Polygenic Risk Score Assessment for Coronary Artery Disease in Asian Indians

Area of Science:

• Genetics
• Cardiovascular Disease Research
• Population Health

Background:

• Polygenic risk scores (PRS) are crucial for predicting coronary artery disease (CAD) risk.
• Current PRS models are predominantly derived from European ancestry populations, limiting their applicability in diverse populations.
• The transportability and predictive performance of PRS across different ancestries require thorough evaluation.

Purpose of the Study:

• To evaluate the performance and transportability of various polygenic risk score (PRS) models for coronary artery disease (CAD) in an Asian Indian (AI) cohort.
• To compare the predictive accuracy of PRS models derived from European (EU), South Asian (SA), and AI ancestries.
• To assess the impact of including clinical risk scores and population-specific genetic diversity on PRS performance.

Main Methods:

• Performance evaluation of PRS models using training and test sets from 13,974 subjects of AI ancestry.
• Comparison of predictive performance metrics (e.g., efficiency in extreme quartiles) between different PRS models (EU, SA, AI, EU+AI).
• Assessment of the incremental value of clinical risk scores and population-specific variants in PRS models.

Main Results:

• PRS models derived from AI and SA ancestries, as well as a combined EU+AI model, demonstrated superior predictive performance and transportability compared to the EU-derived model.
• The AI and EU+AI models showed 18% and 22% higher predictive performance, respectively, than the EU model.
• AI and EU+AI models were 2.6 to 4.6 times more efficient in identifying individuals at high CAD risk compared to the EU model. Clinical risk scores did not significantly alter genetic model performance.

Conclusions:

• Polygenic risk score models derived from or including Asian Indian (AI) and South Asian (SA) genetic data significantly improve coronary artery disease (CAD) risk prediction and transportability in AI populations.
• Incorporating population-specific genetic diversity and risk factors into PRS models is essential for refining risk stratification and enhancing clinical utility.
• Developing ancestry-specific or ancestry-informed PRS models is critical for equitable and effective cardiovascular disease risk assessment across diverse populations.