Prevencio's AI blood test predicts cardiovascular risk one year out for people with diabetes: study

By using artificial intelligence to analyze dozens of protein biomarkers at once, Prevencio showed its blood test could be accurate in predicting whether a person with diabetes may have a heart attack or a stroke within the coming year.

The company's HART CVE test could help classify whether patients have a higher or lower risk of a major cardiac event, according to Prevencio. Healthcare providers could then use that information to decide on the best treatment option, with more aggressive therapies for some while avoiding potentially expensive or invasive treatments for others.

Patients with diabetes, which includes more than 34 million people in the U.S., are about twice as likely to also have cardiovascular disease compared to those without diabetes.

Researchers from Massachusetts General Hospital (MGH) followed 450 patients for one year after a single blood draw. The HART CVE test assigned each person a score from one to 10 and grouped them by low, moderate and high risk. 

In the high-risk category, patients with diabetes were 25 times more likely to suffer a heart attack, stroke or cardiovascular death within the year after the test was performed, the company said.

“This multiple-protein, weighted-risk test provides improved accuracy as compared to clinical risk factors and could be particularly useful for patients with diabetes who also have suspected or known stable and acute heart disease, for which there are very few prognostic risk models,” said MGH cardiologist James Januzzi, M.D., a professor of medicine at Harvard Medical School and principal investigator for the trial of Prevencio’s HART CVE test. 

The study’s results were presented earlier this month at the annual scientific sessions of the American College of Cardiology.

“In addition to leading to more appropriate care of patients with diabetes, it could also have a role in ‘enriching’ pharmaceutical cardiac clinical trials, which could lead to lower trial costs and reduced time to complete clinical trials,” Januzzi said.

Prevencio also develops multi-biomarker tests for diagnosing Kawasaki disease and obstructive coronary artery disease, with the latter demonstrating that it could be more accurate than measurements of the heart protein troponin alone.

Last November, the company presented data on its HART CADhs test from patients with diabetes, showing it could predict the presence of blocked coronary arteries with an accuracy of 81%. 

A high sensitivity test for troponin, which is released by the heart muscle under damage or stress, had delivered an accuracy of 64%, the company said, while standard-of-care cardiac stress tests have posted an accuracy of 52%.