Evaluation of early liver graft performance by the indocyanine green plasma disappearance rate

Abstract

A recent issue of Liver Transplantation presented 2 novel studies,1, 2 both of which applied the indocyanine green plasma disappearance rate (ICG-PDR) to evaluate early graft performance after deceased donor liver transplantation (LT). We read both articles with particular interest as our group has been intensively working in the same field. Olmedilla et al.2 performed a large single-center study with 172 liver recipients and compared posttransplant ICG-PDR values with the graft function score (grades I-IV) developed by Greig et al.3 However, Greig et al.'s classification is only one exemplary score among various others and had already been developed in the period of 1985-1988.3 Thereafter, more scores, using different specific patterns of biochemical and clinical parameters of the early posttransplant period, were developed for the same purpose. It has been shown that different scores provide conflicting results.4 As a result, the degree of association between ICG-PDR and graft function could be dependent on the particular score applied. Moreover, such scores provide only a surrogate parameter of graft function and are not necessarily connected to individual complications and outcomes. Therefore, from our perspective, graft function scores should not be applied as references or gold standards for diagnostic studies after LT. We believe that any diagnostic test providing useful information should be evaluated by outcome analysis. In the second study, Levesque et al.1 daily determined ICG-PDR in 72 liver recipients during the initial 5 days after LT. This study was undertaken to investigate the predictive value of ICG-PDR for early postoperative complications. A composite endpoint was chosen, and it included different posttransplant events such as episodes of acute cellular rejection, arterial or portal complications, and hemorrhagic or septic shock. However, this strategy has a certain limitation because it is difficult to draw any specific or precise conclusions for the therapy of patients in the early posttransplantation period. In addition, the post hoc choice of the lowest individual readouts during 6 days for comparison with clinical outcomes is prone to overestimating the discriminative power of ICG-PDR. Furthermore, patients with an episode of acute cellular rejection revealed a significant and exclusive decrease in ICG-PDR readouts during follow-up.1 This is an interesting and potentially valuable finding from our perspective, but declining ICG-PDR cannot be interpreted as a specific parameter of graft rejection. This might also be the case for any vascular complications, any form of bile duct obstruction, or recurrent hepatic disease (eg, hepatitis C virus reinfection). Therefore, further studies with ICG-PDR should report the progression of patients with those complications separately, and this will probably require multicenter studies with larger patient populations. We find it extremely remarkable that ICG-PDR readouts remained static in the group without any complications after LT. We have had the same experience in one of our clinical studies, and this is contrary to the maximal metabolic liver function capacity measured by the LiMAx test.5 Interestingly, living donor liver grafts also have shown full performance with the indocyanine green (ICG) test directly after LT.6 In contrast to these findings, a significant effect of ischemia/reperfusion on ICG-PDR readouts has been demonstrated in previous experimental studies, and a posttransplant regeneration of test readouts could be expected. It somehow remains unclear why there is no progression or increase in ICG-PDR readouts in humans. There are 2 potential explanations: first, the posttransplant hemodynamic situation could have an important influence, or second, ICG excretion is not significantly affected by ischemia/reperfusion in human livers. It is well known that ICG-PDR is dependent on liver perfusion, sinusoidal uptake, adenosine triphosphate-dependent excretion into biliary canaliculi, and unrestricted biliary drainage.7 However, these confounding factors should not be forgotten during the interpretation of individual tests in clinical practice. Poor ICG-PDR readouts could indicate biliary obstruction or liver malperfusion before parameters of hepatic injury or synthesis become significant for diagnosis. Thus, future studies might investigate the effect of biliary obstruction and compare ICG-PDR readouts with hemodynamic cofactors such as blood pressure, cardiac output, and the application of vasoactive drugs (eg, epinephrine or iloprost). Altogether, in view of our current knowledge, it is hard to accept that ICG-PDR with all its limitations can predict an individual patient's outcome in the early phase after LT. However, this would be a prerequisite for clinical decisions based on a diagnostic test. Finally, we would like to encourage the application of new diagnostic research standards, such as the Standards for Reporting Diagnostic Accuracy initiative,8 which may enhance the effective evaluation and further acceptance of novel diagnostic tools in medicine. Martin Stockmann*, Johan Friso Lock*, Maciej Malinowski*, Peter Neuhaus*, * Department of General, Visceral, and Transplantation Surgery, Charité-Universitätsmedizin Berlin, Berlin, Germany.