Long-Term Renal Allograft Function on a Tacrolimus-Based, Pred-Free Maintenance Immunosuppression Comparing Sirolimus vs. MMF


  • Abstract presented at American Transplant Congress, May 2005, Seattle, WA.

* Corresponding author: Lorenzo Gallon, l-gallon@northwestern.edu


It is not known how different steroid-free immunosuppressive combinations affect renal graft survival and long-term kidney transplant function. Here we sought to compare the impact on graft survival and long-term graft function of two tacrolimus (Tac)-based, prednisone-free maintenance immunosuppressive protocols: Tac/Mycophenolate Mofetil (MMF) vs. Tac/Sirolimus (SRL). Renal transplant patients given induction therapy with IL2-RA and methylprednisolone on days 0, 1 and 2 post-transplant were prospectively randomized to two maintenance immunosuppressive regimens with Tac/MMF (n = 45) or Tac/SRL (n = 37). During the 3-year follow-up the following data were collected: patient survival, renal allograft survival, incidence of acute rejection and glomerular filtration rate (GFR) at different time-points post-transplant. Cumulative graft survival was significantly different in the two groups: one kidney loss in the Tac/MMF vs. six kidney losses in the Tac/SRL (log-rank test p = 0.04). GFR at different time-points post-transplant was consistently and statistically better in the Tac/MMF than in the Tac/SRL group. The slope of GFR decline per month was flatter in the Tac/MMF than in the Tac/SRL group. This study showed that renal graft survival and graft function were significantly lower in the combination of Tac/SRL than Tac/MMF.


Advancements in immunosuppression therapy and patient care have reduced the rate of acute rejection and 1-year graft survival in kidney transplantation. However, the long-term graft survival has not increased significantly (1,2). Multiple factors might account for the persistent loss of renal allografts. Chronic steroids as part of maintenance immunosuppression, by causing hyperlipidemia, diabetes and hypertension, can affect long-term graft function. To tackle this factor, our transplant center has been one of the pioneers in the development of steroid-free immunosuppressive protocols (3,4).

Calcineurin inhibitors (CIs), by inducing chronic nephrotoxicity, are also an important cause for kidney graft loss. The introduction of cyclosporine (CsA) in transplant medicine has definitely represented a milestone to improve 1-year rates of graft survival. Nevertheless, the negative impact of CsA on long-term renal function and eventually graft survival is universally recognized (5–12).

Tacrolimus (Tac) appears to be less nephrotoxic than CsA (13–15), which may explain why many centers have now adopted Tac-based immunosuppressive regimens. New potent immunosuppressive agents such as mycophenolate mofetil (MMF) and sirolimus (SRL) are now entering the clinical practice. Their combination with CIs is aimed to reduce CsA or Tac dose and ultimately the related nephrotoxic risk, while maximizing the anti-rejection effect of the immunosuppressive regimen.

A large study in kidney transplant recipients has shown a significant decrease of renal allograft survival in patients treated with SRL/CsA (16) as compared to those given the combination MMF/CsA, both in addition to steroids. Randomized, multi-center clinical trials of MMF and Tac have demonstrated the safety and effectiveness of this combination (17,18) in kidney transplant patients. However, whether this combined immunosuppressive regimen performs better than the combination SRL/Tac on long-term kidney graft function and survival, particularly in a steroid-free regimen, remains ill-defined.

The present study was, therefore, designed in kidney transplant recipients maintained on prednisone-free immunosuppression with the aim to: (a) prospectively compare two maintenance immunosuppressive protocols of Tac/SRL vs. Tac/MMF on renal allograft survival; (b) assess whether possible difference in the long-term efficacy of the two Tac-based regimens was related to difference in their anti-rejection effects and/or on renal function.

