Ranking heart transplant candidates in terms of medical urgency: Is there a better way?

Researchers have developed a new risk score that could help doctors identify which heart transplant candidates face the highest mortality risk if they do not receive a new heart. The group shared its findings in JAMA, noting that their risk score appears to outperform other common methods for evaluating transplant candidates.

“Heart transplant is a definitive life-saving therapy for eligible patients with advanced heart failure, but the treatment is extremely scarce, with less than 4,500 transplants performed annually in the United States,” wrote first author Kevin C. Zhang, MS, with the University of Chicago’s department of medicine, and colleagues. “To make the best use of scarce donor hearts, the U.S. Department of Health and Human Services mandates that the Organ Procurement and Transplantation Network (OPTN) prioritize medically urgent candidates with a high risk of death without transplant. The OPTN currently determines medical urgency in heart candidates with a categorical six-status system, based mainly on treatment intensity.”

The OPTN’s six-status system is flawed, the authors wrote, primarily because it is easy to manipulate and ignores laboratory-established measures of illness severity. Inspired by the French Candidate Risk Score (French-CRS) model, which provides more accurate candidate rankings using laboratory, Zhang et al. decided to develop their own U.S. Candidate Risk Score (US-CRS) for evaluating candidates for adult donor hearts.

The US-CRS model added additional variables to the French-CRS model. Variables included short-term mechanical circulatory support, bilirubin, albumin, sodium, B-type natriuretic peptide, estimated glomerular filtration rate and the use of a durable left ventricular assist device (LVAD).

To train and then test the US-CRS model, the researchers focused on registry data from nearly 17,000 U.S. heart transplant candidates with a mean age of 53 years old. All transplant candidates were initially listed from January 2019 to December 2022. While data from 70% of the participating centers were used to train their US-CRS model, data from the remaining 30% were used to test the model’s efficacy.

Zhang et al. used the test data to compare the US-CRS model, French-CRS model and OPTN’s six-status model. Overall, when focused on predicting six-week mortality, the US-CRS model had an area under the ROC curve of 0.79, higher than the French-CRS model (0.72) and six-status model (0.68).

The researchers did note that they chose not to include hemodynamic data in their final US-CRS model.

“We made this decision because including hemodynamics did not result in major improvements, and hemodynamics are especially susceptible to manipulation,” they wrote. “Centers can use varying measurement techniques and manipulation of existing inotropes or vasoactive drugs to obtain worse hemodynamic values for a patient.”

Future versions of the US-CRS model could include some hemodynamic parameters, the authors added, if it can be done in a way that not prone to manipulation.

Reviewing these findings, Zhang and colleagues believe their risk score showed potential to predict the six-week mortality of heart transplant candidates with more accuracy than other current methods.

“The US-CRS has better discrimination than the current six-status ranking system used for donor heart allocation in the U.S., suggesting that it may be useful for ranking patients by medical urgency,” they concluded. “The OPTN should consider implementing the US-CRS to quantify medical urgency in the upcoming continuous distribution system for donor heart allocation.”

Click here to read the full study in JAMA.

Michael Walter
Michael Walter, Managing Editor

Michael has more than 18 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.

Around the web

Several key trends were evident at the Radiological Society of North America 2024 meeting, including new CT and MR technology and evolving adoption of artificial intelligence.

Ron Blankstein, MD, professor of radiology, Harvard Medical School, explains the use of artificial intelligence to detect heart disease in non-cardiac CT exams.

Eleven medical societies have signed on to a consensus statement aimed at standardizing imaging for suspected cardiovascular infections.