Abstract 17614: Genome-Wide Association Study of Aspirin Responsiveness in African Americans

While familial studies have established that inter-individual variability in platelet response to aspirin (ASA) is partly due to genetic variation, a genome-wide association study (GWAS) has not been reported in African Americans. Fasting blood samples were obtained from healthy African Americans (N=868) with family history of premature coronary artery disease. Optical aggregation was measured in platelet-rich plasma (PRP) after samples were stimulated with collagen (1, 2, 5 µg/mL), ADP (2, 10 µmol/L), or epinephrine (2, 10 µmol/L) before and after 14-days of ASA therapy. A cohort of European ancestry (N=1236) provided replication. In each cohort, age- and sex-adjusted linear mixed effects models adjusting for family structure were used to test for association between each SNP and each phenotype under an additive model. We analyzed data with adjustment for pre-ASA platelet studies to examine the pure effect of ASA and without adjustment which is likely to be closer to the clinically observed effect. We sought region-based replication with adjustment for the number of LD blocks in HapMap CEU population. We observed SNPs in 3 regions (2 intronic and 1 intergenic) that crossed GWAS threshold of significance (p<5 x 10 -8 ). A SNP in the PEAR1 gene, rs12041331, was significantly associated with most doses of the three agonists and this association remained significant after adjusting for pre-ASA platelet studies (Table). In the replication cohort, rs12041331 was also nominally significant. SNPs close to the other two GWAS-significant SNPs were nominally significant in the replication cohort after adjusting for the LD structure. In this first GWAS of platelet response to ASA therapy in African Americans, we show a relationship between post-ASA platelet reactivity and 3 regions, and replicate these findings in an independent sample. Both PEAR1 and CRB1 are expressed in platelets and are good candidates for follow-up studies examining ASA responsiveness.