Abstract
Purpose :
Compare retinal function measured using multifocal electroretinogram (mfERG) with microperimety (MP) to investigate the utility of MP as a surrogate screening test for hydroxychloroquine (HCQ) retinopathy.
Methods :
Retinal function tests were compared between 25 patients referred for evaluation of HCQ retinopathy after ≥ 5 years of HCQ use and 42 normal controls. The 3 innermost central rings of mfERG and MP responses at approximately equivalent distances from the fovea were compared (mfERG - R1:0°-1°, R2:1°-4°, R3:4°-8°) and (MP - MR1: 1°, MR2: 3°, MR3: 5°). mfERG ring ratios were calculated as the average response density of R1, R2 or R3 divided by the mean of R5. Average retinal sensitivity of each MP ring was calculated. Pearson correlations (r) were used to test for a linear association between mfERG and MP measures in HCQ patients and t-tests were used to compare mfERG and MP measures between HCQ patients and normal controls. Gold standard HCQ toxicity was defined as an mfERG ring ratio >2 standard deviations (SD) below the mean observed in the normal sample. A similar calculation was used to define toxicity with MP measures. Sensitivity and specificity to detect HCQ toxicity with MP measures compared to mfERG were calculated, including 95% Wilson confidence intervals (CI).
Results :
46 eyes of 25 HCQ patients and 52 eyes of 42 normal controls were evaluated. HCQ cases were on average 56.6 years old (SD=10.7). They were treated on average for 11.1 years (SD=6.3), with a mean exposure to HCQ of 1440 g (SD=860). mfERG R2 and R3 ring ratios were positively correlated with their corresponding MP ring averages (both r=0.55, P<0.001). mfERG R2 and R3 ratios of HCQ cases were significantly lower than controls (R2: P=0.0105, R3: P=0.0003). All three MP rings for HCQ patients were significantly lower than those of controls (P<0.0001). 2 HCQ patients (4%) were categorized as having toxicity from mfERG R1 ratio, 6 (13%) from R2 ratio, 13 (28%) from R3 ratio, and 14 (30%) if detected by any ring ratio. Sensitivity of MP to detect HCQ toxicity compared to mfERG was 100% (CI:34-100%) for MR1, 100% (CI:61-100%) for MR2, 85% (CI:58-96%) for MR3, and 86% (CI:60-96%) over all rings. Specificity was 61%, 55%, 55%, and 53%, respectively.
Conclusions :
MP shows high sensitivity to detect HCQ toxicity when compared to mfERG. However, the low specificity suggests that MP use will classify many healthy individuals as having HCQ toxicity.
This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.