATHIES

Idiopathic Autonomic Neuropathies

Patients present with an acute, monophasic autonomic neuropathy but little if any somatic disturbance. There may be evidence of preceding viral infection, or the neuropathy may represent a paraneoplastic disorder. Such associations suggest an underlying immunologic disturbance; this is supported by studies showing high titers of ganglionic acetylcholine receptor antibody in 40 to 50 percent of patients.i? In patients with a paraneoplastic etiology, other autoantibodies may be found as well or instead, including antineuronal nuclear antibodv I (ANNA-lor anti-Hu) or 2 (ANNA-2), Purkinje cell antibody 2 (PCA-2), and collapsin response-mediator protein 5 antibody (CRMP-5). Such antibodies may be im portan t in suggesting the likely site of any underlying primary tumor. Patients with a pandysautonomia typically complain of dizziness, lightheadedness, visual blurring or fading out, near-syncope, or even loss of consciousness on standing (from postural hypotension); heat intolerance, anhidrosis, or hypohidrosis; and gastrointestinal disturbances. Clinical examination generally confirms the presence of postural hypotension, but repeated examination is sometimes necessary to do so. Other common findings include a fixed heart rate, dry skin, dilated pupils, urinary retention, and paralytic ileus. Autonomic function tests are abnormal and reflect these clinical findings. In addition, thermoregulatory sweat tests or QSART are abnormal, nerve biopsy may show perivascular round cell infiltration, and skin biopsy reveals loss of epidermal C fibers. Nerve conduct ion studies are usually normal; mild sensory changes are sometimes found. The prognosis is variable, with only about 30 percent making a good recovery, usually within 1 year of onset.?" Treatment is supportive, but immunomodulating therapy may be worthwhile in those with a progressive or life-threatening disorder. Paraneoplastic dysautonomia may remit with treatment of the underlying malignancy. Guillain-Barre Syndrome

The somatic features of the Guillain-Barre syndrome are well recognized, but it is the autonomic disturbances that generally are responsible for the fatal outcome that sometimes occurs. Cardiac arrhythmias (most commonly a sinus tachycardia), paroxysmal hypotension or hypertension, and postural hypotension are the most disabling features." Disturbances of bladder or gastrointestinal function also occur. The

419

postural hypotension is multifactorial and may relate to inactivity and bedrest, efferent sympathetic denervation, baroreceptor deafferentation, volume depletion, cardiac abnormalities, or some combination of these and other factors. Autonomic function studies reveal an abnormal Valsalva response, abnormal heart rate responses to induced hypertension or deep breathing, and abnormal sweat tests. 52,53 The dysautonomia is treated symptomatically and supportively. This may require adrenergic blockade for hypertension or a demand pacemaker for bradyarrhythmias. The dysautonomia typically recovers without long-term sequelae. The acute dysautonomia of Cuillairi-Barre syndrome is distinguished by its somatic accompaniments from the acute dysautonomia that sometimes arises as an isolated phenomenon. This latter disorder may be a variant of Cuillain-Barre syndrome, but this remains to be established. Botulism

Botulism, which typically occurs within 36 hours of the ingestion of contaminated food, results from the toxin produced by the type B strain of Clostridium botulinum, which impairs acetylcholine release from nerve terminals. It is characterized initially by the development of blurred vision, ptosis, dysphagia, and weakness of the extraocular muscles; and then by facial and more widespread weakness associated with cholinergic failure. Autonomic involvement is signaled by xerophthalmia, xerostomia, anhidrosis, constipation, paralytic ileus, and urinary retention, and studies indicate the presence of significant postural hypotension. The electrophysiologic features resemble those of the Lambert-Eaton syndrome. Single-fiber electromyography is abnormal, indicating a disorder of neuromuscular transmission that is presynaptic in nature. In the absence of associated muscle weakness, it is sometimes difficult to distinguish the disorder clinically from acute idiopathic cholinergic autonomic neuropathy. Other Causes of Acute Dysautonomia

Dysautonomic symptoms may accompany the acute motor neuropathy of porphyria, and are a feature of certain iatrogenic and toxic neuropathies (e.g., related to use of cisplatin.v' vinca alkaloids.t" paclitaxel [Taxo!] ;'>6 or amiodarone; or following exposure to organic solvents or acrylamide). The circumstances in which they develop usually suggest the underlying etiology, and treatment is supportive, combined with avoidance of the offending agent. In most instances, autonomic function studies have not been undertaken and would anyway add little to the evaluation and management of patients. Abnormalities of postganglionic sympathetic efferent fibers have been found in patients

420

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

receiving vincristine.F' Organic solvent exposure has been associated with abnormalities in heart rate variation with respiration and the Valsalva ratio.P? CHRONIC AUTONOMIC NEUROPATHIES

Autonomic involvement occurs most commonly in axonal (as opposed to demyelinating) neuropathies, and especially in small-fiber neuropathies. Such smallfiber neuropathies are often difficult to recognize clinically or electrophysiologically. Patients typically present with pain, dysesthesias, and allodynia beginning in the feet and then extending proximally in the legs before coming to involve the upper limbs. Evidence of sympathetic involvement is also present. Clinical examination reveals normal power and tendon reflexes, and the only sensory deficit is an impaired appreciation of pinprick and temperature in the legs. Sudomotor abnormalities may be detected by thermoregulatory sweat tests or QSART.fiR The autonomic nervous system is affected in many peripheral neuropathies, although the clinical manifestations may be mild. Autonomic dysfunction is clinically important in neuropathies associated with diabetes, primary amyloidosis, and familial amyloid polyneuropathy, and in some cases of hereditary sensory neuropathy, particularly the Riley-Day syndrome, as well as in the acute disorders discussed in the preceding section. In most other neuropathies, autonomic dysfunction is usually of only minor clinical importance. In entrapment neuropathies, for example, postganglionic sympathetic fibers mediating the cutaneous vasoconstrictor response to an inspiratory gasp may be involved." indicating that small- and large-diameter fibers are affected, but this is of little diagnostic relevance.

Diabetes

Autonomic neuropathy occurs in about 20 percent of diabetic patients, sometimes as an isolated neurologic finding but more commonly as part of a more widespread neuropathic process. It may present with impotence, postural lightheadedness, postprandial bloating, early satiety, gastrointestinal motility disturbances, impaired bladder control, and disturbances of sweating. Abnormalities in tests of parasympathetic function (heart rate response to breathing, change of posture, and the Valsalva maneuver) occur early, whereas abnormalities of sympathetic efferent function (blood pressure response to change in posture and isometric exercise) generally develop at a later stage. Abnormalities of sudomotor function are well described. The QSART is commonly abnormal and indicates involvement of distal postganglionic sympathetic fibers.t"

Amyloid Neuropathy

Primary amyloidosis and familial amyloid polyneuropathy of Portuguese type (FAP type 1) are often complicated by autonomic failure resulting from loss of predominantly unmyelinated and small myelinated peripheral fibers and cell loss in the intermediolateral columns of the cord" Postural hypotension and impotence are common early symptoms; alternating constipation and diarrhea, distal anhidrosis, urinary retention, and cardiac arrhythmias are also common. Tests of sympathetic and parasympathetic function are typically abnormal. 50 Familial Dysautonomia

Familial dysautonomia (Riley-Day syndrome or hereditary sensory and autonomic neuropathy type III) is manifest in childhood. Autonomic disturbances include impaired lacrimation, hyperhidrosis, postural hypotension, and poor temperature control. Widespread abnormalities of autonomic function may be detected on testing." Postural Orthostatic Tachycardia Syndrome

This syndrome is characterized by postural intolerance accompanied by an increase in heart rate of at least 30 beats per minute on standing or tilt-testing. Postural hypotension does not occur. The disorder is often preceded by a viral illness. Accompanying features include xerostomia, xerophthalmia, nausea, and postprandial bloating. The response to head-up tilt is characterized by a marked tachycardia within about 2 minutes. The response to the Valsalva maneuver indicates impaired peripheral vasoconstriction, but cardiovagal responses are normal/" Thermoregulatory sweat tests commonly reveal distal impairment, and the QSART shows that the lesion is postganglionic. 53 The disorder may result from partial sympathetic denervation in the legs, as suggested by studies of norepinephrine spill-over into the venous circulation of the limbs in response to various stimuli. 54

Other Autonomic Neuropathies

In Adie's syndrome (tonic pupils with areflexia), sympathetic failure is sometimes present and segmental anhidrosis (Ross syndrome) may occur. Pathologic studies have indicated reduced cholinergic sweat gland innervation in areas of hypohidrosis, implying a selective degenerative process of cholinergic sudomotor neurons." A Horner's syndrome is sometimes encountered, however, and other dysautonomic symptoms may also be present, suggesting a generalized abnormality of ganglion cells or their projections.t"

Evaluation of the Autonomic Nervous System

Dysautonomic manifestations other than impaired sweating distally are uncommon in alcoholic peripheral neuropathy except when the neuropathy is severe or coexists with Wernicke's encephalopathy, which is accompanied by central autonomic dysfunction. Abnormalities of cardiovascular parasympathetic function occur before blood pressure control mechanisms are involved." Abnormalities in sudomotor function are common." In patients with chronic renalfadure, the degree of autonomic dysfunction relates to the impairment of nerve conduction.v? Parasympathetic cardiovascular abnormalities are more common than sympathetic abnormalities such as postural hypotension and abnormalities of the SSR. The baroreceptor reflexes may also be impaired. Alcohol-dependent patients admitted to a psychiatric department were found to have a higher incideuce than normal subjects of parasympathetic (vagal) disturbances of cardiovascular function as measured by heart rate variability?" Cardiovascular autonomic dysfunction may occur in alcoholic and nonalcoholic chronic liver disease. In nonalcoholic disease, parasympathetic abnormalities are more common than sympathetic changes. il Autonomic neuropathy signifies a worse prognosis than otherwise.?" In rheumatoid arthritis, involvement of postganglionic sympathetic efferent fibers leads to impaired sweating, and vagal involvement leads to abnormalities in the heart rate response to standing, the Valsalva maneuver, and respiration, especially in patients with peripheral neuropathy. i3 Autonomic abnormalities may occur in other connective tissue diseases including systemic lupus ersihematosus. mixed connective tissue disease, and Sjogren's syndrome. 1~I'Prosy has been associated with autonomic dysfunction. i4 Patchy anhidrosis is the most common finding, but cardiovascular abnormalities including postural hypotension also occur. Abnormalities in cardiovascular parasympathetic tests are more common and occur earlier than cardiovascular sympathetic abnormalities. Nevertheless, abnormalities of cutaneous vasomotor reflexes may be found in patients with leprosy?" and in the apparently healthy contacts of such patients?" and, when combined with electrophysiologic testing of peripheral nerve function, may provide a means of determining the spectrum of involved fibers. rt Syncope, impotence, bladder and bowel dysfunction, and anhidrosis are common in patients with human immunodeficiency virus infection. Abnormal autonomic function test results become more common and more severe in patients with AIDS: up to 80 percent of patients with AIDS have abnorrnalities.P'?" Diphtheria sometimes causes a demyelinating motor polyneuropathy, and this may be associated with abnor-

421

malities in parasympathetic cardiovascular function; sympathetic abnormalities are not found.f" During the chronic phase of Chagas' disease, which is widespread in South America, the gastrointestinal tract and heart may be affected. Clinical manifestations include cardiac arrhythmias, sudden death, and postural hypotension, with abnormal cardiovascular reflexes." It is generally believed that only minor, infrequent disturbances in autonomic function occur in CIDP,82 but a recent study found a variety of abnormalities, including SSR abnormalities, in 50 percent of patients. 83 The abnormalities may involve sympathetic or parasympathetic components; in the former, both vasomotor and sudomotor fibers may be affected. In patients with hereditary neuropathies, pupillary reflexes may be abnormal and sweating is sometimes impaired distally, but cardiovascular reflexes are usually preserved in patients with hereditary motor and sensory neuropathy types I and II. R4 In hereditary sensory and autonomic neuropathies other than Riley-Day syndrome (discussed earlier), anhidrosis is the most common dysautonomic feature. The dysautonomic manifestations of Fabry's disease include hypohidrosis or anhidrosis and gastrointestinal dysfunction with indigestion, nausea, belching, esophageal reflux, abdominal pain, flatus, and diarrhea. Autonomic tests reveal pupillary abnormalities, abnormal tear and saliva production, widespread anhidrosis, reduced cutaneous flare response to scratch and intradermal histamine, and abnormal colonic motility.'" Cardiovascular autonomic tests are normal. Enzyme replacement therapy may lead to improvement or normalization of sweating, as shown by QSART or thermoregulatory sweat tests."

REFERENCES 1. Aminoff Mj: Autonomic nervous system. p. 93. In Fogel BS, Schiffer RB, Rao SM (eds): Neuropsychiatry. Williams & Wilkins, Baltimore, 1996 2. Wallin BG, Elam M: Microneurography and autonomic dysfunction. p. 243. In Low PA (ed): Clinical Autonomic Disorders. Little, Brown, Boston, 1993 3. Wheeler T, Watkins PJ: Cardiac denervation in diabetes. BMj, 4:584,1973 4. Ingal! T], McLeodjG, O'Brien PC: The effect of ageing on autonomic nervous system function. Aust N Z .I Med, 20:570, 1990 5. Stalberg EV, Nogues MA: Automatic analysis of heart rate variation. I. Method and reference values in healthy controls. Muscle Nerve, 12:993, 1989 6. Linden D, Diehl RR: Comparison of standard autonomic tests and power spectral analysis in normal adults. Muscle Nerve, 19:556, 1996 7. Sundkvist G, Almer L-O, Lilja B: Respiratory influence on heart rate in diabetes mellitus. BMj, 1:924, 1979

422

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

8. Freeman R: Autonomic testing. p. 483. In Brown WF, Bolton CF, Aminoff MJ (eds): Neuromuscular Function and Disease. WB Saunders, Philadelphia, 2002 9. Vita G, Princi P, Calabro R et al: Cardiovascular reflex tests. Assessment of age-adjusted normal range. J Neurol Sci, 75:263, 1986 10. Ewing OJ, Campbell IW, Murray A et al: Immediate heart rate response to standing: simple test for autonomic neuropathy in diabetes. BMJ, 1:145, 1978 II. 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, 1995 12. Freeman R, Saul JP, Roberts MS et al: Spectral analysis of heart rate in diabetic autonomic neuropathy. Arch Neurol, 48:185, 1991 13. Pomeranz B, Macaulay RJ, Caudill MA et al: Assessment of autonomic function in humans by heart rate spectral analysis. AmJ Physiol, 248: H151, 1985 14. Hilz MJ: Quantitative autonomic functional testing in clinical trials. p. 1899. In Brown WF, Bolton CF, Aminoff MJ (eds): Neuromuscular Function and Disease. WB Saunders, Philadelphia, 2002 15. Low PA, WalshJC, Huang CYet al: The sympathetic nervous system in diabetic neuropathy: a clinical and pathological study. Brain, 98:341,1975 16. Ewing OJ, Campbell IW, Clarke BF: Assessment of cardiovascular effects in diabetic autonomic neuropathy and prugnostic implications. Ann Intern Med, 92:308, 1980 17. Molhoek GP, Wesseling KH, Settels lJM et al: Evaluation of the Penaz servo-plethysmo-manometer for the continuous, noninvasive measurement of finger blood pressure. Basic Res Cardiol, 79:597, 1984 18. Ewing OJ, Campbell IW, Kerr F et al: Cardiovascular responses to sustained handgrip in normal subjects and in patients with diabetes mellitus: a test of autonomic function. Clin Sci, 46:295, 1974 19. Hulteman E, Sjoholm H: Blood pressure and heart rate response to voluntary and non-voluntary static exercise in man. Acta Physiol Scand, 115:499, 1982 20. McLeodJG: Evaluation of the autonomic nervous system. p. 381. In Aminoff MJ (ed): Electrodiagnosis in Clinical Neurology. 4th Ed. Churchill Livingstone, New York, 1999 2 L. Low PA, Neumann C, Dyck PJ et al: Evaluation of skin vasomotor reflexes by using laser Doppler velocimetry. Mayo Clin Proc, 58:583, 1983 22. Aminoff MJ: Involvement of peripheral vasomotor fibres in carpal tunnel syndrome. J Neurol Neurosurg Psychiatry, 42:649, 1979 23. AminoffM]: Peripheral sympathetic function in patients with a polyneuropathy.J Neurol Sci, 44:213,1980 24. Parati G, DiRienzo M, Bertinieri G et al: Evaluation of the baroreceptor-heart rate reflex by 24-hour intra-arterial blood pressure monitoring in humans. Hypertension, 12:214, 1988 25. Pitzalis MY, Matropasqua F, Passantino A et al: Comparison between noninvasive indices of baroreceptor sensitivity and the phenylephrine method in post-myocardial infarction patients. Circulation, 97:1362, 1998

26. Kennedy WR, Sakuta M, Sutherland 0 et al: Quantitation of the sweating deficiency in diabetes mellitus. Ann Neurol, 15:482, 1984 27. Kyriakidis MK, Noutsis CG, Robinson-Kyriakidis CA et al: Autonomic neuropathy in leprosy. Int J Lepr, 51:331, 1983 28. Fealey RD: Thermoregulatory sweat test. p. 245. In Low PA (ed): Clinical Autonomic Disorders. 2nd Ed. LippincottRaven, Philadelphia, 1997 29. Shahani B, Halperin lJ, Boulu PJ et al: Sympathetic skin response: a method of assessing unmyelinated axon dysfunction in peripheral neuropathies.J Neurol Neurosurg Psychiatry. 47:536, 1984 30. Maselli RA, Jaspan JB, Soliven BC et al: Comparison of sympathetic skin response with quantitative sudomotor axon reflex test in diabetic neuropathy. Muscle Nerve, 12:420, 1989 31. Toyokura M, Murakami K: Reproducibility of sympathetic skin response. Muscle Nerve, 19:1481, 1996 32. Low PA, Caskey PE, Tuck RR et al: Quantitative sudomotor axon reflex test in normal and neuropathic subjects. Ann Neurol, 14:573, 1983 33. Polinsky RJ: Multiple system atrophy: clinical aspects, pathophysiology, and treatment. Neurol Clin, 2:487,1984 34. Fagius J, Wallin BG: Long-term variability and reproducibility of testing human muscle nerve sympathetic activity at rest, as re-assessed after a decade. Clin Autonom Res, 3:201, 1993 35. Wallin BG, Elam M: Microneurography and autonomic dysfunction. p. 233. In Low PA (ed): Clinical Autonomic Disorders. 2nd Ed. Lippincott-Raven, Philadelphia, 1997 36. Dotson R, OchoaJ, Marchettini P et al: Sympathetic neural outflow directly recorded in patients with primary autonomic failure: clinical observations, microneurography, and histopathology. Neurology, 40:1079,1990 37. Nahm F, Freeman R: Sphincter electromyography and multiple system atrophy. Muscle Nerve, 28:18, 2003 38. American Academy of Neurology: Assessment: clinical autonomic testing report of the Therapeutic and Technology Assessment Sub-Committee of the American Academy of Neurology. Neurology, 46:873, 1996 39. Ewing DJ: Recent advances in the non-invasive investigation of diabetic autonomic neuropathy. p. 667. In Bannister R (ed): Autonomic Failure. 2nd Ed. Oxford University Press, Oxford, 1987 40. Low PA, Zimmerman BR, Dyck PJ: Comparison of distal sympathetic with vagal function in diabetic neuropathy. Muscle Nerve, 9:592, 1986 41. McLeod JG, Tuck RR: Disorders of the autonomic nervous system, Parts I and II. Ann Neurol, 21:419, 519,1987 42. Low PA, Bannister R: Multiple system atrophy and pure autonomic failure. p. 555. In Low PA (ed): Clinical Autonomic Disorders. 2nd Ed. Lippincott-Raven, Philadelphia, 1997 43. Aminoff MJ: Other extrapyramidal disorders. p. 577. In Low PA (ed): Clinical Autonomic Disorders. 2nd Ed. Lippincott-Raven, Philadelphia, 1997 44. Goldstein OS: Dysautonomia in Parkinson's disease: neurocardiological abnormalities. Lancet Neurol, 2:669, 2003

Evaluation ofthe Autonomic Nervous System

45. Goldstein DS, Holmes CS, Dendi R et al: Orthostatic hypotension from sympathetic denervation in Parkinson's disease. Neurology, 58:1247, 2002 46. Zakrzewska-Pniewska B,Jamrozik Z: Are electrophysiological autonomic tests useful in the assessment of dysautonomia in Parkinson's disease? Parkinsonism Relat Disord, 9:179, 2003 47. Haapaniemi TH, Korpelainen JT, Tolonen U et al: Suppressed sympathetic skin response in Parkinson disease. Clin Auton Res, 10:337, 2000 48. Holmberg B, Kallio M, Johnels B et al: Cardiovascular reflex testing contributes to clinical evaluation and differential diagnosis of parkinsonian syndromes. Mov Disord, 16:217,2001 49. Vernino S, Low PA, Fealey RD et al: Autoantibodies to ganglionic acetylcholine receptors in autoimmune autonomic neuropathies. N EnglJ Med, 343:847, 2000 50. Suarez GA, Fealey RD, Camilleri M et al: Idiopathic autonomic neuropathy: clinical, neurophysiologic, and followlip studies on 27 patients. Neurology, 44:1675, 1994 51. Zochodne DW: Autonomic involvement in Guillain-Barre syndrome: a review. Muscle Nerve, 17:1145,1996 52. Tuck RR, McLeod JG: Autonomic dysfunction in Cuillain-Barre syndrome.J Neurol Neurosurg Psychiatry, 44:983,1981 53. Persson A, Solders G: R-R variation in Guillain-Barre syndrome. Acta Neurol Scand, 67:294,1983 54. Rosenfeld CS, Broder LE: Cisplatin-induced autonomic neuropathy. Cancer Treat Rep, 68:659, 1984 55. Hancock BW, Naysmith A: Vincristine-induced autonomic neuropathy. BMJ, 3:207,1975 56. Jerian SM, Sarosy GA, Link CJ et al: Incapacitating autonomic neuropathy precipitated by taxo!' Gynecol Oncol, !11:277,1993 57. Matikainen E, Juntunen J: Autonomic nervous system dysfunction in workers exposed to organic solvents. J Neurol Neurosurg Psychiatry, 48:1021,1985 58. Stewart JD, Low PA, Fealey RD: Distal small fiber neuropathy: results of tests of sweating and autonomic cardiovascular reflexes. Muscle Nerve, 15:661, 1992 59. Ando Y, Suhr OB: Autonomic dysfunction in familial amyloidotic polyneuropathy (FAP). Amyloid, 5:288, 1998 60. Bernardi L, Passino C, Porta C et al: Widespread cardiovascular autonomic dysfunction in primary amyloidosis: does spontaneous hyperventilation have a compensatory role against postural hypotension? Heart, 88:615, 2002 61. Axelrod FB: Familial dysautonomia. Muscle Nerve, 29:352, 2004 62. Sandroni P, Novak V, Opfer-Gehrking TL et al: Mechanism of blood pressure alterations in response to the Valsalva maneuver in postural tachycardia syndrome. Clin Auton Res, 10:1, 2000 63. Low PA, Vernino S, Suarez G: Autonomic dysfunction in peripheral nerve disease. Muscle Nerve, 27:646, 2003 64. Jacob G, Costa F, ShannonJR et al: The neuropathic postural tachycardia syndrome. N Engl J Med, 343:1008, 2000 65. Sommer C, Lindenlaub T, Zillikens D et al: Selective loss of cholinergic sudomotor fibers causes anhidrosis in Ross syndrome. Ann Neurol, 52:247, 2002

423

66. Shin RK, Galetta SL, Ting TY et al: Ross syndrome plus: beyond Horner, Holmes-Adie, and harlequin. Neurology, 55:1841,2000 67. Duncan G,Johnson RH, Lambie DG et al: Evidence of vagal neuropathy in chronic alcoholics. Lancet, 2:1053, 1980 68. Low PA, WalshJC, Huang C-Yet al: The sympathetic nervous system in alcoholic neuropathy: a clinical and pathological study. Brain, 98:357, 1975 69. Wang SJ, Liao KK, Liou HH et al: Sympathetic skin response and R-R interval variation in chronic uremic patients. Muscle Nerve, 17:4Il, 1995 70. Rechlin T, Orbes I, Weis M et al: Autonomic cardiac abnormalities in alcohol-dependent patients admitted to a psychiatric department. Clin Auton Res, 6:Il9, 1996 71. Thuluvath PJ, Triger DR: Autonomic neuropathy in chronic liver disease. QJ Med, 72:737, 1989 72. Hendrickse MT, Thuluvath PJ, Triger DR: Natural history of autonomic neuropathy in chronic liver disease. Lancet, 339:1462,1992 73. Edmonds ME, Jones TC, Saunders WA et al: Autonomic neuropathy in rheumatoid arthritis. BMJ, 2:173, 1979 74. Ramachadran A, Neelan PN: Autonomic neuropathy in leprosy. IndianJ Lepr, 59:277,1987 75. Wilder-Smith EP, Wilder-Smith AJ, Nirkko AC: Skin and muscle vasomotor reflexes in detecting autonomic dysfunction in leprosy. Muscle Nerve, 23:1105, 2000 76. Wilder-Smith E, Wilder-Smith A, Egger M: Peripheral autonomic nerve dysfunction in asymptomatic leprosy contacts. J Neurol Sci, 150:33, 1997 77. Abbot NC, BeckJS, Mostofi S et al: Sympathetic vasomotor dysfunction in leprosy patients: comparison with electrophysiological measurement and qualitative sensation testing. Neurosci Lett, 206:57, 1996 78. Ruttimann S, Hilti P, Spinas GA et al: High frequency of human immunodeficiency virus-associated autonomic neuropathy and more severe involvement in advanced stages of human immunodeficiency virus disease. Arch Intern Med, 151:2441, 1991 79. Welby SB, Rogerson SJ, Beeching NJ: Autonomic neuropathy is common in human immunodeficiency virus infection. J Infect, 23:123, 1991 80. Idiaquez J: Autonomic dysfunction in diphtheritic neuropathy.J Neurol Neurosurg Psychiatry, 55:159,1992 81. Fernandez A, Hontebeyrie M, Said G: Autonomic neuropathy and immunological abnormalities in Chagas' disease. Clin Autonom Res, 2:409, 1992 82. Ingall TJ, McLeod JG, Tamura N: Autonomic function and unmyelinated fibers in chronic inflammatory neuropathy. Muscle Nerve, 13:70, 1989 83. Lyu RK, Tang LM, Wu YRet al: Cardiovascular autonomic function and sympathetic skin response in chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve, 26:669, 2002 84. Ingall TJ, McLeodJG: Autonomic dysfunction in hereditary motor and sensory neuropathy (Charcot-Marie-Tooth disease). Muscle Nerve, 14:1080, 1991 85. Cable \\1L, Kolodny EH, Adams RD: Fabry disease: impaired autonomic function. Neurology, 32:498, 1982 86. Schiffmann R, Floeter MK, Dambrosia JM et al: Enzyme replacement therapy improves peripheral nerve and sweat function in Fabry disease. Muscle Nerve, 28:703, 2003

CHAPTER

21

Visual Evoked Potentials in Clinical Neurology GASTONE G. CELESIA

BASIC TECHNOLOGY

MULTIFOCAL VEPs

NORMATIVE DATA

CONCLUDING COMMENTS

ABNORMAL VEPs Retinopathies and Maculopathies Disorders of the Optic Nerve and Chiasm Retrochiasmatic Disorders

Electrophysiologic recording of visual evoked potentials (VEPs) has been very useful in evaluating visual function.v" Evoked potential techniques are noninvasive and have excellent temporal resolution (in the range of milliseconds), thus permitting the study of dynamic changes occurring in the nervous system.l-! It is important to review some of the fundamental characteristics of visual function before discussing the clinical applications ofVEP studies. The visual system processes information along multiple parallel channels.r" Separation of visual information starts at the neuronal circuitry of the retina, where particular features such as color, contrast, luminance, and other parameters of the stimulus are extracted and processed.l" It has been suggested that at least seven parallel channels of ganglion cells process visual information in the primate retina.'? Recent work has demonstrated several classes of neurons that differ in their visual properties and roles. These retinal neurons send central projections to the lateral geniculate body (LGB). Two separate classes of ganglion cells have been described in the macaque monkey: (1) P cells (also called midget cells); and (2) M cells (also called parasol cells)." These two neuronal groups in the retina have distinct physiologic properties and project separately to segregated brain regions. 7,9- 11 M (magnocellular) pathways can be activated independently ofP (parvocellular) pathways. P cells send their projections to the parvocellular laminae of the LGB, whereas M cells project to the magnocellular laminae." From the LGB, visual information is transmitted to striate area 17. Several projections link area 17 with areas 18, 19, and MT (midtemporal)."!! The output from areas 19 and MT goes to visual areas in the posterior parietal cortex.9,12-18

The magnocellular system is involved primarily with motion analysis. The parvocellular system is associated with color selectivity and shows a preference for high spatial-frequency stimuli l4-18; it overlaps, however, with the magnocellular system in selectivity for orientation. The parvocellular system projects heavily to the inferior temporal cortex. The two major processing systems, directed toward posterior parietal and inferior temporal cortex, exist in all primates. 9,1l,14-18 It is presumed that a similar organization of the visual system exists in humans. At least 10 cortical visual areas have been described in humans,12,13,16 with over one-third of the cerebral cortex devoted primarily to visual function. 12,13 An interesting explanation for the multiplicity of visual cortical areas has been proposed by Marr.I? He suggested that any large computation should be broken down into a collection of smaller independent modules. The multiplicity of cortical representations may provide the structural modules on which new information capabilities have developed in the course of evolution. 12,13,20,21 Thus, increasing the number of visual areas may increase the number of visual abilities and simplify the problem of interconnecting functionally related groups of neurons.s' This brief summary of present knowledge of the visual pathways emphasizes the extent to which the neocortex is involved in parallel visual processing. It also offers opportunities for evaluating the system. Two pragmatic principles can be derived from the data just presented: 1. Visual stimuli not only activate the occipital lobes

but also involve large areas of the temporal and parietal lobes. 453

454

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

2. Different structures of the retina and visual pathways can be preferentially activated by changing the characteristics of the visual stimulus.f The first principle indicates that YEPs can be recorded from a large region of the scalp, essentially from the vertex to the inion. Accordingly, the referential electrode should be located anterior to the vertex or distant from these active regions. The second principle indicates the importance of selecting the proper visual stimulus for the specific clinical problem to be addressed. The use of selective visual stimulation and recording techniques permits the analysis of specific functions within the visual pathways.2,22,23 Selection of the appropriate test requires an understanding of the suspected site of pathology. For instance, if a patient has been referred because of difficulty with night vision and suspected retinopathy, the rod retinal receptor system requires assessment and the appropriate test is full-field flash electroretinography (ERG; discussed in Chapter 20). Conversely, a patient with suspected retrobulbar neuritis should first undergo a pattern-reversal YEP, the most sensitive electrophysiologic test to assess dysfunction of the optic nerve. Only a logical approach to specific diagnostic questions will permit proper utilization of a variety of electrophysiologic tests.

BASIC TECHNOLOGY To improve clinical care and allow comparison between different laboratories it is recommended that a standardized methodology be used. The recommended standards of the International Federation of Clinical Neurophysiology (IFCN) and the International Society for Clinical Electrophysiology ofVision (ISCEV) are followed in this brief description of the basic technology for recording YEPs. The principles of recording the YEP are simple. A visual stimulus is presented to the subject for a selected number of times, and the cerebral responses are amplified, averaged by a computer, and displayed on an oscilloscope screen or printed out on paper. For neurologic purposes, YEPs are generally elicited by monocular stimulation of each eye while the other is covered with a patch. Visual stimuli can be either patterned or unpatterned. Unpatterned stimuli most commonly consist of stroboscopic flashes. Patterned stimuli consist of a specific pattern (e.g., checks or bars) on which the subject is required to fixate. Whenever a pattern stimulus is used, it is important to define the type of pattern, size of the pattern elements, total field size, method of presentation of the pattern, rate of presentation of the pattern, stimulus intensity or luminance, background luminance, and contrast. Any change in these parame-

ters may profoundly modify the response. I- 3,22-25 The two most commonly used patterns are checks and gratings. The pattern should be achromatic (black and white). The size of the individual checks should be expressed in terms of visual angle. The visual angle p is expressed as

P= tan"! (

;

) x 120

where P is the visual angle in minutes of are, W is the width of the checks in millimeters, and D is the distance of the pattern from the corneal surface in millimeters (Fig. 21-1). Although checks are customarily reported in minutes of arc ('), gratings are usually reported in cycles per degree. Measurements in cycles per degree (abbreviated as c/deg) define the spatial frequency of the stimulus. Measurements of the visual angle in minutes of arc can be converted to cycles per degree by the formula 30 c/deg=-W where W is the width of the grating in minutes. In the case of checks, however, W represents the diagonal measure of the checks in minutes of arc. The most common method of presentation of the stimulus is by reversal of a checkerboard pattern; the black checks become white, and vice versa, so there is no change in total luminance (isoluminance) of the pattern. Isoluminance is important in preventing light scatter in the retina. The YEP is affected by the stimulus intensity. The intensity of a visual stimulus is defined as luminance. Luminance is measured by a photometer and is expressed in candela per square meter (cdz'm"). The mean luminance of the field of stimulation is expressed by the formula Lmax + Lmin

2 where Lmax indicates the maximum and Lmin the minimum luminance of the field. A desirable mean field luminance is at or above 100 cd/m 2. Another important parameter that may modify the YEP is contrast. Contrast is defined as the difference between the bright and dark portions of a pattern. It is expressed by the formula C = Lmax - Lmin x 100 Lmax+Lmin where C is the contrast in percent and Lmax and Lmin are the maximum and minimum luminance of the pattern.

Visual Evoked Potentials inClinical Neurology

455

Smm

0.5 meter

1 meter

~

= Tan- 1(S/(2X1000))x120 = 27' .50 = 27' 30"

w= Tan- 1(S/(2X500)) X 120 = 54' .99 = 54' 59"

III Geometry of the visualangle and the relationship between absolute object size and distance. The size of the field (in this example a field containing four checks and having a width of 8 mm) is expressed as visual angle ~ or W. Note that although the absolute size of the field remains 8 mm, the size of the image on the retina changes with the distance. This relationship is reflected by the value of the visual angle, which changes from 27'30" at 1 m to 54'59" at 0.5 m. Further details are provided in the text.

FIGURE 21-1

The reader is referred to the recent guidelines of ISCEV for further information on measurements and calibration of visual stimuli.P" The IFCN recommends that a minimum of three stimuli be used for testing, and suggests the following parameters: 1. Pattern stimuli consisting of either checks or

gratings 2. Size of the pattern elements: 14' to 16',28' to 32', and 56' to 64' 3. Full-field size of at least 8 degrees (l degree = 60 minutes) 4. Contrast between 50 and 80 percent 5. Rate of presentation of 1 Hz (producing a reversal every 500 msec) 6. Mean luminance of the center field of at least 100 cd/m2 7. Background luminance under photopic conditions of at least 30 to 50 cd/m2 A variety of montages have been advocated for recording VEPs. The IFCN suggests a two-channel montage, Oz-Fpz and Oz-A1-A2 (linked ears), with the ground placed at Cz. Reproducible VEPs are ensured with this montage even in cases with a prominent potential gradient at the vertex. The bandpass should be 1 to 250 or 300 Hz. The subject fixates on the center of the pattern during stimulation. The responses are recorded at least twice to ensure their replicability.

NORMATIVE DATA The issue of normative data is a complex one, and two aspects merit particular attention because of their importance and clinical relevance: (1) whether it is appropriate to use the normative data obtained by others; and (2) the effect of changes in stimulus parameters on the responses. VEPs are electric potentials evoked from visual stimuli and recorded from the human scalp. Therefore, if the same stimulation and recording parameters are used, the same response will be recorded. The normative data obtained in one laboratory can be used in another, provided that the physical characteristics of the visual stimulus and the recording arrangements are identical or equivalent in the two laboratories. It may, however, be desirable to verify the validity of the adopted normal values by testing at least 10 normal subjects. The values of the new data should fall within the boundary of normality of the adopted data. VEPs to a pattern-reversing checkerboard (the most commonly used stimulus in clinical laboratories) consist of a set of sequential waveforms (Fig. 21-2). The waveforms are alternately positive and negative and are designated in accordance with their polarity and latency. Positive waves are designated P, followed by a number indicating the peak latency in milliseconds (e.g., P60 and PlOO); negative waves are designated N, followed by a number indicating the peak latency

456

ElECTRODIAGNOSIS INClINICAL NEUROlOGY

P92 P99

D~

A

Log retinal illuminance

2.86

N74

Pupil

P98

::JJ\~

N;S.5V

Dilated

P103.5

P103

2.44

Jv

Natural

P\16

1.73

P1Q7.5

~ constricted~ [ 5 !LV

I

I

I

I

I

I

I

'

,

,

I

25 msec/div FIGURE 21·] • Pattern-reversal visual evoked potentials to 15'

2.5 !LV

1+,

I

,

I

I

I

I

I

,

I

t+, ,

25 msec/div FIGURE 21·2 • Examples of visual evoked potentials in normal

subjects with 20/20 visual acuity. The visualstimulus consisted of pattern-reversing checks subtending 31' at the eye. The total field of the stimulus subtended 22 degrees and had a mean luminance of 116 cd/m 2• The most frequent morphology is shown in A, B, and C. Less common variations are shown in D, E, and F. Note that P50 and even N70 may be absent in normal subjects. The morphology of the response shown in E and F is often referred to as a W configuration. In this and all other illustrations, two trials are superimposed to demonstrate the reproducibility of the potentials.

(e.g., N70 and NI45). There is considerable variation in the morphology of normal VEPs, but the dominant wave is the PIOO component. Figure 21-2 shows the variations in VEP morphology observed in 150 normal subjects. It can be seen that in some subjects the initial negativity (N70) is absent, whereas in other subjects N70 is as large as PIOO. In fewer than 0.5 percent of normal cases, PlOO has a W-shaped configuration (i.e., PIOO is subdivided into two peaks). In these normal subjects, both peaks have latencies within the boundaries of normality. The best way to determine which peak corresponds to PIOO is to obtain VEPs to patterns of three different sizes. Usually the larger checks will yield only one PIOO peak. What is not sufficiently appreciated is the need to maintain constant the luminance, contrast, patternreversal rate, and size of the pattern elements to obtain reproducible and reliable VEPs. Any change in these parameters will inevitably affect the VEPs. As shown in Figure 21-3, a reduction in the amount of light reach-

checks and their relationship to pupil diameter and retinal illuminance. The effect of changes in retinal illuminance is shown in the left column and that of pupil diameter on the right. Note the increase in latency ofPI00 and the change in morphology of N70 with a decline in retinal illuminance or pupillary constriction.

ing the retina will produce a reduction in amplitude and a prolongation in latency of the response. Retinal illuminance (I) is measured in trolands and is calculated by the formula

I=LxA where L is the mean luminance in cd/m 2 and A is the pupil area in mm". Pupil constriction affects the amplitude and the latency of N70 and PIOO in the same manner as decreased luminance of the stimulus does.i? In Figure 21-3, the dilated pupil had a diameter of 7.5 mm with an area of 44.17 mm", whereas the constricted pupil had a diameter of 1.8 mm with an area of 2.54 mm", The stimulus luminance was 26.9 cd/m 2 for both conditions. The variation in pupil size had drastic effects on retinal illuminance, which was 1188.2 trolands with the dilated pupil and 68.3 trolands with the constricted pupil. The latency of PIOO changed from 96 to 107.5 msec, depending on pupillary size. Check size also influences the latency and amplitude of the responses.Pv" A decrease in check size is usually associated with a prolongation ofN70 and PIOO latency. However, the relationship between pattern size and PIOO latency is not linear; certain studies suggest that as check size increases above 30', the latency of PlOO also increases.P This complex relationship indicates the need to establish normative data for every check size used. Age is another important variable that influences the VEP.30 As shown in Figure 21-4, PIOO latency increases

457

Visual Evoked Potentials inClinical Neurology

Check size 15'

115 o 110

Ql Ul

E 105

~t.~~~

•

c

';' 100 o C 95

~

0

0 ,....

a,

90

______

•• • • •

• •

85 80 10

20

30

40

•

~--

--~

• •••• I •

•

•

~--

t'

.1.

•••

•• ••

•

50

•

•

.

• •

70

60

80

Check size 31' 110 ~

105

Ul

E 100 .s

g 95

~ o

• Males

* Females ----=::::;-.-._._._._._.;.-._._._;.a

_.-.-. *...

• •• *..lII'Ii.¥ .*

•

....

*

* • ·II·,.I!'·· * •• •..· "., __ ..._ *_"'_.._ '*- _ 1.

.. *

I

...,. **

• "'"••..·..•~

·~·

. a

JI.

* •• * 90 • * U o • * I · * * • *. a:: 85 • *• * ** * 80 '=----::J=----::"::--'*""-~-_=_=-~--= 10

20

30

40 50 Age in years

60

70

80

FIGURE 21·4 • Effect of age and gender on PlOD latency. Visual evoked potentials were elicited by pattern reversal in 112 normal subjects. There was no statistically significant difference between sexes for responses to 15' checks; thus both men and women are represented by dots. Gender was a significant influence on responses to 31' checks; asterisks refer to women and squares to men. In the upper graph the solid line represents the weighted regression line, and the dashed line the 98 percent confidence boundary calculated so that the probability that the next observation will fall below the boundaries is 0.98. In the lower graph the dashed line and the dotted line represent the weighted regression line for men and women. The dot-dash line and the solid line are the 98 percent confidence boundary for women and men, respectively. The women have shorter latencies than the men.

in older normal subjects. Celesia and co-workers simultaneously recorded pattern ERGs and VEPs in 112 normal subjects and demonstrated that aging influenced the responses at the retinal level.30 The increased latency of the B wave in the pattern ERG was mostly caused by ocular changes related to senescence and only partially caused by aging of the neuronal circuitry in the retina. The effect of aging was more prominent when small checks were used to elicit the responses. There was also an increase in latency of the N70 and PIOO components of the VEP, and this change was not

simply a result of the delayed retinal response. The retinocortical time, expressed as the interpeak interval between the B wave of the pattern ERG and either the N70 or the P100 peak of the VEP, reflects events occurring outside the retina and at some point in the optic nerve, optic pathways, or visual cortices. The retinocortical time was prolonged for responses evoked with 15' checks, but not by 31' checks. The age-related increase in latency of VEP was therefore related to changes in the visual pathways or cortex. Ganglion cell loss,31 dysmyelination, axonal swelling, and nerve fiber loss have been described in the optic nerve. 31.32 Shorter-latency (see Fig. 21-4) and larger-amplitude VEPs than in males have been described in females. 30,33,34 Whether these changes are related to the smaller anatomic size of the head of females or to hormonal factors is unclear. Uncorrected refractory errors may affect the amplitude and latency of the VEP, especially for patterns of small size. 35,36 The effect of defocusing on the VEP is shown in Figure 21-5. Blurring of the pattern stimulus not only prolongs PlOO latency but often drastically changes VEP morphology, with elimination of N70 and broadening of the PlOO wave. Two diopters of blur reduces Snellen visual acuity from 20/20 to 20/120.:17 Patients to be tested not infrequently have 1 or 2 diopters of uncorrected myopia. Certain precautions must be taken in the laboratory before and during the recording of VEPs to prevent changes in the amplitude and latency of VEPs because of ocular factors. Caution should be taken to prevent misinterpretation ofVEP abnormalities resulting from

15' checks

31' checks P103.5

P118

+3D~ P115

~

+2D~~

OD~~

5 ILV

tIl

1

I

1

1

I

I

I

I

tIl

I

I

I

I

I

I

I

I

25 msec/div FIGURE 21·5 • Effect of optic blur on pattern-reversal visual evoked potentials in a normal subject with 20/20 vision. In the left columns are responses to 15' checks and in the right columns are responses to 31' checks. 00 indicates no blur; +20 and +30 indicate that a lens of +2 or +3 diopters was placed in front of the subject's eye to cause visual blurring. The blurring influenced both the latency and morphology of the response. Note the destructive effect of blur on the N70 wave.

