Tomographic Predictors of Foveal Versus Extrafoveal Atrophy Onset in Intermediate Age-Related Macular Degeneration.

Abstract

To evaluate whether baseline foveal tomographic lesions are associated with foveal-first versus extrafoveal-first onset of complete retinal pigment epithelial and outer retinal atrophy (cRORA) in intermediate age-related macular degeneration (iAMD). Retrospective cohort study. A total of 129 eyes of 89 patients with iAMD developing cRORA. Consecutive eyes with iAMD progressing to cRORA were identified through back-tracking grading on spectral-domain optical coherence tomography. Eligible eyes had at least 3 years of follow-up before cRORA onset, with no pre-cRORA inter-visit gap exceeding 6 months. The foveal region was operationally defined as the 1-mm Early Treatment Diabetic Retinopathy Study central subfield. Baseline foveal lesions included soft drusen, subretinal drusenoid deposits (SDD), drusenoid pigment epithelial detachment (dPED), acquired vitelliform lesions (AVL), and incomplete retinal pigment epithelium and outer retinal atrophy (iRORA). Hyperreflective foci (HRF) were recorded as present/absent. Foveal-first versus extrafoveal-first cRORA onset; interval-censored time from extrafoveal cRORA onset to foveal involvement. Over a mean follow-up of 10.4 ± 3.4 years, 108 eyes (83.7%) developed foveal cRORA. cRORA onset was foveal-first in 60 eyes (46.5%) and extrafoveal-first in 69 eyes (53.5%). Onset pattern differed across baseline lesions (P < .001). SDD and soft drusen were more frequently associated with extrafoveal-first onset (20/22 [91%] and 25/40 [63%]), whereas AVL, iRORA, and dPED were more frequently associated with foveal-first onset (16/23 [70%], 7/10 [70%], and 20/34 [59%]). Among 48 extrafoveal-first eyes that reached the fovea, time to foveal involvement was similar across lesion types (median, 1.5-1.8 years; P = .97). Compared with soft drusen, the time to foveal cRORA was shorter for iRORA, dPED, and AVL (time ratios: 0.12, 0.40, and 0.50; 95% CIs: 0.06-0.23, 0.30-0.55, and 0.35-0.71). Foveal HRF independently predicted faster progression (time ratio, 0.66; 95% CI, 0.51-0.85; P = .001). In iAMD eyes progressing to cRORA, distinct tomographic lesions at the fovea were associated with a different propensity to foveal-first versus extrafoveal-first atrophic onset. Once extrafoveal cRORA was established, time to foveal involvement did not differ across tomographic lesions. These findings may inform lesion-specific risk stratification for emerging dry AMD therapies.