As graft and patient survival after liver transplantation (LT) continue to improve, chronic medical complications from long-term immunosuppression have grown to dominate the landscape of posttransplant care. Chronic renal dysfunction, in particular, has become a mounting burden, with as many as 18% of LT recipients developing chronic kidney disease by the fifth postoperative year.1 Many of these patients will deteriorate into end-stage renal disease, an endpoint clearly shown to increase morbidity and shorten patient survival.2 Though often multifactorial, the chronic decline in renal function in LT recipients is, to a great extent, driven by calcineurin inhibitors (CNIs), the current standard in immunosuppression. CNIs result in renal vasoconstriction, which, over time, can progress to tubulointerstitial fibrosis and unrecoverable renal injury.3 Strategies that minimize CNI use have thus been sought to ameliorate the commonly observed decline in kidney function that occurs after LT.

Sirolimus (SRL), a mammalian target of rapamycin inhibitor with potent antiproliferative and immunosuppressive properties, has been touted within the solid organ transplant community for its freedom from significant renal toxicity or neurotoxicity.4 Although SRL is generally avoided in the immediate postoperative period because of reported increased rates of hepatic artery thrombosis and impaired wound healing,5 as many as 14% of liver transplant recipients receive SRL during the maintenance phase of immunosuppression, generally for either the prevention or amelioration of CNI-related complications.6 Multiple studies have suggested a renal sparing effect in patients with CNI-induced nephrotoxicity converted to either SRL alone or SRL used in conjunction with a low-dose CNI regimen.7-13 By and large, these have been small, uncontrolled, single-center reports. As such, the efficacy of SRL with respect to the endpoint of renal preservation has been at once suggestive but inconclusive. More recent data suggest that the reported advantage of SRL-based immunosuppression on long-term kidney function may be overstated. This journal recently published the results of an open-label, randomized controlled trial of conventional immunosuppression versus SRL-based immunosuppression in patients with post-LT, CNI-related nephrotoxicity. In patients receiving SRL, a significant, if modest, improvement in the change in the glomerular filtration rate (GFR) was shown at 1 year from the time of conversion, but there was no significant difference in absolute GFR.14 Shenoy and colleagues,15 in a prospective comparison of SRL conversion with unmodified, conventional immunosuppression in a similar population, showed an early, significant improvement in measured creatinine clearance that, by 1 year of follow-up, was not maintained. A case-control study from Campbell and colleagues16 examined the effect of SRL-based immunosuppression in LT recipients with relatively preserved renal function. Patients switched to SRL were no less likely to develop renal dysfunction than the CNI immunosuppression control group after a median follow-up of 1 year. Subgroup analysis according to the duration of SRL exposure and interval from the time of LT to the initiation of SRL did not substantially alter this outcome.

The evolving literature on SRL in LT mirrors closely the emerging data in kidney transplantation, in which CNI conversion to SRL-based regimens has provided similarly mixed results with respect to renal preservation. A large, multicenter, randomized controlled study demonstrated that withdrawal of cyclosporine A at 3 months post–kidney transplantation (in patients receiving cyclosporine A/SRL/prednisone) led to improvements in GFR.17 These findings were supported by a number of smaller trials suggesting improved renal function but an increased risk of acute rejection following CNI conversion to SRL within 3 to 6 months of transplantation.18 At later periods post-transplantation, improvement in GFR has not been universally observed, however, in that the greater the renal dysfunction is at the time of transition, the poorer the outcome is with SRL-based regimens, particularly in patients with significant proteinuria at the time of transition.19, 20 The SRL CONVERT trial highlighted these findings, insofar as only those patients converted from CNIs to SRL with a preconversion GFR > 40 mL/minute and minimal proteinuria demonstrated benefit in GFR at 12 months.21

In this month's issue of Liver Transplantation, Dubay and colleagues22 provide the latest contribution to the growing literature on SRL conversion for CNI-related nephrotoxicity after LT. Fifty-seven LT recipients with CNI-related chronic kidney dysfunction received SRL a minimum of 90 days and a median of 45 months postoperatively. SRL was maintained for a median duration of 18 months. All but 5 of the SRL-treated patients were completely withdrawn from CNI therapy. Calculated creatinine clearance at 1 year of follow-up stabilized but did not improve. Comparison to a well-matched, CNI-reduction control arm showed no significant difference in creatinine clearance values at 1 year post-intervention or in rates of progression to renal replacement therapy, rejection, or death. When examined according to the change in creatinine clearance with respect to the time of SRL initiation, patients receiving SRL actually did worse than their CNI comparators. One-year creatinine clearance values remained statistically equivalent when stratified according to both the time interval between transplantation and the initiation of SRL and the calculated GFR at the time of medication conversion. The prevalence of side effects was significantly higher in patients receiving SRL, although these were tolerable in most patients. The authors conclude that with no statistical advantage and a higher rate of complications vis-à-vis the conventional strategy of CNI-dose reduction, SRL conversion cannot be recommended for the purpose of ameliorating CNI-induced nephrotoxicity.

The results of Dubay et al.'s experience,22 like those of other recent studies in this area, stand in contrast to what once seemed an uncontested tide of evidentiary support for SRL conversion for amelioration of CNI-induced renal toxicity after LT. Part and parcel of these disparate outcomes are differences in study methodology, in particular, Dubay and coworkers' use of a carefully selected control arm, an element missing from much of the SRL conversion literature until recently. This alone signifies the Dubay study as one of the better studies to date on SRL conversion for renal preservation. A number of caveats deserve consideration before their conclusions are accepted at face value, however. The sample size, though large by comparison to previous studies, is still fairly small, and this raises the question of the study's power to detect a significant difference between study and control arms with respect to GFR. Noteworthy is the trend favoring SRL in individuals with the highest creatinine clearance at baseline (45-60 mL/minute), the group with presumably the lowest CNI exposure and therefore most likely to recover renal function via conversion to SRL. Although this difference was nonsignificant, it raises the possibility that a statistical advantage might emerge in the setting of a larger scale trial. Also important to note is the calculation of GFR according to the creatinine-based equation of Cockcroft-Gault. Inasmuch as GFR estimation equations lack both precision and accuracy in liver transplant recipients, caution must be exercised in interpreting their results.23 This applies not only to the study by Dubay et al. but also to the literature on SRL conversion for renal preservation at large. Iohexol clearance methods, the currently accepted gold standard in GFR determination, would have been desirable but admittedly cumbersome to perform. Finally, unmeasured bias and confounding, inherent concerns in retrospective studies, particularly single-center ones, cannot be discounted.

These mitigating concerns aside, the conclusions of Dubay and colleagues22 add another important piece to the puzzle of SRL conversion for long-term renal preservation. Interpreted within the context of the accumulated literature on SRL conversion, the following conclusions can be made: Despite a strong physiologic rationale, at the present time, no convincing evidence exists that SRL improves long-term renal outcomes in LT recipients. If indeed there is an effect, it is likely small and apt to be seen early in the posttransplant period before the irreversible effects of CNI-mediated renal injury have taken hold. Ultimately, a large-scale, multicenter, randomized controlled trial is needed to clarify what role, if any, SRL may have in ameliorating CNI-induced renal toxicity post-LT.


  1. Top of page