Materials and Methods

Study design and aim

This was a pilot, long-term, open-label, prospective randomized single-center trial. The primary aim of the study was to evaluate the impact on long-term graft function of two Tac-based prednisone-free maintenance immunosuppressive protocols (see below). The primary efficacy end-point was the 3-year patient and graft survival rates. This report summarized the results at 36 months post-transplant. The secondary end-points were: (i) incidence and severity (by histologic grade) of acute renal allograft rejection at 3 years post-transplant; (ii) graft function during the 3-year follow-up determined with the Modification of Diet in Renal Disease (MDRD) equation; (iii) percentage of patients still remaining off steroids at 3 years post-transplant; (iv) incidence and severity of hyperlipidemia and incidence of infections.

All patients provided written informed consent at the randomization, according to the Declaration of Helsinki Guidelines. The protocol was approved by the Institutional Review Board of Northwestern University, Chicago, IL. All patients were recruited at Northwestern University.

Patient recruitment

The recruitment period was 12 months. Since this was a non-inferiority confirmatory trial, the power calculations were based on an expected kidney survival rate in both groups of not less than 80%. Accordingly, it was suggested that recruitment of 47 patients to each group should give an 80% chance of detecting a significant result at one-sided p < 0.05. The sample size was planned by assuming equal exponential survival at the theoretical median follow-up interval (i.e. at 18 months), that is, a null hypothesis for non-inferiority of the experimental regimen of Tac/SRL vs. the standard one of Tac/MMF. This implied a minimum hazard ratio between the two exponential parameters of the expected cumulative survival functions of 7.46 (i.e. a risk ratio of 1.16). For this purpose, 94 patients in total had to be randomized. The patients included in the trial were transplanted between October 2000 and September 2001. All 94 patients were recruited and randomized to one of the two regimens right before the transplantation. One patient in the Tac/MMF and three patients in the Tac/SRL group did not provide informed consent for participation. Therefore, during the recruitment period, 46 patients were enrolled in the Tac/MMF group and 44 patients in the Tac/SRL group. At the end of the recruitment period, no more patients were included in this study. Patients were randomized 1:1 to the two immunosuppressive protocols, Tac/SRL or Tac/MMF (Figure 1). The inpatient nurse performed the randomization after informed consent was obtained.

Figure 1.

CONSORT flow chart of the trial.

One patient discontinued MMF within the first month post-transplant secondary to severe diarrhea, and he was not included in the analysis. Seven patients discontinued SRL within the first month post-transplant secondary to severe hyperlipidemia (n = 1) and wound healing complications (n = 6). The baseline analysis was conducted on 90 patients, while the final analysis was conducted on 45 patients in the Tac/MMF arm and 37 patients in the Tac/SRL arm.

Inclusion criteria:  The inclusion criteria were: all living- and deceased-donor kidney transplant recipients, age 30–70, to have been screened and transplanted at Northwestern University, Chicago.

Exclusion criteria:  The exclusion criteria were: pediatric recipients, patients receiving ABO-incompatible or a positive donor–recipient cross-match transplant; recipients of multi-organ transplants, of a kidney from a non-heart-beating donor, or other expanded criteria donor; pre-transplant fasting serum cholesterol level greater than 350 mg/dL, and patients with BMI greater than 35 kg/m2. Additional exclusion criteria included patients with known sensitivity to Tac, SRL or MMF. Patients who were pregnant or lactating and patients known to carry HIV were excluded from the study.

Immunosuppressive regimens

All patients received peri-operative intravenous corticosteroids therapy of methylprednisolone of 500 mg on day 0, 250 mg on day 1 and 125 mg on day 2. No further steroids were given (rapid steroid elimination). All patients received the anti-IL2 receptor antagonist Basiliximab at day 0 and day 2 at the dose of 20 mg i.v. Tac (Prograf) was started on post-operative day 1. Target 12-h trough levels for Tac were 8–10 ng/mL during the first 3 months, 7–9 ng/mL from 4 to 6 months post-transplant and 6–8 ng/mL thereafter. MMF (Cellcept) was started on post-operative day 1 with a dose of 1000 mg twice/day. MMF doses were adjusted as indicated for leukopenia.

