Abstract
Purpose :
Geographic atrophy (GA) secondary to AMD is a progressive,irreversible degenerative condition leading to visual impairment and blindness over time. While studies have made significant strides in identifying genetic predictors associated with risk of GA incidence,particularly variants within the complement pathway and in ARMS2/HTRA1,our grasp of genetic risk factors for rate of GA progression is incomplete. Acknowledging this gap, our study is the largest GWAS to date towards understanding the complex genetic underpinnings of GA progression.
Methods :
We employed an additive genetic model to analyze longitudinal data in GA patients (N=2,472), comprised of data from 3 GA-based clinical trials (CHROMA, SPECTRI and MAHALO) and 7 longitudinal cohorts or clinical trials for AMD (AREDS1/2, FAM, GATE, PROXIMA A/B, and WUE). Standard sample and variant QC were performed. Variants were restricted to MAF > 1%. We calculated the slope of change in GA area (square root) over a two-year follow-up period, and we excluded participants with CNV in the study eye, aiming to minimize variability between studies. REGENIE was used for analyzing the individual cohorts, and we integrated these results using METAL for a meta-analysis.
Results :
A total of 9,532,089 SNPs were included in the meta-analysis. None passed the genome-wide associated threshold (p<5E-8). The top hit was intronic to GSTM4 (P=5E-7). We evaluated AMD risk loci identified in the analysis by Fritsche et al. However, in our analysis of GA progression, none of these 34 loci showed any evidence of association after adjusting for multiple testing. Similarly, a polygenic score for AMD risk showed no evidence of association. We also conducted the analysis using other metrics for quantifying lesion growth, such as calculating percentage growth, however, this approach did not yield statistically significant results.
Conclusions :
Our study failed to identify new genetic loci associated with GA lesion growth, nor did it observe previously known AMD risk factors to have a meaningful impact on GA progression. This may seem surprising, given the substantial genetic component of risk for GA incidence. It suggests: substantial genetic heterogeneity underlies GA lesion growth, genetic factors do not significantly influence the progression of this disease or that our study lacked the statistical power necessary to detect such associations.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.