Abstract
Purpose :
Fixed grid microperimetry provides limited information in scotomatous areas. Feasibility and test-retest reliability of retinal sensitivity were determined in perilesional zones using a patient-customized microperimetry grid.
Methods :
Two sites in the United States enrolled patients with geographic atrophy (n=11). Scotopic and mesopic microperimetry were done with an S-MAIA at Baseline, Day 7, Month 3, and Month 6. Placement of test points were as follows: 9 points on the patient's preferred retinal locus; ~50 evenly distributed points in two concentric rings 200 μm (Zone 1) and 450 μm (Zone 2) from the lesion border (Figure 1). Spearman correlation, Bland-Altman analyses, and the coefficient of repeatability (CR) quantified test-retest (Baseline-Day 7) reliability. Mean sensitivity readings per zone and illumination condition were used to estimate retinal zone sensitivities (decibels, dB) and annual progression rates (dB/year). All analyses were conducted using R Software (v4.2.2) under RStudio.
Results :
Baseline characteristics are reported in a previous publication (Issa M, et al. IOVS 2022;63:ARVO E-Abstract 4544). Compared to scotopic, mesopic microperimetry resulted in stronger, positive correlations (p<0.05) and smaller Bland-Altman limits of agreement across the zones. Likewise, the range of CRs in all zones for scotopic was 14.5-31.7 dB vs. 5.0-7.8 dB for mesopic. Table 1 shows mean zone sensitivity by visit. For both illumination conditions, the progression rates typically showed a decrease in mean zone sensitivity over time.
Conclusions :
Scotopic microperimetry, in general, showed higher variability and reduced reliability than mesopic microperimetry, appearing unreliable in this patient population. Under mesopic conditions, reliability assessments indicated that changes greater than ~8 dB could be considered meaningful. In conclusion, this study demonstrates the feasibility of patient-customized grids, quantifies the reliability of the assessments, and provides the means for evaluation of therapeutic effects of drugs at diverse perilesional locations.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.