Coronary artery calcium score, not polygenic risk score, best predicts heart disease risk

Key takeaways:

• In a head-to-head comparison, coronary artery calcium score improved CV risk discrimination and a polygenic risk score did not.
• There was no added predictive utility when combining the two scores.

Data show CV risk discrimination improved when adding a coronary artery calcium score to a CHD prediction model based on traditional risk factors; however, there were no changes when adding a polygenic risk score.

The analysis of two large cohorts of middle-age and older adults with long-term follow-up — all with available genetic data and coronary CT scans — addresses a key knowledge gap identified in a recent American Heart Association scientific statement expressing a need to directly compare polygenic risk scores with CAC scores for predicting CHD risk, according to Sadiya Sana Khan, MD, MSc, FACC, FAHA, assistant professor of medicine and preventive medicine, associate program director of the cardiovascular disease fellowship and director of research in the section of heart failure at Northwestern University Feinberg School of Medicine.

“There has been a lot of interest in each of the scores individually for improving risk prediction, but what isn’t clear is whether one would be better than the other or whether the combination of them might be better together,” Khan told Healio. “It was very helpful to see this comparison because CAC added an improved risk prediction and helped reclassify people, but the polygenic score did not. The combination did not do better than CAC alone.”

Changes in risk discrimination

Sadiya Sana Khan

Khan and colleagues analyzed data from two observational population-based studies involving adults aged 45 to 79 years who were of European ancestry and free of clinical CHD at baseline: the Multi-Ethnic Study of Atherosclerosis (MESA; n = 1,991; median age, 61 years) and the Rotterdam Study (n = 1,217; median age, 67 years). Researchers used traditional risk factors to calculate CHD risk, including pooled cohort equations, CT for the CAC score and genotyped samples for a validated polygenic risk score. The primary outcome was model discrimination, calibration and net reclassification improvement (at the recommended risk threshold of 7.5%) for prediction of incident CHD events.

The findings were published in JAMA.

The median predicted atherosclerotic CVD risk based on traditional risk factors was 6.99% for the MESA cohort and 5.93% for the Rotterdam study. During available follow-up for the MESA (median, 16 years) and Rotterdam study (median, 14.2 years), incident CHD occurred in 9.4% and 8.1% of adults, respectively.

In the MESA cohort, both CAC score and polygenic risk score were significantly associated with 10-year risk for incident CHD, with HRs of 2.6 (95% CI, 2.08-3.26) for CAC score and 1.43 (95% CI, 1.2-1.7) for polygenic risk score.

“Higher rates of CHD were observed in categories with higher CAC scores with a dose-response relationship in both studies,” the researchers wrote. “In contrast, rates of CHD were similar with overlapping 95% CIs between a polygenic risk score of less than 50% and between 50% and 80% but was higher for those with scores higher than 80%.”

C statistics were 0.76 for the CAC score (95% CI, 0.71-0.79) and 0.69 for the polygenic risk score (95% CI, 0.63-0.71). The change in the C statistic when each score was added to the pooled cohort equations was 0.09 (95% CI, 0.06-0.13) for the CAC score, 0.02 (95% CI, 0-0.04) for the polygenic risk score and 0.1 (95% CI, 0.07-0.14) for both.

The researchers found that overall categorical net reclassification improvement was significant when the CAC score was added to the pooled cohort equation (0.19; 95% CI, 0.06-0.28), but reclassification did not improve when adding the polygenic risk score to the equation.

The findings were consistent in subanalyses with younger and older participants when stratified by the median age in MESA and the Rotterdam Study.

Longer follow-up needed

“The two scores measure different things,” Khan said. “The genetic risk, or inherited risk, is a probabilistic estimate of what we think that some commonly occurring genetic changes may put you at risk for. Whereas the risk factors measured themselves are directly related to risk for MI. With CT, we know that measuring or visualizing if there is calcium in the arteries is a very clear picture that the disease process is already happening.”

Khan said going forward, studies with longer-term follow-up with more diverse cohorts will be key, as the two cohorts analyzed were predominantly white individuals. The authors also noted that the two scores may be relevant at different life stages.

“CT scans can be helpful in middle age; however, in younger adults, we don’t have the utility of CT scans because it is unlikely calcium has developed. But that does not mean those younger adults are not at risk,” Khan said. “So, it is possible that a polygenic risk score may be more helpful for younger adults, but that is not clear yet. Partly because you need a very large sample of young people and very long follow-up.”

For more information:

Sadiya Sana Khan, MD, MSc, FACC, FAHA, can be reached at [email protected].