Perilesional FAF Patterns Evolve Over Time in Geographic Atrophy

With recent FDA approvals of complement inhibitors targeting geographic atrophy (GA), there is renewed interest in identifying imaging biomarkers that predict GA enlargement rates. In a recent study appearing in Ophthalmology Science, researchers sought to examine how perilesional fundus autofluorescence (FAF) patterns evolve over time in patients with foveal-sparing GA and whether these changes correlate with GA enlargement rate, aiming to address limitations in existing FAF classification systems. Its findings challenge the previously held assumption that FAF patterns are static, instead highlighting the dynamic nature of FAF morphology in GA.Included in the retrospective, longitudinal study were 143 eyes from 99 patients who underwent BCVA assessments, along with blue-light FAF imaging and spectral-domain OCT across multiple visits. The median follow-up period was approximately 3.4 years, during which 23 out of 106 eyes (~22%) exhibited transitions in their perilesional FAF patterns after a median duration of three years (range: 1.74 to 4.10 years). The primary outcomes sought to establish the frequency and timing of these FAF pattern transitions and their association with the growth rate of GA lesions. In patients with foveal-sparing GA, longitudinal analysis of FAF images showed that diffuse-trickling and banded patterns were associated with more extensive atrophy, consistent with prior studies. Eyes with a banded pattern were also found to have worse baseline BCVA, although FAF phenotype showed no correlation with rate of visual decline.This graphic from the paper presents the authors’ proposed scheme of perilesional FAF pattern evolution in GA. Arrows indicate the directions of transitions that were documented in this cohort. Most eyes evolved from none or focal to diffuse non-trickling patterns and from diffuse non-trickling to banded configurations. In contrast, banded patterns remained stable throughout follow-up, likely representing the natural endpoint of FAF evolution. Consistent with this sequence, none and focal patterns were typically observed in eyes with smaller GA areas, diffuse non-trickling patterns in eyes with small-to-intermediate lesions and banded patterns in eyes with the largest atrophic areas. The researchers propose that FAF pattern transitions occur as macular atrophy enlarges and reaches the typical GA size range associated with the next perilesional FAF configuration. Photo: Antropoli A, et al. Ophthalmol Sci. November 2025. Click image to enlarge. The results indicated that the mean square root–transformed GA growth rate was 0.40mm/year, revealing consistent progression across the cohort, which aligns with findings from major clinical trials investigating complement inhibitors. Among the baseline FAF phenotypes, the diffuse non-trickling pattern exhibited a modestly significant association with an increased progression rate (+0.06mm/year). Interestingly, eyes that experienced any transitions in FAF patterns during the follow-up had a slower overall GA enlargement rate of -0.12mm/year, particularly in transitions from a "none" pattern to diffuse or banded patterns.“These observations support a staged model of perilesional FAF evolution, in which the perilesional hyper-FAF area changes over time, and eyes progress from no detectable junctional hyperautofluorescence to focal patterns, followed by diffuse non-trickling, and finally to the banded configuration,” the researchers explained in their paper. Since FAF patterns were found to typically evolve at intermediate lesion sizes and more advanced disease stages, the researchers emphasize “the importance of temporal context when interpreting perilesional FAF morphology.”The exploratory OCT analysis showed that structural changes in the perilesional zone often occurred before FAF pattern transitions, though not consistently. Ellipsoid zone disruption was the most common finding at baseline, while external limiting membrane disruption and retinal pigment epithelium (RPE) thickening varied by transition type. RPE thickening was particularly notable in eyes transitioning to the diffuse non-trickling pattern, suggesting structural remodeling. However, some eyes with measurable GA margin displacement showed no abnormalities in newly exposed perilesional areas, highlighting a disconnect between visible damage and future atrophic expansion. These findings, the researchers noted in their paper, “suggest that OCT-based markers may capture some aspects of local vulnerability but may not detect earlier molecular or subclinical changes that precede morphological degeneration.”While this study confirmed that baseline FAF phenotype does not have predictive power for functional visual acuity decline, the authors argue that it may be beneficial to consider the dynamic nature of FAF patterns in the design of future clinical trials for GA therapies. “Incorporating longitudinal FAF imaging during study follow-up may offer valuable insight into whether treatment modifies not only lesion enlargement but also the tempo or direction of phenotypic evolution,” the team concluded. They added, “Post hoc subgroup analyses comparing treatment effects across evolving vs. stable patterns, and across banded vs. diffuse configurations, may further enhance our understanding of response heterogeneity.”Click here for the journal source. Antropoli A, Vacirca F, Introini U, et al. Perilesional FAF patterns are not static: longitudinal transitions in geographic atrophy and association with disease progression. Ophthalmol Sci. November 2025. [Epub ahead of print]. This article was developed by the editorial staff in conjunction with experts in the field. In the process, AI may have been among the editorial tools used to meet the goals of human editors, who approved all content.