Daniel R Muth, Abinaya P Venkataraman, Alberto Dominguez-Vicent, Goran Petrovski, Pete A Williams, Laurence Quérat, Filippo Locri, Sandrine A Zweifel, Anders Kvanta
In this longitudinal cohort, fovea-sparing lesions showed faster total and regional growth than fovea-involving lesions, mainly caused by a faster outward expansion component. Location 500 μm or closer to the fovea resulted in slower growth, which supports a foveal self-protecting mechanism. These are relevant findings for patients with fovea-sparing lesions.
PURPOSE: Fovea-sparing geographic atrophy (GA) patients retain near-normal central vision. To better understand GA progression in relation to the fovea, we performed a semi-automated image analysis of regional atrophy growth in a long-term GA natural-history cohort.
METHODS: Prospective-observational, single-centre-study (02/2013-07/2025) at the Division of Eye and Vision, Karolinska Institutet/St. Erik Eye Hospital, Stockholm, Sweden. Longitudinal square-root-transformed GA area and distance-growth rates [mm/year] in all retinal quadrants were calculated from fundus autofluorescence images. Eyes were stratified based on best-corrected visual acuity: fovea-sparing (≥65 ETDRS) and fovea-involving (<65 ETDRS).
RESULTS: 65 eyes (65 patients) presented fovea-sparing and 91 eyes (91 patients) fovea-involving. Median follow-up was 18 months (range: 5-125). The fovea-involving group decreased in BCVA (Wilcoxon, p < 0.001). Total area growth was faster for fovea-sparing than fovea-involving lesions: 0.34 vs. 0.26 (Mann-Whitney U, p = 0.0019). The same was true for the superior, inferior, and temporal quadrants (Mann-Whitney U, all p < 0.0472). Fovea-sparing lesions showed slower distance growth towards the fovea than outwards in all quadrants: superior 0.09 vs. 0.21; nasal 0.08 vs. 0.18; inferior 0.06 vs. 0.17; temporal 0.09 vs. 0.20 (Wilcoxon, all p < 0.001). Lesions closer to the fovea (≤500 μm) at baseline showed decreased regional growth towards the fovea. Outwards growth was not related to lesion location.
CONCLUSIONS: In this longitudinal cohort, fovea-sparing lesions showed faster total and regional growth than fovea-involving lesions, mainly caused by a faster outward expansion component. Location 500 μm or closer to the fovea resulted in slower growth, which supports a foveal self-protecting mechanism. These are relevant findings for patients with fovea-sparing lesions.