OUP user menu

The evolving role of sirolimus in renal transplantation

P. Dupont, A.N. Warrens
DOI: http://dx.doi.org/10.1093/qjmed/hcg072 401-409 First published online: 1 June 2003


Sirolimus (rapamycin) is a macrolide antibiotic isolated from the fungus Streptomyces hygroscopicus1 first identified in soil samples from Easter Island. Structurally similar to the calcineurin inhibitor (CNI) tacrolimus (Figure 1), it shares its potent immunosuppressive properties, but with a novel mechanism of action (Figure 2). Initial clinical trials have focussed on its use as adjunctive therapy, substituting for the anti-metabolite azathioprine. However, given that it lacks the nephrotoxicity of the CNIs (cyclosporin and tacrolimus), it has the potential to supplant these agents as baseline immunosuppressive therapy following renal transplantation, and offers the added potential benefit of promoting immunological tolerance.

Figure 1.

Structure of sirolimus. Sirolimus (rapamycin) is a macrocyclic lactone structurally similar to tacrolimus (FK506).

Figure 2.

Mechanism of action of sirolimus (rapamycin). CD4-positive T cells recognize a complex of class II MHC and peptide present on the surface of antigen-presenting cells (APCs). This results in signalling via the T cell receptor (TCR) and activation of calcineurin, a phosphatase. Calcineurin dephosphorylates NFAT (nuclear factor of activated T cells) facilitating its transfer into the nucleus where it acts as a transcription factor regulating IL-2 production. IL-2, and other cytokines, in turn promote further T-cell activation and proliferation, following interaction with its cell surface receptor (IL-2R). Of note, T-cell receptor engagement alone is insufficient to produce an immune response. Co-stimulatory signals (prototypically via CD28) are also necessary to produce and sustain a T-cell response. T-cell receptor engagement in the absence of co-stimulatory signals is thought to produce tolerance by inducing non-responsiveness to specific antigens and promoting apoptotic deletion of the activated cells (activation-induced cell death, AICD). Cyclosporin and tacrolimus (FK506) act by binding to their respective immunophilins (cyclophilin/FK-binding protein, FKBP-12) with the resultant complexes producing calcineurin inhibition. The net effect is blockade of IL-2 production, resulting in inhibition of T-cell activation. Sirolimus (SRL) also binds the same cytosolic receptor—FKBP-12. However, in this case the complex binds the mammalian target of rapamycin (mTOR), resulting in disruption of IL-2 receptor signalling and inhibition of T-cell proliferation (by inducing cell cycle arrest). Sirolimus also blocks co-stimulatory signals generated by the engagement of CD28, which may be beneficial in inducing tolerance.


Sirolimus binds to the same cytosolic receptor as tacrolimus, namely the immunophilin, FK binding protein-12 (FK BP12).2 This complex then binds to the mammalian target of rapamycin (mTOR)—a key regulatory kinase. Disruption of mTOR impedes cytokine signalling, resulting in the inhibition of lymphocyte growth and differentiation. In IL-2-stimulated T cells, sirolimus impedes progression through the G1/S transition of the proliferation cycle, resulting in a mid-to-late G1 arrest.3,,4 Two key targets have been identified for the sirolimus-FKBP12-mTOR complex. The first is the p70 S6 kinase, which plays a crucial role in cytokine-induced proliferation. By inhibiting this enzyme, sirolimus reduces the translation of certain mRNAs encoding for ribosomal proteins (the major substrate for p70 S6 kinase is the 40S ribosomal subunit S6 protein) and elongation factors, thereby decreasing protein synthesis. A second, later effect of sirolimus in IL-2-stimulated T cells is an inhibition of the enzymic activity of the cyclin-dependent kinase cdk2-cyclin E complex, which functions as a key regulator of G1/S transition.

By contrast with the CNIs, tacrolimus and cyclosporin, which inhibit production of IL-2 and other cytokines,5 sirolimus acts further downstream to block cytokine-driven proliferation. In addition, sirolimus inhibits the downregulation of IκBα that occurs following the engagement of CD28,6 possibly the most important ‘co-stimulatory’ molecule on T lymphocytes. Sirolimus thus inhibits T-cell activation in response to both antigenic and cytokine stimulation.7

Because the anti-proliferative effects associated with sirolimus are related to arrest of the cell cycle, this action occurring downstream of IL-2 receptor engagement, sirolimus creates a permissive environment for activation-induced cell death (AICD) by apoptosis.8,,9 By this mechanism, T cells reactive to graft-specific antigens are deleted. This process is thought to be instrumental in the development and maintenance of tolerance.10

So far this discussion has been limited to the effects of sirolimus on T cells. However it also blocks B-cell activation,2,,11 probably by inhibiting the action of mTOR and Ig receptor binding protein-112,,13 (a protein associated with the Igα component of the B-cell receptor involved in signal transduction).

