Abstract 110: Genetic Analysis Implicates LDL Cholesterol Reduction and Plasminogen Activator-inhibitor 1 Antagonism as Therapeutic Interventions for Venous Thromboembolism

Introduction: Venous thromboembolism (VTE) is a significant heritable cause of cardiovascular-related mortality, yet previously published GWAS have only identified 11 genome-wide significant (P<5x10 -8 ) risk loci to date. Hypothesis: Genetic variants affecting multiple biological pathways are associated with VTE risk and may reveal novel therapeutic targets. Methods: Using EHR data, we identified individuals with and without clinical VTE in the Million Veteran Program and UK Biobank participants. Individuals were genotyped on customized Affymetrix Biobank arrays, and we tested 13 million genotyped and imputed DNA variants for association with clinical VTE using logistic regression models adjusting for age, sex and population structure. We then meta-analyzed across the two datasets and set a P<5x10 -8 for statistical significance. In downstream analyses, we examined genetic variant-plasma protein associations at P<5x10 -8 generated from SOMAscan data quantifying 3,622 plasma proteins in 3,301 healthy participants from the INTERVAL study to identify VTE risk variants that alter protein concentrations in human plasma. We then performed a Mendelian randomization (MR) analysis to evaluate the causal role of lipids (HDL cholesterol/LDL cholesterol/triglycerides) on VTE risk using a 222-variant lipid genetic risk score and set P<0.016 (0.05/3 lipid fractions) for statistical significance. Results: We identified 26,066 VTE cases and 624,053 controls. Following meta-analysis, we identified 21 novel VTE loci implicating known VTE risk factors including body mass index ( LMOD ) and hypercoaguability ( VWF, FX, PROC, PROS1 ). Four of the VTE risk variants were associated with changes in protein concentrations in plasma resulting in a pro-coagulant effect, including an increased expression of 1) Factor Xa, a current VTE therapeutic target, and 2) plasminogen activator-inhibitor 1, a possible novel therapeutic target. Through MR analysis, we provide evidence that LDL cholesterol, may be a causal risk factor for VTE (OR=1.17 per SD increase in LDL cholesterol, P=0.003). Conclusions: Here, we assembled the largest reported cohort of individuals with clinical VTE and genetic data and identify novel associations with therapeutic implications.