Back

Persistent Classical and Atypical Memory B Cells Underlie Heterogeneous Vaccine Responses in Ocrelizumab-Treated Multiple Sclerosis

Curtin, R.; Velmurugu, Y.; Dibba, F.; Hao, Y.; Sreenivasaiah, C.; Khodadadi-Jamayran, A.; Nyovanie, S. T.; Kim, A.; Samanovic-Golden, M.; Mulligan, M. J.; Priest, J.; Cabatingan, M.; Winger, R.; Patskovsky, Y.; Kister, I.; Silverman, G. J.; Krogsgaard, M.

2025-11-04 immunology
10.1101/2025.11.03.686372 bioRxiv
Show abstract

Patients with multiple sclerosis (pwMS) treated with ocrelizumab (OCR), a B-cell-depleting therapy, exhibit heterogeneous humoral responses to SARS-CoV-2 mRNA vaccination. The mechanisms underlying this heterogeneity remain poorly understood. We performed a longitudinal analysis of antigen-specific T and B cell responses in OCR-treated pwMS and non-MS healthy controls following vaccination. Based on post-vaccination anti-Spike IgG titers, pwMS were categorized as super-responders (SR), responders (R), or non-responders (NR). We investigated how immune cell composition, timing of OCR infusion, and lymphocyte subset dynamics influenced humoral response outcomes. While CD4+ and CD8+ T cell populations were largely preserved across all OCR-treated pwMS, distinct differences in the representation of residual B cell composition distinguished responders from non-responders. Notably, CD19+CD27+ classical memory B cells and CD19+CD27-IgD-T-bet+CD11c+CXCR5-DN2-like B cells persisted following OCR infusion and were enriched in the SR group compared to the NR group. Our findings identify persistent memory B cell subsets that escape OCR depletion as key immune correlates and mechanistic mediators of vaccine responsiveness in OCR-treated pwMS, highlighting potential targets to enhance vaccine efficacy in this population. All participating patients were enrolled in clinical trials NCT04843774 and NCT04682548.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.