458

ELEGRODIAGNOSIS INCLINICAL NEUROLOGY

refractive errors, pupillary changes, and lens opacities as related to pathology in the visual pathways. The following precautions are recommended: (1) the pupil diameter and the visual acuity of each eye should be measured; (2) corrective lenses should be worn to compensate for refractive errors; (3) if a deficit in visual acuity greater than 20/100 is present, the technologist should determine whether it can be corrected with a "pinhole"; and (4) the pupils should not be dilated with mydriatics to prevent interference with accommodation.

sometimes be associated with decreased amplitude. The most severe abnormality is an absent YEP. In patients with small monocular delays in N70 and PIOO, it is useful to look at the intereye latency differ-

Checks 15'

ERG

ABNORMAL VEPs Abnormalities of YEPs have been described in many disorders of the optic nerve, chiasm, and retrochiasmatic visual pathways. 1-~,22-25,28,37-42 The YEP can be considered abnormal when the latency of the PIOO wave is outside the 95th to 99th percentile boundaries established for normal individuals or when the PIOO is absent. Delay or absence of the N70 peak is more difficult to evaluate because this component is so variable in different subjects. Various YEP abnormalities are shown in Figure 21-6. The most common abnormality is characterized by a normal amplitude but prolonged latency of N70 and PIOO, but a prolonged latency may

as

00

~~ ---..... RCT =? RCT = 53 P94.5 P113

VEP

~ Checks 31'

A

15' checks

31'checks

P130

P125

854.5

ERG~

~

~

RCT = 47

P114

B

~ P198

C

P101.5

P195

~~

D ,

I

I

,

,

,

I

25 msec/div FIGURE 21·6 • Different types of abnormality of pattern-reversal visual evoked potentials. In the left column are responses to 15' checks and in the right column are responses to 31' checks. A, Delayed PIOO of normal amplitude. B, Prolonged latency and small amplitude to 15' checks, but normal amplitude and prolonged latency of PIOO to 31' checks. C, Very delayed and small responses. D. Absent responses.

VEP~

r

RCT = 45.5 P85.5

N65.5

256 msec FIGURE 21·1 • Electrophysiologic studies in a 35-year-old man with central serous choroidopathy of the right eye. He has normal pattern electroretinograms (ERGs) and visual evoked potentials (VEPs) to stimulation of the unaffected eye (OS). Stimulation of the right eye (OD) shows absent pattern ERG to 15' checks and a delayed PIOO response in the VEP. Stimulation with 31' checks evokes an abnormal pattern ERG with a delayed B wave and mildly delayed PIOO. The retinocortical time (RCT) , however, remains within normal limits at 47 msec, thus indicating that the pathology is at the retinal level. Calibration bar represents 5 uv for ERGs and 10 uv for VEPs.

Visual Evoked Potentials in Clinical Neurology

ence. A latency difference between the two eyes greater than 10 msec is clearly indicative of pathology on the side with the longer latency.

as

00

VA= 20 20

VA = finger counting only

459

Checks 15'

Retinopathies and Maculopathies Diseases of the retina, especially diseases affecting the macular region, are associated with abnormal VEPs,38,43-47 VEPs to pattern-reversal stimulation are evoked predominantly from the central 10 degrees of the visual field and therefore reflect the function of the macular area and related pathways.I-3,7,22,23,25 Maculopathies, including macular edema, central serous choroidoparhy," Stargardt's disease, vitelliruptive macular degeneration.t" and age-related macular degeneration (senile macular degeneration) ,38,43,45,46 affect the VEPs because they disrupt vision. Similarly, diffuse retinopathies, when they involve the macular and perimacular region, are associated with abnormal VEPs. Retinal infarcts and scars involving the macular region, cone dystrophy, diabetic retinopathy with macular edema, cancer-associated retinopathy, and retinitis pigmentosa with paracentral scotoma have all been reported to affect VEPs.38,43,46,47 Celesia and co-workers have shown that in maculopathies, simultaneous recording of pattern ERGs and VEPs may be diagnostically useful in differentiating between involvement of the macula and the optic nerve.i" In maculopathies and in retinopathies involving the macular region, pattern ERGs (Figs. 21-7 and 21-8) are absent or have a prolonged latency of the B wave (often referred as P50). The retinocortical time (see p. 457) remains within normal boundaries, thus indicating that the PI00 delay occurs at the retinal levep8,43 In optic nerve diseases, VEPs may be absent or prolonged, but pattern ERGs are normal; the retinocortical time is thus prolonged, indicating that the pathology is postretinal. Only in cases of optic nerve disorder with large central scotoma or optic atrophy are both pattern ERGs and VEPs absent. In these cases it is not possible to distinguish electrophysiologically between maculopathy and optic nerve disease (Fig. 21-9). Several authors have suggested that the pattern ERG is useful in the assessment of macular and retinal ganglion function. 4B-50 Holder has shown that pattern ERG B-wave (or P50) abnormalities may be present in the absence of "significant ophthalmoscopic changes."50 He further noted that reduction in pattern ERG N95 with preservation of P50 suggests dysfunction of the optic nerve. This brief review supports the statement of Celesia and Brige1l22 that "electrophysiological responses evoked by visual stimuli cannot and should not be interpreted in isolation or be the sole methodology to assess the function of the visual system." VEPs need to be interpreted in the clinical context in which they were

ERG

P99.5

VEP

NB3

Checks31'

853

ERG

VEP N101.5 N97.5

256 msec FIGURE 21.. • Electrophysiologic studies in a 68-year-old man with senile macular degeneration of the right eye (OD). Pattern electroretinograms (ERGs) are normal on stimulation of the left eye (OS), but visual evoked potentials (VEl's) to 31' checks are small, with a W morphology and a prolonged latency of PlOD (115.5 msec). Pattern stimulation of the affected eye shows an absent ERG and a small, slightly delayed PlOO. Absence of the ERG with preservation of the VEP suggests that the VEP was evoked from stimulation of the perimacular region.

460

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

Disorders of the Optic Nerve and Chiasm

250 31' checks v

ClJ200 0

v

cQ)

v v

fIl

E150

.s E

o

i= 100

:'b

v

v

0

Q)

v

v

0

v 0

o

o

0

100

150

200

250

300

Time in msec 00 FIGUIE11-t. Scatterplot of latency of the PlOOcomponentof

pattern-reversal visual evoked potentials elicited by monocular stimulation of the right (OD) and left (OS) eyes with 31' checks in patients with maculopathies (open circles) and multiple sclerosis (triangles). The findings are compared with those of agematched controls (not shown in the scatter to avoid cluttering, but all falling inside the shaded rectangle). The solid line of the shaded rectangle represents the boundaries of normality. The open rectangle at 145 msec represents the boundary that separates maculopathies from optic nerve demyelination. Note that although J45 msec seems to separate the two groups, there is considerable overlap. The PlOD latency of the patients with multiple sclerosis falls in all three groups.

obtained and correlated with the funduscopic and visual field examination. Abnormalities of VEPs in patients with a normal funduscopic examination (including a normal macula) suggest dysfunction in the optic nerve or visual pathways.

Halliday's group was the first to indicate the sensitivity of pattern-reversal VEPs in the diagnosis of demyelination of the optic nerve." The ability of VEPs to detect clinically silent optic nerve lesions remains a very useful tool for the diagnosis of multiple sclerosis. In a review of 180 patients with multiple sclerosis (Fig. 21-10), 73 percent had abnormal VEPs; more importantly, VEPs were abnormal in 54 percent of the patients without any symptoms or signs referable to the visual system. Such patients are often in the early stages of the disease, when diagnosis is difficult even with modern neuroimaging techniques. Several studies have compared VEPs and magnetic resonance imaging (MRI) of the optic nerve in optic neuritis. 51-53 MRI demonstrated high-signal lesions in 84 to 88 percent of symptomatic patients,52.53 whereas VEPs were abnormal in 100 percent of cases. Miller and colleagues compared the frequency of abnormalities of the optic nerve elicited by MRI and by VEPs in 30 asymptomatic eyes of patients with a recent or past attack of optic neuritis. MRI lesions were present in 20 percent and VEP abnormalities were present in 27 percent of cases. 52 They concluded that "visual evoked potentials were even more sensitive than MRI in detecting lesions and are still the investigation of choice in suspected demyelinating diseases involving the optic nerve." The incidence of reported VEP abnormalities in studies of patients with multiple sclerosis varies from 75 to 97 percent. 54-56 Although there is a consensus that VEPs are very sensitive in detecting clinically silent lesions of the optic nerve, two issues pertaining to the

Percent

o

20

60

40

80

100 FIGURE 11-10 • Results of visual

evoked potential studies with pattern-reversing 31' checks in 180 patients with multiple sclerosis. With indicates patients with visual dysfunction either at the time of testing or in the past medical history. Without indicates patients without visual complaints and with normal findings on neuro-ophthalmologic examination. Total indicates the sum of both groups of patients.

With

Without

Total

•

Abnormal

o

Normal

Visual Evoked Potentials inClinical Neurology

diagnosis of multiple sclerosis are still unresolved: (l) the problem of incorrect diagnosis; and (2) the issue of VEP specificity. No specific laboratory tests can confirm the diagnosis of multiple sclerosis, and the disorder may be missed or diagnosed incorrectly. Rudick and associates identified five features that should suggest the possibility of an alternative diagnosis in patients with suspected multiple sclerosis.57 Among these features was the absence of ocular (including VEP) abnormalities. No adequate study has defined the frequency with which the diagnosis of multiple sclerosis is missed or made erroneously. Some authors suggest that an abnormal VEP in a clinical setting suggestive of multiple sclerosis is sufficient to classify a patient as definitely having that disorder, but a prolonged PlOD latency may lead to the error of overlooking another diagnosis. Tartaglione and co-workers compared 38 patients with multiple sclerosis and 52 patients with neurologic disorders usually not associated with optic pathway involvement and concluded that there was a 30 percent chance of error in attributing a delayed PlOD to multiple sclerosis.P A delayed PlOD or absent VEPs have been reported in many other disorders involving the optic nerve (e.g., systemic lupus erythematosus.t" sarcoidosis.v" vitamin B I2 deficiency,6l,62 neurosyphilis." tropical spastic paraparesis.P' and spinocerebellar ataxia65,66). The initial reports that temporal dispersion of the YEP (defined as the difference in latency between N70 and N135) was characteristic of optic nerve compression'V? were not confirmed. Bodis-Wollner and Onofrj" and the author's group'" have shown that temporal dispersion can be found in both compression and demyelination of the optic nerve. Abnormal VEPs also occur in ischemic optic neuropathy and in toxic optic neuropathy.W? YEPs are useful in determining the diagnosis in patients presenting with transverse myelitis. They can determine whether the disorder is limited to the spinal cord or is multifocal, thus suggesting that the process is probably a form of multiple sclerosis. The Transverse Myelitis Consortium Group of the American Academy of Neurology?" stated: "Brain MRI with gadolinium and visual evoked potentials will determine if there is demyelination elsewhere in the neuraxis, therefore defining the process as multifocal." VEPs have been retained as the only useful sensory evoked potentials in the revised diagnostic criteria for multiple sclerosis." McDonald and colleagues pointed out that VEPs are not necessary in the diagnosis if the patient has had two or more attacks and has evidence of multiple lesions on magnetic resonance imaging." Although a delayed PlOD is not specific for any disease, and any pathologic process involving the optic nerve can produce similar physiologic disturbances,

461

the clinical context in which an abnormal YEP is obtained will have important diagnostic implications. Electrophysiologic studies should also take advantage of the parallel processing of vision by using a variety of visual stimuli. Pathologic processes may interfere with the normal functioning of only some of the parallel channels. A more comprehensive use of VEPs elicited by different stimuli may therefore assess more effectively any visual deficits and at the same time increase the diagnostic yield. Regan and colleagues have identified a group of patients with multiple sclerosis who, despite normal visual acuity, had difficulty in detecting objects of medium size under certain conditions. 72-74 They showed that these patients had a loss of contrast sensitivity limited to midspatial-frequency stimuli. It has also been shown that in patients with multiple sclerosis or optic neuritis, VEPs may be abnormal only to a specific pattern orientation (horizontal versus vertical);" again suggesting that deficits may be limited to selected processing channels. Porciatti and Sartucci compared VEPs to chromatic and luminance stimuli in patients with multiple sclerosis and in patients with unilateral optic neuritis.F" The amplitude losses and latency delays were greater for chromatic than luminance VEPs. This dissociation between luminance and color suggests that the parvocellular pathway is more impaired than is the magnocellular pathway in optic neuritis. Although it is evident that testing the patient with a variety of visual stimuli (e.g., different size patterns or different orientation) increases the number of abnormalities detected, pragmatically it is necessary to "limit the testing sequence to a tolerable duration for the patient. "22 Among the various techniques devised to speed the gathering of data has been the utilization of steady-state VEPs. Steady-state VEPs are responses to visual stimuli of relatively high frequency (usually greater than 3.5 Hz). These responses resemble quasi-sinusoidal oscillations and therefore can be analyzed by fast Fourier analysis. 1,2,76 The response at the frequency of the stimulus is the fundamental or first harmonic, whereas the response component at twice the stimulus frequency is called the second harmonic (2F). Fourier analysis provides the amplitude and phase of each harmonic response. Phase can be used as a measurement similar to latency for transient VEPS.77,78 The author has used steady-state evoked potentials analyzed by fast Fourier transform both in normal subjects?" and in patients with optic neuropathy or multiple sclerosis."? Visual stimuli were computer generated and interfaced with the recording system. The recorded visual spectrum array is shown in Figure 21-11. It took less than 3 minutes to characterize the spatial frequency function to five spatial frequencies. Phase abnormalities, defined as a phase lag outside the mean phase ± 2 circular

462

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

2F

~ 6.0

Ci Q) "0

3.0

:3>-

()

cQ) ::l

C"

2.0

~

]

(ij

~ c, Cf)

1.5

2 J.1V 1.0

Frequency (6.1 Hz/div) FIGURE 21·11 • Amplitude spectra of steady-state visual

evoked potentials at various spatial frequencies. The stimulus consisted of vertical gratings reversing at 4 Hz with a mean luminance of 90 cd/m 2• 2F indicates the second harmonic response (at 8 Hz) and 4F the fourth harmonic response (at 16 Hz).

standard deviations, were noted in 18 (75 percent) of 24 eyes examined in patients with multiple sclerosis (Fig. 21-12). The phase abnormalities were of two types: type 1 abnormality consisted of phase lag present at all spatial frequencies tested, whereas type 2 abnormality consisted of phase lag present only at selected spatial frequencies. Type 2 abnormality could be further subdivided depending on whether phase lag was present at the lower, middle, or higher part of the spatial frequency tested. This technology offers promise in the assessmen t of visual dysfunction and can be used to study amplitude/temporal frequency functions and contrast functions. Another method used to speed the recording and assessment of the visual system is the "sweep technique." With this technique, during repetitive stimulation, rather than averaging each single point in a graphic function under study, a running average is implemented to plot the whole graph. A limited number of single sweeps are then summated to estimate the shape of the curve in the graph. 8o-82 The speed of this technique in comparison to conventional averaging is increased by a factor greater than 15.80 Sweep techniques have been used to study contrast sensitivity'S and to rapidly determine the visual acuity of infants. 83.84

Further studies are needed to confirm the reliability of these newer technologies. Conventional VEPs have been correlated with clinical findings to determine whether a relationship exists between VEP abnormalities and clinical progression or improvement. Although some authors have reported that visual acuity improved in parallel with improvement in latency of the VEP in patients with multiple sclerosis,85-87 others have found no relationship between clinical status and VEPs.sS The author has been impressed by the persistence of prolonged VEP latency following optic neuritis, even when vision has returned to normal. To address the possibility that abnormal pattern ERGs and VEPs may have prognostic usefulness, a prospective study was carried out on 20 patients with optic neuritis.t" The patients were monitored for 12 months and a complete neuro-ophthalmologic examination (i.e., funduscopy, Goldmann perimetry, measurement of visual acuity, pupil evaluation, contrast sensitivity testing, and pseudoisochromatic plate evaluation) was performed together with recording of pattern ERGs and VEPs at the onset and 1, 2, 4, 6, and 12 months later. At the onset of optic neuritis, VEPs to 15' checks were abnormal in all patients, whereas VEPs to 31' and 60' checks were abnormal in 89 and 95 percent of cases, respectively. Color vision was also abnormal in all patients; contrast sensitivity was abnormal in 95 percent and visual acuity in 90 percent. Complete recovery of visual function occurred in 65 percent of cases. Different measures of vision returned to normal in different patients, but visual acuity, color vision, and visual

200 A

A

100 A

Q)

en

ttl

s:

x

0

x

0-

x

-100 -200

0

2

3

4

5

6

7

8

9

Cycles per degree

FIGURE 21·12 • Phase/spatial frequency function in a patient

with multiple sclerosis. Triangles indicate responses from the right eye (OD) and crosses indicate responses from the left eye (OS). The two solid lines above and below the zero line indicate the boundary of normality. Note the normal response for the left eye and very delayed responses from the right eye at most, but not all, spatial frequencies tested.

Visual Evoked Potentials inClinical Neurology

fields returned to normal in 80 percent of patients; contrast sensitivity returned to normal in 79 percent. In contrast to the recovery of vision, VEPs remained abnormal in all patients tested (Fig. 21-13) with stimuli of 15' and in 18 of 19 patients tested with checks of 60'. As shown in Figure 21-13, although considerable improvement in VEPs occurred in a subset of patients with absent VEPs during the acute phase of optic neuritis, VEPs usually remained greatly delayed. Once a well-formed VEP can be recorded, N70 and P100 generally remain unchanged in latency (Fig. 21-14) for many years. The author has recorded a delayed P100 (latency of 133 msec with 15' checks, and 230 msec with 31' checks) in a patient with

463

15' checks

P145.5

1 month 20 20 P144

2 months P100 15' checks 200

20 15 P144

(J

150

Q)

en

E c:: .- 100 (;' c::

~

4 months 20 20

50

I

A

•

Onset

'I!fIIJ 12 months

P100 60' checks 200

(J

c::

50

o B

I

I

I

25 msec/div FIGURE 21-14 .. Pattern-reversal visual evoked potentials

(VEPs) from the left eye in a 36-year-old woman with optic neuritis in that eye. During the acute phase no VEPs were recordable (see also Fig. 21-15). At 1, 2, and 4 months the responses remained prolonged, although her vision had returned to normal.

150

Q)

en E .5 100 (;'

~

I

•

Onset

'I!fIIJ 12 months

FIGURE 21-13 .. Histograms of PI00 latencies in the pattern-

reversal visual evoked potentials (VEPs) elicited by (A) 15' and (B) 60' checks in patients with optic neuritis. The dark bars represent PIOO at the onset of the optic neuritis, and the lighter bars indicate P100values 12 months later. Barsclose to the zero line indicate that no response was present. The solid line at around 100 msec indicates the boundary of normality. At the onset of optic neuritis, many patients had absent VEPs. In most of these patients the VEPlater returned, but it usually remained abnormally delayed.

normal vision who had a well-documented attack of optic neuritis 37 years earlier. Thus, the VEP can be used as a marker of a past episode of optic neuritis. The author's study did not confirm previous reports that absence of pattern ERGs in the acute phase of optic neuropathy is a poor prognostic indicator for visual recovery.SO,90 Three patients had absent VEPs and pattern ERGs, yet two recovered normal vision (Fig. 21-15) and one regained nearly normal vision. It was concluded that the absence ofVEPs and pattern ERG in acute optic neuritis is compatible with full visual recovery. Brusa and colleagues monitored VEPs for 2 years following an episode of optic neuritis.f! They demonstrated an improvement in the latencies ofVEPs in the affected eye, but also noted a significant prolongation of VEP latencies in the asymptomatic fellow eye. This study suggests that optic nerve demyelination is a dynamic process and that demyelination/remyelination may be active even when the patient appears clinically stable.

464

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Initial

1 month

20 20

HM

30' checks

842

I

ERG

~_:::::==--

~ P139

VEP~ 15' checks ERG

VEP

~

~.e:.v

843

..c:Oo=

0

~ ~

~

_:;C)'~"V

is detected by either Goldmann perimetry or Octopus or Humphrey computerized perimetry, approximately 20 to 40 percent of the ganglion cells have been destroyed. 92 •93 VEPs may be smaller in amplitude, and PlOO may be delayed (Fig. 21-16) in glaucoma. Bray and co-workers prospectively studied 49 patients with verified ocular hypertension by recording transient VEPs, visual fields (Octopus perimetry), and intraocular pressure.f" They monitored their patients for an average of 3.2 years. Of the 24 patients with VEP abnormalities at the onset of the study, glaucoma later developed in 7, but it developed in none of the 25 patients with normal VEPs. Horn and co-workers suggested that the blue-sensitive pathways are more susceptible to damage by glaucoma than are other pathways." They evaluated with disc photography, and VEPs to blue-on-yellow pattern, 161 patients with perimetric and 118 patients with preperimetric chronic glaucoma. The patients were followed for a mean of 24 months. They found significantly prolonged VEP peak times in both glaucoma groups compared with the control group and concluded that VEPs to blue-on-yellow pattern may be a useful method to monitor progression of the disease.

P121.5 I

,

I

,

,

I

I

I

,

I

I

25 msec/div FIGURE 21-15 • Left-eye findings in the same patient as in

Figure 21-14. The left column represents the state of the patient during the acute phase of optic neuritis, and the right column showsthe findings 1 month later. At the onset, acuity wasreduced to the detection of hand motion (HM), there was a large central scotoma on Goldmann perimetry, and pattern electroretinograms (ERGs) and visual evoked potentials (VEPs) to 15' and 30' checks were absent. One month later, visualacuity was normal at 20/20, visualfield was normal, pattern ERG had returned to normal, but the VEP, although it had returned and was of normal morphology, was very prolonged in latency to both 15' and 30' checks.

Although VEPs have no prognostic significance in optic neuritis, there is evidence that they may have prognostic value in progressive disorders of the visual system (e.g., ocular hypertension). Ocular hypertension, defined as an intraocular pressure greater than 21 mmHg, is a benign condition until it damages the retinal ganglion cells and becomes glaucoma. It is estimated that glaucoma eventually develops in 35 percent of patients with increased intraocular pressure. Glaucoma can definitely be diagnosed when a visual deficit is documented; however, by the time a visual loss

00

P117.5

as

+ 2.5 j.LV

1 .

I

I

I

25 msec/div FIGURE 21-16 • Pattern-reversal visual evoked potentials

(VEPs) elicited by 15' checks in a 60-year-Qld man with glaucoma of the right eye (documented by arcuate scotoma on Goldmann perimetry) and ocular hypertension (30 mmHg) in the left eye. VEPs have normal morphology, but N70 and PlOD are clearly prolonged bilaterally. (Normal values for age are shown in Fig. 21-4.) The VEP from the symptomatic eye is more abnormal than that from the nonsymptomatic eye.

Visual Evoked Potentials inClinical Neurology

These data suggest that VEPs have a role in determining which patients with increased intraocular pressure are at risk of developing glaucoma, and therefore who requires aggressive therapy. Abnormalities of VEPs have been reported in patients with lesions of the optic chiasm, especially when caused by compression from pituitary tumors40,41,69 or by direct invasion of gliomas. 96,97 Most of the VEP abnormalities were observed with full-field stimulation and probably indicated involvement of the optic nerve either by direct compression or by involvement of the vascular supply of the nerve and chiasm. It is still unclear whether the utilization of half-field stimulation increases the yield of abnormalities. 41,69,98,GG The issue in compressive chiasmatic lesions is to determine when there is danger to the visual system requiring neurosurgical intervention. At present the accepted approach is periodically to assess the visual fields with Goldmann perimetry or with computerized Octopus or Humphrey perimetry. No prospective study has determined whether VEPs can be used in patients with chiasmatic and perichiasmatic lesions as a prognostic indicator for the risk of visual impairment.

Retrochiasmatic Disorders The reliability of VEPs in the diagnosis of retrochiasmatic lesions remains questionable. VEPs to full-field pattern stimulation are usually normal in patients with unilateral hemispheric lesions, even in the presence of a dense homonymous hemianopia. There have been many studies ofVEPs elicited by hemifield stimulation, with various claims of diagnostic reliability made. 98-104 Blumhardt and co-workers detected 18 of 25 (72 percent) cases of homonymous field defects.l'" A slightly better rate for the detection of homonymous field defects by hemifield VEP abnormalities has been reported by others. 98,103 Abnormalities in these studies consisted of either absent potentials to stimulation of the affected hemifield or gross distortion of the normal amplitude distribution over the scalp, In the author's laboratory, 50 consecutive patients with verified ischemic infarcts of the occipital or temporoparietal lobes and homonymous field defects were tested. 101,I02 The relative values of confrontation field examination, Goldmann perimetry, VEPs to hemifield pattern stimulation, and VEPs to steadystate flash stimulation were compared.l'" Visual field examination by confrontation at the bedside demonstrated the field defect in 96 percent of cases, and Goldmann perimetry did so in 100 percent of cases, whereas VEPs to hemifield stimulation were abnormal in only 79 percent of cases and steady-state VEPs to flash were abnormal in 67 percent (Fig. 21-17). Clearly, VEPs to either flash or hemifield pattern stim-

465

ulation were not sufficiently sensitive to be useful in clinical practice. It has been suggested that VEP amplitude distribution based on multiple scalp recordings (19 derivations) during hemifield stimulation (also referred to as topographic mapping) may be a better method to evaluate patients with retrochiasmatic lesions.!" The author has studied the scalp distribution of VEPs to hemifield stimulation in 14 normal subjects and compared the configuration of the topographic maps with the equivalent dipole underlying the surface evoked potentials as estimated by dipole source localization methods.l'" The stimuli consisted of pattern reversal with 60' checks presented in a hemifield pattern. Normal subjects showed considerable variability in the scalp distribution of the major positive peak (PI 00), with either paradoxical localization of the response over the ipsilateral occipital region or more conventional contralateral occipital localization (Fig. 21-18). Dipole source localization, however, always localized the source of the PlOO over the correct hemisphere. Amplitude distribution over the scalp may provide incorrect information about the underlying cortical generators and cannot be used reliably for diagnostic purposes. Most of the explanations regarding the generator source of VEPs are oversimplifications presuming sequential activation of one or two cortical areas. In reality, as was discussed at the beginning of this chapter, once visual information reaches area 17 it is distributed in parallel to at least 10 cortical areas involving the temporal and the parietal lobes, 12-21 Furthermore, whereas processing is implemented in extrastriate cortices, area 17 still remains active. It is therefore not surprising that the complexity of visual processing receives little clarification from the analysis ofVEPs. Bilateral retrochiasmatic lesions, if sufficiently extensive, will produce cortical blindness.105-107 However, in spite of the severity of clinical dysfunction, VEPs are often preserved. 107-110 The VEP to pattern was preserved in 5 of 7 cortically blind patients studied in the author's laboratory.I'" including 1 patient with total blindness. The VEP waveforms were normal, but in some patients P100 was delayed. VEPs were also normal in morphology and latency in cases of bilateral hemianopia (Fig. 21-19). There are two possible explanations for the preservation ofVEPs in patients with cortical blindness and bilateral occipital lobe lesions: VEPs are either generated in extrastriate visual cortex or originate in remnants of cortical area 17. VEPs were correlated with regional cerebral blood flow and/or regional glucose metabolism with positron emission tomography or single-photon emission computed tomography in 5 patients.l"? All 4 patients with preserved VEPs had some spared cerebral activity in area 17, whereas the patient without a recordable VEP had

466

ElEGRODIAGNOSIS INCLINICAL NEUROLOGY

()

OS

()

RT~

~

RO~

~

M~

~

LO~

~

LT~

~

RO

5

2.5 - - - - - - - - - - - - - - - - - - - - -

01--.....---........--.-----,.----. 7 LO 227 Ratio=6.90

o,

msec

250 I

LP

22

Ratio=3.70

71013161922

RO LO RP LP FIGURE 21-17 • Visual evoked potentials (VEPs) to pattern-reversal stimulation and amplitude spectra to repetitive flash stimu-

lation in a 57-year-old man with an ischemic infarct of the left occipital lobe and isolated right homonymous hemianopia. Goldmann perimetry delineated the extent of the visual field defect. The lower half of the figure illustrates the VEP to hemifield stimulation. Note the absence of response to stimulation of the affected hemifield. Stimulation of the intact left hemifield yielded responses with a contralateral occipital distribution and prominent negative response (represented by an upward deflection in this illustration). The right half of the figure illustrates amplitude spectra to flashes at increasing frequency from 8 to 22 per second. Each spectrum represents the average of a 4-second epoch of electroencephalography. Note the greatly depressed spectra over the left occipital and parietal regions. OS, left eye; RT, right temporal; RO, right occipital; MO, midoccipital; LO, left occipital; LT, left temporal; RP, right parietal; LP, left parietal. The amplitude ratio between the right and left occipital and parietal regions is shown in the graphs. Equal amplitude would yield a number close to 1. (From Celesia GG, Meredith JT, Pluff K: Perimetry, visual evoked potentials and visual spectrum array in homonymous hemianopia. Electroencephalogr Clin Neurophysiol, 56:16,1983, with permission.)

no activity in either occipital lobe. It was concluded that VEPs in cortically blind subjects originate in remnants of area 17. These studies further emphasize that there is a dissociation between the pattern VEP and visual function. Consequently, it is unwise to compare visual perception with VEPs. VEPs are electrical phenomena to a visual stimulus, and although they test some of the function of the visual pathways, they should not be confused with the conscious perception of visual stimuli.

MULTI FOCAL VEPs A new technique has shown promise in the study of the visual function: the multifocal VEP.lJI.Jl2 With this technique, VEPs can be recorded simultaneously from many regions of the visual field. The patient views a display containing about 60 sectors, each with a checkerboard pattern. To assess local defects in the visual field, the multifocal VEP responses must be compared with

Visual Evoked Potentials inClinical Neurology

467

FIGURE 21·18 • Topographic distribution of the major positive wave (P90 and P102, respectively) in the visual evoked potential of two normal subjects. Stimuli consisted of 60' checks reversing at 1 Hz. The left hemifield was stimulated in subject 91-7 and the right hemifield in subject 91-4. Recordings were from 20 scalp electrodes. A, Sample ofVEPs from five scalp derivations. B, The topographic map obtained by using a Hjorth derivation. C, The dipole of the plOO for each subject. Note that in case 91-7 both the topographic mapping and the dipole correctly localize PIOOin the right occipital lobe. However, in case 91-4 the topographic mapping suggests paradoxical localization in the right occipital lobe, whereas the dipole is correctly placed in the left hemisphere with the vector toward the right hemisphere. These complex interactions make interpretation of hemifield stimulation very difficult.

normal controls. These comparisons require relatively sophisticated analyses and software. The multifocal VEP has been used for excluding nonorganic visual loss, diagnosing and following patients with optic neuritis/multiple sclerosis, and following disease progression. It can be combined with multifocal ERG, with the hope of differentiating diseases of the macula and retina from diseases of the ganglion cells and optic nerve. The technique is still in its infancy and there are as yet no studies to determine its reliability and sensitivity compared with other methods of investigation. However, it remains a promising methodology.

CONCLUDING COMMENTS In conclusion, VEPs to patterned stimuli are a reliable and sensitive test in the evaluation of macular and optic nerve function but are of questionable value in the assessment of retrochiasmatic disorders. Future work will require the utilization of variable and complex visual stimuli to test selectively the function of the different parallel channels involved in visual processing. The technique of VEPs is a promising methodology that may permit processing of the evoked biologic signals in a wider and yet still reliable way.

468

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

OD~

OS 20 20

20

RO

P115.7 MO

P113.6

~]+

6J.1V

LO

~-~ 256 msec

FIGURE 21-11 • A 61-year-old man with bilateral hemianopia. Note the normal visual acuity and normal visual evoked potential

(YEP) to 30' checks. RO, MO, and LO refer to right, middle, and left occipital electrode placements. Replication ofYEPs was carried out with stimulation of only the right eye (OD). On the right half of the illustration are drawings of the patient's computed tomographic scan. The right side of the brain is shown on the right side of the template. On the upper left of the illustration are the Goldmann visual fields. (From Celesia GG, Bushnell D, Cone Toleikis S et al: Cortical blindness and residual vision: is the "second" visual system in humans capable of more than rudimentary visual perception? Neurology, 41:862, 1991, with permission.)

REFERENCES 1. Regan D: Human Brain Electrophysiology. Evoked Potentials and Evoked Magnetic Fields in Science and Medicine. Elsevier, New York, 1989 2. Celesia GG: Evoked potential techniques in the evaluation of visual function.J Clin Neurophysiol, 1:55, 1984 3. Celesia GG, Tobimatsu S: Electroretinograms to flash and to patterned visual stimuli in retinal and optic nerve disorders. p. 45. In DesmedtJE (ed): Visual Evoked Potentials. Elsevier, Amsterdam, 1990 4. Churchland PS, Sejnowski TJ: Perspectives on cognitive neuroscience. Science, 242:741, 1988 5. Gouras P: Parallel processing of color-contrast detectors in the visual cortex. p. 242. In Rose D, Dobson WG (eds): Models of the Visual Cortex. John Wiley & Sons, Chichester, 1985 6. Cleland BG, Levik WR, Sanderson KJ: Properties of sustained and transient ganglion cells in the cat retina. J Physiol, 228:649, 1973

7. Kaplan E, Lee BB, Shapley RM: New views of primate retinal function. p. 273. In Osborne NN, Chader GJ (eds): Progress in Retinal Research. Vol 9. Pergamon Press, Oxford, 1990 8. Ballard DH: Cortical connections and parallel processing: structure and function. p. 563. In Arbib MA, Hanson AR (eds): Vision, Brain, and Cooperative Computation. MIT Press, Cambridge, MA, 1987 9. Lennie P, Trevarthen C, Van Essen D et al: Parallel processing of visual information. p. 103. In Spillmann L, Werner JS (eds): Visual Perception. The Neurophysiological Foundations. Academic Press, San Diego, 1990 10. Gouras P, Evers HU: The neurocircuitry of primate retina. p. 233. In Gallego A, Gouras P (eds): Neurocircuitry of the Retina. A Cajal Memorial. Elsevier, New York, 1985 11. Martin KAC: From enzymes to visual perception: a bridge too far? Trends Neurosci, 11:380, 1988 12. DeYoe EA, Carman GJ, Bandettini P et al: Mapping striate and extrastriate visual areas in human cerebral cortex. Proc Nat! Acad Sci USA, 93:2382, 1996

Visual Evoked Potentials in Clinical Neurology

13. Tootell RBH, Mendola JD, Hadjikhani NK et al: Functional analysis of V3A and related areas in human visual cortex.J Neurosci, 18:7060, 1997 14. DeYoe EA,Van Essen DC: Concurrent processing streams in monkey visual cortex. Trends Neurosci, 11:219, 1988 15. Sereno MI, Dale AM, Reppas JB et al: Borders of multiple visual areas in humans revealed by functional magnetic resonance imaging. Science, 268:889, 1995 16. Van Essen DC, Drury HA: Structural and functional analyses of human cerebral cortex using a surface based atlas. J Neurosci, 18:7079, 1997 17. Hubel DH, Livingstone MS: Segregation of form, color, and stereopsis in primate area 18. J Neurosci, 7:3378, 1987 18. Schiller PH, Logothetis NK: The color-opponent and broad-band channels of the primate visual system. Trends Neurosci, 13:392, 1990 19. Marr D: Vision. A Computational Investigation into the Human Representation and Processing of Visual Information. Freeman, San Francisco, 1982 20. Barlow HB: Why have multiple cortical areas? Vision Res, 26:81, 1986 21. Kaas JH: Why does the brain have so many visual areas? .J Cogn Neurosci, 1:124, 1989 22. Celesia GG, Brigell MG: Pattern visual stimulation in prechiasmatic lesions. p. 76. In DesmedtJE (ed): Visual Evoked Potentials. Elsevier, Amsterdam, 1990 23. Spekreijse H, Apkarian P: The use of a system analysis approach to electrodiagnostic (ERG and VEP) assessment. Vision Res, 26:195, 1986 24. Porciatti V, Sartucci F: Retinal and cortical evoked responses to chromatic contrast stimuli. Specific losses in both eyes of patients with multiple sclerosis and unilateral optic neuritis. Brain, 119:723, 1996 25. Bodis-Wollner I, Brannan JR, Chilardi MF et al: The importance of physiology to visual evoked potentials. p. 1. In Desmedt JE (ed): Visual Evoked Potentials. Elsevier, Amsterdam, 1990 26. Brigell M, Bach M, Barber C et al: Guidelines for calibration of stimulus and recording parameters used in clinical electrophysiology of vision. Doc Ophthalmol, 107:185,2003 27. Tobimatsu S, Celesia GG, Cone SB: Effects of pupil diameter and luminance changes on pattern electroretinograms and visual evoked potentials. Clin Vis Sci, 2:293, 1988 28. Kuroiwa Y, Celesia GG: Visual Evoked Potentials: Its Clinical Applications. Nishimura, Niigata, Japan, 1989 29. Kurita-Tashima S, Tobimatsu S, Nakayama-Hiromatsu M et al: Effect of check size on the pattern reversal visual evoked potential. Electroencephalogr Clin Neurophysiol, 80:161, 1991 30. Celesia GG, Kaufman D, Cone S: Effects of age and sex on pattern electroretinograms and visual evoked potentials. Electroencephalogr Clin Neurophysiol, 68:161, 1987 31. Dolman CL, McCormick AQ. Drance SM: Aging of the optic nerve. Arch Ophthalmol, 98:2053, 1980 32. Balazsi AG, Rootman, J, Drance SM et al: The effect of age on the nerve fiber population of the human optic nerve. AmJ Ophthalmol, 97:760,1984

469

33. Allison T, Wood CC, Goff WR: Brain stem auditory, pattern-reversal visual, and short-latency somatosensory evoked potentials: latencies in relation to age, sex, and brain and body size. Electroencephalogr Clin Neurophysiol, 55:919, 1983 34. La Marche JA, Dobson WR, Cohn NB et al: Amplitudes ofvisually evoked potentials to patterned stimuli: age and sex comparisons. Electroencephalogr Clin Neurophysiol, 65:81, 1986 35. Collins DWK, Carroll WM, BlackJL et al: Effect ofrefractive error on the visual evoked response. BMJ, 1:231, 1979 36. Van Lith GHM: The application of visually evoked cerebral potentials in ophthalmological diagnosis. Clin Neurol Neurosurg, 82:85, 1980 37. Thorn F, Schwartz F: Effects of dioptric blur on Snellen and grating acuity. Optom Vis Sci, 67:3, 1990 38. Celesia GG, Kaufman D: Pattern ERGs and visual evoked potentials in maculopathies and optic nerve diseases. Invest Ophthalmol Vis Sci, 26:726, 1985 39. Celesia GG, Kaufman D, Cone SB: Simultaneous recording of pattern electroretinography and visual evoked potentials in multiple sclerosis. A method to separate demyelination from axonal damage to the optic nerve. Arch Neurol, 43:1247,1986 40. Gott PS, Weiss MH, Apuzzo M et al: Checkerboard visual evoked response in evaluation and management of pituitary tumors. Neurosurgery, 5:553, 1979 41. Halliday AM, Halliday E, Kriss A et al: The patternevoked potential in compression of the anterior visual pathways. Brain, 99:357, 1976 42. Halliday AM, McDonald WI, Mushin J: Delayed visual evoked response in optic neuritis. Lancet, 1:982, 1972 43. Kaufman D, Celesia GG: Simultaneous recording of pattern electroretinogram and visual evoked responses in neuro-ophthalmologic disorders. Neurology, 35:644, 1985 44. Arden GB, Hamilton AMP, Wilson-Holt J et al: Pattern electroretinograms become abnormal when background diabetic retinopathy deteriorates to a pre-proliferative stage: possible use as a screening test. Br J Ophthalmol, 70:330, 1986 45. Fish GE, Birch DG, Fuller DG et al: A comparison of visual function tests in eyes with maculopathy. Ophthalmology, 93:1177,1986 46. Lennerstrand G: Delayed visual evoked cortical potentials in retinal disease. Acta OphthalmoI60:497, 1982 47. Thirkill CE, Roth AM, Keltner JL: Cancer-associated retinopathy. Arch Ophthalmol, 105:372, 1987 48. Bach M, Hawlina M, Holder GE et al: Standard for pattern electroretinography. Doc Ophthalmol, 101:11, 2000 49. Celesia GG, Brigell MG: Recommended standards for pattern electroretinograms and visual evoked potentials. Electroencephalogr Clin Neurophysiol Suppl, 52:53, 1999 50. Holder GE: Pattern ERG and an integrated approach to visual pathway diagnosis. Prog Retin Eye Res, 20:531, 2001 51. Ormerod IEC, Miller DH, McDonald WI et al: The role of NMR imaging in the assessment of multiple sclerosis and isolated neurological lesions. Brain, 110:1579, 1988

