Trust in an organ allocation system is predicated on fairness, a belief that organs will be directed to the most deserving patients first. To this end, the Organ Procurement and Transplantation Network contractor, the United Network for Organ Sharing (UNOS), created a system that prioritizes the allocation of hearts to patients with the highest likelihood of dying while waiting for a donor organ. Revised in 2006 to incorporate broader regional sharing, this allocation system has succeeded in directing hearts to a greater proportion of high-urgency patients (UNOS 1A) and in so doing reduced waiting list mortality without compromising post-transplant outcomes (1).

This new allocation system left in place a provision for the elective use of 30 days of high-urgency status, commonly referred to as 1A time, for recipients of implantable left ventricular assist devices (LVAD). The rationale for granting this specialized status stemmed from an earlier experience with pulsatile, implantable LVADs in which the mortality in the first 3 weeks after LVAD surgery was 5% to 10% per week (2). These LVAD recipients were granted 30 days of 1A status from the date of LVAD implant, but some experienced inferior outcomes by proceeding with transplant surgery so soon thereafter. Recognizing that LVAD-supported patients remained at risk of device failure over time, a provision was made to allow them to carry forward these 30 prioritized days to be used at any time at the listing center's discretion. This prioritization has remained even after the emergence of continuous flow LVADs, which are less likely to fail and have produced higher survival rates (3). This has created a perception of stable patients being able to “jump the list” ahead of other, more critically ill patients and has cast doubt as to whether the UNOS donor heart allocation system in its present configuration is indeed fair (4).

To be considered fair and balanced, an organ allocation system must be guided by 2 ethical principles, maximizing utility and distributing resources justly. These 2 ideals are occasionally at odds with one another whereby prioritizing one compromises the other. If one wanted simply to maximize the utility of a transplanted heart by focusing on the number of years of life gained, older recipients or those with certain comorbidities would be passed over in favor of younger, more robust recipients with a greater life expectancy after transplant. Conversely, in prioritizing justice, one would choose to ensure equal access to lifesaving organs, usually in the face of greatest need, regardless of the outcome.

Heart transplant programs must also be guided by these principles when deciding the appropriateness of transplant for an individual candidate and when considering how best to manage deserving patients while awaiting transplantation. Such decisions boil down to answering 4 practical questions: First, what are the expected outcomes with and without a transplant? Second, what are the expected outcomes while waiting for a donor heart? Third, what are the expected outcomes after LVAD surgery? Finally, what is the expected survival after transplant, whether bridged with medical therapy or with an LVAD?

In this issue of the Journal, we are provided some insight into 2 of these questions. In the first paper, Teuteberg et al. (5) evaluated the discriminatory value of the Destination Therapy Risk Score (DTRS) in patients receiving a continuous flow LVAD, the HeartMate II (Thoratec Corporation, Pleasanton, California). They retrospectively analyzed prospectively collected data from 2 mechanical circulatory support trials including >1000 patients. They discovered that the DTRS was a poor mortality risk discriminator for bridge to transplant recipients and a modest discriminator for destination therapy patients. Furthermore, the score failed to characterize a population in whom mechanical support would be futile. The authors are to be congratulated for providing us with a cautionary tale about prematurely adopting risk predictor models into clinical decision making. Even though this DTRS has been widely applied and almost universally accepted, this score was never sufficiently prospectively validated. Not only did the DTRS fail to risk-stratify recipients of a continuous flow pump, but it failed to effectively risk-stratify destination therapy recipients of a HeartMate XVE, a population similar to the derivation cohort. This is an important and timely observation for those working in the field of mechanical support, a field that depends on accurate and effective risk predictor models to advise patients and inform clinical decision making.

In an adjoining paper, Dardas et al. (6) examined the outcomes for wait-listed registrants to examine whether disparities in risk exist within and between UNOS status designations. They reviewed data on >15,000 patients collected by the Scientific Registry of Transplant Recipients from 2005 to 2010, a period of time characterized by the transition from first-generation pulsatile LVADs to smaller continuous flow devices. They report that the chance of experiencing an adverse event (death or wait-list removal for ineligibility) varied significantly within status 1A indications. Those with the lowest risk were LVAD patients using elective 30-day 1A time (1% cumulative hazard) followed by those on medical therapy (6%), LVAD complication (6%), and 1A paracorporeal ventricular assist devices (15%). Over this time period, candidates listed status 1A with an implanted LVAD without complication increased from 11% to 26%, while those on medical support decreased from 44% to 39%. There were no significant differences in survival after status 1A transplantation except for those patients who were previously ventilator dependent.

Should clinically stable LVAD patients continue to receive prioritization for donor hearts? Dardas et al. (6) contend that they should not because doing so violates the justice principle. First, they cite the higher risk for adverse events in medically supported or mechanically ventilated patients and suggest that these patients may be disenfranchised by preferentially allocating hearts to stable LVAD patients who have the lowest risk of wait-list mortality. Second, they argue that there is no utilitarian reason to prioritize stable LVAD recipients because transplant outcomes are not superior in this group compared with other 1A or 1B subgroups. Third, they suggest that in an ideal system there should be no disparity and that all 1A patients should be at equal risk.

These are compelling arguments but not persuasive. Consider the following in rebuttal. First, there is no direct evidence that prioritizing stable LVAD patients has prevented other 1A patients from receiving timely transplants. In fact, wait-list mortality has actually gone down (1). It is likely that centers are exercising good judgment by timing these upgrades to avoid competition with other 1A patients, thereby preserving their exposure to donor hearts. Second, before concluding that in the absence of a utilitarian reason there is no rationale to justify 1A prioritization one should be reminded of the German experience (7). In the Eurotransplant system, there is no prioritization for stable LVAD patients, which effectively eliminates the likelihood of their receiving a transplant in the absence of developing a device complication. Choosing to remain in urgent status on medical therapy instead of accepting an LVAD may increase the odds of receiving a transplant but does so at the risk of dying or becoming ineligible while waiting. Patients who opt for watchful waiting but end up requiring bailout LVAD placement have inferior survival while on the waiting list and, among the few who receive a heart, after transplant. Finally, even if one were to craft an entirely new allocation system, disparities in risk would still exist. Consider the authors' own data. Patients supported with paracorporeal ventricular assist devices or mechanical ventilation had risk profiles exceeding those of the other 1A categories, including those who are medically supported or have an LVAD complication.

Despite these differences in opinion, there is agreement that the current allocation system needs further refinement. The widespread adoption of smaller, continuous flow LVADs has begun to deliver on the promise of minimizing the risk of mechanical support and maximizing its beneficial outcomes. They are also improving the short- and mid-term survival of transplant-eligible recipients. How UNOS adjusts to this changing landscape is not yet clear. The adoption of a heart allocation score, similar in principle to the model for end-stage liver disease or the lung allocation score, is one possibility. Another is to expand the number of prioritization categories matching individual risk profiles. Regardless of which system emerges, facilitating this change will require robust data collection and analysis similar to the ones published today to achieve a fair and balanced system for our patients.