Indirect comparison of biological treatments in refractory rheumatoid arthritis.

ABSTRACT

WHAT IS KNOWN AND OBJECTIVE: A number of biological treatments are available for rheumatoid arthritis. They are effective some patients but their comparative efficacy is inadequately evaluated. Our aim was to compare the efficacy of adalimumab, etanercept, infliximab, abatacept, tocilizumab, golimumab and certolizumab pegol in rheumatoid arthritis, refractory to disease-modifying antirheumatic drugs (DMARDs), through a systematic review of published trials. METHODS: As there were no direct comparisons, we searched for studies with similar characteristics to identify trials with results suitable for indirect comparison. Randomized, placebo-controlled pivotal clinical trials, with reported American College of Rheumatology ACR50 data at 24/30 weeks as efficacy endpoint, approved clinical doses and patients resistant to DMARDs who had not previously received other biological treatments were included. ACR50 was defined as the primary endpoint for the indirect comparison, with ACR20 and ACR70 as secondary endpoints. When two or more trials on one same drug were available, and a combined analysis was performed when appropriate. In the indirect comparison, the Bucher adjusted method was used with etanercept as reference drug. In the equivalence study, the equivalence window was a response efficacy difference of 15% between the alternatives. RESULTS AND DISCUSSION: Ten trials were found suitable for detailed analysis. In the clinical trials, all the biological drugs were seen to be more effective than placebo. Indirect comparison based on the ACR50 efficacy criterion all biological treatments showed similar results within the defined equivalence Δ value. The absolute efficacy difference (reduction of absolute risk, RAR) versus etanercept being 2·6% with adalimumab, 14% with infliximab, 11·6% with abatacept, 3% with tocilizumab, 12·4% with golimumab and 6·5% with certolizumab pegol. WHAT IS NEW AND CONCLUSION: The biological drugs used in rheumatoid arthritis are no different in efficacy. Their therapeutic positioning depends on their relative safety and convenience profiles.