The Discrepancy Between Visual Acuity Decline and Foveal Involvement in Geographic Atrophy.

Abstract

Purpose: To investigate the discrepancy between visual acuity (VA) decline and foveal involvement in geographic atrophy (GA) secondary to nonexudative age-related macular degeneration (AMD), and to explore how early retinal changes impact the progression of visual impairment.

Design: Retrospective, longitudinal cohort study.

Subjects: This study evaluated 80 eyes from 60 patients (mean age, 74.2 ± 10 years) with progressing non-neovascular AMD.

Methods: Blue-light fundus autofluorescence (FAF) and spectral-domain OCT (SD-OCT) were utilized to monitor GA progression and the onset of foveal involvement. The study analyzed VA changes over an average follow-up of 60 ± 26.4 months, encompassing 785 observations. Mixed-effects models with natural splines assessed the effects of demographic and ocular characteristics on baseline VA and its rate of decline. Survival analyses compared the timing of anatomic changes with the most rapid functional declines, indicated by the highest first derivative of VA trajectories. Discrepancies between visual and anatomic changes were explored using generalized linear mixed-effects models.

Main outcome measures: Monthly VA changes, onset and impact of foveal involvement, and factors influencing baseline VA and rate of decline.

Results: Visual acuity declined consistently by an average of 0.010 logarithm of the minimum angle of resolution (LogMAR) per month (standard error [SE], 0.0003; P < 0.001). The onset of foveal involvement significantly exacerbated this decline, adding an average loss of 0.15 LogMAR (SE, 0.02; P < 0.001). Stabilization of VA typically occurred around 41 months post-foveal involvement. Significant factors associated with worse baseline VA were older age, female gender, unifocal GA morphology, and drusen-associated forms of GA (P < 0.05). The most rapid declines in VA typically occurred about 9 months (interquartile range, 0-27 months) before detectable subfoveal changes. The reticular FAF pattern (27/46 [59%] vs. 2/13 [15%], P = 0.02) and smaller baseline GA lesions (P = 0.01) were associated with faster deterioration preceding visible foveal damage.

Conclusions: This study demonstrates that significant VA loss in GA can precede detectable foveal involvement, suggesting a window for early interventions to slow the progression of visual impairment. Identifying specific GA characteristics and FAF patterns as predictors of rapid VA decline supports the need for personalized treatment strategies to optimize outcomes for patients with nonexudative AMD.

Financial disclosure(s): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.