Population Heterogeneity and Selection of Coronary Artery Disease Polygenic Scores

Area of Science:

• Genetics
• Cardiology
• Population Health

Background:

• Identifying individuals at high risk for coronary artery disease (CAD) is essential for timely clinical intervention.
• Numerous polygenic scores (PGSs) exist to estimate genetic risk for CAD.
• Existing PGSs often lack representation from diverse populations, necessitating evaluation in specific ethnic groups like the Italian population.

Purpose of the Study:

• To evaluate the performance of various polygenic scores (PGSs) for coronary artery disease (CAD) risk prediction in the Italian population.
• To identify the most accurate PGS for CAD in Italians.
• To assess the impact of incorporating conventional CAD risk factors on PGS performance.

Main Methods:

• Utilized two independent Italian cohorts: EPICOR (576 individuals) and ATVB (3359 individuals).
• Evaluated 266 cardiovascular disease risk PGSs from the PGS Catalog, narrowing down to 51 for CAD.
• Assessed PGS performance using area under the curve (AUC) and analyzed the effect of adding traditional risk factors.

Main Results:

• Significant differences in distributions between patients and controls were observed for 49 out of 51 evaluated PGSs (p < 0.01).
• Only five PGSs were specifically trained and tested on European populations.
• PGS003727 showed the highest accuracy in independent evaluations (EPICOR AUC = 0.68; ATVB AUC = 0.80).
• Integration of conventional CAD risk factors improved model performance, especially in the ATVB cohort (p = 0.0003).

Conclusions:

• European-derived CAD PGSs may yield varying risk estimates in distinct populations, such as the Italian population.
• Geographical and ethnic variations necessitate careful consideration when applying PGSs.
• Further validation studies are recommended to ensure the clinical applicability of CAD PGSs across diverse populations.