470

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

52. Miller DH, Newton MR, van der PoeljC et al: Magnetic resonance imaging of the optic nerve in optic neuritis. Neurology, 38:175, 1988 53. Filippini G, Comi GC, Cosi V et al: Sensitivities and predictive values of paraclinical tests for diagnosing multiple sclerosis.] Neurol, 241:132, 1994 54. Matthews WB, Small DG, Small M et al: Pattern reversal evoked visual potential in the diagnosis of multiple sclerosis.] Neurol Neurosurg Psychiatry, 40:1009,1977 55. Becker V\]", Richards 1M: Serial pattern shift visual evoked potentials in multiple sclerosis. Canj Neurol Sci, 11:53,1984 56. Bottcher j, Trojaborg W: Follow-up of patients with suspected multiple sclerosis: a clinical and electrophysiological study.] Neurol Neurosurg Psychiatry, 45:809,1982 57. Rudick RA, Schiffer RB, Schwetz KM et al: Multiple sclerosis: the problem of incorrect diagnosis. Arch Neurol, 43:578, 1986 58. Tartaglione A, Oneto A, Bandini F et al: Electrophysiological detection of "silent" plaques in the optic pathways. Acta Neurol Scand, 76:246, 1987 59. jabs DA, Miller NR, Newman SA et al: Optic neuropathy in systemic lupus erythematosus. Arch Ophthalmol, 104:564, 1986 60. Streletz L], Chambers RA, Bae SH et al: Visual evoked potentials in sarcoidosis. Neurology, 31:1545, 1981 61. Fine E;J, Hallett M: Neurophysiological study of subacute combined degeneration.j Neurol Sci, 45:331,1980 62. Krumholtz A, Weiss HD, Goldstein Pj et al: Evoked responses in vitamin B12 deficiency. Ann Neurol, 9:407, 1981 63. Conrad B, Benecke R, Musers H et al: Visual evoked potentials in neurosyphilis.] Neurol Neurosurg Psychiatry, 46:23,1983 64. Newton M, Cruikshank K, Miller D et al: Antibody to human T-Iymphotropic virus type 1 in West-Indian-born UK residents with spastic paraparesis. Lancet, 1:415, 1987 65. Pedersen L, Trojaborg W: Visual, auditory and somatosensory pathway involvement in hereditary cerebellar ataxia, Friedreich's ataxia and familial spastic paraplegia. Electroencephalogr Clin Neurophysiol, 52:283, 1981 66. Livingstone IR, Mastaglia FL, Edis R et al: Pattern visual evoked responses in hereditary spastic paraplegia. J Neurol Neurosurg Psychiatry, 44:176, 1981 67. Holder GE: Pattern visual evoked potential in patients with posteriorly situated space-occupying lesions. Doc Ophthalmol, 59:121, 1985 68. Bodis-Wollner I, Onofrj M: System diseases and visual evoked potential diagnosis in neurology: changes due to synaptic malfunction. Ann N YAcad Sci, 388:327, 1982 69. Celesia GG, Daly RF: Visual electroencephalographic computer analysis (VECA): a new electrophysiologic test for the diagnosis of optic nerve lesions. Neurology, 27:637,1977 70. Transverse Myelitis Consortium Group: Proposed diagnostic criteria and nosology of acute transverse myelitis. Neurology, 59: 499, 2002. 71. McDonald WI, Compston A, Edan Get al: Recommended diagnostic criteria for multiple sclerosis: guidelines from

the International Panel on the Diagnosis of Multiple Sclerosis. Ann Neurol, 50:121, 2001 72. Regan D, Raymond ], Ginsburg AP et al: Contrast sensitivity, visual acuity and the discrimination of Snellen letters in multiple sclerosis. Brain, 104:333, 1981 73. Regan D: Visual information channeling in normal and disordered vision. Psychol Rev, 89:407, 1982 74. Neima D, Regan D: Pattern visual evoked potentials and spatial vision in retrobulbar neuritis and multiple sclerosis. Arch Neurol, 41:198, 1984 75. Coupland SG, Kirkham TH: Orientation-specific visual evoked potential deficits in multiple sclerosis. Can j Neurol Sci, 9:331,1982 76. Strasburger H, Scheidler W, Rentschler I: Amplitude and phase characteristics of the steady-state visual evoked potential. Appl Opt, 27:1069, 1988 77. Tobimatsu S, Tashima-Kurita S, Nakayama-Hiromatsu M et al: Clinical relevance of phase of steady-state VEPs to P100 latency of transient VEPs. Electroencephalogr Clin Neurophysiol, 80:89, 1991 78. Tomoda H, Celesia GG, Cone Toleikis S: Effect of spatial frequency on simultaneous recorded steady-state pattern electroretinograms and visual evoked potentials. Electroencephalogr Clin Neurophysiol, 80:81, 1991 79. Celesia GG: Visual evoked potentials. p. 45. In Barber C, Taylor Mj (eds): Evoked Potentials Review. IEPS, Nottingham, 1991 80. Regan D: Rapid objective refraction using evoked brain potentials. Invest Ophthalmol, 12:669, 1973 81. Regan D: Colour coding of pattern responses in man investigated by evoked potential feedback and direct plot techniques. Vision Res, 15:175, 1975 82. Seiple W, Kupersmith M, Nelson ] et al: The assessment of evoked potential contrast thresholds using real-time retrieval. Invest Ophthalmol Vis Sci, 25:627, 1984 83. Tyler CW, Apkarian P, Levi DM et al: Rapid assessment of visual function: an electronic sweep technique for the pattern visual evoked potential. Invest Ophthalmol Vis Sci, 18:703, 1979 84. Norcia AM, Tyler CW: Spatial frequency sweep VEP: visual acuity during the first year of life. Vision Res, 25:1399, 1985 85. Matthews WB, Small DG: Serial recording of visual and somatosensory evoked potentials in multiple sclerosis. J Neurol Sci, 40:11, 1979 86. Diener HC, Scheibler H: Follow-up studies of visual potentials in multiple sclerosis evoked by checkerboard and foveal stimulation. Electroencephalogr Clin Neurophysiol, 49:490, 1980 87. Nuwer MR, PackwoodJW, Myers LW et al: Evoked potentials predict the clinical changes in a multiple sclerosis drug study. Neurology, 37:1754, 1987 88. Aminoff M], Davis SL, Panitch HS: Serial evoked potential studies in patients with definite multiple sclerosis. Arch Neurol, 41:1197, 1984 89. Celesia GG, Kaufman DI, Brigell M et al: Optic neuritis: a prospective study. Neurology, 40:919, 1990 90. Porciatti V, von Berger GP: Pattern electroretinogram and visual evoked potential in optic nerve disease: early diagnosis and prognosis. Doc Ophthalmol Proc Ser, 40:101,1984

Visual Evoked Potentials in Clinical Neurology

91. Brusa A, Jones SJ, Plant GT: Long-term remyelination after optic neuritis. A 2-year visual evoked potential and psychophysical serial study. Brain, 124:468, 2001 92. Quigley HA, Addicks EM, Green WR: Optic nerve damage in human glaucoma. III. Quantitative correlation of nerve fiber loss and visual field defect in glaucoma, ischemic neuropathy, papilledema, and toxic neuropathy. Arch Ophthalmol, 100:135, 1982 93. Quigley HA, Dunkelberger GR, Green WR: Retinal ganglion cell atrophy correlated with automated perimetry in human eyes with glaucoma. Am J Ophthalmol, 107:453, 1989 94. Bray LC, Mitchell KW, HoweJW: Prognostic significance of the pattern visual evoked potentials in ocular hypertension. Br J Ophthalmol, 75:79, 1991 9!i. Horn FK, Jonas JB, Budde WM et al: Monitoring glaucoma progression with visual evoked potentials of the blue-sensitive pathway. Invest Ophthalmol Vis Sci 43:1828, 2002 96. Kupersmith MJ, Siegel 1M, Carr RE et al: Visual evoked potentials in chiasmal gliomas in four adults. Arch Neurol, 38:362, 1981 97. Groswasser Z, Kriss A, Halliday AM et al: Pattern and flash evoked potentials in the assessment and management of optic nerve gliomas. J Neurol Neurosurg Psychiatry, 48:1125, 1985 98. Maitland CG, Aminoff MJ, Kennard C et al: Evoked potentials in the evaluation of visual field defects due to chiasmal or retrochiasmal lesions. Neurology, 32:986, 1982 99. Muller-jensen A, Zschocke S, Dannheim F: VER analysis of the chiasmal syndrome. J Neurol, 225:33, 1981 100. Blumhardt LD, Barrett G, Kriss A et al: The pattern evoked potential in lesions of the posterior visual pathways. Ann N YAcad Sci, 388:264, 1982 101. Kuroiwa Y, Celesia GG: Visual evoked potentials with hemifield pattern stimulation: their use in the diagnosis of retrochiasmatic lesions. Arch Neurol, 38:86, 1981

471

102. Celesia GG, Meredith JT, Pluff K: Perimetry, visual evoked potentials and visual evoked spectrum array in homonymous hemianopsia. Electraencephalogr Clin Neurophysiol, 56:16,1983 103. Haimovic lC, Pedley TA: Hemi-field pattern reversal visual evoked potentials. 11. Lesions of the chiasm and posterior visual pathways. Electraencephalogr Clin Neurophysiol, 54:121,1982 104. Onofrj M, Bodis-Wollner I, Mylin L: Visual evoked potential diagnosis of field defects in patients with chiasmatic and retrochiasmatic lesions.J Neural Neurosurg Psychiatry, 45:294, 1982 105. Onofrj M: Generators of pattern visual evoked potentials in normals and in patients with retrachiasmatic lesions. p. 87. In Desmedt JE (ed): Visual Evoked Potentials. Elsevier, Amsterdam, 1990 106. Brigell M, Rubboli G, Celesia GG: Application of a single equivalent dipole model to the hemifield pattern VEP. Invest Ophthalmol Vis Sci, 32: Suppl, 910, 1991 107. Celesia GG, Bushnell D, Cone Toleikis S et al: Cortical blindness and residual vision: Is the "second" visual system in humans capable of more than rudimen tary visual perception? Neurology, 41:862,1991 108. Bodis-Wollner I, Atkin A, Raab E et al: Visual association cortex and vision in man: patterned evoked potentials in a blind boy. Science, 198:629, 1977 109. Celesia GG, Archer CR, Kuroiwa Yet al: Visual function of the extrageniculo-calcarine system in man. Arch Neurol, 37:704, 1980 110. Spehlmann R, Grass RA, Ho SU et al: Visual evoked potentials and postmortem findings in a case of cortical blindness. Ann Neurol, 2:531,1977 111. Hood DC, Odel JG, Winn BJ: The multifocal visual evoked potential.J Neuro-ophthalmol, 23:279, 2003 112. Balachandran C, Klistorner AI, Graham SL: Effect of stimulus check size on multifocal visual evoked potentials. Doc Ophthalmol, 106:183,2003

CHAPTER

Visual Evoked Potentials in Infants and Children

22

EILEEN E. BIRCH and RAIN G. BOSWORTH

TRANSIENT LUMINANCE (FLASH) VEPs PATTERN ONSET/OFFSET VEPs PATTERN-REVERSAL VEPs NORMAL MATURATION OFVEPs Amplitude and Latency of Transient Luminance (Flash) VEPs Amplitude and Latency of Pattern VEPs Acuity Development CLINICAL USE OFTHE VEP IN INFANTS AND CHILDREN Patient and Technical Issues

Visual evoked potentials (VEPs) are massed electrical signals generated by occipital cortical areas 17, 18, and 19 in response to visual stimulation. VEPs differ from the electroencephalogram (EEG) in that the EEG represents ongoing activity of wide areas of the cortex, whereas the VEP is primarily a specific occipital lobe response triggered by a visual stimulus. Thus, VEPs can be used to assess the integrity or maturational state of the visual pathway in infants and preverbal children. The basic methodology for recording VEPs is straightforward and is described in the standard developed by the International Society for Clinical Electrophysiology of Vision (ISCEV). lOne or more active electrodes are placed over the occipital lobes. A reference electrode and a ground electrode are also placed on the scalp. Signals are led from the electrodes through a preamplifier that is located near the infant to boost signal amplitude before further contamination by outside noise sources. The preamplifier also typically acts as a bandpass filter (e.g., between 1 and 30 Hz) to eliminate most of the 60-Hz noise that may contaminate the VEP at the electrode sites. From the preamplifier the signal is led into a device capable of averaging VEPs (e.g., a dedicated microprocessor or a personal computer equipped with appropriate hardware and software). The main purpose of filtering and signal averaging is to improve the

Disorders of the Optic Nerve, Chiasm, and Tract Cortical Visual Impairment Perinatal Asphyxia Delayed Visual Maturation Antiepileptic Medications Cerebral White Matter Disorders Phenylketonuria Prenatal Substance Exposure Diverse Neurologic Disorders Visual Loss of Unknown Etiology and Malingering

signal-to-noise ratio. The VEP is contaminated by EEG as well as by outside noise sources. However, because the VEP in response to a given pattern is fairly constant in amplitude and latency (or phase), whereas the EEG occurs randomly with respect to the visual stimulus, stimulus averaging will decrease the unwanted contamination by EEG and other noise sources in proportion to the square root of the number ofVEPs averaged. For some stimulus conditions, digital filtering using computer software can further improve the signal-to-noise ratio. Most commercial equipment for recording the VEP is not specifically designed for use with infants and young children and therefore requires several modifications for use in this patient population. Because it is important that the infant be visually alert and attentive to patterned stimuli, it is helpful to have some small toys that can be dangled in front of the video display when collecting pattern VEPs. The majority of the VEP response is generated by the cortical projection of the macular area (i.e., the central 6 to 8 degrees of the visual field)2.3; therefore, the toys must be very small or open in the center so as not to block the stimulus from reaching the macula. Infants cannot be instructed to attend to the pattern, and an observer must therefore watch the infant and signal to the averaging equipment when the infant is looking at the pattern and when the

473

474

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

infant is looking elsewhere. Both hand-held buttons and foot pedals can be constructed for this purpose. Sudden head and body movements (e.g., yawns, crying, or vigorous sucking) can produce large-amplitude broadband electrical artifacts. Accordingly, the averaging or analysis equipment often contains hardware or software to reject these artifacts because averaging alone is not sufficient to eliminate their contribution to the recorded response. Alternatively, trials containing artifacts can be deleted manually from the signal average. Minimum standards for reporting VEPs have been published. 1 At least two averages should be obtained to demonstrate reproducibility of waveforms. Traces of VEP recordings should have a clear indication of polarity, time in milliseconds, and amplitude in microvolts. The ISCEV recommends that VEP traces be presented as positive upwards, although in many clinical neurophysiology laboratories the traces are presented with a downward deflection representing positivity at the active electrode. Normative values and normal tolerance limits for amplitude and latency should be reported along with the results. Because VEP latency, amplitude, and waveform change with age, comparison to same-age normative values is useful in clinical practice. VEPs are elicited by a temporal change in visual stimulation. Three types of visual stimuli are commonly used to elicit VEPs: (1) luminance (light) flashes; (2) pattern onset/offset; and (3) pattern contrast reversal. Each of these stimuli may be presented as a single event or in a repetitive manner. When the light flash, pattern onset/offset, or contrast reversal occurs infrequently (i.e., at 1 Hz or less), the entire VEP waveform can be seen; this is called a transient response. Peak-to-peak amplitudes and latencies of several major peaks can be measured (Fig. 22-1). When the light flashes or when pattern reversals are repeated frequently at regular intervals (i.e., at 10 Hz or higher), a simpler periodic waveform is seen; this is called a steady-state response. Response amplitude and phase are measured.

TRANSIENT LUMINANCE (FLASH) VEPs Transient luminance VEPs are obtained in response to a strobe light or a flashing light-emitting diode (LED) display or goggles. The flash should subtend at least 20 degrees and should have a stimulus strength of 1.5 to 3 cd/m 2 . 1 The transient luminance flash VEP is a complex waveform with multiple negative and positive changes in voltage. Various approaches have been taken to naming the peaks and troughs in this response, which has led to some difficulty in interpreting differences among studies. Recently, recommendations for standardized reporting ofVEP data have been made by ISCEV.1 As shown in Figure 22-1, peaks are

10 IJV

o

-10

o

A

100

200

300 msec

200

300 msec

200

300 msec

C3

C2

-10

o

100 P100

o... __r ~

-10

c

o

100

FIGURE ]2·1 • Visual evoked potentials. A, Normal transient flash. B, Transient pattern onset/offset. C, Transient patternreversal. The standardization for naming of positive and negative peaks proposed by the International Society for Clinical Electrophysiology of Vision is indicated.

designated as positive and negative in numerical sequence (PI, P2, P3 and NI, N2, N3). The most commonly reported amplitude is the N2-P2 peak-to-peak amplitude. Using both source localization techniques in humans and intracortical recording in primates, several recent studies suggest that the transient luminance flash VEP primarily reflects the activity of striate and extrastriate cortex.r" In addition, some wavelets in the VEP appear to be subcortical in origin.5- 7 These subcortical wavelets are not major components of the flash VEPs obtained from patients with healthy striate and

Visual Evoked Potentials in Infants and Children

extrastriate cortical areas. However, flash YEPs have been recorded in newborns who lack functional striate and extrastriate cortex.s" These VEPs may have reflected subcortical function that was more apparent in the YEP when the cortical components of the response were missing. Thus, the presence of a flash YEP response cannot be taken as unequivocal evidence of cortical function.

PAnERN ONSET/OFFSET VEPs Pattern onset/offset YEPs are obtained in response to a pattern abruptly exchanged with a diffuse background. The pattern should subtend at least 15 degrees, and the pattern and diffuse fields should be well matched in mean luminance.' In most cases, a minimum of two pattern element sizes should be included, 1 degree and 15 minutes, 1 but normal newborns may not respond to the 15-minute pattern element size. Pattern VEPs primarily reflect activity of the striate and extrastriate cortex. 1O- 13 The pattern onset/offset VEP contains three peaks in adults: a positive peak followed by a negative peak and a second positive peak. As shown in Figure 22-1, ISCEV proposes that these components be termed Cl, C2, and C3. Amplitude is measured from the preceding negative peak. Using MRI in humans, a recent study located the source of the C1 peak in striate cortex."

PAnERN-REVERSAL VEPs The pattern-reversal YEP is less variable in appearance than the pattern onset/offset or transient luminance VEP, In adults, the transient pattern reversal VEP typically contains a small negative peak followed by a large positive peak and a second negative peak. As shown in Figure 22-1, ISCEV proposes that these components be termed N75, PlOD, and N135 to indicate their polarity and their approximate latency (in milliseconds) in normal adults. The most commonly reported amplitude is the N75-PIOO peak-to-peak amplitude. The steady-state pattern-reversal VEP has a relatively simple, almost sinusoidal waveform; amplitude and phase of the response are typically reported. Because the mean luminance of the pattern remains constant throughout the contrast reversals, pattern-reversal YEPs reflect pattern sensitivity rather than light sensitivity. Using intracerebral recording in awake humans, Ducati and co-workers" found that PlOD appears to be generated by the pyramidal cells in layer IV of area 17. However, imaging studies in humans point to the source of the early phase of the PIOO peak as being in dorsal extrastriate cortex of the middle occipital gyrus, whereas the late phase of PlOD appears to be generated by the ventral extrastriate cortex of the fusiform

475

gyrus." There have been no reports of pattern-reversal VEPs being recorded in newborns who lack functional striate and extrastriate cortex. Thus, the presence of a pattern-reversal VEP response may be a good indicator of the integrity of cortical function.

NORMAL MATURATION OF VEPs Visual responses have been documented in preterm infants as young as 24 weeks gestational age (GA), but these responses are rudimentary. Infants tested at 22 to 23 weeks GA had very poor or absent VEPs.14 Considerable visual development occurs during the third trimester of gestation and the first post-term year. Although on funduscopic examination the fovea appears mature soon after term birth, detailed anatomic studies have shown that neither the migration of cone photoreceptors toward the foveal pit nor the movement of ganglion cells away from the foveal pit is complete during the first months of life. 15 Moreover, the fine anatomic structure of the foveal cone photoreceptors, which subserve fine-detail vision, is not mature until at least 4 years of age. 15 Myelination of the optic nerve and tract is incomplete at term birth and continues to increase for 2 years postnatally!" Although the number of cells in the primary visual cortex appears to be complete at birth. considerable increases in cell size, synaptic structure, and dendritic density take place during the first 6 to 8 months oflife.l7· 18 One approach to monitoring the progress of anatomic maturation has been to evaluate developmental changes in the VEPs of healthy alert infants. These data also provide a baseline for assessment of the degree of visual impairment in pediatric patients. Flash VEP amplitudes may be influenced by arousal state, but latency is less affected.P:" The mean latency of N3 across various arousal states (awake, drowsy, active sleep, and quiet sleep) was within 15 msec in preterm infants at 30 to 37 weeks GA, but sleep significantly reduced N3 amplitude.V Similarly, the mean latencies of PI for the alert state and the sleep state were within 15 msec at 40 weeks GA.23 Pattern VEPs are affected by arousal state; good quality pattern YEP recordings require an alert, attentive infant whose eyes are focused on the pattern.

Amplitude and Latency of Transient Luminance (Flash) VEPs Flash YEPs can be recorded from preterm infants as early as 24 weeks GA. In preterm infants less than 30 weeks GA, a single long-latency (about 300 msec) negative peak is the most prominent component of the flash VEP (Fig. 22-2) .14.23-28 This peak has been

476

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

400 31

• Pike et al • Shepherd et al23 • Kraemer et al24 + Tsuneishi et al20 • Harding et al34 21 I) Groseet al " Pryds et al1 21 v Tayloret al1 4 \J Ellingson et al35 • Engel & Fay122 o Ferriss et al1 23

350 U300 Q) en

.s>c c

250

.l!!

~

..lo::

200

ell Q)

c..

~

1aCl Q)

z

150 100 50 0 -20

'!'.r ~

-,

DO

o 0

Q7 0

... ..V"'.... ........ ". •

•••

o

""

20

'\l

40

60

FIGURE 11·1 • Latency of the major negative components of the flash visual evoked potential in preterm and term infants. Age at 0 weeks corresponds to 40 weeks gestational age (term). Two exponential decay fits are plotted; the solid line summarizes the maturational trend for N3 latency. Note that, in the study of Tsuneishi and colleagues.s" the initial negative peak was bifid in many infants and the latency of each component is reported separately. The dotted line summarizes the maturational trend for what is believed to be earlier components Nl or N2 starting at 34 weeks GA.

Adult

Age (weeks)

identified by some authors as Nl because it is the earliest negative peak observed in young infants. However, other authors have suggested that this peak corresponds to the N3 component of the mature waveform and that the Nl and N2 peaks are "missing. "14.29 Some difficulty in establishing correspondence between peaks in infant and adult waveforms occurs because the early negative peak in young infants is often bifid, with two sub-peaks that mayor may not be symmetrical in amplitude. 14.22.24.30 In Figure 22-2, flash YEP latencies from various studies are plotted to show the maturation of the early negative peak, which appears to correspond to the adult's N3 (adult latency, 150 msec). Tsuneishi and colleagues20. 30 assessed changes in the early negative peak longitudinally and found that latency decreased at a rate of 4.6 msec per week between 30 and 40 weeks GA. Pike and colleagues" report a similar rate of 5.5 msec per week between 28 and 42 weeks GA. There is some evidence that the change in latency does not occur smoothly but instead occurs in "spurts" of at least 6 msec,20.30 possibly reflecting the myelination process in the visual pathway. At term (40 weeks GA), earlier negative components are present in the waveform that shorten in latency with age; these appear to correspond to N2 (adult latency, 75 msec) and Nl (adult latency, 40 msec). As seen in Figure 22-2, Pike and associates'" report the earliest negative components, including an NO wave at about 85 msec and Nl at about 155 msec, can be obtained from infants as young as 34 weeks GA, whereas all other studies indicate that these components arise at term (40 weeks GA). Yet, even in the study of Pike and co-

workers." only at term age did all infants simultaneously have both an early NO and Nl. One possible explanation for this finding is that Pike and colleagues" included only infants who later had normal neurologic examinations over the first 2 years of life. The youngest infants reported to show an early positive peak with a latency of approximately 200 msec were between 30 and 35 weeks GA.29.32 This positive peak likely corresponds to the P2 peak in the adult waveform (latency, 100 msec). P2 is consistently present in all normal neonates from 37 weeks GA on, and is present in 90 percent of neurologically normal infants by 35 to 36 weeks GA.33 When present, the latency of the positive component decreases from about 220 msec at 34 weeks GA to 150 to 190 msec at term and to 100 msec by 8 to12 weeks post-term age (48 to 52 weeks GA)20.21.24,34.35 (Fig. 22-3). The P2 peak is clearly identifiable in the YEPs of healthy infants by 6 weeks post-term (46 weeks GA), and its amplitude exceeds that ofN3 by 8 weeks postterm (48 weeks GA).29.32.35 A rapid increase in the amplitudes of most of the flash YEP peaks occurs during early childhood, with the largest amplitudes present at about 6 years of age. Amplitudes of the various peaks reach adult levels by about 16 years of age. The maturational changes in amplitude, however, are age trends that are present despite very large individual differences in amplitude within any given age group. This variability limits the utility of flash YEP amplitude in detecting developmental visual abnormalities. Clinical evaluation has largely focused on the latency of the negative component and appearance of the first positive peak (which

Visual Evoked Potentials inInfants andChildren

477

250 • Shepherd et al23 ~ Pike et al31 • Kraemer et al24 • Tsuneishi et al20 6. Harding et al34 o Grose et al21 o Taylor et aj14 'V Ellingson et aj35 [] Ferriss at al123

200 FIGURE

]2.] • Latency of the major positive component of the transient flash visual evoked potential in preterm and term infants. Age at 0 weeks corresponds to 40 weeks gestational age (term). An exponential decay fit summarizes the mean maturational trend.

o Ellingson32

0+----r---...,..---,.----,---r--'"'7'P<---------r-20 o 20 40 60 80 Adult Age (weeks)

corresponds to P2 in adults). Prolonged latency of the initial negative peak (greater than 370 msec) and the absence of the positive peak at 40 weeks GA or older have both been linked to poor outcomes in preterm infants." Steady-state flash YEPs have been used to investigate the maturation of temporal resolution during infancy.'" The data suggest three phases in maturation. During the first month of life (40 to 44 weeks GA), temporal resolution shows little maturation, improving from 15 Hz to 19 Hz. During the next 5 months, temporal resolution matures rapidly to 45 Hz, and then undergoes further maturation slowly, approaching the adult level (55 Hz) by 9 months of age.

Amplitude and Latency of Pattern VEPs With large pattern-element sizes, transient pattern-reversal VEPs can be recorded from preterm infants as early as 30 weeks GA.21.34 At this age the pattern-reversal YEP to a large-element checkerboard (each check subtending about 2 degrees of visual angle) has a simple waveform consisting of a single positive peak with a latency of approximately 330 msec. As shown in Figure 22-4, latency for large-element patterns decreases to about 250 msec by 40 weeks GA (at term), 150 msec by 10 weeks post-term (50 weeks GA), and 110 msec by 25 weeks postterm (65 weeks GA).21.34,37 For small pattern-element sizes (less than 15 minutes), transient pattern-reversal YEPs may not be recordable up to 50 to 54 weeks GA.

Beyond 4 weeks post-term (44 weeks GA), the YEP waveform changes from a simple one to a more complex waveform with multiple peaks, and peak latency grows progressively shorter (Fig. 22-5) .38 Latency is dependent on pattern element size, with latency to large-element patterns decreasing rapidly over the first months of life and latency to small-element patterns decreasing more gradually (see Fig. 22-4 and Fig. 22-5) .37-41 Similar changes in latency have been reported for transient pattern onset/offset YEPS.42 Steady-state pattern YEPs can be recorded from preterrn infants as early as 35 weeks GA when large pattern-element sizes and relatively slow pattern alternation rates are used.P As infants mature, responses can be recorded to progressively smaller pattern-element sizes and progressively faster pattern alternation rates. 43.44

Acuity Development The pattern YEP has been used to estimate visual acuity in infants for 25 years.t" Just as a standard eye chart for measuring visual acuity contains letters ranging from very large to quite small, a pattern YEP acuity test includes a group of checkerboard or striped grating patterns with elements that range from coarse to fine. In both cases, the goal is to determine the finest pattern that the patient's visual system can resolve. When the YEP is used, visual acuity is most often estimated by examining the relationship between amplitude and pattern-element size. Ideally, many pattern-element sizes

478

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

• ., • • o

350

300

0Q) III

"1

[] o 250

t:,.

200

• • •

§.

is' c Q)

~

A

Q)

>

E

III

0

• 150

Crognale et al 42 - 163 min Harding et al 34 - 120 min McCulloch & Skarf 37 - 120 min Malcolm & McCulloch 126 • 120 min Grose et al 21 - 60 min Moskowitz & Sokol 38 - 60, 48 min Malcolm & McCulloch 126·60 min Fiorentini & Trimarchi 124 - 60 min Roy et al 125 - 60 min Malcolm & McCulloch 126 - 24,30 min Moskowitz & Sokol 38 - 12, 15 min McCulloch & Skarf 37 - 7.5 min Malcolm & McCulloch 126 - 12 min Malcolm & McCulloch 126 • 6 min

•••

a. 100

50 -20

a

20

40

60

80

FIGURE 11-4 • Latency of the major positive component of the transient pattern-reversal visual evoked potential in preterm and term infants. 21,34,37,38,42,J24-J26 Age at 0 weeks corresponds to 40 weeks gestational age (term). Studies are presented in inset box along with the pattern element size used (in minutes of arc): large (dark symbols and dotted line), medium (white symbols and thin line) and small (shaded symbols and thick line) checks. The lines are exponential decay fits to the data. All studies used checkerboard patterns except that of Fiorentini and Trimarchi.P' which used black and white stripes.

Adult

Age (weeks)

Small checks

Large checks

~

-r..

1 mo.

.--..--

2mo.

,

3 mo.

,, ,

·

l\T' -s.>: · ~ ,,, ,, ,,, ,

·

~

6 mo.

s mo,

1 yr.

,,

4 yr.

Adult

··,, I

~~O"V :

·

100 msec

FIGURE 22·5 • Pattern-reversal transient visual evoked potentials obtained from visually normal infants, children, and an adult in response to large (60-minute arc) and small (15minute arc) checks. (From Moskowitz A, Sokol S: Developmental changes in the human visual system as reflected by the latency of the pattern reversal YEP. Electroencephalogr Clin Neurophysiol, 56:1, 1983, with permission.)

would be evaluated to pinpoint the exact size at which a response can no longer be recorded. With the limited attention span of infants and young children, typically a limited set of 4 to 6 patterns is presented. Particularly in clinical settings, where there is often little prior information on which to base a choice of the optimal pattern-element sizes, logarithmic spacing between pattern-element sizes in the test series is used (so that, for example, each pattern size is one-half that of the previous pattern in the series). Logarithmic spacing of a wide range of pattern-element sizes despite a small number of steps maximizes the possibility that acuity can be estimated from the data obtained. Linear regression of amplitude on pattern-element size is used to extrapolate the pattern-element size that corresponds to O.O-J.lV amplitude in order to provide the visual acuity estimate. Many pediatric laboratories and clinics have adopted the sweep YEP for acuity testing in recent years.t" Sweep YEP protocols present pattern-element sizes to the infant in rapid succession during a lQ.-second sweep. Using Fourier analytic techniques to extract the YEP responses to each of the brief stimuli (specifically, the amplitude and phase of the harmonics of the stimulation rate), sufficient information can be obtained from the YEP records to estimate visual acuity from only a few brief test trials. This technique has three significant advantages. First, test time is reduced. Second, the infant's behavioral state changes little during the brief recording session. Third, because many more pattern-element sizes are included in the test protocol, linear spacing in the series can be used and a more accurate estimate of acuity can be obtained (Fig. 22-6).

Visual Evoked Potentials in Infants and Children

4

::;2:

3

Q)

"0

:E

2

Q.

E

-c

'0', ..0. ·0

0 180 Ci Q) ~

120

Q) (f)

III

s: a,

60 0 4

7

10

13

16

19

Spatial frequency (c/deg)

In the sweep visual evoked potential protocol, the amplitude vs. linear pattern-element size function may have a nonmonotonic shape with multiple peaks. Linear regression of amplitude on log pattern-element size and extrapolation to 0 ~V is performed on the final descending limb of the function. Sample data from a lO-month-old healthy infant are shown. Filled circles show the amplitude and phase of the visual response to each pattern-element size (spatial frequency). Open circles show the amplitude of noise during the same recording period. The diagonal line (arrow) illustrates the linear regression; acuity is estimated to be 14.5 c/deg, which corresponds to a Snellen equivalent of approximately 20/40. FIGURE 21·6

11/

1.2 1.0

~

0.8

,

Two classic studies of the maturation of visual acuity during infancy using transient pattern YEPs were conducted during the mid-1970s by Sokol'" and by Marg and associates.'? From an initial visual acuity of 1.30 10gMAR (20/400 Snellen equivalent) at 4 weeks of age, visual acuity rapidly matured to nearly adult levels (0.2 to 0.0 10gMAR; 20/30 to 20/20 Snellen equivalent) by 6 to 7 months of age. MAR represents the minimum angle of resolution. Although some controversy remains, it appears that visual acuity matures according to age corrected for preterrn birth (GA) rather than age from birth (Fig. 22-7). In fact, the latency of the transient pattern YEP has been proposed as a method for determining the "neurologic age" of the infant.t" Over the past 10 years, a substantial normative data set has been gathered for monocular and binocular sweep YEP acuity maturation during the first 2 years of life. 45 Overall, sweep YEP acuity improves from about 0.7510gMAR (20/105 Snellen equivalent) at 6 weeks of age to about 0.37 10gMAR (20/45) at 26 weeks of age and 0.25 10gMAR (20/35) by 55 weeks of age (see Fig. 22-7). The acuity improvements parallel postnatal anatomic maturation, including the migration of cone photoreceptors toward the foveal pit, changes in the fine anatomic structure of the foveal cone photoreceptors,15,48 myelination of the optic nerve and tract, 16 increases in cortical cell size, synaptic structure and dendritic density,17,18 and elimination of supernumerary synapses in visual cortex.!? As shown in Figure 22-7, acuity measured by pattern onset-offset YEP matures more rapidly than acuity measured by pattern-reversal YEP, which may be the result of different aspects of pattern vision maturing at different rates.

Sweep VEP Acuity • ()

,

<3

\

o

"

\l

o -

's 0

4>

III

a: 0.6 -c

Birch 45 - Monocular Birch 45 - Binocular Hamer et al. 127 - Monocular Norcia & Tyler46 - Binocular Norcia et al. 128 - Binocular Orel-Bixler & Norcia 129 - Binocular Orel-Blxler & Norcia 129 - Pattern onset/offset Birch 45 - Preterm monocular

g> 0.4 0.2 0.0 -20

0

20

40

60

80

Age (weeks)

100

II Monocular (open symbols) and binocular (filled symbols) sweep visual evoked potential (YEP) acuity as a function of age in weeks.45,46,127- 129 Age at 0 weeks corresponds to 40 weeks gestational age (term) . All researchers used pattern-reversal except Orel-Bixler and Norcia.l'" who used pattern onset!offset (thick line). Exponential decay fits are plotted for monocular (dotted line) and binocular (thin line) sweep YEP acuity. Note that binocular sweep YEP acuity is slightly better (lower log MAR) than monocular sweep YEP acuity after 15 weeks of age. Preterm monocular acuity (shaded diamonds) development matches term infant acuity development when age is adjusted for pre term birth.

FIGURE 21·7

::E ..J

479

Adult

480

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

After 15 weeks of age, mean binocular sweep VEP acuity is better than mean monocular sweep VEP acuity by about 0.03 10gMAR (equivalent to about a half-line on an eye chart). The variability of binocular sweep VEP acuity within an age group (e.g., 17-week-old infants), defined as 95 percent of the normal distribution, is about ±0,23 10gMAR around the mean. Variability is slightly greater for monocular sweep VEP acuity at ±0,29 10gMAR. Within the normative sample, mean interocular acuity differences are small in each age group, averaging less than 0.1 10gMAR (equivalent to about 1 line on an eye chart). However, the range of interocular differences covering 95 percent of the normal distribution varies with age. At 6 weeks, 95 percent of interocular differences fall within ±0.29 10gMAR (equivalent to about 3 lines on an eye chart), whereas by 46 weeks post-term age (86 weeks GA), 95 percent of interocular differences fall within ±0.18 10gMAR (equivalent to about 2 lines on an eye chart). Normative longitudinal measurements of sweep VEP acuity over the first 18 months of life have also been reported for preterm and full-term infants. 49 ,5o These data have direct clinical utility in that they provide a method for determining whether a change in an individual's acuity over time represents a significant change. In other words, such data provide a baseline for the evaluation of progression, recovery, or response to treatment in individual patients. The rate of sweep VEP acuity development (slope) ranged from -0.34 to -0.89 10gMAR/log weeks in a normative sample of 53 healthy term infants tested on 5 occasions during the first 18 months of life. Preliminary studies have suggested that the rate of acuity development during infancy may be more predictive oflong-term visual acuity outcome than is the infant's acuity determined from a single acuity test."!

CLINICAL USE OF THE YEP IN INFANTS AND CHILDREN With few exceptions, the clinical value of VEP testing does not lie in the detection or differential diagnosis of pediatric diseases. 52 Instead, the strength of VEP assessment lies in the quantitative measurement that it provides of the degree of visual impairment. Determination ofVEP visual acuity is of great value in establishing eligibility for educational and social services, for providing parents with a clear picture of their child's abilities, and for defining baseline visual function before progression or intervention. The VEP may be especially useful in conditions where the ophthalmoscopic appearance of the fundus is variable. Especially in very young infants, the optic disc may appear pale or the macular reflex may appear indistinct, so that it may be

difficult to distinguish normal variants from signs of optic nerve or retinal disorders. VEP responses also may provide prognostic information in several pediatric visual disorders. In some limited cases, it is useful to conduct both VEP and electroretinographic (ERG) testing to clarify the differential diagnosis. In particular, tandem VEP and ERG testing may be helpful in establishing whether the visual dysfunction is attributable to a retinal or an optic nerve disorder. One such case is the distinction between optic nerve hypoplasia and Leber's amaurosis. In optic nerve hypoplasia, the ERG is completely normal, whereas the VEP is abnormal or absent. In Leber's amaurosis, by contrast, severely abnormal ERGs are found or, more often, the ERG is absent. In other cases, the VEP may provide an important adjunct to the ophthalmologic examination, In cases of oculomotor apraxia, for example, the classic compensatory head thrusting may not be present during early infancy. Instead, the infant may simply be unable to fixate or track and appear blind. The VEP can document normal acuity in this condition.

Patient and Technical Issues In the projection from the retina to the occipital cortex, the central 10 degrees of the visual field is predominant, with the peripheral visual field more sparsely represented. The electrical signals recorded from the scalp over the occipital lobe even more strongly reflect this central visual field area because the central projection is on the exposed surface of the occipital cortex, whereas the peripheral projection lies deep within the calcarine fissure. This situation enhances the sensitivity ofVEP protocols to disruption of central vision, which subserves the ability to fixate and to perform the many finely detailed visual tasks that make up daily life (e.g., reading or face recognition). However, a child with poor central vision or noncentral fixation may fail to produce a reliable VEP response despite substantial residual visual function. Several patient characteristics limit the ability to obtain VEPs. In general, patients must be alert and quiet, able to look at the center of the stimulus, and able to focus on the stimulus. Sleepy or sedated patients, as well as patients under anesthesia, produce poor or variable VEP results at best and paradoxical and misleading VEP results at worst. Head and body movements (e.g., yawning, crying, and vigorous sucking) can lead to muscle artifacts in the VEP records, which are sometimes so large that the VEP signal cannot be recognized. Patients who are unable to maintain gaze at the center of the stimulus have variable or absent VEP responses even when significant visual function is present. Nystagmus is a particular problem

Visual Evoked Potentials in Infants and Children

481

because central fixation is impaired and the involuntary eye movements themselves may generate electrical artifacts. Failure to accommodate to bring pattern stimuli into focus on the retina or uncorrected refractive error substantially reduces the amplitudes of pattern VEP responses and may lead to nondetectable VEPs in children who see quite well when the stimuli are in proper focus. Patients with neurologic diseases present some additional problems for VEP testing. Gross disorganization of the EEG (e.g., hypsarrhythmia) or critical electrolyte imbalances usually preclude successful VEP testing. 52.53 Accurate electrode placement, which is accomplished by careful measurement relative to skull landmarks, may not be possible in patients with abnormal brain anatomy. Artificial alterations in brain anatomy (e.g., the placement of shunts) may interfere with the ability to record a signal at some sites on the scalp.

dren with neurofibromatosis type 1 is an annual ophthalmologic examination, this method is often unreliable and inaccurate in young children. Magnetic resonance imaging is very sensitive, but cost and the need for sedation preclude its routine use. VEPs have been proposed as a possible alternative or aid in identitying which patients need MRI examinations. Ng and North 58 found 100 percent sensitivity of the VEP for optic gliomas. Specificity of the VEP ranges from 60 to 83 percent.58-62 Multiple sclerosis (MS) in children is rare, but VEPs may be helpful in diagnosis. Abnormal VEPs typical of MS, delayed but well preserved in waveform, have been reported in all children with MS and optic neuritls.f" Among children diagnosed with MS but without symptoms of optic neuritis, changes in the VEP indicating clinically silent lesions of the visual pathway were found in 86 percent.63

Disorders of the Optic Nerve, Chiasm, and Trad

Cortical Visual Impairment

Optic nerve hypoplasia is an ocular malformation diagnosed clinically on the basis of an abnormally small optic nerve head, optic nerve pallor, the "double-ring sign," and tortuous retinal vessels. Functional vision outcome varies widely in this group, and outcome is not always clearly associated with the appearance of the optic nerve head. Weiss and Kelly» have shown that visual outcome in infants with bilateral optic nerve hypoplasia can be predicted from a weighted combination of initial acuity measurement and estimated optic disc diameter, and that the prediction is improved by waiting until 6 months of age when the possible impact of visual immaturity and delayed visual maturation is decreased. Colobomas of the optic nerve are a rare malformation in which the optic disc is excavated. Like optic nerve hypoplasia, visual function outcomes with coloboma are highly variable. Transient pattern-reversal VEPs appear to be more suitable for analyzing vision in these patients than flash VEPs are in predicting functional vision, although flash VEPs are also altered in cases of serious malformation. 55 The carrier state of Leber's hereditary optic neuropathy, a maternally inherited disease associated with mitochondrial point mutations that leads to profound visual impairment, may be detectable via transient pattern VEPS.56 Prolonged PlOO and Nl35 latencies for 15-minute checks and N135 latencies for 60-minute checks have been reported in asymptomatic carriers. VEPs become abnormal immediately after the onset of acut.e visual loss, and may even be abnormal before symptoms appear in some patients.f? Although the current recommendation for screening and monitoring of optic pathway gliomas in chil-

Cortical visual impairment, often referred to as "cortical blindness" before residual visual function was established in many patients.P' is characterized by severe visual impairment, normal pupillary responses, normal fundus, and no nystagmus. Cortical visual impairment rarely occurs in isolation; almost all patients have developmental delays and neurologic abnormalities. In early VEP studies of cortical visual impairment, VEPs ranged from normal to nondetectable, which suggested that VEP testing was of little use in evaluating this group of children. These wide-ranging results were probably caused by differences in stimulation and recording protocols, in eligibility criteria for the patient cohort, and in timing of the test relative to the onset of visual loss. More recent studies document that virtually all children with carefully defined cortical visual impairment have either abnormal flash or pattern VEPs, or both, and that topographic recording can further increase the sensitivity of VEP protocols in this context. 65,66 A recent study showed that VEP acuity could be measured reliably in a cohort of children with moderate to severe cortical visual impairment, and that the acuity test result was correlated with the Huo scale (a clinical scale of visual funcnonr." Nonetheless, two-thirds of infants showed improved acuity with increasing age. The catchup phase was short and highly accelerated in some infants, but for other infants improvement was protracted. 68 ,69 It should be noted that there are also reports of abnormal VEPs in children who recover;" of behaviorally blind children with normal VEPs who show recovery,71·72 and of patients with normal VEPs who remain blind because of damage at higher levels of the visual system beyond area 17. 73.74 VEP testing may also be

482

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

useful in distinguishing gaze disturbances that limit the infant's ability to fix and follow a visual stimulus (e.g., oculomotor apraxia) from cortical visual impairment.59 In infants with shaken baby syndrome, abnormal flash VEPs are common at presentation, even in the absence of retinal hemorrhages'"; however, these abnormalities are not prognostic because many infants with visual impairment have normal VEPs and show visual recovery on a later visit. Flash VEPs are prognostic in cases of acute-onset cortical blindness following surgery, trauma, infectious disease, or hypoxia. 7l>-78 Within this carefully defined subset of patients with cortical visual impairment, tested immediately after the onset of visual impairment, over 90 percent with normal transient flash VEPs recovered normal visual function. Among those with abnormal or absent transient flash VEPs, over 90 percent had longterm visual impairment or blindness. Thus, unlike the diverse outcomes found in the broader context of cortical visual impairment, traditional flash VEPs may provide clinically useful information concerning outcome when recorded immediately after acute onset of cortical visual loss in pediatric patients.

use of pattern VEPs in delayed visual maturation is not diagnostic but as a method to monitor visual acuity development during the first year of life.

Antiepileptic Medications In epileptic patients receiving different antiepileptic drugs, VEPs are used widely for the evaluation of sensory pathways.84-87 It has been suggested that some antiepileptic drugs can significantly affect VEPs,87,88 although other authors did not find any difference in evoked potentials between epileptic patients and controls. 84 The controversy may arise from the study ofvery different cohorts, with some authors examining patients who have just started therapy, but the large majority examining adult patients who have received poly therapy for a long time. Infants with infantile spasms treated with vigabatrin may suffer visual loss because of retinal toxicity. When tested with VEPs, treated infants show reductions in peak contrast sensitivity'" and visual field defects.?" Specialized VEP methods for assessing contrast sensitivity and visual fields in preverbal children may be especially useful in assessing drug toxicity.90

Perinatal Asphyxia Serial measurements of transient flash VEPs during the first week of life may have prognostic value in cases of perinatal asphyxia. McCulloch and colleagues'? reported that neonates with normal flash VEPs or only transient abnormalities during the first week of life had no long-term visual dysfunction. Among neonates who had no recordable flash VEP response during the first week of life, only I achieved normal vision, whereas 3 were blind and 2 had significant visual impairment.

