Clinical and cost-effectiveness of newer immunosuppressive regimens in renal transplantation: a systematic review and modelling study

OBJECTIVES: To examine the clinical effectiveness and cost-effectiveness of the newer immunosuppressive drugs for renal transplantation: basiliximab, daclizumab, tacrolimus, mycophenolate (mofetil and sodium) and sirolimus. DATA SOURCES: Electronic databases. Industry submissions. Current Clinical Trials register. Cochrane Collaboration Renal Disease Group. REVIEW METHODS: The review followed the InterTASC standards. Each of the five company submissions to the National Institute for Clinical Excellence (NICE) contained cost-effectiveness models, which were evaluated by using a critique covering (1) model checking, (2) a detailed model description and (3) model rerunning. RESULTS: For induction therapy, three randomised controlled trials (RCTs) found that daclizumab significantly reduced the incidence of biopsy-confirmed acute rejection and patient survival at 6 months/1 year compared with placebo, but not compared with the monoclonal antibody OKT3. There was no significant gain in patient survival or graft loss at 3 years. The incidence of side-effects with daclizumab reduced compared to OKT3. Eight RCTs found that basiliximab significantly improved 6-month/1-year biopsy-confirmed acute rejection compared to placebo, but not compared to either ATG or OKT3. There was no significant gain in either 1-year patient survival or graft loss. The incidence of side-effects with basiliximab was not significantly different compared to OKT3/ATG. For initial/maintenance therapy, 13 RCTs found that tacrolimus reduced the 6-month/1-year incidence of biopsy-proven acute rejection compared to ciclosporin. There was no significant improvement in either 1-year or long-term (up to 5 years) graft loss or patient survival. The acute rejection benefit of tacrolimus over ciclosporin appeared to be equivalent for Sandimmun and Neoral. There were important differences in the side-effect profile of tacrolimus and ciclosporin. Seven RCTs found that mycophenolate mofetil (MMF) reduced the incidence of acute rejection. There was no significant difference in patient survival or graft loss at 1-year or 3-year follow-up. There appeared to be differences in the side-effect profiles of MMF and azathioprine (AZA). No RCTs comparing MMF with AZA were identified. One RCT compared mycophenolate sodium (MPS) to MMF and reported no difference between the two drugs in 1-year acute rejection rate, graft survival, patient survival or side-effect profile. Two RCTs suggest that addition of sirolimus to a ciclosporin-based initial/maintenance therapy reduces 1-year acute rejections in comparison to a ciclosporin (Neoral) dual therapy alone and substituting azathioprine with sirolimus in initial/maintenance therapy reduces the incidence of acute rejection. Graft and patient survival were not significantly different with either sirolimus regimen. Adding sirolimus increases the incidence of side-effects. The side-effect profiles of azathioprine and sirolimus appear to be different. For the treatment of acute rejection, three RCTs suggested that both tacrolimus and MMF reduce the incidence of subsequent acute rejection and the need for additional drug therapy. Only one RCT and one subgroup analysis in children (<18 years) were identified comparing ciclosporin to tacrolimus and sirolimus, respectively. CONCLUSIONS: The newer immunosuppressant drugs (basiliximab, daclizumab, tacrolimus and MMF) consistently reduced the incidence of short-term (1-year) acute rejection compared with conventional immunosuppressive therapy. The independent use of basiliximab, daclizumab, tacrolimus and MMF was associated with a similar absolute reduction in 1-year acute rejection rate (approximately 15%). However, the effects of these drugs did not appear to be additive (e.g. benefit of tacrolimus with adjuvant MMF was 5% reduction in acute rejection rate compared with 15% reduction with adjuvant AZA). Thus, the addition of one of these drugs to a baseline immunosuppressant regimen was likely to affect adversely the incremental cost-effectiveness of the addition of another. The trials did not assess how the improvement in short-term outcomes (e.g. acute rejection rate or measures of graft function), together with the side-effect profile associated with each drug, translated into changes in patient-related quality of life. Moreover, given the relatively short duration of trials, the impact of the newer immunosuppressants on long-term graft loss and patient survival remains uncertain. The absence of both long-term outcome and quality of life from trial data makes assessment of the clinical and cost-effectiveness on the newer immunosuppressants contingent on modelling based on extrapolations from short-term trial outcomes. The choice of the most appropriate short-term outcome (e.g. acute rejection rate or measures of graft function) for such modelling remains a matter of clinical and scientific debate. The decision to use acute rejection in the meta-model in this report was based on the findings of a systematic review of the literature of predictors of long-term graft outcome. Only a very small proportion of the RCTs identified in this review assessed patient-focused outcomes such as quality of life. Since immunosuppressive drugs have both clinical benefits and specific side-effects, the balance of these harms and benefits could best be quantified through future trials using quality of life measures. The design of future trials should be considered with a view to the impact of drugs on particular renal transplant groups, particularly higher risk individuals and children. Finally, there is a need for improved reporting of methodological details of future trials, such as the method of randomisation and allocation concealment. A number of issues exist around registry data, for example the use of multiple drug regimens and the need to assess the long-term outcomes. An option is the use of observational registry data including, if possible, prospective data on all consecutive UK renal transplant patients. Data capture for each patient should include immunosuppressant regimens, clinical and patient-related outcomes and patient demographics.