Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
The ERG and Sites and Mechanisms of Retinal Disease, Adaptation, and Development DONALD C. HOOD INTRODUCTION What do ambient lights, retinal disease, and the development of the neonatal retinal have in common? One answer is that each can alter the retina's sensitivity to light. A second is that the same noninvasive techniques have been used with human subjects to assess the sites and mechanisms of disease, adaptation, and developmental processes. One technique is the recording of the electroretinogram (ERG). This paper presents a general approach for using ERG data to assess the sites and mechanisms of a change in retinal processing. The approach is illustrated using data from an adaptation paradigm. Application to clinical and developmental questions is considered. Since the receptor is the focus of this symposium, the emphasis here is on testing the hypothesis that sensitivity changes that accompany a disease, an adaptation process, or a developing retina have their locus at the receptors. The ERG is a gross potential recorded from the eye. Figure 1A shows ERGs recorded from a normal adult. Each record is the response of the eye to a flash of a different intensity. For higher flash intensities, the characteristic a- and lo-waves of the ERG can be seen. The ERG is a complex potential. The potential measured is actually the algebraic sum of a number of individual components. Figure 2 shows an analysis of the ERG by Brown (1968~. The analysis is similar to Granit's classic analysis (Granit, 1947~. One of the two main potentials contributing to the ERG is a corneally negative potential generated by the receptors. This component was called p-III by Granit and is labeled the rod late receptor potential in Figure 2. The two corneally positive potentials, labeled D.C. component and lo-wave in Figure 2, were shown 41 -
42 DONALD C HOOD Log I -3.36 -3.03 -2.80 -2~ -2~34 -2~16- -1 .99~ -1.76 _ 1 e 51 __ ,1 .49 _ -1~15- -1 .08 -0~93 -0~66 _ -0~26- me\ -1/15\\'~ 50 msec lOOyV stimulus FIGURE 1A ERG responses from a normal adult. Each record shows the response to a single flash of light. The log of the flash intensity (log cd/m2)is shown next to each record. SOURCE: Modified from Massof et al. (1984~.
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT 600 500 > ~ 400 - . _ - ~ it, 200 (at 300 100 n - 2 R m f I. . R=~ R (') I +/ 7 o , ~ , , -40 -35 L91 by R m 1 1 1 -30 -25 -20 -1 5 -1 0 -05 00 05 log I ( cd -see /m2 ) 43 FIGURE 1B Intensity-response functions for the lo-wave. The filled symbols are the peak-to-trough lo-wave amplitudes plotted against the log intensity of the flash. SOURCE: Courtesy of M. Johnson. by Brown to be generated by cells in the inner nuclear layer. These potentials comprise Granit's p-II. The peak-to-trough amplitude of the lo-wave provides a reasonably good measure of the magnitude of the po- tentials generated in the inner nuclear layer. Some investigators measure the a-wave to obtain information about the receptors. Use of the a-wave as a measure of the magnitude of the receptor response is more risly or rather requires additional assumptions. The ERG was the focus of a great deal of research in the 194Qs, 1950s, and 1960s. Many of the basic functions of the retina were inferred from this wore As a laboratory technique, it has been supplanted for most purposes by intracellular recording. In the clinic, however, the ERG still provides a powerful tool for diagnosing the type and progression of retinal disease. The development of the focal ERG and the use of quantitative analyses have enhanced clinical interest in the ERG. Figure 1B illustrates one quantitative approach. The filled symbols depict intensity-response data from recordings such as those in Figure IN The smooth curve through the data is given by Equation (1) in Figure 1, where R is the peak-to-trough lo-wave amplitude, I is the flash intensity, K is the semisaturation constant, and Rm is the maximum response. The exponent n is usually close to 1. To simplifier the presentation, it is assumed that n is equal to 1.0. Retinal diseases produce both an increase in K and a decrease in Rm. Mary Johnson, Bob Massof, and I have been developing an approach to
