OBJECTIVETo evaluate the efficacy and safety of cyclosporine A(CsA) in the treatment of refractory nephrotic syndrome (RNS) in children.

METHODSThe Cochrane library, PubMed, EMBASE, CBMdisk, CNKI and VIP were searched from the time when the databases were established to December 31, 2008. Reports on RCTs on treating RNS in children with CsA were collected. Data were extracted and assessed independently by three reviewers. The methodological quality of included RCTs was assessed by the revised Jadad-scale (including randomization, allocation concealment, blinding method and withdrawal). Meta-analysis of homogenous RCTs was managed by using RevMan4.2.3.

RESULTNine RCTs involving 293 participants were included. Six RCTs were assessed as high-quality studies with scores from 4 to 7 and 3 RCTs were assessed as low-quality studies with scores from 1 to 3. Sub-category meta-analysis was based on different clinical types and interventions of RNS in children. Meta-analysis based on included RCTs showed the following results. (1) In children with steroid-dependent or frequent relapse nephrotic syndrome: the short-term efficacy of CsA plus prednisone was better than that of prednisone alone [OR 0.14, 95% CI (0.03, 0.71)]; the short-term efficacy of CsA, cyclophosphamide (CTX) and mycophenolate mofetil had no significant differences, but compared with chlorambucil, CsA had a worse short-term efficacy [OR 6.93, 95% CI (1.53, 31.38)] and a higher relapse rate [OR 0.06, 95% CI (0.01, 0.58)]; maintaining a blood level of CsA between 60 and 80 microg/L during remission period could reduce the long term relapse rate [OR 6.43, 95% CI (1.21, 34.19)]; the incidence of end-stage renal disease (ESRD) or mortality was zero in both groups. (2) In children with steroid-resistant nephrotic syndrome, the short-term efficacy of CsA was better than that of placebo or supportive treatment and CTX, OR and 95% CI were 0.15 (0.02, 0.96) and 0.41 (0.03, 5.00), respectively, but no significant differences were found in the relapse rate and the incidence of ESRD or mortality. (3) Side effects of CsA: the incidence of nephrotoxicity, hypertrichosis and gum hypertrophy was higher in the CsA group than in that of control group, OR and 95% CI were 0.19 (0.05, 0.79), 0.06 (0.02, 0.19), 0.05 (0.02, 0.18), respectively, but no significant differences were found in the incidence of hypertension and liver toxicity.

CONCLUSIONSAvailable evidence showed that CsA could improve short term efficacy in RNS in children, but could not improve long term and endpoint efficacy, therefore CsA could be one of the ideal second-line drugs for RNS in children. There was a trend that the effect of CsA on steroid-dependent or frequent relapse nephrotic syndrome was superior to that on steroid-resistant nephrotic syndrome.