RESUMO
OBJECTIVE: To evaluate the safety of biological treatments for rheumatoid arthritis (RA) using results from randomised controlled trials (RCT). METHODS: The literature was searched to December 2007 for RCT evaluating inhibitors of tumour necrosis factor alpha (anti-TNF) for RA. Safety data were abstracted and risk estimates were calculated using three approaches, meta-analysis with and without adjustment for exposure and simple exposure-adjusted pooling. RESULTS: Eighteen randomised trials involving 8808 RA subjects were included. Treatment with recommended doses of anti-TNF found no increase in the odds of death (odds ratio (OR) 1.39; 95% CI 0.74 to 2.62), serious adverse events (OR 1.11; 95% CI 0.94 to 1.32), serious infection (OR 1.21; 95% CI 0.89 to 1.63), lymphoma (OR 1.26; 95% CI 0.52 to 3.06), non-melanoma skin cancers (OR 1.27; 95% CI 0.67 to 2.42) or the composite endpoint of non-cutaneous cancers plus melanomas (OR 1.31; 95% CI 0.69 to 2.48) when evaluated using the unadjusted meta-analytic method. Risk estimates were similar with the other methods. For subjects who received two to three times the recommended doses of anti-TNF the risk of serious infection was increased with the unadjusted meta-analytic and pooled analysis, (OR 2.07; 95% CI 1.31 to 3.26) and (risk ratio (RR) 1.83; 95% CI 1.18 to 2.85), respectively, but not increased in the exposure-adjusted meta-analysis (RR 1.99; 95% CI 0.90 to 4.37). Meta-regression identified that the risk of serious infection with anti-TNF therapy decreases with increasing trial duration (p = 0.035). CONCLUSION: Meta-analytic and exposure-adjusted pooled analyses on over 8800 RA subjects in RCT treated over an average of 0.8 years did not identify an increased risk of serious adverse events with recommended doses. High-dose anti-TNF therapy was associated with a twofold increase in the risk of serious infections.