Brett Jeffrey, PhD - US grants

Specific Aim 1: To examine the sensitivity of the Cambridge Color Test (CCT) and the low vision CCT (LvCCT) to the severity of retinal disease in inherited retinal degeneration(IRD) by comparing color vision status with changes in photoreceptor structure and function. We have recruited 55 subjects with IRDs. Changes in color vision differ between participants with retinitis pigmentosa (RP) and those with cone-rod degenerations (CRD). In RP, all participants with visual acuity from 20/25 to 20/40 had tritan changes in color vision (i.e. confusion along the blue-yellow axis). With worsening visual acuity, the level of dyschromatopsia progressed in RP but the type of color defect were non-specific ( i.e. did not align with classical lines of red-green or blue-yellow color confusion). Our measure of dyschromatopsia, achromatic area (AA) correlated with logMAR acuity (p<0.001), central retinal thickness (p<0.001) and log cycle x cycle ERG amplitude (p<0.004) in RP. In participants with CRD and Stargardt Disease, the relationship between changes in color vision and structural and other functional measures was more complex than for RP. As for RP, achromatic area was inversely correlated with logMAR (P<0.0001; R2 = 0.3). However, there was wide variation in achromatic area for a given visual acuity. For example, CRD participants with 20/125 visual acuity ranged from normal color vision to those who were functionally achromatic (i.e. unable to discriminate any color). Conversely, normal color vision was associated with visual acuities ranging from 20/16 to 20/160. Ten of thirteen subjects with visual acuity of 20/320 or worse had measureable color vision highlighting the utility of the LvCCT in the low vision population. Islands of retinal preservation at or near the fovea were associated with better color vision even in the presence of widespread geographic atrophy. In CRD participants, scotopic, photopic and 30 Hz ERG amplitudes were inversely correlated (P=0.010; P<0.0001; P=0.003) with achromatic area. Similarly scotopic, photopic and 30 Hz ERG implicit times were positively correlated with logAA (P=0.021; P<0.007; P<0.007). In summary, our experience with the LvCCT in patients with retinal degenerations has shown the utility of this test in low vision patients. Many patients with advanced RP may have good visual acuity but narrow visual fields and subsequently no measurable ERG response. Conversely, CRD patients often have reduced acuity in addition to poor and eccentric fixation which, makes classical field testing difficult. We have found color vision is altered in most participants with retinal degeneration and that the LvCCT can be used to measure color vision in most patients with advanced retinal disease. We propose that the test would be good as an outcome measure in clinical trials for retinal degenerations. Specific Aim 2: Examine the effects of eccentric fixation and reduction in visual acuity on the color discrimination thresholds obtained with the CCT and LvCCT. Two subjects have completed the experiment examining the effect of eccentric fixation on color discrimination thresholds. Color thresholds were measured from the fovea and 5, 10 and 15 deg superior to the fovea. Achromatic area increased by 64% at 10 deg eccentricity but was still within the normal range described for healthy volunteers fixing at the fovea. Color discrimination worsened at 15 deg eccentricity, with achromatic area more than double (AA = 75) that obtained with foveal fixation (AA=33). However participants with CRD with fixation >10 deg from the fovea had achromatic areas ranging from 200 to 2472. These results indicate that eccentric fixation alone in CRD participants cannot account for changes in color results. The poorer color discrimination in these participants results from changes in retinal function and structure. Five subjects have completed the experiments examining the effects of reduction in visual acuity on color discrimination thresholds. Experiments with the commercial version of CCT indicate the presence of edge and/or luminance cues that allow participants to achieve substantially better color discrimination thresholds when acuity was reduced down to 20/200- 20/400 using optical blur compared with thresholds at 20/20 acuity. In contrast, with the low vision version of the CCT (LvCCT) that we implemented on a ViSaGe system using custom written software, showed a small 20% increase in color thresholds when acuity was blurred to 20/200-20/400. Further optical blur of visual acuity to 20/800 resulted in a total 40% increase in color thresholds compared with 20/20 acuity. These experiments provide strong evidence that color discrimination can be measured down to 20/800 acuity with the LvCCT and further that absolute thresholds are not greatly affected by the loss of acuity alone. Specific Aim 3: Establish normal ranges for the CCT and LvCCT and determine the inter-session and intra-session variabilities for these two tests. Aim 3a: Normal range: Color discrimination thresholds have been measured from 24 participants aged 5.9 to 60.6 years (median = 25.2 years). Mean ( S.D.) achromatic area for normal subjects was 1.19 0.27 log UV (linear units: mean = 15.6 UV, range of 4.9 54.3 UV). Achromatic area was not correlated with age. Aim 3bi: Intra-session variability: Color discrimination thresholds have been measured in 10 participants twice from the dominant eye in on session. Intra-session variability was 0.165 log UV (linear units = 1.5) which equated to a coefficient of variability of 10%.