44 DONALD C HOOD S t i mulus ( 1 sec ) Predominantly ~ I \ rod ERG I,/ ~ Off-response Rod l ~ t e receptor potent i at D. C. component- \ b-~ave ~ Recant negat ivi ty 1 _- FIGURE 2 An analysis of the components of the dark-adapted rod ERG. Repnnted with permission from Vision Research, 8, Kenneth T. Brown, "The Electroretinogram: Its Components and Their Origin," (if) 1968, Pergamon Press. inferring the retinal site and mechanism of disease action from changes in K and Rm (Johnson and Hood, 1988; Johnson and Massof, 1988~. To evaluate this approach, I have analyzed adaptation data from Fulton and Rushton (1978~. This analysis is summarized below. INFERRING THE SITE OF ADAPTATION FROM ERG DATA In the adaptation paradigm, intensity response data (as in Figure 1) are collected for flashes presented on steady ambient lights. Fulton and Rushton (1978) obtained intensity-response data for lo-waves recorded on steady adapting fields ranging in intensity from no field (dark adapted) to 3.2 log scotopic trolands. They fitted Equation (1), with n set equal to 1.0, to each set of intensity-response data. 1b illustrate our approach, assume that Fulton and Rushton's data are well fitted by Equation (1) and that the rod system has been successfully isolated. As the adapting intensity is increased, their intensity-response curves move down and to the right when plotted as in Figure 1B. In terms of Equation (1), the maximum response, Rm, decreases and the value of the semisaturation constant, K, increases with increases in the intensity of the steady adapting field. In Figure 3 the change in log Rm is plotted against the change in log K. Each data point is for a different adapting field intensity. The values of Rm and K are expressed relative to their dark- adapted values. The value 0 log relative K or Rm represents no change from the dark-adapted value (dashed lines). Data for the higher-adapting intensities were omitted to help assure that the rod system was isolated and to keep the range of sensitivity changes close to those seen below for retinal diseases.
ERG, ATONAL DISEASE, ORATION, AD DE~LOPME 0.5 ~ 0.0 ~ a: ~ -ns- ._ _ - ~v - lo -1 .0 ~ -1.5 ~ ~1 d.a. I ! 1 ll . ~ ~ ~ ~ .8 -.4 .2 -2.0- I , . . . . . -0.5 0.0 0.5 1.0 1.5 2.0 Log Relative K 45 FIGURE 3 Log relative Rm versus log relative K is shown for a range of adapting intensities from the Fulton and Rushton (1978) study. The values of Rm and K, estimated from their fit of Equation (1), are expressed relative to the dark-adapted values. A Model of the ERG Ambient lights produce a small decrease in Rm and a large increase in K. How many of these changes can be attributed to the rod receptor and how many to the cells of the inner nuclear layer? A simplified version (Johnson and Hood, 1988) of the model we are developing is summarized in Table 1. We assume that the pooled rod receptor response, A, is given by Equation (2), where Ka is the semisaturation constant of the receptors and Am is the maximum pooled response. Assume further that the response, B. of the lo-wave generators in the inner nuclear layer is also described by a function of the same form, where the pooled receptor response A is the input to the lo-wave generator [Equation (3~. The parameters Kb and Bm are the semisaturation constant and maximum amplitude of the lo-wave generators. Finally, we assume that the peak-to-trough lo-wave amplitude measured is proportional to the response of the lo-wave generators [Equation (4~. After appropriate substitution and algebraic manipulation, Equation (5) shows
i 46 TABLE 1 Static Model of ERG DONALD C HOOD Pooled response of rod receptors: Response of lo-wave generatom: ERG lo-wave response: by substitution where In in + ran m B = A + ~; Bm R= pB where ,0 is a constant ln l n ~ In m Bum Rm l+(I~i'b/Am) I' via (2) (3) (4) (5) (6) (7) that R. the lo-wave response one measures, has the form of Equation (1), where Rm and K are given by Equations (6) and (7), respectively. Notice that the semisaturation K and the maximum response Rm of the lo-wave depends on three of the four parameters of the model. In other words, changes in the receptor's semisaturation constant, Ka, only affect K and changes in Bm only affect Rm. Both K and Rm are affected by the ratio of Am to Kb. The dashed vertical and horizontal lines in Figure 4 show the predicted changes in K and Rm for increases in Ka (a decrease in receptor sensitivity) or a decrease in Bm (a decrease in the response amplitude of the cells generating the lo-wave). The semisaturation constant, Kb, of the lo-wave generator and the maximum amplitude of the receptor potential, Am, both contribute to the measured values of K and Rm [see Equations (6) and (7~. 1b obtain a measure of the influence of Am and Kb on K and Rm, we can set the ratio of Am/Kb in Equations (6) and (7~. A reasonable estimate of this
