The Pattern ERG in Man Following Surgica Resection of the Optic Nerve

Transcription

1 The Pattern ERG in Man Following Surgica Resection of the Optic Nerve Joseph M. Harrison, Patrick 5. O'Connor, Rockefeller S. L. Young,* Marilyn Kincaid, and Robert Bentley Pattern electroretinogram (PERG) results recorded in different laboratories from patients with unilateral traumatic transections of the optic nerve have led investigators to opposite conclusions about the sources of this response. There was no absolute demonstration of complete transection in any of these studies. In the present study, PERGs and flash ERGs were recorded from a patient who, 30 months earlier, had undergone surgical resection of the right optic nerve to remove a glioma. The histological section of the biopsied nerve confirmed complete optic nerve transection. Ophthalmoscopically and angiographically, the right eye was normal except for marked optic atrophy. PERGs were produced by 10 Hz reversal of high contrast checks with check widths from 13 deg 30 min to 12 min arc. Field size was 27 deg X 21 deg and space-averaged screen luminance was 110 cd/m 2. Smaller checks (3 deg 23 min to 12 min) produced responses in both eyes, but the responses in the right eye were much smaller than those in the left eye. Large checks and diffuse flashes produced approximately equal responses in the two eyes. The implicit times of the PERGs produced by stimulation of the right eye with smaller checks were shorter than those of the left eye. The authors conclude that, in humans, there is a contribution to the high contrast pattern reversal ERG from cells which are not dependent upon the integrity of the ganglion cell layer. These cells and cells dependent upon ganglion cells may both contribute to the high contrast PERG in the normal human eye. Invest Ophthalmol Vis Sci 28: , 1987 Riggs, et al 1 first demonstrated that electroretinograms (ERGs) could be produced by moving pattern stimuli. The fact that such patterns stimulate only localized regions of the retina and that the responses they produce are dominated by cone contributions 2 are both important clinical advantages. Subsequently, the pattern electroretinogram (PERG) was the subject of intense laboratory investigation. 3 " 8 The demonstration in humans 9 and cats 10 that the pattern-evoked, but not the flash-evoked, response of the retina was lost after optic nerve transection generated renewed research efforts and heightened clinical interest because of the potential usefulness of noninvasive measurement of ganglion cell activity. These results were taken to indicate that the retinal response produced by pattern was a separate entity from the ERG produced by flashes and that the pattern responses were dependent upon From the Department of Ophthalmology, University of Texas Health Science Center at San Antonio, San Antonio, Texas, and the *Department of Ophthalmology and Visual Science, Texas Tech Health Science Center, Lubbock, Texas. Supported in part by unrestricted development grants from the Research to Prevent Blindness, Inc. Submitted for publication: January 14, Reprint requests: Joseph M. Harrison, PhD, Department of Ophthalmology, UTHSCSA, 7703 Floyd Curl Dr., San Antonio, TX the integrity of the ganglion cell layer. These issues are in dispute. One study to investigate this issue has been carried out in monkeys.'' However data from human patients can often provide significant and immediately relevant evidence, if the pathology is sufficiently selective. There have been only four papers dealing with the effects of unilateral optic nerve transection on the PERG in humans All have been cases of traumatic transection. The conclusions have been contradictory. Two studies 912 have demonstrated total abolition of the PERG with transection of the optic nerve and two 1314 by the same author, reported no change in the PERG following traumatic transection of the optic nerve. We recorded the PERG from a young patient with a surgically resected optic nerve. The fact that both eyes were angiographically, ophthalmoscopically, and otherwise normal and that the optic nerve was resected surgically with minimal trauma to the globe, and that the resection was documented histologically, provided an unprecedented opportunity to investigate the effect of complete optic nerve resection with consequent ganglion cell degeneration on the PERG. Materials and Methods Subject The subject was a 19 year-old man. Two years previous to surgery, visual acuity was 20/15 in both eyes. 492

