乌鸦传媒

To more fairly rank candidates for heart transplant, researchers developed the U.S. candidate risk score (US-CRS) and found it to better discriminate risk for people with advanced heart failure than the current six-status system for donor heart allocation.

The algorithm turned out to be more accurate at predicting death without transplant within 6 weeks in a test dataset of U.S. adult heart transplant candidates listed from 2019 to 2022. The area under the curve (AUC) for US-CRS was 0.79 (95% CI 0.75-0.83) compared with the current model's 0.68 (95% CI 0.62-0.73). Overall C-index reached 0.76 (95% CI 0.73-0.80) and 0.67 (95% CI 0.63-0.71), respectively, reported William Parker, MD, MS, PhD, of University of Chicago, and colleagues.

Notably, US-CRS leaves out intra-aortic balloon pump (IABP) and percutaneous ventricular assist device (VAD) in considering the risk associated with short-term mechanical circulatory support (MCS). People with durable left ventricular assist devices (LVADs) would also get lower medical urgency scores. Developers also opted not to include hemodynamics in the new model.

Converting risk scores to a 50-point medical urgency scale resulted in one in four candidates currently categorized at low priority status (Statuses 3 to 6) being sent to the top decile of risk of death, the investigators reported in .

"By assigning these high-risk candidates higher scores and less urgent candidates lower scores, the US-CRS model provides a clinically meaningful improvement to promote fair donor heart allocation," commented Michelle Kittleson, MD, PhD, of Cedars-Sinai Medical Center in Los Angeles.

"Remarkably, this model had significantly higher sensitivity for predicting waiting list death and specificity for predicting waiting list survival than any of the 6-status system thresholds," she wrote in an .

The new risk algorithm arrives at a time when the United Network for Organ Sharing (UNOS) and Organ Procurement and Transplantation Network (OPTN) are preparing to move donor heart allocation to continuous distribution. In this system, each person's medical urgency score is expected to weigh heavily in his or her listing priority, with other factors including expected post-transplant survival and placement considerations.

With this update, the beleaguered six-tier system will be no more. There is evidence that since its implementation in 2018, transplant physicians have been gaming the system with overtreatment and exception requests. For example, some suggest an unexpected crowding of Status 2 can be related to people getting IABPs and other devices, despite not being so ill, in order to move to the near top of the transplant waitlist.

This is presumably harder to do with the US-CRS system, which incorporates time-varying short-term MCS (ventricular assist-extracorporeal membrane oxygenation [ECMO] or temporary surgical VAD), the log of bilirubin, estimated glomerular filtration rate, the log of B-type natriuretic peptide, albumin, sodium, and durable left ventricular assist device as factors of medical urgency.

"While some variables in the US-CRS do rely on physician decisions, namely use of [ECMO] and surgically placed temporary MCS, these interventions have been subject to less variation since implementation of the 2018 allocation system, likely due to their increased invasiveness and attendant risk," Kittleson wrote.

As for leaving out hemodynamics, Parker and colleagues explained that they "made this decision because including hemodynamics did not result in major improvements, and hemodynamics are especially susceptible to manipulation. Centers can use varying measurement techniques and manipulation of existing inotropes or vasoactive drugs to obtain worse hemodynamic values for a patient."

In the current system, patients on ECMO have the highest priority, followed by those receiving temporary MCS, followed by those on inotropic support.

"Over the last 5 decades, heart transplant has emerged as the preferred therapy for select patients with advanced heart failure ... Yet demand for donor hearts far exceeds supply, and more than one-third of candidates die or are removed from the waiting list without receiving a transplant," the editorialist noted.

"Hopefully, the UNOS/OPTN will use this elegant analysis as the framework for a continuous distribution model for heart transplant allocation," Kittleson urged. "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."

She nevertheless said that questions remain for how the continuous allocation score would rank transplant candidates, especially those with antibody sensitization, women, underrepresented minorities, and others with residual risk not addressed in an algorithm.

US-CRS was derived from the French Candidate Risk Score model of medical urgency.

Parker and colleagues had training and test cohorts altogether totaling nearly 17,000 transplant candidates from the Scientific Registry of Transplant Recipients (mean age 53, 73% men, 58% white).

Ultimately, 4.7% of the cohort died without a transplant.

Registry data were center-reported and therefore subject to potential recall or misclassification bias, study authors acknowledged. In addition, deaths may have been undercounted in the database, they said.

Comment