Finally, sirolimus inhibits growth factor-stimulated proliferation of parenchymal cells, such as fibroblasts and smooth muscle cells.14 In the clinical setting, this last property may translate into a reduction in chronic allograft nephropathy (CAN),15 which remains the most important cause of a transplant patient returning to dialysis.16

Pre-clinical data

Effectiveness of sirolimus in transplantation

Sirolimus has been used in animal studies of transplantation for over 10 years, either alone, or in combination with other immunosuppressive drugs. Calne et al.17 and Morris et al.18 published the first in vivo studies in 1989. In a rat cardiac transplantation model, both groups noted longer graft and animal survival rates in SRL-treated recipients compared with non-immunosuppressed controls. Similar results have been shown for skin, renal, small bowel and pancreas allografts in a number of species, including mice, rats, rabbits, pigs, dogs, and primates.19–,30 In addition to its use as prophylaxis in organ transplantation, SRL has also been investigated as a potential treatment for ongoing acute rejection. Using a rat model, Chen et al. found that ‘rescue therapy’ with SRL prolonged the survival of cardiac, renal and pancreatic allografts.31

Stepkowski et al. demonstrated that there might be synergistic effects between sirolimus and CNIs.32 In a rat cardiac and skin transplantation model, sub-therapeutic doses of SRL and CyA prolonged allograft survival compared with either drug alone or the additive effect of a combination of both. Similar results were observed in studies of canine and mouse kidney allotransplantation.29,,33 A synergistic effect has also been shown with tacrolimus.34

Sirolimus and vascular disease

Calcineurin inhibitors have not been an unalloyed blessing in renal transplantation; although they have improved early graft survival, they also have considerable nephrotoxicity and promote vasculopathy and graft fibrosis.35 As mentioned above, chronic vasculopathy, a major component of CAN, remains a significant cause of graft failure. One of the most attractive properties to emerge from pre-clinical studies of sirolimus is its ability to inhibit vascular smooth muscle cell activation, migration, and proliferation.36 In a study by Ikonen et al.,37 sirolimus inhibited the progression of graft vascular disease in cynomolgus monkey recipients of aortic allografts. By contrast, monkeys treated with cyclosporin or placebo showed progressive arterial intimal thickening. This has already been exploited clinically, as is discussed below.

Sirolimus and tolerance

The generation of transplantation tolerance (i.e. antigen-specific non-responsiveness) would be an ideal way to prevent graft rejection while minimizing the adverse effects of immunosuppression. Indeed, for years tolerance has been the ‘holy grail’ of transplantation biology. In the first instance, the use of SRL in place of CNIs would carry an advantage for tolerance induction. The blocking of signals generated following engagement of the TCR, which is how CNIs work, not only inhibits T-cell activation but also the active induction of tolerance.38,,39 However, unlike CNIs,38–,42 sirolimus does not interfere with the development of transplantation tolerance. Rather, SRL allows the transmission that antigen has been recognized, but prevents consequent proliferation, which has been shown to be the ideal scenario for tolerance induction.43

A second reason for SRL favouring tolerance is its promotion of activation-induced cell death (AICD) in alloreactive T cells. One of the most effective mechanisms for achieving tolerance is by the depletion of activated T cells. Stimulated T cells that have begun cell division are particularly susceptible to apoptosis. Sensitivity to AICD is abolished by CNIs but not by SRL.39,,42 Indeed, in a stringent murine skin allograft model,44 SRL was synergistic with the blockade of the non-antigen-specific ‘co-stimulatory’ signals (in this model, the interactions of B7 with CD28 and CD40 with CD40L) in promoting tolerance. In contrast to the effects of CNIs, co-stimulation blockade or SRL alone, only SRL plus co-stimulation blockade produced long-term graft survival.

Clinical experience in renal transplantation

The development of cyclosporin in the mid-1980s markedly improved renal allograft survival when added to the then conventional steroid and azathioprine regimen. More recently, tacrolimus has replaced cyclosporin as the CNI of choice in many units, and azathioprine has given way to the newer anti-metabolite agent mycophenolate mofetil.

