Rod and Cone Photoreceptor Function in Patients with Cone Dystrophy

Abstract

purpose. To determine the extent of rod and cone photoreceptor dysfunction in patients with cone dystrophy using psychophysical and electrophysiological tests.

methods. Ten patients with cone dystrophy participated. Rod and cone system psychophysical thresholds were measured as a function of retinal eccentricity. Bright-flash full-field electroretinograms were obtained under dark-adapted (rod-mediated) and light-adapted (cone-mediated) conditions. The a-wave data were fitted with a model based on photopigment transduction to obtain values for log R _max (maximum response) and log S (sensitivity). b-Wave parameters were also examined by fitting a nonlinear, saturating function (the Naka-Rushton equation) to the rod-mediated responses. Oscillatory potentials were measured to the cone-mediated high-intensity flashes.

results. On average, the rod-mediated psychophysical thresholds were elevated by 0.5 log unit. These threshold elevations did not differ significantly with retinal eccentricity. In contrast, cone-mediated psychophysical thresholds were elevated up to 3.0 log units. Threshold elevation was greatest in the central retinal locations. For rod-mediated conditions, the a-wave R _max parameter was significantly reduced in three patients; the a-wave log S parameter was within normal limits. The rod-mediated b-wave R _max parameter was reduced in six patients; log k was abnormal in one patient. For cone-mediated conditions, the a-wave R _max parameter was reduced in six patients and the a-wave log S parameter was reduced in two patients. The cone system oscillatory potentials were abnormal in nine patients.

conclusions. Patients with cone dystrophy show different patterns of psychophysical rod versus cone system sensitivity losses with retinal eccentricity. The full-field electrophysiological data indicate that most of the patients had abnormal cone photoreceptor function. Some patients also showed rod photoreceptor abnormalities. The rod system changes were smaller than the cone system changes.

Cone dystrophy (CD) is an inherited retinal degenerative disease that affects cone system function. This is evidenced by the typical symptoms of CD, which include reduced visual acuity, photophobia, sensitivity to glare, and abnormal color vision. ^{1 2 3 4 5} In addition, psychophysical dark-adapted cone system thresholds may be elevated by more than 1.0 log unit ^{2 3 5} and cone-mediated full-field and multifocal ERGS show large reductions in amplitude and delayed implicit times. ^{1 2 3 4 5 6 7 8 9 10}  

The rod system deficits in patients with CD are less severe than are the cone system deficits. For example, psychophysical dark-adapted rod system thresholds are typically elevated by approximately 0.5 log unit. ^{2 3 5} Rod-mediated full-field ERGs range from normal to approximately 50% reductions in b-wave amplitude; likewise, rod-mediated implicit times can be either normal or delayed. ^{1 2 3 4 5 6 7 10} Rod multifocal (mf)ERG responses show relatively isolated losses, with many responses within normal limits. ¹⁰  

In addition to the difference in the magnitude of the cone and rod system losses, there are differences in the pattern of disease-related changes with retinal eccentricity for the cone and rod systems. ¹⁰ In our previous work, we found that the cone-mediated mfERG amplitudes and psychophysical threshold elevations showed a pattern of impairment that was greatest at the central retina and monotonically decreased with eccentricity. The rod system losses were essentially invariant with retinal location. ¹⁰ Because both the severity of the deficits and the spatial topography of loss differ in the cone and rod systems, it is possible that the underlying retinal site(s) of the losses may also differ. In fact, we know very little about the retinal sites of cone system losses or rod system losses in this disease. Therefore, in the present study, we examined the site(s) of disease-related changes in patients with CD. To do this, we applied models to examine cone and rod photoreceptor function. ^{11 12 13} We also examined full-field ERG b-waves and oscillatory potentials (OPs). In addition, cone and rod system psychophysical thresholds were obtained to provide topographical information. 

