Ultra-low doses of rituximab for continued treatment of rheumatoid arthritis (REDO study): a randomised controlled non-inferiority trial

Background: Rituximab is an effective treatment for rheumatoid arthritis, given as either two doses of 1000 mg (2 weeks apart) every 6 months (the dose recommended by the US Food and Drug Administration and European Medicines Agency) or two doses of 500 mg (2 weeks apart) or one dose of 1000 mg (a standard low dose) every 6 months. Findings of several small uncontrolled studies suggest that doses lower than the recommended dose or standard low dose might be sufficient for maintenance treatment, potentially improving safety and reducing costs. Therefore, we aimed to compare the efficacy of ultra-low doses of rituximab (one dose of 500 mg or 200 mg) with a standard low dose of rituximab (one dose of 1000 mg) for patients with rheumatoid arthritis who respond to standard doses of rituximab. Methods: The REDO study is a randomised, double-blind, non-inferiority trial done at five centres in the Netherlands. Adults (aged ≥18 years) with rheumatoid arthritis responding well to rituximab were randomly allocated (1:2:2) to receive intravenous rituximab as one dose of either 1000 mg, 500 mg, or 200 mg, respectively. Volumes of all doses were equal to achieve masking. Randomisation lists were computer-generated and stratified by rheumatoid factor or anti-citrullinated protein antibody status (positive or negative) and concomitant use of conventional synthetic disease modifying antirheumatic drugs (yes or no). The primary analysis was a per-protocol hierarchical testing procedure comparing ultra-low doses with a standard low dose (500 mg vs 1000 mg at 3 months, followed by 500 mg vs 1000 mg at 6 months, 200 mg vs 1000 mg at 3 months, and 200 mg vs 1000 mg at 6 months), using a non-inferiority margin of 0·60 on change from baseline in the 28-joint disease activity score based on C-reactive protein levels (DAS28-CRP). The study is registered at www.trialregister.nl, NTR6117. Findings: Between Dec 15, 2016, and Sept 20, 2018, 142 patients were randomly allocated to either 1000 mg rituximab (n=29), 500 mg rituximab (n=58), or 200 mg rituximab (n=55). The 500 mg dose was non-inferior to 1000 mg at 3 months (mean change from baseline in DAS28-CRP, −0·07, 95% CI −0·41 to 0·27) but not at 6 months (0·29, −0·08 to 0·65). Because of the hierarchical testing procedure, non-inferiority could not be tested for the 200 mg dose. 13 patients had serious adverse events, three (10%) in the 1000 mg group, six (10%) in the 500 mg group, and four (7%) in the 200 mg group. The most frequently reported serious adverse events were cardiovascular. No deaths occurred during the study. A significantly lower incidence of infections was seen in the ultra-low-dose groups compared with the standard dose group (1·24 infections per patient-year with the 1000 mg dose vs 0·52 per patient-year with the 500 mg dose and 0·55 per patient-year with the 200 mg dose; rate ratio 0·42, 95% CI 0·21–0·83; p=0·013 for 500 mg vs 1000 mg; 0·44, 0·22–0·88; p=0·019 for 200 mg vs 1000 mg). Interpretation: Our study did not show non-inferiority of ultra-low doses of rituximab for continued treatment of patients with rheumatoid arthritis. Nonetheless, in clinical practice, a strategy with an ultra-low dose of rituximab might be considered after evaluation of risks and benefits, although further studies are needed to establish non-inferiority. Further analyses and a 2-year observational extension are ongoing and should provide further insight into efficacy and safety. Funding: Menzis and Centraal Ziekenfonds