Integration of a polygenic score into guideline-recommended prediction of cardiovascular disease

Li, Ling; Pang, Shichao; Starnecker, Fabian; Mueller-Myhsok, Bertram und Schunkert, Heribert (2024): Integration of a polygenic score into guideline-recommended prediction of cardiovascular disease. In: European Heart Journal, Bd. 45, Nr. 20: S. 1843-1852

Abstract

Background and Aims It is not clear how a polygenic risk score (PRS) can be best combined with guideline-recommended tools for cardiovascular disease (CVD) risk prediction, e.g. SCORE2.Methods A PRS for coronary artery disease (CAD) was calculated in participants of UK Biobank (n = 432 981). Within each tenth of the PRS distribution, the odds ratios (ORs)-referred to as PRS-factor-for CVD (i.e. CAD or stroke) were compared between the entire population and subgroups representing the spectrum of clinical risk. Replication was performed in the combined Framingham/Atherosclerosis Risk in Communities (ARIC) populations (n = 10 757). The clinical suitability of a multiplicative model 'SCORE2 x PRS-factor' was tested by risk reclassification.Results In subgroups with highly different clinical risks, CVD ORs were stable within each PRS tenth. SCORE2 and PRS showed no significant interactive effects on CVD risk, which qualified them as multiplicative factors: SCORE2 x PRS-factor = total risk. In UK Biobank, the multiplicative model moved 9.55% of the intermediate (n = 145 337) to high-risk group increasing the individuals in this category by 56.6%. Incident CVD occurred in 8.08% of individuals reclassified by the PRS-factor from intermediate to high risk, which was about two-fold of those remained at intermediate risk (4.08%). Likewise, the PRS-factor shifted 8.29% of individuals from moderate to high risk in Framingham/ARIC.Conclusions This study demonstrates that absolute CVD risk, determined by a clinical risk score, and relative genetic risk, determined by a PRS, provide independent information. The two components may form a simple multiplicative model improving precision of guideline-recommended tools in predicting incident CVD. Structured Graphical Abstract The left panel shows that relative risk related to polygenic disposition [1-10 tenth of the polygenic risk score (PRS) distribution] is independent of having a low, intermediate, or high clinical risk by SCORE2 in the UK Biobank. Thereby, the odds ratio from the PRS (relative risk) can be used as a genetic factor (PRS-factor) to multiply the absolute clinical risk estimate from SCORE2. If this is being done, similar to 10% of the intermediate-risk group is being upgraded to a higher total risk (right panel). Subgroups that were genetically upgraded had incident event rates that were similar to those in the original high-risk groups and substantially increased the overall numbers of individuals correctly assessed to be at high risk. CVD, cardiovascular disease.

• Dokumententyp: Zeitschriftenartikel
• Fakultät: Medizin > Munich Cluster for Systems Neurology (SyNergy)
• Themengebiete: 600 Technik, Medizin, angewandte Wissenschaften > 610 Medizin und Gesundheit
• Sprache: Englisch
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 390857198