In its clinical development as induction immunosuppression, sirolimus has been evaluated in two roles, either as a replacement for the CNI45 or together with a CNI46 (in comparison with placebo47 or azathioprine48). Although replacing a CNI potentially offers greater benefits, it represents a more radical departure from current treatment and, as such, the transplant community has been slower to experiment with this approach. Hence more data are available for the use of sirolimus in conjunction with a CNI.

As an adjunct to a CNI

In a phase II trial involving 149 renal allograft recipients, Kahan et al.46 reported the combined use of SRL, steroids and either full-dose or reduced dose CyA for prophylaxis of acute renal allograft rejection. They found a lower incidence of acute rejection episodes in the group receiving full-dose CyA and SRL, compared with the group receiving full-dose CyA and placebo (SRL 8.5%; placebo 32%). In addition, sirolimus permitted a 50% reduction in CyA trough levels without an increased risk of rejection.

There have been two large phase III studies reported: the ‘Global study’47 and the ‘US study’.48 The data from these trials are summarized in Table 1. In the Global study, 576 renal allograft recipients were randomized to receive either SRL (2 mg or 5 mg) or placebo, in addition to baseline immunosuppression with CyA and corticosteroids. At 6 months, acute rejection rates were significantly lower in the SRL groups compared with placebo (SRL 2 mg 24.7%, SRL 5 mg 19.2%, placebo 41.5%). There was also a statistically significant reduction in episodes of steroid-resistant rejection in the SRL 5 mg group vs. placebo (SRL 2 mg 4%, SRL 5 mg 3.2%, placebo 8.5%). There was no significant increase in the incidence of malignancy between the groups (SRL 2 mg 2.2%, SRL 5 mg 4.1%, placebo 1.5%). Infection rates across the groups were similar with the exception of an increase in mucosal herpes simplex virus infection in the 5 mg SRL group. Hyperlipidaemia (SRL 2 mg 35%, SRL 5 mg 50%, placebo 18%) and thrombocytopenia (SRL 2 mg 11%, SRL 5 mg 23%, placebo 3%) were more common in the SRL groups. These observations were statistically significant.

View this table:
Table 1

Use of sirolimus with cyclosporin
 a Global study (cyclosporin/prednisolone baseline immunosuppression)

AR at 6 monthsSteroid-resistant ARMalignancyLow plateletsHyperlipidaemia
Sirolimus 2mg24.7%4%2.2%11%35%
Sirolimus 5mg19.2%3.2%4.1%23%50%
Placebo41.5%8.5%1.5% 3%18%
b US study (cyclosporin/prednisolone baseline immunosuppression)
1-year graft survival1-year patient survivalARMean serum creatinine at 12 months
Sirolimus 2mg97.2%94.7%16.9%160 µmol/l
Sirolimus 5mg96%92.7%12%171 µmol/l
Azathioprine98.1%93.8%29.8%133 µmol/l
  • AR, acute rejection.

In the US study of 719 renal allograft recipients with good initial graft function, SRL (2 mg or 5 mg) was compared to azathioprine (Aza). Use of SRL was associated with a significant reduction in acute rejection at one year (SRL 2 mg 16.9%; SRL 5 mg 12.0%; Aza 29.8%). There were significantly fewer moderate and severe histological grades of rejection in the SRL groups compared with the azathioprine-treated group (SRL 2 mg 9.2%; SRL 5 mg 4.4%; Aza 17.4%). Of note, the African-American subgroup of patients showed benefit only at the 5 mg dose of SRL. At 12 months, data were similar in all groups for graft survival (97.2%; 96.0%; 98.1%) and patient survival (94.7%; 92.7%; 93.8%). Rates of infection and malignant disorders were similar in all groups. At 6 and 12 months, the mean serum creatinine concentrations were significantly higher in patients in the two sirolimus groups than in those in the azathioprine group (p < 0.001) and the mean creatinine clearance was lower (2 mg p < 0.01; 5 mg p < 0.001). Although this was not associated with higher cyclosporin whole blood trough concentrations, it nevertheless seems likely that this difference was due to potentiation of cyclosporin toxicity. There is known to be an interaction between the two drugs, and the authors reported evidence of exacerbation of other cyclosporin side-effects, and observed that significantly lower cyclosporin doses were required to achieve whole blood target concentrations among sirolimus-treated patients than among azathioprine-treated patients. There is also evidence from animal models that sirolimus may increase cyclosporin partitioning into renal tissue to a greater extent than it increases whole-blood concentrations.49