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT 0.5 - E °°~ ~ -0.5 - ._ co -1 .0 - o -1 .5 - -2.0 - -0.5 o.o 0.5 1.0 1.5 2.0 i Ka ~ `~,Bm 4' Log Relative K 47 Am W orKbt FIGURE 4 The data from Figure 3 are shown with theoretical curves based on changes in the parameters of the model in Table 1. ratio is about 102 5; it was obtained in the following way: The value of K typically measured for the lo-wave is on the order of 102 5 times smaller than the semisaturation of the receptors, Ka, measured for the a-wave (Fulton and Rushton, 1978; M.A. Johnson, personal communication) or inferred from human psychophysics (e.g., Sakitt, 1976; Hood and Greenstein, 1988a, 1988b). Substituting (lo-2 5) x Ka for K2 in Equation (7) provides a value of the ratio Kb/Am. It too will be approximately 10-2 5. By specifying the value of Kb/Am, Equations (6) and (7) provide estimates for changes in Rm and K, with increases in Ka and decreases in Am. The three solid curves in Figure 4 show the predicted changes in log K and log Rm for increases in Kb or decreases in Am for three values of the ratio Kb/Am. The middle curve was derived assuming a K/Ka ratio of 10-2 5. The two other solid curves are for K/K a ratios of 10-3 (upper solid curve) and 10-2 (lower solid curved. Under these assumptions, an increase in Kb or a decrease in Am largely affects K and has little effect on Rm. [For a quantitatively similar analysis of the ERG, see Arden et al. (1983~. For a quantitatively similar analysis of psychophysical data, see Hood and Greenstein (Lomb). For the general notion that the rods are linear over a wide range of conditions, see Rushton (1965~.]
48 DONALD C HOOD How do we answer the question: Are the changes in K and Rm produced by the steady adapting field occurring at the receptors? There are two approaches one can take to answer this question: (1) an explicit model of the receptors can be assumed or (2) the a-wave can be used to obtain estimates of Ka and Am. Let us consider the first approach. A Model of Adaptation of the Rod Receptor The prevailing model of adaptation of the human rod receptor is response compression.) Response compression as a model of receptor adaptation was first described by Boynton and Whitten (1970~. More recently, Baylor et al. (1984) have shown that this model fits the data from single primate rods. Figure S is a summary of a response compression model. The solid curve in the top panel is the dark-adapted intensity- response function. Adapting fields, shown by dashed lines, are assumed to produce responses as if they were flashes of light. The adapting-field- induced desensitization of the receptor occurs because of the restricted or compressed response range. When the total response is plotted versus the total intensity of the adapting field plus the flash (see Figure SA), there is no change in K or Rm. In the Fulton and Rushton experiment, an incremental response, the ERG, was measured as a function of an incremental light, the flash. The bottom panel in Figure 5B shows the predictions of a response compression model in terms of incremental responses and incremental light intensities. For this incremental response function, response compression yields an increase in K and a decrease in Rm (Normann and Perlman, 1979~. If a particular value of the dark-adapted Ka is assumed, the response compression model provides predicted values of Ka and Am as a function of adapting intensity. By substituting these values of Ka and Am in Equations (6) and (7), the changes in Rm and K, relative to their dark-adapted values, can be derived. The solid curves in Figure 6 show the predicted changes in Rm (panel B) and K (panel A) assuming a response compression model and a dark-adapted value for Ka of 102 3 scotopic trolands (td). The open symbols, from Fulton and Rushton and Figure 3, are the changes in log K (panel A) and log Rm (panel B) plotted as a function of adapting field intensity. We can reject the receptor as the site of most of the sensitivity 1 In the clinical literature, "response compression" is often used to signify a decrease in response amplitude without a change in K Here we use the term to mean a decrease in responsiveness secondary to a decrease in the polarization of the receptor. The response compression model described in Figure 5 results in an equal change in (log Ka) and ~log Am) from their dark- adapted values.
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT Response Compression O- 49 0 o In 2 - CJ) J In o ~ _, to it c ID ~ _ 3 O _ ~ 3 - 2 _, _ / ,, A,. 1 Log Total Intensity = 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ~ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ~0 // // // // i/ // _, o 1 2 3 Log Incremental (flash) Intensity ~ s 6 FIGURE 5 Response compression. The solid curve in the upper panel and the left-most curve in the lower panel are the dark-adapted intensity-response functions for a hypothetical receptor. The upper panel plots the total response amplitude versus the total light, flash plus adapting field, for adapting fields of different intensities (dotted lines). The lower panel shows the intensity-response functions in terms of the log of the incremental response above the response of the adapting field versus the log of the incremental intensity above the intensity of the adapting field. SOURCE: Adapted from Hood and F~nkelstein (1986~.
