Objectives(1) To plot the trajectory of humoral and cellular immune responses to the primary (two‐dose) COVID‐19 mRNA series and the third/booster dose in B‐cell‐depleted multiple sclerosis (MS) patients up to 2 years post‐vaccination; (2) to identify predictors of immune responses to vaccination; and (3) to assess the impact of intercurrent COVID‐19 infections on SARS CoV‐2‐specific immunity.MethodsSixty ocrelizumab‐treated MS patients were enrolled from NYU (New York) and University of Colorado (Anschutz) MS Centers. Samples were collected pre‐vaccination, and then 4, 12, 24, and 48 weeks post‐primary series, and 4, 12, 24, and 48 weeks post‐booster. Binding anti‐Spike antibody responses were assessed with multiplex bead‐based immunoassay (MBI) and electrochemiluminescence (Elecsys®, Roche Diagnostics), and neutralizing antibody responses with live‐virus immunofluorescence‐based microneutralization assay. Spike‐specific cellular responses were assessed with IFNγ/IL‐2 ELISpot (Invitrogen) and, in a subset, by sequencing complementarity determining regions (CDR)‐3 within T‐cell receptors (Adaptive Biotechnologies). A linear mixed‐effect model was used to compare antibody and cytokine levels across time points. Multivariate analyses identified predictors of immune responses.ResultsThe primary vaccination induced an 11‐ to 208‐fold increase in binding and neutralizing antibody levels and a 3‐ to 4‐fold increase in IFNγ/IL‐2 responses, followed by a modest decline in antibody but not cytokine responses. Booster dose induced a further 3‐ to 5‐fold increase in binding antibodies and 4‐ to 5‐fold increase in IFNγ/IL‐2, which were maintained for up to 1 year. Infections had a variable impact on immunity.InterpretationHumoral and cellular benefits of COVID‐19 vaccination in B‐cell‐depleted MS patients were sustained for up to 2 years when booster doses were administered.