AI Risk Prediction Tool Improves Treatment of Cancer Patients after Heart Attack

Because of this complexity, they have often been excluded from clinical trials and standard cardiovascular risk scores, leaving clinicians without clear guidance. This gap has made treatment decisions particularly difficult and inconsistent. A new study now shows that integrating cancer-specific factors with cardiovascular data can significantly improve risk prediction and treatment decisions for this vulnerable group.

An international team of researchers, led by the University of Zurich (Zurich, Switzerland), has developed the first risk prediction model designed specifically for cancer patients who suffer a heart attack. The artificial intelligence (AI)-based tool combines traditional cardiovascular indicators with cancer-related variables, such as tumor characteristics and treatment-related risks, to better reflect the real-world clinical complexity of these patients.

The model, called ONCO-ACS, was created using artificial intelligence to predict three major outcomes within six months of a heart attack: death, major bleeding, and recurrent cardiovascular events. Rather than treating cancer and heart disease as separate conditions, the tool captures their interaction. This allows clinicians to better balance competing risks, particularly when deciding on invasive procedures or antiplatelet therapies.

The study analyzed data from more than one million heart attack patients in England, Sweden, and Switzerland, including over 47,000 patients with cancer. Results showed that outcomes in cancer patients were markedly worse, with nearly one in three dying within six months and high rates of bleeding and recurrent events. The findings, published in The Lancet, demonstrated that ONCO-ACS significantly improves individualized risk prediction compared with existing tools.

The findings highlight the close biological and clinical link between cancer and cardiovascular disease. By providing a validated, personalized risk assessment, ONCO-ACS can help doctors determine which patients are likely to benefit from aggressive treatment and which face greater harm. The researchers aim to integrate the tool into routine clinical practice and use it to guide catheter-based interventions, antiplatelet therapy, and the design of future clinical trials focused on cancer patients with heart attacks.

“By accounting for both cancer and heart disease, ONCO-ACS marks a step towards truly personalized medicine,” said senior author Professor Thomas F. Lüscher. “It can help doctors decide who benefits from invasive procedures and intensive drug therapy, and who may be at greater risk of harm.”

Related Links:
University of Zurich