in Disorders of the Retina, Optic Nerve,and Visual Pathway

SECOND EDITION

OPHTHALMOLOGY MONOGRAPHS COMMITTEE

H. Sprague Eustis, MD, ChairAntonio Capone, Jr, MDWilliam W. Culbertson, MDJames C. Fleming, MDCarol L. Karp, MD

]D THE FOUNDATION\V OF THE AMERICAN ACADEMY

OF OPHTHALMOLOGY

655 Beach Street

P.O. Box 7424

San Francisco, CA 94120-7424

CLINICAL EDUCATION SECRETARIES

Thomas J. Liesegang, MDSenior Secretary

Gregory L. Skuta, MDSecretary for Continuing Education

CLINICAL EDUCATION STAFF

Kathryn A. Hecht, EdDVice President

Hal StrausDirector of Publications

Pearl C. VapnekManaging Editor

Ruth K. ModricProduction Manager

Each author states that he has no significant financialinterest or other reiationship with the manufacturer ofany commercial product discussed in the chapters thathe contributed to this publication or with the manufac-turer of any competing commerciai product, with the ex-ception of David G. Birch, PhD, who has received con-sulting fees and research support from Pfizer, Inc., andTAP Pharmaceuticals.

DesignChristy Butterfield Design

ProductionJanet Bollow Associates

Reference to certain drugs, instruments, and other products inthis publication is made for illustrative purposes only and is notintended to constitute an endorsement of such. Some materialmay include information on applications that are not consideredcommunity standard, that reflect indications not included in ap-proved FDA iabeling, or that are approved for use only in re-stricted research settings. The FDA has stated that it is the re-sponsibility of the physician to determine the FDA status ofeach drug or device he or she wishes to use, and to use themwith appropriate patient consent incompliance with applicablelaw. The Academy specifically disclaims any and all liability forinjury or other damages of any kind, from negligence or other-wise, for any and all claims that may arise from the use of anyrecommendations or other information contained herein.

The Academy provides this material for educationai pur-poses only. It Is not intended to represent the only orbest method or procedure in every case, nor to replace aphysician's own judgment or give specific advice for casemanagement. Including ail indications, contralndications,side effects, and aiternative agents for each drug ortreatment is beyond the scope of this materiai. All infor-mation and recommendations should be verified, prior touse, with current information included in the manufactur-ers' package inserts or other independent sources, andconsidered in iight of the patient's condition and history.

Ophthalmology Monograph 2, Elec-trophysiologic listing in Disor4ers ofthe Retina, Optic Nerve, and Visual

~IFELONG Pathway, Second Edition, Is oneEDUCATION FOR THEOPHTirALMOLOGIST~ component of the Lifelong Educa-

tion for the Ophthalmologist (LEO)framework, which assists members in planning theircontinuing medical education. LEO includes an array ofclinical education products and programs that members

may select to form individualized, self-directed learningplans for updating their clinical knowledge. Active mem-bers or fellows who use LEO components may accumu-late sufficient CME credits to earn the LEO Award.Contact the Academy's Clinical Education Division for

further information on LEO.

Copyright @ 2001 Gerald Allen Fishman, MDU sed by permission.

Copyright @ 1990 American Academy of OphthalmologyAll rights reserved.

,ibrary of Congress Cataloging-in-Publication Data

Electrophysiologic testing in disorders of the retina, optic nerve, and visual pathway IGerald Allen Fishman ...[et al.].-2nd ed.

p. ; cm. -(Ophthalmology monographs; 2)Includes bibliographical references and index.ISBN 1-56055-198-41. Electroretinography. 2. Electrooculography. 3. Visual evoked response. I. Fishman,

Gerald Allen, 1943- II. Series.[DNLM: 1. Retinal Diseases-diagnosis. 2. Electrooculography. 3. Electroretinography.

4. Evoked Potentials, Visual. 5. Optic Nerve Diseases-diagnosis. WW 270 E38 2001]RE79.E4 F57 2001617.7'307547-dc21

00-056208

)5 04 03 02 5 4 3 2 1

Printed in Singapore

Contents

Preface Xttt

Acknowledgments XV!

Chapter 1 1THE ELECTRORETINOGRAM

Gerald Allen Fishman, MD

1-1 Components and Origins of the ERG 21-1-1 a-, b-, and c-Waves and Off-Responses 21-1-2 Scotopic and Photopic Threshold Responses 61-1-3 Early Receptor Potential 71-1-4 Oscillatory Potentials 8

1-2 Measurement of the ERG Components 101-3 Recording Procedure 111-4 Recording Electrodes 12

1-4-1 Burian-Allen Electrode 121-4-2 Dawson-Trick-Litzkow Electrode 121-4-3 ERG-Jet Electrode 131-4-4 Mylar Electrode 131-4-5 Skin Electrode 131-4-6 Cotton-Wick Electrode 131-4-7 Hawlina-Konec Electrode 14

1-5 The ERG Under Light- and Dark-Adapted Conditions 141-5-1 With a Constant-lntensity Stimulus 171-5-2 With Variable-lntensity Stimuli 181-5-3 With Different-Wavelength Stimuli 201-5-4 With Variable-Frequency Stimuli 21

1-6 Other Factors Affecting the ERG 231-6-1 Duration of Stimulus 231-6-2 Size of Retinal Area Illuminated 241-6-3 Interval Between Stimuli 241-6-4 Size of Pupil 251-6-5 Circulation and Drugs 251-6-6 Development of Retina 251-6-7 Clarity of Ocular Media 271-6-8 Age, Sex, and Refractive Error 271-6-9 Anesthesia 271-6-10 Diurnal Fluctuations 28

1-7 The ERG in Retinal Disorders 28

v

Contentsvi

1-8 Diffuse Photoreceptor Dystrophies 291-8-1 Rod-{;one Dystrophies 30

1-8-1-1 Retinitis Pigmentosa 301-8-1-2 Allied Disorders 35

1-8-2 Cone-Rod Dystrophies 391-9 Stationary Cone Dysfunction Disorders 43

1-9-1 Congenital Achromatopsia 431-9-2 Congenital Red-Green Color Deficiency 44

1-10 Stationary Night-Blinding Disorders 451-10-1 Congenital Stationary Night Blindness 451-10-2 Oguchi Disease 501-10-3 Fundus Albipunctatus 501-10-4 Fleck Retina of Kandori 53

1-11 Hereditary Macular Dystrophies 541-11-1 Stargardt Macular Dystrophy 541-11-2 Best Macular Dystrophy 561-11-3 Pattern Dystrophies 581-11-4 North Carolina Macular Dystrophy 591-11-5 Progressive Bifocal Chorioretinal Atrophy 591-11-6 Cone Dystrophies 601-11-7 Sheen Retinal Dystrophies 64

1-12 Chorioretinal Dystrophies 651-12-1 Choroidal Atrophy 651-12-2 Gyrate Atrophy 651-12-3 Choroideremia 66

1-13 Angioid Streaks 691-14 Hereditary Vitreoretinal Disorders 69

1-14-1 X-Iinked Juvenile Retinoschisis 691-14-2 Goldmann-Favre Syndrome 721-14-3 Wagner Disease 721-14-4 Stickler Syndrome 731-14-5 Autosomal Dominant Neovascular Inflammatory Vitreoretinopathy 731-14-6 Autosomal Dominant Vitreoretinochoroidopathy 731-14-7 Familial Exudative Vitreoretinopathy 74

1-15 Inflammatory Conditions 751-15-1 Birdshot Retinochoroidopathy 761-15-2 Multiple Evanescent White-Dot Syndrome 781-15-3 Acute Zonal Occult Outer Retinopathy 821-15-4 Pseudo-Presumed Ocular Histoplasmosis Syndrome 831-15-5 Behget Disease 83

1-16 Circulatory Deficiencies 841-16-1 Sickle Cell Retinopathy 841-16-2 Takayasu Disease 841-16-3 Carotid Artery Occlusion 84

viiContents

1-16-4 Central and Branch Artery and Vein Occlusions 851-16-5 Hypertension and Arteriosclerosis 88

1-17 Toxic Conditions 881-17-1 Chloroquine and Hydroxychloroquine 881-17-2 Chlorpromazine 901-17-3 Thioridazine 911-17-4 Indomethacin 931-17-5 Quinine 931-17-6 Methanol 951-17-7 Gentamicin 951-17-8 Ethyl-m-Aminobenzoic Acid Methanesuifonate 971-17-9 Cisplatin 971-17-10 Glycine 971-17-11 Canthaxanthin 971-17-12 Vigabatrin 981-17-13 Deferoxamine 981-17-14 Sildenafil 99

1-18 Vitamin A Deficiency and Retinoids 991-19 Optic Nerve and Ganglion Cell Disease 1001-20 Opaque Lens or Vitreous 1021-21 Diabetic Retinopathy 1021-22 Miscellaneous Conditions 104

1-22-1 Retinal Detachment 1041-22-2 Silicone oil and Sulfurhexafluoride Gas 1041-22-3 Thyroid and Other Metabolic Dysfunctions 1051-22-4 Parkinson Disease 1061-22-5 Myotonic Dystrophy 1061-22-6 Duchenne and Becker Muscular Dystrophies 1061-22-7 Kearns-Sayre Syndrome 1071-22-8 Neuronal Ceroid Lipofuscinoses 1101-22-9 Retinal Degeneration With Spinocerebellar Ataxia 1121-22-10 Tay-Sachs Disease 1141-22-11 Creutzfeldt-Jakob Disease 1141-22-12 Intraocular Foreign Bodies 1151-22-13 Myopia 1161-22-14 Albinism 1171-22-15 Taurine Deficiency 1171-22-16 Bietti Crystalline Dystrophy 1171-22-17 Cancer-Associated Retinopathy 1181-22-18 Enhanced S-Cone Syndrome 1201-22-19 Pigmented Paravenous Retinochoroidal Atrophy 120

1-23 Summary 124References 124Suggested Readings 152

x Contents

4-3-2 Optic Nerve Disease 2144-3-2-1 Optic Nerve Demyelination 2164-3-2-2 Optic Nerve Compression 2194-3-2-3 Optic Atrophy 2194-3-2-4 Miscellaneous Optic Nerve Dysfunction 221

4-3-3 Glaucoma 2214-3-4 VEP Abnormalities and the PERG 2264-3-5 Nonorganic Visual Loss 226

4-4 Summary 227References 227Suggested Readings 235

Chapter 5 237THE VISUAL EVOKED POTENTIAL

Mitchell G. Brigell, PhD

5-1 Origins of the VEP 2395-2 VEP Stimulus Parameters 2405-3 VEP Response Parameters 2425-4 Clinical Use of the VEP in Adults 245