The first two Aims of this protocol require establishing the normal ranges for dark-adapted retinal sensitivity and the kinetics of dark adaptation using a new instrument, the Medmont dark-adapted chromatic (DAC) perimeter. Specific Aim 1: To establish the normal ranges of dark-adapted retinal sensitivities for the Medmont red and blue stimuli in healthy volunteers. With 2-color dark-adapted perimetry, calculation of the difference in dark-adapted retinal sensitivity to blue and red stimuli is fundamental to determining whether a sensitivity at a given retinal location is mediated by rods and/or cones. We have established preliminary normal values for mean ( SD) dark-adapted retinal sensitivities for the blue (57.9 1.9 dB) and red (35.5 1.6 dB) stimuli, and for the difference in sensitivity between these two colors (22.5 1.7 dB). The mean difference obtained with the Medmont DAC is close to previously reported difference values of 18 dB and 19 dB for LCD and tungsten sources respectively. The direct implication of our result is that a difference value of less than 19.1 dB (mean 2SD) would indicate abnormal rod function in a patient. Dark-adapted retinal sensitivity was not significantly correlated with age although there was a trend for lower blue sensitivity with age. This question will be more thoroughly evaluated in the coming year by recording from a greater number of healthy volunteers and from older subjects (oldest healthy volunteer to date was 63 years). Specific Aim 2: To establish the normal ranges for the kinetics of dark adaptation for the Medmont DAC blue and red stimuli in healthy volunteers. The time course of dark adaptation has historically been measured clinically following exposure to an achromatic background light that bleaches approximately 50-100% of rhodopsin. Given that dark adaptation can be very delayed (several hours) in patients with AMD and some genetic disorders (e.g. Fundus Albipunctatus) and many patients with retinal degenerations have marked photo aversion, we are seeking to examine the kinetics of dark adaptation to smaller bleaches. We examined the kinetics of dark adaptation to a range of bleaches (30-60%) obtained by varying background light exposures (duration, intensity, color). To date, a 30% rhodopsin bleach obtained by a 5-minute exposure to 505 nm light has been tolerated by all participants and this will be our default for AMD subjects (see below). Specific Aim 3: To quantify local changes in rod and cone photoreceptor function across the retina in participants with retinal disease. Dark adaptation kinetics are known to be delayed in AMD patients. In preparation for a larger study, we examined DA kinetics in 5 participants with varying levels of AMD severity. We examined two spatial patterns and used vary levels of rhodopsin bleaching (30%-60%). We found that the DA kinetics vary with spatial eccentricity and that a 30% bleach will likely be sufficient to characterize AMD patients. In two of the five patients tested, we saw no evidence of rod recovery over the 30 min following a bleach. We also measured dark-adapted retinal sensitivity in six patients with ABCA4 retinopathy and relatively good acuity (20/40 or better). We are seeking to determine whether peripheral rod function is altered in this disease but will need to test a greater number of subjects before we are able to examine this question.

Specific Aim 1: Develop a method for recording the focal EOG in response to a centrally presented light stimulus. A total of 15 healthy volunteers over the course of 1-3 study visits have participated in the study. Initial experiments to induce a dark trough before generating the light peak were shown to be highly variable. We shifted towards elimination of the dark trough in favor of generating a baseline dark amplitude during a long dark adaptation period of 40 minutes. For the light peak, we tested various stimulus sized and found that a 20o central stimulus was insufficient to generate a response from the target area, but a 40o central stimulus could produce the target response. Therefore, our current method is dark adaptation for 25 minutes, recording in the dark (to establish the baseline amplitude) for 20 minutes, then presentation of the light stimulus for 15 minutes. This method produces a recording lasting 35 minutes, which has been well-tolerated by all subjects. In order to test the response to centrally presented 40o light stimuli, volunteers were presented various central stimuli ranging from 10 to 150 cd/m2 in intensity. While 5/9 (56%) of eyes showed a light rise to 10 cd/m2, all 36 eyes tested showed light rises in response to intensities 40 cd/m2. Light peak:baseline averaged 1.370.19. Therefore, 40 cd/m2 appears to be the minimum 40 stimulus necessary and sufficient to generate a light rise. In order to test the variation in response to scattered light from central stimuli, various background intensities ranging from 0.4 cd/m2 to 1.1 cd/m2 were presented to volunteers. At the lowest background tested, participants had no or minimal light rise. The results above indicate that our method is able to measure an EOG light rise using a centrally presented 40 deg stimulus. Specific Aim 2: Testing in healthy volunteers to determine intra- and inter-session variability. We have not yet attempted to measure variability in the focal EOG response. Specific Aim 3: Examine the focal EOG in participants with macular disease. Our results from testing healthy volunteers suggests that method provides the capability to measure an EOG light rise from a centrally presented 40 deg stimulus. We then conducted proof-of-principle experiments in 6 patients with ABCA4 retinopathy who had foveal preservation with good acuity and whom had contiguous atrophy across the macular but a normal peripheral retina. We hypothesized that these patients would have a normal full-field EOG but no focal EOG, if indeed there is no light scatter from the 40 deg stimulus. Two subjects did have normal full-field EOG with no focal EOG as predicted which supports our hypothesis. However, three other patients with ABCA4 patients had a light rise with both full-field and focal EOGs. There are two possible explanations of these results: First, the level of atrophy across the macula was not sufficient to reduce the focal EOG. Second, that the apparent focal EOG light rise was caused by scattered light. Therefore, these experiments did not provide conclusive evidence that the focal EOG is being driven by the central retina without a contribution from scattered light. Therefore, we are now enrolling patients with uveal colobomas larger than 40 deg in diameter and which do not involve the fovea. These patients have a clear cut loss of retina/RPE complex with surrounding normal retina and therefore, should provide a better test of whether scattered light confounds recording of the focal EOG.