The advancement of liver transplantation (LT) from a novel procedure with poor outcomes to a nearly routine operation has been one of the great biomedical achievements and has required advances in medical management, patient selection, and surgical techniques. Although LT was first performed in 1963, none of the early recipients survived more than 3 weeks. Indeed, the results were viewed as so poor that a worldwide moratorium on further LT was enacted for more than 3 years. In the current era, overall 3-year patient survival following LT in the United States exceeds 80%, with a 10-year survival rate of approximately 50%. Among the many factors contributing to the success of LT, the development of effective immunosuppressive agents—primarily calcineurin inhibitors (CNIs)—has been one of the most important. Currently, more than 95% of LT recipients receive immunosuppression with a CNI-based protocol. The decline in the frequency of chronic rejection, with currently less than 4% of livers lost to rejection, has been associated with sequential overall improvements in 3-year patient survival. The most common causes of death beyond the first postoperative year among LT recipients are hepatic in nature (dominated by the recurrence of hepatitis C virus), malignancies, cardiovascular events, infections, and renal failure. Renal-related deaths, which are strongly attributable to exposure to CNIs, increase more than deaths attributable to any other cause after the fifth postoperative year. Furthermore, renal failure is the strongest predictor of late post-LT mortality.[2, 3] The impact and importance of posttransplant renal insufficiency are increasing as the proportion of recipients with poor renal function at the time of transplantation increases in step with the Model for End-Stage Liver Disease score needed by recipients to undergo LT. The widely recognized rise in the importance of kidney function has led to a transition in the focus of immunosuppression development from improving efficacy to improving tolerability (chiefly renal). Enter sirolimus (SRL).
SRL (rapamycin), a macrolide antibiotic, was discovered by Brazilian researchers as a product of the bacterium Streptomyces hygroscopicus on Rapa Nui (Easter Island). The soil sample from which the Rapa Nui Streptomyces was cultured was harvested in January 1965 near Rano Kau, where a plaque was subsequently placed that commemorated the unlikely contribution of Rapa Nui to modern medicine. SRL was approved by the Food and Drug Administration (FDA) in September 1999, some 34 years after the original isolation of SRL from S. hygroscopicus; this provides an inkling of the imagination, foresight, patience, and investment that can be required to translate an idea into a pharmaceutical product. Although SRL has been studied in many contexts and used in others (eg, drug-eluting endovascular stents), the sole FDA-approved indication for oral SRL is “the prophylaxis of organ rejection in patients aged ≥13 years receiving renal transplants.”
A pivotal phase 3 study of SRL in LT ended badly: a higher incidence of graft loss, death, and hepatic artery thrombosis (HAT) with SRL was observed, and this led to the premature termination of the study and a black-box warning from the FDA. Since the early worldwide moratorium on LT, the nature of many of those in the clinical trenches of LT has been to make note of bad results and press onward. Thus, although SRL has been used only in exceptional situations in the great majority of centers, others have considered its use in LT recipients almost de rigueur, often on the basis of a presumed renal-sparing effect. This is despite an FDA label for SRL that, in addition to the black-box warning, includes a specific precaution for patients with renal insufficiency: “Renal function should be closely monitored during the co-administration of sirolimus with cyclosporine, because long-term administration of the combination has been associated with deterioration of renal function.” Enough reports of studies of the renal-sparing effect of SRL were generated to facilitate a meta-analysis of the published results of off-label use of SRL in LT recipients. This concluded that the conversion from CNIs to SRL resulted in a “signal” for improvement in renal function among LT recipients with CNI nephrotoxicity. Scarcely an LT conference goes by without a debate about the relative efficacy and toxicity of SRL, which frequently concludes with “the results of the Spare-the-Nephron study are eagerly awaited.” This brings us to the Spare-the-Nephron study by Teperman et al. published in the July issue of Liver Transplantation.
