Critically Appraising the 2018 UNOS Donor Allocation Policy: Adding Life Boats or Rearranging the Deck Chairs?

Abstract

Purpose of review:

Due to the growing mismatch between donor supply and the demand as well as unacceptably high transplant waitlist mortality, the heart organ allocation system has been revised in October 2018. This review gives an overview of the changes in the new heart organ allocation system and its impact on heart transplant practice and outcomes in the United States.

Recent Findings:

2018 heart allocation system offers a 6-tiered policy and therefore prioritizes sickest patients on the transplant waitlist. Patients supported with temporary mechanical circulatory support devices are prioritized as Status 1 or Status 2, which have resulted in increased utilization of this strategy. Patients support with durable left ventricular assist devices have been prioritized as Status 3 or 4, which have resulted in decreased utilization of this strategy. Broader geographic sharing in the new heart allocation system has resulted in prolonged donor ischemic times. Overall, the new heart allocation system has resulted in significantly lower candidate waitlist mortality, shorter waitlist times, and higher incidence of transplantation.

Summary:

The new UNOS allocation policy confers significant advantages over the prior algorithm, allowing for decreased waitlist times and improved waitlist mortality without major impact on post-transplant survival.

“The only source of knowledge is experience”

– Albert Einstein

Introduction

The United Network for Organ Sharing (UNOS) began the process of critically appraising its donor allocation policy for heart transplantation in 2014¹. The impetus for this review was a combination of worsening mismatch between organ supply and demand, high waitlist mortality for the sickest patients, and an increasing number of heart transplant candidates being bridged to transplant with left ventricular assist devices (LVAD). In response to these issues, the thoracic subcommittee elected to expand the number of priority tiers within the revised allocation policy to provide more granular risk stratification and decrease the number of status exceptions. It is important to note that while the subcommittee considered the impact of the changes on post-transplant survival, the primary objective of the updated allocation system was to improve waitlist mortality.

The new allocation policy was implemented in October 2018 and consisted of a 6-tiered system with broader geographic sharing for those in the highest priority statuses (Table 1)². To date, over 10,000 patients have been listed with the new policy, with more than 7,000 surviving to transplant. While multiple reports have praised the policy change for its improvement in waitlist mortality and time to transplantation, others have criticized its impact on post-transplant survival, decreased priority for patients on LVAD support, longer travel times for procurement teams with associated longer ischemic times, and its impact on center behavior regarding listing strategies. In this contemporary review, we critically appraise the new UNOS allocation system for heart transplantation – evaluating the advantages and disadvantages of this recent change – and propose areas for ongoing improvement and innovation.

Table 1

Allocation Tiers	Description
1	• VA-ECMO • Non-dischargeable BiVAD • MCS with arrhythmia
2	• Dischargeable RVAD/BiVAD • Non-dischargeable LVAD • IABP or other percutaneous MCS • MCS with malfunction
3	• Continuous IV inotrope with hemodynamic monitoring • 30-day exception time for LVAD • MCS with complication
4	• Continuous IV inotrope without hemodynamic monitoring • Stable LVAD • Congenital Heart Disease • Restrictive Cardiomyopathy • Re-transplant
5	• Multi-organ transplant
6	• All other candidates

Intents and Purposes: Examining the Impact of the New Allocation System on Waitlist Outcomes

Three years after its implementation, multiple studies have confirmed that waitlist outcomes have improved under the new allocation policy. Kilic et al. examined 5,280 patients listed after October 18, 2018 and compared the competing waitlist outcomes of death or delisting versus transplantation to those listed between October 1, 2015 and October 1, 2018³. Risk of death or delisting for worsening clinical status was much lower in the current era [sHR 0.6 (95% CI 0.52-0.69; p<0.001)] and the odds of successful transplantation had dramatically increased [sHR 1.38 (95% CI 1.32-1.45; p<0.001)]. Even after adjustment for covariates including recipient age, gender, body mass index, diabetes, smoking status, and pre-transplant dialysis, the risk of death or delisting was lower and likelihood of transplant higher under the new policy⁴. These observations were likely a result of shorter waitlist times that were, in part, facilitated by broader sharing, as evidenced by significant increases in donor to recipient hospital distances³. Thus, to date, all data from the new allocation era suggests that the policy change has achieved its intended purpose of improving waitlist mortality and time to transplantation, particularly for the most medically urgent candidates⁵.