Delayed Visual Maturation In many cases of delayed visual maturation, transient flash VEPs are reduced in amplitude and delayed in timing. 26,80.81 Because these findings are also characteristic of infants with permanent cortical visual impairment, it is not possible to use the flash VEP to distinguish infants who will recover from those who will have long-term visual impairment. Delayed myelination of the optic nerves may underlie the flash VEP abnormalities.F Pattern VEPs are normal in cases of delayed visual maturation and may provide an avenue for establishing prognosis.I" Caution is needed when in terpreting normal pattern VEPs as good prognostic indicators in a child who appears blind, however, because there are rare reports of children with normal pattern VEPs who are blind because of dysfunction of higher visual areas beyond area 17. 73 Perhaps the best

Cerebral White Matter Disorders Cerebral white matter disorders include demyelination, hypomyelination, and dysmyelination of white matter from neurodegenerative, metabolic, inflammatory, or developmental disorders. Flash VEP abnormalities (e.g., increased latencies or generally altered waveforms) are present in many but not all children with cerebral white matter disorders." Very few patients with progressive disease have shown a normal flash VEP. In general, increased latency correlates with progressive disease, and highly abnormal or non-recordable flash VEPs are associated with severe disability and a poor prognosis."

Phenylketonuria Most studies of children and adolescents with phenylketonuria report increased latency of the PIOO wave, either in terms of mean latency relative to age-matched controls or in terms of a higher frequency of abnormalities in patients with phenylketonuria, or both. 92-98 There are some reports of associations between prolonged PlOD latency and recent exposure to high phenylalanine levels, the mean phenylalanine level during the first decade of life, the level of phenylalanine measured on the day of the VEP recording, and the age at discontinuation of dietary therapy for phenylketonuria. However, others have found no association

Visual Evoked Potentials in Infants and Children

with clinical or metabolic measures. Correlations between prolonged latency and severity of cerebral white matter abnormalities on MRI are also controversial. An alternative hypothesis for the pathogenesis of the VEP abnormality is that it may be related to sensitivity of retinal dopaminergic neurons to increased phenylalanine levels.P?

Prenatal Substance Exposure Alterations in the flash or pattern VEPs of children prenatally exposed to moderate levels of alcohol, marijuana, tobacco, or other illicit drugs have been noted.I?" Most striking is a delay in PI latency during the first weeks following birth. Methylmercury can produce Widespread adverse effects on the development and functioning of the human central nervous system, especially in some fishing- communities where infants are exposed prenatally or during early childhood.!" Prolonged PlOO latency has been reported, 102,103 and neuropathologic studies confirm that lesions affect the visual cortex. 104,105 VEP abnormalities have also been associated with prenatal exposure to organic solventsl'" and with chronic exposure to toxic molds!"? and lead. lOS

Diverse Neurologic Disorders Routine behavioral assessment of visual capacity in infants and children with neurologic disorders is often complicated by motor abnormalities that limit their ability to fix and follow. Although the caveats discussed earlier urge caution in the interpretation of VEP responses in such cases, VEPs can nevertheless provide useful information about visual cortical function. Pattern-reversal VEPs are small and irregular in infants with hypsarrythmia. However, with medical therapy to normalize the EEG, good-quality VEPs can be recorded from pediatric patients with infantile spasms, and preliminary evidence suggests that pattern VEPs show abnormalities only with moderate patternelement sizes.l?" Flash VEPs have been used to assess the degree of visual impairment in high-risk neonates. Abnormalities in nash VEPs have been demonstrated in infants with grade III intraventricular hemorrhage but not in infants with milder hemorrhage; the abnormal responses may reflect subcortical rather than cortical visual function.P Abnormal flash VEPs are common in infants with periventricular low-density masses'"; absence of one or more components of the flash VEP or increased latency of the components has been found in all infants with periventricular leukomalacia. Normal flash VEPs were found in 88 percent of very low birth-

483

weight infants with normal computed tomographic scans.!? O'Connor and colleagues reported that sweep VEP acuity at 7 to 12 months, but not at I to 6 months, was predictive of acuity outcome at age 4 to 8 years in a cohort of children with birth weights less than 2000 grams. I 10 Infants with posthemorrhagic hydrocephalus may have delays in latency of the primary positive peak of the flash VEP; these delays increase with increasing intracranial pressure. Jll 113 Some evidence suggests that the changes in latency occur only in hydrocephalic infants with brain damage and not in those who are neurologically normal. 113 Vertical nystagmus and asymmetric nystagmus are commonly associated with serious intracranial pathology and are considered indications for neuroimaging. Visual evoked potentials, together with electroretinography, are also useful tools for evaluating such cases because vertical or asymmetric nystagmus may be associated with hereditary retinal dystrophies, albinism, or spasmus nutans.114.115

Visual Loss of Unknown Etiology and Malingering VEPs can be a useful acljunct in evaluating children with visual loss of unknown etiology, particularly in distinguishing visual impairment from malingering in children who have a normal ophthalmologic examination. ll6.l 17 Children who are familiar with eye charts are unfamiliar with striped patterns on video displays and often provide evidence of normal visual acuity in a VEP protocol. Abnormal VEP responses are not necessarily indicative of visual impairment because patients can willfully degrade VEP responses by defocusing the patterns. 118-120

REFERENCES 1. OdomJV, Bach M, Barber C et al: Visual evoked potentials standard. International Society for Clinical Electrophysiology of Vision. Doc Opthalmol, 108:115, 2004 2. Sakaue H, Katsumi 0, Mehta M et a1: Simultaneous pattern reversal ERG and YER recordings. Effect of stimulus field and central scotoma. Invest Ophthalmo1 Vis Sci, 31:506, 1990 3. Rabin ], Swittkes E, Crognale M et al: Visual evoked potentials in three-dimensional color space: Correlates of spatiochromatic processing. Vision Res, 34:2657, 1994 4. van del' Mare1 EH, Dagnelie G, Spekreijse H: Subdurally recorded pattern and luminance EPs in the alert rhesus monkey. Electroencephalogr Clin Neurophysiol, 57:354, 1984 5. Kraut MA, Arezzo jC, Vaughan HG jr: Intracortical generators of the flash YEP in monkeys. Electroencephalogr Clin Neurophysiol, 62:300, 1985

484

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

6. Ducati A, Fava E, Motti EDF: Neuronal generators of the visual evoked potentials: intracerebral recording in awake humans. Electroencephalogr Clin Neurophysiol, 71:89,1988 7. Schanel-Klitsch E, Siegfried JB, Kavanagh V: Developmental differences in visual evoked potential slow waves and high frequency wavelets across the lifespan: effects ofluminance variation of flash and background. p. 198. In Barber C, Blum T (eds): Evoked Potentials III. Butterworth, Boston, 1987 8. Snyder R, Hata SK, Brann BS et al: Subcortical visual function in the newborn. Pediatr Neurol, 6:333, 1990 9. Dubowitz LMS, Mushin J, DeVries L et al: Visual function in the newborn infant: is it cortically mediated? Lancet, 1:1139, 1986 10. Spekreijse H, Dagnelie G, Maier Jet al: Flicker and movement constituents of the pattern reversal response. Vision Res, 25:1297, 1985 11. Dagnelie G, DeVries M, Maier J et al: Pattern reversal stimuli: motion or contrast? Doc Ophthalmol, 31:343, 1986 12. Schroeder CE, Tenke CE, Givre SJ et al: Striate cortical contribution to the surface-recorded pattern-reversal VEP in the alert monkey. Vision Res, 31:1143, 1991 13. Di Russo F, Martinez A, Sereno MI et al: Cortical sources of the early components of the visual evoked potential. Hum Brain Map, 15:95, 2001 14. Taylor MJ, Menzies R, MacMillan LJ et al: VEPs in normal full-term and premature neonates: longitudinal versus cross-sectional data. Electroencephalogr Clin Neurophysiol, 68:20, 1987 15. Hendrickson A: Morphological development of the primate retina. p. 287. In Simons K (ed): Early Visual Development: Normal and Abnormal. Oxford University Press, New York, 1993 16. Magoon EH, Robb RM: Development of myelin in human optic nerve and tract. A light and electron microscope study. Arch Ophthalmol, 99:655, 1981 17. Huttenlocher PR, de Courten C: The development of synapses in striate cortex of man. Hum Neurobiol, 6:1, 1987 18. Garey LJ: Structural development of the visual system of man. Hum Neurobiol, 3:75,1984 19. Kurtzberg D, Vaughan HG: Electrophysiological assessment of auditory and visual function in the newborn. Clin Perinatol, 12:277, 1985 20. Tsuneishi S, Casaer P, FockJM et al: Establishment of normal values for flash visual evoked potentials (VEPs) in preterrn infants: a longitudinal study with special reference to two components of the Nl wave. Electroencephalogr Clin Neurophysiol, 96:291, 1995 21. Grose J, Harding GFA, Wilton AYet al: The maturation of the pattern reversal VEP and flash ERG in pre-term infants. Clin Vision Sci, 4:239, 1989 22. Whyte HE, Pearce JM, Taylor MJ: Changes in the VEP in preterm neonates with arousal states, as assessed by EEG monitoring. Electroencephalogr Clin Neurophysiol, 68:223, 1987 23. Shepherd AJ, Saunders KJ, McCulloch DL et al: Prognostic value of flash visual evoked potentials in preterm infants. Dev Med Child Neurol, 41:9, 1999

24. Kraemer M, Abrahamsson M, Sjostrom A: The neonatal development of the light flash visual evoked potential. Doc Ophthalmol, 99:21, 1999 25. Ellingson RJ: Variability of visual evoked responses in the human newborn. Electroencephalogr Clin Neurophysiol, 29:10, 1970 26. Fielder AR, Russell-Eggitt IR, Dodd KL et al: Delayed visual maturation. Trans Ophthalmol Soc, 104:653, 1985 27. Mushin J, Hogg CR, Dubowitz LMS et al: Visual evoked responses to light emitting diode (LED) photostimulation in newborn infants. Electroencephalogr Clin Neurophysiol, 58:317, 1984 28. Chin KC, Taylor MJ, Menzies R et al: Development of visual evoked potentials in neonates. A study using light emitting diode goggles. Arch Dis Child, 60:1166, 1985 29. Ellingson RJ: Electroencephalograms of normal full-term infants immediately after birth with observations on arousal and visual evoked responses. Electroencephalogr Clin Neurophysiol, 10:31, 1958 30. Tsuneishi S, Casaer P: Stepwise decrease in VEP latencies and the process of myelination in the human visual pathway. Brain Dev, 19:547, 1997 31. Pike AA, Marlow N, Reber C: Maturation of flash visual evoked potentials in preterrn infants. Early Hum Dev, 54:215, 1999 32. Ellingson RJ: Cortical electrical responses to visual stimulation in the human infant. Electroencephalogr Clin Neurophysiol, 12:663, 1960 33. Placzek M, Mushin J, Dubowitz LMS: Maturation of the visual evoked response and its correlation with visual acuity development in preterrn infants. Dev Med Child Neurol, 27:448, 1985 34. Harding GF, Grose J, Wilton A et al: The pattern reversal VEP in short-gestation infants. Electroencephalogr Clin Neurophysiol, 74:76, 1989 35. Ellingson RJ, Lathrup GH, Danahy T et al: Variability of visual evoked potentials in human infants and adults. Electroencephalogr Clin Neurophysiol, 34:113, 1973 36. Apkarian P: Temporal frequency responsivity shows multiple maturational phases: state-dependent visual evoked potential luminance flicker fusion from birth to 9 months. Vis Neurosci, 10:1007, 1993 37. McCulloch DL, Skarf B: Development of the human visual system: monocular and binocular pattern VEP latency. Invest Ophthalmol Vis Sci, 32:2372, 1991 38. Moskowitz A, Sokol S: Developmental changes in the human visual system as reflected by the latency of the pattern reversal VEP. Electroencephalogr Clin Neurophysiol, 56:1,1983 39. Sokol S: Measurement of infant visual acuity from pattern reversal evoked potentials. Vision Res, 18:33, 1978 40. Porciatti V, Vizzoni L, von Berger GP: Neurological age determination by evoked potentials. p. 345. In Francois J, Maione M (eds): Pediatric Ophthalmology. John Wiley & Sons, New York, 1982 41. McCulloch DL, Orbach H, Skarf B: Maturation of the pattern-reversal VEP in human infants: a theoretical framework. Vision Res, 39:3673, 1999 42. Crognale MA, KellyJP. Chang BS et al: Development of pattern visual evoked potentials: longitudinal measurements in human infants. Optom Vis Sci, 74:808, 1997

Visual Evoked Potentials in Infants and Children

43. Birch EE, Birch DG, Hoffman DR et al: Dietary essential fatty acid supply and visual acuity development. Invest Ophthalmol Vis Sci, 33:3242, 1992 44. Porciatti V: Temporal and spatial properties of the pattern-reversal VEPs in infants below 2 months of age. Hum Neurobiol, 3:97,1984 45. Birch EE: Assessing infant acuity, fusion, and stereopsis with visual evoked potentials. In Heckenlively J (ed): Principles and Practice of Clinical Electrophysiology. Mosby Year Book, St. Louis, in press. 46. Norcia AM, Tyler CW: Spatial frequency sweep VEP: visual acuity during the first year of life. Vision Res, 25:1399,1985 47. Marg E, Freeman DN, Petzman P: Visual acuity development in human infants: evoked potential measurements. Invest Ophthalmol Vis Sci, 15:150, 1976 48. Hendrickson AE: Primate foveal development; a microcosm of current questions in neurobiology. Invest Ophthalmol Vis Sci 35:3129, 1994 49. Salornao SR, Berezovsky A, Cinoto RW et al: Longitudinal visual acuity development in healthy preterm infants. Invest Ophthalmol Vis Sci, 41:A3879, 2000 50. Salornao SR, Birch EE: Individual growth curves for infant visual acuity measured by sweep VEP and FPL in the first 18 months of/ife. Submitted for publication 51. Jeffrey BG, Weems lJ, Salornao SR et al: Prediction of visual acuity outcome following perinatal cortical insult American Association for Pediatric Ophthalmology and Strabismus, 2001 52. Hoyt CS: The clinical usefulness of the visual evoked response.] Pediatr Ophthalmol Strabismus, 21:231, 1984 53. Smith DN: The clinical usefulness of the visual evoked response.J Pediatr Ophthalmol Strabismus 21:235,1984 54. Weiss AH, KellyJP: Acuity, ophthalmoscopy, and visually evoked potentials in the prediction of visual outcome in infants with bilateral optic nerve hypoplasia. J AAPOS, 7:108,2003 55. Tormene AP, Riva C: Electroretinogram and visualevoked potentials in children with optic nerve coloboma. Doc Ophthalmol, 96:347,1999 56. Salomao SR, Carelli V, Sacai PYet al: Pattern-reversal visually evoked potential (VEP) in asymptomatic carriers from an extensive Brazilian pedigree with 11778 Leber's hereditary optic neuropathy (LHON). Invest Ophthalmol Vis Sci, 44:E-Abstract 936, 2003 57. Hung HL, Kao Y, Huang CC: Clinical features of Leber's hereditary optic neuropathy with the 11778 mitochondrial DNA mutation in Taiwanese patients. Chang Gung Med, 26:41, 2003 58. Ng Y, North KN: Visual evoked potentials in the assessment of optic gliomas. Pediatr Neurol, 24:44, 2002 59. North KN, Cochineas C, Tang E et al: Optic gliomas in neurofibromatosis type 1: role of visual evoked potentials. Pediatr Neurol, 10:117, 1994 60. Jabbari B, Maitland CG, Morris LM et al: The value of visual evoked potentials as a screening test in neurofibromatosis. Ann Neurol, 42:1072, 1985 61. Groswasser Z, Kriss A, Halliday AM et al: Pattern and flash-evoked potentials in the assessment and management of optic nerve gliomas. J Neurol Neurosurg Psychiatry, 48:1125,1985

485

62. Lund AM, Skovby F: Optic gliomas in children with neurofibromatosis type 1. Eur J Paediatr, 150:835, 1991 63. Zgorzalewicz M, Michalowska-Wender G, LosyJ et al: The value ofVEP assessment and immunological CSF analysis in the diagnosis of childhood and juvenile MS. Przegl Lek, 60:1, 2003 64. JanJE, Farrel K, Wong PK et al: Eye and head movements of visually impaired children. Dev Med Child Neurol, 28:285, 1986 65. Whiting S, Jan JE, Wong PKH et al: Permanent visual cortical impairment in children. Dev Med Child Neurol, 27:730, 1985 66. Frank Y, Kurtzberg D, Kreuzer JA et al: Flash and pattern-reversal visual evoked potential abnormalities in infants and children with cortical blindness. Dev Med Child Neurol, 34:305, 1992 67. Good WV: Development of a quantitative method to measure vision in children with chronic cortical visual impairment. Trans Am Ophthalmol Soc, 99:253, 2001 68. Clarke MP, Mitchell KW, Gibson M: The prognostic value of flash visual evoked potentials in the assessment of non-ocular visual impairment in infancy. Eye, 11:398, 1997 69. Weiss AH, KellyJP, PhillipsJO: The infant who is visually unresponsive on a cortical basis. Ophthalmology, 108:2076, 2001 70. Kupersmith MJ, NelsonJl: Preserved visual evoked potential in infant cortical blindness. Neuroophthalmology, 6:85,1986 71. Sokol S, Hedges TR, Moskowitz A: Pattern VEPs and preferential looking acuity in infantile traumatic blindness. Clin Vision Sci, 2:59, 1987 72. Regan D, Regal DM, TibblesJAR: Evoked potentials during recovery from blindness recorded serially from an infant and his normally sighted twin. Electroencephalogr Clin Neurophysiol, 54:465, 1982 73. Bodis-Wollner I, Atkin A, Raab E et al: Visual association cortex and vision in man: pattern-evoked occipital potentials in a blind boy. Science, 198:638, 1977 74. Spehlmann R, Gross RA, Ho SU et al: Visual evoked potentials and postmortem findings in a case of cortical blindness. Ann Neurol, 2:531,1977 75. Brennan R, Clarke MP, Redfearn A et al: Visual impairment following shaking/impact injury in infancy: A clinical and electrophysiological study. Invest Ophthalmol Vis Sci, 44:E-Abstract 2717, 2003 76. Taylor MJ, McCulloch DL: Prognostic value of VEPs in young children with acute onset of cortical blindness. Pediatr Neurol, 7:111, 1991 77. Barnet AB, Manson JI, Wilner E: Acute cerebral blindness in childhood: six cases studied clinically and electrophysiologically. Neurology, 20:1147,1970 78. Kupersmith MJ, Nelson JI, Carr RE: The visual evoked potential as a prognosticator in childhood cortical blindness. Ann Neurol, 14:146, 1983 79. McCulloch DL, Taylor M, Whyte HE: Visual evoked potentials and visual prognosis following perinatal asphyxia. Arch Ophthalmol, 109:229, 1991 80. Harel S, Holtzman M, Feinsod M: Delayed visual maturation. Arch Dis Child, 58:298, 1983

486

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

81. Mellor DH, Fielder AR: Dissociated visual development: electrodiagnostic studies in infants that are slow to see. Dev Med Child Neurol, 22:327, 1980 82. Baker RS, Schmeisser ET, Epstein AD: Visual system electrodiagnosis in neurologic disease of childhood. Pedriatr Neurol, 12:99, 1995 83. Lambert SR, Kriss A, Taylor D: Delayed visual maturation: A longitudinal clinical and electrophysiological assessment. Ophthalmology, 96:524, 1989 84. Harding GF, Herrick CE,]eavons PM: A controlled study of the effect of sodium valproate on photosensitive epilepsy and its prognosis. Epilepsia, 19:555, 1987 85. Cosi V, Callieco R, Galimberti CA et al: Effect of vigabatrin on evoked potentials recordings in patients with epilepsy. Br] Clin Pharmacol, 27: Suppll, 61S, 1989 86. Drake ME ]r, Pakalnis A, Padamadan H et al: Effect of anti-epileptic drug monotherapy and polypharmacy on visual and auditory evoked potentials. Electromyogr Clin Neurophysiol, 29:55, 1989 87. Martinovic Z, Ristanovic D, Dokic-Ristanovic D et al: Pattern reversal evoked potentials recorded in children with generalized epilepsy. Clin Electroencephalogr, 21:233, 1990 88. Cosi V,Callieco R, Galimberti C et al: Effect of vigabatrin: on visual, brainstem auditory, and somatosensory evoked potentials in epileptic patients. Eur Neurol, 28:42, 1988 89. Morang SE, Westall CA, Buncic R] et al: Sweep visual evoked potentials in infants with infantile spasms before and during vigabatrin treatment. Invest Ophthalmol Vis Sci, 44:E-Abstract 2720, 2003 90. Harding GF, Robertson K, Spencer EL et al: Vigabatrin: its effect on the electrophysiology of vision. Doc Ophthalmol, 104:213, 2002 91. Kristjansdottir R, Sjostrom A, Uverbrant P: Ophthalmological abnormalities in children with cerebral white matter disorders. Eur j Pediatr Neurol, 6:25, 2002 92. Jones Sj, Turano G, Kriss A et al: Visual evoked potentials in phenylketonuria: association with brain MRl, dietary state, and IQ. j Neurol Neurosurg Psychiatry, 59:260, 1995 93. LeuzziV,Cardona F,Antonozzi I et al: Visual, auditory, and somatosensorial evoked potentials in early and late treated adolescents with phenylketonuria. j Clin Neuraphysiol, 11:602, 1994 94. Ludolph AC, Ullrich K, Nedjat S et al: Neurological outcome in 22 treated adolescents with hyperphenylalaninemia. A clinical and electrophysiological study. Acta Neurol Scand, 85:243, 1992 95. Ludolph AC, Vetter U, Ullrich K; Studies of multimodal evoked potentials in treated phenylketonuria: the pattern of vulnerability. Eur J Pediatr, 155: Suppl I, S64, 1996 96. Cleary MA, Walter jH, Wraith jE et al: Magnetic resonance imaging of the brain in phenylketonuria. Lancet, 344:87, 1994 97. Giovannini M, Valsasina R, Villani R et al: Pattern reversal visual evoked potentials in phenylketonuria.] Inherit Metab Dis, ll:416, 1986 98. Korinthenberg R, Ullrich K, Fullenkemper F: Evoked potentials and electroencephalography in adolescents with phenylketonuria. Neuropediatrics, 19:175, 1988

99. Leuzzi V,Rinaldussi S, Chiarotti F et al: Subclinical visual impairment in phenylketonuria. A neurophysiological study (VEP-P) with clinical, biochemical, and neuroradiological (MRl) correlations. J Inherit Metab Dis, 21:351, 1998 100. Scher MS, Richardson GA, Robles Net al: Effects of prenatal substance exposure: altered maturation of visual evoked potentials. Pediatr Neurol, 18:236, 1998 101. Igata A: Epidemiological and clinical features of Minamata disease. Environ Res, 63:157, 1993 102. Murata K, Weihe P, Renzoni A et al: Delayed evoked potentials in children exposed to methylmercury from seafood. Neurotoxicol Teratol, 21:343, 1999 103. Inayoshi S, Okajima T, Imai H: Electrophysiological studies on Minamata disease. jpn] EEG EMG, 16:48, 1988 104. Choi BH: The effects of methylmercury on the developing brain. Prog Neurobiol, 32:447, 1989 105. Tokuomi H, Uchino M, Imamura S et al: Minimata disease (organic mercury poisoning): neuroradiologic and electrophysiologic studies. Neurology, 32:1369, 1982 106. Till C, Rovet]F, Koren G et al: Assessment of visual functions following prenatal exposure to organic solvents. Neurotoxicology, 24:725, 2003 107. Anyanwu EC, Campbell AW, Vojdani A: Neurophysiological effects of chronic indoor environmental toxic mold exposure on children. Sci World], 3:281, 2003 108. Altmann L, Sveinsson K, Kramer U et al: Visual functions in 6-year-old children in relation to lead and mercury levels. Neurotoxicol Teratol, 20:9, 1998 109. Taddeucci G, Fiorentini A, Pirchio M et al: Pattern reversal evoked potentials in infantile spasms. Hum Neurobiol 3:153,1984 llO. O'Connor AR, Birch EE, Leffler] et al: Infant visual evoked potential (VEP) and preferential-looking (PL) grating acuity as predictors for recognition acuity in low birth weight children. Invest Ophthalmol Vis Sci, 44: E-Abstract 2712, 2003 Ill. Coupland SG, Cochrane DD: Visual evoked potentials, intracranial pressure and ventricular size in hydrocephalus. Doc Ophthalmol, 66:321,1987 112. Desch LW: Longitudinal stability of visual evoked potentials in children and adolescents with hydrocephalus. Dev Med Child Neurol, 43:ll3, 2001 113. Guthkelch AN, Sclabassi R], Hirsch RP et al: Visual evoked potentials in hydrocephalus: relationship to head size, shunting, and mental development. Neurosurgery, 14:283, 1984 114. Shawkat FS, Kriss A, Thompson D et al: Vertical or asymmetric nystagmus need not imply neurological disease. Br J Ophthalmol, 84:175, 2000 115. Shaw FS, Kriss A, Russel-Eggitt I et al: Diagnosing children presenting with asymmetric pendular nystagmus. Dev Med Child Neurol, 43:622, 2001 ll6. Sokol S: The pattern visual evoked potential in the evaluation of pediatric patients. p. ll5. In DesmedtjE (ed): Visual Evoked Potentials. Elsevier, Amsterdam, 1990 ll7. Holder G: Electrodiagnostic testing in malingering and hysteria. p. 573. In Heckenlively]R, Arden GB (eds): Principles and Practice of Clinical Electrophysiology. Mosby Year Book, St. Louis, 1991

Visual Evoked Potentials in Infants and Children

11S. Morgan RK, Nugent B, Harrison]M et al: Voluntary alteration of pattern visual evoked responses. Ophthalmology, 92:1356, 1985 I 19. Chiappa KH, Yiannikas C: Voluntary alteration of visual evoked potentials. Ann Neurol, 12:496, 1982 120. Baumgartner F, Epstein CM: Voluntary alteration of visual evoked potentials. Ann Neurol, 12:475, 1982 121. Pryds 0, Trojaborg W, Carlsen] et al: Determinants of visual-evoked potentials in preterm infants. Early Human Dev 19:117,1989 122. Engel R, Fay W: Visual evoked responses at birth, verbal scores at 3 years, and IQ at 4 years. Dev Med Child NeuroI14:283,1972 123. Ferriss GS, Davis DD, Dorsen MM et al: Changes in latency and form of the photically induced average evoked responses in human infants. Electroencephalogr Clin Neurophysiol, 22:305, 1967 124. Fiorentini A, Trimarchi C: Development of temporal properties of pattern electroretinogram and

125.

126.

127.

128.

129.

487

visual evoked potentials in infants. Vision Res, 32:1609, 1992 Roy M-S, Barsoum-Hornsy M, Orquin] et al: Maturation of binocular pattern visual evoked potentials in normal full-term and preterm infants from 1 to 6 months of age. Pediatr Res, 37:140, 1995 Malcolm CA, McCulloch DL, Shepherd A]: Patternreversal visual evoked potentials in infants: gender differences during early visual maturation. Dev Med Child Neurol, 44:345, 2002 Hamer RD, Norcia AM, Tyler CW et al: The development of monocular and binocular VEP acuity. Vision Res, 29:397, 1989 Norcia AM, Tyler CW, Piecuch R et al: Visual acuity development in normal and abnormal preterm human infants.] Pediatr Ophthalmo1 Strabismus, 24:70,1987. Orel-Bixler DA, Norcia AM: Differential growth of acuity for steady-state pattern-reversal and transient pattern onset-offset VEPs. Clin Vision Sci 2:1, 1987

CHAPTER

Brainstem Auditory Evoked Potentials: Methodology, Interpretation, and Clinical Application

23

ALAN D. LEGATT

OVERVIEW OF AEPs STANDARD BAEP RECORDING TECHNIQUES Stimulation Recording Waveform Identification and Measurement Patient Relaxation and Sedation MODIFICATIONS TO BAEP RECORDING TECHNIQUES Reducing the Stimulus Artifact Improving the Resolution of Specific Components Stimulation at Several Intensities Rapid Stimulation Alternative Auditory Stimuli BAEPs to Electrical Stimuli EFFECTS OF VARYING STIMULUS PARAMETERS Stimulus Polarity Stimulus Intensity Stimulus Rate GENERATORS AND SCALP TOPOGRAPHIES OF BAEPs Wave I Wave IN Wave II Wave liN Wave III Wave IV

Following a transient acoustic stimulus, such as a click or a brief tone pip, the ear and parts of the nervous system generate a series of electrical signals with latencies ranging from milliseconds to hundreds of milliseconds. These auditory evoked potentials (AEPs) are conducted through the body tissues and can be recorded from electrodes placed on the skin to evaluate noninvasively the function of the ear and portions of the nervous system activated by the acoustic stimulation. These short-latency or brainstem auditory evoked potentials (BAEPs) have proven to be valuable tools for hearing assessment, diagnosis of neurologic disorders, and intraoperative monitoring.

Wave Wave Wave Wave

V VI VII VN

CLINICAL INTERPRETATION OF BAEPs Normative Data Delay Versus Absence of Components Significance of Specific BAEP Abnormalities Abnormalities of Wave I Abnormalities of the I-III Interpeak Interval Abnormalities of the III-V Interpeak Interval Abnormalities of the IVIV:I AmplitudeRatio Portion of the Auditory System Assessed by BAEPs Functional Subsets Crossed and Uncrossed Pathways Rostrocaudal Extent Role of Descending Pathways Relationship Between BAEPs and Hearing Categorization of Abnormalities Peripheral Hearing Loss Central AuditoryPathway Abnormalities BAEPs in Specific Neurologic Conditions Acoustic Neuroma Other Posterior Fossa Tumors Cerebrovascular Disease Demyelinating Disease Coma and Brain Death

OVERVIEW OF AEPs Auditory evoked potentials have been divided into short-latency components, with latencies of under 10 msec in adults; long-latency AEPs, with latencies exceeding 50 msec; and middle-latency AEPs, with in termediate latencies (Fig. 23-1). The earliest components derive from electrical processes within the inner ear and action potentials in the auditory nerve. AEP components generated within the brainstem may reflect both action potentials and postsynaptic potentials. Auditory-evoked neural activity becomes increasingly affected by temporal dispersion as the poststimulus latency increases and as the contribution of

489

490

ELEORODIAGNOSIS IN CLINICAL NEUROLOGY

II

VI [ 0.21LV

III IV V

Nb

and analyzing stimulus features (Fig. 23-2). They have therefore been used as probes of cognitive processes, but their variability as well as uncertainly about the precise identity of their cortical generators limit their utility for neurologic diagnosis. Middle-latency AEPs are small; are subject to contamination by myogenic signals; and are rather variable from subject to subject, which also limits their clinical application. Both middleand long-latency AEPs are affected prominently by surgical anesthesia. Short-latency AEPs have achieved the greatest clinical utility because they are relatively easy to record and their waveforms and latencies are highly consistent across normal subjects. They are unaffected by the subject's degree of attention to the stimuli and are almost identical in the waking and sleeping states.' aside from minor differences related to changes in body ternperature.!

( 0.41LV

~---------'---------'" 50

Attend

Ignore

msec

1--------12.5 msecl""--...............................---.........

0.5 sec Short-latency (top), middle-latency (middle), and long-latency (bottom) auditory evoked potentials (AEPs) elicited by monaural click stimuli and recorded in a vertexmastoid channel with vertex positivity plotted as a downward deflection. Note the differing epoch durations of the three AEPs. (From Picton TW, Hillyard SA, Krausz HI et al: Human auditory evoked potentials. 1. Evaluation of components. Electroencephalogr Clin Neurophysiol, 36:179, 1974, with permission.) ' - - _ - ' " - _ - - " -_ _. L - -........-

FIGURE 23-1

........

II

short-duration electrical phenomena (e.g., action potentials) is eliminated. Thus, AEP components that are longer in latency are also wider, and the middleand long-latency AEP components are predominantly generated by postsynaptic potentials within areas of cerebral cortex that are activated by the acoustic stimulus. AEP components are also increasingly affected by the state of the subject and by anesthesia as their latency increases. Long-latency AEPs are profoundly affected by the degree to which the subject is attending to the stimuli

~

L-

50 msect---------

'-0.5 sec "'-

_

FIGURE 23-2 • Effect of attention on auditory evoked poten-

tials. Trains of clicksthat were occasionally lower in intensity were presented, and the subjectwasasked to either ignore the clicks (left) or to attend to them and count the number of softer ones (right). Recording conditions were otherwise the same as Figure 23-1. (From Picton TW, Hillyard SA: Human auditory evoked potentials. II: Effects of attention. Electroencephalogr Clin Neurophysiol, 36:191, 1974, with permission.)

Brainstem Auditory Evoked Potentials

Sedation also produces only minor changes in BAEPs,3 and has been employed during BAEP recordings. However, the use of sedation for evoked potential recordings has been markedly reduced owing to concerns about monitoring and the care of patients during conscious sedation." Because a typical surgical level of anesthesia produces only minor alterations in BAEP S5.6 (Fig. 23-3), they can be used for intraoperative monitoring of the ears and the auditory pathways. The earliest electrical signals produced by the auditory system in response to a transient stimulus, constituting the electrocochleogram (ECochG), were initially recorded by electrodes placed directly in the middle ear. 7 Extratympanic recordings of the ECochG8 yielded smaller signals that required signal averaging to achieve an adequate signal-to-noise ratio. Signal averaging also permitted recording of small time-locked signals originating from other locations. Additional deflections with latencies of several milliseconds after auditory stimulation were first recorded in humans during signal-averaged ECochG studies by Sohmer and Feinrnesser," Jewett and co-workers'P-'! identified the short-latency scalprecorded AEPs as far-field potentials volume-conducted from the brainstem, described the components and their properties, and established the Roman numeral labeling of the peaks that is used in most laboratories (Fig. 23-4). A far-field potential is a potential (voltage) recorded at a sufficiently large distance from its source that small movements of the recording electrode have no significant effect on the waveform. Although short-latency AEPs are commonly called brainstem auditory evoked potentials, this term is not completely accurate because the roster of generators clearly includes the distal (with respect to the brainstem) cochlear nerve and may also include the thalamocortical auditory radiations, neither of which is within the brainstem. Nonetheless, the designation EAEP is used in this chapter because it is the most widely used and

491

Fp1-C3 - - - - - - - - - - C3-01 Fp2-C4 - - - - - - - - - - C4-02 - - - - - - - - - - T3-Cz - - - - - - - - - - Cz-T4 EMG/ECG - - - - - - - . . -_ _~_ _

BAEP

60 dBHl click

+L-...II.--l-.---L_...L----L.--' I I I I I

2

4

6

8

10

12

msec Brainstem auditory evoked potentials recorded during surgery in a patient anesthetized with isoflurane at a concentration sufficient to render the electroencephalogram isoelectric. Although the amplitudes of waves IV, V, VI, and VII were reduced by anesthesia, the latencies of waves I through V were not significantly different from those recorded in this patient in the unanesthetized state. (From Stockard Il- Pope-Stockard JE, Sharbrough FW: Brainstem auditory evoked potentials in neurology: methodology, interpretation, and clinical application. p. 514. In Aminoff MJ [ed]: Electrodiagnosis in Clinical Neurology. 3rd Ed. Churchil\ Livingstone, New York, 1992, with permission.) FIGURE

13-3

II

III

FIGURE 13-4 II Normal brainstem auditory evoked potential recorded between the vertex (Cz) and the right earlobe (A2) following right ear stimulation in a 23-year-old woman. Two averages of 2,000 sweeps each are superimposed. Vertex-positive peaks, shown here as upward deflections, are labeled with Roman numerals according to the convention of Jewett and Williston."! Downgoing peaks after waves I and V are labeled IN and VN, respectively. An electrical stimulus artifact appears at the beginning of the tracings.

O.2.VL IN

VN

1 msec

492

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

understood term. Other synonyms or related designations include auditory brainstem response, far-field electrocochleography, and brainstem audiometry.

STANDARD BAEP RECORDING TECHNIQUES This section describes the standard techniques used to record BAEPs in adult subjects. BAEP recording techniques for infants and children are described in Chapter 24, and intraoperative BAEP monitoring is discussed in Chapter 29. The American Electroencephalographic Society (now the American Clinical Neurophysiology Society) has published guidelines for clinical'! and intraoperativel'' BAEP recordings.

tralateral stimulation from occurring and possibly being misinterpreted as a BAEP arising from stimulation of the ipsilateral ear, the contralateral ear is masked with continuous white noise at an intensity 30 to 40 dB below that of the BAEP stimulus. Acoustic crosstalk also occurs with ear-insert transducers, though the signal reaching the opposite ear is attenuated to an even greater extent, typically 70 to 100 dB.15 If the electrical square pulse causes the diaphragm of the acoustic transducer to move toward the patient's ear, a propagating wave of increased air pressure, termed a compression click (also called a condensation click) is produced. Reversing the polarity of the electrical square pulse that activates the transducer produces a rarefaction click. BAEPs to rarefaction and compression clicks may differ16 ,17 (most prominently in patients with

Stimulation BAEPs are most commonly elicited by brief acoustic click stimuli that are produced by delivering monophasic square pulses of 100-Ilsec duration to headphones or other electromechanical transducers at a rate of about 10 Hz. A rate of exactly 10 Hz or another submultiple of the power line frequency should be avoided; otherwise, the inevitable line frequency artifact will be timelocked to the stimuli and will not be removed by the averaging process. Audiometric headphones having a relatively flat frequency response are desirable so that "broad-band" clicks, whose energy is spread over a wide frequency range, will be produced. The stimulus intensity should be loud enough to elicit a clear BAEP waveform without causing discomfort or ear damage; 60 to 65 dB HL is a typical level. If hearing loss is present, stimulus intensity may be acljusted accordingly so that stimulation is at 60 to 65 dB SL. (It should be noted that dB HL is decibels relative to the threshold of a normal population, dB SL is decibels relative to the threshold of the ear being tested, and dB nHL is decibels relative to the threshold of the specific control population used to establish a laboratory's normative database.) Reduced stimulus intensities are also useful during BAEP recordings, as discussed later in this chapter. The subjective click threshold should be measured in all subjects in whom this is possible, to recognize hearing loss and to determine the stimulus level corresponding to 0 dB SL. Stimuli are delivered monaurally, so that a normal BAEP to stimulation of one ear does not obscure the presence of an abnormal response to stimulation of the other ear. An acoustic stimulus delivered to one ear via headphones can reach the other ear via air and bone conduction with a volume attenuation of 40 to 70 dB14,15 and generate an evoked potential by stimulation of the contralateral ear (Fig. 23-5). To prevent this con-

III

V

II

t A

[ 0.25 !LV

V

t

B

t

c

I

I

I

I

5 msec FIGURE U·5 • Brainstem auditory evoked potentials to monaural stimulation from a patient with one nonfunctioning ear. A, Stimulation of the unaffected ear elicits a normal BAEP. B, When contralateral noise masking is not used, stimulation of the nonfunctioning ear produces a delayed wave V because of air and bone conduction of the stimulus to the other ear, e, When masking noise is delivered to the unaffected ear, stimulation of the nonfunctioning ear does not elicit any reproducible BAEPs. (Modified from Chiappa K, Gladstone Kj, Young RR: Brain stem auditory evoked responses. Studies of waveform variations in 50 normal human subjects. Arch Neurol, 36:81, 1979, with permission.)

Brainstem Auditory Evoked Potentials

a cochlear high-frequency hearing loss), and averaging the responses to the two click polarities together may produce a composite waveform with less diagnostic utility. Therefore, a single stimulus polarity should be used unless alternating polarities are necessary for canceling the electrical stimulus artifact or cochlear microphonic. Rarefaction clicks are generally preferable because they tend to yield BAEPs with better definition of the components; this may be because the initial cochlear movement.s produced by rarefaction clicks are in a direct.ion that depolarizes the hair cells.'? In evaluat.ing a patient's BAEPs, the normative data used should have been acquired with the same stimulus polarity used to test the patient.

Recording Recording electrodes are typically placed at t.he vertex (location Cz of the International 10-20 System) and at both earlobes (the earlobes ipsilateral and contralateral to the stimulated ear are labeled Ai and Ac, respectively). Electrodes at the mastoids (labeled Mi and Mc) may be substituted for the earlobe placements, although wave I tends to be smaller. Such placement, in combination with increased pickup of muscle noise, yields a poorer signal-to-noise ratio for wave I with mastoid leads than with earlobe leads. The ground electrode is often placed on the forehead, but its precise location is not critical. Metal cup or pellet electrodes may be used; needle electrodes should be avoided. Electrode impedance should be less than 5 kohm. Optimally, the same type of electrode should be used at all recording positions, and electrode impedance should be as consistent as possible across all recording electrodes because mismatched electrode impedance can increase the amount of noise in the BAEP data. 18 BAEPs should be recorded between Cz and either Ai or Mi. A minimum of a two recording-channel system, with Cz-Ac or Cz-Mc in the second channel, has been recommended 12 because this channel may aid in the identification of waves IV and V, which may be fused in the channel 1 waveform. An Ac-Ai or Mc-Mi channel may assist in the identification of wave I, and can be substituted in channel 2 when necessary, or can be recorded as a standard part of all tests if more than two recording channels are available. The raw analog data are amplified by high inputimpedance differential amplifiers with a common mode rejection ratio of at least 80 dB (10,000:1).J2 A typical analog filter bandpass is 100 Hz or 150 Hz to 3,000 Hz (-3 dB points). The analog gain depends on the input window of the analog-to-digital converter; a value of approximately 100,000 is typical. Data are typically digitized over an epoch durat.ion or analysis time of approximately 10 msec (the analysis

493

time in some recording systems is actually 10.24 msec). However, a longer analysis t.ime of 15 msec may be required for recording pathologically delayed BAEPs, BAEPs to lowered stimulus intensities (as when recording a lat.ency-intensity study), BAEPs in children, and BAEPs during intraoperative monitoring. The Nyquist criterion requires analog-to-digital conversion with a sampling rate of at least 6,000 Hz (2 x 3,000 Hz) in each channel to avoid aliasing, but higher sampling rates are required for accurate reproduction of the BAEP waveshape and precise measurement of peak latencies. The an alog-to-digital conversion should use at least 256 points per epoch; sampling of a 10.24-msec epoch at 256 time points corresponds to a sampling interval of 0.4 msec and a sampling rate of 25,000 Hz. Far-field BAEPs are too small to be visible in unaveraged raw data, and so signal averaging is required. The improvement in the signal-to-noise ratio is proportional to the square root of the number of data epochs included in the average. Automatic artifact rejection is used to exclude sweeps with high-amplitude noise from the average. The number of epochs per trial is typically 2,000, although a larger number may be required if the signal-to-noise ratio is poor (usually reflecting a lowamplitude BAEP or noisy raw data). At least two separate averages should be recorded and superimposed to assess reproducibility of the BAEP waveforms. Latency replication to within 1 percent of the sweep time and amplitude replication to within 15 percent of the peakto-peak amplitude have been recommended as standards for adequate reproducibility.!"

Waveform Identification and Measurement When recording far-field potentials such as the BAEPs, the two electrodes connected to the inputs of a differential amplifier cannot be considered active and reference electrodes. The voltage distributions of the BAEPs extend over most, if not all, of the head, and no cephalic electrode can be considered to be truly inactive for all components. For example, wave V in the Cz-Ai BAEP waveform largely reflects positivity at the vertex, whereas wave I is derived from a negativity around the stimulated ear that is picked up by the Ai electrode. 19 .20 Thus, both input. electrodes must be specified when describing a BAEP waveform. The Cz-Ai BAEP is typically displayed so that positivity at the vertex relative to the stimulated ear is displayed as an upward deflection, and the upward-pointing peaks are labeled with Roman numerals according to the convention established by Jewett and Williston II (see Fig. 23-4; Fig. 23-6). The downward-pointing peaks are labeled with the suffix N according to the peak that they follow; for example, downward peak IN follows

494

ElECTRODIAGNOSIS IN ClINICAL NEUROlOGY

Cz-Ai III Brainstem auditory evoked potentials recorded simultaneously from three different recording electrode linkages following monaural stimulation. The vertical dashed lines indicate the peak latenciesof waves IVand Yin the Cz-Ai waveforms. Note the decreased wave IVlatencyand the increased wave V latency in the Cz-Acwaveforms.