50 DONALD C. HOOD changes occurring with light adaptation over the range of adapting intensi- ties shown.2 From this analysis the changes observed in the semisaturation constant, K, of the ERG can be largely attributed to an increase in Kb. The small change in Rm can be attributed to changes in Bm.3 Using A-Wave Data to Estimate Ka and Am As an alternative to assuming a model of receptor adaptation, we can assume that the a-wave provides a measure of the receptor's response. Fulton and Rushton (1978, their Figure 3a) measured intensity-response functions for the a-wave. The a-wave amplitude is not easily measured in all subjects or at all intensities. As a measure of the a-wave amplitude, Fulton and Rushton used the slope of the a-wave. If we assume that this measure is a good substitute for the rod receptor's response amplitude, then their intensi~-response functions provide a measure of the change in Ka and Am for each adapting intensity. Using Equations (6) and (7) and the Fulton and Rushton values of Ka and Am for each adapting intensity, the predicted changes in K and Rm attributable to receptor changes were calculated. The small crosses in Figure 6 show the expected changes in log K and Rm based on the observed changes in Ka and Am of the a-wave. The conclusions from this analysis are the same; most of the change in K is caused by an increase in Kb and the small change in Rm by a decrease in Bm. (The agreement between the Fulton and Rushton a-wave data and the response compression model of the receptors is better than it appears in Figure 6. If the model were fitted to the a-wave data, the solid curve in Figure 6A would be shifted vertically. Most of the variation from the model is caused by the dark-adapted data point.) Assessing the Approach With both procedures for estimating Ka and Am, we reject the receptor as the primary site of sensitivity change for adapting fields up to 1.4 log scotopic trolands. In fact, we can conclude that the small changes in Rm are attributable to changes in Bm of the inner nuclear layer lo-wave generators and that the change in log K is largely due to an increase in Kb (cellular adaptation in the inner nuclear layer). This is the conclusion we should have expected from previous physiological and psychophysical work (see Shapley and Enroth-Cugell, 1985, and Walraven et al., 1989, for reviews). 2This conclusion is restricted to the range of adapting intensities in Figure 6. At higher adapting field values, the changes in Ka andAm make major contributions to changes in both K and Rm. 3According to the model, the difference between the solid curve and the data provide measures of log Kb (panel A) and log Bm (panel B).
ERG, RETINAL DISEASE, ADAPTATION' AND DEVELOPMENT A. 3.0 Y 2.0- ._ it_ 1.0- ~: o ._ o B. o.o - -1.0 0.5 ~ 0.0 ~ -0.5 ~ -1 .0 - -1.5 ~ -2.0 m LO to to ~ + - 11 - ~ to + ~ d.a. -2.0 -1.0 o.o 1.0 Log Adapting Intensity (td) ~ ~ m b O . . .. . . . . . . . d.a. -2.0 -1.0 o.o 1.0 Log Adapting Intensity (td) 51 FIGURE 6 (A) Log relative Rm and (B) log relative K versus log adapting field intensities. The values O log K or Bm represents a dark-adapted value. The open symbols are from the Fulton and Rushton (1978) study. The solid curves are the predicted changes in log K and log Rm based on the response compression model of the rod receptors. The crosses show the predicted change in log K and log Rm derived from the Fulton and Rushton a-wave data and Equations (6) and (7~.
