Clinical utility and implementation of polygenic risk scores for predicting cardiovascular disease: A clinical consensus statement of the ESC Council on Cardiovascular Genomics, the ESC Cardiovascular Risk Collaboration, and the European Association of Preventive Cardiology

Abstract

Genome-wide association studies have revealed hundreds of genetic variants associated with cardiovascular diseases (CVD). Polygenic risk scores (PRS) can capture this information in a single metric and hold promise for use in CVD risk prediction. Importantly, PRS information can reflect the causally mediated risk to which the individual is exposed throughout life. Although European Society of Cardiology guidelines do not currently advocate their use in routine clinical practice, PRS are commercially available and increasingly sought by clinicians, health systems, and members of the public to inform personalized health care decision-making. This clinical consensus statement provides an overview of the scientific basis of PRS and evidence to date on their role in CVD risk prediction for the purposes of disease prevention. It provides the reader with a summary of the opportunities and challenges for implementation and identifies current gaps in supporting evidence. The document also lays out a potential roadmap by which the scientific and clinical community can navigate any future transition of PRS into routine clinical care. Finally, clinical scenarios are presented where information from PRS may hold most value and discuss organizational frameworks to enable responsible use of PRS testing while more evidence is being generated by clinical studies.

Keywords: Genetics; Polygenic risk score; Risk prediction.

Graphical Abstract

Considerations for polygenic risk score (PRS) testing in the context of guideline-based assessment of cardiovascular disease (CVD) risk.

Figure 1

(A) The cumulative risk to suffer from coronary artery disease in men from the UK Biobank depending on their quintile of a polygenic risk score. The green line indicates the first quintle and the red line indicates the fifth quintile. The marked difference between the first and fifth quintile of the distribution indicates that polygenic variants determine a large proportion of coronary artery disease risk. (B) The risk to suffer from coronary artery disease in women and men from the UK Biobank, however, here people are grouped depending on the number of genome-wide significant risk alleles they carry. The three middle quintiles are shaded in grey, i.e. the 0%–20% group is left from the grey box, and the 80%–100% group is on its right side. The figure illustrates that the majority of the population in the grey box is separated by only a small number of risk alleles—with only moderate separation in their lifetime prevalence of coronary artery disease explaining why c-statistics or similar measures indicate—on average—only moderate improvements. GRS, genomic risk score; HR, hazard ratio

Figure 2

Data from the UK Biobank indicating that the relative difference of coronary artery disease between the first and tenth decile of a coronary artery disease polygenic risk score is always about three-fold, irrespectively of other risk factors. As the absolute risk increases with the number of (modifiable) conventional risk factors, the absolute difference in coronary artery disease prevalence between the first and tenth decile of a coronary artery disease polygenic risk score grows

Figure 3

Absolute reduction of cardiovascular disease events achieved with lipid-lowering medications as observed in various randomized trials. While the risk of cardiovascular disease events was higher in subjects with a high polygenic risk score, the absolute risk reduction achieved with lipid-lowering medications was about three times larger in the group with a high polygenic risk score as compared

Figure 4

Challenges for implementation of polygenic risk score-based counselling for cardiovascular disease risk