Genome-wide association studies of geographic atrophy: a systematic review of susceptibility, progression, and pharmacogenomic evidence

Abstract

Abstract Background Age-related macular degeneration (AMD) is the leading cause of irreversible central vision loss in older adults. Geographic atrophy (GA), its advanced atrophic form, now has two approved complement inhibitors as disease-modifying therapy, yet whether genetic variation predicts treatment response, and whether current genetic evidence applies beyond European populations, remains unclear. We systematically synthesized genome-wide association study (GWAS) evidence on genetic determinants of GA to address these two gaps and their implications for complement-targeted therapy. Methods A systematic search of PubMed, Scopus, and the NHGRI-EBI GWAS Catalog was conducted from database inception through July 2026 following PRISMA guidelines. Eligible studies included human GWAS reporting GA-related genetic associations. Studies were classified as providing direct (GA-specific) or indirect evidence. Two reviewers independently performed study selection, data extraction, and methodological quality assessment using the Q-Genie instrument, and rated confidence in each finding qualitatively by directness of evidence, statistical significance, and replication (GRADE was not applicable to unpooled genome-wide evidence). Owing to heterogeneity in study design, phenotypes, and reported effect measures, findings were synthesized narratively following the Synthesis Without Meta-analysis (SWiM) guideline. Results Five GWAS, all conducted in predominantly European-ancestry populations, met the inclusion criteria. Two studies provided direct GA-specific evidence, and three provided indirect evidence (late AMD progression, a methodological pharmacogenomic study, and drusen traits). Complement pathway genes ( CFH , C3 , C2 , CFB , CFI ) showed the strongest associations with GA susceptibility in the single GA-specific case–control GWAS (direct evidence), with concordant but indirect (late-AMD) evidence for progression; the CFH rs1061170 (Y402H ) variant had the largest reported effect (odds ratio, 2.78), while the ARMS2/HTRA1 locus was the principal non-complement susceptibility locus. In the one study that compared GA with neovascular AMD, complement variants did not discriminate between subtypes, whereas ARMS2/HTRA1 favored neovascular AMD. Longitudinal analysis of progression to late AMD (indirect evidence) identified seven loci (four established and three novel, marginally significant, model-dependent loci not supported by case-control data), and the only GA lesion-growth GWAS (direct evidence) identified PRMT6 and LSS without replication and without a complement signal. All studies were rated good methodological quality (Q-Genie), but confidence in individual findings was qualified by directness, significance, and replication: it was higher only for the complement and ARMS2/HTRA1 susceptibility loci, and no GA-specific locus was replicated in an independent GA cohort within the included studies. Only one study tested a genotype-treatment interaction, and this was a methodological study of the AREDS antioxidant/zinc formulation, not of a complement inhibitor and not GA-specific; none evaluated genotype-stratified response to pegcetacoplan or avacincaptad pegol, the two complement inhibitors currently approved for GA. No GA-specific GWAS were identified in Asian, African, or Hispanic populations. Conclusions Despite two decades of consistent evidence implicating complement dysregulation in AMD, with GA-specific GWAS evidence limited to two studies, no validated pharmacogenomic biomarkers are currently established to guide genotype-based selection between approved complement inhibitors, and no GA-specific genetic evidence exists outside European populations. The absence of validated biomarkers should not be interpreted as definitive evidence that genetic variation cannot influence treatment response. These evidence gaps, rather than the established genetic architecture itself, define the current limits of precision medicine in GA. Future studies should prioritize pharmacogenomic sub-analyses of existing and future complement-inhibitor trials, alongside multi-ancestry GWAS. GA is a complex disease shaped by genetic and non-genetic factors, so genotype should not be used as the sole basis for treatment decisions. Clinical trial number Not applicable.

The authors' abstract, as published at the source. International Journal of Retina and Vitreous, 2026 · DOI ↗