52 DONALD C HOOD The main purpose here, however, was not to test hypotheses about rod system adaptation, but rather to evaluate an approach to inferring sites and mechanisms of disease-related sensitivity changes from ERG data. If the results of this analysis turned out to be inconsistent with previous work, the ERG would not have been the topic of this paper. The results are, however, encouraging. Let us consider the application of the approach to the problem of retinal disease. RETINAL DISEASE AND ERG CHANGES Various retinal diseases have been shown to increase log K and de- crease log Rm of the lo-wave. Figure 7 shows data from three clinical studies. Each data point represents the log Rm and log K for a single eye from a single patient. Open symbols are from two studies (Birch and Fish, 1987; Massof et al., 1984) measuring ERGs of patients with retinitis pigmentosa (RP), and the filled symbols are from a study examining the ERGs of patients with central vein occlusion (Johnson and Hood, 1988~. Both diseases decrease the maximum lo-wave response and increase the value of K. However, the patterns of the results direr substantially. The patients with central vein occlusion show relatively larger changes in K than the patients with RP. ~ determine whether the changes in K and Rm are due to receptor changes, the approach described above can be applied. We can assume a model of receptor change and ask if it describes the changes in K and Rm, or we can use a-wave recordings to obtain estimates of Ka and Am. RP will be used to illustrate the first approach and central vein occlusion the second. Consider RP (filled symbols) first. RP is an inherited disease that affects the pigment epithelium and thus the receptors. A loss in visual pigment has been documented with both retinal densitometry and psy- chophysical matching techniques. Histopathological studies show outer segments that have been shortened and receptors that are twisted and dis- torted. Ultimately the disease results in the loss of receptor cells. Ibble 2 lists seven hypothesized actions of RP on the receptors (labeled 2 through 8~. The other columns show the effects on the model's parameters (column 2) and the predicted changes in ~ and Rm (column 3~. Notice that except for hypothesis 8 the receptor hypotheses translate into changes in Ka and Am. And these changes in Ha and Am result largely in changes in K. Based on these hypotheses, there should be little or no change in Rm. Are the changes in K and Rm produced by RP occurring at the receptors? Suppose we assume an explicit model of the receptor. If we assume a value of Ka between 1.8 and 2.8 log scotopic trolands, the curves labeled increased Ka and decreased Am in Figure 4 show the range of
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT 0.5 - E o.o ~ ~.5- ~_ - a: -1.0 - a -1.5 -2.0 - -0.5 o.o _ 1 porma _ , ,_ _ _ . - O. ~ .10 ~ on 0= ~ ° ~ ~ a O0 0 0 ~ a o a a a · , · I · I to 0.5 1.0 1.5 2.0 Log Relative K 53 FIGURE 7 Log relative Rm versus log relative K for ERG intensity-response functions recorded from patients with retinal disease. Each data point is for a single eye of a single patient. The filled symbols represent data from patients with central vein occlusion (Johnson and Massof, 1988~. The open symbols represent data from RP patients from two studies (Birch and Fish, 1987 squares; Massof et al., 1984~tnangles). The Rm and K values are divided lay mean normal data for the RP patients and by the parameters for the unaffected eye for the central vein occlusion patients. The 0.0 value represents normal mean value or the value for the normal eye. predicted changes in Rm and K for hypotheses 2 through 7 in Able 2. The envelope of these curves is reproduced in Figure 7. We cannot reject the receptor as the major cause of the change in log K. Can we, however, reject the receptor as the only cause of the changes in the ERG with RP? Hypotheses 1 through 7 can be rejected, as each predicts little or no change in log Rm. A loss of visual pigment, shortened or distorted receptors, or a decrease in the responsiveness of the receptors cannot account for the ERG results. Only one receptor hypothesis is capable of explaining the large decrease in log Rm. Hypothesis ~ in Table 2, if coupled with one or more of the other receptor hypotheses, could in theory account for the changes in both K and Rm. Large patches of receptor
54 TABLE 2 Hypothesized Disease Action and Predicted ERG Changes DONALD C HOOD Disease Action Change in Model Predicted lo-Wave K Rm Prereceptoral 1. Preretinal screening (e.g., cataract, increase in macular pigment, blood) Receptor 2. Ha Decreased quantal catch Doss of pigment) 3. Decreased quantal catch (misaligned or misshapened receptors) 4. Shortening of outer segments 5. Random loss of receptors 6. Increased membrane potential (response compression) 7. Decrease in response amplitude 8. Large regional loss of receptors IKa Ha Ha ]'Am? MA MA = OK m a lAm JIB m normal (n) n n An An An An n ~1 loss could lead to a loss of lo-wave generators and a decrease in both Bm and Rm. However, the rod visual fields in these patients are probably too large to produce a sufficient decrease in Bm based simply on a loss of patches of receptors4 (Birch et al., 1987~. Although additional analyses of field data must be done and electrical shunting must be considered, it is likely that the ERG data will be consistent with recent psychophysical data (Greenstein and Hood, 1986; Hood and Greenstein, 1988b) in supporting an inner nuclear component to the sensitivity loss observed in RP. 4To generate predictions for a regional or patchy loss of receptors, a spatial dimension must be added to the model in Table 1. Assume that there are a finite number of lo-wave generators and that the response of each is given by Bi = A + K Bm for all i, where Ai is the pooled response of the receptors feeding into the ith lo-wave generator. Note that we are assuming that all lo-wave generators have the same Kb and Bm. Assuming an equal contribution of all lo-wave generators to the lo-wave recoded at the electrode, Equation (4) becomes R = ~ ~ pi i A patchy loss of receptors removes a subset of lo-wave generators by removing their receptor pools. A loss of all the receptors contributing to A: eliminates the response of Bi. Under these assumptions a regional loss of one-half of the visual field, by itself, would reduce log Rm by 0.3 log unit.
