Purpose: To evaluate the association between spectrally resolved fundus autofluorescence (Sr-FAF) signals and four macular lesion types identified on spectral-domain optical coherence tomography (SD-OCT): (1) drusen, (2) incomplete retinal pigment epithelial and outer retinal atrophy (iRORA) and complete retinal pigment epithelium and outer retinal atrophy (cRORA), (3) retinal pigment epithelium (RPE) and outer retinal atrophy, and (4) nascent geographic atrophy (nGA) in eyes with intermediate and atrophic age-related macular degeneration (AMD). Methods: This cross-sectional observational study analyzed 775 SD-OCT B-scans from 62 eyes (62 patients) with AMD. Lesions were classified according to established OCT criteria (290 drusen, 49 iRORA, 390 cRORA, 46 nGA). All eyes underwent SD-OCT imaging (97 B-scans) and Sr-FAF using a 450-nm wavelength. Quantitative Sr-FAF metrics (average red emission fluorescence component REFC and average green emission fluorescence component GEFC) were determined using custom software, and lesion groups were compared via one-way analysis of variance with Scheffé post hoc testing. Results: Significant differences among lesion types were observed for REFC (F = 59.3, P < 0.001) and GEFC (F = 13.5, P < 0.001). Drusen exhibited higher REFC intensities than iRORA, cRORA, and nGA (all P < 0.001), and higher GEFC intensities than cRORA (P < 0.001) and nGA (P = 0.002). Wavelength (nm) progressively decreased from drusen to iRORA, nGA, and cRORA. Lesion size (in pixels) differed significantly (iRORA < nGA < drusen < cRORA; P < 0.001) and was the strongest predictor (β = −0.31, P < 0.001) in the regression analysis. Conclusions: Sr-FAF reveals distinct metabolic signatures across AMD lesion types. Higher fluorophore signals in drusen suggest preserved photoreceptor–RPE activity, whereas reduced signals in cRORA and nGA reflect advanced atrophy. Combined SD-OCT and Sr-FAF may enhance the detection of early atrophic changes and improve AMD risk stratification.
Vujosevic et al. (Thu,) studied this question.