Nineteen RCTs (n=3,143 patients) were included in the review. Trial quality was poor, with only two trials scoring 3 or more on the Jadad scale. Only two trials reported adequate allocation concealment and seven used an intention-to-treat analysis. Duration of follow-up ranged from six to 60 months.
Acute rejection: Mycophenolate mofetil significantly reduced the risk of acute rejection compared with azathioprine (RR 0.62, 95% CI 0.55 to 0.70; 17 RCTs; I2=6.8%). Subgroup analyses showed that the risk reduction of acute rejection was greater in patients receiving sandimmune (RR 0.53) than microemulsions (RR 0.70); this was also the case in patients receiving higher doses (3g/day) of mycophenolate mofetil (RR 0.48) compared with lower doses (2g/day or less) (RR 0.66). Sensitivity analyses of higher quality trials did not alter the results. There was some evidence of publication bias.
Patient and graft survival: Mycophenolate mofetil reduced the hazard of graft loss (HR 0.76, 95% CI 0.59 to 0.98; 11 RCTs; I2=0%) There were no significant differences for patient survival or graft loss, or between subgroups. There was no evidence of publication bias. When the analysis was repeated for higher quality trials, the results were no longer statistically significant, but only three trials were included. There was no evidence of a difference between mycophenolate mofetil and azathioprine for patient survival or graft loss, excluding patient death with a functioning graft.
Graft function: Graft function measured by serum creatinine showed no overall difference between groups, but the observed heterogeneity was moderate (I2=40.3%). There was no evidence of a difference in the glomerular filtration rate.
Side effects: Mycophenolate mofetil increased the risk of diarrhoea (RR 1.57, 95%CI 1.33 to 1.86; six RCTs; I2=1.5%), but there were no differences between mycophenolate mofetil and azathioprine for vomiting or nausea or for more severe side effects (infections, anaemia, leukopenia or malignancy).