Sirolimus Does Not Exhibit Nephrotoxicity Compared to Cyclosporine in Renal Transplant Recipients


* Corresponding author: José M. Morales,


Sirolimus and cyclosporine (CsA) prevent acute rejection in man when used as primary therapies in triple drug regimens. Sirolimus does not act via the calcineurin pathway and therefore is not expected to produce the same renal side-effects. This paper presents the pooled 2-year data analysis of renal function parameters from two open-label, randomized, multicenter studies. Patients (18–68 years) receiving a primary renal allograft were randomized to receive concentration-controlled sirolimus (n = 81) or CsA (n = 80), in combination with azathioprine and steroids (n = 83), or mycophenolate mofetil (MMF) and steroids (n = 78). From week 10 through year 2, calculated glomerular filtration rate (GFR) was significantly higher in sirolimus- than in CsA-treated patients (69.3 vs. 56.8 mL/min, at 2 years, p = 0.004). Serum uric acid was significantly higher in the CsA-treated patients and magnesium was significantly lower; these parameters were more likely to be within normal limits in the sirolimus group. Mean serum potassium and phosphorus were lower in sirolimus-treated patients. In conclusion, sirolimus, when administered as primary therapy in combination with azathioprine or MMF, has a favorable safety profile compared to CsA with regards to renal function.


The acute and chronic nephrotoxicities associated with the calcineurin inhibitors are suggested to be risk factors for chronic allograft nephropathy (1, 2). Acute cyclosporine (CsA) toxicity has been characterized by a decrease in glomerular filtration rate (GFR) and renal blood flow, hypomagnesemia and tubular injury (3). The biochemical mechanism of these hemodynamic changes is not well understood, but is linked in part to the inhibition of renal phosphatase calcineurin (4). Chronic CsA toxicity is a progressive state of renal dysfunction characterized by interstitial fibrosis, tubular atrophy and vascular changes such as arterial ischemia (5). Recent studies indicate that the molecular signal leading to progressive renal fibrosis in this chronic process may be related to the up-regulation of transforming growth factor β (TGF-β) (6).

Sirolimus (SRL, Rapamune®, rapamycin) is a macrocyclic lactone isolated from Streptomyces hygroscopicus. It is a potent immunosuppressive agent with a mechanism of action distinct from that of CsA and tacrolimus (FK 506), despite being structurally homologous to tacrolimus (7). Sirolimus does not inhibit calcineurin; thus it is anticipated that sirolimus will lack the acute nephrotoxic profile of the calcineurin inhibitors. In addition to its action on cells of the immune system, sirolimus inhibits growth-factor-induced proliferation of fibroblasts, endothelial cells, hepatocytes and smooth muscle cells (8–10).

For both of these unique features, non-calcineurin inhibitor immunosuppressive effect and suppression of growth-factor-induced smooth-muscle-cell proliferation and migration, sirolimus is considered as a good candidate to fulfill the need for a non-nephrotoxic immunosuppressive agent. Preliminary experimental data indicate that sirolimus shows less nephrotoxicity compared to CsA and tacrolimus. In a rat model, it has been demonstrated that sirolimus does not induce nephrotoxicity at doses 3 times higher than its effective immunosuppressive doses (11). In addition, infusion of therapeutic doses of sirolimus had no deleterious effects on renal function in pigs (12). In a more recent report, it was shown that only calcineurin inhibitors resulted in glomerular dysfunction in an acute experimental rat model of nephrotoxicity (4).

Phase I and II clinical trials have shown that the combination of sirolimus with CsA reduces the incidence of rejection episodes when compared to CsA alone, without deleterious effects on renal function (13, 14). In the larger phase trials that followed, however, mild renal impairment was observed when sirolimus 2 or 5 mg was added to standard doses of CsA (15, 16). Recently, two subsequent open-label, randomized phase II clinical trials, comparing the efficacy of sirolimus vs. CsA-based immunosuppression in renal transplantation were carried out (17, 18). These trials allowed for the first time the direct comparison of the effects of sirolimus and CsA on tubular function and glomerular filtration parameters in human renal transplant recipients. In the present report, we provide a pooled data analysis of the 2-year data of these two phase II studies.

