Mesopic microperimetry has added benefit over visual acuity in explaining self-reported visual function in age-related macular degeneration: a cross-sectional study

Abstract

Purpose : Mesopic Microperimetry (MMP) is a promising tool to evaluate retinal function in clinical trials. Visual Function (VF) expresses the ability to cope with vision related tasks. It relates strongly to Visual Acuity (VA) in patients with Age-related Macular Degeneration (AMD). However, it is still unclear to what extent MMP is a reflection of VF. Therefore, we performed this study which investigates the relationship between VF, retinal sensitivity, and VA.

Methods : We performed this cross-sectional study with data collected from January 2018 through April 2019. We selected AMD patients with available MMP and VA data, who contributed a National Eye Institute VF questionnaire. MMP was performed with the Macular Integrity Assessment Microperimeter (S-MAIA) (CenterVue, Padova, Italy) using a 4–2 strategy with 10° diameter grid containing 37 loci. We computed the Long Form Visual Function Score (LFVFS-39) with questionnaire data. We statistically modelled VF three separate times via linear regression controlling for VA of the best seeing eye and worse seeing eye, sex, and retinal sensitivity. The LFVFS-39 served as the dependent variable. In each iteration of the same model we used a different value for retinal sensitivity: median, geometric mean, mean. The mean was calculated by reconverting the 37 decibel values back to candela per square meter prior to averaging. Our outcomes include proportion of variance explained (R²) (F-test) as well as the standardized beta to evaluate the relative predictive strength of MMP. A p value below .05 was considered significant.

Results : We included 104 AMD patients. The majority (90/104) had late AMD in at least one eye. MMP changed VF (beta .303 p < .001 for mean sensitivity). Visual acuity (beta -.410 p < .001) changed VF for visual acuity of the best seeing eye. Visual acuity of the best seeing eye explained 46.4% of the total variance (p <.001). Median (R² .027 p = .012), geometric mean (R² .056 p < .001), and mean sensitivity (R² .077 p < .001) all explained additional variance in their models, while mean sensitivity performed best .

Conclusions : In this study MMP is in accordance with self-reported VF. MMP has complementary merit to VA in providing an objective measure of VF in AMD. This finding suggests that MMP has added value as secondary or primary endpoints in upcoming AMD trials.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.