5-4-1 Media Opacities 2455-4-2 Retinal Diseases 246

5-4-2-1 Central Serous Retinopathy 2465-4-2-2 Macular Disease 2465-4-2-3 Glaucoma 246

5-4-3 Demyelinative Optic Neuropathies 2465-4-3-1 Optic Neuritis 2465-4-3-2 Multiple Sclerosis 2475-4-3-3 Leukodystrophies 2485-4-3-4 Nutritional Amblyopia 248

5-4-4 Compressive Optic Neuropathies 2495-4-4-1 Dysthyroid Optic Neuropathy 2505-4-4-2 Idiopathic Intracranial Hypertension 250

5-4-5 Anterior Ischemic Optic Neuropathies 2505-4-6 Traumatic Optic Neuropathies 2515-4-7 Toxic Optic Neuropathies 251

5-4-7-1 Ethambutol 2515-4-7-2 Cisplatin 2515-4-7-3 Deferoxamine 252

5-4-8 Inflammatory Optic Neuropathies 2525-4-9 Hereditary Optic Neuropathies 252

5-4-9-1 Autosomal Dominant Optic Atrophy 2525-4-9-2 Leber Hereditary Optic Neuropathy 2525-4-9-3 Mitochondrial Myopathies 253

5-4-10 Operating Room Monitoring 253

Contents xi

5-4-11 Neurodegenerative Diseases 2545-4-11-1 Alzheimer Disease and Other Dementias 2545-4-11-2 Spinocerebellar Degeneration 254

5-4-12 Functional Disorders 2545-4-13 Disorders of the Optic Chiasm, Optic Radiations, and Visual Cortex 256

5-4-13-1 The Hemifield VEP and Paradoxical Lateralization 2565-4-13-2 Chiasmal Compression 2575-4-13-3 Albinism 2575-4-13-4 Dysfunction of the Optic Radiations and Visual Cortex 2575-4-13-5 Migraine 2575-4-13-6 Developmental Dyslexia 2585-4-13-7 Cortical Blindness 258

5-5 Clinical Use of the VEP in Children 2585-5-1 Estimation of Visual Acuity in Infants 2595-5-2 Childhood Amblyopia and Binocular Function 2605-5-3 Oculomotor Disorders 2625-5-4 Delayed Visual Maturation 2625-5-5 Optic Nerve Hypoplasia 2625-5-6 Optic Nerve Gliomas 2625-5-7 Prognostic Value of the VEP in Neurologic Diseases 263

5-6 Practical Guidelines 2645-6-1 Instrumentation 2645-6-2 Patient Preparation 2645-6-3 Adult Protocol 2645-6-4 Pediatric Protocol 266

5- 7 Summary 267References 268Suggested Readings 278

281

282

283

291

CME Credit

Answer Sheet

Self-Study Examination

Index

Gerald Allen Fishman, MD

The full-field (ganzfeld) light-evoked elec-troretinogram (ERG) is the record of a dif-fuse electrical response generated by neuraland nonneuronal cells in the retina. The re-sponse occurs as the result of light-inducedchanges in the transretinal movements ofions, principally sodium and potassium, inthe extracellular space. This retinal poten-tial can be recorded in all vertebrates and inmany invertebrates. It was first identified inrecordings from a frog eye in 1865 by theSwedish physiologist Alarik Frithiof Holm-gren, who initially misinterpreted the wave-form as arising from action potentials in theoptic nerve. Although James Oewarl ofScotland recorded this potential in humansas early as 1877, electroretinography did notfind widespread clinical application until1941, when the American psychologist Lor-rin RiggsZ introduced a contact-Iens elec-trode for human use. In 1945, Gosta Karpe3reported the results of a study on 64 normaland 87 abnormal human eyes, providing theinitial groundwork for clinical investigation.

The ERG represents the combined elec-trical activity of different cells in the retina.In 1908, Einthoven and Jolly4 reported onthree components of the ERG. An initialnegative deflection was designated by theletter "a"; a subsequent positive compo-nent, normally greater in amplitude, wastermed "b"; and a final, prolonged positivecomponent was referred to as "c." In 1933,Ragnar Granit,5 working with the dark-

adapted, rod-dominated cat retina, demon-strated that the ERG consisted of threeprocesses, appropriately called PI, PII, andPIll. The Roman numerals designated thtorder in which a gradual increase in ethernarcosis suppressed the various ERG com-ponents in the cat. The PI, PII, and PIllprocesses of Granit correspond to the c-, b-,and a-waves, respectively, by which thecomponents of the ERG are now conven-tionally named.

Subsequently, the PIll component wasfound to consist of two separate compo-nents, or phases, that arise from two differ-ent classes of retinal cells (Figure 1-1 ). Theinitial phase (termed the receptor potential or

fast PIll), the leading edge of which sub-stantially forms the a-wave, reflects the ac-tivity of photoreceptor cells and arises fromlight-evoked closure of sodium channelsalong the plasma membrane of receptor-cellouter segments (Figure l-IA). The second,more slowly developing phase ( termed theslow PIll) has its origin from a hyperpolar-ization in the distal end of Muller cells.This response of the Muller cells is thoughtto be stimulated by a light-induced de-crease in extracellular potassium in thevicinity of photoreceptor-cell inner seg-ments developing as a consequence of pho-toreceptor-cell activity (Figure 1-1 B ).

1

The Electroretinogram2

BA

Pigment epithelial potential

bc

Time

a-#

Slow Pill potential

2000 50 100Time (ms)

150Light

Figure 1-1 ( A) ERG a- and b-wave response to lO-msflash in dark-adapted human eye. Also illustrated isdashed-curve fit of receptor model, which providesestimate of receptoral contribution to ERG. Note thatpeak a-wave amplitude does not represent peak recep-tor response, because rod a-wave of human ERG tomoderately intense flashes is partly postreceptoral inorigin. (B) ERG a-, b-, and c-wave response to 4-sstimulus in cat eye, as well as slow PIlI and retinalpigment epithelial potentials, whose origins are Mullercells and retinal pigment epithelium, respectively.Pat1 A modified with permission of Kluwer Academic Pub-lishers from l{ood DC, Birch DG: Assessing abnormal rodphotoreceptor activity with the a-wave of the electroretino-gram: applications and methods. Doc Ophthalmol1997 ;92:253-267. Pat1 B modified with permission of ElsevierScience from Ripps H, Witkovsky P: Neuron-glia interactionin the brain and retina. In: Osborne NN, Chader GJ, eds:Progress in Retinal Research. Elmsford, NY:. PergamonPress; 1985;4:181-219. Copyright 1985, Pergamon PressPLC.

COMPONENTS AND ORIGINS OF THE ERG

1.1.1 a., b-, and c-Waves and Off-Responses

The site of origin of the ERG componentshas been a subject of investigation for morethan half a century. Various approaches forinvestigation have compared the results from

1. The predominantly cone or rod retinas of..

vanous species2. Chemical destruction, with toxins knownfrom histologic data to destroy specific reti-nal cells63. Microelectrode investigations of individ-ual cell layers4. Observations of the ERG components'variation with various stimulus intensitiesand adaptation levels

In vertebrate retinas, the absorption oflight by visual pigment in photoreceptor

1-1 Components and Origins of the ERG 3

minals. This modulation of neurotransmit-ter release in turn causes a depolarization orhyperpolarization of the postsynaptic bipo-lar and horizontal cells. Mainly as a conse-quence of bipolar-cell depolarization, an in-crease in extracellular potassium, primarilyin the postreceptoral outer plexiform layer,causes a depolarization of Muller (glial)cells.8 The resulting transretinal currentflowing along the length of the radially ori-ented Muller cells appears to contributesubstantially to the corneal positive b-waveof the clinical ERG.9-11 A more proximal in-crease in extracellular potassium that occu'1sat the level of the inner plexiform layer fol-lowing light stimulation appears to derivefrom depolarization of amacrine, bipolar,and ganglion cells.12 The more distal in-crease in extracellular potassium in theouter plexiform layer contributes apprecia-bly more to the extracellular current, andthus to the b-wave, than does the proximalpotassium increase.13

A decrease in extracellular potassiumaround photoreceptor-cell outer segmentsfollowing light absorption also alters astanding electrical potential that exists be-tween the apical and basal surfaces of reti-nal pigment epithelial (RPE) cells. A flashof light induces a transient hyperpolariza-tion at the apical surface of RPE cells and ahyperpolarization of Muller cells, whichcombine to form a monophasic, corneal pos-itive deflection that follows the b-wave, re-ferred to as the (-wave. Thus, the c-waverepresents the algebraic summation of acorneal positive component, resulting froma change in the transepithelial potential in-duced by hyperpolarization at the apicalmembrane of the RPE cells, and the

outer segments initiates a sequence of mo-lecular events localized to these segmentsthat generates a wave of hyperpolarizationof the photoreceptors. In rod photorecep-tors, the photo-activated visual pigmentrhodopsin activates a protein, called trans-ducin (or G-protein), which in turn activatesthe enzyme cyclic guanosine monophos-phate phosphodiesterase (cGMP PDE), ul-timately resulting in a reduction in photore-ceptor-cell cGMP levels. This reduction incGMP levels causes the closure of sodiumion channels in the outer-segment mem-brane, which are normally permeable tothese ions in darkness. A hyperpolarization(negative change in intracellular electricalpotential) results from the light-induceddecrease of the inward-directed sodiumcurrent across the plasma membrane ofphotoreceptor outer segments (more pre-cisely, a decrease in sodium conductanceof the plasma membrane). The electricalchange thus generated can be measured asthe corneal negative a-wave of the ERG.

There is a suitable amount of evidencefrom different sources that the scotopic(dark-adapted) ERG a-wave reflects essen-tially solely rod photoreceptor-cell activity.By investigating the monkey photopicERG during the intravitreal administrationof glutamate analogs, Bush and Sieving7were able to demonstrate a proximal retinalcomponent in the primate photopic (light-adapted) ERG a-wave. They concludedthat at the flash intensities commonly usedto elicit the ERG a-wave in clinical diag-nostic testing, this component is likely tocontain a significant contribution from reti-nal activity postsynaptic to cone photOre-ceptor cells.

The light-induced hyperpolarization ofphotoreceptor cells diminishes the releaseof a neurotransmitter at their synaptic ter-

The Electroretinogram4

...

corneal negative component generated byhyperpolarization at the distal portion ofthe Muller cells, the slow PIll (see Figure1-lB).14 Diffuse degeneration or a malfunc-tion of RPE cells reduces the c-wave ampli-tude of the ERG. Similarly, diffuse photo-receptor-cell degeneration causes a reductionin ERG c-wave amplitude. Ample evidenceexists that the c-wave, although reflectingalteration of electrical activity at the level ofRPE cells, depends on rod receptor-cell ac-tivity. The c-wave has rod spectral sensitiv-ity, is lost when the retina is light-adapted,and is absent in cone-dominant retinas.15However, Tomita et al16 reported a cone-specific c-wave found in cone-dominantretinas.

