Retransplantation for Early Liver Graft Failure: A Call for Standardization of Concepts

The issue of liver retransplantation (re-LT) is complex. From an individual perspective, re-LT is always more difficult than the first transplant, both in terms of the surgery itself and the perioperative period, and the results are less favorable. From a collective point of view, using 2 grafts for 1 patient remains an ethical problem in the context of organ shortage. Defining criteria beyond which re-LT is futile, as well as identifying grafts that will not allow survival after transplantation, are essential elements. In this issue of Transplantation, Robin et al1 developed a decision-support score (Early Liver Retransplantation Score) to assess the risk of failure in re-LT for early liver graft failure. Using the French national database (CRISTAL registry), they assessed the risk of graft loss (defined by death or re-LT) in 477 patients who underwent re-LT within 90 d of a first LT between 2007 and 2022. In this cohort, the 1-y posttransplant graft survival after re-LT was only 64.3% (95% confidence interval, 59.9%-68.5%), highlighting the need to enhance decision-support tools to avoid significant loss of liver grafts in a context of graft shortage. The Early Liver Retransplantation Score was developed on the basis of a multivariable model that included 8 recipient-related or donor-related variables. This score appears to provide a good correlation between observed and predicted 1-y posttransplant graft loss (r = 0.8), enabling patients to be classified into low-, medium-, and high-risk categories for graft loss. These data are very interesting and fundamental to improving the fairness of organ allocation. However, they also raise issues of concept definition, which we believe should be standardized in the future. The concept of early liver graft failure requires clarification and standardization in several respects. First, the pathophysiology of the condition must be clarified. In the present study, the 2 main causes of liver graft failure were primary nonfunction (PNF; 43%) and hepatic artery thrombosis (HAT; 38%). Re-LT for early HAT may differ from re-LT for PNF, and lumping them into the same category of “re-LT for early graft failure” is highly questionable. The clinical presentation, severity, and urgency of re-LT are usually different, and as the study by Robin et al showed, the prognosis is also very different. The outcome after re-LT for HAT is usually good, in contrast to re-LT for PNF. Unfortunately, the term “liver graft failure” has never been precisely defined, so it may be justified to associate different pathophysiological mechanisms under the same term. Early liver graft failure is usually described as a continuum from early allograft dysfunction (EAD) to early nonfunction2,3 although definitions of EAD and nonfunction per se are not unanimous. Diagnostic criteria for EAD have been proposed, notably by Olthoff et al4 and Agopian et al.5 However, none of these criteria have been proven to be sufficiently accurate, and they are often difficult and complex to use. Similarly, the definition of nonfunction is unclear in the literature. Under the term “PNF,” some include all causes of graft nonfunction, whereas others exclude secondary causes such as sepsis, early vascular, or biliary complications and reserve the term PNF for nonfunction related to ischemia/reperfusion injury.2,6 For some, PNF is defined as graft failure requiring re-LT or leading to death, whereas others have proposed to use of biological criteria such as aspartate aminotransferase, international normalized ratio, lactate concentration, and blood pH.6 Scores to define PNF have also been developed.2,6,7 Depending on the definitions used, patients retransplanted for HAT would or would not be included in studies such as that by Robin et al and patients with nonfunction could be identified as either PNF or secondary nonfunction. Second, the concepts of early allograft failure, EAD, and PNF need to be standardized in terms of timing, as there is currently no consensus on how to differentiate between early and late liver graft failure. The timeframes usually used to define EAD or PNF vary in the literature from 7 to 30 d.2,6 In the study by Robin et al, early allograft failure was defined as the need for re-LT within 90 d post-LT. As for pathophysiology, re-LT within the first week or at 3 mo is not equivalent (both in terms of cause and outcome), and their association within the same cohort induces heterogeneity and may be questionable. In fact, in their study, >60% of patients were retransplanted within the first 8 d, and time to re-LT was a prognostic factor in multivariate analysis. This highlights the importance of considering timing in defining these concepts. In France, for example, patients relisted within the first 7 d after LT are prioritized in case of PNF or HAT. After 7 d, they are listed on the basis of the Model for End-Stage Liver Disease score, which can impact the outcome. Considering re-LT up to day 90 is justified to encompass all “early” complications after LT. However, it runs the risk of mixing patients with very different mechanisms and reasons for re-LT who would not belong to the same diagnostic framework. Finally, the standardization of these concepts is essential for future research. A major issue for studies on re-LT is that they are inherently outcome-based (including re-LT or deceased patients), meaning they can only be conducted retrospectively. This limits their potential for informing clinical decision-making. By defining and standardizing these concepts in terms of pathophysiology and timing, it would be possible to include patients prospectively in studies according to diagnosis, to assess the benefits of re-LT. In conclusion, further research into early re-LT is absolutely essential to decrease the risk of liver graft loss and improve graft allocation from a distributive justice perspective. To this end, it is crucial to standardize the concepts of EAD, early graft failure, and PNF in terms of pathophysiology and timing to improve the quality of research in this field.