Development of a Machine Learning Score to Predict Risk of Sudden Cardiac Death in Patients with Heart Failure with Preserved Ejection Fraction

Lead Investigator: Barry Greenberg, University of California, San Diego
Title of Proposal Research: Development of a Machine Learning Score to Predict Risk of Sudden Cardiac Death in Patients with Heart Failure with Preserved Ejection Fraction
Vivli Data Request: 9012
Funding Source: None
Potential Conflicts of Interest: None

Summary of the Proposed Research:

Heart failure is a common and deadly disease that affects millions of patients around the world. In the majority of patients with heart failure, the heart muscle has normal or near normal ability to pump blood while the patient is at rest. However, in these cases patients have stiff hearts which don’t relax properly. This limits the hearts’ ability to fill with blood. As a result, blood backs up in the patients’ lungs and in the rest of their body. This causes fluid to leak out of the blood vessels into the surrounding tissue and for congestion to occur. One of the major and most feared problems in patients with heart failure is that they are risk of sudden death. This devastating event is most often caused by an electrical problem in the heart that results in a very rapid heart rate and uncoordinated rhythm that can abruptly limit the heart’s ability to pump blood to the rest of the body. Patients who develop these heart rhythm problems often die within a matter of minutes and there is little chance to provide treatment for the abnormal rhythm in time to save them.

We know from earlier studies that about 1-2% of patients with heart failure with normal or near normal heart function at rest die suddenly but there are no good ways of determining whether an individual patient is at high or low risk of this devastating event occurring at some future date. A result of our inability to identify which patients are at high risk, there are no recommendations for how to treat patients so that sudden death can be avoided. Consequently, special pacemakers that can recognize and treat cardiac rhythm that develop, thereby preventing sudden death, are not offered to patients with heart failure with normal pump function since we would have to treat 100 patients with these devices to save 1-2 lives each year. Since the pacemakers are costly, have associated serious side effects including infection or delivering an inappropriate electrical shock their benefit to risk ratio is not high enough for recommending them to the entire population of patients with heart failure and normal pump function.

The goal of our project is to use patient information from data collected in a clinical trial which carefully identified instances of sudden death to develop a risk score that can predict the future likelihood of this event. If we are successful in developing such a score it would allow selection of high-risk patients who are likely to benefit from having one of these specialized pacemakers implanted in their heart while avoiding pacemaker implantation and exposure to the side effects associated with these devices in patients who at low risk. It is anticipated that by being able to identify patients with heart failure with normal ejection fraction who are at high risk of dying suddenly, tens of thousands of lives can be saved each year.

Requested Studies:

A Phase III Randomised, Double-blind Trial to Evaluate Efficacy and Safety of Once Daily Empagliflozin 10 mg Compared to Placebo, in Patients With Chronic Heart Failure With Preserved Ejection Fraction (HFpEF)
Data Contributor: Boehringer Ingelheim
Study ID: NCT03057951
Sponsor ID: 1245.110