The c-wave, which in humans is re-corded with a fully dilated pupil in a dark-adapted eye, generally peaks within 2 to10 s following the onset ofa flash stimulus,depending on flash intensity and duration.Both the c-wave amplitude and the timefrom stimulus to response peak increasewith stimulus intensity or stimulus dura-tion. Because the response develops overseveral seconds, it is subject to interfer-ence from electrode drift, eye movements,and blinks. Special recording electrodes,DC amplification (rather than AC amplifi-cation, which is used for standard ERGrecordings), and both high-luminance andextended-duration light stimuli are neces-sary for an optimal recording of the ERGc-wave.17 These procedural constraints, inaddition to the wide physiologic variabilityin c-wave amplitude and waveform (somenormal individuals appear to lack a record-able c-wave entirely), have limited the clin-ical application of c-wave measurements.

A study by Sieving et al18 on monkeyretina led them to conclude that the pho-topic ERG b-wave results from opposingfield potentials generated by depolarizingand hyperpolarizing bipolar and horizontalcells. A "push-pull model" was proposedwhereby depolarizing bipolar-cell activitymakes an essential contribution to theb-wave amplitude, which is appreciablymodified in amplitude and shape by electri-cal activity generated from hyperpolarizingbipolar/horizontal cells. The depolarizingand hyperpolarizing bipolar cells could par-ticipate in generating the photopic ERGb-waveform by modulating the extracellularpotassium that reaches the Muller cells.The b-wave voltage would then be gener-ated from depolarizing Muller cells as aresult of the changes in extracellularpotassium.

Subsequent investigationsl9-22 haveled to the conclusion that the total dark-adapted ERG in response to a brief flashstimulus can be considered as the sum ofphotoreceptor- and bipolar-cell compo-nents, although some contribution fromMuller cells is not precluded. The initialnegative portion (rod a-wave) of the dark-adapted response to relatively intense stim-uli, including the return to baseline, canbe considered as the sum of a negative-directed contribution from the photore-ceptors and a somewhat delayed, but ulti-mately larger positive-directed contribution(rod b-wave) from the rod on-bipolar cells.2The initial 15 ms of the a-wave seems toapproximately represent the photoreceptorresponses (see Figure l-lA).21 Rod recep-tors synapse selectively onto on-bipolarcells, which depolarize in response to light,but, unlike cone receptor cells, rods makeno direct contact onto off-bipolar cells.There are also negative inner-retinal com-

I-I Components and Origins of the ERG 5

likely generate the "true" (on-response)photopic b-wave by increasing the concen-tration of extracellular potassium. Photopicon- and off-pathway abnormalities in vari-ous retinal dystrophies were described bySieving.24 A small positive wave at the endof the descending portion of the ERGb-wave response to a high-Iuminance flashis also an off-response component, whichhas been designated as the "i-wave" to sug-gest that the character of the wave is the re-sult of interference between on- and off-response components.2.1 ...

Electrical events within ganglion cells oroptic nerve fibers do not contribute to theflash-elicited a- or b-waves of the ERG.Thus, disorders such as glaucoma and vari-ous types of optic atrophy, which selectivelyaffect ganglion cells and/or optic nerveaxons, do not ordinarily reduce ERG a- orb-wave amplitudes. Although isolated re-ports on both animals and humans showthat ERG amplitudes increase following

ponents that likely contribute to the dark-adapted a-wave at high stimulus intensities,including a probable contribution fromamacrine cells.2 In the cat, at maximalstimulus intensities, this can amount toabout one third of the overall dark-adapteda-wave amplitude.2

Not readily recognized is how apprecia-bly an off-response can contribute to thephotopic b-wave amplitude. The tradition-ally very brief strobe-flash stimuli used forERG recordings preclude isolation of thatportion of the off-response that is part ofthe photopic b-wave amplitude. However,with more prolonged flashes, it can bemade apparent that a substantial portion ofthe photopic b-wave is actually an off-re-sponse component, d-wave (Figure 1-2). Asindicated, hyperpolarizing bipolar cells,12probably in addition to other cells such asthe photoreceptors,23 likely generate theoff-response component of the b-wave am-plitude, while depolarizing bipolar cells

Figure 1.2 With longer-duration flashes, pho-topic b-wave can bedemonstrated to con-tain substantialoff-response ( d-wave);i-wave is also off-response component.COU11esy Neal S. Peachey,PhD.

d~/{' ~-- 25 msV-r--- 20 ms

~ -~--~ 15 ms i I-0-C"'

"'

i:i:

10 ms

5ms

0 20 40 60 80 100Time (ms)

120 140

6 The Electroretinogram

and 60 of photocoagulation caused de-creases in ERG amplitude that were pro-portionally related to the destroyed area.

Thus, normal full-field ERG recordingsmay be obtained in patients with markedimpairment in central vision resulting fromdisorders that affect the visual system at orcentral to the retinal ganglion cells or fromphotoreceptor-cell degeneration limited tothe fovea. Conversely, markedly reduced oreven nondetectable ERG amplitudes canbe found in the presence of 20/20 acuity, asseen in some cases of retinitis pigmentosa.

1-1-2 Scotopic and PhotopicThreshold Responses

The scotopic threshold response (STR) is asmall, negative-directed ERG response thatcan be recorded after a prolonged period ofdark adaptation with the use of very dimlight stimuli.32.33 This response is observedwith light stimuli too dim to elicit conepathway signals or a rod ERG b-wave. TheSTR appears to originate from electrical ac-tivity in retinal amacrine cells, which re-lease potassium upon light stimulation andcause depolarization of Muller cells.34 Inpatients with juvenile X-Iinked retinoschi-sis, a disease in which pathologic changes inMuller cells have been observed,35 the STRis absent.36 This observation implies Mullercell, as well as amacrine cell, involvementin the origin of the STR. Although the clin-ical diagnostic value of the STR has yet tobe fully defined, it is said to be absent insome forms of congenital stationary nightblindness and is diminished in glaucoma.34

A somewhat similar, delayed corneal neg-ative response, elicited from cones underphotopic conditions, has been termed thephotopic negative response (PhNR).37 The am-plitude of this response was found to be re-duced in a macaque model of experimental

section of the intracranial portion of theoptic nerve,26,27 this has not been a consis-tent finding.28 Because the ERG b-wavenecessarily depends on electrochemicalevents that generate the ERG a-wave, anyretinal disorder that prevents generation ofa normal a-wave will also affect the devel-opment of a normal b-wave, Examples in-clude retinitis pigmentosa, retinal detach-ment, and ophthalmic artery occlusion. Theconverse, however, is not true. Disordersthat result in a diffuse degeneration or dys-function of cells in the inner nuclear layer(Muller or bipolar cells) can selectively de-crease the ERG b-wave amplitude withoutdiminishing the ERG a-wave. A notable ex-ample is central retinal artery occlusion,where the choroidal circulation maintainsblood flow to the photoreceptor cells andsustains the biochemical events that gener-ate the a-wave.

To reduce a- and/or b-wave amplitudes,a disorder must affect a large area of retinaltissue. Th us, focal lesions of the fovea ( de-fined as a region the approximate size ofthe optic disc, 1.5 mm in diameter, cen-tered at the foveola) do not affect the a- andb-wave amplitudes elicited bya full-fieldflash stimulus. The work of Armington etal29 suggests that a loss of one half the pho-toreceptors across the entire retina may re-sult in approximately a 50% reduction inERG amplitude. Fran~ois and de Rouck30observed that a macular lesion, up to a sizeof 3 disc diameters, produced by photOCO-agulation did not modify ERG amplitudes.This finding was in accord with the investi-gations by Schuurmans et al,31 who demon-strated in rabbits that photocoagulation of a20 retinal area of the posterior pole did notdecrease the ERG amplitude. Between 20

1-1 Components and Origins of the ERG 7

glaucoma.37 Ganglion cells or their axons,and possibly amacrine cells, are involved ingenerating the response. This tlash-evokedelectrical signal is similar to the photopicthreshold response (PTR) described bySpileers et al.38

1.1.3 Early Receptor Potential

The early receptor potential (ERP) is arapid, transient waveform recorded almostimmediately after a light-flash stimulus,particularly after one of high intensity in adark-adapted eye. The response, which iscomplete within 1.5 ms, originates from thebleaching of photopigments at the level ofthe photoreceptor outer segments. This po-tential, first described by Brown and Mu-rakami39 in 1964, was found by Pak andCone40 to consist of two components, desig-nated R1 and Rz. The initial, corneal posi-tive portion (R1) has a peak time of approxi-mately 100 microseconds (Jls) and appearsto be primarily a cone response. The sec-ond, corneal negative component (Rz) has apeak time of approximately 900 JlS and con-sists of contributions from both rods andcones. This corneal negative componentis essentially complete by the time thea-wave begins (Figure 1-3). Both positiveand negative components of the ERP resistanoxia. The initial, positive phase can stillbe observed in the frozen eye, while thesecond, negative phase disappears on cool-ing. This potential appears to reflect molec-ular events within photoreceptor visualpigments and corresponds to their photo-chemical kinetics. The corneal positivecomponent, R1, has been associated withthe conversion of lumirhodopsin to meta-rhodopsin I during the bleaching of visualpigment, while the corneal negative com-ponent, Rz, is generated by the conversionof metarhodopsin I to metarhodopsin II.

a.wavedescending limb

-#

Figure 1.3 Note positive (+) Rl and negative (-) Rzportions of ERR ERP occurs immediately after high-luminance stimulus. Its negative portion blends witha-wave of ERG.