Means to an End: The Changing Landscape of Mechanical Circulatory Support Under the New UNOS Allocation System

Under the new allocation system, patients bridged to transplant with temporary mechanical circulatory support (MCS) devices such as venoarterial extracorporeal membrane oxygenation (VA-ECMO), non-dischargeable or percutaneous ventricular assist devices, and intra-aortic balloon pumps (IABP) who meet strict hemodynamic criteria are prioritized as Status 1 or 2, allowing access to an increased number of donors through broader geographic sharing. It was immediately recognized that waitlist IABP utilization increased dramatically after policy change, with estimates suggesting the proportion of patients bridged to transplant with IABP increased from under 10% to close to 25%⁶. While patient characteristics of those supported by IABP prior to transplant did not appear to be significantly different than in the prior era, patients supported by IABP on the waitlist under the new policy experienced a dramatic decrease in days on the waitlist (22 days v. 52 days), higher rates of transplantation (89.9% v. 76.4%), and lower rates of death (2.3% v. 8.5%). After transplant, there was no difference in short term survival of patients supported by IABP in the current versus previous era. The increase in IABP utilization appeared to be localized to those cardiac intensive care units at transplant centers and did not appear to reflect an overall paradigm shift in cardiac intensive care management of decompensated heart failure⁷.

While the use of IABP has increased dramatically post-policy change, rates of bridge to transplant with durable LVAD have steadily declined⁸. Between 2018 and 2019 alone, the proportion of patients with durable LVAD at the time of transplant decreased from 45% to 34%. The most recent data from the Scientific Registry of Transplant Recipients, including patients in 2020 and 2021, suggest that the proportion of patients on LVAD support at the time of transplant under the new allocation policy is now less than 20%⁹. BTT LVAD patients transplanted in the new policy appear to be more acutely ill than in the old allocation system, with increased utilization of temporary right ventricular support devices associated with pre-transplant ICU admission and poorer functional status¹⁰. Given the shorter waitlist times in the new heart allocation system, this data suggests that there has been a paradigm shift in pre-transplant support strategies, with temporary MCS being favored over durable LVAD as a way to avoid an additional sternotomy and the associated risks of LVAD as a bridge to transplant¹¹. In addition, LVAD supported patients are not eligible for status upgrade unless they develop a device-related complication, which may delay time-to-transplantation and increase risk for post-transplant complications. While prioritizing stable LVAD patients as Status 4 with upgrade for device complication was an intention of the new policy , the increase in Status 2 upgrades for IABP or other temporary MCS, may prevent LVAD patients with complications from being transplanted in a timely fashion.

Exceptions are the Rule

Another goal of the policy revision was to decrease the overall number of patients assigned to statuses as a result of exceptions. This was accomplished by the addition of more granular status tiers to accommodate common exception requests in the prior era, including device complications and congenital heart disease. In addition, under the new policy, status exception requests were no longer reviewed within their UNOS region of origin to prevent competition for donors and avoid regional differences in practice patterns¹. Despite these changes, up to 30% of patients continue to be assigned exception status in the new era, with over half of those being granted an exception at the time of initial listing¹². Specifically, 30% of patients listed as Status 2, 3, and 4 were listed with an exception, as were 20% of those listed as Status 1.. Significant region- and center-level differences are present in exception status utilization suggesting that center behavior is a key determinant of this strategy. Importantly, patients listed with status exceptions experienced higher rates of transplant (83.5% v. 60.0%, HR: 1.25, p<0.001). These findings suggest that the new heart allocation system may not fully capture medical urgency and complexity of waitlisted patients resulting in ongoing high utilization of exception status. These data also suggest that status exception review outside the requesting transplant center may not be an effective strategy in minimizing variability in status exception approvals or denials.

Lifeboats or Deck Chairs? Post-Transplant Survival in the New Allocation System

While early reports of post-transplant outcomes under the new allocation policy suggested a signal toward decreased post-transplant survival, more recent publications with larger number of patients and longer follow up time have been unable to identify an association between the policy change and increased post-transplant mortality^13–15. Moreover, certain high-risk groups, such as those with congenital heart disease and restrictive/hypertrophic myopathies also appear to have similar post-transplant survival to the pre-policy change era ¹⁶. Data on post-transplant survival in patients bridged to transplant with LVAD in the new allocation remains conflicting. While some studies report no association between post-transplant survival and policy era¹⁷, other suggest a significantly worse 1-year post-transplant survival under the current policy (91.7% v. 83.7%, HR: 2.0 (95%CI 1.4-2.9), p<0.001)¹⁸. In both studies, BTT LVAD patients in the new policy era were more likely to receive a Hepatitis C positive organ and experienced significantly greater ischemic times, likely owing to broader sharing. Thus, the overall impact of the policy change on post-transplant survival appears to be net neutral, but larger number of transplant patients with longer follow up data, particularly in those patients bridged with temporary or durable MCS, are needed.

