New Risk Score Better for Capturing Medical Urgency for Heart Transplantation

A novel candidate risk score (CRS), one that includes clinical and laboratory variables, is able to accurately predict the risk of death in patients awaiting heart transplantation, according to results of a new study.

The “continuous distribution” risk score outperformed the existing risk classification system, as well as another score currently used in France, and should provide clinicians with an objective means to assess medical urgency for the allocation of scarce donor hearts, researchers say.

“These patients on the heart list are very sick. They all need heart transplants and would have a substantial survival benefit from the procedure . . . but there are some people who have a higher risk of death than others,” senior investigator William F. Parker, MD, PhD (University of Chicago, IL), told TCTMD. “What our objective score does is use known laboratory markers that have strong associations with the severity of heart failure to predict who's going to die. That’s the big step forward. The current system is based mostly on treatment choices.”

In the United States, there were approximately 4,500 heart transplantations performed in 2023, a number that has been steadily climbing for at least 10 years. Despite the increasing number of transplantations, donor hearts remain hard to come by, which has led the Organ Procurement and Transplantation Network (OPTN) to prioritize patients who have a high risk of dying unless they receive a new heart.

Medical urgency is currently defined using a tiered status system that relies mostly on the intensity of treatment, ranging from status 1, which includes patients who are critically ill and on mechanical support, to status 6, which includes stable patients not requiring mechanical support or intravenous medications. Prior work by Parker and colleagues has shown that the current scoring system fails to identify patients with the most urgent medical need. Moreover, the current system can be manipulated with overtreatment or exception requests.

“The issue with relying on transplant programs to rank-order patients into these categories is that the subjective choice to escalate therapy to a more intensive level is what sets the priority relative to other patients rather than their underlying severity of advanced heart failure,” said Parker.

In an editorial, Michelle Kittleson, MD, PhD (Cedars-Sinai Medical Center, Los Angeles, CA), said the current system is “imperfect with the potential for manipulation, however well-intentioned, to garner higher priority for some candidates without appropriate medical justification.” The development of a continuous allocation risk score can provide a more accurate assessment of patient urgency, she added. “This will be an important step away from subjective physician decision-making toward a system that balances beneficence and justice to optimize equitable access to transplant for all patients with advanced heart failure,” writes Kittleson.

The new study, with lead author Kevin Zhang, MS (University of Chicago), and editorial were published in the February 13, 2024, issue of JAMA.

Objectively Assessing Medical Urgency

For all organ transplantations, the OPTN is shifting towards a continuous distribution system to prioritize the sickest patients first. As opposed to the 6-status system for heart transplantation, which puts patients into broad buckets based mostly on their treatment, continuous distribution considers multiple factors simultaneously to create a score based on medical urgency, patient outcomes, biological makeup, and other factors, such as how quickly an eligible patient can get to the hospital for transplantation. Continuous distribution scoring has already been adopted for lung transplantation and the OPTN will soon implement it for other organs.

“What we're hoping is that our model, or something like it, some improved version of it, will end up being the medical urgency component of a continuous distribution system for hearts,” said Parker.

The CRS, known as the US-CRS, was developed by building on an existing risk score used in France (FR-CRS). To develop the risk score, the investigators selected 46 demographic, clinical, laboratory, and hemodynamic variables and tested these in a logistic regression analysis. The final US-CRS includes use of extracorporeal membrane oxygenation (ECMO), surgically placed mechanical circulatory support (MCS), measurements of bilirubin, albumin, sodium, B-type natriuretic peptide, and estimated glomerular filtration rate, and use of a durable left ventricular assist device (LVAD).

“We’re also collecting laboratory values more than once,” said Parker. “It’s not the first snapshot of when the person gets on the waitlist, but when [the predictor values] change over time.”

The study cohort included 12,362 candidates for heart transplantation at 97 centers, which formed the dataset to train the risk model, and 4,543 candidates from 41 centers that served as the test dataset. Of the nearly 17,000 total patients (mean age 53 years; 73% male), 58% were white, 26% were Black or African American, 10% were Hispanic or Latino, 4% were Asian, 0.4% were American Indian or Alaska Native, and 0.4% were Hawaiian/other Pacific Islander. Among these, 4.7% died while on the waiting list or within 6 weeks of removal from the list while 71.3% had a transplant. Nearly 24.0% were either still on the waiting list or had been removed from the list as of March 2023. 

The current 6-status model had an area under the curve (AUC) of 0.68 for death within 6 weeks without a transplant compared with AUCs of 0.72 and 0.79 for the FR-CRS and US-CRS, respectively. The C-index was 0.67, 0.69, and 0.76 for the 6-status model, FR-CRS, and US-CRS, respectively.

Moving From Low to High Priority

To TCTMD, Parker said that they constructed a 50-point medical urgency score for the final US-CRS model by mapping the predicted mortality in the next 6 weeks to percentiles. A US-CRS score of 50 means that patients would be in the 98th or 99th percentile of risk, for example. One “startling” finding from the study was that 25.8% of patients considered to have a low priority for transplant based on the current 6-status model were at high risk for death.

“There are candidates who currently are status 3 or status 4 in the current system who nevertheless are in the top 20% for risk of death,” he said. “That points to the limitation of this current system to identify patients who are sick based on their laboratory values, but who aren't being treated with the most intense therapies. As a result, they don't get bumped up to status 2.”

Such patients might not be treated with more intense therapy because clinicians may think it would be ineffective, but the patient is penalized under the current system for not receiving it. With the proposed US-CRS, physicians can provide the therapy they consider the most appropriate without worrying about the priority status for transplantation. “Right now, I know the status system weighs heavily on the minds of heart transplant cardiologists and surgeons when they're making treatment choices,” said Parker.  

He noted that if patients had an LVAD, they had a much lower risk of death at 6 weeks. Future risk models will need to account for the time patients spend on LVAD so that they eventually undergo transplant, he said. “They’re always going to have low medical urgency scores because the devices work really well,” said Parker. “Even if they have a complication, often the complication doesn’t increase the risk of death much.”

Parker also noted that the US-CRS did not include intra-aortic balloon pumps or percutaneous ventricular assist devices, such as Impella (Abiomed).

“The reason is because when we included those in the definition, the coefficient moved towards zero,” said Parker. “That basically means that the patients who are on these devices, balloon pumps and Impellas, are not as sick as patients receiving ECMO and surgically placed mechanical circulatory support. So, that's something, if this score is implemented, that would help reverse the trend away from use of intra-aortic balloon pumps and percutaneous VADs because there's not going to be any priority bump with using those devices.”