FIGURE ]]-6

Cz-Ac

Ac-Ai

upward wave I. The downward deflection following wave V, which has also been labeled the slow negativity (SN), is typically wider than the positive components and the earlier negative peaks. Recognition and identification of the components in BAEP waveforms obtained with standard recording techniques will now be described. Modifications in the recording paradigm that can be used to enhance specific BAEP components are discussed later in this chapter. The BAEP waveform typically begins with an electrical stimulus artifact that is synchronous with stimulus production at the transducer. Wave I is the first major upgoing peak of the Cz-Ai BAEP. It appears as an upgoing peak of similar amplitude in the Ac-Ai waveform and is markedly attenuated or absent in the Cz-Ac waveform (see Fig. 23-6). The cochlear microphonic may be visible as a separate peak preceding wave I, especially if the stimulus artifact is small. Its scalp distribution is similar to that of wave I. They may be distinguished by reversing the stimulus polarity, which will reverse the polarity of the cochlear microphonic; wave I may show a latency shift, but will not reverse polarity. A bifid wave I is occasionally present and represents contributions to wave I from different portions of the cochlea. The earlier of the two peaks, which reflects activation of the base of the cochlea, corresponds to the single wave I that is typically present in the Cz-Ai waveform. Reversal of stimulus polarity can be used to distinguish a bifid wave I from a cochlear microphonic followed by (a single) wave I. In contrast to wave I, wave IN is present at substantial amplitude in the Cz-Ac channel (see Fig. 23-6). This downgoing deflection is usually the earliest BAEP component in that waveform.

Wave II is typically the first major upward deflection in the Cz-Ac waveform because wave I is markedly attenuated or absent there. When present, wave II is usually of similar amplitude in the Cz-Ai and Cz-Ac channels (see Fig. 23-6). However, wave II may be small and difficult to identify in some normal subjects. A substantial wave III is usually present in both the Cz-Ai and Ac-Ai channels. Wave III in the Cz-Ac waveform is usually substantially smaller than that in the Cz-Ai waveform (see Fig. 23-6). This difference helps to distinguish it from wave II, whose amplitude is little changed by the change in the ear electrode. The peak latency of wave III decreases slightly, whereas that of wave II increases slightly when the inverting amplifier input is changed from Ai to Ac; thus, the Cz-Ai waveform tends to give better separation of waves II and III than does the Cz-Ac waveform. A bifid wave III is occasionally observed as a normal variant (Fig. 23-7); the wave III latency in such waveforms can be scored as midway between the peak latencies of the two subcomponents. Rarely, wave III may be poorly formed or absent in a patient with a clear wave V and a normal I-V interpeak interval; this finding is best interpreted as a normal variant waveform. Waves IV and V are often fused into a IVIV complex whose morphology varies from one subject to another, and may differ between the two ears in the same person (Fig. 23-8). The IVIV complex is often the most prominent component in the BAEP waveform. It is usually followed by a large negative deflection that lasts several milliseconds and brings the waveform to a point below the pre-stimulus baseline. Occasionally, the most negative point in the waveform follows waveVI (see Fig. 23-7), which could lead to misinterpretation of wave VI as an abnormally delayed wave V. The identity of wave V

Brainstem Auditory Evoked Potentials

495

IV III II

C clicks Cz-Ai

VVI

t A

0.21J.V [

D

R clicks Cz-Ai

1

1

B

Simultaneous Cz-Ac I

I

I

I

6 10 8 msec FIGURE 11·7 Ii Brainstem auditory evoked potentials recorded from a normal subject following stimulation with compression (C) and rarefaction (R) clicks. Prominent differences in wave shape are evident between the two stimulus polarities. The latency ofwave VI and the relative amplitudes ofwaves IV and V are markedly affected by the click polarity. The lowest point in the waveform after wave V precedes wave VI with compression clicks and follows wave VI with rarefaction clicks. A bifid wave III is present in the Cz-Ai waveform elicited by compression clicks. (By permission of the Mayo Foundation.) 2

E

4

can be clarified by changes in recording montage or click polarity (see Figs. 23-6 and 23-7) or by reducing the stimulus intensity and/or increasing the stimulus rate to attenuate wave VI relative to wave V. In distinguishing between a totally fused IVIV complex and a single wave IV or V, Epstein notes that the former has a "base" that is greater than 1.5 msec in duration, whereas the width of a single wave is less than 1.5 msec." When waves IV and V overlap in the Cz-Ai waveform, the wave V latency measurement used for BAEP interpretation should be taken from the second subcomponent of the IVIV complex even if this is not the highest peak (in contrast to the amplitude measurement used to calculate the IVIV:I amplitude ratio, which is taken from the highest point in the complex). Measurement of the peak latency of wave V may be inaccurate if V appears only as an inflection on the falling edge of wave IV (see Fig. 23-8) and it may be impossible if they are smoothly fused. Two approaches may be used in such cases. The first involves measure-

t

t

c

F I

!

I

!

I

,

5 msec FIGURE 11-8 III Various IV/V complex morphologies in Cz-Ai waveforms recorded in normal subjects. (From Chiappa K, Gladstone KJ, Young RR: Brain stem auditory evoked responses. Studies of waveform variations in 50 normal human subjects. Arch Neurol, 36:81, 1979, with permission.)

ment of wave V latency in a Cz-Ac recording channel. The overlapping peaks are more clearly separated there because the latency of wave IV is typically earlier, and that of wave V is later, than in the Cz-Ai waveform (see Fig. 23-6). However, because of these latency shifts, wave V latency values measured in a Cz-Ac waveform should be compared with normative data in which the latency of wave V was also measured in a Cz-Ac recording. The other approach is to reduce the stimulus intensity to attenuate wave IV relative to wave V and permit accurate measurement of the peak latency of wave V. That latency value cannot be compared with normative data obtained at a higher stimulus intensity, but the I-V interpeak interval can be evaluated because it is minimally affected by changes in stimulus intensity. It is important to distinguish wave V from wave IV. If wave V were abnormally delayed, but an earlier and larger wave IV (which dominated the IVIV complex) was mistaken for wave V, the BAEP abnormality might be missed. If the latency of an apparent wave V is abnormally short, efforts should be made to determine whether this peak is in fact a dominant wave IV. Lesions that affect wave V almost always also affect wave IV,22-24 but rarely wave IV may be unaltered (Fig. 23-9).

496

ELECTRODIAGNOSIS IN CUNICAL NEUROLOGY

v

A o.1Il- V

L 1 msec

FIGURE 23·1 • BAEPs to left ear stimulation recorded (Cz-Ai) during surgery for a basilar artery aneurysm. The aneurysm ruptured and the basilar artery was transiently dipped to control the bleeding. The patient suffered a brainstem infarct. A, Clear waves I through VI were present in these BAEPs recorded just before the aneurysm ruptured. B, BAEPs recorded after the clip was removed show a loss of wavesV and VI. Waves I through IV were unaffected. (Modified from Legatt AD, Arezzo JC, Vaughan HG Jr: The anatomic and physiologic bases of brain stem auditory evoked potentials. Neural CIin, 6:681,1988, with permission.)

B

Clinical interpretation of BAEPs is based predominantly on the latencies ofwaves I, III, and V. Once these peaks have been identified and their latencies measured, the I-III, III-V, and I-V interpeak intervals are calculated. The latency of wave I has also been labeled the peripheral transmission time (PTT) and the I-V interpeak interval has been called the central transmission time (CTT). Differences of component latencies and interpeak intervals with stimulation of the right and left ears are also calculated. The amplitudes of waves I and the IVIV complex are measured, each with respect to the most negative point that follows it in the waveform (I to IN and IVIV to VN), and their ratio is calculated. An excessively small IVIV:l amplitude ratio can identify as abnormal some BAEP waveforms in which all component latencies and interpeak intervals are normal (Fig. 23-10). It has sometimes been called the V:I amplitude ratio. However, most studies establishing the value of this quantity24-26

measured the amplitude from the highest peak of the IVIV complex to VN, not from the wave V peak to VN. Measuring the latter may give an abnormally small ratio in normal waveforms in which the peak of wave IV is much higher than that of wave V.

Patient Relaxation and Sedation The amplifier bandpass used for BAEP recordings filters out all of the delta, theta, alpha, and beta bands of the electroencephalogram, and the biologically derived noise in the recordings is predominantly derived from muscle activity. Therefore, patient relaxation during the recording session is essential to get "clean" waveforms with a good signal-to-noise ratio. Patients are usually tested while lying comfortably so that their neck musculature is relaxed. Patients should be requested to let their mouth hang open if the muscles of mastication

II

05.vL 1 msec

1.54

3.62

5.36

FIGURE :Z:S·IO • Brainstem auditory evoked potential (Cz-Ai) in a 27-year-oldwoman with probable multiple sclerosis. The IVIV:I amplitude ratio is 0.28; all absolute latencies and interpeak intervals are normal. The stimulus intensity was 65 dB nHL.

Brainstem Auditory Evoked Potentials

are tensed. They are encouraged to sleep during testing because this aids relaxation and will not alter the BAEPs.27 If the patient cannot relax sufficiently, sedation can be induced with agents such as chloral hydrate, a shortacting barbiturate, or a benzodiazepine; these have little or no effect on BAEPs in the usual sedative doses."

MODIFICATIONS TO BAEP RECORDING TECHNIQUES Reducing the Stimulus Artifad The transient electrical current in the transducer that generates the acoustic stimulus produces a transient electromagnetic field, which in turn induces a voltage in the patient and in the recording leads. This voltage creates an electrical stimulus artifact that, if large and prolonged, may overlap with wave I and impair the identification and measurement of that component. Using shielded headphones and headphones with piezoelectric transducers/" instead of voice coil transducers can reduce this artifact. Transducers that are connected to an ear insert by flexible plastic tubing several centimeters in length may also be used to mitigate this problem. The BAEPs, including wave I, are delayed by the time required for the acoustic stimulus to propagate through the plastic tube, but the electrical stimulus artifact remains simultaneous with transducer activation; the increased temporal separation between them permits greater decay of the electrical artifact before wave I. The greater distance between the transducer and the earlobe or mastoid recording site also serves to reduce the amplitude of the electrical stimulus artifact. The delay introduced by the plastic tube must be considered when one interprets BAEP latencies; passage through the tube may also change the acoustic properties of the stimulus. Ideally, the normative data used for BAEP interpretation should have been acquired with the same techniques that are used to test the patients. When ear inserts are used, they may be covered with metal foil to serve as near-field recording electrodes for wave I, thereby yielding a larger wave I. Responses to compression and rarefaction clicks may be averaged together to reduce the electrical stimulus artifact by cancellation. Most evoked potential recording svstems have an option for alternating the click polarity during stimulus delivery. This option should be avoided if possible during diagnostic testing because the BAEPs elicited by the two click polarities may differ.!""? Stimulus artifact is usually more of a problem in the operating room environment, and alternating click polarity is often necessary during intraoperative BAEP monitoring. In this setting, the BAEPs are being compared with signals recorded earlier in the same patient

497

with the identical stimulus paradigm, so the admixture of responses to the two stimulus polarities is less of a problem.

Improving the Resolution of Specific Components Wave I may be small and difficult to record in some patients, especially if hearing loss is present. Several techniques can be used to obtain a clearer wave I. Because wave I is a near-field potential, relatively small movements of the Ai/Mi recording electrode can have a substantial effect on its amplitude. Thus, alternative electrode positions around the stimulated ear can be used. An electrode within the external auditory canal yields an even larger wave I, and may take the form of a metal foil covering on an ear insert or a spring-leaf electrode that makes contact with the wall of the canal. Although wave I appears predominantly because of the presence of a near-field negativity at the Ai electrode, the horizontal orientation of its dipole projects a small positivity to the contralateral ear. Accordingly, an Ac-Ai recording channel can yield a somewhat larger and clearer wave I than that in the standard Cz-Ai recording. Modifications in stimulus parameters are also useful in obtaining a clearer wave l. Alternating stimulus polarity can be useful by attenuating a large stimulus artifact that is obscuring wave I or by helping to differentiate wave I from the cochlear microphonic. A reduction in the stimulus repetition rate may yield a clearer wave I. Increasing the stimulus intensity is particularly helpful in improving the clarity of wave I and may be used if that peak is not clearly identifiable at standard intensities in the presence of hearing loss. One caveat is that in patients with a high-frequency hearing loss of cochlear origin, this maneuver yields interpeak intervals that are shorter than those of normal subjects.F' This difference in interpeak intervals probably reflects contributions to wave I from different portions of the cochlea. At normal intensities, the earliest peak, which reflects activation of the base of the cochlea, predominates. At higher intensities, a somewhat longer-latency component becomes increasingly prominent. If the earlier component is missing because of high-frequency hearing loss, and if stimulus intensity is increased, the wave I that appears may predominantly reflect the longer-latency contribution; the I-III and I-V interpeak intervals are thus shorter than those that would have been measured from the usual wave I. Measurement of the PTT from this longer-latency subcomponent would decrease the sensitivity of the test for recognition of prolonged I-III or I-V interpeak intervals. Wave V is often fused with wave IV into a IVIV complex in the Cz-Ai BAEP waveform. A Cz-Ac recording

498

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

channel helps to identify wave V by increasing the sep· aration between the peaks of these two components; the latency of wave IV is shorter and that of wave V is longer than in the Cz-Ai recording (see Fig. 23-6). Because of this latency shift, the latency of wave V in a patient's Cz-Ac waveform should be compared with normal values derived from Cz-Ac recordings in control subjects. Modifications of the stimulus parameters can also be used to identify wave V if it is unclear when standard recording techniques are used. Wave V is the BAEP component most resistant to the effects of decreasing stimulus intensity (Fig. 23·11) or increasing stimulus rate. If either of these stimulus modifications is performed progressively until only one component remains, that peak can be identified as wave V and then traced back through the series of waveforms to identify wave V in the BAEP recorded with the standard stimulus. Occasionally, wave V may be present following stimulation with one click polarity but not the other (Fig. 23-12). Therefore, recording a BAEP with the opposite stimulus polarity may be useful if wave V is not identifiable with the standard laboratory protocol.

Stimulation at Several Intensities In a patient with a conductive hearing loss, the stimulus intensity reaching the cochlea is less than that delivered to the external ear, and an abnormal BAEP with a delayed or absent wave I may result. If the stimulus intensity is increased to compensate for the conductive hearing loss, and no coexisting sensorineural hearing loss is present, a normal BAEP will be recorded. In contrast, BAEPs that are delayed as a result of abnormally slowed neural conduction do not normalize when the stimulus intensity is increased (Fig. 23·13). Thus, increasing the stimulus intensity can help to differentiate peripheral from neural abnormalities, especially when wave I is not clear. The degree of hearing loss can be estimated by the elevation of the subjective auditory click threshold; increasing the stimulus intensity to compensate for this elevation is thus equivalent to stimulating at a specific intensity above that threshold (e.g., at 60 to 65 dB SL rather than at 60 to 65 dB HL). The normative data to which such recordings are compared should also have been recorded at a specific intensity in dB SL, rather than db HL.

8.0

.................. : I

,

~ III

7.5

-:

~

:

,.

.

-

.

., •

I

· ... , ··,

-

,

·

0

c: 6.5 CD

...

..

-.,..

..

., . ,

, , , ,

>.

.

..

·, " ., .................................

..s 7.0

:

. . ., . .-.-.. . - - .. -.-.. .. _

.. .. .. ..

.

.

. . ,. . .. . . . ...

19

> 6.0 CD ~

3: 5.5 5.0

+--+------1~-t--+-_+--+---l

o

10

20

30

40

50

60

70

Stimulus intensity (dB nHL)

5 FIGURE 23-11 • Brainstem auditory evoked potentials (BAEPs) recorded at progressively lower stimulus intensities are shown on the left. Wave V latencies (arrows) are measured and then plotted as a function of stimulus intensity to give the latency-intensity curve on the right. This 35-year-old woman with dizziness and tinnitus had normal BAEPs and normal magnetic resonance imaging findings.

Brainstem Auditory Evoked Potentials

499

AD

R 70 dB SL

R 70 dB SL

R 55 dB SL

C 70 dB SL

t

C 70 dB SL

t

FIGURE 23-12 • Brainstem auditory evoked potentials elicited by rarefaction (R) and compression (C) clicks of several different intensities. Wave V is absent following left ear stimulation with 70 dB SL rarefaction clicks but appears when the stimulus intensity is decreased to 55 dB SL or when stimulus polarity is reversed. A clear wave V is present following right ear stimulation with 70 dB SL clicks of either polarity. The calibration bars are 0.25 IlVand 1.0 msec. (From Emerson RG, Brooks EB, Parker SW et al: Effects of click polarity on brainstem auditory evoked potentials in normal subjects and patients: unexpected sensitivity ofwave V. Ann NY Acad Sci, 388:710,1982, with permission.)

If each ear is stimulated at several different in tensities, and the latency of wave V is graphed as a function of stimulus intensity, a latency-intensity curve is produced (see Fig. 23-11). Examination of latencyintensity curves may help to classify a patient's hearing loss.t" A shift of the curve to a higher intensity level without a change in its shape suggests conductive hearing loss, whereas a change in the shape of the curve with an increased slope suggests sensorineural hearing loss. Latency-intensity curves may also reveal abnormalities that are not demonstrated by BAEP recordings at the standard, relatively high stimulus intensity typically used in BAEP studies performed for neurologic diagnosis."

Rapid Stimulation A stimulus rate of approximately 10 Hz is used for routine clinical testing; this is because some of the BAEP components become attenuated and less clearly defined and the interpeak intervals lengthen as the rate is increased substantially above this. Measurement of the BAEP threshold is based solely on the presence or absence of wave V because it is the last BAEP component to disappear as the stimulus intensity is reduced. Because wave V is relatively resistant to the effects of rapid stimulation, recordings made solely for threshold

measurement can be accomplished more rapidly by increasing the stimulus rate to 50 to 70 Hz. However, rapid stimulation can make BAEPs undetectable, especially in premature infants. Thus, infants who appear to have a hearing loss on BAEP screening using rapid stimulation should be retested using a slower stimulation rate. 32 When stimuli are delivered at a rate of about 10 Hz, the neural elements generating the BAEPs have approximately 100 msec to recover after responding to one stimulus before they must respond to the next. It has been speculated that in some pathologic states, action potential propagation or synaptic transmission occurs normally with recovery times of this magnitude but is delayed when recovery times are shorter. If this were the case, more rapid stimulation would demonstrate BAEP abnormalities in some patients with normal BAEPs to the standard stimuli, thus increasing the sensitivity of the test. Rapid stimulation has been reported to increase the test sensitivity of BAEPs in patients with neurologic abnormalities in some, but not all, studies. 33.34

Altemative Auditory Stimuli The standard BAEP stimulus is a "broad-band" click produced when an electrical square pulse is delivered

500

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

4.54 I

6.38 I

A 5.80 2.903.82 I I

4.32 I

I

1.66 I

o.1/l-VL 1 msec

8

D FIGURE 2:1-15 • Brainstem auditory evoked potentials (BAEPs) elicited by two different stimulus intensities in

two patients and recorded in Cz-Ai montage. A, Left ear stimulation at 70 dB nHL elicits an abnormally delayed wave V in a 3&-year-old woman with a peripheral hearing loss in the left ear. Wave I is not visible. B, When the stimulus intensity is increased to 85 dB nHL, wave I becomes identifiable and the latency of waveV is markedly decreased. Interpeak intervals are normal. C, Left ear stimulation at 65 dB nHL in a 14-year-old boy with a fourth ventricular tumor elicits clear BAEPswith a normal wave I latency and an abnormally delayed waveV.The I-III and I-V interpeak intervals are both abnormally prolonged. D, An increase in the stimulus intensity to 85 dB nHL causes a slight decrease in all component latencies. Wave V remains abnormally delayed, and the I-III and I-V interpeak intervals remain abnormally prolonged.

to the headphone or other electromechanical transducer. It is generated predominantly by the region of the cochlea responding to 2000- to 4000-Hz sounds,35,36 although wave V may also receive contributions from lower-frequency regions of the cochlea. BAEPs have also been recorded following stimulation with brief tone pips, in an effort to probe specific parts of the cochlea. This method can be used to determine thresholds at different frequencies for BAEP audiometry. Tone pips are most often generated with stimuli consisting of brief bursts of sine waves. The burst cannot be abruptly stopped and started because an audible "click" containing many frequencies would be produced. Instead, the sine wave stimulus is amplitude-modulated with a rise time, plateau, and fall time to reduce (although not eliminate) the energy content of the stimulus at other frequencies. A filtered click, which is obtained by passing a broad-band click through a bandpass filter, has also been used as a frequency-specific stimulus. Another technique that can be used to obtain frequency-specific information from BAEPs is acoustic

masking. In one approach, the broad-band clicks used to elicit the BAEPs are superimposed on white noise that has been highpass-filtered by using different cutoff frequencies in different runs. Each BAEP waveform is generated predominantly by the unmasked region of the cochlea and represents a response to frequencies lower than the cutoff frequency used during that run. Subtraction of one such response from another yields a "derived" response that estimates the response to the frequency band between the cutoff frequencies of the two runs (Fig. 23-14). The subtraction process yields waveforms with relatively poor signal-to-noise ratios,

however." Frequency-specific stimuli and frequency-specific masking can be combined by embedding tone pips in continuous notched noise. The latter is white noise that has been notch-filtered to remove the basic frequency of the tone pip. Shaping a pure tone into a pip with an onset and an offset introduces power at other frequencies into its frequency spectrum, which would impair the frequency specificity of the test, but the notched noise masks the responses to frequencies other than

Brainstem Auditory Evoked Potentials Click • high-pass noise

.. A

Derived narrow-band response

~~

1200nV

OO~A~p.~12

12J~!\~

~7

~8~:~~D~5.2 -6 .../_. I\v ~ f4

~200

_ A-."va--

I

t>

153 ~_~ ~ ~

312

f\

'g

~1.5~~

i1~ "'-- ~ ~ 0.75~ - v - -or 1

5

10

msec

15

,/v··

f\

f\

V V~3.5

~ 2.6

.. -.-.A.A -""\/ V~ 1.8 :0

. .'v-"AM._ V- y--.1.3

~.9

p ~.65

0.5~200 o

/\J\f\;..

to 15 msec each) produces an evoked potential that contains oscillatory components at the frequency of the tone (Fig. 23-15). Intracranial recordings in animals'" have shown that this frequencyfollowing response does reflect neuronal activity in the brainstem, although care must be taken to avoid contamination by the cochlear microphonic.P? The frequency-following response and the BAEPs do not appear to have the same neural generators.t? Whereas the BAEPs reflect activity originating in the base of the cochlea,35,36 the frequency-following response predominantly reflects activity originating in lower-frequency regions of the cochlea.t'<" However, the clinical utility of frequency-following responses for assessing hearing in the lower frequencies has been limited by the technical difficulties associated with recording them and by their relatively high thresholds. BAEPs may also be elicited by bone-eonducted stimuli. 43,44 This is most useful in assessing patients who may have conductive hearing losses, such as neonates in whom BAEPs performed with air-conducted stimuli are suggestive of a hearing loss.

~.45 o

5

10

15

msec

FIGURE 1]·14 • Derivation of frequency-specific brainstem

auditory evoked potentials (BAEPs) with high-pass noise masking. Alternating-polarity, 170-llsec-duration clicks at an intensity of 60 dB SL were delivered at a rate of 13 Hz. The noise, which wasdelivered to the same ear, was highpass-filtered with cascaded filters to produce a filter function with a slope of 96 dB per octave. The BAEPs elicited by the clicks in the presence of the masking noise are shown in the left column; the number at the left of each waveform indicates the cutoff frequency of the noise filter. The uppermost waveform (marked "00") is the conventional BAEP, without masking noise. The derived responses, obtained by calculating the differences between pairs of adjacent BAEPs from the left column, are shown in the right column. CF, central frequency assigned to each derived waveform, based on the characteristics of the noise filter. (From Don M, Eggermont1J: Analysisof the clickevoked brainstem potentials in man using high-pass noise masking.] Acoust Soc Am, 63:1084,1978, with permission.)

BAEPs to Eledrical Stimuli BAEPs can also be recorded following electrical stimulation of eighth nerve fibers through the electrodes of

Tone

A

B the basic frequency of the tone pip. BAEP audiograms produced with such stimuli are similar to behavioral audiograms in the same subject." Wave V is broader in the frequency-specific BAEPs elicited by the lower frequencies; when these are recorded, reducing the lowcut (high-pass) analog filter in the evoked potential averager to 30 Hz may yield a clearer wave V and more reliable results.t" Stimulation with relatively prolonged, low-frequency tone bursts (e.g., 200- to 500-Hz tone bursts lasting 10

501

10 msec [25 IJ..V I

I

FIGURE 23·15 • Averaged frequency-following response to a

500-Hz tone burst, recorded between the vertex and linked ears with amplifier filter bandpass of (A) 8 Hz to 10,000 Hz and (B) 200 Hz to 1,000 Hz. The acoustic stimulus, recorded by a microphone at the subject's earpiece, is shown at the top. Note the latency shifts between the onset and the offset of the stimulus and the response. (From Marsh]T, Brown WS,Smith ]C: Far-field recorded frequency-following responses: correlates of low pitch auditory perception in humans. Electroencephalogr Clin Neurophysiol, 38:113,1975, with permission.)

502

ElECTRODIAGNOSIS INCLINICAL NEUROLOGY

a cochlear prosthesis. This can be used to assess the proximity of these electrodes to the spiral ganglion during implantation and the adequacy of eighth nerve stimulation during programming of the processor.v" though recording of the evoked eighth nerve compound action potential through a different set of electrodes may prove to be a more useful tool for the latter application.f

EFFEa5 OF VARYING STIMULUS PARAMETERS

Stimulus Polarity Responses to rarefaction and compression clicks delivered to the same ear may differ (see Fig. 23-7). Component latencies following rarefaction clicks are usually, although not always, shorter than those following compression clicks.I" The BAEP latency differences between the two stimulus polarities tend to be greatest in patients with cochlear high-frequency hearing 10ss.29 BAEP peaks tend to be clearer with rarefaction clicks because of fusion of adjacent components when compression clicks are used," although again there are exceptions. The presence of wave V following one click polarity, but not the opposite, has been reported as an uncommon finding in patients referred for BAEP studies'" (see Fig. 23-12); the significance of this finding is unclear.

Stimulus Intensity As the stimulus intensity is decreased, the latencies of all BAEP components increase (see Fig. 23-11). The increase is predominantly caused by a latency shift of wave I; the interpeak intervals are relatively constant until the stimulus intensity nears threshold. Component amplitudes decrease as the stimulus intensity is lowered, but with different patterns. Wave I shows the most rapid attenuation as the stimulus intensity is lowered, and it is usually lost before waves III and V. Wave V is attenuated to a lesser degree at lower stimulus intensities and is the last peak to disappear; its disappearance defines the BAEP threshold. At lower intensities, the IVIV:I amplitude ratio is increased because of the greater degree of attenuation of wave I.

most resistant to these effects. Thus, a stimulation rate of approximately 10 Hz is preferable for routine clinical testing, but more rapid rates may be used to facilitate recordings to measure the wave V threshold.

GENERATORS AND SCALP TOPOGRAPHIES OF RAEPs In their first detailed description of BAEPs, Jewett and Williston hypothesized that most of the BAEP components represented composites of contributions of multiple generators. I I This premise is supported by the bifid wave III seen in some normal BAEPs (see Fig. 23-7), which contains two subcomponents that overlap in most subjects. The effects of variations in stimulus parameters on the topography of wave II suggest that this component arises from more than one generator. 4H Shifts in the latencies of many BAEP peaks across various scalp recording locations-" (see Fig. 23-6) indicate that these peaks represent summations of overlapping components. Intraoperative recordings in humans'? and detailed intracranial mapping studies in animals 50.51 have also demonstrated multiple generators for many BAEP components. The complexity of the generators of human BAEPs (Fig. 23-16) derives in part from the pattern of connections within the auditory pathways, with ascending fibers both synapsing in and bypassing various relay nuclei.52-54 It also reflects the presence of two bursts of activity in the auditory nerve, seen as the Nl and N2 action potentials of the ECochG, which can drive the more rostral pathways. Because of both of these factors, several different structures within the infratentorial auditory pathways may be active and may generate field potentials simultaneously. Although most early reports of generators based on clinicopathologic correlations stated that their conclusions and summary figures were probably simplifications,22-24 many subsequent authors and investigators have assumed a one-peak to one-generator correspondence. Inasmuch as human intracranial and clinical data tend to identify only the major generator of each wave, it is important to realize that other generators may provide significant contributions that may be unmasked if the major contribution is lost. If these contributions were construed as the contribution of the major generator, the clinical interpretation of these BAEPs could be erroneous.

Stimulus Rate Increasing the stimulus rate increases both the absolute latencies of the BAEP components and the interpeak in tervals. As the stimulus rate is increased above approximately 10 per sec, component amplitudes decrease and the peaks tend to become less well defined. Wave V is

Wave I Wave I arises from the first volley of action potentials in the auditory nerve at the most distal (i.e., closest to the cochlea) portion of the nerve. It represents the same

Brainstem Auditory Evoked Potentials

503

FIGURE 2J-16 III Diagram showing the probable generators of the human brainstem auditory evoked potentials. SN, slow negativity after wave V; AC, auditory cortex; AR, auditory radiations; BIC, brachium of the inferior colliculus; CN, cochlear nucleus; Ie. inferior colliculus; LL, lateral lemniscus; MGN, medial geniculate nucleus; SOC, superior olivary complex. (From LegattAD, Arezzo JC, Vaughan HG Jr: The anatomic and physiologic bases of brain stem auditory evoked potentials. Neurol CHn, 6:681, 1988, with permission.)

electrical phenomenon as the Nl component of the eighth nerve compound action potential in the ECochG, as confirmed by simultaneous BAEP and ECochG recordings.55 There may also be a small contribution from the cochlear summating potential that precedes and overlaps with the Nl in ECochG recordings from the human external auditory meatus.P" The origin of wave I in the most distal portion of the auditory nerve is demonstrated by its presence in some patients who fulfill clinical and electroencephalographic criteria for brain death 57-59 (Figs. 23-17 and 23-18) and by its occasional persistence after section of the auditory nerve during acoustic neuroma surgery6°,61 (Fig. 23-19). The lateral location of the generator accounts for its surface distribution, a circumscribed negativity around to the stimulated ear. 20 Wave I therefore appears in Cz-Ai and Ac-Ai recordings but is minimal or absent in Cz-Ac recordings.

in the auditory nerve generates wave IN as it passes the internal auditory meatus and moves from a nerve encased in bone to one surrounded by cerebrospinal fluid. Human intracranial recordings at the internal auditory meatus contain a large peak synchronous with IN49 (Fig. 23-20), and experimental inactivation of the intracranial eighth nerve reduces or eliminates wave IN but spares wave I in anirnals." Because IN originates in a cranial nerve, it may persist and may be even larger than wave I in brain-dead patients'? (see Fig. 23-17) and in patients in whom an acoustic neuroma has eliminated all centrally generated BAEP components (Fig. 23-21). Just as the upward-pointing wave I corresponds to a negativity around the stimulated ear, the downwardpointing wave IN is synchronous with a positivity at the mastoid'" and in the ECochG.65 However, the field of IN also includes a far-field negativity around the vertex. 20 Thus, in contrast to wave I, wave IN is prominent in Cz-Ac BAEP waveforms (see Fig. 23-6).

Wave IN Propagating action potentials can produce far-field potentials at points where the impedance of the tissue surrounding the nerve changes. 62.63 The afferent volley

Wave II Sounds at the intensity levels used for BAEP recordings elicit two volleys of activity within the auditory nerve,

504

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

3 trials of 4,000

[ 0.1 IJ-V

I

I

!

I

2

4

6

8

I 10

msec fiGURE 23-18 • Brainstem auditory evoked potentials in a 52-

year-old woman with complete brain death, including necrosis of the cochlear nuclei, verified at postmortem examination. Waves I and II are both present. Simultaneous electrocochleography suggested that the latter originated in the N2 component of the eighth nerve action potential. (By permission of the Mayo Foundation.)

2 msec fiGURE 13·17 • Brainstem auditory evoked potentials from

four patients fulfilling clinical and electroencephalographic criteria for brain death. Wave I is present in the upper three waveforms, and an even larger wave IN is present in the middle two. (Modified from Starr A: Auditory brain-stem responses in brain death. Brain, 99:543, 1976, with permission.)

corresponding to the N1 and N2 components of the eighth nerve action potential in the ECochG. When the N2 component begins in the distal auditory nerve, the activity of the first volley has propagated to the proximal auditory nerve or cochlear nucleus, and both contribute to the generation of wave II. The contribution from N2 in the distal auditory nerve is confirmed by simultaneous BAEP and ECochG recordings'" and by recordings from the intracranial eighth nerve 49 ,66 (see Fig. 23-20). It arises from the same location and is generated in the same wayas wave I, and has a similar topography.''? The N2 contribution accounts for the presence of wave II in patients with acoustic neuromas that have destroyed the proximal eighth nerve and eliminated all later BAEP components 23,68 (see Fig. 23-21), in a patient with a lower brainstem lesion that compromised the proximal eighth nerve as far as the internal meatus (wave IN was eliminated) ,69 and in some brain-dead patients59 ,68 (see Fig. 23-18). The wave II in Figure 23-22, recorded in a patient with an acoustic neuroma that markedly

delayed wave V, can also be attributed to the distal eighth nerve. Waves I and II were equally delayed, a reflection of peripheral auditory dysfunction. Wave V was delayed to a much greater extent, the difference representing slowed conduction through the eighth nerve. Because wave II did not show this additional delay, it must have originated distal to the tumor. The origin of the more rostrally generated part of wave II is controversial. It is the earliest component affected by pontomedullary cerebrovascular accidents involving the cochlear nucleus. 22 ,70 This finding has been interpreted as implying a generator within the brainstem, specifically the cochlear nucleus or its outflow. Dissection within the pons can eliminate the farfield wave II in the Cz-Ac channel, whereas the nearfield wave II recorded at the ipsilateral ear persists, indicating that the former is generated within the brainstem.?! Intracranial recordings over the pons and from within the ventricular system have also been interpreted as indicating a generator within the brainsterrr" (Fig. 23-23). However, when the propagating N1 volley is recorded at multiple sites along the intracranial eighth nerve, the peak latency of the compound action potential at the brainstem end of the nerve approximates that of wave II, thus suggesting a contribution from the proximal eighth nerve 49 •72 (see Fig. 23-20). Meller and jannetta'" estimated auditory nerve conduction times and believed that timing considerations ruled out a contribution of postsynaptic cochlear nucleus activity to wave II. These data can be reconciled." and it is likely that activity within the proximal end of the auditory nerve and postsynaptic activity within the cochlear nucleus both contribute to wave II. Because the auditory nerve terminals are within the substance of the cochlear nucleus, the distinction between a proximal auditory nerve generator and a cochlear nucleus generator does

505

Brainstem Auditory Evoked Potentials

FIGURE 1J-19 • Intraoperative brainstem auditory evoked potential (Cz-Ai recording) to left ear stimulation showing persistence ofwave I after sacrifice of the intracranial eighth nerve during resection of a left-sided acoustic neuroma. (Modifed from Legatt AD, Arezzo JC, Vaughan HG Jr: The anatomic and physiologic bases of brain stem auditory evoked potentials. Neural Clin, 6:681, I9HH, with permission.)

-'. 1 msec

not have a major impact on the anatomic localization of the cause of a wave II abnormality. The brainstem generator is usually predominant and produces a scalp topography with maximal amplitude over the dorsal part of the head and a clear wave II in the Cz-Ac waveform.f" The presence of an additional generator in the distal

auditory nerve is clinically important, however, because it may produce a wave II in a patient whose entire brain, including the brainstem, is not functioning (see Fig. 23-18). If wave II were regarded as originating entirely within the brainstem, its presence would be interpreted as incompatible with a diagnosis of brain death. IV/v III II

VI

VII

N1 Normal lAM

[ 7.5 !LV CPA

C3

[ 0.25 !LV Neuroma

o

5

10

msec FIGURE 23-20 • Brainstem auditory evoked potentials (BAEPs) recorded directly from the human eighth nerve at the internal auditory meatus (top) and at the cerebellopontine angle (middle), as well as far-field scalp BAEPs (bottom). The onset and peak of PI are coincident with that of the scalp-recorded wave I, whereas Nl corresponds to component IN. Negativity with respect to the scalp muscle reference is displayed as an upward deflection in this figure. (From Hashimoto I, Ishivama Y, Yoshimoto T et al: Brain-stem auditory evoked potentials recorded directly from human brain-stem and thalamus. Brain, 104:841, 1981, with permission.)

025·~L 2 msec FIGURE 23-21 • Presence of waves I, IN, II, and lIN but absence of all subsequent waves in a patient with an acoustic neuroma. A normal Cz-Ai brainstem auditory evoked potential is shown for comparison. (Modified from Starr A, Hamilton AE: Correlation between confirmed sites of neurological lesions and abnormalities of far-field auditory brainstem responses. Electroencephalogr Clin Neurophysiol, 41:595, 1976, with permission.)

506

ELEaRODIAGNOSIS INCLINICAL NEUROLOGY

v FIGURE 23·22 • Brainstem auditory evoked

II

potentials (Cz-Ai) to right ear stimulation in a 36-year-old man with a right-sided acoustic neuroma. The stimulus intensity has been increased to 85 dB nHL. The absolute latencies of waves I and II are abnormally delayed, and the I-V interpeak interval is abnormally prolonged.

O.05.VL 2.3

1 msec

3.3

8.3

Wave liN N3

A

N4

As it passes the internal auditory meatus, the N2 volley of the eighth nerve compound action potential can generate a downgoing peak, wave lIN, in the BAEP waveform in the same manner as the Nl volley gives rise to wave IN. Human intracranial recordings at the internal auditory meatus have demonstrated compound action potential peaks corresponding to both IN and IIN. 49 Like IN, wave lIN may be recorded from patients in whom an acoustic neuroma has eliminated all centrally generated BAEP components (see Fig. 23-21).

B

Wave III Wave III also probably represents a composite of contributions from multiple generators. It occasionally has a bifid waveform (see Fig. 23-7), and may be more abnormal either ipsilateral''v" or contralateral/" to the major pathology in patients with asymmetric lesions. Human lesion data localize the major generators of v wave III to the caudal pontine tegmentum. Most patients with lesions in this region that involve the superior olivary complex have a normal wave II but abnor10 msec mal wave 111. 22 ,24,76 Wave III is present and is usually FIGURE 13-1] • Human brainstem auditory evoked potentials \ normal in patients with lesions confined to the middle (BAEPs) to binaural stimulation recorded during surgery fro~ . or upper pons or mesencephalon. 22- 25 ,77 Intracranial the midline dorsal pons (A)at the level of the facial colliculi and, recordings are also consistent with a pontine genera(B) the midline dorsal mesencephalon at the level of the infe49 (see Fig. 23-23). Latency shifts as a function of rostor rior colliculi, as well as (C) the far-field BAEP. Negativity with trocaudal position 49 ,78 suggest a contribution from respect to the scalp muscle reference is displayed as an upward activity ascending in the lateral lemniscus as well. deflection in this figure. (From Hashimoto I, Ishiyama y, There is also evidence for generation of wave III Yoshimoto T et al: Brain-stem auditory evoked potentials within the cochlear nucleus.l" Although wave III may recorded directly from human brain-stem and thalamus. Brain, 104:841,1981, with permission.) receive a contribution from cochlear nucleus activity

c

Brainstem Auditory Evoked Potentials

driven by the N2 volley of the eighth nerve compound action potential, the available data suggest that it is predominantly generated in the superior olivary complexes or their outflow within the lateral lemniscus. The scalp distribution of wave 111 20 includes both an amplitude maximum over the dorsal portion of the scalp and a substantial horizontally oriented component; thus, wave III is prominent in Cz-Ai, Cz-Ac, and Ac-Ai BAEP waveforms.

Wave IV Wave IV is affected by tumors or cerebrovascular acciden ts of the midpons or rostral pons. 22-24 Its generators are close to or overlapping with, but not identical to, those of wave V; waves IV and V are usually either both

III

c

507

affected or both unaffected by brainstem lesions, but they may be differentially affected by multilevel demyelination'< or a brainstem infarct" (see Fig. 23-9). Intracranial recordings demonstrate a large positivity coincident with wave IV over the dorsal pons'? (see Fig. 23-23). Latency shifts observed at serial intracranial recording positions are consistent with propagating action potentials within the ascending brainstem auditory pathways. Also, wave IV may persist in the presence of a lesion of the inferior colliculus that eliminates wave yli0 (Fig. 23-24) Thus, wave IV appears to reflect activity predominantly in ascending auditory fibers within the dorsal and rostral pons, caudal to the inferior colliculus. Although the nucleus of the laterallemniscus has been implicated as a BAEP generator in animal studies,81.82 human data are insufficient to confirm its contribution to wave IV.

FIGURE 13·14. A, Axial and B, sagittal MRI in a patient who developed left-sided hearing impairment following gammaknife treatment for a midbrain arteriovenous malformation that had bled. The treatment caused a lesion in the right inferior colliculus (bl,ack arrows). C, Brainstem auditory evoked potentials (BAEPs) recorded in this patient at two different stimulus intensities (70 and 90 dB). BAEPs to left ear (LE) stimulation are shown as solid lines, whereas those to right ear (RE) stimulation are shown as dashed lines. Wave IV is present for both stimulus lateralities, but wave V and following downgoing peak, VN, are absent following left ear stimulation. (Modified from Hirsch BE, Durrant JD, Yetiser S et al: Localizing retrocochlear hearing loss. Am J Otol, 17:537, 1996, with permission.)

IVN

1 msec/div

508

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Wave V Waves IV and V are the earliest components that are absent and usually are the earliest that are abnormal in patients with lesions of the midpons, rostral pons, or mesencephalon. 22-25,76.77.83-85 Occasionally, waves II and III are delayed in latency with rostral brainstem tumors 23.25 (Fig. 23-25); this delay may be caused by pressure on more caudal structures or it may reflect effects on descending pathways." In intracranial recordings, the positive peak corresponding to waveV is largest near the inferior colliculus contralateral to the stimulated ear78.87 (Fig. 23-26). It inverts to a negativity over the dorsal pons, in contrast to earlier waves, which do not display an inversion between the pons and the mesencephalon (see Fig. 23-23). Ipsilateral near-field responses recorded near the inferior colliculus are smaller and do not match the latency of wave V as well. These data are consistent with generation of wave V at the level of the mesencephalon, either from the inferior

N

.",

f/ ..

1\./

A

/V

'vo\

II \" "oJ \I

III I I I

\

v

B

III

~~ A L

o

2

4

6

8

10

12

14

16

18

20

Time (msec) FIGURE n·ll • Brainstem auditory evoked potentials (BAEPs) recorded during surgery from the human inferior colliculus (solid lines) compared with the far-field BAEP (dashed lines). A, Low-pass filtered waveforms. B, Digitally filtered waveforms. Negativity with respect to the reference is displayed as an upward deflection in this figure. Calibration bars: 1 I!Vfor inferior colliculus recordings and 0.2 I!V for far-field BAEP. (From M011er AR, Jannetta PJ: Interpretation of brainstem auditory evoked potentials: results from intracranial recordings in humans. ScandAudiol, 12:125, 1983, with permission.)

R

L

O.25 ..