SRL (Rapamune) was started on post-operative day 1 at 3 mg once/day. The target 24-h trough levels for SRL were 7–10 ng/mL during the whole post-transplant period. After transplantation, trough levels for Tac and SRL were monitored daily during the first week and twice weekly during the first month, once a week in months 2 to 3, and on a monthly basis thereafter.

All cytomegalovirus (CMV) seronegative and seropositive patients receiving a kidney from a CMV-positive donor were treated with CMV prophylaxis with valganciclovir 450 mg p.o. once/day for 6 months. Seronegative recipients receiving a kidney from a CMV-negative donor did not receive CMV prophylaxis. Prophylactic therapy for Pneumocystis carinii pneumonia was administered to all patients for up to a year post-transplant.

Diagnosis and treatment of acute rejections

All rejection episodes were biopsy-proven. All biopsies were graded using the Banff 97 classification. Acute cellular rejections were treated, based on severity, with methylprednisolone 500 mg i.v. for 3 days, followed by a week's course of prednisone taper or with an anti-lymphocyte antibody therapy (Thymoglobulin, Atgam) for a 14-day course. No specific immunosuppressive therapy or modification of the immunosuppressive regimen was offered to patients with biopsy-proven chronic allograft nephropathy. Patients who developed acute cellular rejection requiring a course of methylprednisolone or anti-lymphocyte antibody therapy were then maintained on chronic immunosuppressive maintenance of corticosteroids with prednisone at the dose of 5 mg once/day.

Renal allograft function measurement

Glomerular filtration rate (GFR) was calculated at different time-points post-transplant (6, 12, 24 and 36 months) using the abbreviated MDRD equation (19).

Efficacy end-points

Primary efficacy end-point of the study was: graft survival at 3 years post-transplant.

Secondary end-points were: incidence and severity—determined by histologic grade—of acute renal allograft rejection at 3 years post-transplant; graft function during the 3-year follow-up determined with the MDRD equation; percentage of patients still remaining off steroids at 3 years post-transplant.

Safety variables

Safety was assessed by monitoring adverse events as reported during the study follow-up.

In particular we monitored the incidence and severity of hyperlipidemia, infections and hematological abnormalities. Discontinuation of MMF or SRL therapy was also monitored as a safe parameter.

Statistical analysis

Continuous data are reported as mean ± SD, unless stated otherwise. Two sample t-tests or non-parametric Mann-Whitney U test were used to compare continuous variables (e.g. GFR, etc.) of the groups on the two immunosuppressive maintenance combinations (Tac/SRL vs. Tac/MMF).

The incidence of events during the follow-up was compared using the Pearson chi-square test or Fisher's exact test. Laboratory data between specific time-points were analyzed by the paired t-test or Wilcoxon rank-sum test as appropriate. Patient survival, graft survival and occurrence of acute rejection were determined using the Kaplan–Meier survival estimates, whereas the cumulative curves were compared by log-rank test. ANOVA and multiple parametric regression models were used to compare and predict GFR slopes. A p-value of 0.05 or less was considered significant.


Patient demographics and donor–recipient characteristics

Between October 2000 and September 2001, a total of 82 patients were analyzed in this prospective study. During this period, 37 patients were included and followed as randomized to the Tac/SRL maintenance combination, and 45 patients to the Tac/MMF maintenance combination. There were no significant differences between the two groups when comparing the recipient age, gender, ethnicity, incidence of delay graft function (DGF), sensitization, pre-existing diabetes mellitus, ratio of living donor vs. deceased donor kidneys (Table 1). The mean age of donors in the Tac/MMF group was significantly lower than that in the Tac/SRL group (33.2 ± 11.4 vs. 39.9 ± 12.9 years, p = 0.04). The mean number of donor–recipient mismatched HLA alleles was significantly higher in the Tac/MMF group than the mean of donor–recipient mismatches in the Tac/SRL group (3.6 ± 1.8 vs. 3.1 ± 1.9, p = 0.02).