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT 55 The open symbols in Figure 7 represent data from patients with central vein occlusion. In this disease the blood flow to the inner nuclear layer of one eye is disrupted. The changes in K and Rm of this eye, relative to the normal eye, are shown by the open square. Ophthalmologists must determine which of these patients will go on to develop neovascularization of the iris. Neovascularization not only leads to further loss of vision but can also result in the loss of the eye. Johnson et al. (1988) have shown that changes in the ERG, especially log K, are four times more sensitive than fluorescein angiography in identifying patients at risk. Are the large increases in log K due to receptor changes (Ka or Am) or to changes in the inner nuclear layer (Kb)? Johnson and Massof (19~) have analyzed a-wave data and found that the increase in log K is largely due to an increase in Ka; that is, the increase in the semisaturation constant of the lo-wave in these patients is largely due to an increase in the semisaturation constant of the receptors. It appears that changes in receptor sensitivity may be a good predictor of which patients will develop neovascularization. This is particularly interesting since central vein occlusion disrupts blood flow to the inner nuclear layer while leaving the choroidal blood supply of the receptors intact. Disrupting the blood flow to the inner nuclear layer appears to compromise receptor responsiveness. DEVELOPMENT AND ERG CHANGES Over the first 6 months or more of life the most fundamental of visual capabilities, the detection of light, continues to improve. Fulton and Hansen (1982) have measured ERG intensity-response functions of neonates. The changes they measured in log Rm and log K for infants ranging in age from 1.5 to 12 months are shown in Figure 8 (filled symbols). The values of Rm and K are expressed relative to the mean for a group of adults. The adaptation data (open squares) from Figure 3 are shown for comparison. The developmental data show larger changes in Rm for equivalent changes in sensitivity than do the adaptation data. The theoretical curves are the same as those in Figure 4. From these data and the analysis above, we can reject the receptor as the only cause for the developmental changes. Since the changes in Rm are largely attributable to an increase in the response amplitude of the inner nuclear layer cells, these ERG data suggest some postnatal development of the inner nuclear layer. The nature of the postreceptoral development can be further constrained by testing explicit hypotheses about the development of the inner nuclear layer and by analyzing behavioral data (Hood, 1988~.
56 DONALD G HOOD 0.5 ~ 0.0~ ~ -0.5 CC -1 .0 - lo -1 .5 ~ -2.0 ~ adult 6mo ~ 1 i ll 1.5 mo '' ~ ' 1 ' I ' 1 ' I -0.5 0.0 0.5 1.0 1.5 2.0 Log Relative K FIGURE 8 Log relative Rm versus log relative K for ERG intensity-response functions recorded from infants. The filled symbols are for infants 1.5 to 12 months old and for adults (from Fulton and Hansen, 1982~. The open symbols represent the adult adaptation data from Figure 3. The smooth theoretical curves are the same as in Figure 4. SUMMARY The ERG is a complex potential; we should be cautious in interpreting changes it brought about by disease or development. However, carefully measured intensity-response functions combined with explicit models of the ERG and with data from psychophysical paradigms (e.g., visual fields, matching, increment threshold) provide the possibility of identifying sites and mechanisms of disease action for specific diseases and individual pa- tients. . . ACKNOWLEDGMENTS This research was supported by National Eye Institute grant EY-02115 to D.C. Hood and V. Greenstein. I appreciate the helpful comments