~L 2 msec

FIGURE 2J-25 • In a patient in whom a germinoma (stiPPks) destroyed most of the mesencephalon and pons, including the right but not the left lateral lemniscus, waves I through III were present following stimulation of either the right (R) or the left (L) ear, but later brainstem auditory evoked potential (BAEP) components were absent. The latencies of waves II and III were abnormally delayed, but the waveI latencies were normal. A normal Cz-Ai BAEPis shown at the top for comparison. IC, inferior colliculus; LL, lateral lemniscus; PCN, posterior cochlear nucleus; ACN, anterior cochlear nucleus; SO, superior olivary nucleus. (From Starr A, Hamilton AE: Correlation between confirmed sites of neurological lesions and abnormalities offar-field auditory brainstem responses. Electroencephalogr Clin Neurophysiol, 41:595,1976, with permission.)

colliculus itself or, as some authors have suggested,79,88 from the fibers in the rostral portion of the lateral lemniscus as they terminate in the inferior colliculus. Normal BAEPs have been reported following damage in the region of the inferior colliculus caused by brainstern stroke,89 head trauma." and injury during resection of pineal region tumors,91.92 but it is not clear that the colliculi were totally destroyed in these cases. In contrast, Hirsch and associates reported a case in which a highly focal lesion involving the inferior colliculus eliminated wave V of the BAEP (see Fig. 23-24) .80 Intracranial data also suggest that the mesencephalon contralateral to the stimulated ear is the major generator of wave V. Clinically, however, unilateral abnormalities of wave V are most often associated

Brainstem Auditory Evoked Potentials

with ipsilateral pathology,74-76,H3,93 although there are exceptions.v'v" On the scalp, the amplitude of wave V is maximal in the midline.t? and a large wave V is present in hoth the Cz-Ai and the Cz-Ac waveforms. The monkey homologue of the human wave V,95 which likewise predominantly reflects activity in the mesencephalon (contralateral more than ipsilateral) ,51 also contains contributions from other structures, including the medial geniculate nucleus, auditory radiations, and lateral lemniscus within the pons. The lemniscal contribution reflects the multiple bursts of activity that are present at all levels of the auditory system. Because wave V may be eliminated by mesencephalic lesions,39,41,80 the lemniscal contribution is most likely not of clinical significance in humans. Depth recordings within the human medial geniculate nucleus suggest that activity in this structure contributes to wave V49,96,97 (Fig. 23-27), but this contribution does not affect the clinical interpretation of BAEPs. The normal wave V recorded in patients with rostral midbrain or supratentorial lesions 22,98 is gener-

MGB

1 mm

8mm

9mm

[ 0.5 /LV

C3

o

5

10

msec

FIGURE 13·27 • Brainstem auditory evoked potentials recorded by a thalamic depth electrode passed near or through the medial geniculate in a human patient. The two large positive peaks can be traced to the scalp (bottom waveform). Binaural stimulation was used. Negativity with respect to the scalp muscle reference is displayed as an upward deflection in this figure. (From Hashimoto I, Ishiyama Y, Yoshimoto T et al: Brain-stem auditory evoked potentials recorded directly from human brainstem and thalamus. Brain, 104:841, 1981, with permission.)

509

ated predominantly in the mesencephalon, just as it is in normal individuals; absence of the geniculate contribution is not apparent. Conversely, a lesion that eliminates the mesencephalic generators of wave V will remove the auditory input to the medial geniculate nucleus and eliminate its contribution to the BAEP waveform as well.

Wave VI Abnormalities of wave VI, with normal waves I through V, have been described in patients with tumors of the rostral midbrain and caudal thalamus at the level of the medical geniculate nucleus and the brachium of the inferior colliculus.F Depth recordings within the human medial geniculate demonstrate large near-field peaks coincident with wave VI; both positive 49,99 (see Fig. 23-27) and negative'" peaks have been recorded. The polarity differences most likely reflect differences in the position of the electrode relative to the equivalent dipole of the wave VI generator within the geniculate. The scalp topography of wave VI also provides evidence for generation ofwave V within the medial geniculate nuclei or their outflow tracts.?" The monkey homologue of the human wave VI95 also contains a contribution from the inferior colliculus." a reflection of the multiple bursts of activity within the auditory system. A large wave VI peak is visible in recordings from electrodes placed directly on the human inferior colliculus when the large slow-wave components in them are removed by filtering''? (see Fig. 23-26). Thus, the inferior colliculus may contribute to the human wave VI. This mesencephalic contribution could generate a wave VI in a patient with bilateral geniculate damage. Conversely, wave VI is absent in Cz-Ai and Cz-Ac recordings in some normal individuals. Thus, neither the presence of a normal wave VI nor the complete absence of this component provides clinically useful information about the status of the medial geniculate nuclei. BAEPs cannot be used to assess the status of the auditory pathways rostral to the mesencephalon.

Wave VII Wave VII is so often absent in conventionally recorded normal BAEPsloO that its absence in pathologic states cannot be used to ascertain its generators. A normal wave VI but a delayed wave VII was described in a single patient who was deaf with "evidence of deep hemispheric damage bilaterally" and absent middle-latency AEP S22 and was interpreted as being consistent with a generator within the auditory radiations. Waves I through VII were all normal in another cortically deaf

510

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

patient with bilateral temporal lobe infarctions.'?' thus indicating that such a generator would not lie within the most distal portion of the radiations. The scalp topography of wave VII also supports its generation near the auditory cortex, predominantly contralarerally.'" Intracranial recordings demonstrate a large nearfield peak at latency ofwaveVII or its simian homologue'" in the region of inferior colliculus in both humans''? (see Fig. 23-26) and monkeys." again most likely originating in the repeated bursts of activity that occur in this structure following a single auditory stimulus. As was the case with wave VI, the common absence of wave VII in normal individuals and the possibility of a mesencephalic contribution (which could produce a wave VII in the face of more rostral damage) mean that assessment of wave VII does not provide clinically useful information about the status of the auditory pathways.

WaveVN The large downward deflection following wave V has a slower time-course than do the preceding upwardpointing peaks; this, most likely, predominantly reflects postsynaptic potentials within brainstem auditory nuclei, primarily the inferior colliculus."

CLINICAL INTERPRETATION OF BAEPs Waves II, IV, VI, and VII are sometimes not identifiable in normal individuals, and their peak latencies display more interindividual variability than do the peak latencies of waves I, III, and V. Amplitude measurements of the individual components are also highly variable

III

IV

across subjects, but the ratio between the amplitude of the IVIV complex and that of wave I has proved to be a clinically useful measure. Therefore, clinical interpretation ofa patient's BAEPs is based predominantly on the presence or absence of waves I, III, and V; and on measurements of the wave I latency (PTT) , the I-V interpeak interval (CIT), the I-III and III-V interpeak intervals, the right-left differences of these values, and the IVIV:I amplitude ratios. Measurement of right-left differences increases test sensitivity because the intersubject variability of these measures is less than that of the absolute component latencies and interpeak intervals from which they were derived (Fig. 23-28).

Normative Data The control data used to derive normal values should have been acquired under the same conditions used to test the patient, including the polarity, rate, and intensity of the stimulus and the filter settings used for data recording, because alteration in any of these parameters can change the latencies or amplitudes of BAEP components. As is the case with all evoked potentials, calculation of normal ranges must take into account the distribution of values in the control population. 12.102.103 A value that is not normally distributed can be transformed into a normal or gaussian distribution to calculate the mean and standard deviation, or a nonparametric statistical measure such as percentiles can be used. The limits of the normal range are typically set at 2.5 or 3 standard deviations from the mean of normally distributed data. A BAEP may be classified as abnormal if any of several measurements is abnormal. The performance of multiple tests increases the possibility of a false-positive result.l'" so the more conservative limit

V

II FIGURE 2J-28. Brainstem auditory evoked potentials (Cz-Ai) to (A) left ear and (B) right ear stimulation in a 23-year-old man with a left-sided acoustic neuroma. The I-III, III-V, and I-V interpeak intervals are within normal limits bilaterally, but the right-left differences of the I-III and I-V interpeak intervals are abnormally large.

A III

IV

o.2J,LVL 8

1 msec

Brainstem Auditory Evoked Potentials

(3 standard deviations) is preferable for evaluation of each measured value. Because the I-V and III-V interpeak intervals are, on average, shorter in women than in men,16,1O!\,I06 some laboratories establish separate normal values for the two genders

Delay Versus Absence of Components Evoked potentials represent the summated activity of large populations of neurons firing in synchrony; the electrical signal produced by a single cell is too small to be seen at the scalp. If the timing of neuronal activity is delayed uniformly across the cell population, a delayed evoked potential component will result. If the delay is nonuniform and the electrical signals are desynchronized (a process called temporal dispersion), the summation may not produce a recognizable evoked potential component. Because the same pathophysiologic process (i.e., delay in neural activity) can cause either delay or absence of a BAEP peak, both of these findings should be interpreted similarly. They both indicate dysfunction, but not necessarily complete loss of activity, in a part of the infratentorial auditory pathways.

Significance of Specific BAEP Abnormalities ABNORMALITIES OF WAVE

511

internal auditory canal alongside the eighth nerve. Cochlear ischemia or infarction may result from compression of the internal auditory artery within the canal or from occlusion of its parent vesse1. 3 1,I08,I09 Thus, wave I may be delayed or absent in patients with basilar artery thrombosis or other posterior circulation vascular disease,89,109-112 acoustic nerve tumors60,l l3 (see Fig. 23-22; Fig. 23-29), or brain death 57-59,1l 4 (see Fig. 23-17) caused by interference with the blood supply of the cochlea. With mild cochlear ischemia, BAEPs may be normal to standard, high-intensity stimuli but become abnormally delayed or absent as the stimulus intensity is lowered. Examination of latency-intensity curves may increase the sensitivity of BAEPs for detecting small acoustic neuromas." The latency-intensity curve may normalize following resection of a small acoustic neuroma (Fig. 23-30), thereby demonstrating the reversibility of the cochlear ischemia.

v

A

I

Because wave I originates in the distal portion of the auditory nerve, abnormalities (delay or absence) of wave I usually reflect peripheral auditory dysfunction, either conductive or cochlear; or pathology involving the most distal portion of the eighth nerve. An audiogram is useful to identity and quantity the degree of hearing loss; a BAEP waveform with a poorly formed or absent wave I but a clear wave V may reflect highfrequency hearing loss. If all BAEP components are absent with the standard stimulus, an attempt should be made to elicit them by using a longer analysis time (typically 15 msec) and a higher stimulus intensity. The increased stimulus artifact caused by the latter may necessitate the use of alternating stimulus polarity. Because peripheral auditory dysfunction is sufficient to cause complete absence of BAEPs, a waveform in which no BAEP components are identifiable provides no information about the status of the brainstem auditory pathways. Cochlear dysfunction may reflect intracranial pathology because the cochlea receives its blood supply from the intracranial circulation via the internal auditory artery. This vessel, which is usually a branch of the anterior inferior cerebellar artery,107 passes through the

B

c

v

7.30 msec

o.2fl. V

D

L

1 msec

E FIGURE ]J-29 • Brainstern auditory evoked potentials recorded (Cz-Ai) in five patients (A to E) with acoustic neuromas following stimulation ipsilateral to the tumor. The stimulus intensity was increased to 85 dB nHL in all cases.

512

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

10.0

t

10.0

t

NR

NR

9.0

9.0 oQ)

()

Q)

en

en

E 8.0 :>. o

E 8.0

~

c: Q)

c

§

7.0

~

>Q)

>

s:

til

7.0

~ 6.0

~ 6.0

s:

5.0

5.0

30

40

50

60

70

80

30

40

50

60

70

80

Stimulus intensity, dB nHL Stimulus intensity, dB nHL FIGURE U·30 • (Left) Latency-intensity curves to left ear (solid line) and right ear (dashed line) stimulation recorded before surgery in a 52-year-old woman with a left-sided intracanalicular acoustic neuroma. The wave V latencyto 70-<1B SL (85-dB nHL) stimulation of the left ear wasnormal, but as the stimulus intensitywasdecreased the latency-intensity curve went outside the normal range (shaded) and then the BAEPs disappeared (NR, no response). The tumor was completelyresected with intraoperative monitoring ofBAEPs and facialnerve function. (Right) After the resection, the patient's latency-intensity curves were normal bilaterally. Her audiogram wasnormal and her speech discrimination score in the left ear was 100 percent. (From Legatt AD, Pedley TA, Emerson RG et al: Normal brain-stem auditory evoked potentials with abnormal latency-intensity studies in patients with acoustic neuromas. Arch Neurol, 45:1326, 1988,with permission.)

ABNORMALITIES OF THE I-III INTERPEAK INTERVAL

Prolongation of the I-III interpeak interval, either absolute (in comparison to normal limits) or relative (excessive right-left difference) in the presence of a prolonged CTT (I-V interpeak interval) reflects an abnormality within the neural auditory pathways between the distal eighth nerve on the stimulated side and the lower pons. Absence of waves III and V carries the same significance. However, the rare absence of wave III in the presence of a clear wave V with a normal I-V interpeak interval should not be interpreted as an abnormality. Interpretation of a prolonged I-III or III-V interpeak interval when the I-V interpeak interval is normal is less clear. Owing to the complexity of the brainstem auditory pathways, the neural activity that generates wave V may not arise from propagation of the same neural activity that generates wave III; the III-V interpeak interval would therefore not actually measure propagation of neural signals between the two sets of anatomic generators. Thus, it is prudent to refrain from interpreting a prolonged III-V interpeak interval as an abnormality in the presence of a normal CIT. In contrast, the I-III interpeak interval does reflect propagation of neural signals from the generator of wave I to the generator(s) of wave III because all auditory afferent activity must pass through the distal eighth

nerve on its way to the brainstem. Thus, it is logical to interpret a prolonged I-III interpeak interval as an abnormality whether or not the corresponding CTT is also prolonged. Abnormality of the I-III interpeak interval is the characteristic BAEP finding in eighth nerve lesions such as acoustic neuromas (see Fig. 23-29), although there may be a simultaneous peripheral abnormality if the internal auditory artery has been compromised. Abnormal I-III interpeak intervals can also result from other processes such as demyelinating disease, brainstem tumors, or vascular lesions of the brainstern. In a patient with unilateral auditory nerve pathology, prolongation of the I-III interpeak interval will be found on stimulation of the ear on the side of the lesion. In patients with unilateral brainstem lesions and unilateral I-III abnormalities (i.e., abnormal BAEPs to stimulation of one ear and normal BAEPs to stimulation of the other), the ear in which stimulation produces the abnormal BAEP waveform is most often ipsilateral to the lesion,74.75,93 but there are rare exceptions.V ABNORMALITIES OF THE III-V INTERPEAK INTERVAL

Prolongation of both the I-V and III-V interpeak intervals, or complete absence of the IVIV complex in the presence of a wave III, reflects an abnormality within

Brainstem Auditory Evoked Potentials

the neural auditory pathways between the lower pons and the mesencephalon. As explained earlier, prolongation of the III-V interpeak interval is best not interpreted as an abnormality if the I-V interpeak interval is normal. Abnormalities in the III-V interpeak interval are seen in a variety of disease processes involving the brainstem, including demyelination, tumor, and vascular disease. If the disease process also involves the lower pOllS or eighth nerve, both the I-III and the III-V interpeak intervals may be prolonged within the same waveform (Fig. 23-31). In patients with unilateral brainstem lesions and unilateral III-V interpeak interval abnormalities, the ear in which stimulation produces the abnormal BAEP waveform is most often, although not always, ipsilateral to the lesion. 74-76.83,93,94 ABNORMALITIES OF THE

IV/V:I

513

recorded as a far-field potential, small displacements of the Cz recording electrode will not substantially attenuate the IVIV complex.

Portion of the Auditory System Assessed by BAEPs All primary afferent auditory neurons synapse in the cochlear nucleus, but the ascending auditory pathways after that structure are extremely complex. The cochlear nucleus projects to several different brainstem nuclei both ipsilaterally and contralaterally, and projections from subsequent auditory nuclei (e.g., the superior olivary complex) also both synapse in and bypass more rostral nuclei. 52- .?4 These various projections do not all serve the same function, and it appears that they do not all contribute equally to the BAEPs.

AMPLITUDE RATIO

An abnormally small IVIV:I amplitude ratio (see Fig. 23-10) reflects dysfunction within the auditory pathways between the distal eighth nerve and the mesencephalon. Because waves I and V are differentially affected by stimulus intensity, the amplitude ratio should be measured at the same intensity used to establish the normative data. Decreasing the stimulus intensity will attenuate wave I more than wave V, and thus inflate the ratio and perhaps cause the abnormality to be missed. Conversely, increasing the stimulus intensity beyond the normal level could increase wave I and give a false-positive result. Suboptimal placement of the Ai recording electrode mav decrease the amplitude of wave I because that component is recorded as a near-field negativity around the stimulated ear. Such placement may artifactually increase the IVIV:I amplitude ratio and perhaps cause an abnormality to be missed. Fortunately, electrode position shifts are unlikely to cause a false-positive result; because the IVIV complex is predominantly

FUNCTIONAL SUBSETS

Because the BAEP peaks are narrow, the neuronal populations that generate them must be extremely well synchronized. Single-unit recordings in the cat'" have shown that BAEPs are generated by a subset of subcortical auditory neurons characterized by the shortest and most consistent latencies, reflecting a high degree of synaptic security that minimizes temporal dispersion. Some neurons within the brainstem auditory system have morphologic and physiologic features that minimize temporal dispersion.J'" Neurons with such characteristics are required for sound localization. An anatomically distinguishable neuronal subsystem for sound localization is present within the brainstem, Jl6 and it appears that these are the neurons that generate the BAEPs. In human patients, BAEP abnormalities are highly correlated with difficulty in correctly lateralizing click pairs with interaural time delays.!17-119 In contrast, pure tone audiometry and speech discrimination may be normal

III

FIGURE 2:1-:11 II Brainstem auditory evoked potentials (BAEPs) (Cz-Ai) to stimulation of each ear in a 35-year-old woman with multiple sclerosis. A, BAEPs to left ear stimulation are normal. B, The I-III and III-V interpeak intervals are both abnormally prolonged following right ear stimulation.

V L

A

O.2 fL II

B

V

1 msec

514

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

in patients with grossly abnormal BAEPs.1l8,120,121 Also, normal BAEPs may be present in patients with infratentoriallesions causing profound impairment of speech perception,91,92 though this finding could conceivably reflect interruption of the auditory pathways rostral to the inferior colliculi. Dipole analysis of BAEPs provides another indication of their relationship to the sound localization system; some components show changes that depend on the position of the sound source. 122

decompression of the brainstem led to an increase in the amplitude of wave V and a decrease in the amplitude of wave I. 126 In a patient with a mesencephalic lesion and an attenuated IV IV complex, enlargement of wave I because of disinhibition will decrease the IVIV:I amplitude ratio even further. This mechanism therefore acts to enhance the sensitivity of the IVIV:I amplitude ratio in the detection of brainstem damage.

Relationship Between BAEPs and Hearing

CROSSED AND UNCROSSED PATHWAYS

Ascending projections from the cochlear nucleus are bilateral but are more extensive contralaterally than ipsilaterally.52-54 Despite this anatomic asymmetry, the BAEPs appear predominantly to reflect activity in the ipsilateral ascending pathways because unilateral brainstem lesions that produce unilateral abnormalities involving either the I-III or the III-V interpeak interval usually do so on stimulation of the ear ipsilateral to the lesion. 74-76,83,93,11O

ROSTROCAUDAL EXTENT

Although some of the BAEP components may receive contributions from the medial geniculate nucleus and auditory radiations, BAEPs as currently performed do not provide reliable, clinically useful assessment of these structures. BAEPs can only be used to assess the status of the ear, auditory nerve, and brainstem auditory pathways up through the level of the mesencephalon. ROLE OF DESCENDING PATHWAYS

Waves I and II may be quite large in patients with rostral brainstem pathology that causes abnormalities of wave V (see Figs. 23-10 and 23-25). This increase in waves I and II probably reflects loss of activity in descending inhibitory pathways originating in or traversing the region of the inferior colliculus.P" Growth of wave I concurrent with loss of wave V has been observed in serial BAEP studies of preterminal patients in whom brain death subsequently developed l23,124 and following placement of a lesion in the rostral brainstem in a monkey.I'" In a patient with an extra-axial tumor compressing the brainstem, removal of the tumor with

CATEGORIZATION OF ABNORMALITIES

BAEP abnormalities can be divided into a "central" pattern, in which the CTT (I-V interpeak interval) is prolonged; and a "peripheral" pattern, in which wave I is delayed. A single waveform may contain both abnormalities (see Fig. 23-22). Clinically, hearing losses are divided into conductive and sensorineural categories. These categories do not correspond to the BAEP categories, however; the relationship among them and the anatomic location of the abnormality are shown in Table 23-1. Although much of the I-III interpeak interval represents the conduction time of the afferent volley along the auditory nerve, abnormalities of the I-III interpeak interval are common in patients with multiple sclerosis (MS). This observation is not surprising in light of the histologic structure of the auditory nerve. Most of the other cranial nerves (except I and II, which are actually central nervous system tracts) are histologically peripheral nerves; their myelin is produced by Schwarm cells, except for a short segment at the nerve root. In contrast, eighth nerve fibers are ensheathed along most of their lengths by central-type myelin, produced by oligodendrocytes; the transition to peripheral-type myelin produced by Schwarm cells occurs near the distal end of the nerve.l'? Therefore, in MS the eighth nerve is vulnerable to demyelination along most of its length, and magnetic resonance imaging has demonstrated involvement of the eighth nerve in MS.128,129 Although it might at first glance seem inappropriate to include the conduction time along the eighth cranial nerve in the "central" conduction time (I-V interpeak interval), it is actually reasonable to do so because most

TABLE 23·1 • Classification of Hearln. Loss Clinical Classification Conductive hearing loss Sensorineural hearing loss

BAEP Classification

Location of Pathology External ormiddle ear Inner ear (cochlea) { Eighth nerve orbrainstem (retrocochlear)

}

Peripheral hearing loss "Central" hearing loss

Brainstem Auditory Evoked Potentials

of that conduction is through fibers with central-type myelin. PERIPHERAL HEARING

Loss

The BAEP waveform to a broad-band click predominantly reflects activity originating in the base of the cochlea. It is mediated by neurons with characteristic frequencies in the 2000- to 400Q..Hz range. 35,36 Thus, signiflcanr high-frequency hearing losses in this frequency band typically produce BAEP abnormalities, whereas BAEPs are relatively insensitive to isolated lowfrequency hearing losses. As described earlier, frequencyspecific stimulus and acoustic masking paradigms can increase the accuracy with which BAEP recordings detect and quantify hearing loss in several frequency ranges. In patients with peripheral auditory dysfunction, high stimulus intensities can partially or completely correct BAEP abnormalities that become obvious at lower stimulus intensities (see Figs. 23-13 and 23-30). Thus, BAEP studies that are used to screen for hearing loss should be performed with multiple stimulus intensities, and latency-intensity curves should be constructed. BAEP studies intended to diagnose conduction abnormalities within the neural auditory pathways can be performed with a single, relatively high (e.g., 60 to 65 dB nHL) stimulus intensity.

515

The possibility of normal hearing in the presence of abnormal BAEPs enhances their clinical utility. It is precisely because they can detect subtle neuronal dysfunction that is not clinically apparen t on the neurologic and audiologic examination that BAEPs (and visual and somatosensory evoked potentials as well) are a valuable addendum to the clinical evaluation of patients with suspected neurologic disease. In patients with auditory symptoms that are suspected of being functional, an abnormal BAEP study demonstrates the existence of pathology within the auditory system. However, a normal study does not prove that the symptoms are psychogenic. The lesion in such a patient may have spared the short-latency, highly synchronized brainstem subsystem that generates the BAEPs but damaged other parts of the auditory system enough to produce symptoms.

BAEP in Specific Neurologic Conditions The final section of this chapter describes BAEP findings in the neurologic conditions for which diagnostic BAEP recordings are most commonly performed in adults. Screening for hearing impairment and evaluation for possible neurodegenerative disorders are important additional indications for BAEP testing in the pediatric population (discussed in Chapter 24).

CENTRAL AUDnoRY PATHWAY ABNORMALITIES

Because BAEPs reflect activity in only a subset of the auditory pathways, dysfunction in another portion of the auditory system can affect hearing without altering the BAEPs. For example, patients with bilateral temporal lobe infarctions involving the auditory cortex may be deaf and yet have completely normal BAEPs.101.130.131 Conversely, BAEPs may be abnormal in patients with brainstem disease who have normal hearing. Unilateral brainstem lesions only rarely cause hearing losSl32 because the ascending projections from each ear are bilateral. In addition, lesions (whether unilateral or bilateral) that affect the short-latency, highly synchronized subsystem involved in sound localization 116 may spare other portions of the brainstem auditory pathways that are sufficient to maintain normal perception. Even if the disease process affects the brainstem auditory pathways globally, abnormal BAEPs do not necessarily imply hearing loss. If BAEPs are present but delayed, the information that a suprathreshold auditorv stimulus has occurred still reaches higher auditory centers; it just does so a few milliseconds later. Thus, the audiogram may be normal. As previously discussed, the absence of a component may reflect temporal dispersion rather than conduction block, so hearing may even be present when there is no identifiable wave V 133

Acoumc NEUROMA

Many different BAEP patterns are seen in patients with acoustic neuromas (see Fig. 23-29). In several large published series,134-137 abnormal BAEPs to standard high-intensity stimuli were found in more than 95 percent of patients with acoustic neuromas. The probability of abnormal BAEPs is less in patients with small (less than I cm) tumors, though some investigators have report sensitivities greater than 90 percent. 1:18,139 In patients with small, intracanalicular tumors in whom BAEPs to standard high-intensity stimuli are normal, latency-intensity studies may reveal abnormal cochlear function resulting from compression of the internal auditory artery" (see Fig. 23-30). Acoustic neuromas typically originate from the distal vestibular nerve at the vestibular ganglion.v''' and the auditory portion of the nerve may be unaffected early in the course of the disease. As an acoustic neuroma enlarges, it begins to compress the auditory nerve. Such compression produces prolongation of the I-III interpeak interval (see Figs. 23-28 and 23-29, A) and eventually complete eradication of wave III and subsequent BAEP components (see Fig. 23-29, D). In some cases, wave V becomes delayed but persists whereas wave III (see Fig. 23-29, B) or waves

516

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

I and III (see Fig, 23-29, C) are lost. Wave II may be relatively spared, a reflection of the contribution to that component originating in the distal eighth nerve (see Fig. 23-22). Simultaneous with I-III interpeak interval abnormalities, wave I may become delayed as the degree of cochlear ischemia increases (see Fig. 23-22). Infarction of the cochlea may cause elimination of all BAEPs (see Fig. 23-29), a common finding with large tumors in patients who have major hearing loss or who are completely deaf. As acoustic neuromas expand within the posterior fossa, they begin to compress the brainstem. Prolongation of the III-V interpeak interval in response to stimulation of the ear contralateral to the tumor (Fig. 23-32) is indicative of brainstem compression by a large tumor. The pattern of BAEP abnormality on the side of the tumor can be used to predict the likelihood of hearing preservation during surgical resection of the tumor.!" OTHER POSTERIOR FOSSA TUMORS

BAEPsare almost always abnormal in brainstem gliomas and other intrinsic brainstem tumors. 22 ,24,25,98,142- 144 Exceptions are tumors that are located entirely within the medulla, neither directly involving the auditory pathways nor affecting them by compression or deformation. Abnormalities in the I-III or III-V interpeak interval, or a combination of both, may be present depending on the location ofthe tumor and the extent to which it affects surrounding brainstem tissue. Serial recordings may show deterioration of the BAEPs because of tumor growth. 22 When a brainstem tumor shrinks in response to treatment, the BAEPs may demonstrate an improvement in conduction within the brainstem auditory pathways.l'" Extra-axial posterior fossa tumors can also produce BAEP abnormalities because of brainstem compres-

III II

sion; the pattern of the abnormality depends on the location of the tumor and on which structures are being compressed. Although the earliest BAEP changes seen with acoustic neuromas (which usually arise distally within the internal auditory canal 140) are typically I-III interpeak interval abnormalities and wave I abnormalities to stimulation ipsilateral to the tumor, posterior fossa tumors arising in other locations may first produce abnormalities in the III-V interpeak interval because of brainstem compression without evidence of eighth nerve compromise. BAEPs may show improvement following decompression of the brainstem because of tumor resection or tumor shrinkage in response to treatment. CEREBROVASCULAR DISEASE

The effect of a brainstem stroke on BAEPs depends on the location and type of the stroke and on the extent of involvement of the auditory pathways, and may be quite limited (see Fig. 23-9). BAEPs are usually normal in patients with medullary infarcts 75,89,145 or lesions confined to the basis pontis, cerebral peduncles, and cerebellar hemispheres'": however, they may be affected if edema or ischemia compromises brainstem function outside the area of infarction. Shimbo and co-workers'" studied BAEPs in 55 patients with posterior circulation strokes and found that they were abnormal in all patients with lesions of the pontine tegmentum or the cerebellar peduncles. BAEPs may have prognostic significance; the presence of BAEP abnormalities is associated with an adverse clinical outcome. 89,110 Abnormalities in the I-III or III-V interpeak interval, or a combination of both, may be seen in patients with brainstem strokes. Posterior circulation vascular disease can also interfere with the blood supply to the cochlea by the internal auditory artery, If a cochlear stroke

IV FIGURE ]J·n. Brainstem auditory evoked potentials (Cz-Ai) to stimulation of the ear contralateral to a large cerebellopontine angle tumor that was compressing the brainstem. The I-V and III-V interpeak intervals are both abnormally prolonged. This tumor was a meningioma, but a large acoustic neuroma could produce the same changes by the same mechanism.

2.20 msec_~-2.50 msec

Brainstem Auditory Evoked Potentials

accompanies the brainstem stroke, all BAEP components will be absent following stimulation of that ear. BAEP abnormalities are found in most patients with vertebrobasilar transient ischemic attacks if they are recorded acutely. 146 These findings tend to resolve over time,146,147 and the yield of abnormalities was lower in studies in which the BAEPs were recorded more than a week after the last transient ischemic attack. 148 Persistent BAEP changes may represent small infarcts that are clinically silent. DEMYELINATING DISEASE

One of the major clinical applications of evoked potential testing is in patients suspected of having a demyelinating disease such as MS. The diagnosis of definite MS requires a demonstration that multiple sites within the nervous system are involved. Evoked potentials can detect conduction slowing and temporal dispersion caused by subclinical demyelination and thus help to demonstrate the involvement of areas besides those accounting for the patient's clinical signs. Following an episode of demyelination, signs and symptoms may clear because of remyelination or the establishment of nonsaltatory conduction, but the conduction velocity usually does not return to baseline. Nonsaltatory conduction in a completely demyelinated fiber is slow.149 In remyelination, the new myelin is usually thinner than the originall'" and the internodes are shorter'P': both of these anatomic changes contribute to a slowed conduction velocity. Thus, evoked potentials can demonstrate a residual abnormality related to a prior symptom that has cleared and confirm that it is related to demyelination. The reported sensitivity of BAEPs in patients with MS varies widely among published studies. The discrepancies are in part caused by differences in the criteria used to determine abnormality of BAEPs, as well as differences in the patient populations being studied (e.g., definite rather than probable MS). As an example of the latter, in one study of 202 patients with the disease,152 abnormal BAEPs were found in approximately 20 percent ofthe patients with possible or probable MS and in 47 percent of the patients with clinically definite MS. The abnormality rate was even higher (57 percent) in a subgroup of patients with definite MS and symptoms or signs of a brainstem lesion. In a meta-analysis by Chiappa'F' covering 1,000 patients with MS (in various categories) reported in the literature, abnormal BAEPs were found in 46 percent. When subdivided by categories of disease, the rate of abnormal BAEPs was 67 percent in patients with definite MS, 41 percent in probable MS, and 30 percent in possible MS. BAEPs may display various patterns of abnormality in MS depending on the areas that are involved by the

517

disease process. Abnormalities in the I-III or III-V interpeak interval, or a combination of both, may be seen. Another pattern seen in MS is an abnormally small IVIV:I amplitude ratio in the presence of normal component latencies and interpeak intervals (see Fig. 23-10). As noted earlier, MS may affect the eighth nerve itself. Involvement of the eighth nerve would be expected to cause a I-III interpeak interval abnormality of the BAEPs. However, prolongations of the PTT (wave I latency) have also been noted in patients with MS.154 Multiple audiologic tests failed to reveal an otologic cause for the BAEP abnormalities in these patients, and the latency prolongations increased or decreased coincident with exacerbation or remission of the patients' clinical symptoms. Thus, it seems most likely that demyelinating disease itself can cause wave I abnormalities, and it should be considered as a possible cause of a PIT prolongation found in a patient being evaluated for possible MS. COMA AND BRAIN DEATH

Coma can be caused by bilateral cerebral hemispheric dysfunction, brainstem dysfunction, or metabolic abnormalities that cause diffuse brain dysfunction. 155 The BAEP findings depend on the cause of the coma and, for structural lesions, on the location of the pathology. BAEPs are typically normal in patients with coma caused entirely by supratentorial disease,156,157 although they may subsequently deteriorate because of transtentorial herniation.123.124,158 Abnormal BAEPs in patients with supratentorial infarctions or hemorrhages are correlated with poor clinical outcornes.P'' Patients who are comatose because of brainstem lesions will have BAEP abnormalities and clinical findings that depend on the location of the pathology.156,157 The finding of normal-appearing BAEPs in a patient whose examination shows widespread brainstem dysfunction should prompt suspicion of a metabolic etiology such as a drug overdose; BAEPs are highly resistant to central nervous system depressant drugs and may be nearly normal in patients who clinically have no brainstem function (Fig. 23-33). There have been many published reports of BAEPs in coma. Most of the patients studied were comatose following head trauma, although brainstem strokes were also included in some series. In several studies,157,160-163 BAEPs were shown to have prognostic value: patients in whom they were normal or who showed relatively mild latency or amplitude abnormalities were more likely to survive with good functional outcomes than were those in whom BAEPs were grossly abnormal or absent. However, BAEPs were not thought to provide useful prognostic information in other reports,59,164,165 predominantly because many patients with normal or

518

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

Cz+

EKG IV-V ~

Fp2-C4

[0.1 IJ-V

I

III

C4-02

VII

Fp1-C3 -~--------­ C3-01

.,

,I

T4-Cz - - - - - - - - - Cz-T3 - - - - - - - - - -

&--...J 15 IJ-V

t

I

2

6

10

14

msec

1 sec

FIGURE 13·JJ • Electroencephalogram (EEG) and brainstem auditory evoked potentials (BAEPs) from a 35-year-old woman who was comatose following a mixed drug overdose with central nervous system depressant drugs and a respiratory arrest. The clinical examination wasconsistent with brain death, and the EEG showed periods of complete suppression of electrical cerebral activity (left) lasting up to 18 minutes. The patient subsequently made a full neurologic recovery, and her EEG became normal. BAEPs recorded during a 13-minute EEG suppression while she was comatose (solid lines) and after she had recovered (dotted line) were both normal. (From StockardlJ, Sharbrough FW: Unique contributions of short-latency auditory and somatosensory evoked potentials to neurologic diagnosis. p. 231. In Desmedt JE led]: Clinical Uses of Cerebral, Brainstem and Spinal Somatosensory Evoked Potentials. Karger, Basel, 1980, with permission.)

nearly normal BAEPs died or had persistent severe neurologic impairment. This is not surprising; the unfavorable outcomes in these patients may have reflected brain damage outside the region of the brainstem generating the BAEPs. All of the studies showed that patients with markedly abnormal BAEPs are likely to have poor neurologic outcomes. These outcomes are attributable to the brainstem damage demonstrated by the BAEPs. Several technical considerations should be kept in mind when performing and interpreting BAEP studies in comatose patients. Endotracheal tubes and positivepressure ventilation interfere with eustachian tube function, and in patients who are chronically intubated, a middle ear effusion or pressure change may develop and cause conductive hearing 10ss,166 which appears as a wave I abnormality in the BAEP recording. Body temperature should be noted when testing comatose patients because hypothermia causes alteration of the BAEPs; both interpeak latencies and the PTT will show reversible increases as the patient's temperature declines.167.168 Hypothermia is commonly encountered (and may be profound) during intraoperative monitoring. Some comatose patients are treated with barbiturate anesthesia. BAEPs can persist with only minor

changes despite high doses of barbiturates in both human patientsl69.170 and animals.'?' High doses of intravenous lidocaine can transiently abolish BAEPs, however. 172 The locked-in syndrome typically results from infarction or demyelination affecting the motor tracts in the ventral pons. Patients with the locked-in syndrome are by definition not in coma because they are aware and can respond with eye movements, but a comatose state may be diagnosed unless the clinician is aware of the possibility of the syndrome and checks for it. BAEPs may be either normall73-175 or abnormaP73.176 in patients with the locked-in syndrome, depending on the extent to which the lesion extends outside the ventral pons and involves the auditory pathways. Typically, the BAEP recording in brain death contains no identifiable components, consists of wave I alone, or contains only a wave I followed by a wave IN57-!i9 (see Fig. 23-17). Reversal of stimulus polarity may be necessary to distinguish an apparent persistent wave I from the cochlear microphonic. Rarely, waves II and lIN may also be present and reflect the contribution to these components from the auditory nerve 59.68 (see Fig. 23-18). The presence ofwave III is not compatible with total cessation of function of the entire brainstem because this component is generated within the lower pons, with a possible

Brainstem Auditory Evoked Potentials

contribution from the cochlear nucleus. Rarely, persistence of wave III, and possibly of wave V as well, demonstrate persistent function ofsome brainstem structures in comatose patients meeting all clinical and EEG criteria for brain death, though none of the patients in whom this has been reported has survived. I 14 The absence of wave I in a comatose patient reflects ear or eighth nerve damage. Loss of the eighth nerve compound action potential eliminates input to the brainstem auditory pathways. Thus, a waveform in which no BAEP components are identifiable does not provide an assessment of the brainstem auditory pathways. Although consistent with brain death, it cannot be used as evidence that the brainstem is nonfunctional.

