Abstract
AIMS: Recent aspirin primary prevention trials failed to identify a net benefit of aspirin for preventing cardiovascular disease vs. the harms of bleeding. This study aimed to investigate whether a high-risk subgroup, individuals with elevated genetic predisposition to coronary artery disease (CAD), might derive more benefit than harm with aspirin, compared to those with lower genetic risk. METHODS AND RESULTS: We performed genetic risk stratification of the Aspirin in Reducing Events in the Elderly (ASPREE) randomized controlled trial using a CAD polygenic risk score (GPSMult). For 12 031 genotyped participants (5974 aspirin, 6057 placebo) overall, we stratified them by GPSMult quintiles (q1-5), then examined risk of CAD (composite of myocardial infarction and coronary heart disease death) and bleeding events using Cox models. During a median 4.6 years of follow-up with randomization to 100 mg/day aspirin vs. placebo, 234 (1.9%) participants had CAD and 373 (3.1%) had bleeding events. In the overall cohort, aspirin resulted in higher bleeding risk [adjusted Hazard ratio (aHR) = 1.30 (1.06-1.61), P = 0.01] but no significant CAD reduction [aHR = 0.84 (0.64-1.09), P = 0.19]. However, among the highest quintile of polygenic risk (q5, top 20% of the GPSMult distribution), there was a 47% reduction in risk of CAD events with aspirin [aHR = 0.53 (0.31-0.90), P = 0.02] without increased bleeding risk [aHR = 1.05 (0.60-1.82), P = 0.88]. Interaction between the GPSMult and aspirin was significant for CAD (q5 vs. q1, P = 0.02) but not bleeding (P = 0.80). CONCLUSION: The balance between net benefit and harm on aspirin in the primary prevention setting shifts favourably in individuals with an elevated genetic predisposition.