Background Since 2006, rituximab (RTX) is approved in combination with methotrexate (MTX) for the treatment of rheumatoid arthritis (RA) after failure of tumor necrosis factor-α (TNFα) blocking therapy. A substantial number of patients, especially among those with severe disease, do not tolerate methotrexate and/or other synthetic DMARDs. Therefore off-label use of RTX in monotherapy or with leflunomide (LEF) is common in daily practice but rarely investigated.
Objectives We compared treatment regimens with RTX+MTX, RTX+LEF or RTX monotherapy regarding safety and effectiveness in a 2-year follow-up, taking discontinuation rates, drop-outs and concomitant use of glucocorticoids into account.
Methods We included patients enrolled in the German biologics register RABBIT between 2007 and 2011 with a minimum of one follow-up visit (N=709). Therapy discontinuation and drop-out processes were examined in Cox-proportional hazard models. Linear mixed effects models were applied to capture the influence of concomitant use of glucocorticoids and of clinical effectiveness, measured by the disease activity score (DAS28). We applied multiple imputation techniques to address missing data.
Results Patients of all treatment variants showed similar clinical baseline characteristics except for a reduced functional capacity (FFbH, p<0.01) and a significantly higher number of comorbidities (p<0.01) in patients treated with RTX monotherapy. We did not find differences regarding therapy discontinuation, the occurrence of serious adverse events or the number and timing of re-treatment with RTX. The average doses of concomitant glucocorticoids over time were similar in RTX+MTX and RTX monotherapy but significantly decreased by 1.1mg/d (p≤0.01) in patients treated with RTX+LEF. The decreases in the DAS28 were analogous between therapy variants and statistically not different (Table 1). Rheumatoid factor negative patients had a two-fold increased hazard to drop-out if treated in RTX monotherapy (2.15, p≤0.01). This exceeded drop-out of rheumatoid factor negative patients in RTX monotherapy raised a spurious additional effect of this therapy variant at month 24 (Table 1). Further significant predictors of disease activity were the numbers of comorbidities and of TNFα failures, i.e. an increase in these predictors led to an attenuated decrease of the DAS28.
Conclusions Off-label uses of RTX in combination with LEF or of RTX in monotherapy are not inferior to RTX+MTX therapy regarding safety, effectiveness and therapy adherence. RTX+LEF or RTX monotherapy are useful alternatives for patients being intolerant to MTX, with the particular feature that RTX monotherapy should be restricted to rheumatoid factor positive patients to ensure therapy adherence.
Disclosure of Interest: A. Richter Grant/research support from: Supported by a joint, unconditional grant from AbbVie, Bristol-Myers Squibb, MSD SHARP & DOHME, Pfizer, Roche, and UCB., A. Strangfeld Grant/research support from: Supported by a joint, unconditional grant from AbbVie, Bristol-Myers Squibb, MSD SHARP & DOHME, Pfizer, Roche, and UCB., P. Herzer: None Declared, E. Wilden: None Declared, A. Bussmann: None Declared, J. Listing Grant/research support from: Supported by a joint, unconditional grant from AbbVie, Bristol-Myers Squibb, MSD SHARP & DOHME, Pfizer, Roche, and UCB., A. Zink Grant/research support from: Supported by a joint, unconditional grant from AbbVie, Bristol-Myers Squibb, MSD SHARP & DOHME, Pfizer, Roche, and UCB.