We read with interest the article by Watt et al. about the impact of sirolimus (SRL)-based immunosuppression on patient and graft survival among recipients of liver transplantation (LT) with hepatitis C virus (HCV), and we commend their efforts in trying to shed some light on the mechanisms linking immunosuppression to the risk of HCV-related graft failure. However, we are convinced that the results of their analysis do not provide evidence against the use of SRL and mammalian target of rapamycin (mTOR) inhibitors in general in HCV-positive LT recipients.
Despite Watt et al.'s attempt, capturing the complexity of posttransplant immunosuppression is a challenging task that cannot be remedied by the use of propensity score matching analysis alone. Concerns about the use of propensity score matching have been raised recently: hidden bias may increase because the matching of patient samples by observed variables (ie, HCV versus non-HCV) may generate further bias on account of latent, unobserved, and/or underestimated confounders. Because of the Food and Drug Administration's black-box warning for SRL in LT, its use may deliberately have been reserved for patients with poorer prognoses and/or multiple risk factors beyond HCV infection, such as renal failure, expanded criteria donors, hepatocellular carcinoma (eg, grade, stage, and microvascular invasion), de novo and recurrent malignancies, and cardiovascular/metabolic complications at any time during the posttransplant period. Furthermore, information on SRL trough levels, overall patient exposure, and the mode of calcineurin inhibitor withdrawal is missing in Watt et al.'s analysis, as are histological data on the severity of HCV recurrence.
Evidence against the dismissal of mTOR inhibitors in HCV-positive LT recipients is accumulating in the international literature. A recent retrospective analysis of HCV patients on SRL showed significantly less advanced fibrosis (stage ≥ 2) in comparison with an HCV control group at 1 and 2 years (30.1% versus 50.5%, P = 0.001), with no difference in patient survival at 1 year (92.5% versus 87.9%, P = 0.15) and 5 years. A recent prospective randomized controlled trial of using everolimus (EVR) to minimize tacrolimus (TAC) in de novo LT recipients has shown that the use of EVR is associated with fewer and less severe acute rejection episodes, similar patient and graft survival, and superior renal function at 12 months in comparison with standard-exposure TAC. HCV-positive patients on EVR fared as the overall population did with respect to efficacy and renal function, and there were comparable changes in the viral load from the time of randomization to month 12 in comparison with the TAC group. EVR with minimized TAC was also associated with numerically (but not statistically) more stable liver fibrosis and comparable necroinflammatory scores.[4, 5] Altogether, these results challenge Watt et al.'s conclusions and argue for cautiously interpreting data derived from large databases and for addressing the need for more properly designed longitudinal studies before the use of mTOR inhibitors in HCV-positive patients is dismissed. We believe that transplant physicians have acquired experience with mTOR inhibitors over the last years and that the results of Watt et al.'s study reflect a learning-curve effect and an initial misuse of mTOR inhibitors in the patients with poorest outcomes.
Paolo De Simone, M.D.1
Faouzi Saliba, M.D., Ph.D.2
Lutz Fischer, M.D.3
1Hepato-Biliary Surgery and Liver Transplantation
University of Pisa Medical School Hospital
Paul Brousse Hospital (Public Hospital System of Paris)
University of Paris-Sud
3Department of Hepatobiliary and Transplant Surgery
University Medical Center