223. Autoreactive plasmablasts after B cell depletion with rituximab and relapses in ANCA-associated vasculitis

Background: Autoreactive B cells are responsible for ANCA production in ANCA-associated vasculitis (AAV). Rituximab depletes circulating B cells including the autoreactive ones. We aimed to evaluate changes and associations with relapse of the circulating autoreactive B cell pool following therapeutic B cell depletion in AAV.

Methods: Sequential flow-cytometry was performed on 148 samples of peripheral blood mononuclear cells from 23 patients with proteinase-3 (PR3)-ANCA⁺ AAV treated with rituximab for remission-induction and monitored off-therapy during long-term follow-up in a prospective clinical trial. PR3 was used as ligand to target autoreactive B cells (PR3⁺ B cells). B cell recurrence was considered to occur at the time of the first blood sample with ≥10 B cells/μL after rituximab.

Results: At B cell recurrence, autoreactive B cell frequency among B cells was higher than at baseline (p<0.01). Frequencies of transitional and naïve subsets were higher, while memory subsets were lower at B cell recurrence than at baseline within both autoreactive and total B cells (p<0.001 all comparisons).  At B cell recurrence, frequencies of B cells and subsets did not differ between relapsers and non-relapsers. In contrast, the frequency of plasmablasts within the autoreactive B cell pool (CD19⁺CD27⁺CD38⁺PR3⁺) was higher in relapsers (median [25-75%IQR]; 1.95% [1.315-3.845] vs 0.84% [0.05-1.66], p=0.022) and severe relapsers vs non-severe relapsers (2.165% [1.66-4.315] vs 0.84% [0.1-1.74], p=0.015). Time-to-relapse and time-to severe-relapse were significantly shorter in patients with circulating PR3⁺ plasmablasts higher than the median value of the cohort at B cell reconstitution (1.6%, PR3⁺PB in Figure A; cumulative rate for relapse and severe-relapse represented in Figure B-C). In addition, levels of PR3⁺ plasmablasts higher than baseline were more likely to be found in patients who relapsed within the following 12 months compared to non-relapsers (p<0.05).

Conclusions: The composition of autoreactive B cell pool varies significantly following treatment of AAV with rituximab, and an early enrichment of plasmablasts within the autoreactive pool is associated with future relapses.

Disclosures: PB: employee of Genentech during the conduct of the RAVE trial, received salary and Genentech stock. PM: Consulting: AbbVie, AstraZeneca, Biogen, Boeringher-Ingelheim, BMS, Celgene, ChemoCentryx, CSL Behring, Forbius, Genentech/Roche, Genzyme/Sanofi, GlaxoSmithKline, InflaRx, Insmed, Jannsen, Kiniksa, Kyverna, Magenta, Novartis, Pfizer, Sparrow, Takeda, Talaris. Research Support: AstraZeneca, Boeringher-Ingelheim, BMS, Celgene, ChemoCentryx, Forbius, Genentech/Roche, Genzyme/Sanofi, GlaxoSmithKline, InflaRx. Royalties: UpToDate. CL: Genentech: research support. RS: research funding from InflaRx, Chemocentryx, Roche-Genetech, consulting for Chemocentryx and Roche-Genentech. US, FCF, JHS: research grants from Roche-Genentech.