REFERENCES 1. Campbell KB,Bell I, Bastien C: Evoked potential measures of information processing during natural sleep. P: 89. In Broughton RJ, Ogilvie RD (OOs): Sleep, Arousal, and Performance. Birkhauser, Boston, 1992 2. Litscher G: Continuous brainstem auditory evoked potential monitoring during nocturnal sleep. Int J Neurosci, R2:135, 1995 3. Loughnan BL, Sebel PS, Thomas 0 et al: Evoked potentials following diazepam or fentanyl. Anaesthesia, 42:195, 1987 4. American Academy of Pediatrics Committee on Drugs: Guidelines for monitoring and management of pediatric patients during and after sedation for diagnostic and therapeutic procedures. Pediatrics, 89:1110,1992 5. Banoub M, Tetzlaff JE, Schubert A: Pharmacologic and physiologic influences affecting sensory evoked potentials: implications for perioperative monitoring. Anesthesiology, 99:716, 2003 6. Legatt AD: Mechanisms of intraoperative brainstern auditory evoked potential changes. J Clin Neurophysiol, 19:396, 2002 7. Ruben R], Bordley JE, Lieberman AT: Cochlear potentials in man. Laryngoscope, 71:1141,1961 8. \bshie N, Ohashi T, Suzuki T: Non-surgical recording of auditory nerve action potentials in man. Laryngoscope, 77:76,1967 9. Sohmer H, Feinmesser M: Cochlear action potentials recorded from the external ear in man. Ann Otol Rhinol Laryngol, 76:427, 1967 10. Jewett DL, Romano MN, Williston JS: Human auditory evoked potentials: possible brain stem components detected at the scalp. Science, 167:1517, 1970 11. Jewett DL, Williston JS: Auditory-evoked far fields averaged from the scalp of humans. Brain, 94:681,1971 12. American Electroencephalographic Society: Guideline nine: Guidelines on evoked potentials. J Clin Neurophysiol, 11:40, 1994 13. American Electroencephalographic Society: Guideline eleven: Guidelines for intraoperative monitoring of sensory evoked potentials.J Clin Neurophysiol, 11:77, 1994 14. Chiappa KH: Brain stem auditory evoked potentials: methodology. P: 157. In Chiappa KH (ed): Evoked

519

Potentials in Clinical Medicine. 3rd Ed. Raven Press, New York, 1997 15. Roeser RJ, ClarkJL: Clinical masking. p. 253. In Roeser RJ, Valente M, Hosford-Dunn H (eds): Audiology: Diagnosis. Thieme, New York, 2000 16. Emerson RG, Brooks EB, Parker SW et al: Effects of click polarity on brainstem auditory evoked potentials in normal subjects and patients: unexpected sensitivity of wave V. Ann N YAcad Sci, 388:710,1982 17. Schwartz OM, Morris MD, Spydell JD et al: Influence of click polarity on the brainstern auditory evoked response (BAER) revisited. Electroencephalogr Clin Neurophysiol, 77:445, 1990 18. Legatt AD: Impairment of common-mode rejection by mismatched electrode impedances: quantitative analysis. Am J EEG Technol, 35:296, 1995 19. Grandori F: Field analysis of auditory evoked brainstem potentials. Hearing Res, 21:51, 1986 20. Hughes JR, Fino.IJ: A review of generators of the brainstem auditory evoked potential: contribution of an experimental study. J Clin Neurophysiol, 2:355, 1985 21. Epstein CM: The use of brain stem auditory evoked potentials in the evaluation of the central nervous system. Neurol Clin, 6:771, 1988 22. Stockard], Rossiter VS: Clinical and pathologic correlates of brain stem auditory response abnormalities. Neurology, 27:316, 1977 23. Starr A, Hamilton AE: Correlation between confirmed sites of neurological lesions and abnormalities of farfield auditory brainstem responses. Electroencephalogr Clin Neurophysiol, 41:595, 1976 24. Stockard], StockardJE, Sharbrough FW: Detection and localization of occult lesions with brainstem auditory responses. Mayo Clin Proc, 52:761,1977 25. Starr A, Achor LJ: Auditory brain stem responses in neurological disease. Arch Neurol, 32:761,1975 26. Hecox KE, Cone B, Blaw ME: Brainstem auditory evoked response in the diagnosis of pediatric neurologic diseases. Neurology, 31:832, 1981 27. Campbell KB, Bartoli EA: Human auditory evoked potentials during natural sleep: the early components. Electroencephalogr Clin Neurophysiol, 65:142, 1986 28. Hughes JR, Fino J: Usefulness of piezoelectric earphones in recording the brain stem auditory evoked potentials: a new early deflection. Electroencephalogr Clin Neurophysiol, 48:357, 1980 29. Coats AC, Martin JL: Human auditory nerve action potentials and brain stem evoked responses: Effects of audiogram shape and lesion location. Arch Otolaryngol, 103:605, 1977 30. Arnold SA: The auditory brain stem response. p. 451. In Roeser R], Valente M, Hosford-Dunn H (eds): Audiology: Diagnosis. Thieme, New York, 2000 31. Legatt AD, Pedley TA, Emerson RG et al: Normal brainstem auditory evoked potentials with abnormal latency-intensity studies in patients with acoustic neuromas. Arch Neurol, 45:1326, 1988 32. Klein AJ, Alvarez ED, Cowburn CA: The effects of stimulus rate on detectability of the auditory brain stem response in infants. Ear Hearing, 13:401, 1992

520

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

33. Pratt H, Ben-David Y, Peled R et al: Auditory brain stem evoked potentials: clinical promise of increasing stimulus rate. Electroencephalogr Clin Neurophysiol, 51:80, 1981 34. Tanaka H, Komatsuzaki A, Hentona H: Usefulness of auditory brainstem responses at high stimulus rates in the diagnosis of acoustic neuroma. ORL J Otorhinolaryngol Relat Spec, 58:224,1996 35. van der DriftJFC, Brocaar MP, van Zan ten GA: The relation between the pure tone audiogram and the click auditory brainstern response threshold in cochlear hearing loss. Audiology, 26:1, 1987 36. Gorga MP, Worthington DW, Reiland JK et al: Some comparisons between auditory brain stem response thresholds, latencies, and the pure tone audiogram. Ear Hearing, 6:105, 1985 37. Stapells DR, Picton TW, Durieux-Smith A et al: Thresholds for short-latency auditory-evoked potentials to tones in notched noise in normal-hearing and hearing-impaired subjects. Audiology, 29:262, 1990 38. Davis RL, Britt RH: Analysis of the frequency following response in the cat. Hearing Res, 15:29, 1984 39. Chimento TC, Schreiner CE: The frequency-following response and the onset response: evaluation of frequency specificity using a forward-masking paradigm. Electroencephalogr Clin Neurophysiol, 75:88, 1990 40. Hoormann J, Falkenstein M, Hohnsbein J et al: The human frequency-following response (FFR): normal variability and relation to the click-evoked brainstem response. Hearing Res, 59:179, 1992 41. Huis iri't Veld F, Osterhammel P, Terkildsen K: The frequency selectivity of the 500 Hz frequency following response. Scand Audiol, 6:35, 1977 42. Yamada 0, Kodera K, Hink RF et al: Cochlear distribution of frequency-following response initiation. A highpass masking noise study. Audiology, 18:381, 1979 43. Cone-Wesson B, Ramirez GM: Hearing sensitivity in newborns estimated from ABRs to bone-conducted sounds. J Am Acad Audiol, 8:299, 1997 44. Yang EY, Stuart A, Mencher GT et al: Auditory brain stem responses to air- and bone-conducted clicks in the audiological assessment of at-risk infants. Ear Hearing, 14:175,1993 45. Wackym PA, FirsztJB, Gaggl W et al: Electrophysiologic effects of placing cochlear implant electrodes in a perimodiolar position in young children. Laryngoscope, 114:71,2004 46. Waring MD: Intraoperative electrophysiologic monitoring t.o assist placement of auditory brainstem implant. Ann Otol Rhinol Laryngol, 106: Suppl166, 33, 1995 47. Brown C]: Clinical uses of electrically evoked auditory nerve and brainstem responses. Curr Opin Otolaryngol Head Neck Surg, 11:383,2003 48. Ananthanarayan AK, DurrantJD: On the origin of wave II of the auditory brain stem evoked response. Ear Hearing, 12:174, 1991 49. Hashimoto I, Ishiyama Y, Yoshimoto T et al: Brain-stem auditory evoked potentials recorded directly from human brain stem and thalamus. Brain, 104:841, 1981 50. Achor LJ, Starr A; Auditory brain stem responses in the cat. 1. Intracranial and extracranial recordings. Electroencephalogr Clin Neurophysiol, 48:154,1980

51. Legatt AD, Arezzo JC, Vaughan HG Jr: Short-latency auditory evoked potentials in the monkey. II. Intracranial generators. Electroencephalogr Clin Neurophysiol, 64:53,1986 52. Strominger NL, Strominger AI: Ascending brain stern projections of the anteroventral cochlear nucleus in the rhesus monkey.J Comp Neurol, 143:217, 1971 53. Strominger NL: The origins, course and distribution of the dorsal and intermediate acoustic striae in the rhesus monkey.J Comp Neurol, 147:209, 1973 54. Strominger NL, Nelson LR, Dougherty \\]': Secondorder auditory pathways in the chimpanzee. J Comp Neurol, 172:349, 1977 55. Gersdorff MCR: Simultaneous recordings of human auditory potentials: transtympanic electrocochleography (ECoG) and brainstem-evoked responses (BER). Arch Otorhinolaryngol, 234:15, 1982 56. Chatrian GE, Wirch AL, Edwards KH et al: Cochlear summating potential recorded from the external auditory meatus of normal humans. Amplitude-intensity functions and relationships to auditory nerve compound action potential. Electroencephalogr Clio Neurophysiol, 59:396, 1984 57. Starr A: Auditory brain-stem responses in brain death. Brain, 99:543, 1976 58. Steinhart CM, Weiss IP: Use of brainstem auditory evoked potentials in pediatric brain death. Crit Care Med, 13:560, 1985 59. Goldie WD, Chiappa KH, Young RR et al: Brainstem auditory and short-latency somatosensory evoked responses in brain death. Neurology, 31:248, 1981 60. Raudzens PA, Shetter AG: Intraoperative monitoring of brain-stem auditory evoked potentials. J Neurosurg, 57:341,1982 61. Legatt AD, Pedley TA, Emerson RG et al: Electrophysiological monitoring of seventh and eighth nerve function during surgery for acoustic neuromas. Electroencephalogr Clin Neurophysiol, 64:30P, 1986 62. Chimento TC, Williston JS, Jewett DL et al: The 3-channel Lissajous' trajectory of the auditory brain-stern response. VIII. Isolated frog sciatic nerve in a volume conductor. Electroencephalogr Clin Neurophysiol, 68:380, 1987 63. KimuraJ, Kimura A, Ishida T et al: What determines the latency and amplitude of stationary peaks in far-field recordings? Ann Neurol, 19:479, 1986 64. Legouix JP, Pierson A: Investigations on the sources of whole-nerve action potentials recorded from various places in the guinea pig cochlea. J Acoust Soc Am, 56:1222,1974 65. Portmann M, Cazals Y, Negrevergne M et al: Transtympanic and surface recordings in the diagnosis of retrocochlear disorders. Acta Otolaryngol, 89:362, 1980 66. SpireJP, Dohrmann GJ, Prieto PS: Correlation of brainstem evoked response with direct acoustic nerve potential. p. 159. In CourjonJ, Mauguiere F, Revol M (eds): Clinical Applications of Evoked Potentials in Neurology. Raven Press, New York, 1982 67. Scherg M, Von Cramon D: A new interpretation of the generators of BAEPwavesI-V: Results of a spatia-temporal

Brainstem Auditory Evoked Potentials

dipole model. Electroencephalogr Clin Neurophysiol, 62:290, 1985 68. Stockard 11, Pope-Stockard JE, Sharbrough FW: Brainstem auditory evoked potentials in neurology: Methodology, interpretation, and clinical application. p. 503. In Aminoff MJ (ed): Electrodiagnosis in Clinical Neurology. 3rd Ed. Churchill Livingstone, New York, 1992 69. Curio G, Oppel F, Scherg M: Peripheral origin of BAEP wave II in a case with unilateral pontine pathology: a comparison of intracranial and scalp recordings. Electroencephalogr Clin Neurophysiol, 66:29, 1987 70. Hopf HC: Die Generatoren der AEP-Wellen-II-III. Flogerungen aus Befunden bei definierten Hirnstammprozessen. Akt Neurol, 12:58, 1985 71. Legatt AD, Gordon DS, Flamm ES: BAEP changes during surgery for an intra-pontine cavernous hemangioma: evidence that wave II is composite. Muscle Nerve, 28:S81, 2003 72. Moller AR, jannetta PJ: Compound action potentials recorded intracranially from the auditory nerve in man. Exp Neurol, 74:862, 1981 73. Legatt AD, Arezzo JC, Vaughan HG Jr: The anatomic and physiologic bases of brain stem auditory evoked potentials. Neurol Clin, 6:681, 1988 74. Brown RH Jr, Chiappa KH, Brooks E: Brain stem auditory evoked responses in 22 patients with intrinsic brain stem lesions: implications for clinical interpretations. Electroencephalogr Clin Neurophysiol, 52:38P, 1981 75. Faught E, Oh SJ: Brainstem auditory evoked responses in brainstem infarction. Stroke, 16:701, 1985 76. York DH: Correlation between a unilateral midbrain-pontine lesion and abnormalities of brain-stem auditory evoked potential. Electroencephalogr Clin Neurophysiol, 65:282, 1986 77. Cohen M, Luxon L, Rudge P: Auditory deficits and hearing loss associated with focal brainstem haemorrhage. Scand Audiol, 25:133, 1996 78. Curio G, Oppel F: Intraparenchymatous pontomesencephalic field distribution of brain-stem auditory evoked potentials in man. Electroencephalogr Clin Neurophysiol, 69:259,1988 79. Meller AR, Jannetta PJ: Auditory evoked potentials recorded from the cochlear nucleus and its vicinity in man.J Neurosurg, 59:1013,1983 80, Hirsch BE, DurrantJD, Yetiser S et al: Localizing retrocochlear hearing loss. AmJ Otol, 17:537, 1996 81. Huang C-M, Buchwald .IS: Interpretation of the vertex short-latency acoustic response: a study of single neurons in the brain stem. Brain Res, 137:291, 1977 82. Buchwald JS, Huang C-M: Far-field acoustic response: origins in the cat. Science, 189:382, 1975 83. Scaioli V, Savoiardo M, Bussone Get al: Brain-stem auditory evoked potentials (BAEPs) and magnetic resonance imaging (MRI) in a case of facial myokymia. Electroencephalogr Clin Neurophysiol, 71:153,1988 84. Fischer C, Bognar L, Turjman F et al: Auditory evoked potentials in a patient with a unilateral lesion of the inferior colliculus and medial geniculate body. Electroencephalogr Clin Neurophysiol, 96:261,1995 85. Chu N-S: Brainstem auditory evoked potentials: correlation between CT midbrain-pontine lesion sites and

521

abolition of wave V or the IV-V complex. .J Neurol Sci, 91: 165, 1989 86. Musick FE: Neuroanatomy, neurophysiology, and central auditory assessment. Part III: Corpus callosum and efferent pathways. Ear Hearing, 7:349, 1986 87. M011er AR,.Jannetta PJ: Evoked potentials from the inferior colliculus in man. Electroencephalogr Clin Neurophysiol, 53:612,1982 88. M011er AR, .Jannetta PJ: Interpretation of brainstem auditory evoked potentials: results from intracranial recordings in humans. Scand Audiol, 12: 125, 1983 89. Shimbo Y, Sakata M, Hayano M et al: Topographical relationships between the brainstem auditory and somatosensory evoked potentials and the location of lesions in posterior fossa stroke. Neurol Med Chir, 43:282, 2003 90. Hu C:J, Chan K-Y, Lin T:J et al: Traumatic brainstem deafness with normal brainstem auditory evoked potentials. Neurology, 48:1448, 1997 91. Masuda S, Takeuchi 1(, Tsuruoka H et al: Word deafness after resection of a pineal body tumor in the presence of normal wave latencies of the auditory brain stem response. Ann Otol Rhinol Laryngol, 109: 1107, 2000 92. Meyer B, Kral T, Zentner J: Pure word deafness after resection of a tectal plate glioma with preservation of wave V of brain stem auditory evoked potentials..J Neurol Neurosurg Psychiatry, 61:423,1996 93. Oh S.J, Kuba T, Soyer A et al: Lateralization of brainstem lesions by brainstem auditory evoked potentials. Neurology, 31:14,1981 94. Zanette G, Carteri A, Cusumano S: Reappearance of brain-stem auditory evoked potentials after surgical treatment of a brain-stem hemorrhage: contributions to the question of wave generation. Electroencephalogr Clin Neurophysiol, 77: 140, 1990 95. Legatt AD, Arezzo JC, Vaughan HG .Jr: Short-latency auditory evoked potentials in the monkey. I. Wave shape and surface topography. Electroencephalogr Clin Neurophysiol, 64:41, 1986 96. Velasco M, Velasco F, Almanza X et al: Subcortical correlates of the auditory brain stem potentials in man: bipolar EEG and multiple unit activity and electrical stimulation. Eleetroencephalogr Clin Neurophysiol, 53:133,1982 97. Cohen.J, Harner RN, Sussman NM et al: Origin of late components of the brain-stem auditory evoked response (BAER). Eleetroencephalogr Clin Neurophysiol, 61:22P, 1985 98. Hashimoto 1, Ishiyama Y, Tozuka G: Bilaterally recorded brainstem auditory evoked responses: their asymmetric abnormalities and lesions of the brainstem, Arch Neurol, 36:161,1979 99. Hashimoto I: Auditory evoked potentials from the human midbrain: slow brain stem responses. Electroencephalogr Clin Neurophysiol, 53:652, 1982 100. Fabiani M, Casini A, Marullo T: Wave VII recognizability in brainstem auditory-evoked potentials. Acta Neurol Scand, 67:312, 1983 101. Ozdamar 0, Kraus N, Curry F: Auditory brain stem and middle latency responses in a patient with cortical deafness. Electroencephalogr Clin Neurophysiol, 53:224,1982

522

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

102. Oken BS: Statistics for evoked potentials. p. 565. In Chiappa KH (ed): Evoked Potentials in Clinical Medicine. 3rd Ed. Raven Press, New York, 1997 103. Sand T: Statistical properties of ABR amplitudes and latencies. Implications for computation of reference limits and relation to click phase. Scand Audiol, 19:131, 1990 104. van DijkJG, Jennekens-Schinkel A, Caekebeke JFV et al: What is the validity of an "abnormal" evoked or eventrelated potential in MS?Auditory and visual evoked and event-related potentials in multiple sclerosis patients and normal subjects.J Neurol Sci, 109:11, 1992 105. Chan YW', Woo EK, Hammond SR et al: The interaction between sex and click polarity in brain-stem auditory potentials evoked from control subjects of Oriental and Caucasian origin. Electroencephalogr Clin Neurophysiol, 71:77, 1988 106. Lopez-Escamez JA, Salguero G, Salinero J: Age and sex differences in latencies of waves I, III and V in auditory brains tern response of normal hearing subjects. Acta Otorhinolaryngol Belg, 53:109, 1999 107. Kim HN, Kim \H, Park IY et al: Variability of the surgical anatomy of the neurovascular complex of the cerebellopontine angle. Ann Otol Rhinol Laryngol, 99:288, 1990 108. Eggermont .Il- Don M: Mechanisms of central conduction time prolongation in brainstern auditory evoked potentials. Arch Neurol, 43:116, 1986 109. Ferbert A, Buchner H, Bruckmann H et al: Evoked potentials in basilar artery thrombosis: correlation with clinical and angiographic findings. Electroencephalogr Clin Neurophysiol, 69:136, 1988 110. Stern BJ, Krumholz A, Weiss HD et al: Evaluation of brainstem stroke using brainstem auditory evoked responses. Stroke, 13:705, 1982 Ill. Rao TH, Libman RB: When is isolated vertigo a harbinger of stroke? Ear Nose ThroatJ, 74:33, 1995 112. Yamasoba T, Kikuchi S, Higo R: Deafness associated with vertebrobasilar insufficiency. J Neurol Sci, 187:69, 2001 113. Rosenhall U: Brain stem electrical responses in cerebello-pontine angle tumours. J Laryngol Otol, 95:931, 1981 114. Facco E, Munari M, Gallo F, et al: Role of short latency evoked potentials in the diagnosis of brain death. Clin Neurophysiol, 113:1855, 2002 115. Trussell LO: Synaptic mechanisms for coding timing in auditory neurons. Annu Rev Physiol, 61:477, 1999 116. Evans EF, Nelson PG: On the functional relationship between the dorsal and ventral divisions of the cochlear nucleus of the cat. Exp Brain Res, 17:428, 1973 117. Zerlin S, Mowry HJ: Click lateralization and the auditory brain stem response. Audiology, 19:346, 1980 118. Hausler R, Levine RA: Brain stem auditory evoked potentials are related to interaural time discrimination in patients with multiple sclerosis. Brain Res, 191:589, 1980 119. Levine RA, Gardner JC, Fullerton BC et al: Effects of multiple sclerosis brainstem lesions on sound lateralization and brainstem auditory evoked potentials. Hearing Res, 68:73, 1993

120. Musiek FE, Geurkink NA: Auditory brain stem response and central auditory test findings for patients with brain stem lesions: a preliminary report. Laryngoscope, 92:891, 1982 121. Kaga K, Shin do M, Tanaka Y: Auditory brain stem responses and nonsense monosyllable perception test findings for patients with auditory nerve and brain stem lesions. Laryngoscope, 96:1272, 1986 122. Polyakov A, Pratt H: Electrophysiologic correlates of direction and elevation cues for sound localization in the human brainstem. IntJ Audiol, 42:140, 2003 123. Garcia-Larrea L, Bertrand 0, Artru F et al: Brain-stem monitoring. II. Preterminal BAEP changes observed until brain death in deeply comatose patients. Electroencephalogr Clin Neurophysiol, 68:446, 1987 124. Hall JW III, Mackey-Hargadine JR, Kim EE: Auditory brain-stem response in determination of brain death. Arch Otolaryngol, 111:613, 1985 125. Legatt AD: Short Latency Auditory Evoked Potentials in the Monkey. Doctoral Dissertation, Albert Einstein College of Medicine, 1981 126. Musiek FE, Weider DJ, Mueller RJ: Reversible audiological results in a patient with an extra-axial brainstem tumor. Ear Hearing, 4:169, 1983 127. Skinner HA: Some histologic features of the cranial nerves. Arch Neurol Psychiatry, 25:356, 1931 128. Yamasoba T, Sakai K, Sakurai M: Role of acute cochlear neuritis in sudden hearing loss in multiple sclerosis. JNeurol Sci, 146:179, 1997 129. Bergamaschi R, Romani A, Zappoli F et al: MRI and brainstem auditory evoked potential evidence of eighth cranial nerve involvement in multiple sclerosis. Neurology, 48:270, 1997 130. Ho KJ, Kileny P, Paccioretti D et al: Neurologic, audiologic, and electrophysiologic sequelae of bilateral temporallobe lesions. Arch Neurol, 44:982, 1987 131. Bahls FR, Chatrian GE, Mesher RA et al: A case of persistent cortical deafness: clinical, neurophysiologic, and neuropathologic observations. Neurology, 38:1490,1988 132. Kumar A: Unilateral sensory neural hearing loss: analysis of 200 consecutive cases. Laryngoscope, 96:14, 1986 133. Kraus N, Ozdamar 0, Stein L et al: Absent auditory brain stem response: peripheral hearing loss or brain stem dysfunction? Laryngoscope, 94:400, 1984 134. Barrs DM, Brackmann DE, Olson JE et al: Changing concepts of acoustic neuroma diagnosis. Arch Otolaryngol, Ill: 17, 1985 135. Brackmann DE: A review of acoustic tumors: 1979-1982. Am J Otol, 5:233, 1984 136. House WF, Brackmann DE: Brainstem audiometry in neurotologic diagnosis. Arch Otolaryngol, 105:305, 1979 137. ZappiajJ, O'Connor CA, Wiet RJ et al: Rethinking the use of auditory brainstern response in acoustic neuroma screening. Laryngoscope, 107:1388, 1997 138. El-Kashlan HK, Eisenmann D, Kileny PR: Auditory brain stem response in small acoustic neuromas. Ear Hearing, 21:257, 2000 139. Dornhoffer jt, Helms J, Hoehmann DH: Presentation and diagnosis of small acoustic tumors. Otolaryngol Head Neck Surg, 111:232, 1994

Brainstem Auditory Evoked Potentials

140. Sterkers JM, Perry J, Viala P et al: The origin of acoustic neuromas. Acta Otolaryngologica, 103:427, 1987 141. Matthies C, Samii M: Management of vestibular schwannomas (acoustic neuromas): the value of neurophysiology for evaluation and prediction of auditory function in 420 cases. Neurosurgery, 40:919,1997 142. Jerger J, Neely G, Jerger S: Speech, impedance, and auditory brainstern response audiometry in brainstem tumors: importance of a multiple-test strategy. Arch Otolaryngol, 106:218, 1980 143. Weston PF, Manson JI, Abbott KJ: Auditory brainsternevoked response in childhood brainstern glioma. Childs Nerv Syst, 2:301, 1986 144. ~aer M:Localizing brain stem lesions with brain stem auditory evoked potentials. Acta Neurol Scand, 61:265, 1980 145. Itoh A, Kim YS, Yoshioka K et al: Clinical study ofvestibular-evoked myogenic potentials and auditory brainstem responses in patients with brainstem lesions. Acta Otolaryngol Suppl, 545: 116, 200 I 146. Factor SA, Dentinger MP: Early brain-stem auditory evoked responses in vertebrobasilar transient ischemic attacks. Arch Neurol, 44:544, 1987 147. Rossi L, Amantini A, Bindi A et al: Electrophysiological investigations of the brainstern in the vertebrobasilar reversible attacks. Eur Neurol, 22:371, 1983 14H. Benna P, Bianco C, Costa P et al: Visual evoked potentials and brainstem auditory evoked potentials in migraine and transient ischemic attacks. Cephalalgia, 5:Suppl 2, 53, 1985 149. Moll C, Mourre C, Lazdunski Met al: Increase of sodium channels in demyelinated lesions of multiple sclerosis. Brain Res, 556:311, 1991 150. Rodriguez M, Scheithauer B: Ultrastructure of multiple sclerosis. Ultrastruct Pathol, 18:3, 1994 151. Blakemore WF, Murray JA: Quantitative examination of internodal length of remyelinated nerve fibres in the central nervous system. J Neurol Sci, 49:273, 1981 152. Chiappa KH, Harrison JL, Brooks EB et al: Brainstem auditory evoked responses in 200 patients with multiple sclerosis. Ann Neurol, 7:135, 1980 153. Chiappa KH: Use of evoked potentials for diagnosis of multiple sclerosis. Neurol Clin, 6:861, 1988 154. Rimpel J, Geyer D, Hopf HC: Changes in the blink responses to combined trigeminal, acoustic and visual repetitive stimulation, studied in the human subject. Electroencephalogr Clin Neurophysiol, 54:552, 1982 155. Plum F, Posner JB: The Diagnosis of Stupor and Coma. 3rd Ed. FA Davis, Philadelphia, 1980. 156. Uziel A, Benezech J: Auditory brain-stem responses in comatose patients: relationship with brain-stem reflexes and level of coma. Electroencephalogr Clin Neurophysiol, 45:515, 1978 157. Tsubokawa T, Nishimoto H, Yamamoto T et al: Assessment of brainstem damage by the auditory brainstem responses in acute severe head injury.J Neurol Neurosurg Psychiatry, 43:1005,1980 158. Feblot P, Uziel A: Detection of acoustic neuromas with brainstem auditory evoked potentials: comparison between cochlear and retrocochlear abnormalities. p. 169. In Courjon J, Mauguiere F, Revol M (eds): Clinical Applications of Evoked Potentials in Neurology. Raven Press, New York, 1982

523

159. Haupt WF, Birkmann C, Halber M: Serial evoked potentials and outcome in cerebrovascular critical care patients. J Clin Neurophysiol, 17:326, 2000 160. Karnaze DS, Marshall LF, McCarthy CS et al: Localizing and prognostic value of auditory evoked responses in coma after closed head injury. Neurology, 32:299, 1982 161. Ottaviani F, Almadori G, Calderazzo AB et al: Auditory brain-stem (ABRs) and middle latency auditory responses (MLRs) in the prognosis of severely head-injured patients, Electroencephalogr Clin Neurophysiol, 65:196,1986 162. Facco E, Martini A, Zuccarello M et al: Is the auditory brain-stem response (ABR) effective in the assessment of post-traumatic coma? Electroencephalogr Clin Neurophysiol, 62:332, 1985 163. Krieger D, Adams H-P, Rieke K et al: Prospective evaluation of the prognostic significance of evoked potentials in acute basilar occlusion. Crit Care Med, 21:1169, 1993 164. Cant BR, Hume AL, Judson JA et al: The assessment of severe head injury by short-latency somatosensory and brain-stem auditory evoked potentials. Electroencephalogr Clin Neurophysiol, 65:188, 1986 165. Anderson DC, Bundli S, Rockswold GL: Multimodality EPs in closed head trauma. Arch Neurol, 41:369,1984 166. Starr A: Clinical relevance of brain stem auditory evoked potentials in brain stem disorders in man. p. 45. In DesmedtJE (ed): Auditory Evoked Potentials in Man. Psychopharmacology Correlates of Evoked Potentials. Karger, Basel, 1977 167. Markand ON, Lee BI, Warren C et al: Effects of hypothermia on brainstem auditory evoked potentials in humans. Ann Neurol, 22:507, 1987 168. Rodriguez RA, Edmonds HL Jr, Auden SM et al: Auditory brainstem evoked responses and temperature monitoring during pediatric cardiopulmonary bypass. CanJ Anaesth, 46:832,1999 169. Drummond JC, Todd MM, U HS: The effect of high dose sodium thiopental on brain stem auditory and median nerve somatosensory evoked responses in man. Anesthesiology, 63:249, 1985 170. Newlon PG, Greenberg RP, Enas GG et al: Effects of therapeutic pentobarbital coma on multimodality evoked potentials recorded from severely head-injured patients. Neurosurgery, 12:613, 1983 171. Sutton LN, Frewen T, Marsh Ret al: The effects of deep barbiturate coma on multimodality evoked potentials. J Neurosurg, 57:178, 1982 172. Garcia-Larrea L, Artu F, Bertrand a et al: Transient drug-induced abolition of BAEPs in coma. Neurology, 38:1487,1988 173. Bassetti C, Mathis J, Hess CW: Multimodal electrophysiological studies including motor evoked potentials in patients with locked-in syndrome: report of six patients. J Neurol Neurosurg Psychiatry, 57:1403,1994 174. Landi A, Fornezza U, De Luca G et al: Brain stem and motor evoked responses in "locked-in" syndrome. .J Neurosurg Sci, 38:123, 1994 175. Budak F,Ilhan A, Ozmenoglu M et al: Locked-in syndrome: a case report. Clin Electroencephalogr, 25:40, 1994 176. Seales DM, Torkelson RD, Shuman RM et al: Abnormal brainstem auditory evoked potentials and neuropathology in "locked-in" syndrome. Neurology, 31:893,1981

CHAPTER

24

Brainstem Auditory Evoked Potentials in Infants and Children TERENCE W. PICTON, MARGOT J. TAYLOR, and ANDREE DURIEUX-SMITH

BAEP Evaluation of Specific Hearing Disorders Treatment of Sensorineural Hearing Loss Neurology of Hearing

SPECIAL TECHNICAL CONSIDERATIONS IN CHILDHOOD Sleep Stimuli Recording DEVELOPMENTAL CHANGES Normal Neonates Premature Infants Postnatal Maturation EVALUATION OF HEARING Hearing Impairment in Infancy Newborn Hearing Assessment with BAEPs Evoked Potential Audiometry

Brainstem auditory evoked potentials (BAEPs) are used in pediatrics to detect and measure hearing loss in children who cannot be tested behaviorally and to evaluate the auditory brainstem pathways in children who may have neurologic problems. Recording pediatric BAEPs requires close cooperation between audiologists and neurologists because it is impossible to interpret these responses correctly without paying careful attention to both the ear and the brain.

SPECIAL TECHNICAL CONSIDERATIONS IN CHILDHOOD

NEUROLOGIC APPLICATIONS Tumors Developmental Disorders Myelin Disorders Metabolic and Degenerative Disorders Anoxia Head Injury, Coma, and Brain Death Cognitive Disorders CONCLUDING COMMENTS

than 18 months of age will usually be quiet for the procedure, provided that it is introduced to them slowly and gently. If necessary, sleep may be assisted by oral diazepam (0.2 to 0.3 mg/kg) or chloral hydrate (30 to 50 mg/kg), but only if adequate facilities for resuscitation are immediately available. General anesthesia is occasionally necessary in extremely disturbed children and those with severe involuntary movement disorders. Most anesthetics cause small dose-related changes in the latency and amplitude of the responses but do not affect the detection of the different waves or the assessment of thresholds.P

Sleep

Stimuli

BAEPs in children are smaller than in adults, and the background electrical noise from the electroencephalogram (EEG) and scalp muscles is often higher. Whenever possible, infants and young children should be tested while asleep because sleep reduces both EEG and muscle artifact. Newborn infants quickly fall asleep after feeding. Most older infants will sleep through a l-hour recording session if they are awakened early on the day of the test. After the electrodes are applied, the infant is left with the mother in a darkened, soundattenuated room to feed and fall asleep. Children older

BAEPs are most commonly evoked by clicks. Each recording should be made with clicks of only one polarity. In patients with high-frequency hearing loss, latency differences between the responses to condensation and rarefaction clicks may distort the responses when using clicks with alternating polarity.' Recording separate responses to condensation and rarefaction clicks allows brainstem responses to be distinguished from stimulus artifacts and cochlear microphonics." For most purposes, responses are recorded using monaural stimuli. If binaural stimuli are used, the examiner

525

526

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

should be aware that the binaural response approximates the sum of the responses from each ear. A unilateral hearing loss may thus not be noticed. If the threshold for one ear differs by more than 40 dB from the other ear, contralateral masking to prevent cross hearing is necessary. Some systems automatically present masking in the contralateral ear. The intensity of the click is calibrated relative to normal hearing thresholds (nHL) for adults. The acoustic intensity of 0 dB nHL is approximately 30 dB peak-topeak equivalent sound pressure level (peSPL).5 The earphones should be placed or held so that they do not slip off the ear or occlude the ear canal. Some BAEP systems use an earphone with a light circumaural seal and a transparent cover that allows the examiner to observe the ear canal." Insert earphones are perhaps the best way of attenuating the stimulus artifact and preventing collapse of the ear canal." The stimuli are presented at rates between 10 and 100 Hz. Wave I is best recorded at rates between 10 and 20 Hz. Wave V is most efficiently recorded at rapid rates because it does not decrease in amplitude sufficiently to offset the more rapid decline in the levels of averaged noise at rapid rates. Maximum length sequences provide an effective way of presenting stimuli at rates greater than 100 Hz without running into problems of overlap.S? Clicks contain energy at all audiometric frequencies, and simple click-evoked BAEPs cannot assess auditory thresholds at different frequencies. Thresholds for the click-evoked BAEP are most closely related to behavioral thresholds between 2,000 and 4,000 HZ.1O However, children with high-frequency hearing losses and normal hearing at 1,000 Hz can still show BAEPs (albeit delayed) down to nearly normal thresholds. Furthermore, although absent click-evoked potentials usually indicate severe high-frequency loss, they provide very little information about the extent of lowfrequency hearing. Procedures for assessing frequencyspecific thresholds use two main techniques: (1) limiting the responsiveness of the auditory system by masking; and (2) concentrating the acoustic energy in the stimuli by using tones.'! The derived-response technique records the responses to clicks in high-pass masking noise. Different high-pass cutoff frequencies are used for the masking noise, and sequential subtraction of the recordings yields the derived responses to the frequencies between the filter settings. This technique has been used to evaluate the hearing of infants and the development of the cochlea.P Brief tones are another way of evaluating frequencyspecific thresholds. The most commonly used tones have linear rise and fall times of 2 cycles each and a plateau of I cycle. Brief tones have significant energy in frequencies other than their nominal frequency, and

BAEPs to brief tones may be evoked by this spectral splatter. Notched noise or broad-band masking may be used to limit responsiveness to the frequencies of the tone and may provide important information about hearing in infancy and childhood.P'!' Continuous tones with sinusoidal amplitude modulation have energy limited to the carrier frequency and two side bands separated from the carrier by the frequency of modulation. These frequency-specific stimuli can evoke steady-state responses at the frequency of the modulation.!" In adults and older children, prominent steady-state responses can be recorded at frequencies near 40 Hz. Unfortunately, these responses are attenuated by sleep and are difficult to record in infants and young children. Responses at frequencies of 70 to 100 Hz are probably the steady-state version of the transient BAEP and can be reliably recorded in sleeping infants.l" Multiple responses can be recorded simultaneously if each stimulus is characterized by its own signature modulation frequency.'? BAEP thresholds to bone-conducted stimuli are important when assessing middle ear function. The bone vibrator should be positioned over the temporal bone posterior and superior to the ear and held in place with a pressure of 400 to 500 g.l8,19 The skull bones of infants are not tightly meshed together, and bone-conducted stimuli are not equally distributed through the skull. BAEPs recorded between the vertex and the ipsilateral mastoid in infants are larger and have slightly lower thresholds than those recorded between the vertex and the contralateral mastoid.i" This asymmetry may aid in the evaluation of whether inner ear function differs between the ears in infants with bilateral conductive hearing loss.

Recording BAEPs are recorded between surface electrodes placed on the vertex and the ipsilateral mastoid or earlobe. If the scalp electrode is located on the forehead, wave V will be smaller but still recognizable. In young infants the vertex electrode is best located in the midline anterior to the fontanelle. In newborn infants, the electrodes should be attached to the scalp with nonirritating tape and saline jelly. In older children, an adhesive paste or collodion and gauze may be used. Collodion should not be used in neonates because of skin sensitivity. The authors recommend recording simultaneously from two channels: (1) vertex to ipsilateral mastoid; and (2) vertex to contralateral mastoid. It is then unnecessary to switch electrodes when the baby rolls over during sleep and the examiner decides to stimulate the upper ear rather than risk wakefulness by rolling the baby back. A mastoid-to-mastoid recording, which can sometimes make wave I easier to recognize,

Brainstem Auditory Evoked Potentials in Infants and Children

can be obtained by subtracting one recording from the other. Because the infant's response is slower than that of the adult, the recording sweep should be longer (15 or 20 msec) and the low-frequency cutoff of the filters should be lower (20 to 30 Hz) than in adults. 21.22 This adjustment is particularly important when responses to low-intensity stimuli are being examined. Settings similar to those for adults (10-msec sweeps and 100- to 3,000-Hz filter bandpass) are used in much of the literature and are acceptable for older children. Normative data for latencies and amplitudes are specific to the filter settings.

DEVELOPMENTAL CHANGES Normal Neonates The BAEP of infants is probably best introduced through the response of the normal newborn (Fig. 24-1). In general, the neonatal BAEP is about onehalf the size of the adult BAEP.23 Wave V is particularly small, and the average V/1 amplitude ratio in the

Adult

+ ++ Neonate

-~ +O.5 IJ-V

J

' - -_ _- - L -'--____ 15 msec 70 dB 11/sec FIGURE 14·1 l1li Typical brainstem auditory evoked potentials from a normal adult and a normal neonate (40 weeks' gestational age). In the adult the responses were recorded between electrodes on the vertex and the ipsilateral mastoid. In the newborn infant the responses were recorded between an electrode just anterior to the anterior fontanelle and an electrode on the ipsilateral mastoid. The polarity convention of this and all subsequent figures is that positivity at the scalp relative to the mastoid is represented by an upward deflection. For both subjects the stimulus was a 70-dB nHL rarefaction click presented at a rate of 11 per second. Each tracing represents the average of 2,000 individual responses. The stimulus artifacts recorded in the first 0.5 msec of the sweep have been removed from the recordings. The arrows indicate waves I, III, and V for both the adult and the neonatal response.

527

normal newborn is about 1.5, whereas in adults it is over 2. These numbers are based on a low-frequency cutoff of 20 or 30 Hz; raising this setting will decrease the V/1 ratio. The V/1 ratio may decrease with increasing stimulus intensity because wave I increases more than wave V. The general morphology of the response differs in several other ways: wave I is often doublepeaked; a prominent negative wave follows wave I; the negative wave after wave III is small. The literature on normal latencies for the neonatal BAEP is extensive. 23-26 Exact values differ from study to study because of differences in filter settings, stimulus polarity and intensity, and rate of presentation. All latencies increase with decreasing intensity, and the I-V latency increases with increasing rates of stimulation. A rarefaction click usually evokes an earlier wave I than a condensation click. Because the wave I of newborns is often double-peaked, latency values involving this wave will differ with the rules used for its measurement. Benchmark latencies for the BAEP in neonates are a I-V latency of 5 msec and a wave V latency of 7 msec at 70 dB nHL. The scalp distribution of the neonatal BAEP is very different from that of the adult response.F''" This difference may be related to the different orientations of the neonatal auditory pathways or to incomplete myelination in parts of the pathway. In the neonate, wave I is often larger on a mastoid-to-mastoid recording than on a vertex-to-mastoid recording. The dipoles underlying waves III and V are also more laterally oriented. On a recording between the vertex and the contralateral mastoid, waves III and V are small and appear to have opposite polarity to those recorded on the ipsilateral montage (Fig. 24-2). The small size of the contralateral recording in the neonate makes the response very difficult to recognize at low intensities, and when assessing thresholds care must be taken to ensure that the electrode montage is correct. The scalp distribution of the BAEP becomes similar to that of the adult BAEP by the end of the first year of life. 29 The scalp distribution of the BAEP can also be represented by using vector plots and 3-channel Lissajous' trajectories." The BAEP can be recorded in normal newborn infants with stimulus intensities as low as 30 dB nHL provided that the infant is asleep, sufficient averaging is performed, and the acoustic noise in the environment is low. If the recording parameters are optimized, responses can be recorded with stimuli as low as 10 dB nHL.6,14 Figure 24-3 shows BAEPs recorded near threshold in a normal newborn infant.

Premature Infants BAEPs can be recorded in premature infants as early as 26 weeks' gestational age. 3J.32 In these infants they are

528

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

70 dB 11/8ec

30 dB 61/8ec

-r-.: • •

Fz-Mi

+ 0.5 !LV ] '--_---J._ _- ' -_ _. 15 msec FIGURE 24·2 • Ipsilateral and contralateral recording mon-

tages in the neonatal brainstem auditory evoked potential (BAEP). These responses were recorded from a newborn infant at 38 weeks' gestational age. On the left are responses to 7O-dB nHL clicks presented at a rate of 11Hz, and on the right are responses to 3O-dB nHL clicks presented at a rate of 61 Hz. The upper tracings represent recordings between the forehead and the mastoid ipsilateral to stimulation, whereas the lower tracings represent recordings between the forehead and the mastoid contralateral to stimulation. In the BAEP recorded at high intensity with the ipsilateral montage, waves I, III, and V are clearly visible (filled triangles). On the contralateral montage, waves III and V are much smaller in amplitude and have an opposite polarity (open triangles). At the lower intensity, waveV is recognizable on the ipsilateral montage (filled triangle), and there is no clear response on the contralateral montage.

evoked only by stimuli presented at high intensity and slow rates. The amplitude of the response, particularly wave V, is smaller than that in full-term neonates. The latencies of all the components of the response decrease with increasing conceptional age. Wave V shows a greater change with age than does wave I, and thus the I-V interpeak latency also decreases with age. From 36 to 40 weeks' gestational age the latency of wave I decreases by about 0.1 msec, whereas the latency of wave V decreases by 0.4 msec. The increase in latency and decrease in amplitude of wave V that occur with increasing stimulus rate are more marked in premature infants than in full-term infants. 33•M Figure 24-4 shows the BAEPs from a normal premature infant and a term infant. The BAEPs of a premature infant and a full-term infant who have reached the same conceptional age are similar in morphology and interpeak latencies. However, the absolute latencies of specific peaks in the BAEP of premature infants are longer by about 0.3 msec." This difference persists over the first 2 years of life. It might be caused by the greater incidence of otitis media in babies born prematurely. In premature infants who are small for gestational age, wave V is shorter in latency than in normal premature infants, but the latency does not change as much with increasing age. 36

Premature (32 wk) Newborn screening

4-month follow-up

-r-; <: -r-:»

11/sec

.. ..~

70

~

60

~

40

~

30

~-.".

20

..

..

.. 61/sec"'~

. ~

lOdB

+0.5 fLV

1

~_-'-_""",,_---.J

. ...~ ..

... ~

+0.5 fLY

~

10~

dB nHL

Term (41 wk)

15 msec

FIGURE 24·:5 • Threshold studies during infancy. On the left

are shown the responses recorded in a newborn infant at 37 weeks' gestational age. On the right are shown follow-up results taken 4 months later. Responses were evoked by rarefaction clicks presented at a rate of 61 Hz. For the neonatal recordings, wave V is clearly recognizable (triangle) down to 30 dB nHL. Four months later, the responses can be recognized down to 20 dB nHL.

~............~.........._] 15 msec

.........

FIGURE 24-4 • Brainstem auditory evoked potentials (BAEPs) in premature infants. On the left are shown BAEPs recorded from a premature infant (32 weeks' gestational age), and on the right are shown responses from a term infant (41 weeks' gestational age). The upper tracings represent the responses to 70-<1B nHL clicks presented at a rate of 11 Hz, and the lower tracings represent responses recorded at a stimulus rate of 61 Hz. The response of the premature infan t is smaller and has longer latencies than the response of the term infant; increasing the stimulus rate to 61 Hz causes a greater change in the latency and amplitude of wave V (triangles).

Brainstem Auditory Evoked Potentials in Infants and Children

Postnatal Maturation BAEPs recorded from infants within the first few hours after birth differ significantly from those recorded a day after birth.:J7.~H Wave I is delayed just after birth by about 0.8 msec, probably because of residual fluid in the middle ear. However, there is also a significant difference in the I-V interval of 0.2 msec, which indicates some central changes over the first day of life. In normal children the BAEP to monaural stimulation matures to the adult pattern by the age of about 3 years.2~.~9-44 After this age, the child's BAEP has latencies that are similar to those of the adult, although some subtle differences remain. Table 24-1 presents normative developmental data from the authors' laboratories in Ottawa and Toronto. These data are probably sufficient for most clinical applications. The different components of the BAEP mature differently. Wave I latency reaches adult values by 2 mon ths, Waves 111 and V show a rapid decrease in latency over the first several months and then a slower decrease to reach adult normal values at the end of the third year. The amplitude of wave V shows a marked increase after 6 months but does not reach adult values until about 5 years. The developmental changes in the latencies of the BAEP can be interpreted in terms of the differential maturation of axonal conduction time and synaptic transmission, and by the changes in the length of the auditory pathways.v" Binaural interaction in the BAEP is present at birth,"? with latencies and amplitudes that are the same relative to the monaural response as those in adults. This finding suggests that the binaural systems of the brainstem probably mature similarly to the monaural systems. The threshold for detecting the BAEP decreases by about 10 dB in the first 3 months of life and by a further 5 dB by the end of the first year. 2:J.4H These changes are probably related to several factors including the resolution of neonatal conductive hearing loss

and maturation of the cochlea and brainstem. Figure 24-3 illustrates this improvement in threshold. Gender causes significant differences in BAEPs, but there is some controversy in the literature about when these differences become significant. Some studies of neonates have found a significantly shorter latency of wave V in female babies,36.49 whereas other studies have found no differences.P-" More consistent latency differences between males and females show up somewhere between a few months of age and puberty.23.50-53 The major differences involve the I-Ill and I-V interpeak latencies. These effects are in part related to differences in the length of the auditory pathway, particularly the auditory nerve. The amplitude of wave V is larger in girls,5~ with this difference becoming apparent by 3 years of age. 2~

EVALUATION OF HEARING Hearing Impairment in Infancy Between 1 and 3 per 1,000 children are born with a sensorineural hearing loss that requires trearment.Pv"? The cost of hearing impairment to the individual and to society is a result of the decreased communication ability of the hearing-impaired individual. Even mild hearing loss can significantly impede the normal development of speech and language. Conductive hearing losses can be treated medically or surgically; sensorineural hearing losses can be treated with hearing aids, cochlear implants, and communication development training (e.g., "aural habilitation"). A major determinant of how well a hearing-impaired individual ultimately communicates is the age at which the impairment is detected and treatment instituted. It is therefore essential to identity and assess hearingimpaired infants as soon as possible.t" There are clearly defined risk factors for hearing loss in the newborn period (Table 24-2) .59 Most of the risk factors for hearing loss (except family history) are

TABLE 24·1 • Normative Values for Pediatric BAEPS· Latency (msec) Age Premature (36-wk gestational age) Full-term neonate 6wk 3 mo 6mo 12 mo 2 yr

2.1 (0.3) 2.0 (0.3) 1.8 (0.2) 1.7 (0.2) 1.7 (0.2) 1.7 (0.2) 1.7 (0.2)

529

Amplitude (jJ.V)

III

V

I-V

V

VJI

5.0 (0.4) 4.B (0.3) 4.4 (0.3) 4.3 (0.3) 4.1 (0.3) 4.0 (0.3) 3.8 (0.2)

7.4 (0.4) 7.0 (0.3) 6.6 (0.3) 6.4 (0.3) 6.2 (0.3) 6.0 (0.3) 5.7 (0.2)

5.3 (0.3) 5.0 (0.3) 4.9 (0.3) 4.7 (0.3) 4.6 (0.3) 4.3 (0.2) 4.0 (0.2)

0.4 0.5 0.5 0.6 0.6 0.6 0.6

1.5 1.6 1.6 1.6 1.8 2.0 2.2

'These values arefor 8AEPs elicited by70-
530

ElEGRODIAGNOSIS INCLINICAL NEUROLOGY

TABLE 24·2 • Risk Factors for Hearing Loss inNeonates 1. Family history ofhereditary childhood hearing loss 2. Congenital perinatal infection (e.g., cytomegalovirus, rubella, herpes,

toxoplasmosis, syphilis) 3. Anatomic malformations involving the head orneck

4. Birth weight <1,500 g 5. Hyperbilirubinemia atlevels requiring exchange transfusion 6. ototoxic medications 7. Bacterial meningitis 8. Apgar scores of0-4 at1min or0-6at5min 9. Mechanical ventilation forlonger than 5days 10. Stigmata ofsyndromes known toinclude sensorineural hearing loss (e.g., Waardenburg's orUsher's syndrome)

present in infants treated in a neonatal intensive care unit (NICU), and between 30 and 90 percent of children with significant hearing loss were admitted to an NICU.56.60.61 One approach to the detection of hearing impairment has therefore been to record BAEPs in babies at risk for hearing impairment and/or babies being discharged from an NICU. 62-64 Unfortunately, between 20 and 50 percent of infants with significant hearing loss do not show any risk factors. 56.60,61 A consensus panel of the National Institutes of Health in the United States therefore recommended universal screening of all newborn infants for hearing impairrnent/" Behavioral testing is not an effective approach to the evaluation of hearing in the first few months of life. Some babies have reflex responses to sound, but sleepy or sick babies may not. Furthermore, because behavioral responses are mainly elicited by loud sounds, hearing losses of mild or moderate degree may not show any abnormality on behavioral testing. BAEPs and otoacoustic emissions are objective techniques for evaluating auditory function in infants. Two main approaches are now used for newborn hearing screening. 55,66 One approach is to screen all infants with otoacoustic emissions and then to use BAEPs to assess infants who lack otoacoustic emissions. Another approach is to use automatic techniques for recording BAEPs to low-intensity sounds in the initial screening test. Otoacoustic emissions are sounds that originate in the cochlea and can be recorded in the external ear canal using a sensitive microphone. They can be evoked either by presenting a brief stimulus (click or tone pip) and recording the acoustic energy in the external ear canal over the subsequent 20 msec; or by presenting two tones and recording the acoustic energy at the distortion products of these tones. Because they take less time than BAEPs, otoacoustic emissions are widely used for hearing screening. 67-
ciently normal that treatment is not indicated. If they are absent, hearing loss is present, but the severity of this loss is not indicated. Infants who fail this test should be assessed with BAEPs to confirm a persistent hearing loss and to assess its severity. Screening with otoacoustic emissions will miss infants with auditory neuropathy (see p. 534), who usually have normal cochlear function. Hearing impairment can develop during the first few years of life. Any parental concern about the hearing of a child should lead to an audiologic evaluation. Infants with certain congenital disorders such as rubella or cytomegalovirus may show a deterioration in hearing after birth. All children exposed to postnatal risk factors (e.g., meningitis, encephalitis, skull fracture, and ototoxic medication) should have their hearing tested. It is also important to monitor the auditory status of infants who have risk factors for chronic middle ear effusions (e.g., prematurity, Down syndrome, cleft palate, unilateral atresia of the external auditory canal, or other craniofacial malformations). All children with delayed intellectual development should also be tested because hearing loss could explain or exacerbate their retardation.