Methods

Subjects

Ten patients (mean age, 31.6 ± 14.1 years) with CD recruited from the practice of one of the authors (REC) participated in this study. The diagnosis of CD was based on the patient’s history, visual acuity, color vision (assessed with the FM-100 hue and D-15 tests using a standard illuminant-C lamp), full-field ERG results (according to International Society for Clinical Electrophysiology of Vision [ISCEV] standards) and fundus examination (Table 1) . The patients had no evidence of any other ocular or systemic diseases. The patients’ best corrected visual acuity ranged from 20/25 to 20/200 (see Table 1 ). Patients (P)1, P4, and P7 were siblings from a family with dominantly inherited disease; patients P5 and P6 were the offspring of P7. None of the other patients were from the same family, nor did any of the other patients have a family history of CD. The control group (mean age, 44.1 ± 5.6 years) consisted of seven age-similar observers with best corrected visual acuity of 20/20 or more, normal findings in ophthalmic examination, and no evidence of any other ocular or systemic diseases. All subjects gave informed consent to participate after a full explanation of the procedure, and the protocol complied with the tenets of the Declaration of Helsinki. The institutional human experimentation committee at New York University School of Medicine approved the research. 

Apparatus and Procedure

In all subjects, the eye with the better visual acuity was tested. If visual acuity was equivalent in the two eyes, the right eye was tested. The contralateral eye was patched. The tests were conducted in the order in which they are presented. 

Cone System Threshold Visual Fields.

Cone system threshold visual fields were measured using a Humphrey perimeter (Carl Zeis Meditec, Dublin, CA). The standard program was modified to assess thresholds at 103 locations, including a foveal threshold location. ¹⁴ Each test spot subtended 26 minutes (0.43°) and the test locations extended to 23° (see Fig. 1B for a schematic of the display). The background luminance of the display was 10 cd/m². Each subject’s vision was best corrected for the viewing distance of 32 cm. 

Rod System Threshold Visual Fields.

After pupil dilation (1% tropicamide and 2.5% phenylephrine hydrochloride), the tested eye was patched and dark adapted for 40 minutes. Dark-adapted, two-color, rod system thresholds were measured with a modified perimeter (Humphrey; Carl Zeiss Meditec). ¹⁵ Thresholds were measured for both 500- and 650-nm lights as a function of eccentricity from 6° to 72° at 73 test locations. Each test spot subtended 1.7° (see Fig. 1A for a schematic of the display). 

Full-Field Flash ERGs.

The full-field ERGs were recorded after dark adaptation. Rod- and cone-mediated full-field flash ERGs were recorded with a bipolar contact lens (Doran Instruments, Littleton, MA). The earlobe served as the ground. The signal was amplified (1 K; preamplifier P511; Grass Instruments, Boston, MA) and filtered (1–10 kHz). The light flashes were generated with a photostimulator (set at 150 W/s; model 600; Novatron of Dallas, Dallas, TX). To obtain responses for the dark-adapted b-wave analysis, ERGs were recorded with a filter (Wratten 47B; Eastman Kodak, Rochester, NY). Dark-adapted rod-mediated ERGs were recorded as a function of increasing stimulus intensity at stimulus intensities from −0.70 to 3.0 log scotopic troland [td]-s. To isolate rod responses, cone-matched responses were recorded to the filter (Wratten 47B; Eastman Kodak) under light-adapted conditions, and these responses were subtracted from the dark-adapted responses. ^{16 17} To obtain responses for the a-wave analysis, ERGs were recorded to a series of white-light flashes as a function of stimulus intensity. The procedure used to isolate rod and cone responses has been previously described. ¹³ Under dark-adapted conditions, ERGs were recorded to stimulus intensities of 3.5 to 5.0 log scotopic td-s. To avoid light-adaptation effects with the brighter flashes, the minimum interstimulus interval was 60 seconds. To ensure that there was no cone intrusion in these responses, the ERG responses to the same flashes were recorded under light-adapted conditions and subtracted from the responses recorded under dark-adapted conditions. The subjects then adapted to a steady Ganzfeld background (3.1 log photopic td) for 10 minutes. After light adaptation, cone-mediated ERGs were recorded for the photopic a-wave analysis to a series of white-light flashes as a function of flash intensity (3.1–4.6 log photopic td-s). 

Analysis.

