Time-dependent increased risk for serious infection from continuous use of tumor necrosis factor antagonists over three years in patients with rheumatoid arthritis.

ABSTRACT

OBJECTIVE: To investigate associations between continuous treatments with tumor necrosis factor (TNF) antagonists and risk for developing serious infections (SIs) over 3 years in Japanese patients with rheumatoid arthritis (RA) enrolled in the Registry of Japanese RA Patients for Long-Term Safety (REAL) database. METHODS: We analyzed 727 RA patients who had started either infliximab or etanercept (the anti-TNF group; 1,480.1 patient-years [PY]) and 571 RA patients who had started conventional nonbiologic disease-modifying antirheumatic drugs (the unexposed group; 1,104.1 PY) at the time of enrollment in the REAL. We assessed the occurrence of SIs within a 3-year observation period, including the period after switching to other TNF antagonists, and all SIs, unlimited to the first one in each patient as reported in other studies, to evaluate the real safety of TNF antagonists in daily practice. RESULTS: The incidence rate of SIs per 100 PY was 5.54 (95% confidence interval [95% CI] 4.44-6.84) in the anti-TNF group and 2.72 (95% CI 1.87-3.83) in the unexposed group. Poisson regression analysis revealed that the relative risk (RR) of continuous use of TNF antagonists for SIs after adjusting for baseline and time-dependent covariates was significantly elevated both overall (1.97, 95% CI 1.25-3.19) and for the first year (2.40, 95% CI 1.20-5.03), but not for the second and third years combined (1.38, 95% CI 0.80-2.43). The adjusted RR for SIs of etanercept compared to infliximab was not significantly elevated. CONCLUSION: Continuous anti-TNF therapy was significantly associated with increased risks for developing SIs during, but not after, the first year.