New AI tool predicts repeat heart problems in cancer patients

Researchers developed ONCO-ACS, an AI model that combines cancer-related and clinical data to predict six-month risks of another heart attack, major bleeding, or death; the analysis used nearly 50,000 patients and was published in The Lancet.

Researchers have developed an artificial intelligence tool to estimate short-term cardiological risks in people with cancer. Cancer treatments such as chemotherapy and radiation can affect the heart, a condition known as cardiotoxicity that raises the chance of further cardiac events. The University of Leicester team created ONCO-ACS to combine cancer-related factors with routine clinical data and predict the risk of another heart attack, major bleeding, or death within six months. The model and its results were reported in The Lancet after analysis of data from nearly 50,000 patients across England, Sweden, and Switzerland. Key findings: - ONCO-ACS uses cancer-specific and clinical variables to provide six-month risk estimates for repeat heart attack, major bleeding, and death. - The study analyzed records from nearly 50,000 people with cancer and acute coronary syndrome across England, Sweden, and Switzerland. - The findings were published in The Lancet and the lead author, Florian A. Wenzl, M.D., commented that this population has been underrepresented in research. - The Cleveland Clinic is cited noting that some cancer treatments, including chest radiation, are associated with higher cardiotoxicity risk. Summary: The paper reports that ONCO-ACS can give doctors information to help balance benefits and harms when treating cancer patients who have had a heart event. Undetermined at this time.