Mass General Brigham Develops Two Polygenic Risk Score Tools for Cardiovascular and Metabolic Conditions

Two new polygenic risk score tools aim to transform heart and metabolic disease prediction.

Researchers at Mass General Brigham Heart and Vascular Institute and collaborating institutions have developed and validated two distinct polygenic risk score (PRS) tools. One assesses inherited risk for eight cardiovascular conditions, and the other predicts risk for obesity and type 2 diabetes. Both sets of findings were published in peer-reviewed journals.

Cardiovascular Polygenic Risk Score

The cardiovascular PRS tool estimates genetic risk for eight conditions: coronary artery disease, atrial fibrillation, type 2 diabetes, venous thromboembolism, thoracic aortic aneurysm, extreme hypertension, severe hypercholesterolemia, and elevated lipoprotein(a).

The tool combines multiple genetic risk models into a single integrated score using DNA from one genotyping test. It was developed using PRSmix, an elastic-net approach that combined previously published polygenic risk scores from the Polygenic Score Catalog.

The tool was trained on genotype and clinical data from 245,394 participants in the National Institutes of Health's All of Us Research Program. Validation was performed on 53,306 individuals from the Mass General Brigham Biobank.

In the validation cohort, individuals in the top 10% of genetic risk showed significantly increased odds of disease compared to average-risk individuals:

• Coronary artery disease: odds ratio (OR) 3.7
• Type 2 diabetes: OR 3.1
• Atrial fibrillation: OR 3.0
• Severe hypercholesterolemia: OR 4.1
• Extreme hypertension: OR 2.1
• Venous thromboembolism: OR 1.9
• Elevated lipoprotein(a): OR 41.0

The integrated PRS improved risk classification when incorporated into clinical prediction models, particularly for individuals near clinical decision thresholds.

• Predictive strength was reduced in individuals with greater genetic variation from European ancestry, reflecting limitations in current genomic reference datasets.
• Many risk scores used in the tool were derived from populations of primarily European ancestry.

The cardiovascular PRS is currently available through Mass General Brigham Laboratory for Molecular Medicine and Broad Clinical Labs. The report provides risk levels (high, average, or low) for each condition with explanations and graphical comparisons to the general population. It integrates into electronic health records and patient portals.

Researchers plan further prospective validation across diverse populations, cost-effectiveness evaluation, and research on clinical decision-making influenced by genetic risk information. The study was published in the Journal of the American College of Cardiology and supported by NIH grants, including the Polygenic Risk Methods in Diverse Populations (PRIMED) Consortium.

Metabolic Polygenic Risk Scores

A separate set of advanced PRSs was developed for predicting obesity and type 2 diabetes.

These PRSs integrate genetic findings from multiple large biobanks, encompassing over 8.5 million participants globally. The scores use genome-wide association studies (GWAS) data with a focus on non-European populations.

They assess risk beyond body mass index (BMI) by targeting genes associated with 20 different metabolic traits, including fat distribution, insulin control, and glucose regulation.

• The risk scores identified individuals at high risk for clinical outcomes such as cardiovascular disease and stroke.
• Individuals with high PRS who were initially healthy were approximately twice as likely to receive GLP-1 agonist medications or bariatric surgery over a median follow-up of 5.5 years, compared to those with mid-range PRS scores.
• By utilizing multi-ancestry GWAS data, the obesity and type 2 diabetes risk scores outperformed previous PRS models in African, East Asian, and South Asian individuals.

Researchers stated plans to continue refining the understanding of genetic subtypes of type 2 diabetes and obesity. This work is intended to improve patient classification and stratification for clinical trials.

The stated goal is to facilitate early identification of individuals with higher genetic susceptibility to poor metabolic health, moving beyond BMI as the primary indicator.