Measurement and Meaning – The “Heisenberg Uncertainty Principle” in Action

The more precisely we know a datapoint, the less likely we may be able to know its momentum. This concept might be applied to new allocation, where centers may create the status for the patient, rather than patients naturally falling into predefined status criteria. Even prior to the implementation of the 2018 policy change, leaders in the field of transplantation were concerned about the impact of the new allocation system on physician and transplant center behaviors¹⁹. In particular, there was concern that the hemodynamic criteria for status escalation were too lax, and would classify a large number of ambulatory heart failure patients as “in shock” to justify deployment of device therapies and higher priority listing. This could greatly increase the volume of patients transplanted from high status tiers, which would make it difficult to separate meaningful gains for patients and their transplant centers²⁰. These behavior changes could also potentially invalidate the thoracic simulation allocation modeling upon which the new policy was built and tested, thereby adding a degree of uncertainty to the expected impact of the policy change. Indeed, these trends have played out in contemporary assessments of listing strategies before and after the policy change, where there has been a 5-fold increase in the number of patients listed in high priority statuses – primarily via IABP or status exception - without an associated change in patient characteristics or acuity²¹. These changes varied markedly by region and by transplant center, suggesting marked heterogeneity in the adoption of more aggressive listing strategies and further complicating the assessment of the new UNOS policy. The reality is, however, that as long as the heart allocation policy is based upon therapies deployed, changes in policy will result in behavioral changes that may or may not underlie the fundamental intentions of that policy.

In the Eye of the Beholder: Critical Appraisal of the UNOS 2018 Policy Change

We believe that 2018 change to the organ allocation policy for heart transplantation has been successful overall. It has achieved its intended goal of improving wait list outcomes for the sickest patients by allowing for broader geographic sharing and significantly decreasing overall waitlist times. It has done so without significantly impacting overall post-transplant mortality, although ongoing data collection and analysis to ensure this remains the case is of paramount importance. However, it also has significant flaws that warrant both intellectual scrutiny and future policy changes to ensure continued equitable distribution of organs and optimal patient outcomes. First, as previously mentioned the hemodynamic criteria that must be met for justification of Status 2 listing may not be enough, alone, to discriminate those patients with advanced, but compensated, ambulatory heart failure from those with evidence of hypoperfusion and deterioration or refractory shock. Thus, there is currently no safeguard against centers placing patients with abnormal hemodynamics on tMCS support to facilitate transplant in the absence of elevated lactate or worsening end-organ function that would justify the deployment of these therapies. This too can be demonstrated from SRTR data, where patients with Status 2 upgrades for tMCS that met hemodynamic criteria alone or were upgraded due to an exception had significantly better outcomes than those upgraded to Status 2 with tMCS and evidence of end organ dysfunction (Dr. Clerkin – personal communication). In order to decrease the “crowding” of patients into this higher tiered status and to guard against unwarranted escalation of therapy, the UNOS thoracic committee must impose stricter criteria for Status 2 upgrade that go beyond hemodynamics to include clear evidence of end-organ decompensation or clinical instability.

In addition, as a direct result of broader geographic sharing, ischemic time – a well-documented risk factor for primary graft dysfunction and post-transplant mortality– has increased significantly²². In one single center study, rates of severe primary graft dysfunction (PGD) increased from 5.4% to 18.7% in association with greater travel distance and time²³. The combination of increased ischemic time and inherent recipient risk for PGD may explain some of the trends in high early post-transplant mortality in BTT LVAD patients, though additional studies are needed to confirm this hypothesis²⁴.

With shorter wait times and ease of status upgrade, the inertia to move patients toward heart transplantation is greater than ever before. And while patients generally do well after surgery, we cannot forget that organ longevity is not infinite. If we choose to take a young non-ischemic patient with their first presentation of decompensated heart failure and/or cardiogenic shock directly and quickly to transplant, are we losing the opportunity to stabilize them and deploy guideline directed medical therapies which may ultimately lead to recovery? And in doing so are we shortening their life expectancy? Data suggests that in the new allocation system, rates of delisting for improvement in status have fallen dramatically likely as a result of the very short time to transplant. This raises the question of whether are we doing enough to prioritize those patients – even those in shock – who have potential for recovery with left ventricular unloading and aggressive GDMT toward LVAD therapy? These questions are particularly important in young recipients, where the median survival of 13 years following heart transplantation will not allow them to reach mid-life without re-transplantation – their candidacy for which of course would depend on the impact of immunosuppression on their renal function, development of malignancy, and post-transplant sensitization²⁵. The transplant community, as a whole, must grapple with these difficult questions and must work to come to consensus on how to approach this new dilemma. In particular, additional efforts must be made to distinguish those patients who will derive benefit from the advantages of rapid transplant in the new allocation system from those in whom longevity and life expectancy may be reduced as a result.

