Annual biosimilar rituximab in seropositive RA: a four-year real-world study

This note points to a peer-reviewed paper I co-authored with clinicians at Apollo Hospitals, now published in Clinical Rheumatology.

Citation. Syamasis Bandyopadhyay, Aheli Ghosh Dastidar, Subhendu Datta Bhowmik, Sandip Kumar Chandra, Aloke Dutta, Avishek Mukherjee. Long-term effectiveness and safety of annual biosimilar Rituximab retreatment in seropositive Rheumatoid Arthritis: A 4-year real-world study. Clinical Rheumatology (2026). DOI: 10.1007/s10067-026-08194-9 · also on my ORCID record.

Why the question exists

Rituximab is an established biologic for rheumatoid arthritis when conventional DMARDs are not enough — especially in seropositive disease. In settings where cost constrains access, biosimilar rituximab (bRTX) is often the practical choice. What remains thinner than the originator literature is long-horizon, routine-care evidence on annual retreatment with the biosimilar: does control hold, and does the safety profile stay acceptable, year after year?

That is a scientific question with an engineering flavour. A trial tells you what can work under protocol. A four-year outpatient cohort tells you what actually holds when the clinic is the laboratory.

What we did

We ran a retrospective, single-centre observational cohort of seropositive, biologic-naïve RA patients with inadequate response to conventional DMARDs who received at least four annual cycles of biosimilar rituximab. Each cycle was two intravenous infusions of 1000 mg, fourteen days apart. Disease activity was tracked with DAS28-ESR/CRP, SDAI and ACR response criteria; adverse events were recorded. Longitudinal comparisons used repeated-measures ANOVA.

What held over four years

Thirty-five patients started therapy; 27 completed four years of annual retreatment. The cohort was predominantly female (71.4%), mean age 46, and started from high disease activity (mean DAS28-ESR 7.66).

Across six-monthly follow-ups, reductions in DAS28-ESR, DAS28-CRP and SDAI versus baseline were significant and sustained (p < 0.05). An ACR20 response of ≥ 80% was seen after each treatment course. In this cohort we did not observe serious adverse events, serious infections, or treatment discontinuations attributable to the regimen.

Why I care about this kind of paper

I spend most of my professional life on enterprise systems and AI platforms. Clinical real-world studies are a reminder that evidence is a product with constraints: sample size, follow-up discipline, outcome definitions, and the honesty to state what the design cannot claim. The same discipline — state the hypothesis, measure against it, publish the limits — is what I want from production AI evaluation.

If the paper is useful to you, please cite the journal version rather than this summary. The full methods, statistics and discussion live at the DOI.