2 No. 3 HUMAN PATTERN ERG AFTER RESECTION OF THE OPTIC NERVE / Horrison er ol. 490 Fig. 1. This is a cross section of the intraorbital optic nerve from a region 6-12 mm behind the globe. This cross section demonstrates complete transection of the optic nerve. The filamentous structure detached from the nerve on the far left but more closely apposed to the nerve on the right is the arachnoid. The intact membrane closely surrounding the nerve (seen slightly detached at the bottom center) is the pia. The tumor is roughly outlined by the central light gray area. Blood vessels are seen centrally as darkly stained spots just below the center of the light gray area, but none are large enough to be the central retinal artery or vein. At the bottom center, four lumina of one or more blood vessels are seen in an elongated structure partially detached from the nerve between the arachnoid and the pia (Verhoeffvan Gieson stain, original magnification XI2). One year before surgery, the patient noted blurring and decreased vision in the right eye. Visual acuity was 20/30 at this time. By the end of that year visual acuity had decreased to 20/100, but the patient was orthophoric. He was diagnosed to have an optic nerve glioma and underwent surgery to remove it. A right frontal craniotomy with unroofing of therightoptic canal and orbit was peformed. The optic nerve was sectioned in two places, just distal to the chiasm and flush with the globe, the nerve between these points (approximately 3.5 cm in length was removed intact within the meninges, and the remaining stump on the globe was treated with heavy bipolar cautery. A section of the nerve containing the tumor is seen in Figure 1. Despite the resection of the optic nerve at the globe and cauterization of the distal stump, fluorescein angiography revealed normal choroidal and retinal circulation during arterial, venous, and late phase (Fig. 2). Opthalmodynamometry readings were not significantly different in the two eyes (90/60 OD and 100/ 65 OS). There was an amaurotic pupil and vision was "no light perception" in the right eye. Fundus photos taken 20 months after surgery showed marked optic atrophy with normal appearing retinal arteries and veins in both eyes. Red-free photography and subsequent analysis of nervefiberlayer were not performed. Fig. 2. Fluorescein angiograms of the right eye taken 2 weeks after surgery documenting a normal choroidal and retinal circulation during A, arterial; B, venous; and C, late phases of the angiography.

3 494 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Morch 1987 Vol. 28 The normal retinal circulation in the right eye suggests that large anastomoses can develop between the central retinal artery and the laminar and prelaminar posterior ciliary artery circulation when the nerve and central retinal artery are compressed by tumor growth. 15 There were 15 prism diopters of exotropia in the right eye 8 months after surgery. This increased to 25 diopters at the time of the experiment, 30 months after surgery. The PERG experiment was explained to the subject and he gave full consent to participate in it. Stimulus and Recording A checkerboard pattern stimulus was generated on the screen of a television monitor. The checkerboard was alternated at a rate of 10 reversals per sec. The mean luminance of the TV screen was 110 cd/m 2 and the contrast (L * - L min /L max + 1 ) of the checks was The screen subtended a visual angle of 27 deg horizontally by 21 deg vertically at the viewing distance of 0.5 m. The checks were rectangular in the same proportion as the dimensions of the screen. For all check sizes, there was always an equal and integral number of black and white checks in the display. This reduced the effects of stray light. The test was conducted in a darkened room with no other ambient lighting. The PERG was recorded from both eyes with active (ERG-jet, Universo, S.A., La Chauxde Fonds, Switzerland) electrodes on the corneae and reference, gold cup, electrodes on the outer canthi. The ground was a gold cup electrode placed on the right mastoid. We chose to use ERG-jet lens electrodes because they allow recording of large PERGs with high signal to noise ratios in every subject. Also, these electrodes do not pick up volume-conducted potentials, 16 so the seeing eye could be used for fixation and positioning the other eye. DTL 17 and gold-foil electrodes 18 would not interfere with the optics of the eye, but we have found them unsatisfactory for recording responses with large amplitude and signal to noise ratio. These latter electrodes are susceptible to picking up volume conducted potentials from the other eye or from the visual cortex. 