Abstract 125: Trans-ancestry Genome-wide Association Study Of Varicose Veins In >1 Million Individuals Reveals Circulating Effectors Of Venous Disease

Abstract

Background: Varicose veins (VV) represent a common cause of cardiovascular morbidity and healthcare expenditures, with limited available medical therapies. The molecular and genetic basis of VV remains uncertain. Methods: We identified individuals of diverse ancestry with and without diagnosis of VV in the VA Million Veteran Program (MVP). Genome-wide association studies (GWAS) were performed using logistic regression adjusted for age, sex, and population structure. Trans-ancestry meta-analysis was then performed with 6 additional VV GWAS among participants of UK Biobank (UKB), FinnGen, eMERGE, and Biobank Japan. Downstream fine-mapping, phenome-wide association (pheWAS), colocalization, and Mendelian randomization (MR) analyses were performed to identify putative causal genetic variants, clinical traits, and circulating proteins that share associations with VV. Results: GWAS meta-analysis included 49,765 VV cases and 1,334,301 controls, identifying 151 independent genetic loci (p < 5 x 10 -8 ) associated with VV. Fine-mapping revealed 11 loci where a single causal variant was prioritized with high confidence (posterior probability > 0.7). PheWAS of lead variants revealed associations at 101 unique loci with 90 circulating proteins and 992 clinical traits/measurements assessed in UKB. Colocalization identified shared genetic signals (posterior probability > 0.7) between VV and clusters of proteins (eg. a cluster of vascular proteins at the ABO locus including vascular proteins like VEFGR2/3, and ICAM1/4/5, among others), and clusters of clinical traits (eg. a cluster of anthropometric associations at the DLC1 locus). To separate shared genetics from causal associations we performed proteome-wide MR across 738 circulating proteins. We found significant causal associations (p < 0.05/738) between 15 proteins and VV, including vasoactive proteins like ADM; extracellular matrix and connective proteins like MFAP2, POSTN, ASPN, ECM1; and hormone-proteins like SHBG. Conclusions: We assembled the largest trans-ancestry GWAS of diagnosed VV to date. These results highlight shared associations between VV and anthropometric traits, and identify causal circulating proteins that may represent biomarkers or therapeutic targets.