8 The Electroretinogram

The ERP is thought to originate as theresult of charge displacements that occurwithin photoreceptor outer segments dur-ing the photochemical reactions describedabove. Its recovery rate has been correlatedwith the regeneration rates of these visualpigments. The human ERP is probablygenerated predominantly by the cones. Byevaluation of cone-deficient subjects, therod contribution to the ERP has been esti-mated to be between 20% and 49%.41.42This response in humans is technically con-siderably more difficult to record than theERG, and it is therefore not seen in routineERG recordings. It requires the use of ahigh-intensity flash stimulus, the deliveryof the stimulus to the eye preferably inMaxwellian view (requiring the use of acondensing lens and careful optical focus-ing), and isolation of the recording elec-trode from the flash stimulus to avoid aphotovoltaic artifact that would obviaterecording of the ERP response. Measure-ment of the ERP can be useful in the studyof human retinal disorders that cause pho-toreceptor-cell degeneration because it re-flects a direct measure of visual photopig-ment activity. Its amplitude is proportionalto the quantity of bleached pigment mole-cules. A review of the ERP and its meas-urement in various human retinal disorderscan be found in Muller and Topke.43

1.1.4 Oscillatory Potentials

In 1954, Cobb and Morton44 first reportedthe presence of a series of oscillatorywavelets superimposed on the ascendinglimb of the ERG b-wave after stimulation

by an intense light flash. Yonernura45 subse-quently termed these wavelets oscillatorypotentials (OPs). These potentials are high-frequency, low-amplitude components ofthe ERG, with a frequency of about 100 to160 Hz, to which both rod and cone sys-tems can contribute.46,47 By comparison, thea- and b-waves are dominated by frequencycomponents of about 25 Hz.46

The cellular origin of OPs in the retinais somewhat uncertain, although it is likelythey are generated by cellular elementsother than those that generate the a- andb-waves. Current information supports theconclusion that cells of the inner retina,supplied by the retinal circulation, such asthe amacrine or possibly interplexiformcells, are the generators of these potentials.lntravitreal injection of glycine, which pro-duces morphologic changes in amacrinecells,48 results in a loss of OPs. Drugs thatserve as antagonists to the inhibitory neuro-transmitters gamma-aminobutyric acid(GABA), glycine, and dopamine can selec-tively reduce OPS.49,50

Reduction in amplitude of OPs becomesapparent in the presence of retinal ische-mia, such as that seen in patients with dia-betic retinopathy, central retinal vein occlu-sion, and sickle cell retinopathy. Reductionin or amplitudes has also been reported inpatients with X-Iinked juvenile retinoschi-siS,51 in some patients with congenital sta-tionary night blindness,52 and in patientswith Beh

1-1 Components and Origins of the ERG 9

A Blue-light stimulus White-Iight stimulus

0:Oin>

'6~

0L{)N

0:Oij)>

~

C)"'

"'

~

B Rod OPs Cone OPs

2cov;>

'6

'f..""

N

0=Ou;

.>'6>

"-

LO"'

...

0 80 020 40 60Time (ms)

100 20 40 60

Time (ms)

80 100

Figure 1-4 ( A) Single-flash and (8) OP responses toblue- and white-Iight stimuli in dark-adapted eye.OPs to blue stimulus are from rod system, whilethose to white stimulus are from cone system.Courtesy Neal S. Pealhey, PhD.

ther, a "conditioning" flash should be pre-sented to a dark-adapted eye approximately15 to 30 s prior to averaging a subsequentset of three or four responses to additionalflashes presented at about 15-s intervals.Responses obtained under these stimulusconditions are from the cone system (theconditioning flash having adapted the rodsystem so that it does not contribute to thesubsequent flash-elicited OPS).47 OPs fromthe rod system can be obtained with low-intensity blue light (Figure 1-4).

Although OPs are generated by cellsof the inner retina, they will be reducedin amplitude by disorders that affect outer,more distal retinal cells, because the distalcells provide electrical signals that com-prise the input to the more proximal cellsthat generate the OPs. Thus, diseases thatseem to affect primarily the outer retina,such as cone dystrophy or retinitis pigmen-tosa, reduce OPs as well as a- and b-waveamplitudes.

10 The Electroretinogram

BA

Figure 1-5 Methods of measuring ERG ( A) waveformamplitudes and (B) time relationships. Most examin-ers measure b-wave amplitude from trough of a-waveto peak of b-wave. Latency of response refers to timefrom stimulus onset to beginning of a-wave, while im-plicit time ( b-wave response illustrated) is measuredfrom stimulus onset to peak of b-wave.

MEASUREMENT OF THE ERG COMPONENTS

An evaluation of ERG components in-cludes the measurement of both ampli-tude and timing characteristics. Latency( the time from the onset of the stimulusuntil the beginning of ~he response), im-plicit time (the time from the onset of thelight stimulus until the peak amplitude re-sponse), and amplitude are all depicted inFigure 1-5. Note that the a-wave peak am-plitude is measured from the baseline tothe trough of the a-wave, while the b-waveis traditionally measured from the trough ofthe a-wave to the peak of the b-wave. Ifonly the a- and b-waves are included, theduration of the ERG response is less than0.25 s. Table 1-1lists the author's normalranges of a- and b-wave amplitudes and im-plicit times to single-tlash stimuli underscotopic and photopic conditions. Thesevalues will vary with the intensity of thelight stimulus, the state of retinal adapta-tion, and several other variables (describedin Sections 1-5 and 1-6).

1-3 Recording Procedure 11

TABLE 1-1

Range of ERG Values Obtained on 100 Normal Control SubjectsFrom Authors Electrophysiology Laboratory

MaximalDark-Adapted(Scotopic)273-684 ~V

12-15 ms

489-908 ~V

33-54 ms

Rod-lsolated(Scotopic)(Photopic)

70-137 ~V

12-15 ms

132-320 ~V

31-38 ms

Light-Adapted32.Hz Flicker

a-wave amplitudea-wave implicit timeb-wave amplitude

b-wave implicit time273-684 ~V

56-70 ms74-137IJV22-32 ms

-#

The ERG amplitude responses from theright and left eyes of normal subjects undercontrolled standard conditions using a full-field stimulus are usually within 10% ofeach other, with a maximal difference of ap-proximately 20% to 24%. The 24% repre-sents the maximal difference caused byshifts in the position of a contact lens evenunder controlled conditions. Thus, a differ-ence in amplitude between the two eyes isprobably pathologic if it is distinguishableby between 20% and 24% and likely to bepathologic if it exceeds 24%. Some investi-gators indicate that a reduction in ERG am-plitude during serial followup examinationsis significant, at the 99% confidence limit,if a decline greater than 31% occurs to asingle-flash white stimulus or greater than44% to a 30-Hz flicker stimulus.54

Because actual ERG amplitudes, as meas-ured in microvolts, do not follow a normalgaussian distribution, nonparametric meth-ods are likely most suitable for obtaining arange of normal results for clinical testing.55Converting ERG amplitudes into log valuesbetter simulates a normal distribution ofamplitude data.55

RECORDING PROCEDURE

The most basic aspect of obtaining an ERGis an adequate light stimulus that allowsvariation in stimulus intensity over a rangesuitable for clinical electroretinography.The specific stimulus is probably less im-portant for clinical use than is the ability tocalibrate the stimulus and maintain it at aconstant luminance. Many examiners use aGrass xenon-arc photostimulator, with aflash duration of 10 ].1s.

A contact-Iens electrode with a lid spec-ulum is most often used to record retinalresponses to the light stimulus. A topicalanesthetic is administered to the eye to re-duce any difficulty and discomfort associ-ated with lens insertion. A ground electrodeis generally placed on the patient's earlobewith electrode paste. A reference, or "inac-tive," electrode is placed centrally on thepatient's forehead slightly above the supra-orbital rims or on the mastoid region or ear-lobe. The reference electrode serves as the

12 The Electroretinogram

more negative pole, having been placedcloser to the electrically negative posteriorpole of the eye. (In the Burian-Allen bipo-lar electrode, discussed in Section 1-4-1, asilver reference electrode is incorporatedwithin the lid speculum, obviating the needfor a separate reference electrode.)

The corneal contact lens (active elec-trode ), ground electrode, and referenceelectrode all connect with a junctional box,from which the signals are delivered to ad-ditional recording components for amplifi-cation and, finally, display. Most often, dis-play and amplification systems are con-tained within a single unit. A differentialamplifier is used to detect and amplify thedifference in voltage between the activecorneal electrode and the inactive referenceelectrode that develops following lightstimulation of the retina.

controlling, or at least limiting, the effectsof blinking and lid closure during photicstimulation. Images viewed through thecentral portion of the lens, made of poly-methylmethacrylate (PMMA), show vari-able degrees of blur. Thus, the lens is notoptimal for use when stimulus image clarityis vital, for example, when recording thepattern ERG.

The recording electrode consists of anannular ring of stainless steel surroundingthe central PMMA contact-Iens core. In thebipolar version of the Burian-Allen lens, aconductive coating of silver granules sus-pended in polymerized plastic is painted onthe outer surface of the lid speculum. Thisconductive coating serves as a referenceelectrode, obviating the need for an addi-tional separate silver-silver chloride wire, asis necessary with the unipolar model.57 It isnoteworthy that the ERG amplitudes ob-tained with a bipolar electrode will be pre-dictably smaller than those obtained witha monopolar electrode.RECORDING ELECTRODES

Important features of ERG recording elec-trodes include

1. Quality components with low intrinsicnoise levels, which facilitate stability of re-sponses2. Patient tolerance, with limited irritationof the corneal surface

3. Availability at a reasonable cost

1-4-1 Burian-Allen Electrode

The Burian-Allen electrode has lens sizessuitable for adults to premature infants.56These lenses have the advantage of a lidspeculum, which promotes a more consis-tent input of light entering the pupil by

1-4-2 Dawson-Trick-Litzkow Electrode

The Dawson-l'rick-Litzkow (DTL) elec-trode consists of a low-mass conductiveMylar thread whose individual fibers (ap-proximately 50 pm in diameter) are impreg-nated with metallic silver. The conductivethread virtually floats on the corneal filmsurface.58

When compared to the Burian-Allenelectrode, the amplitudes of the ERG re-sponses recorded with the DTL electrodeare, on average, reduced 10% to 13%. How-ever, the variance in signal amplitude withthe DTL electrode was found to be lessthan with the Burian-Allen electrode. In ad-dition, the noise characteristics of the DTLelectrode were favorable, as was its toler-ance for extended recording periods.59

..

1-4 Recording Electrodes 13

tact-Iens-style recording electrodes. Both ofthese electrodes are probably unsuitable forDC amplification measurements because ofrather large low-frequency drift. Borda etal65 reported a 34% to 44% reduction in am-plitude with a gold-coated Mylar electrodethan with a contact-lens electrode.

Major advantages of the DTL electrodeinclude its low cost, patient acceptance, andstimulus image clarity. However, eyelidblinks following flashes can cause move-ment artifacts, which will sometimes ob-scure ERG components. Proper placementof the DTL electrode thread deep andloose within the lower-Iid conjunctival saccan minimize electrode displacement fromeye blinks or eye movements and thusallow highly reproducible retinal potentialsto be recorded.6 While facilitating greaterreproducibility, positioning the electrodedeep within the conjunctival sac, as op-posed to having it ride on the lower-Iid mar-gin, results in the measurement of apprecia-bly lower ERG amplitudes.61

143 ERG-Jet Electrode

The ERG-jet electrode lens is made of aplastic material that is gold-plated at theperipheral circumference of its concave surface.62 Its advantages include sterility (thelens is packaged and can be discarded aftereach use), simplicity in design, and ease ofinsertion, which has potential benefit forsensitive eyes. Movement of the lens dur-ing recordings and absence of a lid specu-lum could contribute to variability ofrecordings in individual instances.