In a further study of African-American renal transplant recipients50 (n = 137), Kahan reported that sirolimus improved 2-year graft survival when added to baseline immunosuppression with CyA and Prednisolone (SRL/CyA/Pred 97.9% vs. CyA/Pred 85.6%). Patient survival was similar in the two groups (SRL/CyA/Pred 95.7% vs. CyA/Pred 97.8%). Acute rejection rates were significantly lower in those African-American patients who received sirolimus than in those treated with CyA/Pred alone (SRL/CyA/Pred 19.2% vs. CyA/Pred 43.3%). Graft survival in the sirolimus treated cohort of African-American patients was comparable to that in a cohort (n = 120) of similarly treated Caucasian patients (2-year graft survival 91.7%).

SRL has not yet been fully compared to MMF in any study reported in anything other than abstract form.

CNI replacement (primary therapy)

Groth et al.45 reported the use of sirolimus (SRL) vs. cyclosporin (CyA) as base therapy (in conjunction with steroids and azathioprine) for the prophylaxis of acute rejection in 83 first cadaveric renal allograft recipients in a multi-centre trial. At 12 months, graft survival (SRL 98% vs. CyA 90%), patient survival (100% vs. 98%), and incidence of biopsy-confirmed acute rejection (41% vs. 38%) were similar in the two groups (Table 2). Also of interest, serum creatinine levels were consistently lower in the SRL group. This difference was statistically significant at 12 weeks (SRL 126.2±11.4 µmol/l; CyA 159.2 ± 11.2 µmol/l) and 16 weeks (SRL 123.8 ±&!thinsp; 10.5 µmol/l; CyA 153.4±9.6 µmol/l) although not at 12 months. It is tempting to speculate that this difference in graft function reflects lower nephrotoxicity of sirolimus compared to CNIs, and that this might translate into a reduction in CAN and prolonged graft survival. Future studies with longer duration of follow-up will hopefully address this important question.

View this table:
Table 2

Use of SRL as an alternative to cyclosporin

1-year graft survival1-year patient survivalAcute rejectionSerum creatinine
CyA/Pred/Aza90% 98%38%133.5±7.7
  • SRL, sirolimus; Pred, prednisolone; Aza, azathioprine; CyA, cyclosporin. Data from Groth et al.45

Other roles for sirolimus—drug-eluting coronary stents

The inhibitory effect of sirolimus on vascular smooth muscle has also found application in the treatment of coronary artery disease.51 Human studies have shown that coating coronary stents with sirolimus prevents neo-intimal proliferation, thereby dramatically reducing restenosis rates.52 The RAVEL study53 showed a 0% restenosis rate at 6 months in patients receiving sirolimus-coated stents, vs. 26.6% in the standard stent group. More recently, the SIRIUS study reported preliminary results in 400 randomized patients receiving sirolimus-coated stents. Nine-month data showed an in-stent restenosis rate of 2%, compared with 32.3% in the standard stent group.54

Adverse effects

The principal side-effects of sirolimus are bone marrow suppression and hyperlipidaemia. The former usually manifests as a mild thrombocytopenia, and seldom necessitates discontinuation of treatment. Severe thrombocytopenia is rare, and there have been no reports to date of bleeding episodes.45–48,,57 The phenomenon also appears to be dose-related,46,,47 often resolving with reduction to maintenance levels. Hyperlipidaemia is more common among sirolimus-treated patients, but the magnitude of the effect is small and it is seldom a cause for discontinuation of therapy. Most patients respond to conventional lipid-lowering treatment.

Other reported side-effects include hypokalaemia, hyperglycaemia, diarrhoea and abnormal liver function tests. Of note, in contrast to CNIs, sirolimus tends to lower uric acid levels, which may reduce the incidence of gout following transplantation. More significantly, sirolimus lacks the nephrotoxicity and neurotoxicity seen with CNIs.45–,48 There are some data to suggest that sirolimus may enhance the nephrotoxicity associated with cyclosporin, probably by a pharmacokinetic interaction which results in elevated blood cyclosporin levels.55 This interaction is not seen with tacrolimus.56

