Association between seropositivity and discontinuation of tumor necrosis factor inhibitors due to ineffectiveness in rheumatoid arthritis.

INTRODUCTION/

OBJECTIVES

Discontinuation of biologic therapy in rheumatoid arthritis is attributable to various reasons, with the most important cause being insufficient response. In this study, we investigated the association between rheumatoid factor (RF) and anti-citrullinated protein autoantibody (ACPA) status and the discontinuation of tumor necrosis factor inhibitors (TNFi) therapy due to insufficient response in bio-naïve rheumatoid arthritis (RA) patients.

METHOD: This study included patients enrolled in the Tsurumai Biologic Communication Registry in Japan. The crude comparison of TNFi discontinuation due to ineffectiveness between seropositive and seronegative patients was analyzed using the cumulative incidence function of competing events and Gray test. We assessed the associations between baseline patient characteristics and discontinuation of TNFi therapy due to insufficient response using Fine-Gray proportional hazard regression. Fine-Gray proportional hazard analysis considered competing events of interest, including insufficient response, adverse event, palliation, and personal reasons.

RESULTS

Of 1237 patients evaluated, 79.3% were positive for RF and 85.4% for ACPA; 72.6% were double positive and 11.1% were double negative. TNFi therapy had been discontinued because of insufficient response at 200 weeks in 19.8% RF-positive, 16.7% RF-negative, 23.0% ACPA-positive, and 13.8% ACPA-negative patients. There was a significantly higher discontinuation rate due to insufficient response in ACPA-positive patients than in ACPA-negative patients using Gray test, with a similar trend as that for RF status. RF positivity was significantly predictive of the discontinuation of TNFi therapy due to ineffectiveness using Fine-Gray proportional hazard regression analysis after adjusting for baseline characteristics, including age, sex, stage, class, disease activity at baseline, methotrexate use, and prednisolone use [hazard ratio 1.73 (95% confidence interval 1.07-2.80)].

CONCLUSIONS

Using Fine-Gray proportional hazard regression, we demonstrated that RF positivity was related to a higher discontinuation rate of TNFi therapy due to ineffectiveness in bio-naïve RA patients. Key Points • RF positivity is related to a higher discontinuation rate of TNFi therapy due to ineffectiveness. • ACPA is not predictive of a discontinuation of TNFi therapy due to ineffectiveness.