145 Skin Electrode

For specific purposes-for instance, record-ing ERGs from infants and young children-the corneal electrode can be replaced by anelectrode placed on the skin over the infra,.orbital ridge near the lower eyelid. Recordedamplitudes are 10 to 100 times smaller thanthose obtained using a corneal electrode.66.67Coupland and Janaky68 observed that theERG amplitudes recorded with skin elec-trodes ranged from between 43% and 73%of those obtained with DTL electrodes.Both the lower amplitudes and the variabil-ity in responses with the use of skin elec-trodes preclude their use for purposes otherthan screening. Whenever possible, it ispreferable to select a smaller recordingelectrode with a lid speculum, such as aninfant Burian-Allen electrode, which can beused successfully in children and infants.

146 Cotton-Wick Electrode

Sieving et al69 introduced an electrode con-sisting of a Burian-Allen electrode shell fit-ted with a cotton wick. This electrode, aswell as earlier variants of a cotton-wick con-figuration, is of value in recording ERG am.plitudes to a high-luminance flash and forrecording the ERP, because the lens is freeof a photovoltaic artifact, which can occurwhen intense light illuminates the metalpresent in other recording electrodes.

1-4-4 Mylar Electrode

Aluminized or gold-coated Mylar has alsobeen used for recording ERG potentials.63.64The gold-foil electrode is preferable to thealuminum Mylar because it provides a morestable baseline and is less likely to dislodgeor flake.65 The gold-foil electrode, whichfits over the lower eyelid, when correctlypositioned, is less sensitive to eye move-ment than is the DTL electrode, yet moresensitive to such movement than are con-

14 The Electroretinogram

147 Hawlina-Konec Electrode

The Hawlina-Konec (HK) loop electrode isa noncorneal electrode, consisting of a thinmetal wire (silver, gold, or platinum) that ismolded to fit into the lower conjunctivalsac. The wire is Teflon-insulated, exceptin its central region, where three windows,3 mm in length, are formed on one side.Electrical contact is made with the scleralconjunctiva through these small uninsu-lated portions. The HK electrode may beused without topical anesthetic, althoughthis may not be advisable in sensitive pa-tients or certain young children. PatternERG recordings with this electrode are inthe same amplitude range as observed withthe gold-foil electrode, while flash ERGamplitudes are about two thirds thoserecordable with corneal electrodes.

THE ERG UNDER LIGHT -AND DARK-ADAPTED CONDITIONS

The recording of the photopic (cone) ERGpotential can be done before or after darkadaptation, with the patient light-adaptedto a background light intensity of at least5 to 10 foot-lamberts, or 17 to 34 cd.m-2(candela per meter squared), which facili-tates the recording of a response exclu-sively from the cone systein.70 Regardlessof which sequence is adopted, the responseobtained to a high-Iuminance stimulus(Figure 1-6) is relatively small in amplitude,with a short implicit time and a brief dura-tion. The amplitude of the cone b-wavegrows with stimulus intensity and tends toapproach a maximal limiting value at the

highest intensities. The cone b-wave im-plicit time increases slightly with increasingstimulus luminance.71

If single-flash or 30-Hz flicker cone re-sponses are obtained after a period of darkadaptation, it is vital to wait approximately10 to 12 min for light adaptation before ob-taining final photopic amplitudes, becausethere is a progressive increase in ERG am-plitude, the extent of which depends onthe level of the background adapting lightand the stimulus intensity.71-78 More in-tense backgrounds and stimuli will be asso-ciated with greater increases in cone ERGamplitudes during the period of light adap-tation. With certain stimulus and back-ground intensities, amplitudes can approxi-mately double (Figure 1-7). A progressivedecrease in cone b-wave implicit time mayalso be noted during the course of lightadaptation.76 In normal individuals, a rod-cone interaction appears to influence thelight-adapted cone b-wave implicit time.Light adaptation of the rods shortens theimplicit time of the cone b-wave.79

Figure 1-6 also shows scotopic responseselicited with a white-Iight stimulus of highluminance after 30 min of dark adaptation.Note the increase in ERG amplitude andimplicit time as compared to the light-adapted recording. This response, althoughevoked under dark-adapted conditions, hasboth rod and cone components. However,the rod component is dominant and is themajor contributor to both the increased am-plitude and the increased implicit time.This is understandable because the rodpopulation is appreciably greater than thatof the cones (about 17 rods to 1 cone) andthe rod cells are intrinsically more sensitiveto light. An isolated rod response can beevoked by a low-intensity short-wavelength(blue) stimulus (see Figure 1-6).

1-5 The ERG Under Light- and Dark-Adapted Conditions 15

Normal ERG

Photopic white

5-min scotopic white

15-min scotopic white

Figure 1-6 Normal ERGresponses under pho-topic and scotopic con-ditions. Cone responsesare generally isolatedunder proper photopicconditions and with ei-ther single-/lash or 30-Hz flicker stimulus,while rod function canbe isolated with use oflow-Iuminance, shot1-wavelength (blue) stim-ulus as either single ...flash or flicker at 10Hz. High-Iuminancewhite-Iight flash underdark-adapted condi-tions measures combinedrod and cone response,which, in normal subject,is predominantly fromrod system. lirms 14and 18 refer to stimulusintensity. In thisfigure,as well as in subsequentfigures, time calibrationof 20 ms refers only tosingle-/lash recordings.Vet1icalline indicatesstimulus onset.

30-min scotopic white

30-min scotopic blue

14 blue flicker 10 Hz

18 red flicker 30 Hl

16 The Electroretinogram

B

c:Oin>

'6~

a

."'

"0E":9-

0

t~"05.c

160

Figure 1-7 (A) Light-adapted single1lash white and(B) -,O-Hz white flicker cone responses at 1 and 20min of light adaptation subsequent to period of ap-proximately 30 min of dark adaptation. Note im-portance of allowingfor suitable period of light adaptation to properly determine maximal cone amplitu&responses. Also note changes in implicit time duringlight adaptation.Courtesy Neal S. Peachey, PhD.

Rods and cones contribute independ-ently to the scotopic ERG amplitude. Thatis, the rod contribution to the scotopicb-wave is the same whether the cone con-tribution is present or absent. As will beemphasized in later sections, excluding theinfluence of retinal or choroidal disease,those recording conditions that determinethe relative cone and rod contributions tothe ERG response include the intensity,wavelength, and frequency of the lightstimulus, in addition to the state of retinaladaptation. Therefore, a comprehensiveevaluation of retinal function includes ob-taining ERG responses under testing conditions that include

1,

2,

3,

4,

Evaluation of the ERG with a constantstimulus intensity during dark adaptationEvaluation of the dark-adapted retinawith various stimulus intensitiesThe ERG response to stimuli of differentwavelengthsThe ERG response to different stimulusfrequencies

1-5 The ERG Under Light- and Dark-Adapted Conditions 17

The ERG response with the retina light-adapted is also part of the evaluation. 100,.V L

20 ms

bl

1 min

at

6min

-#

10min

15 min

1-5-1 -' --::::" -:~ J Figure 1-8 is a schematic representation of

the ERG during dark adaptation with theuse of a high-Iuminance flash. Traditionally,the b-wave includes two subcomponents,bl and bz. Note the immediate increase inamplitude of b1 from that seen under pho-topic conditions; it occurs with less than1 min of dark adaptation. The initial growthof b1 amplitude is likely related to a "neu-ral" network adaptation of primarily the rodresponse to the removal of the adaptinglight used under photopic conditions and/orto a receptor adaptation within the outersegments. Subsequently, both b1 (a com-bined cone and rod response when elicitedby a moderate or high-intensity flash stimu-Ius) and bz (a presumed rod component ofthe b-wave) increase in amplitude as thephotoreceptor cells dark-adapt and becomeprogressively more sensitive.

The rate at which bz develops relates tothe duration and intensity of previous lightexposure. The major increase in amplitudeof b1, as well as the entire development ofbz, occurs because of the increase in rod-re-ceptor response. Note that the implicit timeof both a- and b-waves also increases as theERG changes from a light-adapted to a dark-adapted response. Also note the presence ofa biphasic a-wave with the high-Iuminanceflash, as well as the appearance of OPs onthe ascending limb of the b-wave duringprogressive dark adaptation. It is likely thatan OP divides the a-wave into its a1 and azcomponents.80 The a-wave amplitude alsoprogressively increases during dark adapta-tion as photoreceptor cells gradually recovertheir sensitivity.

25 min

82

Figure 1-8 ERG responses during dark adaptationwith use of high-Iuminance stimulus. Increases in botha- and b-wave amplitudes with progressive darkadaptation are apparent; bl wave has both rod- andcone-system components, while b2 wave is likely exclu-sively rod-system component. OPs are also more ap-parent with progressive dark adaptation.

18 The Electroretinogram

kxaminers need to be aware of an artifactthat may be present, particularly with theuse of a bipolar recording electrode, in theclinical ERG of some patients. This artifactoccurs with a latency that is fast enough tointerfere with the scotopic b-wave of theERG. Because most of the artifact is proba-bly due to a reflex contraction of the orbic-ularis muscle, it has been termed the photo-myoclonic reflex. 81 This artifact is less likelyto be apparent with the use of unipolar thanwith bipolar lenses.

1-5-2 With Variable-lntensity Stimuli

Figure 1-9 shows dark-adapted ERG re-sponses to flash stimuli of minimal, moder-ate, and high luminance. Generally, the am-plitudes of both a- and b-waves increaseand the implicit times decrease as a func-tion of stimulus luminance; in both cases,the curves approach a plateau at high stim-ulus intensities. It is noteworthy that, as al-ready indicated, the b-wave implicit timeappears to increase with stimulus intensityunder light-adapted conditions. As a refer-ence, a standard flash stimulus strength hasbeen defined as one that produces at least1.5 to 3.0 cds.m-2 (candela-second permeter squared) at the surface of a full-fieldbowl (3.43 cds.m-2 = 1 foot-lambert).7o

Figure 1-10 shows a series of ERG re-sponses to flashes of increasing intensity.The lowest-intensity flashes elicit an ex-clusively rod response, while high- andmedium-intensity stimuli evoke responsesthat, as already noted, are rod-dominant,but contain both rod and cone components.Note the absence of the a-wave at thelower stimulus intensities. Because there isa large magnification of electrical activity assignals are transmitted from photoreceptorcells to inner retinal neurons, the b-wave

Figure 1-9 ERG responses in dark-adapted eye withstimuli of different white-Iight intensities. Note shot1erimplicit times and larger amplitudes as stimulus in-tensity is increased. OPs are most apparent with high-intensity stimulus.