16 The signals from the eyes were amplified, digitized (12 bits) and averaged by a commercial system (UTAS- E 1000, LKC Systems, Inc., Gaithersburg, MD). The analysis time for each response was 225 msec. Fifty responses were averaged for each stimulus condition. The signals were filtered with 3 db bandpass limits at 1 and 30 Hz. An artifact rejection routine rejected traces containing peak to peak voltages larger than 100 uv, thus eliminating large transients associated with movement and muscle activity. Seven check-widths from 12 min to 13 deg 30 min were presented in random order. Controls consisted of presentation of 12 min and 6 deg 40 min checks with +20 D lenses placed in the trial frame in front of the eyes. Signals were also recorded with the right eye covered using 50 min checks as the stimulus. The flash ERG was produced by ganzfeld stimulation in the dark. No mydriatics were used. The amplitude of the PERG was measured between the peak positive wave and the trough of the following negative wave. The flash ERG was measured in the traditional manner; from baseline to the trough of the a-wave for a-wave amplitude, and from the trough of the a-wave to the peak of the b- wave for the b-wave amplitude. Implicit time for the PERG was the time from the last reversal to the peak of the major positive wave. Implicit time for the b- wave of the flash ERG was measured from the onset of the flash to the peak of the b-wave. Techniques The eyes were anesthetized with 2 drops of Alcaine (proparacaine HC1, Alcon, Fort Worth, TX) and the electrodes wetted on the concave surface with Goniosol (methylcellulose, Looper Vision, San German, Puerto Rico) were placed on the eyes. Base-in prisms (8 prism diopters) were placed in the trial frames in front of both eyes resulting in a total of 16 prism diopters correction to reduce the disparity of the image position on the two retinae due to the 25 prism diopter exotropia. The resulting disparity would be about 5 deg visual angle. Both eyes were uncovered except for one control condition in which the right eye was covered. No attempt was made to measure or correct the spherical error induced by the electrode lens. 16 However, the position of the electrode lens on the left eye was adjusted on the cornea until the subject noted maximally sharp and clear vision. The electrode on the right eye was positioned symmetrically. The subject sat facing the screen with his back against the chair to stabilize head position. The left eye was always uncovered and the subject used this eye to fixate a small black dot in the center of the screen. Results PERGS produced by stimulation of the right and left eyes with 50 min check widths are seen in the left panel of Figure 3. The amplitude of the PERG produced by stimulation of the left eye was about three times larger than that produced by stimulation of the right eye (6.2 vs 2.1 /xv) The implicit time of the major positive wave was 6 msec greater for the response of the left eye than that for the right eye. When this eye was covered (right panel), no PERG waveform was detectable while the PERG of the left eye was not affected. This rules out the possibility that the PERG

4 No. 3 HUMAN PATTERN ERG AFTER RESECTION OF THE OPTIC NERVE / Horrison er ol Min - Eyes Uncovered OS Fig. 3. PERGs produced by 50 min checks in the left and right eyes (top and bottom traces). Left panel: both eyes uncovered; Right panel: right eye covered. Vertical lines mark reversal. Positive at the active electrode on the cornea is indicated by an upward deflection in this and the following figures. Fig. 4. PERGs produced by 12 min checks with normal viewing (left panel) and with +20 diopter lenses in front of the eyes (right panel). from this eye was a volume-conducted potential from the left eye or from the occipital area. This is consistent with the conclusions of a previous study. 16 With 12 min checks, the PERG waveform produced by stimulation of the right eye was less striking; however, it is clearly detectable, especially for the response produced by the "second" reversal (Fig. 4, left). The implicit time of the major positive wave of the PERG in the left eye was about 12 msec longer than that in the right eye for this check size. When the retinal image was defocussed by placing +20 diopter lenses in front of the eyes, no PERG waveform was detectable in either eye. There were baseline perturbations in all traces, and 60 Hz contamination, which was seen intermittently, but these could not be mistaken for a true response. If there had been changes of overall luminance with reversal of the checkerboard, it would not be affected by the 20 diopter lenses. The lack of responses with the +20 diopter lenses eliminates the possibility of significant luminance contamination of the stimulus producing the response. Figure 5 shows that the PERGs produced by large checks (6 deg 40 min wide by 5 deg 11 min high; 2 white and 2 black checks,both horizontally and vertically) without defocus had approximately the same amplitudes and implicit times in the two eyes (4.3 nv, 50 msec OS; 3.8 /xv 48 msec OD). Defocussing the image of large checks with the +20 diopter lens had no effect on the amplitudes of