TOPLINE:

Researchers developed and validated integrated polygenic risk scores (PRS) for eight cardiovascular conditions and found that individuals with high scores had increased odds of these conditions, particularly coronary artery disease (CAD) and elevated lipoprotein(a) levels; adding these scores to existing clinical tools improved risk prediction.

METHODOLOGY:

• Researchers developed and validated trait-specific integrated PRS for eight cardiovascular conditions and tested whether adding genetic risk improved the classification of clinical risk.
• They used PRSmix tool to integrate PRS from numerous published and publicly available scores for eight traits — atrial fibrillation, CAD, type 2 diabetes, elevated lipoprotein(a) levels, severe hypercholesterolemia, extreme hypertension, thoracic aortic aneurysm, and venous thromboembolism.
• The integrated scores were trained on genetic and electronic health record data from 245,394 adults in the All of Us research program (mean age, 51.7 years; 59.3% women); the lipoprotein(a) score was trained in a separate population.
• Researchers externally validated the integrated PRS in 53,306 genotyped patients from the Mass General Brigham Biobank (mean age, 53.4 years; 55.6% women).
• Performance of trait-specific integrated PRS was compared between individuals in the high-risk (91st-100th percentile) and average-risk (26th-75th percentile) categories, and net reclassification improvement was evaluated after adding PRS to existing risk tools. New cardiovascular events were tracked.

TAKEAWAY:

• The integrated PRS generally outperformed or matched individual published scores for all eight traits, achieving good discrimination between individuals at high risk and average risk.
• Individuals at high risk for CAD were 3.73-fold more likely to develop the condition than those at average risk (P < .0001). Those at high risk for elevated lipoprotein(a) levels were 40.98-fold more likely to develop the condition (P < .0001).
• Adding the integrated PRS to the Pooled Cohort Equations produced a net reclassification improvement of 0.17 among individuals at borderline and intermediate 10-year risk for atherosclerotic cardiovascular disease.
• Over a median follow-up duration of 7.63 years, individuals with higher risk per integrated PRS had more than twice the risk for incident CAD, atrial fibrillation, and type 2 diabetes (P < .0001 for all).

IN PRACTICE:

“Our approach demonstrates that a consensus score can consolidate predictive information while improving stability in individual-level risk classification, which is essential for clinical deployment,” the researchers wrote.

SOURCE:

The study was led by Anika Misra of Broad Institute of MIT and Harvard in Cambridge, Massachusetts. It was published online on April 29 in the Journal of the American College of Cardiology.

LIMITATIONS:

The study drew on biobank volunteers and clinic patients; thus, the high-risk group may not have reflected the general population. Most genetic scores were made in individuals of European descent and lost accuracy in other groups. Excluding individuals who lacked the required health measurements may have introduced selection bias.

DISCLOSURES:

The study received support from the US National Institutes of Health for the Polygenic Risk Methods in Diverse Populations Consortium, with grant funding for study site FFAIRR-PRS and the coordinating center. Additional funding was provided by grants from the National Heart, Lung, and Blood Institute and the American Heart Association. Several authors reported receiving consulting fees or honoraria from biotechnology and pharmaceutical companies. Some reported receiving research funding or holding equity interests in commercial genetics and diagnostics firms. One author reported a family member’s employment in the pharmaceutical industry.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.