How can we improve prioritization for liver transplantation in patients with hepatocellular carcinoma?

Abstract

Potential conflict of interest: Nothing to report. This article was supported by Research and Development, Veterans Affairs Puget Sound Healthcare System, Seattle, WA. See Article on Page 1343 Since the introduction of the Model for End‐Stage Liver Disease (MELD)–based allocation system, the transplant community has been struggling to determine how to best incorporate hepatocellular carcinoma (HCC) patients. This was a particular challenge because many have a low biological MELD score. The “solution” that was adopted in 2002 was to assign an artificially high MELD score to patients with HCC at the time of listing, regardless of their natural MELD score, and to increase that MELD score every 3 months, a concept known as the “MELD elevator,” until transplantation was achieved. The expressed aim of this system was to lead to an equal rate of wait‐list death or dropout between patients with and without HCC. It quickly became evident that too much priority was allocated to patients with HCC under this system. First, multiple studies showed that patients with HCC had much lower wait‐list mortality and dropout rates and had much higher transplantation rates than patients without HCC.1 Second, patients with HCC came to dominate liver transplant waiting lists and to account for an increasing proportion of all liver transplant recipients.2 This was the result both of the excessive prioritization awarded to patients with HCC and of the relentless increase in HCC incidence since 2002, which is not expected to peak until 2020. Third, the HCC MELD exception system contributed greatly to a steady increase in MELD scores at the time of transplantation.2 In response to these issues, the United Network for Organ Sharing (UNOS) has adjusted the HCC MELD exception system multiple times since 2002 in an attempt to reduce disparities between patients with and patients without HCC in wait‐list mortality and dropout rates. The priority MELD score assigned to patients with stage II HCC at the time of listing was 29 in 2002 when the MELD‐based system was introduced and was reduced to 24 points in 2003 and 22 points in 2005. This system was in place until October 2015 when 2 additional changes were introduced. First, the “HCC Delay” policy mandated a 6‐month delay after listing for transplantation before assigning an exception MELD score of 28, followed by the scheduled progression of exception points on the “MELD elevator.” This policy was meant to serve a dual purpose of both slightly reducing access to transplantation in patients with HCC (in regions where transplantation occurs at MELD scores <28) and to identify patients with very aggressive tumor biology before transplantation. Second, the “Cap HCC” policy capped the scheduled progression on the “MELD elevator” to a maximum assigned score of 34, so that these patients would not be candidates for default regional sharing under the Share 35 policy introduced in June 2013.3 The Cap HCC policy may reduce disparities between HCC and non‐HCC patients by reducing the very high wait‐list mortality and dropout rates of non‐HCC patients with MELD scores ≥35 because they will not have to compete with HCC patients and by increasing the wait‐list mortality and dropout rates of HCC patients in regions where MELD scores at transplant can exceed 34. These adjustments introduced by UNOS are clearly steps in the right direction that may reduce the disparities between HCC and non‐HCC patients in adverse wait‐list outcomes—studies of their actual impact are eagerly awaited. However, the study by Alver et al.4 in this month's issue of Liver Transplantation and other similar recent studies5 prompt us to critically reexamine all 3 of the fundamental principles on which the current HCC MELD exception prioritization system is based. First, is the current principle of assigning the same MELD exception priority score regardless of the patient's natural MELD scores or tumor characteristics appropriate? The most important predictor of wait‐list mortality or dropout in patients with HCC is the natural MELD score followed by tumor burden (size and number) and serum AFP level. Therefore, if our aim is to reduce wait‐list mortality or dropout rate, then we should use these 3 characteristics to assign a priority score rather than assigning all HCC patients the same priority score. It does not make sense, for example, that a patient with a biological MELD score of 29 and a 5‐cm HCC has the same priority score as a patient with a biological MELD score of 7 and a 2.5‐cm HCC, after they have been on the waiting list for 9 months, even though the first patient has a dramatically higher likelihood of death or dropout. Alver et al.4 and others5 have developed models using primarily natural MELD score, tumor burden, and serum AFP level to predict wait‐list outcomes in patients with HCC that yield a score that is equivalent to the natural MELD score of patients without HCC in terms of the risk of wait‐list death or dropout. Although these models require further refinement and validation, in principle, they are superior to the current practice of assigning the same MELD priority score to all HCC patients at listing, and they could eliminate disparities between HCC and non‐HCC patients in wait‐list outcomes. Second is the principle of the “MELD elevator.” As a result of the “MELD elevator,” length of time on the waiting list is the only characteristic that determines prioritization among patients with HCC. This is unreasonable because time on the waiting list does not directly or strongly predict adverse wait‐list outcomes. In fact, patients with HCC who are stable for a long time on the waiting list may have a lower risk of dropout than patients who are recently listed. Historically, it is interesting to consider that prior to 2002, priority for non‐HCC patients was based primarily on the length of time on the waiting list, a system that was replaced by the current MELD‐based system, and yet, a time‐based priority system was at the same time introduced for patients with HCC! Third is the principle of prioritization based on urgency. Urgency‐based prioritization aims to reduce wait‐list mortality and dropout. Although this is a worthwhile goal, it completely ignores posttransplantation survival. This is critically important in patients with HCC because the very same patients who have the highest urgency (large tumors, high serum alpha‐fetoprotein [AFP]) may also have a significantly reduced posttransplantation survival due to the increased likelihood of tumor recurrence after transplantation. A better metric is prioritization based on transplant‐related survival benefit, which is the difference between posttransplantation and wait‐list life expectancy.9 Patients with the highest survival benefit are those with the highest urgency (ie, the highest wait‐list mortality or dropout rate corresponding to low wait‐list life expectancy) who also have a moderate or high posttransplant life expectancy. Maximizing the survival benefit avoids the trap of prioritizing patients with very high urgency who also have unacceptably low posttransplant survival (eg, very large tumors, very high AFP) or conversely of prioritizing patients who have high posttransplant survival but low urgency (eg, low natural MELD, tumor burden, and AFP), which is routinely done under our current system.9 In summary, real improvement in the prioritization of HCC patients for liver transplantation can occur, but it will require a radical rethinking of current policies. Prioritization scores in patients with HCC have to be a function of the natural MELD score primarily together with a smaller contribution from serum AFP level and tumor burden, until better, readily available markers of tumor biology and microvascular invasion are identified. Difficulties do exist, such as how locoregional treatments and the response to such treatments should affect prioritization scores, but these difficulties can be overcome. Prioritization scores should be modeled to predict survival benefit rather than wait‐list outcomes or, at the very least, to predict wait‐list outcomes (ie, urgency) with provisions for excluding patients with unacceptably low posttransplant survival. Although scores that predict wait‐list outcomes or survival benefit are far from perfect, they are clearly superior to the current practice of assigning an arbitrary score that has no association with wait‐list outcomes or survival benefit. The potential benefits of improving our prioritization strategies for HCC patients are enormous: studies have estimated that very little survival benefit has been achieved by transplanting thousands of patients with HCC since 2002 due to the current policies that prioritize HCC patients arbitrarily with no regard to their urgency or, more importantly, their survival benefit.9 Introducing improved prioritization policies will significantly improve the overall survival of all patients in need of transplantation in the United States.