Human heart showing surrounding veins and arteries to represent cardiovascular disease as predicted by polygenic risk scores
Credit: LEONELLO CALVETTI/SCIENCE PHOTO LIBRARY

A nationwide research team led by University of Utah (U of U) Health investigators has developed a calculator that can predict a person’s individualized risk of complications following mechanical heart pump surgery. The surgery to implant the mechanical heart pump is often indicated for patients with late-stage failure of the left side of the heart, a condition that is often fatal and affects hundreds of thousand of patients in the U.S. alone. As vital as the surgery may be, it is also risky and can lead to right heart failure where the heart is no longer able to pump enough blood into the lungs.

The team’s research was published Wednesday in the journal JAMA Oncology.

Traditionally, identifying which patients were more at risk for complications resulting from this surgery has been difficult. For patients who undergo surgery to implant a left heart pump, the risk of subsequent right heart failure is very high, between 15% and 30%. But a variety of factors contribute to higher risk and predicting this has been “exceptionally difficult” according to first author of the study Iosif Taleb, MD, a cardiology fellow at the University of California, San Diego (UCSD).

“Each patient is unique with different health conditions and heart characteristics,” adds Taleb who helped develop the calculator at U of U Health while he was a clinical research fellow. “Heart pumps also have specific traits, and the combination of these factors makes predictions tough.”

Professor of cardiology at U of U health and senior author of the study Stavros Drakos, MD, PhD, noted that attempts have been made in the past to better predict which patients would not do well after getting a heart pump—also called a left ventricular assist device, or LVAD—but they did not perform well in real-world clinical setting. Oddly, some models that appeared to be giving accurate predictions at one hospital didn’t perform to the same level of accuracy at other hospitals.

To solve this, the Utah Health-led team used data from 1,125 patients from six different health centers and took into account variables from pre-existing conditions, medications, and demographics. They used machine learning algorithms to develop different risk models for these patients to find the one that performed best at predicting post-surgery outcomes. The selected risk calculator, now named STOP-RVF, was then applied retrospectively to patient data from another hospital to check that its capabilities were transferrable. When they compared the calculator’s predictions with the real-world outcomes, they found it still accurately predicted right heart failure risk.

Working across a set of geographically dispersed health systems was essential for the development of STOP-RVF. Drakos said: “It’s important because we live in a very diverse country. By basing this analysis in multiple sites all over the country—the Washington, DC, area, the Detroit area, California, Utah, and the broader Mountain West area—it’s representative of a large part of our country. It strengthens the generalizability of the work.”

The calculator is now in use by surgeons, cardiologists, and nurse coordinators of the heart failure and LVAD team at U of U Health. “It helps tailor the risk assessment for each patient, allowing for better preparation before surgery,” Taleb explains. “For patients who have a high risk of right heart failure, doctors can delay the surgery, use different medications to improve patients’ odds of recovery, or consider alternative treatments.”

Also of Interest