Table 1.  Patient demographics and donor–recipient characteristics
  1. Tac = tacrolimus; SRL = sirolimus; MMF = mycophenolate mofetil; PRA = panel reactive antibody; DGF = delay graft function; HLA = human leukocyte antigen; DM = diabetes mellitus; Tx = transplantation.

Age (years)46.3 ± 12.642.3 ± 11.9ns
Gender (female)41%37%ns
 African American30%20%ns
PRA >25%7%0%ns
Prior Tx0%0%ns
Age donor39.9 ± 12.933.2 ± 11.40.04
DM (pre-existing)17%8%ns
Mean HLA-mismatch3.1 ± 1.93.6 ± 1.80.02
Donor source
 Deceased donor27%33%ns

Patient and graft survival

Three-year Kaplan–Meier patient survival rates were not significantly different between the Tac/SRL vs. Tac/MMF group (98% vs. 100%, p = 0.43, respectively). The cause of death in the Tac/SRL group was lung adenocarcinoma (n = 1).

The Kaplan–Meier graft survival rates at 3 years post-transplant were significantly different (84% graft survival in the Tac/SRL vs. 98% in the Tac/MMF group, p = 0.04, Figure 2). Specifically, in the Tac/SRL group there were six kidney losses vs. one kidney loss in the Tac/MMF group. The causes of graft loss in the Tac/SRL group included: chronic allograft nephropathy (n = 4), severe acute cellular rejection (n = 1), acute rejection in setting of medical non-compliance (n = 1) and death with functioning graft (n = 1). The cause of graft loss in the Tac/MMF group was severe acute rejection in setting of medical non-compliance (n = 1). Graft survival was still significantly lower in the Tac/SRL group after normalization for the different donor age in the two groups.

Figure 2.

Cumulative renal allograft survival in Tac/MMF vs. Tac/SRL at 3 years post-transplant. The Kaplan–Meier curves are different by the log-rank test (p = 0.0439).

Drug monitoring and incidence of acute rejection

Dosing and blood trough levels of Tac during the 3 years post-transplant follow-up were similar in both groups (Table 2).

Table 2.  Tac trough levels and adverse events during 3 years post-Tx follow-up
  1. Tac = tacrolimus; SRL = sirolimus; MMF = mycophenolate mofetil; UTI = urinary tract infections; CMV = cytomegalovirus; DM = diabetes mellitus; WBC = white blood cell count; Tx = transplantation.

Tac levels:
 3 months post-Tx8.2 ± 2.08.3 ± 1.6ns
 6 months post-Tx6.9 ± 1.57.8 ± 2.4ns
 24 months post-Tx7.0 ± 2.06.9 ± 1.7ns
 36 months post-Tx6.4 ± 1.66.9 ± 1.3ns
Anemia (hemoglobin <10 g/dL)16%17%ns
Leukopenia (WBC <2000/μL)5%13% 
Hyperlipidemia (total cholesterol >200 mg/dL)53%35%0.0001
New onset DM5%3%ns

Eleven patients (29.7%) in the Tac/SRL group experienced acute cellular rejection compared to eight patients (17.7%) in the Tac/MMF group. This difference was not statistically significant (p = 0.31). Seventy percent of the acute rejections occurred early (<30 days post-transplant) in both groups. There was no statistically significant difference in the severity (using the Banff 97 classification) of the acute rejections between the two groups (Figure 3).

Figure 3.

Banff 97 classification of acute rejection episodes according to the study groups (p = 0.31).

