Ten RCTs were included for review (n=847). Sample sizes ranged from 20 to 370. Seven studies had a Jadad score of 1 or 2 and three had a score of 3. Four studies had adequate allocation concealment.
Mycophenolate mofetil did not significantly differ from cyclophosphamide in effects on complete or partial renal remission at six months (RR 1.05, 95% CI 0.95 to 1.17; eight studies, n=742), all-cause mortality (RR 0.71, 95% CI 0.37 to 1.35; eight studies, n=783) and ESRF (RR 0.45, 95% CI 0.18 to 1.12; four studies, n=306).
Sensitivity analyses that included only high-quality trials did not significantly alter the findings (three studies, n=245). There was evidence of moderate statistical heterogeneity for renal remission, but no significant statistical heterogeneity for mortality or ESRF.
Significantly fewer participants who received mycophenolate mofetil developed amenorrhoea (RR 0.21, 95% CI 0.09 to 0.48; six studies, n=357), leukopenia (RR 0.47, 95% CI 0.27 to 0.83; eight studies, n=434), herpes infection (RR 0.7, 95% CI 0.39 to 1.25; eight studies, n=434) or pneumonia (RR 0.57, 95% CI 0.24 to 1.36; five studies, n=329) compared to cyclophosphamide. Significantly more patients who received mycophenolate mofetil experienced diarrhoea (RR 2.08, 95% CI 0.98 to 4.40; five studies, n=280). There was no evidence of statistical heterogeneity for all adverse events except diarrhoea (I2=23%).
In a meta-regression the proportion of non-white or non-Asian participants in a study was the only variable to significantly impact on heterogeneity (p=0.05).