The cone and rod system threshold visual field data were converted to log threshold elevation, computed as the difference between the patient’s log threshold and the average log threshold in the control group, with values greater than 0.0 indicative of elevated thresholds. A version of the Lamb and Pugh ¹⁸ model of phototransduction was fitted to the rod and cone-mediated a-wave responses. ^{11 12 13} To minimize intrusion from the bipolar cells, the first 8 to 12 ms of the leading edge of the a-wave responses were used for analysis. The a-wave data were fitted with models that yielded values for three parameters: log R _max (maximum response amplitude), log S (sensitivity), and td (time delay). For all the fits, td was held constant at the average value of the control group. 

Rod-mediated b-wave amplitudes were measured from the trough of the a-wave to the peak of the b-wave. Amplitudes as a function of flash intensity were fit with a nonlinear saturating function (the Naka-Rushton equation) to obtain values of R _max (a maximum response amplitude parameter) and log k (a sensitivity parameter that represents the flash intensity that produces one-half the maximum response amplitude). To assess cone-mediated b-wave function, it was not possible to fit Naka-Rushton functions to b-wave amplitude, because the responses to these high-intensity “white” flashes do not contain the single prominent peak that is typically measured as the b-wave. Instead, the response is composed of a number of OPs. Therefore, the amplitude of these individual oscillatory components was measured as a function of flash intensity (3.1–4.6 log phot td-s). The amplitude of each OP was measured from the baseline to the peak of the response. 

Results

Rod System Thresholds

To summarize the data, we averaged the thresholds and collapsed them into rings as a function of eccentricity (see Fig. 1A for a schematic of the visual field test locations for the right eye and the rings used for averaging). Figure 2A shows the results for the 500-nm test light and Figure 2B for the 650-nm test light. The results were analyzed by computing the difference between the log threshold values in the patients and the averaged log threshold values in the control group. These difference scores were plotted as log threshold elevation, and the results for each patient are shown as separate symbols. The brackets show the 95% confidence intervals of the control subjects. For both stimulus wavelengths, 7 of the 10 patients had normal thresholds at all the eccentricities examined. In one patient (P8), thresholds were elevated at test locations from 10° to 70°. In two other patients (P1 and P4), thresholds were significantly elevated only at test locations from 50° to 70°. In the test locations between the fovea and 30° (the locations for which cone system thresholds were examined), 9 of the 10 patients had thresholds within normal limits with both stimulus wavelengths. A repeated-measures analysis of variance on the group data indicated that there were no significant differences between the 500- and the 650-nm results nor a significant effect as a function of eccentricity. There was also no significant interaction between test wavelength and eccentricity. 

Rod-Mediated Full-Field ERG Parameters

Figure 3A shows examples of rod-mediated a-wave data and the individual fits of the model in a representative control subject (left) and in a representative patient (right; patient P5 in Table 1 ). For both the rod- and cone-mediated a-wave data, the data were fitted to the first 12 ms (or less) at the lower flash intensities and the first 8 to 10 ms (or less) at the higher flash intensities. Both sets of rod-mediated ERG data are well fitted by the model of rod receptor activity. 

Figure 4 shows the rod-mediated log R _max and log S results derived from the fit of the a-wave model. The results are shown as log difference from the averaged value in the control group, and the results from each patient are shown as separate symbols. The brackets show the 95% confidence intervals from the control subjects. Three patients (P6, P8, and P9) had significant reductions in the R _max parameter (average loss = 0.34 log unit); log S was not significantly reduced in any of the patients. 

Figure 5 shows the log R _max and log k results derived from the fit of the Naka-Rushton equation to the rod-mediated b-wave amplitude data. Again, the brackets show the 95% confidence intervals in the control subjects. For the b-wave data, six patients (P1, P5, P6, P7, P8, and P10) had significant reductions (average loss = 0.38 log unit) in the rod-mediated R _max parameter; log k was significantly abnormal (0.26 log unit increase) in one patient (P3). 