Renal allograft function

GFR measured at different time-points post-transplantation (6, 12, 24 and 36 months), using the abbreviated MDRD equation, was consistently lower in the Tac/SRL than in Tac/MMF group (Figure 4). Even after excluding from the analysis all patients with history of acute cellular rejection, the graft function was significantly lower in the Tac/SRL versus Tac/MMF throughout the 3 years of the follow-up (Figure 5). The slope of GFR decline per month (mL/min/1.73 m2/month), even after adjustment for donor age and HLA mismatch, still favored the use of Tac/MMF as compared to Tac/SRL (−0.069 ± 0.73 mL/min/1.73 m2/month vs. −0.38 ± 1.22 mL/min/1.73 m2/month, p = 0.07), and could likely be predicted by the treatment combination (Figure 6).

Figure 4.

Time course of GFR calculated by MDRD equation in kidney transplant patients on Tac/SRL or Tac/MMF therapy.

Figure 5.

Time course of GFR calculated by MDRD equation in kidney transplant patients with no rejection on Tac/SRL or Tac/MMF therapy.

Figure 6.

Slope of GFR decline as predicted by the treatment (Tac/MMF vs. Tac/SRL) and adjusted by donor age and degree of mismatch (p = 0.07). Median (thick horizontal line) and interquartile range (white and gray areas, respectively) for each treatment group are indicated on the box plots.

Adverse events

There were no significant differences between the two groups in regards to anemia (hemoglobin <10 g/dL), infections including CMV, and new onset of DM (Table 2). There was a trend toward more frequent leukopenia (WBC < 2000/μL) in patients receiving Tac/MMF compared to patients receiving Tac/SRL (13% vs. 5%). The incidence of hyperlipidemia (total cholesterol >200 mg/dL) was significantly higher in the Tac/SRL group than in the Tac/MMF (53% vs. 35%, p = 0.0001).

Discontinuation or change of immunosuppressive medications

MMF dose changes occurred due to diarrhea in seven patients (15%) and to leukopenia in six patients (13%). MMF was discontinued in one patient (2%) of the Tac/MMF group. Reasons for MMF discontinuations included diarrhea and leukopenia.

SRL was discontinued in seven patients (16%) of the Tac/SRL group. The reasons for discontinuation were hyperlipidemia (n = 1) and wound healing complications (n = 6). At 3 years post-transplant, 86.49% of the patients in the Tac/SRL group were maintained off steroids as compared to 97.78% in the Tac/MMF group (p = 0.056). This difference was secondary to the higher rates of acute cellular rejection in the Tac/SRL group than in Tac/MMF requiring initiation of prednisone during the post-transplant period.


The present prospective randomized pilot study in kidney transplant recipients shows that the combination Tac/SRL in a steroid-free regimen is associated with lower 3-year graft survival and higher incidence of acute graft rejection than the combination Tac/MMF.

During the 3-year follow-up, the renal graft function was also worse in patients given Tac/SRL than Tac/MMF, even after excluding from the analysis those who had acute rejection episodes.

Of note, patients enrolled in this study were low-risk kidney transplant recipients, with comparable low percentages of deceased donors in both the study groups. Moreover, although in animal transplant models and more recently in humans, SRL has been shown to prolong the length of DGF by impairing the recovery of ischemic reperfusion injury (20,21), the incidence as well as the duration of DGF was not significantly different in patients on Tac/SRL versus Tac/MMF immunosuppressive therapy. Together, these findings indicate that the different long-term outcomes with the SRL- and MMF-based immunosuppression were not due to selection bias of patients or difference in early post-transplant events.

The optimal use of immunosuppressive agents when administered as combination therapies is dependent upon their pharmacokinetic characteristics and any potential interaction (22).

Evidence is available that SRL enhances the bioavailability of the calcineurin-inhibitor CsA, ultimately increasing its potential nephrotoxic effect (23). Less consistent are the findings of pharmacokinetic interaction between SRL and Tac, but in kidney transplant recipients data show no change (22) or mild reduction (24) in Tac blood levels, as compared to unchanged Tac levels in the presence of MMF (25). Consistent with these observations are our present findings of comparable Tac trough levels in patients given SRL or MMF during 3-year follow-up. This further suggests that the negative effect of SRL on graft survival and long-term kidney function that we found in our study was not attributable to the increased patient exposure to Tac eventually contributing to the chronic nephrotoxicity of this CI.

