Martin Wagner, Amy K Earley, Angela C Webster, Christopher H Schmid, Ethan M Balk, Katrin Uhlig

Modern immunosuppressive regimens after kidney transplantation usually use a combination of two or three agents of different classes to prevent rejection and maintain graft function. Most frequently, calcineurin-inhibitors (CNI) are combined with corticosteroids and a proliferation-inhibitor, either azathioprine (AZA) or mycophenolic acid (MPA). MPA has largely replaced AZA as a first line agent in primary immunosuppression, as MPA is believed to be of stronger immunosuppressive potency than AZA. However, treatment with MPA is more costly, which calls for a comprehensive assessment of the comparative effects of the two drugs. This review of randomised controlled trials (RCTs) aimed to look at the benefits and harms of MPA versus AZA in primary immunosuppressive regimens after kidney transplantation. Both agents were compared regarding their efficacy for maintaining graft and patient survival, prevention of acute rejection, maintaining graft function, and their safety, including infections, malignancies and other adverse events. Furthermore, we investigated potential effect modifiers, such as transplantation era and the concomitant immunosuppressive regimen in detail. We searched Cochrane Kidney and Transplant's Specialised Register (to 21 September 2015) through contact with the Trials' Search Co-ordinator using search terms relevant to this review. All RCTs about MPA versus AZA in primary immunosuppression after kidney transplantation were included, without restriction on language or publication type. Two authors independently determined study eligibility, assessed risk of bias and extracted data from each study. Statistical analyses were performed using the random-effects model and the results were expressed as risk ratio (RR) for dichotomous outcomes and mean difference (MD) for continuous outcomes with 95% confidence intervals (CI). We included 23 studies (94 reports) that involved 3301 participants. All studies tested mycophenolate mofetil (MMF), an MPA, and 22 studies reported at least one outcome relevant for this review. Assessment of methodological quality indicated that important information on factors used to judge susceptibility for bias was infrequently and inconsistently reported.MMF treatment reduced the risk for graft loss including death (RR 0.82, 95% CI 0.67 to 1.0) and for death-censored graft loss (RR 0.78, 95% CI 0.62 to 0.99, P < 0.05). No statistically significant difference for MMF versus AZA treatment was found for all-cause mortality (16 studies, 2987 participants: RR 0.95, 95% CI 0.70 to 1.29). The risk for any acute rejection (22 studies, 3301 participants: RR 0.65, 95% CI 0.57 to 0.73, P < 0.01), biopsy-proven acute rejection (12 studies, 2696 participants: RR 0.59, 95% CI 0.52 to 0.68) and antibody-treated acute rejection (15 studies, 2914 participants: RR 0.48, 95% CI 0.36 to 0.65, P < 0.01) were reduced in MMF treated patients. Meta-regression analyses suggested that the magnitude of risk reduction of acute rejection may be dependent on the control rate (relative risk reduction (RRR) 0.34, 95% CI 0.10 to 1.09, P = 0.08), AZA dose (RRR 1.01, 95% CI 1.00 to 1.01, P = 0.10) and the use of cyclosporin A micro-emulsion (RRR 1.27, 95% CI 0.98 to 1.65, P = 0.07). Pooled analyses failed to show a significant and meaningful difference between MMF and AZA in kidney function measures.Data on malignancies and infections were sparse, except for cytomegalovirus (CMV) infections. The risk for CMV viraemia/syndrome (13 studies, 2880 participants: RR 1.06, 95% CI 0.85 to 1.32) was not statistically significantly different between MMF and AZA treated patients, whereas the likelihood of tissue-invasive CMV disease was greater with MMF therapy (7 studies, 1510 participants: RR 1.70, 95% CI 1.10 to 2.61). Adverse event profiles varied: gastrointestinal symptoms were more likely in MMF treated patients and thrombocytopenia and elevated liver enzymes were more common in AZA treatment. MMF was superior to AZA for improvement of graft survival and prevention of acute rejection after kidney transplantation. These benefits must be weighed against potential harms such as tissue-invasive CMV disease. However, assessment of the evidence on safety outcomes was limited due to rare events in the observation periods of the studies (e.g. malignancies) and inconsistent reporting and definitions (e.g. infections, adverse events). Thus, balancing benefits and harms of the two drugs remains a major task of the transplant physician to decide which agent the individual patient should be started on.