Materials and Methods

Study design

Both studies were randomized, open-label, parallel-group trials conducted in 19 centers in Europe using triple-therapy with either CsA or sirolimus in combination with corticosteroids and azathioprine [Study A, (17)] or corticosteroids and mycophenolate mofetil [Study B, (18)]. Approvals were obtained from local ethics committees, and written informed consent was obtained from each patient enrolled. The studies were carried out in accordance with the Declaration of Helsinki.

Patient population

Adults receiving a first cadaveric renal graft were centrally randomized in a 1 : 1 ratio to receive CsA or sirolimus. In study A, the randomization was done within 24 h after transplantation, once adequate graft function had been established. In study B, patients were randomly assigned within 24–2 h before transplantation. A total of 161 renal transplant recipients were enrolled into the two studies. Of the 80 patients randomized to CsA, 54 completed 1 year of assigned treatment and 41 completed 2 years. Amongst the 81 patients in the sirolimus group, 44 completed 1 year and 32 completed 2 years. A complete description of the patient population is provided in Table 1.

Table 1. : Baseline characteristics and efficacy outcomes at 2 years
CharacteristicSirolimus (n = 81)Cyclosporine (n = 80)p-value
  • a anova with treatment as factor.

  • b

    Fisher's exact test.

  • c

    ATN is present in demography if occurring within 14 days of transplantation.

  • d

    On-therapy data at 24 months; Banff 1993: grade I, II or III.

  • e

    Intention-to-treat at 24 months. One patient in the sirolimus group was lost to follow-up; this was counted as a graft and patient loss.

Recipient age (years)
 Mean SD45.6 ± 1142.2 ± 11.60.064a
Recipient sex N(%)
 Male57 (70)52 (65)0.503b
Recipient ethnic origin N(%)
 White79 (98)72 (90)0.113b
 Black 1 (1) 1 (1) 
 Oriental 1 (1) 4 (5) 
 Other  3 (4) 
Donor age (years)
 Mean SD42.5 ± 14.640.2 ± 17.70.152a
Donor sex N(%)
 Male51 (63)52 (65)0.424b
Number of Human leukocyte
ntigen (HLA) matches
 0 7 (9) 7 (9)0.530b
 112 (15)13 (16) 
 221 (26)20 (25) 
 320 (25)26 (33) 
 414 (17) 7 (9) 
 5 7 (9) 5 (6) 
 6 2 (3) 
Cold ischemia time (hours)
 Mean SD21.1 ± 8.318.8 ± 6.80.060a
 Range 4–58 5–43 
ATN N(%)c
 Yes 9 (11) 5 (6)0.402b
Biopsy proven rejectiond28 (34.6)23 (28.7)0.50b
Graft survivale73 (90.1)71 (88.8)n.s.
Patient survivale76 (93.8)75 (93.8)n.s.

Immunosuppressive therapy

In both studies, control group patients received CsA (microemulsion, Novartis, Basel, Switzerland) with dosage adjusted to maintain whole-blood trough levels of 200–400 ng/mL for 2 months, and 100–200 ng/mL thereafter (Figure 1A). Sirolimus patients were given initial loading doses ranging from 16 to 24 mg/m2/day (17, 18). Sirolimus doses were then adjusted to achieve steady-state whole-blood trough levels of approximately 30 ng/mL for 2 months, and 15 ng/mL thereafter (Figure 1B) (19). Sirolimus oral solution (Wyeth-Ayerst Research, Philadelphia, PA, USA) was administered once daily in the morning, following dilution with water or orange juice. All patients received corticosteroids.

Figure 1.

Mean sirolimus whole blood trough concentrations (A , determined at a central laboratory by high-performance liquid chromatography with ultraviolet detection) and mean cyclosporine (CsA) whole blood trough concentrations (B , troughs determined at each investigational site by a monoclonal immunoassay).

In study A, patients were treated for up to 2 years with azathioprine. In study B, patients were treated with mycophenolate mofetil (MMF, Cellcept®), 1.0 g twice daily for up to 6 months. MMF was then to be discontinued by tapering the dose over 1 month. At the discretion of the investigator, the patient could then remain on double-therapy or be converted to azathioprine when MMF was discontinued.

