Does MELD need an update?
Maybe, but take it slow
BY JULIE K. HEIMBACH, MD
Even though 2020 was another record year for organ donation in the United States, a truly remarkable feat considering the profound impact of COVID-19 on health care as well as the population at large, there remains a critical shortage of available liver allografts.1 Last year in the U.S., of the approximately 13,000 patients waiting for a liver transplant, just under 9,000 patients underwent liver transplantation from a deceased or a living donor, while 2,345 either died waiting on the list or were removed for being too sick, and the rest remained waiting.1 In a perfect system, we would transplant every wait-listed patient at a time that would provide them the greatest benefit with the least amount of distress. However, because of the shortage of available organs for transplantation, an allocation system to rank wait-listed candidates is required. Because organ transplantation relies on the incredible altruism of individuals and their family members who make this ultimate gift on their behalf, it is crucial both for donor families and for waiting recipients that organ allocation be transparent, as fair and equitable as possible, and compliant with federal law, which is currently determined by the “Final Rule” that states that organ allocation be based in order of urgency.2
Since February 2002, U.S. liver allocation policy has been based on MELD.3 Prior to that time, liver allocation was based in part on the Child-Turcot-Pugh classification of liver disease, which included subjective components (ascites and encephalopathy) that are difficult to measure, as well as increased priority based on admission to the intensive care unit, also subjective and open to interpretation or abuse. Most crucially, the system defaulted to length of waiting time with large numbers of patients in the same category, which led to higher death rates for patients whose disease progressed more quickly or who were referred very late in their disease course.
MELD relies on a simple set of laboratory values that are easily obtained at any clinical lab and are already being routinely monitored as part of standard care for patients with end-stage liver disease.3 MELD initially required just three variables (bilirubin, creatinine, INR) and was updated to include just four variables with the adoption of MELD-Na in 2016, which added sodium levels. The MELD- and MELD-Na–based approach is a highly reliable, accurate way to rank patients who are most at risk of death in the next 3 months, with a C statistic of approximately 0.83-0.84.3,4 Perhaps the greatest testament to the strength of MELD is that, following the adoption of MELD-based liver allocation, MELD has gradually been adopted as the system of liver allocation by most countries around the world.
With the adoption of MELD and subsequently MELD-Na, which prioritize deceased donor liver allografts to the sickest patients first and is therefore compliant with the Final Rule, outcomes for patients waiting for liver transplant have steadily improved.3,4 In addition, MELD has provided an easily obtainable, objective measure to guide decisions about timing of liver transplant, especially in the setting of potential living donor liver transplantation. MELD is also predictive of outcome for patients undergoing nontransplant elective and emergent surgical procedures, and because of the ease in calculating the score, it allows for an objective comparison of patients with cirrhosis across a variety of clinical and research settings.
The MELD system has many additional strengths, though perhaps the most important is that it is adaptable. While the MELD score accurately predicts death from chronic liver disease, the MELD score is not able to predict mortality or risk of wait-list dropout due to disease progression from certain complications of chronic liver disease such as the development of HCC or hepatopulmonary syndrome, in which access to timely transplantation has been proven to be beneficial. This has required an adaption to the system whereby candidates with conditions, such as HCC, that meet specific criteria receive an assigned MELD score, rather than a calculated score. Determining which patients should qualify for MELD exceptions, as well as what the assigned priority score should be, has required careful analysis and ongoing revision. An additional issue for MELD, which was identified more than a decade ago and is overdue for adjustment, is the disparity in access to transplant for women who continue to experience a lower transplant rate (14.4% according to the most recent analysis) and approximately 8.6% higher death rate than men with the same MELD score.5 This is due, in part, to the use of creatinine in the MELD equation, which as a by-product of muscle metabolism, underestimates the degree of renal dysfunction in women and thus underestimates their risk of wait-list mortality.5 A potential modification to the MELD-Na score that corrects for this sex-based disparity is currently being studied by the OPTN Liver-Intestine committee, which is further evidence of the strength and adaptability of a MELD-based allocation system.
While it is tempting to conclude that a system that requires on-going monitoring and revision is best discarded in favor of a new model such as an artificial intelligence–based solution, policy development requires a tremendous amount of time for consensus-building, as well as effort to ensure that unexpected negative effects are not created. Whereas a novel system could be identified and determined to be superior down the road, the amount of effort and expense that would be needed to build consensus around such a new model should not be underestimated. Considering the challenges to health care and the population at large that are already occurring as we emerge from the COVID pandemic, as well as the short-term need to monitor the impact from the recent adoption of the acuity circle model which went live in February 2020 and allocates according to MELD but over a broader geographic area based on a circle around the donor hospital, building consensus around incremental changes to a MELD-based allocation system likely represents the best option in our continued quest for the optimal liver allocation system.
Julie K. Heimbach, MD, is a transplant surgeon and the surgical director of liver transplantation at Mayo Clinic in Rochester, Minn. She has no conflicts to report.
References
1. Organ Procurement and Transplantation Network data. Available at https://optn.transplant.hrsa.gov/data/view-data-reports/national-data. Accessed May 1, 2021.
2. Organ Procurement and Transplantation Network. Final rule. Available at https://optn.transplant.hrsa.gov/governance/about-the-optn/final-rule. Accessed May 1, 2021.
3. Wiesner R et al; United Network for Organ Sharing Liver Disease Severity Score Committee. Gastroenterology. 2003 Jan;124(1):91-6.
4. Nagai S et al. Gastroenterology. 2018 Nov;155(5):1451-62.e3.
5. Locke JE et al. JAMA Surg. 2020 Jul 1;155(7):e201129.
