The Trouble With Exceptional Exceptions

In this month's issue of the American Journal of Transplantation, Braun et al 1 evaluated outcomes for pediatric liver transplant patients as a function of nonstandard exception requests (NSERs) for pediatric end-stage liver disease (PELD) points. However, before delving into the details of their article, we first provide some background information to put the research and conclusions into perspective.

Liver transplantation is an effective treatment for irreversible liver disease, and outcomes have improved due to effective immunosuppression, improved technical expertise, increased availability of interdisciplinary liver transplant care models and better infection prevention strategies. Over the past two decades, the overall demand for liver transplants has grown and far exceeds the deceased donor organ supply. Complex rationing and allocation systems have developed parallel to the growing organ scarcity. The principles for allocation should be clear, equitable, provide utility and be iteratively monitored and modified so as to maintain public trust. The Final Rule was drafted in 1998 to form the foundation for organ allocation policy in the United States 2. Among the critical components of the Final Rule are that (1) allocation of organs from deceased donors shall be based on sound medical judgment; (2) allocation shall be designed to avoid wasting organs, to avoid futile transplants, to promote patient access to transplantation and to promote the efficient management of organ placement; (3) allocation shall not be based on the candidate's place of residence or place of listing; and (4) policies shall be reviewed periodically and revised as appropriate. The United Network for Organ Sharing (UNOS) develops policy, monitors metrics and administers organ allocation in the United States.

While survival and quality of life benefits exist for all transplant recipients, children with irreversible liver disease may be vulnerable to lifelong consequences associated with growth and cognitive deficits if the transplantation is delayed 3. Consequently, pediatric liver transplant candidates face risk and gain benefits from transplantation that are distinct from adult recipients. Approximately 550 deceased donor liver transplants are performed each year in patients less than 17 years old. More than 50 centers perform liver transplants in children in the United States. Children account for less than 5% of patients on the waiting list and account for about 8–10% of deceased donor liver transplants 4. The PELD model is a numerical scale that has been used since 2002 to prioritize children on the waiting list who are less than 12 years old. The PELD score is based on a patient's risk of dying while waiting for a liver transplant and is based on objective and verifiable medical data 5. The score is determined from serum albumin (g/dL), bilirubin (mg/dL), international normalized ratio, growth failure (based on gender, height, and weight) and age at listing. The metrics used to assess organ allocation include pretransplant mortality, rate of transplant and graft and patient survival. Since inception of the PELD scoring system and modification of the allocation system, pretransplant mortality has decreased substantially. The most dramatic decrease in mortality was observed for children less than 1 year old in whom deaths per 100 waitlist years decreased from approximately 45 to 25 4. Viewed from this surface-level perspective, the policies for prioritization and allocation of deceased donor livers to children have been successful. But why is this the case? A deeper look compels us to ask if the decreased mortality is, at least in part, attributable to the nearly linear increase in the number of nonstandard exception requests since 2002.

As early as 2005, the pediatric liver transplant community noted an increased dependence on exception requests for PELD priority points 6. Exception scores have been requested for individual cases when transplant teams judge that the calculated PELD score does not accurately reflect the risk of mortality or morbidity. Standard exception scores are given to candidates with urea cycle defects, hepatoblastoma, hepatocellular carcinoma, primary hyperoxaluria, familial amyloidosis and hepatopulmonary syndrome, conditions for which there is consensus regarding risk. Following the implementation of policy changes to allow for standard exception scores, particularly in those conditions that were either elevated to status 1B after 30 days (urea cycle defects) or immediately listed as 1B (hepatoblastoma), transplantation occurred in a timely fashion. But a challenge remained. Those children with end-stage liver disease for whom allocation was dependent upon calculated PELD scores continued to incur morbidity in the form of frequent hospitalizations, complications of infection or portal hypertension and lost opportunities for social and neurologic development. To address the inherent risk of morbidity and mortality, it was incumbent upon pediatric liver transplant teams to increase allocation priority for children. Consequently, pediatric liver transplant teams increasingly submitted NSERs for PELD points. The requested PELD points were granted by regional review boards on a case-by-case basis.

In the current issue of the Journal, Braun et al (1) note that since 2002, the incidence of NSERs has increased fivefold, with approval of 90% of NSERs1. Braun et al sought further to examine the impact of approval and denial of NSERs on waitlist mortality and posttransplant patient and graft survival. To do so, they queried data from the Scientific Registry of Transplant Recipients (SRTR), focusing their analyses on pediatric liver transplant candidates listed between January 2009 and December 2014. The authors excluded candidates who had standard exception points and those who were listed for multiorgan transplant. The authors found NSERs for 44% of the 2581 candidates and found that 93% of requests were approved. NSER denial increased the risk of pretransplant mortality or removal from the waiting list and decreased posttransplant survival. There were differences in approval rates among regions, ranging between 100% in region 6 to 65% in region 5. Perhaps more concerning was the observation that candidates with public insurance were less likely to have had an NSER than those with private insurance.

Based on the observations made by Braun et al. 1, it is worthwhile to reexamine the current policy for prioritization for pediatric candidates as related to the Final Rule—specifically that allocation of organs from deceased donors shall be based on sound medical judgment and that allocation shall not be based on the candidate's place of residence or place of listing. The increasing use of NSERs suggests that calculated PELD scores and standard exceptions do not effectively prioritize almost half of the candidates. We must be careful when we interpret the effectiveness of the existing pediatric policies, specifically the decrease in waiting list mortality, because the pediatric liver transplant community has developed an effective workaround to address the challenges of the present prioritization. Second, equity is an emerging challenge. Braun et al (1) clearly show that patients who do not receive exceptions are disadvantaged, with higher rates of waitlist mortality and posttransplant graft loss. Even more concerning, Braun et al observed regional variation and disparities based on insurance.

If our goal is an excellent long-term outcome for all pediatric liver transplant recipients, we must offer a solution that allows for all children to have equal increased access to deceased donor livers while preserving the objectivity of disease severity scoring and reducing the role of wait time. It is also critical to determine if changes in prioritization and allocation to reduce morbidity and mortality on the pediatric waitlist while maintaining the principles set forth by the Final Rule can be made without significantly affecting adult outcomes. We ask if the liver transplant community should examine pediatric liver allocation policy to consider strategies that reduce the reliance upon exception scores to achieve a more equitable and transparent allocation policy.

Disclosure

The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.