Multitrait analysis of genome-wide association studies expands eosinophilic esophagitis genetic susceptibility and polygenic risk scores.

Abstract

Eosinophilic esophagitis (EoE) is an atopic disease driven in part by genetic susceptibility, but single-trait genome-wide association study (GWAS) has identified a limited number of genome-wide significant risk loci. We sought to expand discovery of EoE genetic risk loci by leveraging shared genetic architecture with other atopic diseases and to develop a polygenic risk score (PRS) for EoE. We performed a GWAS of 1,757 individuals with EoE and 14,467 population controls. We then applied multitrait analysis of GWAS (MTAG), integrating EoE with other atopic disease GWAS (UK Biobank; >450,000 subjects). Functional analyses were used to nominate candidate EoE risk genes. PRS models derived from MTAG were compared to PRS derived from the EoE-only GWAS. An interactive tool (EGIDExpress; https://egidexpress. cchmc.org/GWAS/) was developed to enable dataset queries and visualization. The EoE-only GWAS identified 11 independent risk variants across 8 loci (P < 5 × 10^-8), including 3 novel loci. MTAG identified 33 independent EoE risk variants across 24 loci, including 14 novel loci. Functional studies nominated 90 candidate EoE risk genes, including genes implicating mechanisms beyond type 2 immunity. A PRS derived from MTAG outperformed a PRS derived from the EoE-only GWAS (OR 11.57 [95% confidence interval, 6.90-19.40] for top vs bottom decile). Leveraging shared atopic disease genetics via MTAG substantially expands the landscape of EoE risk loci and improves EoE polygenic risk prediction, underscoring shared genetic mechanisms across atopic diseases. We further provide a public resource (EGIDExpress; https://egidexpress. cchmc.org/GWAS/) to advance the field.