A meta-analysis of the efficacy and toxicity of combining disease-modifying anti-rheumatic drugs in rheumatoid arthritis based on patient withdrawal

Introduction. Combinations of disease-modifying anti-rheumatic drugs (DMARDs) are increasingly used to treat rheumatoid arthritis (RA). Early trials showed their toxicity while recent trials suggest superior efficacy. Trials of DMARD combinations have enrolled different types of patient (early or established RA), used different designs (step-up, parallel or step-down) and utilized a range of outcome measures. We undertook a systematic review of combination DMARD therapy for RA and carried out a meta-analysis to evaluate the evidence for efficacy and toxicity. Method. Medline, PubMed and EmBase were searched using MESH headlines 'arthritis, rheumatoid', 'drug therapy, combination' and 'randomized controlled trial' (RCT) for papers published from 1975 to April 2004. References from published articles were also searched. Three independent assessors evaluated abstracts and selected trials for detailed examination. Trials were excluded if their quality was poor, were not published in English or studied DMARDs not licensed to treat RA. Two independent assessors extracted data. Efficacy was assessed by the numbers of patients withdrawn due to lack of efficacy. Toxicity was assessed by the numbers of patients withdrawn due to adverse events. Risk ratios (RR) with 95% confidence intervals (CI) were calculated and meta-analysis was carried out based on a random effects model. Sensitivity analyses evaluated different treatment combinations, trial designs, study populations and outcome measures. Results. Fifty-three potentially relevant RCTs were identified. Twelve were excluded due to: using unlicensed DMARDs (n = 3); reporting in journal supplements of RCTs already included (n = 2); follow-up of an earlier RCT, report of biological outcomes or pharmacokinetics (n = 5); and non-English language publications (n = 2). Forty-one RCTs were evaluated in detail and another five excluded (three open-labelled studies and two with high patient attrition); 36 studies were included in the meta-analysis. These comprised 13 step-up, 16 parallel and 7 step-down trials. Nine assessed early RA and 27 established RA. Seven added steroids to DMARD monotherapy and one study added steroids to DMARD combinations. Six assessed methotrexate (MTX) plus tumour necrosis factor (TNF) inhibitors. Overall, combination DMARD therapy was more effective than monotherapy (RR 0.35; 95% CI 0.28, 0.45) although the risk of toxicity was also slightly higher (RR 1.37; 95% CI 1.16, 1.62). Combinations of MTX with TNF inhibitors and MTX with sulphasalazine or anti-malarials showed good efficacy/toxicity ratios. Conclusions. DMARD combinations vary in their efficacy/toxicity ratio. MTX plus sulphasalazine and/or anti-malarials and MTX plus TNF inhibitors have particularly favourable benefit/risk ratios.