Newborn Hearing Assessment with BAEPs The most studied BAEP in newborns is that evoked by clicks. This provides a quick assessment of the general hearing threshold and evaluates the neurologic integrity of the brainstem auditory pathways. This evaluation can be used as an initial hearing test (especially for infants in the neonatal intensive care unit), as a follow-up test after screening with otoacoustic emissions, and as the initial step in a full objective audiometric evaluation. Both BAEPs and otoacoustic emissions must be recorded in any infant at risk for auditory neuropathy (see p. 534). The authors' basic approach 25.64 evaluates the BAEP to monaurally presented clicks. Hearing is assessed by using 30-dB nHL clicks presented at a rate of 61 Hz; 4,000 responses are averaged. If replicated responses are not recognizable, the testing is continued at higher intensities until an auditory threshold is obtained. Replicate BAEPs are also obtained for each ear by using 70-dB nHL clicks presen ted monaurally at a rate of 11 Hz. These responses allow assessment of neurologic function as well as hearing. Any infant not showing responses at 30 dB is retested at the age of 3 to 5 months. By that time, normal thresholds have developed in many of these infants, perhaps because a perinatal conductive loss has resolved. The authors initially hesitated between using 30 and 40 dB nHL as the screening level. Screening at 40 dB greatly reduces

Brainstem Auditory Evoked Potentials in Infants and Children

the number of follow-up tests and should not miss an infant with a sensorineural hearing loss requiring amplification. However, approximately 40 percent of the babies with a 40-dB threshold in the newborn period showed persistent conductive losses on followup. Because it was considered important to monitor these infants, the decision was made to screen at 30 dB nHL. Screening at 3 months is more accurate than screening in the neonatal period because transient neonatal conductive losses have resolved. However, it may still be more efficient to assess infants while they are in the hospital in case they do not return for follow-up.

Figure 24-5 shows the BAEPs obtained on screening and at follow-up for an infant with bilateral sensorineural hearing loss. A large study has evaluated a similar protocol in over 7,000 infants.f? This study used more accurate ways to calibrate sound levels and more precise measures of the signal-to-noise ratio to determine when to decide whether a response was present. It was found that more than 99 percent of newborn infants could be tested, and that more than 90 percent of these passed the test (at 30 dB nHL). The rest were followed up with more extensive audiometric testing, leading to the identification of

Left

Right

'"

"'~70 '" .. ~60

'"

...

~

'"

..~

Neonate .-- ~

50

40

'" ~60

'" '"

~50

3 months 40~

30

'"

60"~ 5 months

531

'"

50~ 40~

dB nHL

+ 0.5 IJ-V '--_---"_ _--'-____] 15 msec

FIGURE 14-5 • Early identification of hearing impairment using the brainstem auditory evoked

potential. This child was born after 38 weeks' gestation with a birth weight of 3,210 g. The postnatal course was complicated by perinatal asphyxia and seizures. The initial screening test done on discharge from the neonatal intensive care unit is shown in the upper part of the figure. Responses to clicks at 61 Hz were recognizable only down to 60 dB nHL (triangles), which is indicative of bilateral moderate hearing impairment. Follow-up studies shown in the middle and lower part of the figure confirmed this impairment with thresholds between 50 and 60 dB nHL. After the second follow-up test, the child was fitted with a hearing aid on the right ear and enrolled in the aural habilitation program. By the age of 2 years the child was able to comprehend well and to speak two-word phrases with good voice quality and intonation.

532

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

sensorineural hearing loss in about 2 percent of the tested population.57 BAEP thresholds in the neonatal period closely predict the high-frequency thresholds of the behavioral audiogram obtained when the child is old enough to provide accurate behavioral thresholds. One study defined significant hearing loss on follow-up as a sensorineural loss with thresholds above 40 dB nHL and found that abnormal BAEP results (neonatal BAEP thresholds above 40 dB nHL) correctly detected 98 percent of the hearing-impaired babies (sensitivity), with only 4 percent false-positive results (96 percent specificityj.?? We defined the target population as those requiring hearing aids and found a sensitivity of 86 percent and specificity of 100 percent." Between 10 and 30 percent of babies in an NICU have a transient conductive hearing loss that resolves within the first few months of life.62,64,71,72 Furthermore, conductive hearing losses may occur by the time of follow-up in a baby who was normal in the neonatal period. Infants who have conductive losses in the neonatal period are more likely to have repeat otitis later than are those with normal neonatal hearing. Some children with normal thresholds for clicks in the newborn period show significant hearing loss at fol10w-up.71 Passing a neonatal screening test is no guarantee that the child's hearing is normal or that it will continue to be normal. In some infants, sensorineural hearing loss may develop after birth. However, most of the hearing-impaired children who showed normal click-evoked BAEPs in infancy have an audiometric pattern with normal thresholds somewhere between 1 and 4 kHz and significant hearing losses at other frequencies (Fig. 24-6). Normal BAEPs occurred because the

broad-band click elicited a response from the frequency region with normal thresholds. Auditory steady-state responses may also be used to screen for hearing loss in newborn infants. 73,74 Either clicks or modulated noise may be used to evoke responses that can be accurately and quickly recorded in the frequency domain. Further work is needed to demonstrate the validity of this testing procedure, but the early results are very promising.

Evoked Potential Audiometry The click-evoked BAEP gives a general indication of hearing but does not provide the frequency-specific thresholds of an "audiogram." This information is essential to fitting hearing aids. By 6 months of age a normal child will turn to look for a faint sound, and this response can be reinforced by such stimuli as an animated toy on the correctly localized speaker. This approach can provide an accurate assessment of hearing in most children aged from 6 months to 2 years. As the child gets older, increasingly refined tests become available. Objective testing is necessary in all children before the age of 6 months and in many older children who cannot give reliable thresholds with behavioral testing (e.g., those with mental retardation, emotional disturbance, or multiple handicaps). The main goal of evoked potential audiometry is to determine thresholds at frequencies between 500 and 4,000 Hz in each ear. Evoked potential audiometry in infants and children usually begins with click stimuli. Thresholds for the BAEP (wave V) are estimated to within 10 dB using rapid stimulus rates (50 to 80 Hz).

125

40

250

500

1K

2K

o

40

60

8K

~ )

1'iV'"

20

30

4K

}./

I Ll.

I "'( 11

80

~~L

:0.2 fLV

J

100

A 15 msec B 120 FIGURE 24-6 • Follow-up audiograms in infants screened for hearing loss in the neonatal period. This child was born at term and was admitted to the neonatal intensive care unit because of perinatal asphyxia. The brainstem auditory evoked potentials in infancy showed normal thresholds. A, Responses to 40- and 3O-dB nHL clicks presented at a rate of 61 Hz to the right ear. Wave V is recognizable (triangle) at both intensities, B, Audiogram obtained at age 3 years, Although the thresholds at 2,000 and 4,000 Hz are within normal limits, the hearing at 500 and 1,000 Hz is significantly impaired. Both ears showed a hearing impairment, and the child was fitted with hearing aids after the audiogram,

Brainstem Auditory Evoked Potentials in Infants and Children

Under reasonable recording conditions (i.e., a quiet child in a quiet room), the BAEP in a child older than 6 months of age should be detected to within 20 dB of the normal adult behavioral threshold for the click. The threshold for the click-evoked BAEP does not give any information about the response to specific frequencies but can provide a general estimate of hearing, particularly at high frequencies. The next step estimates thresholds at different frequencies by using techniques such as derived responses, tone pips in noise, or steady-state responses. The most widely studied technique for estimating the audiogram in infants and young children involves recording the BAEPs to brief tones. 11,13,14,21,75 A metaanalysis of many such studies indicates that thresholds are estimated between 10 and 20 dB above behavioral thresholds for frequencies between 500 and 4,000 Hz.76 The variability of the threshold estimation is such that any individual threshold will be within 10 to 20 dB of this mean difference between the physiologic and behavioral thresholds. The basic protocol presents brief tones with alternating onset polarity at a rate of 40 to 50 per second. BAEPs are recorded with a low-frequency cut-off no greater than 30 Hz to obtain two replicate averages of 2,000 trials each. Ipsilateral notch-noise masking will make the thresholds frequency-specific. If notch-noise masking is not used, the estimated thresholds may be inaccurate if the slope between adjacent frequencies on the audiogram is steep. Bone conduction thresholds for clicks or tones should be obtained if air conduction thresholds are elevated. Alternating polarity is essential to attenuate the stimulus artifact when recording bone conduction responses. Auditory steady-state responses can also be used to estimate the audiogram. 15-17 Because multiple responses can be obtained simultaneously, the time taken to estimate the audiogram is significantly less than that required when using BAEPs to brief tones. More extensive evaluation of how well the test performs in newborn and young infants is still required, but the performance of the test in adults with hearing impairment'? suggests that it may become the most efficient and accurate way of objectively estimating the audiogram.

sient, unilateral, or mild. The incidence of hearing loss in children recovering from meningitis is between 25 and 60 percent when BAEP testing is performed.Pt'" The hearing loss is conductive in about one-quarter, cochlear in one-half, and retrocochlear in the remaining one-quarter of children. The BAEP is essential in the evaluation of infants born with atresia of the external auditory meatus (Fig. 24-7). In a child with unilateral atresia, the status of the nonatretic ear can be assessed easily by BAEPs.82 During the first few years of life, the BAEP may also be helpful if middle ear disease is suspected in the good ear. In a child with bilateral atresia, the BAEP can demonstrate function of the inner ear (see Fig. 24-7) and may suggest whether only one inner ear is functioning. Otitis media and middle ear effusions are very common in the first 3 years of life. It is important to monitor the hearing of children who experience recurrent middle ear disease. Because of the age of the children, monitoring will often require objective tests of auditory function, such as those provided by the BAEP. The conductive hearing losses associated with otitis media show up in the BAEP as a delay in all of the waves. The degree of delay can provide an estimate of the amount of hearing loss,82 although variability in the response may obscure small changes in threshold. BAEPs are very helpful in monitoring infants with cleft palate, who are susceptible to recurrent middle ear effusions.P Ototoxic medications are often used in infancy and childhood, particularly in intensive care nurseries. Because infants and young children are unable to

Bacterial meningitis is the most common cause of acquired sensorineural hearing loss in childhood. All children with meningitis should be assessed audiometricaUy before leaving the hospital. Because the incidence of this disease is high in the first year of life, the BAEP is often a necessary part of this assessment. The BAEP is probably more accurate than behavioral testing, particularly if the hearing impairment is tran-

Bone

Air

70

... ~

20

60

.~

10 ~

50

.-----.-

0

,

~

~

+ 0.5 fJ.V

A

BAEP Evaluation of Specific Hearing Disorders

533

dBnHL

~_~_~_ _]15msec

B

FIGURE 24·7 • Congenital atresia of the external auditory

meatus. This child was born at term with no abnormality other than bilateral absence of the pinna and external auditory canal. She wasreferred for testing at the age of 4 months. A, Responses to air-conducted rarefaction clicks presented at a rate of 61 Hz to the left ear. The triangles indicate wave V. The responses to right ear stimulation showed a similar elevated threshold at 60 dB nHL. B, Responses to an alternating click presented through a bone conduction transducer placed on the forehead. Responses are clearlyvisible down to 10 dB nHL. This pattern indicates that at least one cochlea is functioning well.

534

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

complain of auditory deterioration during the course of treatment, the BAEP may provide an objective means of monitoring hearing and detecting neurotoxicity, particularly if the treatment is prolonged. 84- 87

Treatment of Sensorineural Hearing Loss Once sensorineural hearing loss is detected, it is essential to provide the patient with amplification. Fitting hearing aids normally requires accurate information about the audiometric thresholds and the gain characteristics of the aid. The amplification and compression of the aid are then adjusted so that the frequencies and intensities of the normal speech spectrum are made audible. A final confirmation of a satisfactory hearing aid is the subject's ability to discriminate speech with the aid. The most important use of the BAEPs or auditory steady-state responses in the fitting of hearing aids is to provide an accurate audiogram when this cannot be obtained behaviorally. Other possible uses of physiologic testing are to determine maximum levels of amplification and to demonstrate the discriminability of the aided sounds. BAEPs have been used in the past to assess hearing aids. The general aim was to adjust the hearing aid until the latency or amplitude of wave V reached the normal range. These procedures were limited because (l) the click-evoked BAEP is mainly related to the high-frequency gain; (2) the correlation between wave V and loudness is low, particularly when a sloping hearing loss is present'f; and (3) the brief click can be significantly more distorted than more continuous stimuli in both the sound-field speaker and the hearing aid. 89 As they can be evoked by amplitudemodulated tones, which are frequency specific and stable over time, auditory steady-state responses are unlikely to be distorted by amplification in either a sound-field speaker or a hearing aid. These responses may be used to assess aided thresholds (Fig. 24-8).90 A more important use of the auditory steady-state responses would be to demonstrate the ability of the brain to discriminate changes in the frequency and intensity of amplified sounds.?' Cochlear implants can provide stimulation of auditory nerve fibers in children who are so severely hearing impaired that they get little or no benefit from hearing aids. It is customary to record the BAEPs in response to very loud clicks in children being considered for cochlear implants to determine any residual cochlear function. Some of these children show a vertex negative wave at about 3 msec that probably represents activation of vestibular nerve fibers rather than residual function in the hair cells and auditory nerve fibers.P'' Figure 24-9 shows a waveform that may represent this type of response. Electrically elicited BAEPs

provide essential information for preoperative evaluation of surviving neural elements in the auditory pathways and for intraoperative assessment of the implant's function. 9 3-95 After the operation, BAEPs may be used to check the continuing function of the implanted device, to adjust the intensity levels in the external processor, and to monitor the development of the auditory pathways. The waveform of the electrically evoked response is similar to that of the acoustically evoked BAEP. The latencies of the waves are shorter, and wave I is usually obscured by stimulus artifact. Wave V latency is usually about 4.0 msec and changes by only about 0.5 msec from high to low intensity.

Neurology of Hearing The BAEP does not really assess "hearing." Children with disorders of auditory perception may have normal BAEPs. In these children, normal BAEP thresholds indicate that they will probably not benefit from amplification. Some of these children may have abnormalities of the later auditory evoked potentials. Patients with abnormal BAEPs because of neurologic dysfunction probably do not have normal hearing. However, they may have much more hearing than is apparent from the BAEP findings. For example, demyelination in the auditory pathways may desynchronize neuronal discharges and render it impossible to record a BAEP. Such pathology may interfere with auditory perception that requires accurate timing without significantly affecting the hearing of pure tones or speech. Approximately two-thirds of infants with neurologically abnormal BAEPs will still show normal BAEP thresh01ds.62 However, BAEPs recorded for evaluating auditory thresholds in patients with known neurologic disorders must be interpreted cautiously. Particular attention must be paid to the latency and threshold of wave I, and electrocochleography may aid in the assessment of inner ear function." In all children treated for hearing loss without good behavioral testing, close cooperation is important between the audiologist and the therapist treating the child and the family. The child should not behave as well without the hearing aid as with it. The relationship between hearing and BAEPs (and between audiology and neurology) is particularly prominent in patients with a disorder that has come to be known as auditory neuropathy.96-99 This disorder is characterized by abnormal function in the auditory pathways beginning in the auditory nerve despite normal function in the external hair cells of the cochlea. The auditory nerve problems are demonstrated by absent or severely abnormal BAEPs, beginning at wave I; the relative preservation of hair cell function is demonstrated by otoacoustic emissions or cochlear microphonics. Patients have mild or moderate elevation

Brainstem Auditory Evoked Potentials inInfants and Children Stimulus frequencies (kHz)

Intensity (dB SPL)

0.5,1,2,4

40

Spectrum (60-120 Hz)

500

~ t t

0.5,1,2,4

30

0.5,1,2,4

20

~ t

http',. "d It

In

~ ~

tr

1K

2K

535

4K

Tt.+ .* t .ef-

~

-$-

100 nvJ

0.5,1,2,4

10

-,meer!rk','

Ii'l.

4-

+ +

-i-

Subject A24 Physiologic 40 30 20 30 Left ear Behavioral 25 15 15 15 FIGURE 24·8 II Auditory steady-state responses in the evaluation of aided thresholds. For these recordings four amplitude-modulated tones were presented simultaneously to the subject through a sound-field speaker. The carrier frequencies of the tones were 500, 1,000, 2,000, and 4,000 Hz. Each carrier frequency had its own signature modulation frequency: 81, 89, 97, and 105 Hz for the 500-, 1,000-, 2,000-, and 4,000-Hz carrier frequencies, respectively. The subject was a 12-year-old child with a bilateral sensorineural hearing impairment. The test was performed with a hearing aid in the left ear. The potentials evoked by the amplitude-modulated tones were converted into the frequency domain to provide measurements of amplitude at different frequencies. The middle column shows the amplitude spectrum from 60 to 120 Hz. The locations of the four modulating frequencies are noted on the baseline of the spectra. At the righ t are shown the responses at each of the modulating frequencies (measuring the responses to each of the tones) plotted with polar coordinates to show both amplitude and phase. The circles represent the confidence limits for the response. If the origin of the polar plot is not included within the circle (shaded), the response is significantly different from zero. In this child the physiologic responses are recognizable to within 15 dB of the thresholds that were measured with behavioral techniques. This approach may therefore provide a means for objectively evaluating aided thresholds in younger children who, unlike this child, are unable to provide behavioral responses.

of pure-tone thresholds but a significantly reduced ability to discriminate speech. Other abnormalities are a lack of middle ear muscle reflexes and no evidence of contralateral suppression of the otoacoustic emissions.'?" When recording BAEPs it is important to record separate responses to rarefaction and condensation clicks because the cochlear microphonics may be quite prominent and may mimic abnormal BAEPs (see Fig. 24-9).4 The disorder might be caused by an abnormality of the inner hair cells, of the afferent auditory nerve fibers, or of the synapse between them. Evidence of a more generalized neuropathy in some patients led to the concept of an auditory neuropathy. The disorder

may be idiopathic or occur in association with certain definite pathologies such as hyperbilirubinernia'v'v'" and hereditary sensorimotor neuropathy. 103,104 A report of a temperature-dependent form suggests that demyelination may playa role. 105 In certain cases, as the disorder progresses, the otoacoustic emissions disappear, perhaps owing to withdrawal of cochleotrophic factors. lOB The diagnosis is simple if the emissions are still present, but it is less certain if they are absent. The fact that auditory neuropathy occurs in newborn infants suggests that both otoacoustic emissions and BAEPs should be used in the screening process.l'" Otoacoustic emissions alone would miss these patients,

536

ELECTRODIAGNOSIS INCLINICAL NEUROLOGY

250

, 1000

, 4000 Rarefaction

o 20 40

60

80 100

_--l

"""

./

'\-(

l--_

Condensation

"""i.J'

\ \ \ \

Combined

U

U.I

120

90-dB nHL Clicks

, o

*

,O.~"~ ] 10 msec

FIGURE 14-11 • Absent brainstem auditory evoked potentials (BAEPs) in a patient with only mild hearing loss. This figure represents the audiogram and the BAEPs of a 17-year-old girl with a history ofjaundice at birth and difficulty with speech perception. Results are shown for the left ear. On the left of the figure, the pure-tone audiogram shows mild hearing loss except for frequencies greater than 4,000 Hz. On the right of the figure are shown the BAEPs in response to 90-dB nHL clicks. These BAEPs show small waveforms out to about 4 msec that reverse in polarity when the stimulus is changed from a condensation to a rarefaction click (two-headed arrows). These represent cochlear microphonics and are not neural potentials. When the condensation and rarefaction responses are combined, a vertex negative wave (asterisk) can be noted at a peak latency of approximately 3 msec, which may represent a response of the vestibular rather than the cochlear system.

and the BAEPs alone would indicate that there was a hearing loss but would not recognize its special character. The disorder is difficult to treat because amplification may not improve speech perception and may damage the remaining hair cells. Low-gain frequencymodulated systems may improve the acoustic signal-tonoise ratio, and speech perception may be supported with lip reading or other communication aids. Cochlear implants may help in some patients. 108 •109

NEUROLOGIC APPLICATIONS The BAEP is very useful in investigating neurologic disorders that may damage the brainstem auditory pathway. The BAEP is usually abnormal with brainstem tumors, is almost always abnormal with leukodystrophies, and is often abnormal with degenerative diseases that affect white matter. It is a valuable part of the neurologic workup in children with degenerative processes, encephalopathy, ataxia, or coma. In patients in whom the extent of brain damage has not been fully determined, the BAEP can demonstrate whether the brainstem has been affected significantly. However, the BAEP and the auditory pathways that generate this response are resistant to many neurologic insults and can be normal in a child who is far from normal neurologically. Furthermore, although an abnormal BAEP may demonstrate and localize dysfunction in the auditory pathway, it cannot define the etiology of this dysfunction.

When the BAEP is recorded for neurologic purposes, it is essential not to consider the results in isolation. The BAEP evaluates only the auditory pathways of the brainstem. Diagnosis and prognosis in pediatric neurology require information about the whole nervous system. Multimodality evoked potential studies are very helpful in distinguishing different disorders by the different patterns of abnorrnality.U'v'!' By surveying the extent of brain damage, they can also contribute to a more accurate prognosis. In many pediatric disease processes, abnormalities of the BAEP become apparent only when comparisons are made between groups of affected and normal subjects. Such comparisons may indicate that the disease affects the brainstem auditory pathways, but they do not aid in the diagnosis of individual patients, most of whom have normal BAEPs. Comparison may assist in assessment of the extent or severity of the disorder and thereby help to define the prognosis. Furthermore, if the BAEP is abnormal, monitoring the BAEP may aid in monitoring treatment. BAEPs in children with acquired immunodeficiency syndrome (AIDS) provide an example. Some delays in I-V latency occur, particularly at rapid rates of stimulation.U'' and abnormalities may be improved by treatment.I'" For neurologic assessment, identification of wave I is essential inasmuch as it is the absence or delay of subsequent waves that identifies dysfunction within the brainstem auditory pathways. A low rate of stimulation (10 to 20 Hz) and a relatively high intensity (70 to 90 dB nHL) should be used. Some neurologic disorders

Brainstem Auditory Evoked Potentials in Infants and Children

may become apparent only at more rapid rates of stimulation.!'" but we have not found this to be common. BAEPs to stimulation of the left and right ears should be recorded separately and, as with all evoked potential testing, replication is essential. If wave I is abnormal or absent, it is necessary to evaluate peripheral hearing either behaviorally or by using BAEPs.

Tumors The BAEP is very helpful in the diagnosis of tumors of the posterior fossa, 114.115 which are far more common in children than in adults. Abnormalities of the response can demonstrate a brainstem lesion and assist in its localization. Although computed tomography (CT) scanning and magnetic resonance imaging (MRI) are the most helpful diagnostic procedures in children with suspected tumors of the posterior fossa, the BAEP is useful as an inexpensive test when the index of suspicion is relatively low; as a means of confirming a small tumor or of determining the effect of a questionable radiographic abnormality; and as an effective way of monitoring progression during treatment. BAEP abnormalities beginning after wave I and before wave III indicate a lesion of the pons or auditory nerve (Fig. 24-10). The lateralization of such a lesion can usually be determined by noting from which ear the more abnormal response is obtained. The BAEP is a sensitive means for the early detection of acoustic neuromas. I 16 These tumors are rare in childhood and usually occur bilaterally in association with neurofibromatosis type 2. Genetic studies, MRI, and BAEP recordings are important means of detecting and monitoring children with this disorder.U? Abnormalities of the BAEP beginning after wave III indicate lesions of the upper pons or midbrain (see Fig. 24-10). An increased I-V latency without a clear wave III indicates some dysfunction of the auditory nerve or brainstem but is not more specific in localizing the lesion. Brainstem gliomas are particularly likely to produce abnormal BAEPs, sometimes before the clinical or radiographic evidence becomes definite. One case of normal BAEPs in a child with a brainstem glioma involving the thalamus and upper midbrain has been reported. 115 In our experience, however, the BAEPs have been abnormal in 26 children with tumors involving the pons or midbrain. Furthermore, in only 2 of these patients have the BAEPs not correctly predicted the side predominantly involved. Both children had large mass effects in the brainstem, and the BAEPs may well have reflected functional rather than structural effects of the lesion. In three children with thalamic tumors, we have found a small or delayed wave V,probably reflecting pressure on the midbrain. Far-field somatosensory

537

evoked potentials (SEPs) were more useful in assessing these tumors. In children with cerebellar tumors, the BAEP is usually abnormal, depending on whether the auditory pathway is affected by tumor extension or mass effect. The BAEP may be helpful in monitoring brainstem function during operations on posterior fossa tumors and in assessing the effects of surgical resection, radiation, and chemotherapy.'!" Figure 24-11 shows the BAEP studies of a child before and after resection of a cerebellar cyst. Neurofibromatosis type I (von Recklinghausen's disease) is an inherited disorder that affects the cell growth of neural tissue and can lead to tumor formation. Evoked potentials can help to detect abnormalities and monitor progression of the disease. We have found that 11 of 25 children with neurofibromatosis type I had an abnormally delayed I-III or I-V interpeak interval (Fig. 24-12). One child had an auditory nerve tumor, another had a brainstem glioma, and the others are being monitored. This incidence of abnormal BAEPs is similar to that reported by other investigatorS.1l9.120 The yield of abnormalities is even greater with multi-modality evoked potentials or magnetic resonance imaging. Evoked potentials are clearly useful for detecting cerebral abnormalities in children with neurofibromatosis, but the significance of these abnormalities is uncertain. Some abnormalities may represent tumor formation, but others probably indicate non-neoplastic local areas of brain dysplasia. More long-term follow-up will determine the value of monitoring patients with neurofibromatosis.

Developmental Disorders Hydrocephalus is typically associated with severe BAEP abnormalities, although normal BAEPs may occasionally occur. The most common abnormality is a decrease in the amplitude of the later components, with a significantly reduced V/1 amplitude ratio or an absent wave V.121,122 The relationship of the BAEP abnormality to the hydrocephalus is unclear. Abnormal recordings may be partially related to the ventricular dilatation because the BAEP can improve when the child is successfully treated with a shunt.P" They may also be caused by developmental abnormalities of the brainstem that occur in association with hydrocephalus. The Arnold-Chiari malformation is present in most patients with myelomeningocele, but it may occur occasionally without either myelomeningocele or any associated hydrocephalus. It can cause symptoms related to medullary compression, the most important being swallowing difficulties and apneic episodes. Brainstem decompensation can sometimes lead to respiratory failure and death. Several studies have investigated whether BAEPs in children with the Arnold-Chiari

538

ELEaRODIAGNOSIS IN CLINICAL NEUROLOGY

L

+ R

L

R

L

R

A

B

FIGURE 14-10 • Brainstem auditory evoked potentials (BAEPs) in the evaluation of children with tumors of the posterior fossa. A, Responses to 70-dB nHL click stimuli presented at a rate of 11 Hz. Responses to left (L) and to right (R) ear stimuli were recorded with electrodes on the vertex and ipsilateral mastoid or earlobe. Waves I, III, and V are identified with arrows. The calibration lines represent 0.5 11m and 2 msec. The third set of tracings was recorded with a sweep of 15 msec rather than 10 msec. B, Computed tomographic (CT) scans from these children. All scans were obtained with contrast enhancement. The scans are printed such that the left side of the patient is on the left of the scan. The patient whose tracings are shown at the top of the figure was a If-year-old girl with right-sided hearing loss and headaches. Her audiogram showed a severe right-sided hearing loss for pure tones with no detectable speech discrimination. Only the first component of the response was recognized in the BAEP to right ear stimulation. The CT scan showed a large tumor of the right cerebellopontine angle. An acoustic neuroma was removed at surgery. The tracings in the middle part of the figure were obtained from an 8-year-old girl with ataxia and slurred speech. On examination she showed bilateral sixth nerve palsies and a left seventh nerve palsy. The BAEPs in response to left ear stimulation showed no components after wave III. In response to right ear stimulation, there were questionable wavesNand V of small amplitude. The CT scan showed an enhancing tumor in the left pons. The patient underwent shunting but did poorly and died 6 months later. Pathology revealed a malignant astrocytoma. The bottom tracings are from a patient examined for ataxia and vomiting. The BAEPs show a severely delayed I-V interval (7.9 msec) for right ear stimulation, with no clear wave III. The response to left ear stimulation showed normal components I. II, and III and a delayed wave V. The CT scan showed a contrast-enhancing mass in the right pontine region pushing the fourth ventricle backward and to the left. An exploratory operation revealed an unresectable astrocytoma of the right pons.

Brainstem Auditory Evoked Potentials in Infants and Children

539

Patient 1

7 yr

R 4.5.88

R

~ L

L

Patient 2 15 yr

L 22.8.88

R

L 19.5.91

""-""'r-'--r-r--"-"""T"--..,r--.,--r--,--, I I I I I

o

2

4

6

L

]~ 0.3 fLV

8

msec FIGUIE24·11 • Monitoring tumor progression with brainstern

auditory evoked potentials (BAEPs). These recordings were obtained in a girl aged 7 years who was seen in early 1988 with behavior problems and seizures following a skull fracture in 1985. Her BAEPwas abnormal. With left-ear (L) stimulation with 80-<1B nHL clicks presented at 11 Hz, wave III was not recognizable and wave V was significantly delayed; the response to stimulation of the right ear (R) was normal. Radiologic studies showed a cystic tumor of the left cerebellum, which was removed in June 1988. Subsequen t recordings in August 1988 and May 1991 show improvement of the BAEP to normal. Arrows indicate the various components of the response. malformation can predict the emergence of these clinical symptoms and the need for brainstem decompression. In children with myelomeningocele, the longest central conduction times in occur in those with hydrocephalus and neurologic symptoms.F" Brainstem decompression may lead to resolution of the medullary symptoms and improvement in the BAEP. 125 However, another study found abnormal BAEPs in 72 percent of 18 children older than 5 years of age with myelomeningocele and hydrocephalus, none of whom had clinical syrnptoms.l'" MRJ is probably the most helpful test in determining brainstem compression, but the BAEP may be helpful in evaluating the general neurologic prognosis. We studied BAEPs after myelomeningocele repair in 47 neonates between 1 day and 3 months of age to determine whether the BAEP had prognostic value regarding the child's neurologic outcome.F? In our series, 69 percent had abnormal BAEPs: some had small waves IV and V, whereas others had a normal wave I and poorly defined and delayed later components (Fig. 24-13). BAEPs were abnormal in all infants with symptomatic Arnold-Chiari malformation and in 47 percent of nonsymptomatic patients. More interestingly,

I

o

I

i

I

I

2

4

6

8

msec FIGURE 24·12 • The brainstem auditory evoked potentials

(BAEPs) from two patients with neurofibromatosis. The top pair of traces, from a 7-year-old, show normal responses to 70-dB nHL clicks at 11 Hz for the right ear (R) but delayed III-V and I-V latencies for the left ear (L) (2.4 and 4.7 msec, respectively). Magnetic resonance imaging revealed only a small focal nodule deep in the right cerebellum. Neurofibromatosis type I was diagnosed in this patient at age 4 years, and she had a positive family history and cafe-au-lair spots. Apart from some developmental delay, her neurologic examination is normal. The second pair of traces is from a 15year-old boy with neurofibromatosis type II. Computed tomography showed bilateral acoustic neuromas (L greater than R), a cystic lesion in the upper cervical cord, and a lesion in the left jugular foramen. The BAEPs were recorded in response to 80-dB nHL click stimuli at 11 Hz before surgical excision of the left acoustic neuroma. Only wave I is present on the left; on the right, the peaks after wave 1 are poorly delineated and delayed (I-V interpeak latency of 5.3 msec).

87 percent of the infants with abnormal BAEPs had central neurologic abnormalities on follow-up at 2 years of age, whereas 81 percent of the infants with normal BAEPs had normal cerebral function on follow-up. Thus, the BAEP was highly predictive of neurologic outcome. In this series the BAEP abnormalities were associated with cerebral dysfunction that became evident only after 1 to 2 years, at an age when disturbances in normal development (apart from the typical sequelae of myelomeningocele such as lower limb weakness or paresthesias or neurogenic bladder) are more easily identified. BAEPs are also useful during surgical decompression of ArnoId-Chiari malformations, to help determine the extent of decompression needed and to reduce the risk of neurologic injury.128

540

ElECTRODIAGNOSIS INClINICAL NEUROLOGY

I

I

I

I

2

6

10

14

msec FIGURE 24·13 • Brainstem auditory evoked potentials (BAEPs)

in infants with myelomeningocele and hydrocephalus. These recordings were taken from three different infants after repair of the myelomeningocele and before any shunting. The first infan t had normal BAEPs. An associated Arnold-Chiari malformation was present, but there was no definite hydrocephalus and no shunt was placed. The second infant shows a small delayed wave V. The third infant shows delayed and poorly formed waves III and V. Both the second and the third infants had severe hydrocephalus and underwent shunting at 2 weeks of age. All recordings were in response to 90-<1B nHL monaural clicks at 11 Hz. The responses were similar for stimulation of either the left or the right ear. For clarity, the figure shows only one of the responses for each infant.

Hydranencephalic infants have normal BAEPs, although later auditory responses, visual evoked potentials (VEPs), and cortical SEPs are absent. 129,130

entiate between gray and white matter diseases. Gray matter diseases cause paroxysmal EEG abnormalities and tend not to affect the BAEP. White matter diseases cause diffuse EEG slowing and abnormal BAEPs. The BAEPcan demonstrate the demyelinating nature of the disease process very early in its course and may be helpful in differentiating between the leukodystrophies. The BAEPs are so consistently abnormal in the various leukodystrophies that another diagnosis should be considered if the BAEP is normal. Figure 24-14 shows typical BAEPs recorded in children with different leukodystrophies. In the infantile leukodystrophies (e.g., Krabbe's disease or Pelizaeus-Merzbacher disease), the BAEPs are very abnormal, with only the early waves being recognizable. 131- 135 We have not seen normal BAEPs in any children with these disorders even though the disease has been clinically diagnosed in children as young as 1 month of age. The BAEPs may aid in the differential diagnosis inasmuch as such marked BAEP abnormalities are not seen in other disorders with similar clinical features. In older children with demyelinating disorders, abnormalities of the BAEP are less severe. In metachromatic and adrenal leukodystrophies, the initial abnormality is an increase in interpeak latencies, particularly between waves I and III. The abnormalities increase with clinical severity. Even in patients who are still neurologically normal, the BAEPs are abnormal and useful as an early diagnostic test for leukodystrophy.114,136,137 Some studies have found that carriers of the sex-linked gene for adrenoleukodystrophy also have abnormal BAEPs.138,139 Multimodal evoked potential studies have long been used to evaluate adult demyelinating disorders such as multiple sclerosis. Although far rarer in children, multiple sclerosis is often a difficult diagnostic dilemma because of the fluctuating symptoms and because children are often unreliable observers. As with adults. MRI and multimodal evoked potential studies are very important.lv' In a series of 24 children with multiple sclerosis, we found abnormal BAEPs in 4. Although they were not abnormal in isolation. BAEPs helped to demonstrate multiple lesions.

Myelin Disorders Disorders affecting the myelin sheath initially delay the later components of the BAEP. As the disorder progresses, these later components decrease in amplitude and finally disappear until only wave I is present. Because these findings are probably the result of desynchronization rather than disruption of transmission in the auditory pathway, they usually occur before definite auditory symptoms. Combined use of the EEG and the BAEP in children with progressive neurologic disorders may help to differ-

Metabolic and Degenerative Disorders The BAEP is abnormal in many metabolic disorders. Those that affect the BAEP probably cause significant disruption of cerebral myelin. Inherited metabolic disorders associated with abnormal BAEPs are maple syrup urine disease. pyruvate decarboxylase deficiency, and phenylketonuria. 114,115 Four of five children with nonketotic hyperglycinemia have had prolonged I-V intervals. 141 BAEPs are normal in children

Brainstem Auditory Evoked Potentials in Infants and Children

2 yr

6mo

Pelizaeus-Merzbacher disease

Krabbe's disease

19 mo

+ ]

0.3 fLV

Metachromatic leukodystrophy

3 yr MUltiple sulfatase deficiency

+

17 yr Adrenoleukodystrophy

I 0

I 2

I 4

I

6

I 8

]+0.15fLV I 10

msec FIGURE 24·14 • Brainstem auditory evoked potentials (BAEPs) recorded in patients with various leukodystrophies. Individual components are indicated by arrows. The top traces are from a 2-year-old with Pelizaeus-Merzbacher disease. Only waves I to III are present, at normal latencies. The second traces are from a 6-month-old with Krabbe's disease. The BAEP shows only waves I to III, with delayed interpeak latencies of 3.2 msec. The third traces are from a patient who was asymptomatic but who had metachromatic leukodystrophy confirmed by absent arylsulfatase A activity in enzymatic assays in fibroblasts; the child was investigated because of a symptomatic sibling. The I-III and I-V interpeak latencies are prolonged (2.7 and 4.8 msec, respectively). The fourth trace is from a 3-year-old with multiple sulfatase deficiency. Her BAEPs show prolonged absolute and interpeak latencies (I-III and I-V are 3.3 and 5.4 msec, respectively). The final trace is from a boy with adrenoleukodystrophy. The BAEPs show poorly defined waves, with delayed III-V and I-V intervals (2.5 and 4.8 msec, respectively) and a lowamplitude wave V. All responses were to monaural 80-<1B nHL clicks presented at I I per second. For clarity, only the response from one ear is shown. The upper calibration refers to the top three tracings and the lower to the bottom two tracings.

541

542

ELECTRODIAGNOSIS IN CLINICAL NEUROLOGY

with abetalipoproteinemia, although the VEPs and SEPs are often abnormal. 142.143 We have also found abnormal BAEPs in patients with Leber's disease, propionic acidemia, and Kearn-Sayre syndrome; in five of six patients with cerebrohepatorenal (Zellweger) syndrome; and in one of three patients with Menkes' kinky hair disease. Leigh disease, or subacute necrotizing encephalomyelopathy, is a syndrome that results from a number of metabolic abnormalities in pyruvate metabolism. Within this broad grouping are patients with deficiencies of pyruvate dehydrogenase, cytochrome oxidase, complex 1, and complex 5, as well as other mitochondrial disorders. The BAEPs in these patients are often abnorrnal.l'
5 mo

s rno

15 rno

~.o-"'~] :0.3 !LV 7ma

I

o

i

I

I

i

2

4

6

8

msec FIGURE 24-15 • Brainstem auditory evoked potentials (BAEPs) in Leigh's syndrome. This figure presents recordings from four patients with autopsy-proven Leigh's syndrome in whom the underlying metabolic defects were known. The top traces are from a patient with complex I deficiency, who typically has normal BAEPs. The second set of traces, from a patient with pyruvate dehydrogenase deficiency, shows poorly defined BAEPs with interpeak latencies just at the upper limits of normal for his age. The third traces are from a patient with complex 4 or cytochrome oxidase deficiency with delayed interpeak latencies and low-amplitude waves IV and V. The bottom traces are from a patient with complex 5 deficiency; it shows normal waves I to III, but absent later waves. The upper vertical calibration refers to the upper three responses, which were evoked by 70-<1B nHL clicks presented at 11 Hz. The lower calibration refers to the bottom response, which was elicited by 80-dB nHL clicks. For clarity, only the response from one ear is shown in each case.

The abnormal BAEPs in Friedreich's ataxia (Fig. 24-17) can help to exclude other hereditary ataxias in which the BAEPs are normal. In agreement with these BAEP findings, neuropathologic studies have shown cell loss and gliosis in the brainstem of patients with Friedreich's ataxia but not in other ataxias such as ataxia telangiectasia. The BAEPs in Friedreich's ataxia exhibit a characteristic rostrocaudalloss of waves beginning very early in the course of the disease. 15Q-152 This typical BAEP abnormality can be very helpful because young children with ataxia do not always have the differentiating signs (e.g., telangiectasia or cardiomyopathy) that allow easy discrimination among ataxias in older children and adults. The BAEPs also reflect both

Brainstem Auditory Evoked Potentials in Infants and Children

543

11 me

FIGURE 24·16 II Brainstem auditory evoked potentials (BAEPs) from three different children with Hurler's syndrome. The responses were elicited by 70-dB nHI. clicks presented at I I Hz. The top three recordings were obtained before bone marrow transplantation. The recordings illustrate the types of BAEP abnormalities found in these children. The top trace shows delayed waves beginning at wave I, probably caused by conductive hearing loss (seen in two cases). The second trace shows a complete absence of waves that could represent both hearing loss and brainstem dysfunction (seen in two cases). The third trace shows a prolonged interpeak latency indicating brainstem dysfunction (seen in three cases). The fourth trace shows a normal BAEP following bone marrow transplantation in the same child as in the top trace.

20 me

1 yr

3 yr

~0.5~V 1 msec

the rate of progression and the severity of Friedreich's ataxia.F" Certain disorders such as kernicterus have a specific predilection for the auditory pathways.l'" In acute bilirubin encephalopathy, BAEPs are absent or severely abnormal. BAEP abnormalities vary with the "bilirubin toxicity," an index that takes into account serum bilirubin, albumin binding, and pH. 153 The BAEP is a good technique for detecting and monitorin