the PERGs. For both the focus and the defocus condition, the implicit times of the PERGs produced by stimulation of the left eye with this check size were shorter than for stimulation with smaller check sizes and not greatly different from those produced by stimulation of the right eye with large or small checks. Curiously, however, the implicit time of the PERG from the right eye became prolonged with image defocus so that the implicit time became comparable to that of PERGs produced by stimulation of the left eye with small checks without the +20 diopter lenses. The ERGs produced by ganzfeld flash stimulation are seen in Figure 6. These traces represent the averages of two responses. The a- and b-wave amplitudes were as follows: 145 and 154 nv, respectively, OS; and 120 and 139 /*V OD. Although the amplitudes of the a- and b-waves recorded from the right eye were slightly smaller than those from the left eye, neither the a- nor the b-wave amplitudes from either eye differed significantly from one another. 19 The normal flash ERG indicates that the receptors, the cells of the inner nuclear layer, and the Muller cells were normal, which also implies, as demonstrated by fluorescein angiography, that the retina was well perfused through the choroidal and retinal circulation. Figure 7 shows the amplitude of the PERG from OS and OD as a function of log )0 dominant spatial frequency (i e, (60/checkwidth in min) X >/5) of the checkerboard patterns. The spatial frequency for each stimulus is plotted at a value 10% higher than this to compensate for the higher spatial frequency content of the rectangular checks. Above 0.12 cy/deg, the amplitude of the PERG from the left eye was about three to four times that from the right eye. At 0.12 and 0.06 cy/deg, the amplitudes were less different or the same in the two eyes. The data from the two eyes show some similarity, such as the decline of responsiveness on either side of 0.95 cy/deg, but more striking is the decreased responsiveness of the right eye to medium and high spatial frequencies relative to the left eye. 6 Deg 40 Min + 20 D Lens Fig. 5. PERGs produced by 6 deg 40 min checks with normal viewing (left panel) and with +20 D lenses in front of the eyes (right panel).

5 496 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / March 1987 Vol. 28 Flash ERG (Ganzfeld) 100/xV 25 msec Fig. 6. Ganzfeld flash ERGs recorded from the 2 undilated eyes in a darkened room without prolonged dark adaptation following pattern stimulation. The spatial frequency function of the PERG from the left eye is very similar to that demonstrated recently, 20 with a medium spatial frequency peak for about 0.85 cy/deg, a low spatial frequency trough at 0.18 to 0.35 cy/deg and an increase of amplitude for 0.09 cy/deg. However, direct comparisons are impossible, especially for the low spatial frequencies, since stray light would have appreciably affected our amplitudes produced by large checks because of the lack of a surround in our study. Discussion These data demonstrate that PERGs can be produced by high contrast pattern reversal in an eye long after transection of the optic nerve when retrograde degeneration of its ganglion cells should be complete It might be beneficial to examine how these observations affect the controversy surrounding the source of the PERG. In early investigations, there was evidence that no PERG could be produced with either low 12 or 3 ude (/ e ae < «OS o 0.D iiiiiiti< > 0* / \ o- i Spatial Frequency (Log Cy/Deg) Fig. 7. PERG amplitudes as a function of the log dominant spatial frequency of the checkerboard pattern. Dashed horizontal line is the average noise level. 1.0 high contrast 9 stimuli in an eye with degenerated ganglion cells. Later studies 1314 found PERGs produced by high contrast stimuli in cases of optic nerve trauma, but complete transection of the optic nerve was not demonstrated. In the present study, the PERGs to all except the two largest check sizes in the right eye were very reduced compared to those in the left eye. Since the flash ERG was normal in the affected eye, one might propose that the PERGs produced by the two larger check sizes had a predominant contribution from the mechanism givingriseto theflasherg and that the corresponding PERGs produced in the normal eye, under our conditions, also had a major contribution from the flash ERG mechanism. One might propose that the reduced responses to the smaller checks were due to an associated loss of an active contribution from the ganglion cells in that eye. This has been the major hypothesis considered to account for the loss or diminution of the PERG following optic nerve damage. The fact that the PERG implicit time varies little with contrast when the responses of ganglion cells do vary in response time with contrast is