Renal function measurements

In both studies, patients were followed daily during week 1, weekly through month 3, monthly through month 12 and every 3 months in the second year of treatment for a total of 40 scheduled visits. At each visit, renal function was monitored by serum creatinine, urea, sodium, potassium, chloride, bicarbonate (HCO3 or total CO2), calcium, magnesium and phosphorus. Glomerular filtration was calculated by the method of Nankivell (20). The frequencies of hypo- and hyper- calcemia, phosphatemia, uricemia, kalemia and magnesemia were calculated for each group, before and after 3 months of treatment, by determination of the number of patients with 3 or more laboratory values that were outside the normal limits.

Blood pressure

Supine or sitting blood pressure was taken at each study visit. Data were pooled for the purposes of analyzing systolic and diastolic blood pressure. Hypertension was reported spontaneously by investigators using his or her diagnostic criteria. It was considered treatment emergent if it was not reported previously or if the severity worsened.

Statistical methods

The statistical analyses of this study were based on pooled data from all study centers. Patient and graft survival are reported for all patients, whether or not they remained on assigned therapy. Acute rejection and renal function analyses were performed on data from patients on therapy; this information was not collected beyond 1 month following discontinuation. The use of the word ‘significant’ in connection with the results of a pairwise comparison refers to p-values ≤ 0.05 or ≤ 0.01. All tests of hypotheses were two-sided. Comparisons between treatment groups for laboratory data related to renal function (sodium, potassium, chloride, carbon dioxide, urea, creatinine, calcium, phosphorus, uric acid, magnesium, calculated creatinine clearance) were made using a one-way anova with treatment in the model. In this paper, the use of the expression ‘mean’ refers to any mean derived from an analysis of variance (anova).


Renal function

Glomerular filtration: Analyses of renal function included all patients on therapy, regardless of whether or not they had experienced an acute rejection episode. At 1 year, 10 of the 44 ongoing sirolimus-treated patients (22%) and 15 of the 54 ongoing CsA-treated patients (28%) had experienced a biopsy-confirmed acute rejection. At 2 years, 7 of the 32 ongoing sirolimus-treated patients (22%) and 12 of the 41 ongoing CsA-treated patients (29%) had experienced a biopsy-confirmed acute rejection.

As shown in Figure 2(A), mean serum creatinine levels were consistently lower in the sirolimus group compared to the CsA group as early as 8 weeks after transplantation. These differences were significant at 11, 12, 16, 20, 32 weeks and 1 and 2 years of treatment. From week 20 to 2 years of treatment, the mean values remained in the 120–130-μm range for sirolimus-treated patients, whereas for CsA-treated patients the mean value was approximately 140–150 μm. The calculated GFR was significantly higher in the sirolimus group from week 9 to week 32 and after 1 year of treatment (Figure 2B). At each of these visits, the mean GFR was approximately 10 mL/min better in sirolimus-treated patients. The mean serum urea concentration (Figure 2C) was significantly lower for sirolimus-treated patients starting from the second month of treatment. The mean urea concentration difference between the 2 groups was more than 2 mm after 1 year of treatment.

Figure 2.

Mean serum creatinine (A ), calculated glomerular filtration rate (GFR) (B ) and serum urea (C ), * p ≤ 0.05; **p ≤ 0.01.

In the sirolimus group, no statistically significant differences were observed between rejecters and non-rejecters (data not shown) in the mean creatinine concentration and calculated GFR. In the CsA group, however, the mean creatinine concentration tended to be higher and calculated GFR lower for the rejecters from week 2 to week 20.

Tubular function

Table 2 presents the mean uric acid, potassium, magnesium, phosphorus and calcium concentrations for all patients treated in the study. Table 3 presents the percentage of patients in each group with at least 3 laboratory values above or below the normal range during the first 12 weeks of treatment or from 12 to 52 weeks.