Cone System Thresholds

Figure 6 shows the light-adapted threshold visual field results. The data were averaged in rings as a function of eccentricity out to approximately 25° (see Fig. 1B for a schematic of the visual field test locations and the averaging procedure). All the patients had significantly elevated thresholds within the central 5°; the average threshold elevation in the 10 patients was 0.91 log unit. In all patients, the amount of threshold elevation decreased with retinal eccentricity. For eccentricities between 20° and 30°, only two patients had significantly elevated thresholds. A repeated-measures analysis of variance indicated that there was a significant effect of eccentricity (F_(3,27) = 8.94, P < 0.001). 

Cone-Mediated Full-Field ERG Parameters

Figure 3B shows an example of cone-mediated a-wave data and the individual fits of the model in the same two subjects as in Figure 3A . There were differences in R _max and log S between the patient and the control subject. Figure 7 shows the cone-mediated a-wave log R _max and log S results derived from the fit of the a-wave model. Six of the 10 patients had significant reductions in the R _max parameter; the R _max changes in these six patients averaged 0.74 log units. The log S parameter showed significant reductions (average loss = 0.34 log units) in two patients. 

Figure 8 shows examples of the OPs of the cone-mediated full-field ERG in one control subject (Fig. 8A) and one patient (P6; Fig. 8B ) at three flash intensities. The OPs of the patient were amplified by a factor of three for this figure. Compared with the responses of the control subject, the responses of the patient were reduced in amplitude and delayed. This was true in 9 of the 10 patients. Figure 9 shows the summarized data for the first OP (OP1). The results for the first OP are shown because the later OPs of the patients were sometimes unmeasurable, and the responses from the first OP were the most reliable. Figure 9 shows the amplitude results, plotted in each patient as log amplitude difference from the control subjects. The brackets show the 95% confidence intervals of the control group. Nine of the patients showed significant reductions in log amplitude at all flash intensities. A repeated-measures analysis of variance indicated that there was no significant difference in the amount of amplitude loss as a function of flash intensity (F_(4,32) = 1.52, P = 0.221). 

Discussion

In the present study, cone and rod system function was assessed in a group of patients with CD, by using full-field bright-flash ERG techniques. Based on the nature of the disease, it was expected that these patients would show abnormal cone system function. Consistent with this, 9 of the 10 patients showed abnormal function on at least one of our cone system measures. Previous research has indicated that there are both photoreceptor and postreceptoral abnormalities in patients with CD. ^{19 20 21 22 23 24 25 26 27} The results from genetic analysis have been largely consistent with mutations acting at the level of the photoreceptors. For example, mutations in the gene encoding protein (GCAP-1) involving retinal guanylate cyclase activator GUCA1A, which is active in the dark-adaptation process of photoreceptor cells, have been shown to be associated with autosomal dominant CD. ^{19 20 21} In addition, it has been reported that patients with degenerative cone diseases have a mutation in the GUCY2D gene, which is also related to photoreceptor function. ²² There is also histologic evidence supporting a photoreceptor origin of CD. Gregory-Evans et al. ²³ and To et al. ^{24 25} showed that patients with CD have a pronounced loss of cone photoreceptors throughout the retina. The results of Gregory-Evans et al. ²³ also provide some evidence for cone system postreceptoral changes in patients with CD. Their double-labeling experiments indicate that the Müller cell processes were swollen and pale, and the postsynaptic outer plexiform layer processes were abnormal in these patients. In addition, postreceptoral changes in patients with retinal diseases have been postulated to explain the delays in the standard cone-mediated full-field ERG (e.g., Hood and Birch ²⁶ ) that cannot be explained by photoreceptor outer segment changes (e.g., Hood ²⁷ ) In addition, mfERG data from some patients with RP ¹⁴ and CD ¹⁰ show very large implicit time delays in regions of very poor visual field sensitivity, and these delays have also been suggested to result from postreceptoral effects, perhaps from changes in the outer plexiform layer. ²⁷  