an observation that may be inconsistent with the hypothesis of an active contribution of the ganglion cells to the PERG. 23 One might also propose that there is no active contribution of the ganglion cells to the PERG. Under this hypothesis, the ganglion cells would be playing a passive role, allowing the PERG to be recorded with greater amplitude in the normal eye. This could be true, if the ganglion cells had a sustentacular role for some cells such as the amacrine cells or if they were part of a resistance network that allowed the potential drop of the PERG generator to be recorded at the cornea. This would imply an interaction of the ganglion cells with a select group of cells exclusive of those giving rise to the flash ERG. The remaining PERG produced by smaller checks following optic nerve transection, then, would be either a diminished version of the normal response with normal relative contributions from various cells or could differ in its sensitivity from the PERG in the normal eye because of a selective loss of some subpopulation of cells dependent upon ganglion cells. If the PERG in the affected eye displayed different properties from those of the PERG in the normal eye, then this would be evidence for the loss of contributions from a selective group of cells. Loss or diminution of the PERG following degeneration of the ganglion cells, by itself, cannot distinguish between the hypotheses concerned with active versus passive contributions of the ganglion cells to the PERG. Asymmetry in focus of the two eyes cannot be excluded since refraction over the ERG-Jet lens elec-

6 No. 3 HUMAN PATTERN ERG AFTER RESECTION OF THE OPTIC NERVE / Horrison er ol. 497 trode and prisms was not peformed. A greater defocus in the right eye could have caused a reduction in the PERG amplitude in the right eye relative to that in the left eye. The defocus would have had a greater effect on the PERGS produced by the smaller check sizes. However, if defocus was a significant factor in the reduced amplitude of the PERGs in the right eye, the PERGs would have been even larger with better refraction, which would have been stronger evidence for our primaryfinding:that the PERGs produced by high contrast medium check sizes can survive following total degeneration of the ganglion cells. In our study, contrasting with previous studies which dealt with traumatic transection of the optic nerve, the hypothesis that the remaining PERG is due to the contribution from a few remaining ganglion cells can be excluded. Our data do not give unequivocal support for either of the hypotheses dealing with the possible contribution of the ganglion cells to the PERG produced by high contrast medium check sizes in the normal eye. Since we also cannot exclude some effect of defocus on the PERGs produced by smaller checks in the right eye, we cannot estimate, if the first hypothesis were true, what percentage of the normal response, with our stimulus conditions, would be contributed by the ganglion cells. However, we believe we can give a lowerlevel estimate of the contribution of cells other than ganglion cells, for 12,25, and 50 min checks, the PERG amplitudes in the right eye were 0.12, 0.29, and 0.27 of those in the left eye, with the average noise amplitude subtracted from both. It is clinically important to be aware that PERGs produced by high contrast checks can contain contributions from cells other than ganglion cells. There is considerable evidence that the PERG has no contributions from mechanisms which respond to contrast across borders. 6 ' 20 ' 24 ' 25 However, there is good evidence that the PERG has contributions both from the mechanism giving rise to the focal ERG and from another mechanism more sensitive to the spatial distribution of luminance of the retina Stimulus parameters such as spatial frequency, contrast, and luminance can vary the relative contributions of the two mechanisms. Arden et al 23 found that with check widths less than 2 deg, the PERG appeared as a distinct response from the focal ERG since, with smaller check widths, there was no net luminance change within the summation area of the focal ERG. Bobak, et al, 26 however, cautioned that 50 min checks may generate a flicker response. Arden and Vaegan 20 found, as we did, that in the smaller check size range, check widths about 50 min arc produced the largest PERGs. In our experiment, this was true in both eyes. Our stimulus conditions of high luminance and contrast probably emphasized the focal ERG contribution. The appropriate experiments were not carried out to characterize the relative contributions of the two mechanisms. However, the implicit times of the PERGs produced by the smaller checks in the right eye were shorter than those produced by the same stimuli in the left eye. Arden and Vaegan 