1-5 The ERG Under Light- and Dark-Adapted Conditions 19

-0.5250 "V L

20ms

-1.0

:J

appears to be at least 1 log unit more sensi-tive than the a-wave. Thus, b-wave ampli-tudes can be measured at low stimulus in-tensities, where a-wave responses are notyet apparent.

ERG investigations of patients with reti-nal disorders now more frequently includestudies of scotopic b-wave responses to arange of full-field stimulus intensities toobtain a stimulus/response (S/R), or Naka-Rushton-type, function. The normal S/Rfunction (Figure 1-11) exhibits an increas-ing amplitude over a stimulus range ofabout 2 log units. The S/R relationship canbe summarized by the equation

NE0;"8"iI90)u~(0~

"E.2-5j(0Lr:

R/Rmax = 1"/(1" + a")

-2.5where R is the b-wave amplitude producedby a flash of intensity I (typically, in foot-lamberts per second or cds.m-2) and

20 The Electroretinogram

600Figure 1-11 Flash stimu-Ius versus amplitude response function curvewith equation used todetermine curve fit.

-Rmax500

400

~

"'

-0~ 300""-

E

~

0~

"'

"0~

0-E

'0

~

0~

0>"0E-0.E

=1.

8~

0)"0oE"5.E

."

0 40 80 120 OTime (ms;

40 80 120

B

120 80 120 1600 40 80 160 0 40Time (ms)

1-17 Toxic Conditions 95

With time, the b-wave may slightly increasein amplitude, although rarely to normal val-ues. Moloney et al481 emphasized that coneERG function remains more impaired thanrod function. Generally, the a-wave andthe EGG stay essentially normal, althoughFran

96 The Electroretinogram

Figure 1.64 ERG record-ings under ( A) photopicand (B) scotopic condi-tions in patient withmethano!-induced reti-na! toxicity. Note pre-dominant reductionin b-wave amp!itude,giving negative ERGwaveform.

1-17 Toxic Conditions 97

retinal toxicity occurs has differed betweenreports.491-493 Palifieris et al492 found an in-travitreal injection of as little as 0.1 fig tobe toxic to the rabbit retina.

Kupersmith et al496 described the de-velopment in 3 patients of a maculopathywith both hypopigmentation and hyperpig-mentation in the eye ipsilateral to an intra-arterial infusion of cisplatin. The 8 patientstreated also received an intravenously ad-ministered regimen of carmustine. A dif-fuse-flash ERG showed cone and rod re-sponses that were reported as being withinthe normal range, as was the EOG recordedin 1 patient.

1-17-8 Ethyl-m-Aminobenzoic AcidMethanesulfonate

Ethyl-m-aminobenzoic acid methanesul-fonate (MS-222) is a veterinary anestheticcommonly used for immobilizing fish oramphibians. The finding of retinal toxicitywas observed in an ichthyologist who hadchronic occupational skin exposure to thesubstance.494 Exposure to the drug was as-sociated with decreased vision, photOpho-bia, and photopsia. The retina appearedgrossly normal, with no evidence of pig-mentary change. ERG photopic and sco-topic amplitudes were notably reduced.The b-wave amplitude to a high-intensitystimulus showed a greater reduction com-pared to the a-wave amplitude. OP ampli-tudes were also reduced. After discontinua-tion of contact with MS-222 for 7 months,vision returned to normal and ERG ampli-tudes improved.

1.17.10 Glycine

Creel et al497 reported photopically ob-tained ERG abnormalities in 4 patientswho had undergone transurethral resectionof the prostate when glycine was used as anirrigating fluid. These abnormalities in-cluded a reduction in b-wave amplitude,loss of OPs, and marked reduction in the30-Hz flicker amplitude. The patients ex-perienced a temporary, but marked reduc-tion in visual acuity. The fundus, on exami-nation, was normal. Glycine is known tofunction as an inhibitory neurotransmitter.During prostatectomy, excessive absorptionof the glycine irrigating solution can lead tomarked elevation of serum glycine, withpossible consequences to the visual systemin some patients.

1-17-11 Canthaxanthin

Canthaxanthin is a carotinoid pigment thathas been used as both a food coloring andan oral tanning agent. The ingestion of thissubstance can result in the development ofcrystalline deposits in the retina. Arden etal498 noted the development of a generallysmall, but discernible reduction in the sco-topic and, to a distinctly less noticeable ex-

1.17.9 Cisplatin

Marmor495 reported a negative-type ERGwaveform in a 68-year-old woman withovarian cancer who inadvertently receivedan overdose ( 480 mg) of cisplatin (Platinol),an antineoplastic agent. A negative-typescotopic response, with a normal a-waveamplitude but reduced b-wave, was ob-served. A reduction in the cone b-waveamplitude, as well as scotopically obtainedOPs, Was also observed. Using an extended-duration flash procedure, on-responses wereobserved to be markedly reduced, whileoff-responses were essentially normal. Tox-icity from cisplatin is apparently another con-dition that can cause a negative-type ERG.

98 The Electroretinogram

tent, photopic b-wave amplitude, whichwas dose-related and most notable withhigher-intensity stimuli. These authorsspeculated that the ERG b-wave changewas likely due to a direct effect of this com-pound on Muller cells. Hamois et al499 ob-served that canthaxanthin retinopathy maybe at least partly reversible after the drug isdiscontinued.

1-17-13 Deferoxamine

Deferoxamine (also known as desferriox-amine) is a chelating agent used to treatiron storage disorders, like hemosiderosis,which can result from multiple transfusionssuch as those used in patients afflicted withr3-thalassemia major. The use of, initially,intramuscular or intravenous and, more re-cently, subcutaneous infusion has beenshown to increase the urinary iron excretionsubstantially in these patients. Deferoxa-mine (Desferal) has also been used in pa-tients with rheumatoid arthritis, on thepremise that iron deposits in synovial mem-branes may be contributing to the inflam-matory response and the occurrence of ane-mia in these patients. By removing ironfrom the synovial membranes, deferoxam-ine may decrease the inflammation, replen-ish bone-marrow iron stores, and lessen theanemia.505 The drug may also playa role byinhibiting iron-promoted peroxidation fromfree radicals generated by phagocytic cellsin the inflamed rheumatoid joint.506,507

Ocular signs and symptoms that canoccur with the use of this substance includeblurred or decreased visual acuity, impairedcolor vision, and night blindness. Opticneuropathy, pigmentary retinopathy, or, in-frequently, both have been described. Thedevelopment of cataracts, as well as centraland peripheral visual field defects, has alsobeen reported.506-510 Hearing loss can alsooccur with the use of this drug.510 The reti-nal pigmentary changes involve a stippled,mottled, or "salt-and-pepper-like" appear-ance of the macula, resulting from a mosaicpattern of depigmentation and hyperpig-mentation, as well as areas of pigment clump-ing in the midperiphery. It is relevant that,on occasion, pigmentary retinal changes canoccur in patients who are visually asympto-matic.510 Optic nerve atrophy may become

1-11-12 Vigabatrin

Vigabatrin is an antiepileptic drug that in-creases brain gamma-aminobutyric acid(GABA) by inhibiting GABA transaminase.GABA is an inhibitory neurotransmitter atdifferent postreceptoral retinal sites. Itplays a role in the regulation of horizontal-cell coupling. An accumulation of GABAcan be found in amacrine cells.

Bilateral concentric visual field restric-tion and blurred vision have been describedwith the use of this drug.soO-so3 ERG record-ings show retinal cone-system dysfunction,including a prolonged b-wave implicit timeand preferentially reduced b-wave ampli-tude. Cone a-wave and rod function havebeen reported as normal. However, mark-edly reduced or non detectable cone OPsare reported, consistent with impairment ofhighly GABA-ergic amacrine cells.sol,SO2Oaneshvar et also4 observed ERG abnor-malities in 4 of 10 patients. AI14 showed re-duced b-wave amplitudes to scotopic test-ing. Only 1 patient showed a reduction ofthe 30-Hz photopic flicker ERG, while OPswere normal in all 10 patients. No delays inimplicit time were observed in either pho-topic or scotopic responses. EGG light-peak to dark-trough ratios were reported aseither normalSO1,SO4 or abnormal.so4

1-18 Vitamin A Deficienry and Retinoids 99

observed in either the isolated rod responseor the light-adapted cone b-wave ampli-tude. Implicit times for both dark- andlight-adapted responses were reported tobe normal.

In their study of 6 patients who ingested100 mg of sildenafil, Kretschmann et alS16observed an increase in the dark-adaptedrod b-wave implicit time, with a normal am-plitude 1 to 3 hours after ingestion. Themaximal dark- and light-adapted b-waveamplitudes, as well as the implicit times,were unchanged from baseline. A large number of such patients, treated with dif-ferent dosages, need to be investigated be-fore firm conclusions can be drawn aboutthe type and frequency of ERG changesthat occur with the use of sildenafil. Fur-thermore, the long-term consequences ofthe use of this drug, if any, need to befirmly established.

clinically apparent as a consequence of theoptic neuropathy.

Most significant is that when deferoxa-mine is discontinued, most patients experi-ence either a complete or a partial improve-ment of their visual function.sO6-S13 N ever-theless, a progression in the pigmentaryretinal changes without further deteriora-tion of visual function was observed tooccur despite discontinuation of the drug.so9

Abnormal ERG amplitudes and reducedEOG light-peak to dark-trough ratios havebeen reported.SO6-S09.S11,SI2 Improvementin these measures of visual function areoften seen when deferoxamine is discon-tinued. ERG rod function tends to be moreimpaired than cone function in some pa-tients.so7 Prolonged ERG implicit times, inaddition to reduced amplitudes, have beenobserved.so9

1-17-14 Sildenafil

Sildenafil citrate (Viagra) is an oral treatmentfor erectile dysfunction. The drug acts as arelatively selective inhibitor of phosphodi-esterase type 5 (PDE5), an enzyme in thecorpus cavernosum of the penis that breaksdown cyclic guanosine monophosphate(CGMP).514 In vitro studies have shown thatsildenafil is approximately 10% effective asan inhibitor of the enzyme PDE6, which isfound in retinal photoreceptor cells, com-pared to its effect on PDE5. The recom-mended dosage is 25 to 100 mg.