In our group of patients with CD, six showed abnormalities in the cone a-wave maximum response (R _max), with losses in the range of 0.4 to 1.1 log units. In addition, two patients had significant abnormalities in cone a-wave sensitivity (0.35 log unit range). Therefore, in 7 of the 10 patients, the a-wave results are consistent with abnormal photoreceptor function. Can we attribute the pattern of change in the a-wave parameters to specific abnormalities in the response properties of the photoreceptors? In previous ERG studies, large losses in a-wave maximum amplitude with smaller or no consistent losses in sensitivity (as found in the present study) have been attributed to a combination of an overall loss of photoreceptors with perhaps some outer segment shortening in some of the remaining receptors. ¹¹ Therefore, in 70% of our patients, we found a-wave changes consistent with a loss of cone photoreceptors. ^{23 24 25} In the present study, we also examined postreceptoral measures of cone- and rod-mediated function. For the cone system, we could not examine standard measures of b-wave amplitude; therefore, we examined the cone-mediated OPs. All the patients except P3 had abnormal (or nonmeasurable) OP amplitudes. In two patients (P2 and P4), the cone a-wave maximum amplitude and sensitivity were within normal limits, yet the OPs were significantly reduced in amplitude and delayed. This finding is consistent with additional postreceptoral cone system losses in these two patients. 

Based on previous psychophysical and electrophysiological research, it was predicted that the rod photoreceptors would be normal or near normal in this group of patients. Consistent with this prediction, the rod-mediated a-wave parameters were within normal limits in 7 of the 10 patients. Retinal histology of eyes from patients with CD ^{23 24 25} has shown that the number of rods and the rod morphology are essentially normal throughout the retina, although there is some evidence of a slight shortening of the rod outer segments. Recent results of genetic testing have indicated that some patients with CD show mutations in proteins acting at the level of the rod photoreceptors, specifically those involved in the dark-adaptation process. ^{19 20 21 22} These changes could account for the reduced a-wave maximum amplitude found in three of our patients. The rod system b-wave responses were also examined in the present study. The results from the fits of the Naka-Rushton equation indicated that three patients with normal rod a-wave parameters had losses in the b-wave maximum amplitude parameter and one patient had a significant elevation in the sensitivity (log k) parameter. Again, this could indicate some additional postreceptoral changes in the rod system in these patients. 

Why did three patients with CD have full-field cone-mediated a-wave parameters within normal limits? The visual acuity and color vision measures in these three patients (P2, P3, and P4) were not notably different from those in the other patients with CD. Of course, visual acuity and color vision reflect foveal and parafoveal function and the a-wave parameters reflect widespread retinal function. One possibility for the normal cone a-wave responses is that these patients may have had cone photoreceptor abnormalities that were too small to detect with full-field ERG techniques. Unfortunately, our recording techniques do not allow us to assess local measures of photoreceptor function to test this possibility. 

Five of our 10 patients were from a family with dominantly inherited CD, whereas the other five patients were individuals from families with no history of CD. Therefore, these data offer an opportunity to determine whether the results in patients from a single family are more similar than the results in patients with different pedigrees. Of the five patients from the same family, two had rod-mediated psychophysical threshold abnormalities, one had a rod-mediated a-wave abnormality, four had rod-mediated b-wave abnormalities, and four had cone-mediated a-wave abnormalities. Based on these results, there does not appear to be a consistent pattern to the disease-related changes for the members of this family with dominant disease (nor was there a consistent pattern for the group of patients with isolated disease). These findings indicate that there is either phenotypic variation within this genotype, and/or that the patients are at different stages of the disease. The extent of the losses observed among the members of this family was not related to age. 

For the patients examined in this study, there was disparity between the severity of the cone a-wave changes and the severity of the rod a-wave changes. A pair-wise cross-correlation analysis indicated that there was no significant relationship between any measure of cone system function and any measure of rod system function. In addition, there were differences in the amount of threshold elevation as a function of eccentricity for the cone and rod systems. The rod-mediated psychophysical thresholds were within normal limits in seven of the patients. In contrast, the cone-mediated thresholds showed marked abnormalities and were most impaired in the central retinal areas in all 10 of the patients. This dissociation between rod and cone systems is consistent with our previous results in patients with CD, in which we found differences in rod and cone system topography measured both psychophysically and electrophysiologically. ¹⁰  