20 showed that focal ERGs, for comparable luminance and contrast conditions, had 5 msec shorter implicit times than the PERGs. For the smaller check sizes in our experiment, the implicit time difference between the two eyes might be evidence for dominance by a focal ERG mechanism in the right eye and a large contribution by the other mechanism in the left eye, and perhaps loss or severe diminution of the latter mechanism in the right eye. If the remaining PERG in the right eye arose from the same generator as that of the focal ERG, then the summation area for this generator may be smaller than previously suspected. The spatial distribution of luminance in multispot flashes equated for diffuse and local luminance affects the amplitude of the human flash ERG. 27 The small pattern reversal ERG for smaller check sizes may be a reflection of that mechanism demonstrated in the multispot flash ERG. Another perspective on this subject is provided by considering the work of Baker and Hess 25 who showed that only a nonlinear, even harmonic response was seen in the contrast reversal PERG, but that the mechanism was sensitive to spatial frequency due to a sensitivity to luminance difference across the retina and that it was not purely a local luminance response. They suggested that it is the even harmonic response that is lost following optic nerve section and thatflashergs contain a contribution from this component, but that it is not easily detectable since theflasherg is dominated by a much larger fundamental response. This schema, however, does not account for the remaining PERG following transection of the optic nerve, since the contrast reversal PERG does not have the fundamental component of the flash ERG, and since the odd symmetric components of the local generators of the even symmetric component should cancel. The discrepancy between the present study and that of Dawson, et al 12 is not surprising because of the difference in contrast in the two studies. It could be predicted that reducing contrast by 57% from the 0.99 in the present study to 0.42 in the Dawson et al 12 study would result in PERGs from the affected eye which would be below the noise level in this study since the PERG amplitude varies directly and linearly with contrast up to 0.90 in the normal eye. 20 Slopes of the amplitude versus contrast function have been found not to differ significantly in normal and some abnormal eyes. 28

7 498 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / March 1987 Vol. 28 Although we demonstrated a PERG in an eye with a resected optic nerve, responses of this eye to middle and higher spatial frequencies were significantly reduced compared to those in the normal eye. The responses were, in fact, just above the level of detectability. For this reason, our data do not preclude the possibility of a significant dependence of the PERG on the integrity of the ganglion cell layer. Sherman 13 reported that the PERGs in the affected eyes of patients with unilateral traumatic transection were not significantly different from those in the normal eyes. These data would exclude the possibility of a significant dependence of the PERG on the integrity of the ganglion cell layer. The reason for the discrepancy in the two studies is unclear. A remeasurement of the PERGs from the positive peak to the after-negative peak revealed an approximate 50% reduction in the affected eye which brings his results closer to those of the present study. The fact that appropriate fixation controls were performed and that PERGs were recorded from the affected eye when the simultaneously recorded pattern visual evoked potential was not present seem to exclude the possibility of significant crosstalk from the unaffected eye or the visual cortex. The presence of a few remaining ganglion cells was not absolutely excluded in his study. As mentioned above, available evidence in humans 22 indicates complete retrograde degeneration of ganglion cells within 6 months of destruction of optic nerve fibers. More recent data in the cat indicate that 40% of the ganglion cells remained and appeared normal morphologically 15 months after section of the optic nerve. 