Two reports on ERG findings in patientstreated with sildenafil found different re-sults. In their study of 5 subjects who took100 mg of sildenafil, Vobig et al.115 reporteda 37% and 23% reduction in both a- and b-wave amplitudes, respectively, in the maxi-mal dark-adapted ERG response; 6 hoursafter ingestion, the response returned tobaseline. No reduction in amplitude was

VITAMIN A DEFICIENCY AND RETINOIDS

Vitamin A deficiency and its associationwith night blindness has been recognizedsince the time of ancient Egypt. Ohanda517studied scotopic responses in patients withvitamin A deficiency. Subnormal-to-nonde-tectable responses were noted in patientswith xerosis and night blindness. After vita-min A administration, symptoms resolvedand ERG values returned to normal. Smalldot-Iike white deposits, which may developin the mid peripheral or peripheral retinaand not affect the macula, also usually re-solve after proper vitamin A therapy.518 Ingeneral, children below the age of 15 yearsare particularly prone to develop night blind-ness as a result of vitamin A deficiency.

100 The Electroretinogram

the complaint of poor night vision and a re-duction in ERG amplitudes most apparentunder scotopic conditions. A similar reduc-tion, primarily in rod-mediated ERG ampli-tudes, was reported with the use of fenreti-nide, a synthetic retinoid.525 Abnormal rodphotoreceptor-cell function, observed onERG testing in 2 patients treated for basal-cell carcinoma, returned to normal rapidlyafter cessation of therapy with fenretinide.525

OPTIC NERVE AND GANGLION CELL DISEASE

In albino rats, Dowling519 found an ini-tially more marked reduction of the a-waveto threshold stimuli that was associatedwith a decline in rhodopsin concentrationin the rod outer segments and was eventu-ally followed by a b-wave reduction. Dowl-ing noted a linear relationship between therhodopsin concentration and the logarithmof the stimulus luminance necessary to pro-duce a threshold ERG response. In theearly stages of vitamin A deficiency in hu-mans, a reversal of initially reduced ERGamplitudes occurs when parenteral vitaminA is administered (Figure 1-65). In 1966,Genest520 reported that both a- and b-wavesof the ERG were equally affected by de-creased vitamin A blood serum levels in hu-mans. Genest did not note the initial selec-tive a-wave reduction reported by Dowlingin his experiments with albino rats.

Patients with vitamin A deficiency andnight blindness associated with chronic al-coholism and liver cirrhosis or secondary tomalabsorption do not show the appreciabledelays in b-wave implicit times seen insome patients with retinitis pigmentosa.521,522Further, after vitamin A administration,cone function recovers more quickly thanrod function in the central retina, comparedto the retinal periphery, where rod functionrecovers more quickly, possibly due to rod-cone rivalry for vitamin A.521,523

Retinoids, which are derivatives of vita-min A, are used for the treatment of derma-tologic disorders, such as acne and psoriasis.One such derivative, isotretinoin (13-cis-retinoic acid), was reported by Weleber etal524 to cause abnormal retinal function, with

Some authors claim that the presence ofcentrifugal efferent fibers in the optic nervehas an inhibitory effect on the ERG. Thus,apparently as a consequence of the inter-ruption of this inhibitory effect, the ERGin some optic nerve diseases has been re-ported as supernormal. Other investigatorshave noted that surgical sectioning of theoptic nerve, without disturbance of the reti-nal circulation, has no apparent effect onthe ERG amplitude.

Because the ganglion cells do not con-tribute to the flash-elicited full-field ERGresponse, an essentially normal ERG is ob-tained in most eyes blinded by glaucoma.Further, patients with infantile amauroticfamilial idiocy (Tay-Sachs disease), a diso~-der known to involve the retinal ganglioncells, have a normal ERG.

Some investigators have noted a loweredcritical flicker-fusion frequency associatedwith the retrobulbar neuritis of multiplesclerosis. Furthermore, Fazio et al526 em-phasized that if proper gender and age-similar controls are used, reduced flash-evoked ERG amplitudes and associatedprolonged implicit times can be encoun-tered in eyes with advanced glaucomatousoptic nerve damage. Secondary ocular is-

1-19 Optic Nerve and Ganglion Cell Disease 101

chemia or vascular disease, in addition toretrograde degeneration of bipolar and pho-toreceptor cells (secondary to increased in-traocular pressure), was considered a possi-ble explanation.

Of interest, Weleber and Miyake527 re-ported a negative ERG waveform, mostapparent under scotopic conditions, in twofamilies affected with a familial (likely au-tosomal dominant) form of optic atrophy.

A Figure 1-65 ERG re-sponses from patientwith systemic vitaminA deficiency before and1 week after high-doseoral vitamin A supple-mentation. Note im-provement in ERGamplitudes under both( A) photopic and(B) scotopic conditions.

0:O

";;;>

:0---

>::!.

oo~

0)-0~

-0.E""'

40 80 1200 40 80 120 0 40 80Time (ms)

120 0

B

co";;;>

'0---

>::!.

oo~

0>-0~

-0.E""'

0 40 80 120 1600 40 80 120 160 0 40 80 120 160

Time (ms)

102 The Electroretinogram

ERG recordings have been reported inpatients with optic nerve hypoplasia.s28.s29The majority of these patients showed nor-mal photopic and scotopic ERG amplitudes.Occasionally, some patients show a smallreduction in amplitude, possibly as a con-sequence of accompanying degenerativechorioretinal changes.S28

preoperative ERG responses have been re-ported in patients with dense vitreous opac-ities who showed appreciable improvementin visual function after vitreous surgery.S31,S32Further, results from a bright-flash-elicitedERG need to be interpreted with particularcaution in patients who have undergone ex-tensive panretinal photocoagulation. Expo-sure of eyes with clear media to a very high-intensity flash stimulus should be avoidedbecause it is likely to evoke unwarranteddiscomfort.OPAQUE LENS OR VITREOUS

Particularly dense lens opacities can reducethe ERG a- and b-wave amplitudes. Whenpresent, the amplitude decrease is associ-ated with a comparable increase in implicittime, both relating to the resultant decreasein effective stimulus intensity. It takes amore intense stimulus to reach the sameamplitude and implicit time compared toan eye with clear media. Lens opacities donot appear to modify the OPs. Patients withmild or possibly even moderate degrees ofvitreous hemorrhage generally have normalor only moderately subnormal ERG a- andb-wave amplitudes. As with lens opacities,the implicit times are prolonged because ofthe apparent decrease in stimulus intensityreaching the photoreceptors. With long-standing vitreous hemorrhages, a markedreduction of ERG amplitudes is always pos-sible because of the ever-present threat ofsiderotic changes occurring within the retinasecondary to blood-breakdown products.

Fuller et al530 reported that ERG re-cordings elicited by a very high-intensity(bright-tlash) stimulus could be of value forpredicting postvitrectomy improvement invisual function of diabetic patients with vit-reous opacities. However, nondetectable

DIABETIC RETINOPATHY

The most frequently reported ERG abnor-malities in patients with diabetic retinopa-thy include a reduction in b-wave ampli-tude and a reduction or absence of OPs. Forthe most part, reductions in b-wave ampli-tude are noted in eyes with proliferativeretinopathy. However, a reduction in sco-topic b-wave amplitude may be observedin insulin-dependent diabetic children,even in the absence of clinically or angio-graphically apparent diabetic retinopathy.533These amplitude reductions likely repre-sent a quantitative measure of overall innerretinal ischemia and/or hypoxia. Chung etal534 studied ERG function in 65 patientswith diabetes. They measured ERG ampli-tudes as a function of stimulus intensity,and a Naka-Rushton-type function was fitto b-wave amplitudes in order to measureretinal sensitivity. The ERG-measured reti-nal sensitivity tended to decrease (ie, logK t ) as the severity of diabetic retinopathyincreased. However, the eyes of diabeticpatients without visible retinopathy did notshowa significant ERG sensitivity loss. Sig-nificantly prolonged b-wave implicit timeswere found in all stages of retinopathy and

1-21 Diabetic Retinopathy 103

in eyes of diabetic patients without clini-cally apparent retinopathy.

Reduced OP amplitudes have beennoted at early stages of retinopathy, whenERG a- and b-wave amplitudes are nor-mal,535 while delayed OP implicit timeshave been reported as an early functionalabnormality in eyes with mild or even noretinopathy.536 The most consistent changeis a significant increase in the OP1 implicittime.533 Yonemura and Kawasaki537 haveemphasized a selective delay in the implicittime of OPs, in the absence of amplitudereductions, as an early functional change ofthe retina in diabetic patients. Bresnick andPalta538 noted a delay in the 30-Hz ERGflicker implicit time, which became pro-gressively more prolonged with increasingseverity of retinopathy. These authors alsoshowed that the summed amplitudes ofOPs decrease, and implicit times in some ofthese potentials increase, as the severity ofdiabetic retinopathy increases.538-540 Bres-nick et al541 reported a lO-fold higher rate ofprogression to high-risk characteristics ofdiabetic retinopathy, as defined by the Oia-betic Retinopathy Study, in eyes with re-duced OP amplitudes compared to eyeswith normal amplitudes at study entry.

Later changes in both a- and b-wave am-plitudes, with more severe cases of diabeticretinopathy, are probably related to associ-ated arteriosclerotic changes, vitreous hem-orrhages, and, frequently, to retinal detach-ments present in the advanced stages ofdiabetic retinopathy. A useful review of theERG findings in diabetic patients with vari-ous stages of diabetic retinopathy is avail-able in an article by Tzekov and Arden.533

Oeneault et al542 used ERG amplitudesand implicit times to document the bene-fits of multiple-dose daily insulin therapy toconventional once-a-day injection of insulinin streptozotocin-induced diabetic rats.

The effect of panretinal photocoagula-tion on ERG amplitudes in diabetic pa-tients has been addressed in several publi-cations. Lawwill and O'Connor543 reportedan average 10% reduction in a- and b-waveamplitudes when approximately 20% of theretinal area was photocoagulated. In gen-eral, the percentage decrease in ERG am-plitude was less than the total area photoco-agulated. Wepman et a1544 reported findingsin 10 diabetic patients who had argon laserphotocoagulation burns to 15% to 18% ofthe retina. A mean reduction of 61% in therod b-wave amplitude, a 35% decrease in.the photopic (cone) b-wave response, and a39% decrease in the combined scotopic rodand cone response to a bright-Iight stimuluswere observed. Reduction of the ERG fol-lowing laser treatment was positively corre-lated with the total area of retina destroyed,provided that at least 7% to 10% of theretina had been treated. Patients with largerpretreatment b-wave amplitudes (>300 \lV)showed the greatest decrease in ERG am-plitude, while patients with smaller pre-treatment amplitudes (

104 The Electroretinogram

served by Wepman et a1544 contrasts withthe findings of Lawwill and O'Connor,,"43but is consistent with the findings of Ogdenet al545 and Frank.546

Franchi et al547 reported an improvementin OP amplitude after panretinal argon laserphotocoagulation in patients with prolifera-tive diabetic retinopathy. They advocateddetermining an OP/ERG amplitude quo-tient in patients before and after photoco-agulation treatment to evaluate the func-tional status of the retina following lasertreatment.

