Background Clinical trials have shown that tocilizumab (TCZ) is efficacious as monotherapy and in combination with methotrexate (MTX) or other DMARDs. However, longitudinal TCZ retention data from large registry populations have been missing.
Objectives To examine retention of TCZ administered alone or in combination with DMARDs in rheumatoid arthritis (RA) patients included in the TOcilizumab Collaboration of European Registries in RA (TOCERRA).
Methods RA patients treated with TCZ who had baseline (BL) data and not immediately lost to follow-up were included. Patients were considered as taking TCZ in monotherapy (mono) or combination with DMARDs (combo) based on their BL DMARD co-therapy. Time on TCZ was defined by the time between TCZ start and discontinuation with censoring occurring at date of last visit on TCZ. Crude median TCZ retention with 95% CIs were estimated for mono and combo therapy. The hazard for TCZ discontinuation was modeled using a country-stratified, covariate-adjusted Cox proportional hazards model applied to patients with complete BL covariate information for sex, age, disease duration, number of prior biologics, corticosteroid use, seropositivity (rheumatoid factor or anti-CCP), DAS28, HAQ, and therapy.
Results A total of 1271 eligible treatment courses (TCs) were retrieved from 8 registries by April 2013. Of these, 328 (26%) were started as mono and 943 as combo therapy. For 83% of TCs, DMARD co-therapy was stable over time. In 471 TCs (37%) (of which 124 mono) TCZ was discontinued. Main causes of discontinuation were lack of effectiveness (52% for both therapies) or safety issues (34% and 26% for mono and combo). Crude median TCZ retention was 2.2 years (95% CI: 1.7-3) for mono and 3.5 years (95% CI: 2.7-NA) for combo, (P=0.08). Of the 1271 TCs 856 were included in the country-stratified, covariate-adjusted analysis. Significant results were obtained for seropositivity (HR: 0.64 (pos vs neg), 95% CI: (0.49, 0.84), P=0.001) and HAQ (HR: 1.25 (per unit increase), 95% CI: (1.04, 1.49), P=0.014) at BL. Therapy was not significant (HR: 1.16 (mono vs combo), 95% CI: (0.89, 1.52), P=0.27).
Conclusions In routine care, TCZ retention is better in patients who are seropositive and have lower HAQ at BL, but comparable for mono and combo therapy.
Disclosure of Interest C. Gabay Grant/research support: Roche, M. Riek: None declared, M. Hetland Grant/research support: Roche, E. Hauge Grant/research support: Roche, K. Pavelka Grant/research support: Roche, M. Tomsic Grant/research support: Roche, H. Canhao Grant/research support: Roche, K. Chatzidionysiou Grant/research support: Roche, R. van Vollenhoven Grant/research support: Roche, G. Lukina Grant/research support: Roche, D. Nordström Grant/research support: Roche, E. Lie Grant/research support: Roche, I. Ancuta Grant/research support: Roche, E. Loza Santamaria Grant/research support: Roche, P. van Riel Grant/research support: Roche, T. Kvien Grant/research support: Roche