Early animal studies have shown that SRL caused renal tubular collapse, vacuolization in proximal tubules and nephrocalcinosis, even if this nephrotoxic effect was considerably less than that of CsA or Tac (26). Furthermore, in the same rat model, SRL given in combination with CsA synergized the nephrotoxic effect of the CI (27). In addition, the administration of SRL alone to rats has also been shown to increase the intra-renal production of TGF-β (28), a cytokine whose excessive synthesis has been linked with the development of fibrosis in renal allografts (29). Interestingly, a recent study in de novo kidney transplant patients randomized to maintenance immunosuppression, consisting of CsA/MMF/prednisone vs. SRL/MMF/Prednisone, showed at 2 years post-transplant a significant benefit on renal allograft function in the group of patients who were randomized to SRL compared to the patients randomized to CsA. Furthermore, SRL-treated patients have a decreased incidence of chronic allograft nephropathy compared to the CsA-treated patients (30). Unfortunately, human data, looking at drug-induced nephrotoxicity when CIs are used in combination with SRL, are lacking and most of the current knowledge is based on observations obtained from retrospective analysis and from short-term prospective studies, none of which have information regarding allograft biopsies.

At this point, it is speculative to say that in our patients the chronic nephrotoxic effect of Tac has been amplified with the concomitant use of SRL, contributing to the lower graft survival than in patients given Tac/MMF. This does not occur with MMF, which is free of any nephrotoxic effect and does not per se impair graft function (31).

A recent observation by Meier-Kriesche et al. from a large retrospective analysis of 44 915 adult renal transplant patients, in the Scientific Registry of Transplant Recipients from 2000 to 2004, indicated that the combination Tac/SRL is associated with significantly worse graft survival as compared to Tac/MMF in all subgroups of patients, but particularly in recipients of deceased donor kidneys and expanded criteria donor grafts (32). In another retrospective study of patients treated with Tac or CyA in combination with SRL, withdrawal of SRL and substitution with MMF led to improvement of graft function, suggesting a possible synergistic nephrotoxicity between SRL and CyA or Tac (33).

To date, the only randomized prospective study comparing SRL and MMF in combination with Tac in kidney transplant recipients has been reported by Mendez et al. (34). In this multi-center study, 185 patients were randomized to receive Tac/SRL plus corticosteroids and 176 patients to Tac/MMF plus corticosteroids. Despite a very short follow-up, 12-month post-transplant renal function was significantly lower in the SRL- than in MMF-based group. Our present study extends this initial observation to a longer follow-up, and documents for the first time, in a steroid-free regimen, that the long-term nephrotoxic effect of Tac can be augmented with the concomitant use of SRL.

The limitations of our study are that, despite being a prospective randomized study, this was just limited to a single center. However, our conclusions remain speculative in terms of the potential synergistic nephrotoxicity between SRL and Tac and only histologic finding of increased nephrotoxicity in the Tac/SRL group, when compared to the Tac/MMF group, would have given the final verdict. Our observations are indeed consistent with multiple animal studies and with recent data from large retrospective analysis. If potential bias was inserted during the comparison of the GRFs between the two groups, this would have in fact attenuated the difference since more graft losses and dropouts occurred in the Tac/SRL group compared to the Tac/MMF.

Overall, these results raise caution in adopting a combined immunosuppressive therapy with SRL and Tac in kidney transplant recipients, particularly in the perspective of improving long-term graft survival. On this line, a combination treatment with MMF and Tac, even in a prednisone-free protocol, would offer more benefits.


This study was partially sponsored by industry (Astellas-USA). Funding from the Division of Nephrology and Transplantation at Northwestern University also supported the protocol, including monitoring of the patients and gathering of the data. Final analysis and interpretation of the results were performed in collaboration with the Mario Negri Institute for Pharmacological Research, Bergamo, Italy.