With regard to infectious complications, as has already been stated, sirolimus treatment may be associated with an increased incidence of mucosal herpes simplex infection47,48,,57 (although it has been suggested that the mucosal lesions seen by these authors were due to local drug toxicity rather than HSV infection, as there was no virological confirmation of the diagnosis). In the phase II trial by Kahan et al., there was an excess incidence of Pneumocystis carinii pneumonia (PCP) in the sirolimus group,46 although this finding was not confirmed in subsequent studies. Of note, all cases occurred at a single centre where PCP prophylaxis was not routinely given. Groth et al. also reported an increased incidence of pneumonia in patients receiving sirolimus compared with those receiving cyclosporin (17% vs. 2%), but in this study the pneumonia was not PCP. A similar trend was observed by Kreis et al.58 in the Sirolimus European Renal Transplant Study Group. Rates of CMV infection do not appear to be increased by sirolimus use.46–48,57,,58

With regard to malignant complications, trials to date have not reported any increase in post-tranplantation lymphoproliferative disease (PTLD) or other malignancy associated with sirolimus use.

Pharmacodynamics and drug monitoring

Sirolimus is rapidly absorbed (Tmax 1–2 h) with relatively low oral bioavailability (14%) and linear dose proportionality.59 The half-life is long (62 h) and there is large inter- and intra-subject variability in drug clearance. Monitoring of drug levels may not be essential where sirolimus is being used as adjunctive therapy. However, it may be advisable where sirolimus is used as base therapy, to ensure adequate levels of immunosuppression. The large intra-patient variability observed in trough sirolimus concentrations indicates that dose adjustments should be optimally based on more than a single trough sample. Because of the time required to reach steady state, sirolimus dose adjustments would optimally be based on trough levels obtained >5–7 days after a dose change. A whole-blood sirolimus therapeutic window of 5–15 ng/ml (measured by microparticle enzyme immunoassay) is recommended for patients at standard risk of rejection.

Where sirolimus is taken concomitantly with cyclosporin, the two should be separated by an interval of 4 h to reduce the risk of interaction. It may also be prudent to adjust cyclosporin doses to produce trough concentrations at values lower than the putative therapeutic range for a purely cyclosporin-based regimen.

Potential roles for sirolimus therapy

Due to the very short follow-up in the literature to date, most units do not use sirolimus as a first-line therapy. There are, however, several settings in which sirolimus may be preferred over other agents:

Nephrotoxicity due to calcineurin inhibitors

Nephrotoxicity has long been the Achilles’ heel of treatment with calcineurin inhibitors. A small proportion of patients experience significant impairment of graft function even at conventional doses of calcineurin inhibitor. Sirolimus is an attractive alternative in this setting. The available evidence suggests it is not nephrotoxic on its own, and is effective as base therapy. Early reports suggest that switching to sirolimus is a viable approach to treating patients with CNI toxicity.60

Intolerance of mycophenolate mofetil (MMF)

MMF is commonly used both as induction immunoprophylaxis and as ‘rescue therapy’ in patients following episodes of acute rejection. Gastrointestinal and haematological side-effects are common with this drug, resulting in discontinuation of treatment in up to 20% of patients.61 Sirolimus would be an appropriate alternative if MMF was not tolerated.

Haemolytic uraemic syndrome (HUS)

Patients with HUS are at risk of recurrent disease following transplantation. This risk may be increased if CNIs are used.62 Prognosis in recurrent HUS is poor, frequently resulting in loss of the graft and occasionally life-threatening systemic disease. Hence in these cases, the use of SRL as primary therapy may be appropriate. A recent report of two cases suggests that, in patients with CNI-induced HUS, substituting sirolimus for the CNI permits resolution of the thrombotic microangiopathy without increased risk of graft rejection.63

Delayed graft function (DGF)

DGF due to slowly recovering acute tubular necrosis occurs in up to 40% of renal allografts following transplantation.64 Use of calcineurin inhibitors in this setting may exacerbate the problem because of their intrinsic nephrotoxicity and potential to cause vasoconstriction.65 Nevertheless, potent immunosuppression is still required in this setting, as an ischaemic graft is at increased immunological risk.66 For this reason, many would advocate continuation of CNIs in patients with DGF where possible. Others reduce the dose or stop the drug altogether. An alternative approach would be to substitute sirolimus. This should provide potent immunosuppression without impeding graft recovery.


In summary, sirolimus is an exciting addition to our immunosuppressive armamentarium. There are encouraging early data suggesting that this drug may have significant advantages over existing agents, namely the promotion of tolerance, reduction in chronic allograft vasculopathy and avoidance of toxic effects of other immunosuppressive drugs. If further large clinical studies can confirm its safety and efficacy over a longer duration of follow-up this will no doubt promote its more widespread adoption.


View Abstract