Polygenic Risk, Rare Variants, and Family History: Independent and Additive Effects on Coronary Heart Disease

Background: Genetic factors are not included in prediction models for coronary heart disease (CHD). Objectives: The authors assessed the predictive utility of a polygenic risk score (PRS) for CHD (defined as myocardial infarction, coronary revascularization, or cardiovascular death) and whether the risks due to monogenic familial hypercholesterolemia (FH) and family history (FamHx) are independent of and additive to the PRS. Methods: In UK-biobank participants, PRS_CHD was calculated using metaGRS, and 10-year risk for incident CHD was estimated using the pooled cohort equations (PCE). The area under the curve (AUC) of the receiver operator curve and net reclassification improvement (NRI) were assessed. FH was defined as the presence of a pathogenic or likely pathogenic variant in LDLR, APOB, or PCSK9. FamHx was defined as a diagnosis of CHD in first-degree relatives. Independent and additive effects of PRS_CHD, FH, and FamHx were evaluated in stratified analyses. Results: In 323,373 participants with genotype data, the addition of PRS_CHD to PCE increased the AUC from 0.759 (95% CI: 0.755-0.763) to 0.773 (95% CI: 0.769-0.777). The AUC and NRI_Event for PRS_CHD were higher before the age of 55 years. Of 199,997 participants with exome sequence data, 10,000 had a PRS_CHD ≥95th percentile (PRS_P95), 673 had FH, and 46,163 had FamHx. The CHD risk associated with PRS_P95 was independent of FH and FamHx. The risks associated with combinations of PRS_CHD, FH, and FamHx were additive and comprehensive estimates could be obtained by multiplying the risk from each genetic factor. Conclusions: Incorporating PRS_CHD into the PCE improves risk prediction for CHD, especially at younger ages. The associations of PRS_CHD, FH, and FamHx with CHD were independent and additive.