Table 2. : Mean uric acid, potassium, magnesium, phosphorus and calcium serum concentration at 4, 12, 24, 36, 52 and 104 weeks of treatment
  • a

    Standard error of the mean (SEM);

  • *

    p ≤ 0.05;

  • **

    p ≤ 0.01.

4 weeks0.360.27**4.413.90**0.700.83**0.790.742.362.23**
12 weeks0.440.30**4.133.56**0.750.87**0.960.77**2.452.38*
24 weeks0.440.35**4.003.76**0.760.83**0.980.90*2.462.42
36 weeks0.450.33**4.143.91*0.740.83****
52 weeks0.460.33**4.123.86**0.770.83**1.070.992.432.39
104 weeks0.480.32**4.153.82**0.790.831.110.97*2.442.40
Table 3. : Percentage of patients in each group with at least 3 out of range values
CsA group (%)SRL group (%)CsA group (%)SRL group (%)
  • *

    p < 0.05,

  • **

    p < 0.01.

  • Normal ranges in mM: calcium 2.20–2.56, phosphorus 0.8–1.6, uric acid (upper normal): 0.42, potassium 3.5–5, magnesium (lower normal): 0.8.

Hypocalcemia50.061.7 1.6 1.9
Hypercalcemia21.3 8.6*19.011.1
Hypophosphatemia60.865.4 7.926.4*
Hypokalemia15.053.8** 4.814.8
Hyperkalemia15.010.0 4.8

The mean uric acid serum concentration was significantly higher in the CsA group. After 3 months, hyperuricemia (> 0.42 mm) was 2.8-fold more frequent in CsA-treated patients. In contrast, average serum magnesium concentrations were significantly lower for the patients treated with CsA compared to the sirolimus group. The percentage of patients with hypomagnesemia (< 0.8 mm) was also significantly higher. The mean serum potassium concentration was significantly lower in the sirolimus group. During the first 3 months, 12 (15%) of the sirolimus patients received potassium supplements. From 3 to 6 months, six (10%) sirolimus patients were treated with potassium supplements; no patients required supplements after 9 months of treatment. One patient in the CsA group was treated with potassium supplements.

Serum phosphorus concentration was significantly lower for the patients treated in the sirolimus group. The number of patients experiencing hypophosphatemia was the same in the 2 groups during the first 3 months (61%, CsA vs. 65%, sirolimus) and was slightly higher in the sirolimus group compared to the CsA group after 3 months of treatment (26% vs. 8%, respectively, p ≤ 0.05). These data suggest that sirolimus may result in prolongation of the post-transplant hypophosphatemia usually observed.

Mean calcium concentration was lower for the sirolimus patients when compared to the CsA patients. These differences were small and the mean values of both groups remained well within the normal range (2.20–2.56 mm) during the 2-year treatment period. Hypocalcemia was observed in 62% of the patients in the sirolimus group and in 50% of the patients in the CsA group, almost exclusively during the first 3 months of treatment.

No significant differences in mean sodium, chloride or bicarbonate were observed between the sirolimus- and CsA-treated patients.

Blood pressure

Patients treated with sirolimus had a statistically lower incidence of hypertension than CsA-treated patients (30%, sirolimus vs. 48% CsA, p ≤ 0.05). At 2 years of treatment, sirolimus patients had lower mean systolic and diastolic blood pressures than CsA-treated patients, although the differences were not statistically significant. Mean systolic and diastolic blood pressure for sirolimus patients (n = 31) was 132.0 mmHg and 79.2 mmHg, respectively, whereas CsA patients (n = 34) had mean values of 137.8 mmHg and 82.5 mmHg.


Acute CsA nephrotoxicity is characterized by a decrease in glomerular filtration rate, reduced renal blood flow and hypomagnesemia (3). In these two studies, 81 patients treated with sirolimus based therapy were compared to 80 patients treated with CsA-based therapy. The mean serum creatinine, calculated GFR and urea were consistently and significantly better for sirolimus-treated patients than for CsA-treated patients. These results suggest that sirolimus does not exhibit the acute renal toxicity associated with CsA, most probably related to the renal vasoconstriction mediated by the inhibition of calcineurin. Indeed, hypertension was reported less frequently in sirolimus-treated patients. These results further support the absence of nephrotoxicity seen in an 8-week study in psoriasis patients (21). In that study, the serum creatinine levels of patients treated with CsA 5 mg/kg/day increased over time and were significantly higher than in patients receiving sirolimus monotherapy up to 3 mg/m2/day. Recent data indicate that renal function at 6 or 12 months is a predictor of chronic allograft nephropathy and long-term graft survival (22, 23). Therefore the significant difference observed in the creatinine values between the two groups may indicate a long-term beneficial effect on renal function of sirolimus-based therapy.

