Relapse risk after rituximab discontinuation in neuromyelitis optica spectrum disorders: a multi-center retrospective cohort study.

OBJECTIVE

We aimed to evaluate the association between rituximab (RTX) treatment duration and relapse risk, and explore clinical outcomes following treatment discontinuation in neuromyelitis optica spectrum disorder (NMOSD).

METHODS

We retrospectively collected data from rituximab-treated patients (>1-year follow-up after treatment initiation) with NMOSD at five major clinical centers in China between 2016 and 2023. The main outcome measures were changes in relapse risk based on the RTX treatment duration and clinical outcomes following relapse after RTX discontinuation. The Andersen-Gill model was used to analyze treatment duration-relapse risk associations.

RESULTS

In total, 106 rituximab-treated patients were included (40 patients discontinued and 66 continued RTX). Longer RTX treatment significantly reduced relapse risk (hazard ratio [HR]=0.43, P < 0.001). Among 28 patients who discontinued RTX and were followed-up for >1 year after drug withdrawal, 53.6% (15/28) relapsed at a median interval of 14 months. Patients with >2 years of RTX treatment exhibited lower annual relapse rate (ARR) (mean ARR: 0.97 vs. 0.28, P = 0.020) and less severe relapses (mean ΔEDSS score: 1.42 vs. 0.50, P = 0.021) after discontinuation, compared to pre-treatment levels. Only one of four (25%) longer-treated and clinically stable patients (≥2 years of RTX treatment and no relapses ≥2 years before discontinuation) experienced a non-severe relapse by 50 months. Patients aged >55 years in the discontinuation group had the lowest post-discontinuation relapse rate.

INTERPRETATION

RTX treatment duration strongly correlated with reduced relapse risk in NMOSD patients. Relapse risk after RTX discontinuation was lower than that after discontinuing traditional oral immunosuppressants, particularly in patients with longer treatment and clinical stability.