Teperman et al. conducted a heroic prospective, open-label, multicenter study in which LT recipients maintained on mycophenolate mofetil (MMF) and a CNI were randomized 4 to 12 weeks after transplantation to MMF and SRL (n = 148) or to continue on MMF and a CNI (n = 145). One of the primary endpoints was met: the MMF/SRL combination was associated with significantly greater renal function improvement from the baseline with a mean percentage change in the glomerular filtration rate of 19.7 ± 40.6 for the MMF/SRL group versus 1.2 ± 39.9 for the MMF/CNI group (P = 0.002). The rates for the composite endpoint of biopsy-proven acute rejection, graft loss, death, and loss to follow-up at 12 months were similar in the 2 groups: 16.4% for MMF/SRL and 15.4% for MMF/CNI. The incidence of biopsy-proven acute rejection was significantly greater with MMF/SRL (12.2%) versus MMF/CNI (4.1%, P = 0.02). Graft loss (including death) occurred in 3.4% of the MMF/SRL-treated patients and in 8.3% of the MMF/CNI-treated patients (P = 0.04).
So, is this a late vindication of SRL? Were we wrong to doubt its safety and efficacy? The answers merit some consideration. The first important point to consider is that the higher incidence of graft loss, death, and HAT with SRL use in LT that led to the black-box warning was observed in the context of an earlier initiation of SRL (day 1 after LT). In the Spare-the-Nephron study, SRL was initiated 4 to 12 weeks after transplantation, at which time the great majority of wound healing had occurred; this made HAT, the most catastrophic of SRL complications, theoretically much less likely. The frequency of HAT was reassuringly low in both arms: 2 cases of HAT in the SRL arm and 1 case in the CNI arm (a 100% higher frequency with the same percentage difference–reporting metric used to report the renal-sparing effect of SRL in the Spare-the-Nephron study), and the overall rate of death and graft loss (the most important components of the composite endpoint) was numerically substantially lower in the SRL arm (3.4% versus 8.3%). Perhaps the most remarkable aspect of this study was that 116 patients were able to eliminate CNIs altogether between the 4th and 12th postoperative weeks, with only 12 (10.3%) developing biopsy-proven acute cellular rejection. At several levels, the results of the Spare-the-Nephron study are unequivocally something of a vindication of SRL as a renal-sparing tool. The finding that CNIs can be discontinued so early and with so little in the way of important adverse events is likely to be a landmark observation that affects future trial design. The adverse events that were statistically more common with SRL versus CNIs—leukopenia, pyrexia, edema, rash, hyperlipidemia, and mouth ulcers—are known side effects of SRL therapy. There are some important nuances to the Spare-the-Nephron study. The trough levels of CNIs were higher than those used in many centers currently, but not by much. In addition, the duration of follow-up was relatively short (12 months). SRL undoubtedly causes proteinuria (as it did in Spare-the-Nephron study). The effects, basis, and subsequent impact of proteinuria on renal function cannot be determined within 12 months. Insight into the longer impact of SRL conversion is to be found in a recent, much larger (n = 607) randomized controlled trial by Abdelmalek et al., who similarly evaluated the conversion from CNI-based immunosuppression to SRL-based immunosuppression for the preservation of renal function in LT patients. Their study converted patients later than the Spare-the-Nephron study (eligible patients were 6-144 months past LT), and it followed participants for longer (up to 6 years). The mean glomerular filtration rate in the SRL arm of the study by Abdelmalek et al. was initially superior to the rate in the CNI arm. By the 24th month of follow-up, this relationship was reversed, with the mean glomerular filtration rate numerically superior in the CNI arm. The large sample size and the follow-up duration of the study by Abdelmalek et al. suggest that the findings are probably more robust than those of the Spare-the-Nephron study, and they certainly highlight the need for longer follow-up in renal function studies that include mammalian target of rapamycin inhibitors. Finally, the large randomized controlled trial that led to recent FDA approval of an alternative mammalian target of rapamycin inhibitor, everolimus, for use in LT recipients (without a black-box warning) should make the use of SRL in LT recipients the rarest of events. Everolimus, though sharing a mechanism of action with SRL (binding to FK-binding protein 12 and exclusively inhibiting mammalian target of rapamycin complex 1), is quite distinct from SRL in ways that are likely to account for its superior tolerability. The half-life of SRL (60 hours) is approximately twice that of everolimus (28 hours), and this may be critical in delaying wound healing. Everolimus is also more soluble and bioavailable than SRL, with a 2.1-fold weaker binding affinity for FK-binding protein 12.
The FDA approval of everolimus means that the answer to the question of SRL as a CNI- and renal-sparing agent is neither “all good things come to those who wait” nor “spare the nephron, spoil the patient.” It is more a matter of too little, too late. The original inhabitants of Rapa Nui can scarcely have imagined that the soil failing to support their civilization would be so fertile for ours.
Michael R. Charlton
Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, MN