ERG, RETINAL DISEASE, ADAPTATION, AND DEVELOPMENT 57 provided by M. Johnson, V. Greenstein, and D. Birch on an earlier version of the manuscript and am particularly grateful to M. Johnson for her tutorials and for reacquainting me with the ERG. REFERENCES Arden, G.B., R.M. Carter, C.R. Hogg, D.F. Powell, W.J.K. Ernst, G.M. Clover, ALA Lyness, and M.P. Quinian 1983 A modified ERG technique and the results obtained in x-linked retinitis pigmentosa. British Journal of OpAthalmolo~ 67:419~30. Baylor, D.A., B.J. Nunn, and J.L. Schnapf 1984 The photocurrent, noise, and spectral sensitivity of rods of the monkey Macaca fascicularis. Joumal of Physiology 357:575~07. Birch, D.G., and G.E. Fish 1987 Rod ERGs in retinitis pigmentosa and cone-rod degeneration. Investigative Ophthalmologic and Visual Science 28:14~150. Birch, D.G., WK. Herman, J.M. deFaller, D.T. Disbrow, and E.E. Birch 1987 The relationship between rod perimetric thresholds and full-field ERGs in retinitis pigmentosa. Investigative OpAtha*nology and Visual Science 28:954- 965. Boynton, R.M., and D.N. Whitten 1970 Visual adaptation on monkey cones: recordings of late receptor potentials. Science 170:1423 1426. Brown, K.T. 1968 The electroretinogram: its components and their origin. Vision Research 8:633-678. Fulton, A.B., and R.M. Hansen 1982 Background adaptation in human infants. Documenta Ophthalmolog~ca Proceeding Series 31:191-197. Fulton, A.B., and W.NH. Rushton 1978 The human rod ERG. Correlation with psychophysical responses in light and dark adaptation. Vision Reseach 18:793-800. Granit, R. 1947 Sensory Mechanisms of the Retina. London: Oxford University Press. Greenstein, V.C., and D. C. Hood 1986 Test of the decreased responsiveness hypothesis in retinitis pigmentosa. American Joumal of Optometry and Physiological Optics 63:22-27. Hood, D.C. 1988 Testing hypotheses about development with ERG and incremental threshold data. Joumal of the Optical Sociery of America 5:2159-2165. Hood , D. C. , and M .A. Finkelstein 1986 Visual sensitivity. In Handbook of Perception and Human Perfonnance, vol. 1, K Boff, ~ Kaufman, and J. Thomas, eds. New York: Wiley. Hood, D.C., and V.C. Greenstein 1988a Blues (s) cone vulnerability: a test of a fragile receptor hypothesis. Applied Optics 27:1025-1029. 1988b Increment threshold (tvi) data and the site of disease action. OSA Technical Digest 3:~5.
58 lDONALD C HOOD Johnson, M.A., and D.C. Hood 1988 A theoretical interpretation of ERG abnormalities in central retinal vein occlusion. OSA Technical Digest 3:&~87. Johnson, M.A., and R.W. Massof 1988 Photoreceptor sensitivity loss in patients with central retinal vein occlusion and iris neovasularization. Investigative Ophthalmology and Jovial Science (Suppl.~:179. Johnson, M.A., S. Marcus, MJ. Elman, and TO. McPhee 1988 Electroretinographic abnormalities associated with neovascularization in cen- tral occlusion. Archives of Ophthabnology 75:51~517. Massof, R.W., L. Wu, D. Finkelstein, C. Perry, S.O. Stair, and M.N Johnson 1984 Properties of electroretinographic intensity-response functions in retinitis pigmentosa. Documenta Ophthalmolog~ca 57:279-296. Normann, R.A., and I. Perlman 1979 Evaluating sensitivity changing mechanisms in light-adapted photoreceptors. Vision Research 19:391-394. Rushton, W.NH. 1965 The Ferrier lecture: visual adaptation. Proceeding; of the Royal Society of London B 162:20~6. Sakitt, B. 1976 Psychophysical correlates of photoreceptor activity. Vision Research 16: 12 140. Shapley, R., and C. Enroth-Cugell 1985 Visual adaptation and retinal gain controls. Progress in Retinal Research 3:26~346. Walraven, J., C. Enroth-Cugell, D.C. Hood, D.I.A. MacLeod, and J. Schnapf 1989 The control of visual sensitivity: receptoral and postreceptoral processes. Ch 5 in Visual Perception: The Neurophysiolog~cal Foundations, L. Spillman and J. Werner, eds. New York: Springer-Verlag.
