Cancer patients who survive a heart attack face unusually high risks and few tailored guidelines. A new AI tool, ONCO-ACS, offers doctors a way to personalize care and better balance life-saving treatments with the danger of bleeding and new cardiac events.
Cancer patients who survive a heart attack often find themselves in a medical gray zone: their risks are higher, but the tools doctors use to guide treatment were never built with them in mind.
A new artificial intelligence tool, developed by an international team led by researchers at the University of Zurich, aims to change that by giving clinicians a way to predict key risks and tailor care for this vulnerable group.
The model, called ONCO-ACS, is the first risk prediction tool designed specifically for people with cancer who experience an acute coronary syndrome, a term that includes heart attacks and related emergencies. The study, published in The Lancet, analyzed data from more than 1 million heart attack patients in England, Sweden and Switzerland, including over 47,000 with cancer.
The numbers underscore how precarious recovery can be for these patients. Nearly one in three cancer patients in the study died within six months of their heart attack. About one in 14 suffered a major bleed, and roughly one in six had another heart attack, stroke or cardiovascular death in that same period.
Yet despite these dangers, cancer patients have often been excluded from major cardiology trials and from the risk scores that guide standard care. That has left doctors without a validated way to weigh the benefits and harms of aggressive treatments, such as invasive procedures or powerful blood-thinning drugs, in this group.
“To provide targeted treatment for these patients, clinicians need more accurate tools to assess individual risk profiles,” first author Florian A. Wenzl, from the Center for Molecular Cardiology at the University of Zurich and the National Health Service England, said in a news release.
ONCO-ACS is designed to fill that gap. The AI-based model combines cancer-related information — such as tumor characteristics — with standard clinical data used in cardiology. It then estimates a patient’s chances, over the next six months, of dying, suffering a major bleed, or experiencing another heart-related event.
This approach reflects a growing recognition that cancer and cardiovascular disease are deeply intertwined. Some cancers and cancer treatments can make blood more likely to clot, raising the risk of heart attacks and strokes. Others, or their treatments, can increase bleeding risk. Many patients face both threats at once.
“Depending on the tumor characteristics, cancer patients can be at elevated risk of bleeding, of arterial blood clotting, or both – each requiring different anti-platelet medication for secondary prevention after the acute event,” Wenzl added.
That balancing act is especially challenging after a heart attack. Standard care often involves catheter-based procedures to open blocked arteries and long-term use of antiplatelet drugs, which reduce clotting but increase bleeding risk. For someone with cancer, the wrong choice can be dangerous either way.
By integrating cancer-specific and cardiac data, ONCO-ACS aims to give clinicians a clearer picture of each patient’s risk profile. That, in turn, can help them decide how aggressively to treat, how long to continue certain medications, and which patients might need closer monitoring.
“By accounting for both cancer and heart disease, ONCO-ACS marks a step towards truly personalized medicine. It can help doctors decide who benefits from invasive procedures and intensive drug therapy, and who may be at greater risk of harm,” added senior author Thomas F. Lüscher from the National Heart and Lung Institute, Imperial College London and the Royal Brompton and Harefield Hospitals.
The researchers validated ONCO-ACS using large, real-world datasets from three countries, strengthening the case that it can be applied broadly in clinical practice. Because it is based on routinely collected information, the model could potentially be integrated into hospital electronic systems or decision-support tools that clinicians already use.
If adopted widely, ONCO-ACS could influence several aspects of care. It may guide decisions about whether to perform catheter-based interventions, how to choose and dose antiplatelet therapies, and how to align treatment with a patient’s overall prognosis and cancer plan. It could also help standardize care for a group that has historically fallen outside the scope of traditional guidelines.
Beyond day-to-day practice, the tool offers a way to design better clinical trials. By identifying which cancer patients are at highest risk of death, bleeding or new ischemic events after a heart attack, researchers can more precisely target interventions and measure their impact. That could accelerate the development of therapies and strategies tailored to this complex population.
The study highlights a broader shift in medicine toward using AI not just to automate tasks, but to untangle overlapping conditions that defy one-size-fits-all rules. For patients living with both cancer and heart disease, that shift could mean care that is not only more precise, but also more humane — grounded in a realistic understanding of their risks and priorities.
The next step will be to embed ONCO-ACS into clinical workflows and test how it performs in real-time decision-making. As hospitals and health systems explore AI tools, the model offers a concrete example of how data-driven predictions can support, rather than replace, the judgment of experienced clinicians.
For now, ONCO-ACS gives cardiologists and oncologists something they have long lacked: a dedicated, evidence-based way to navigate some of the toughest treatment choices their patients face after a heart attack.
Source: University of Zurich