Utility of a coronary artery disease polygenic risk score in a low-risk primary prevention cohort.

AIMS: To evaluate whether a coronary artery disease (CAD) polygenic risk score (PRS) identifies clinically low-risk individuals at elevated atherosclerotic cardiovascular disease (ASCVD) risk.

METHODS

We conducted a longitudinal cohort study within the UK Biobank including adults aged 40-74 years, free of ASCVD and diabetes, not on lipid-lowering therapy, and with baseline LDL <190 mg/dL. Participants were enrolled from 2006 to 2010 and followed through 2021. Ten-year ASCVD risk was estimated using pooled cohort equations. A CAD PRS derived from 2.3 million genome-wide variants was categorized as top 1%, 5%, 20%, 50%, and bottom 50%. Lipoprotein(a) and high-sensitivity C-reactive protein were similarly stratified. The primary endpoint was a composite of myocardial infarction, ischemic stroke, or cardiovascular death. Cox proportional hazards models estimated adjusted hazard ratios (HRs) for 10-year ASCVD events, and net reclassification improvement (NRI) was assessed at 5% and 7.5% risk thresholds.

RESULTS

Among 266,917 low-risk individuals (median age 54; 38% male), ASCVD risk increased substantially across higher PRS strata. Compared with the bottom 50%, individuals in the top 50% had more than double the risk (aHR 2.48), the top 20% had over triple the risk (aHR 3.71), and the top 5% and 1% had 6-fold (aHR 6.31) and 9-fold (aHR 9.73) higher risks, respectively (all p < 0.0001). Adding PRS improved model discrimination (AUC 0.76 vs 0.72) and reclassified 12% of low-risk individuals upward-over tenfold more than Lp(a) or hs-CRP.

CONCLUSION

In primary prevention, a CAD PRS identifies individuals with markedly higher ASCVD risk despite low clinical risk estimates and may merit consideration among ASCVD risk enhancers.