Re: “Antropoli A et al. Perilesional fundus autofluorescence patterns are not static: longitudinal transitions in geographic atrophy and association with disease progression” Authors: Marc Biarnés Published: 2026 TO THE EDITOR: I read with interest the paper by Antropoli et al1 regarding potential changes in fundus autofluorescence (FAF) patterns related to the progression of geographic atrophy (GA). The authors followed 106 eyes of 76 patients for a median of 3 years and found that 22% of them showed a change in their graded pattern. Therefore, FAF patterns may represent incidental findings along the natural progression of the disease rather than true phenotypic signatures. The role of FAF patterns on GA progression was initially reported in the Prospective Natural History Study of Fundus Autofluorescence Imaging in Age-related Macular Degeneration (FAM study,2 NCT00393692). In Table 3 of that manuscript,2 the association between FAF patterns and the baseline area of atrophy was not statistically significant based on the Fisher exact test (P = 0.07), supporting that both characteristics were independent.2 However, a reanalysis of that data using the likelihood ratio (an equally valid statistical test in this context) shows a P value = 0.046, which, according to an arbitrary P value cut-off of 0.05, would have led to the opposite conclusion. Taken together, these data show inconclusive results in the relationship between FAF patterns and area of atrophy. Later, it was found that FAF patterns were strongly related to GA atrophy size using latent class analysis3 and that almost 85% of the effect of FAF patterns on GA growth rate was mediated by baseline area of atrophy (NCT01694095). It was also shown that increased FAF was caused by a combination of vertical alignment of migrated retinal pigment epithelium (RPE) cells4 and photoreceptor shortening, and not by increased intracellular lipofuscin. Therefore, it seems that enlargement of GA, coupled with the nonuniform retinal topographic distribution of photoreceptors (rods and cones) and RPE (and its lipofuscin content), creates a set of patterns on FAF imaging in which the prognostic value for GA growth is mostly attributable to the area of atrophy itself, a known predictor of RPE atrophy enlargement. Antropoli et al1 found, however, that GA growth rate in mm/year slowed in the eyes that showed a transition to FAF patterns previously associated with faster growth.2 Three reasons may explain that, and I wonder whether the authors could discuss these possibilities:-First, they stated that if atrophy extended beyond the 30° × 30° image field of view, they measured just the visible lesion, which may have biased growth rates downwards for larger lesions.-Second, growth rate, expressed in mm/year, has been shown to decrease for larger lesions,5 as already noted by the authors.-Third, other factors influencing growth rate may change with time (incident foveal involvement, multifocal lesions becoming unifocal, increased lesion circularity, disappearance of reticular pseudodrusen, etc.). -First, they stated that if atrophy extended beyond the 30° × 30° image field of view, they measured just the visible lesion, which may have biased growth rates downwards for larger lesions. First, they stated that if atrophy extended beyond the 30° × 30° image field of view, they measured just the visible lesion, which may have biased growth rates downwards for larger lesions. -Second, growth rate, expressed in mm/year, has been shown to decrease for larger lesions,5 as already noted by the authors. Second, growth rate, expressed in mm/year, has been shown to decrease for larger lesions,5 as already noted by the authors. -Third, other factors influencing growth rate may change with time (incident foveal involvement, multifocal lesions becoming unifocal, increased lesion circularity, disappearance of reticular pseudodrusen, etc.). Third, other factors influencing growth rate may change with time (incident foveal involvement, multifocal lesions becoming unifocal, increased lesion circularity, disappearance of reticular pseudodrusen, etc.). This study improves our understanding of the interpretation of FAF imaging in GA. The clarification of the role of FAF patterns will be relevant to eligibility criteria in clinical trials in GA and the potential role of specific visual cycle modulators in the therapeutic armamentarium against this disease.