Combining European and U.S. risk prediction models with polygenic risk scores to refine cardiovascular prevention: the CoLaus|PsyCoLaus Study

• Published in: European journal of preventive cardiology Vol. 30; no. 7; pp. 561 - 571
• Main Authors: de La Harpe, Roxane, Thorball, Christian W, Redin, Claire, Fournier, Stephane, Müller, Olivier, Strambo, Davide, Michel, Patrik, Vollenweider, Peter, Marques-Vidal, Pedro, Fellay, Jacques, Vaucher, Julien
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 09.05.2023
• Subjects: Atherosclerosis; Cardiovascular Diseases - diagnosis; Cardiovascular Diseases - epidemiology; Cardiovascular Diseases - genetics; Coronary Artery Disease; Female; Humans; Male; Middle Aged; Prospective Studies; Risk Assessment - methods; Risk Factors; Polygenic risk score; Genetic predisposition to disease; Primary prevention; Risk assessment; Sensitivity and specificity; Adult; Cardiovascular disease; Risk; ROC curve; Predictive value of tests
• ISSN: 2047-4873; 2047-4881
• DOI: 10.1093/eurjpc/zwad012

Abstract Aims A polygenic risk score (PRS) has the potential to improve individual atherosclerotic cardiovascular disease (ASCVD) risk assessment. To determine whether a PRS combined with two clinical risk scores, the Systematic COronary Risk Evaluation 2 (SCORE2) and the Pooled Cohort Equation (PCE) improves the prediction of ASCVD. Methods and results Using a population-based European prospective cohort, with 6733 participants at the baseline (2003–2006), the PRS presenting the best predictive accuracy was combined with SCORE2 and PCE to assess their joint performances for predicting ASCVD Discrimination, calibration, Cox proportional hazard regression, and net reclassification index were assessed. : 4218 subjects (53% women; median age, 53.4 years), with 363 prevalent and incident ASCVD, were used to compare four PRSs. The metaGRS_CAD PRS presented the best predictive capacity (AUROC = 0.77) and was used in the following analyses. 3383 subjects (median follow-up of 14.4 years), with 190 first-incident ASCVD, were employed to test ASCVD risk prediction. The changes in C statistic between SCORE2 and PCE models and those combining metaGRS_CAD with SCORE2 and PCE were 0.008 (95% CI, −0.00008–0.02, P = 0.05) and 0.007 (95% CI, 0.005–0.01, P = 0.03), respectively. Reclassification was improved for people at clinically determined intermediate-risk for both clinical scores [NRI of 9.6% (95% CI, 0.3–18.8) and 12.0% (95% CI, 1.5–22.6) for SCORE2 and PCE, respectively]. Conclusion Combining a PRS with clinical risk scores significantly improved the reclassification of risk for incident ASCVD for subjects in the clinically determined intermediate-risk category. Introducing PRSs in clinical practice may refine cardiovascular prevention for subgroups of patients in whom prevention strategies are uncertain. Lay Summary The aim of this study is to determine whether using polygenic risk scores improves the prediction of atherosclerotic cardiovascular disease risk when combined with clinical scores currently recommended by European and US guidelines on cardiovascular prevention. Graphical Abstract Graphical Abstract Polygenic risk scores, summing the weak to moderate contribution of >1mio of genetic variants derived from genome-wide association studies, are used to predict the genetic predisposition of developing ASCVD. Clinically determined intermediate-risk categories were defined according to each guideline (i.e. European Society of Cardiology for SCORE2 and American College of Cardiology/American Heart Association for PCE) and corresponded to the category where treatment should be considered but not recommended. In the figure on the left, the reclassification of people without ASCVD after integrating the PRS into equations is not shown. ASCVD; atherosclerotic cardiovascular disease; metaGRS_CAD; polygenic risk score from Inouye et al. (in Journal of the American College of Cardiology, 2018, doi: 10.1016/j.jacc.2018.07.079), PRS; polygenic risk score