In contrast to CsA-treated patients, the mean uric acid, and magnesium concentrations were in the normal range for the sirolimus-treated patients. These results contrast with the findings of Andoh et al. (4) who reported that sirolimus, along with CsA and tacrolimus, causes significant hypomagnesemia in a rat model. The hypomagnesemia commonly observed in CsA-treated patients has been implicated in the pathophysiology of neurotoxicity and hypertension (24), and it is proposed to be an important factor in the pathogenesis of CsA nephrotoxicity (25). In the present study, significantly less tremor (data not shown) and hypertension were reported in sirolimus-treated patients compared to CsA-treated patients. It has been recently suggested that a lower incidence of arterial hypertension could have an important clinical implication in the prevention of graft failure (26).

More hypophosphatemia and hypokalemia were observed for the sirolimus patients. The majority of these events occurred during the first 3 months of treatment when the target trough concentration was at its highest (30 ng/mL). The post-3-month sirolimus trough decrease to 15 ng/mL correlates with the reduction in the frequency of low phosphorus and potassium values, and a normalization of the respective mean concentrations. The mechanism associated with this hypophosphatemia is still unclear, and additional studies are needed to determine if this is associated with a specific toxicity or simply to sirolimus patients having a better renal function. None of the sirolimus patients developed pathologies associated with hypophosphatemia or hypokalemia, and none of the patients discontinued treatment because of these laboratory changes. Patients responded well to potassium dietary supplements.

Given that renal function is increasingly recognized as a predictor of long-term graft survival (22, 23), and that sirolimus only results in improved kidney function when given without CsA, should a calcineurin inhibitor-free approach be pursued with sirolimus in renal transplantation? The first pilot study associating sirolimus with MMF and steroids produced satisfactory acute rejection results (18), although the acute rejection rates were somewhat higher than those observed when sirolimus was associated with CsA (15, 16). The acute rejection rate of the sirolimus-MMF-steroid combination has been dramatically improved by the addition of basiliximab induction (27). Early CsA withdrawal from a sirolimus-CsA-steroid regimen was also quite successful (28) and was the basis of the marketing approval for sirolimus in the European Union. Improvement in renal function using sirolimus in calcineurin inhibitor-free regimen is one of the promising new options in renal transplantation and should be pursued to establish whether this therapeutic approach reduces chronic allograft nephropathy and increases graft half-life.


Sirolimus administered as primary therapy, in combination with azathioprine or MMF, had a favorable safety profile compared to CsA as regards renal function. Filtration parameters (creatinine, urea and glomerular filtration rate) were better for the sirolimus-treated patients than for those patients treated with cyclosporine. Over the 2-year treatment period, tubular function parameters (magnesium and uric acid) were normal for the sirolimus-treated patients. Hypophosphatemia and hypokalemia observed for some of the patients appeared to be sirolimus dose-dependent and readily manageable with dietary supplements. These differences suggest that sirolimus does not exhibit the nephrotoxicity characteristic of CsA, and may therefore offer an important alternative to calcineurin pathway drugs to prevent chronic allograft nephropathy.


The authors would like to thank M. Gioud-Paquet, V. Haudiquet, R. Taylor and K. Rutault (Wyeth-Ayerst-Research, Paris, France), for their contribution to study coordination, data analysis and manuscript preparation. This work was supported in part by a grant from Wyeth-Ayerst Research, Philadelphia, USA.

This paper was presented in part at the 18th International Congress of the Transplantation Society, Rome, Italy August 2000 and at the American Society of Nephrology Congress, Toronto, Canada October 2000.