29 ' 30 This is approximately 1 year after PERGs have been reduced to noise level. So apparently even if there were a few surviving ganglion in the affected eye of our subject, they could not account for the PERG produced by the smaller check sizes. However, given the difference in the PERG in humans and cats, 31 these comparisons may not be valid. Key words: pattern-erg, optic nerve, surgical resection, spatial frequency, local luminance Acknowledgments The authors would like to acknowledge and thank Michael A. Morris, OD for his comments and assistance, and Charles S. Ballentine for his technical assistance. References 1. Riggs LA, Johnson EP, and Schick AML: Electrical responses of the human eye to moving stimulus patterns. Science 144:567, Johnson EP, Riggs LA, and Schick AML: Photopic retinal potentials evoked by phase alternation of a barred pattern. In Clinical Electroretinography: Proceedings of the Third International Symposium, Burian AM and Jacobson JH, editors. New York, Pergamon Press, 1966, pp Armington JC, Marsetta R, and Schick AML: Stimulus alternation and low-level response. Vision Res 10:227, Armington JC, Corwin TR, and Marsetta W: Simultaneously recorded retinal and cortical responses to patterned stimuli. J Opt Soc Am 61:1515, Sokol S: An electrodiagnostic index of macular degeneration: Use of a checkerboard pattern. Arch Ophthalmol 808:619, Spekreijse H, Estevez O, and van der Tweel LH: Luminance responses to pattern reversal. Doc Ophthalmol Proc Ser 2:205, Lawwill T: The bar-pattern electroretinogram for clinical evaluation of the central retina. Am J Ophthalmol 78:121, Sokol S and Nadler D: Simultaneous electroretinograms and visually evoked potentials from adult amblyopes in response to a pattern stimulus. Invest Ophthalmol Vis Sci 18:848, Groneborg A and Teping C: Topodiagnostik von sehstorungen durch ableitung retinaler und kortikaler antworten auf umkehrkontrastmuster. Ber Dtsch Ophthalmol Ges 77:409, Maffei L and Fiorentini A: Electroretinographic responses to alternating gratings before and after section of the optic nerve. Science 211:953, Maffei L, Fiorentini A, Bisti S, and Hollander H: Pattern ERG in the monkey after section of the optic nerve. Exp Brain Res 59:423, Dawson WW, Maida TM, and Rubin ML: Human patternevoked retinal responses are altered by optic atrophy. Invest Ophthalmol Vis Sci 22:796, Sherman J: Simultaneous pattern-reversal electroretinograms and visual evoked potentials in diseases of the macula and optic nerve. In Evoked Potentials, Bodis-Wollner I, editor. Ann NY Acad Sci 388:214, Sherman J and Richardson V: What is the origin of the pattern reversal electroretinogram in humans? ARVO Abstracts. Invest Ophthalmol Vis Sci 22(Suppl): 138, Levin ML, O'Connor PS, Aguirre G, and Kincaid, MC: Angiographically normal central retinal artery following the total resection of an optic nerve glioma. J Clin Neuro Ophthalmol, 6: 1, Peachey NS, Sokol S, and Moskowitz A: Recording the contralateral PERG: Effect of different electrodes. Invest Ophthalmol Vis Sci 24:1514, Dawson WW, Trick G, and Litzkow CA: Improved electrode for electroretinography. Invest Ophthalmol Vis Sci 18:988, Arden GB, Carter RM, Hogg CR, Siegel IM, and Margolis SA: A gold foil electrode: Extending the horizons for clinical electroretinography. Invest Ophthalmol Vis Sci 18:421, Fishman GA: The electroretinogram and electro-oculogram in retinal and choroidal disease. American Academy of Ophthalmology, San Francisco, 1975, p Arden GB and Vaegan: Electroretinograms evoked in man by local uniform or patterned stimulation. J Physiol (Lond) 341: 85, Radius R and Anderson D: Retinal ganglion cell degeneration in experimental optic atrophy. Am J Ophthalmol 86:673, Kupfer C: Retinal ganglion cell degeneration following chiasmal lesions in man. Arch Ophthalmol 70:256, Arden GB, Vaegan, and Hogg CR: Clinical and experimental evidence that the pattern electroretinogram (PERG) is generated in more proximal retinal layers than the focal electroretinogram

8 No. 3 HUMAN PATTERN ERG AFTER RESECTION OF THE OPTIC NERVE / Harrison er al. 499 (FERG). In Evoked Potentials, Bodis-Wollner I, editor. Ann NY AcadSci 388:214, Riemslag FCC, Ringo JL, Spekreijse H, and Lunel HFV: The luminance origin of the pattern electroretinogram in man. J Physiol (Lond) 363:191, Baker CL and Hess RF: Linear and nonlinear components of human electroretinogram. J Neurophysiol 51:952, Bobak P, Bodis-Wollner I, Harnois C, Maffei L, Mylin L, Podos S, and Thornton J: Pattern electroretinograms and visual-evoked potentials in glaucoma and multiple sclerosis. Am J Ophthalmology 96:72, Diehl R and Zrenner E: Multispot stimuli reveal spatial organization in the human electroretinogram (ERG). Doc Ophthalmol Proc Ser 23:209, Seiple W, Price MJ, Kupersmith M, Siegel IM, and Carr RE: The pattern electroretinogram in optic nerve disease. Ophthalmology 90:1127, Hollander H, Bisti S, Maffei L, and Hebel R: Electroretinographic responses and retrograde changes of retinal morphology after intracranial optic nerve section. A quantitative analysis in the cat. Exp Brain Res 55:483, Hollander H, Bisti S, and Maffei L: Long-term survival of cat retinal ganglion cells after intracranial optic nerve transection. Exp Brain Res 59:663, Zrenner E, Hess RF, Baker CL, and Schwarzer J: Fundamental differences between pattern electroretinograms of cat and man. ARVO Abstracts. Invest Ophthalmol Vis Sci 26(Suppl): 112, 1985.