Similarities and Differences with the Eurotransplant Heart Allocation System

Eurotransplant, is the non-profit organization responsible for coordinating organ transplants in Austria, Belgium, Croatia, Germany, Hungary, Luxembourg, Netherlands, and Slovenia. Eurotransplant system prioritizes waitlist candidates with international high-urgency (HU) status, then to candidates with national high urgency status, and finally to elective candidates with normal urgency status (Table 2).²⁶ High urgency status patients include those who are refractory for current heart failure treatment and admitted to an intensive care unit, intermediate care unit or a heart failure unit of the transplant center, except for patients supported with assist device. Mandatory data including Swan-Ganz catheter data, laboratory values, and treatment information are submitted to International audit group for request of International HU status. Final decision of the request is based on a majority vote. The national HU status is granted according to national policies of each member state. The ranking of transplant candidates is based on several factors including medical urgency, donor-to-recipient ABO blood group, and donor-to-recipient country balance.

Table 2.

Eurotransplant Heart Allocation System Tiers

International High Urgency Status

● Inotrope dependency

○ Meeting following 3 criteria for cardiac hypoperfusion syndrome

■ Signs and symptoms of terminal heart failure

■ Hemodynamic features

● Cardiac index < 2.0 L/min/m2

● Mixed venous saturation < 50%

● Pulmonary capillary wedge pressure > 15 mm Hg

■ Secondary organ damage (at least one of the following)

● Lactate above normal

● Bilirubin above normal

● Calculated GFR < 60 ml/min

● Serum sodium < 135 mmol/L

○ Meeting one of the following 2 criteria for inotrope dependency

■ When on continuous inotrope infusion with Dobutamine, Milrinone, or Enoximone for at least 72 hours

● Cardiac index stays < 2.0 L/min/m2 OR

● There is proof of decrease in cardiac index below 2.0 L/min/m2 when inotrope is weaned

■ Cardiac hypoperfusion syndrome within 14 days of a single infusion of Levosimendan

● Complications while on assist device

○ Assist device complication requiring device exchange OR

○ Infection on assist device with positive cultures or other proof (driveline infection excluded) OR

○ Repeated assist device related cerebral events demonstrated by CT scan

● Additional Special Cases

○ Primary graft failure within 1 week after transplant

National High Urgency Status

● Granted according to national policies

Normal Urgency Status

● Stable assist device or total artificial heart patients

● Stable heart failure patients not meeting urgency status criteria

While patients with temporary MCS device support are given the highest priority in the new UNOS heart allocation system, Eurotransplant system does not assign priority to these candidates. While both UNOS and Eurotransplant system prioritize complicated LVAD patients, specific criteria and definitions for LVAD complications are more explicit and detailed in the UNOS system. As opposed to UNOS, Eurotransplant system also factors in the donor-recipient age matching into account.

Conclusion

In conclusion, the updated UNOS allocation policy confers significant advantages over the prior algorithm, allowing for decreased waitlist times and improved waitlist mortality without major impact on post-transplant survival. Questions of how to further risk stratify those patients with abnormal hemodynamics remain pressing and the impact of significant increases in ischemic time has yet to be fully realized. Lastly, the transplant community must engage in conversations around identifying those patients in whom rapid listing and transplantation will confer long-term survival benefit and distinguish from them those patients in whom significant recovery potential, with medical or device therapies, exist.

Key Points.

• 2018 heart allocation system offers a 6-tiered status policy and prioritizes sickest patients on the transplant waitlist, which have resulted in significantly lower transplant candidate waitlist mortality, shorter waitlist times, and higher chances of transplantation

• Utilization of temporary MCS devices including VA-ECMO, IABP, and percutaneous LVADs has significantly increased while utilization of durable fully implantable LVADs has decreased in the new heart allocation system.

• Broader geographic sharing in the 2018 heart allocation policy has resulted in significantly increased donor ischemic times, and exception status utilization remains high in the new heart allocation system.