Studies that have examined photoreceptor function in patients with RP have found a different pattern of a-wave changes than we found in our patients with CD. ^{11 26 28 29} In the patients with RP, cone system sensitivity (log S) was affected in most of the patients, whereas the maximum amplitude parameter (R _max) was quite variable and within normal limits in some of the patients. ²⁶ However, in the rod system, the maximum amplitude parameter was primarily affected, with smaller, less-consistent losses in sensitivity. ^{11 28 29} The authors ^{11 28 29} postulated that the changes in maximum amplitude in the rod system were due to the primary degeneration of the rods in RP, including photoreceptor dropout and structural changes, such as shortened outer segments. In the cones, the degeneration is secondary to the rods, and the changes in a-wave sensitivity have been attributed to an abnormal activation phase of the functioning cones. In our patients, the losses in cone a-wave maximum amplitude are consistent with a primary dropout of the cones with relatively normal activation of the remaining photoreceptors. 

 Supported by a grant from the Foundation Fighting Blindness; an unrestricted grant to the Department of Ophthalmology, NYU Medical Center, from Research to Prevent Blindness, Inc., New York, New York; Grant EY02115 from the National Eye Institute; and grants from the Allene Reuss Memorial Trust, New York, New York, and the Helen Hoffritz Foundation.

 Submitted for publication June 20, 2003; revised September 9, 2003; accepted September 13, 2003.

 Disclosure: K. Holopigian, None; V.C. Greenstein, None; W. Seiple, None; D.C. Hood, None; R.E. Carr, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Clinical Characteristics: Cone Dystrophy Patients

Patient	Fundus	Sex	Age	Eye	Acuity	Color Vision
1	Mild pigment mottling	M	36	OD	20/200	Scotopic axis (D15)
2	Normal	F	27	OD	20/200	Tritan axis
3	Macular granularity; no foveal reflex	M	35	OD	20/200	Protan axis
4	Normal	F	29	OS	20/60+	Scotopic axis (D15)
5	Mild pigment mottling	M	12	OS	20/25	Tritan axis (D15)
6	Mild pigment mottling	M	10	OS	20/200	Tritan axis (D15)
7	Normal	F	39	OD	20/40+	Scotopic axis (D15)
8	Narrowed arterioles; loss of macular RPE	M	57	OD	20/25	Normal
9	Pigment mottling	F	27	OD	20/70	Tritan axis
10	Normal fundus; faint foveal reflex	F	44	OD	20/40	Tritan axis

Figure 1.

 

Schematics of the visual field test under (A) dark-adapted and (B) light adapted conditions showing how the data were averaged into rings as a function of eccentricity.

Figure 2.

 

Dark-adapted psychophysical threshold results. Results for the (A) 500- and (B) 650-nm targets. The results are plotted as log threshold elevation from the control. The error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol.

Figure 3.

 

Examples of a-wave data and the individual fits of the model for rod-mediated (A) and cone-mediated (B) responses. The results are shown for a representative control subject (left) and for a representative patient (right) (P5 in Table 1 ).

Figure 4.

 

Rod-isolated log R _max (left) and log S results (right) derived from the fits of the a-wave model. The results are plotted as log difference from the control. Error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol. Individual points have been displaced for clarity.

Figure 5.

 

Rod-mediated log R _max (left) and log k results (right) derived from the fits of the Naka-Rushton equation to the rod-mediated b-wave amplitude data. The results are plotted as log difference from the control. Error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol. Individual points have been displaced for clarity.

Figure 6.

 

Light-adapted psychophysical thresholds, plotted as log threshold elevation from the control. Error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol. Individual points have been displaced for clarity.

Figure 7.

 

Cone-mediated log R _max (left) and log S results (right) derived from the fits of the a-wave model. The results are plotted as log difference from the control. Error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol.

Figure 8.

 

Examples of the cone-mediated OPs of the full-field ERG for one control subject (A) and one patient (patient P6; B) as a function of stimulus intensity.

Figure 9.

 

The first cone-mediated OP (OP1) results as a function of flash intensity in all the patients. The results are shown as log amplitude difference, relative to the control. Error bars show the 95% confidence intervals of the control data. Each patient’s results are represented by a unique symbol. Individual points have been displaced for clarity.