MISCELLANEOUS CONDITIONS

1-22-1 Retinal Detachment

In general, the amplitudes of both a- andb-waves are related to the degree of retinaldetachment. Karpe and Rendahl548 have re-ported subnormal ERG values in the nor-mal eye when the contralateral eye has aretinal detachment. In an eye with a de-tached retina, the ERG deteriorates in rela-tion to both the size of the detached areaand the decrease in retinal function. A non-detectable or minimal response implies atotal or large detachment, respectively, witha poorly functioning retina.

There is, however, some disagreementas to whether these responses imply a poorprognosis for successful reattachment andultimate return of visual function. BothRendahl549 and Jacobson et al550 indicatedthat the amplitude of the ERG response isof some prognostic value, while Schmoger55land Fran~ois and de Rouck552 did not placesignificant prognostic value on the ERGresponse.

1-22-2 Silicone Oil and Sulfurhexafluoride Gas

Silicone-oil injection into the vitreous cav-ity for the management of complicated reti-nal detachment was introduced in the UnitedStates by Cibis et al553 in the early 1960s.Frumar et al554 reported their findings onthe effects of intravitrealliquid silicone or amixture of 20% sulfurhexafluoride (SF6 )gas and air on retinal function in a series ofpatients who underwent vitrectomy for re-cent rhegmatogenous retinal detachment.In the early postoperative period, a reduc-tion of both a- and b-wave amplitudes wasobserved. This reduction was likely partlyrelated to the vitrectomy itself, but mainlyto the silicone oil or SF6 gas. A recovery ofERG amplitudes was observed in all pa-tients and was accelerated by absorption ofthe gas or removal of the liquid silicone.The authors opined that the relative ERGamplitude reductions were due to the insu-lating effect of the tamponading agents andobserved that these reductions were similarfor both gas and liquid silicone.

Similar findings on the effects of siliconeoil on ERG amplitudes were observed byother authors.555"';';6 Of interest, Meredith etal557 did not observe any sustained reduc-tion of ERG amplitudes or significant histo-pathologic changes of the retina in pig-mented rabbits following vitrectomy andintraocular silicone-oil injection. Eyes un-dergoing vitrectomy alone showed a tran-sient decrease in b-wave amplitude of 31 %,a finding similar to that observed by De-clercq et al558 after intraocular surgical in-jury to the rabbit eye.

Doslak5,';9 proposed a theoretical modelfrom which he concluded that it was notuntil at least 50% of the vitreous was re-placed with silicone oil that a small reduc-tion of the ERG amplitude would be ob-served. It was further theorized that thisreduction would increase nonlinearly as the

1-22 Miscellaneous Conditions 105

percentage of silicone oil was increased inthe vitreal cavity. According to the model,if the replacement of vitreous was particular-ly extensive, little, if any, ERG amplitudewould be measurable even if the retinawere otherwise normal.

tients with Gushing disease and patients re-ceiving exogenous corticosteroids can mani-fest a supernormal ERG.562 Zimmerman etal563 recorded an increase of approximately200% to 300% for both photopic and sco-topic a- and b-wave amplitudes in subjectswith normal eyes who received 40 mg ofprednisone daily for 3 weeks. Negi et al564could not demonstrate a similar effect ofbetamethasolie or dexamethasone in theirstudy on albino rabbits. However, these au-thors demonstrated an increase in both theb- and c-wave following the intravenous ad..ministration of AGTH (adrenocorticotropichormone) in the rabbit.565

Aldosterone was found by other investi-gators to increase the ERG b-wave, but de-crease the c-wave amplitude in the rab-bit.566 An increase in ERG amplitude inprimary aldosteronism was also noted byWirth and Tota.562 These investigators alsonoted that an intravenous injection of al-dosterone in rabbits could increase theERG b-wave amplitude by 45%. A sum-mary of ERG abnormalities in various en-docrine and other metabolic disorders isavailable in a review by Wirth.567

Patients afflicted with hepatic failureand accompanying hepatic encephalopathycan manifest functional abnormalities of theretina best demonstrated by ERG record-ings. These functional changes, which cor-relate well with the degree of encephalopa-thy,568 were thought to result from alteredneurotransmitter levels and impaired retinalglial-neuronal interaction as a consequenceof Muller cell damage caused by elevatedammonia levels. The ERG changes includedreduced a- and b-wave amplitudes, withprolonged implicit times for both cone androd responses. Reduced amplitude and de-layed implicit time for OPs were observed

1.22.3 Thyroid and Other Metabolic Dysfunctions

Several investigators have noted the bio-electrical activity of the retina to be in-creased in patients with hyperthyroidismand reduced in patients with myxedema.Particularly in thyrotoxic exophthalmos,both a- and b-waves of the ERG were re-ported as significantly supernormal. Wirth)6substantiated this finding of supernormalresponses and emphasized the generallyhigh correlation with the plasma protein-bound iodine. The ERG is said to be a sen-sitive test in hyperthyroidism, being super-normal even when the basal metabolic rateand protein-bound iodine may still retainnormal values.

Peatlman and Burian561 noted that thesupernormal ERG amplitudes in patientswith hyperthyroidism tend to diminish sig-nificantly in the course of treatment. Con-versely, they reported that myxedematouspatients with initially subnormal ampli-tudes showed an increase in ERG responsewhile receiving thyroid treatment. Theseauthors suggested that the ERG may be areasonably reliable, objective means of esti-mating the success of medical and surgicaltherapy in thyroid disease.

Wirth.)6 stated that the adrenal medullahad no influence on the ERG, noting thatpatients with a pheochromocytoma, as wellas those injected with 1 mg of epinephrine(Adrenalin), still had normal b-wave ampli-tudes. ]T Pearlman, MD, and HM Burian,MD (unpublished data), however, recordeda selective b-wave enlargement in patientswith an adrenal pheochromocytoma. Pa-

106 The Electroretinogram

to be the most sensitive indicators of retinaldysfunction, with abnormal findings occur-ring even in those with less advanced stagesof hepatic encephalopathy. It is noteworthythat patients did not complain of impairedvisual function and that normal serum vi-tamin A levels were maintained. Onlynonspecific RPE mottling and irregularmacular reflexes were observed on fundusexamination.

1-22-4 Parkinson Disease

Ellis et al569 reported their f\ash-evokedERG findings in 7 patients with idiopathicParkinson disease. Subnormal scotopic andphotopic amplitudes were observed in theParkinson patients compared to control pa-tients. An increase in ERG amplitude and areduction in implicit time were observed insome of the Parkinson patients shortly afterthe administration of levodopa.

changes. A reduction in scotopic b-waveamplitude was also reported by Cavallacciet al,572 who additionally found a prolongedscotopic b-wave implicit time. Stanescu andMichiels573 noted a reduction of the pho-topic as well as scotopic b-wave amplitude,in addition to a delayed scotopic b-waveimplicit time. Creel et al574 did not observea reduction in single-flash photopic ampli-tudes or prolonged scotopic implicit timesin their patients with myotonic dystrophy.These authors did note reductions in ERGscotopic b-wave amplitudes to dim blueand red flashes prior to the development ofophthalmoscopically detectable changesand in some neurologically asymptomaticpatients with minimally expressed myo-tonic dystrophy. The reduction in b-waveamplitude in patients with myotonic dys-trophy differs from the normal ERG ampli-tudes found in patients with myotonia con-genita, another autosomal dominant disorderof skeletal muscle.575

1-22-5 Myotonic Dystrophy

Myotonic dystrophy is a dominantly inher-ited systemic disease recognized by weak-ness of facial and distal muscles, sternoclei-domastoid muscle wasting, frontal balding,dysarthria, and percussion or grip myotonia.Ocular findings in Steinert myotonic dys-trophy include cataract, low intraocularpressure, and, less frequently, macularand/or peripheral pigmentary changes. Boththe systemic and the ocular findings resultfrom an expanded and variable number oftrinucleotide (CTG) repeat sequences inthe myotonin protein kinase gene.S70

Burian and BurnsS71 obtained ERG dataon patients with myotonic dystrophy andnoted an average b-wave reduction of40% to 45%, even in those patients withno or minimal ophthalmoscopically visible

1-22-6 Duchenne and BeckerMuscular Dystrophies

Duchenne muscular dystrophy (DMD) isa fatal X-Iinked recessive neuromusculardisease resulting from mutations in a genethat alters the structure and function of a427-kD (kD = kilodalton) cytoskeletal pro-tein termed dystrophin, which is expressedin a number of tissues, including muscle,brain, and retina. Alterations in dystrophincan cause either a severe expression of neu-romuscular disease, such as DMD, or amilder allelic form, Becker muscular dys-trophy (BMD).

Abnormal negative ERG responses havebeen recorded from patients with DMD orBMD. These result from a defect in signaltransmission between photoreceptor cellsand "on" (or depolarizing) bipolar cells,

1-22 Miscellaneous Conditions 107

as also noted in patients with CSNB andmelanoma-associated retinopathy.576,577 Pa-tients with DMD or BMD, however, do nothave any corresponding functional visualcomplaints, such as night blindness. Abnor-mal findings are not apparent on ophthal-mologic examination. Although the pres-ence of a normal a-wave and an impairedb-wave is similar to the morphologic ap-pearance of the ERG waveform seen in pa-tients with some forms of CSNB, patientswith DMD have a more selective impair-ment of OPz, particularly under photopicconditions, while patients with CSNB canshow reduction or absence of all OPS.577Pillers et al578 identified dystrophin in theouter plexiform layer of the human retina.This finding, coincident with the observa-tion of an abnormal b-wave amplitude inpatients with DMD or BMD, suggests thatdystrophin is required for normal retinalelectrophysiology.

The term Oregon eye disease has beenused to refer to a small group of patientswho, in addition to DMD or BMD, alsomanifest a glycerol kinase deficiency andcongenital adrenal hypoplasia. This disor-der seemingly represents a contiguous-genedeletion syndrome resulting from a deletionat Xp21.579 Patients were found to show anegative-configuration ERG with a reducedb-wave amplitude in the dark-adaptedstate. Of 5 reported patients, 1 showed nys-tagmus, an albinotic-appearing fundus, iristransillumination, and macular hypoplasia.

1. Heart block (associated with a cardio-myopathy)2. Cerebellar dysfunction with ataxia3. Cerebrospinal fluid protein above100 mg/dL

Other features, such as weakness of limb,facial, and neck skeletal muscles primarily,exercise intolerance, sensorineural deaf-ness, small stature, less than average intelli-gence, dental anomalies, respiratory distress(Adams-Stokes attacks), abnormal electro-encephalogr