Tumor necrosis factor antagonist responsiveness in a United States rheumatoid arthritis cohort.

objectiveThe study objective was to investigate responsiveness according to whether patients satisfy eligibility criteria from randomized controlled trials of tumor necrosis factor (TNF) antagonists in a multicentered US cohort.

methodsBiologic-naive patients with rheumatoid arthritis who were prescribed TNF antagonists (n=465) in the Consortium of Rheumatology Researchers of North America registry were included. Patients were stratified by whether they met eligibility criteria from 3 major TNF antagonist trials. Two cohorts were examined: Cohort A (n=336) included patients with complete American College of Rheumatology response criteria except acute phase reactants, and cohort B (n=129) included patients with complete response criteria. Study outcomes included modified American College of Rheumatology 20% and 50% improvement responses (cohort A) and standard American College of Rheumatology improvement (cohort B).

resultsA minority of patients (5.4%-19.4%) prescribed TNF antagonists met trial eligibility criteria and predominantly had high disease activity (78.5%-100%). For patients who met eligibility criteria in cohort A, rates of 20% improvement (52.3%-63.6%) and 50% improvement (30.8%-45.5%) were achieved. Among patients failing to meet eligibility criteria, rates of 20% improvement (16.2%-20.4%) and 50% improvement (8.9%-10.8%) were consistently inferior (P<.05 all comparisons). For cohort B, similar differences were observed.

conclusionThis multicentered US cohort study demonstrates that the majority of patients receiving TNF antagonists would not meet trial eligibility criteria and achieve lower clinical responses. These findings highlight the tradeoff between defining treatment responsive populations and achieving results that can be generalized for broader patient populations.