Subcutaneous versus intravenous ultra-low-dose rituximab in rheumatoid arthritis: a randomized controlled PK-PD non-inferiority trial

OBJECTIVES: Ultra‐low‐dose rituximab is an effective treatment in RA. S.c. administration may further facilitate rituximab treatment, reduce infection risk and reduce costs. This study aimed to assess non‐inferiority of s.c. vs i.v. rituximab regarding pharmacokinetic and pharmacodynamic outcomes. METHODS: In this randomized, open‐label, non‐inferiority study, RA patients with stable low disease activity on ultra‐low‐dose rituximab (500 mg or 200 mg every 6 months) were 1:1 randomized to receive rituximab 336 mg s.c. or 200 mg i.v., stratified by previous rituximab dose. Primary outcome was non‐inferiority of s.c. to i.v. rituximab, based on the estimated rituximab serum concentration area under the curve during the first 6 months (AUC0‐6m), with a non‐inferiority margin of 0.8. AUCs were estimated using PK modelling. Secondary outcomes included change in DAS28‐CRP compared with a non‐inferiority margin of 0.6. RESULTS: Thirty‐five patients were randomized (17 i.v., 18 s.c.). The geometric mean ratio for AUC0‐6m was 0.99 (90% CI 0.82‐1.20), with minimal bias [mean predication error: i.v. 13.1 (1.86%); s.c. ‐3.40 (‐0.47%)] but moderate random prediction error [root mean square error: i.v. 140 (19.8%); s.c. 142 (19.5%)] for the AUC estimates. Mean change in DAS28‐CRP was non‐inferior for s.c. rituximab (‐0.39, 95% CI ‐0.77 to ‐0.004). CONCLUSION: Our data suggest non‐inferior pharmacokinetics for s.c. rituximab; however, this cannot be definitively concluded due to uncertainty of the AUC estimates. Nevertheless, based on pharmacokinetic and pharmacodynamic parameters, ultra‐low‐dose s.c. rituximab could be an alternative for ultra‐low‐dose i.v. rituximab.