An observational study of tocilizumab and TNF-α inhibitor use in a Japanese community hospital: different remission rates, similar drug survival and safety

OBJECTIVE To assess the effectiveness, drug survival and safety of tocilizumab compared with TNF-α inhibitors in clinical practice. METHODS Patients in the Cohort of Arthritis Biologic Users at Kameda Institute (CABUKI) registry who were on biologics during July 2003 to October 2010 were included. Remission rates at 6 months, Kaplan-Meier drug survival estimates and serious adverse event (SAE) rates were compared. RESULTS A total of 247 RA patients were analysed. For first-line biologic users, the 6-month 28-joint DAS (DAS-28)-ESR remission rates were 66.7% for tocilizumab vs 25.8% for TNF inhibitors (P < 0.001, Fisher's exact test). This advantage disappeared with the application of the newly suggested Boolean remission criterion for clinical trials: 0% for tocilizumab vs 8.2% for TNF inhibitors (P = 0.367, Fisher's exact test). Tocilizumab users in DAS-28-ESR remission had lower mean ESR (3.9  mm/h for tocilizumab vs 7.9  mm/h for TNF inhibitors, P = 0.026, t-test) and higher mean swollen joint count (2.6 for tocilizumab vs 1.3 for TNF inhibitors, P = 0.036, t-test), thus failing to meet the more stringent Boolean criteria. First- and second-line tocilizumab users showed similar drug survival and SAE rates compared with TNF inhibitor users. CONCLUSION Tocilizumab had drug survival and safety profiles similar to those of TNF inhibitors in this Japanese single-centre registry. Tocilizumab was superior to TNF inhibitors when compared at 6 months by DAS-28-ESR remission. However, the newly suggested Boolean criteria are more appropriate measures of effectiveness as DAS-28-ESR remission by tocilizumab was mainly due to very low ESR in our study population.