Natalizumab differentially affects plasmablasts and B cells in multiple sclerosis

Natalizumab treatment increases the frequencies of B cells in blood but reduces IgG in blood and CSF. Plasmablasts are important in the production of IgG, and the development of plasmablasts is CD49d dependent.

We hypothesized that natalizumab treatment affects the development of plasmablasts.

We retrospectively analyzed frequencies and absolute counts of B cell subsets by flow cytometry from a longitudinal cohort of 9 progressive multiple sclerosis (MS) patients treated with natalizumab for 60 weeks, and a cross-sectional relapsing-remitting MS (RRMS) cohort with 17 untreated and 37 treated with natalizumab (17 stable and 20 unstable patients with relapse activity). Additionally, CD49d expression on B cell subsets was examined in 10 healthy controls, and blood and cerebrospinal fluid (CSF) frequencies of B cell subsets were quantified in untreated and natalizumab treated RRMS patients.

In progressive MS, levels of IgG decreased in plasma (p<0.001) from baseline to 60 weeks follow-up. In the progressive MS and RRMS cohorts we observed that natalizumab treatment significantly increased the frequency of B cells (p=0.004; p<0.0001) and several B cell subsets, most pronounced for memory B cell subsets (p=0.0001; p<0.0001), while there was a decrease in plasmablast frequency (p=0.008; p=0.008). In both progressive MS and RRMS the absolute cell counts of B cells increased (p=0.004; p<0.001), which was explained by a significant increase in all subsets, except for plasmablasts. Furthermore, we found decreased memory B cell counts in unstable compared to stable natalizumab-treated patients (p=0.02). The expression of CD49d was higher on plasmablasts compared to other B cell subsets (p<0.0001). In CSF, plasmablasts could not be detected in patients treated with natalizumab, in contrast to an increased frequency in untreated RRMS patients.

We confirm previous studies showing that natalizumab increases circulating number of B cells, particularly memory cells, concomitant with a decrease in plasma IgG concentrations. Moreover, we demonstrate in two separate cohorts that natalizumab treatment markedly decreases frequencies of plasmablasts while the absolute number is stable. Additionally, plasmablasts have high expression of CD49d, and plasmablasts could not be detected in the CSF of natalizumab-treated patients. Finally, memory B cells were found to be reduced in unstable natalizumab-treated patients, which could possibly indicate increased recruitment to the CNS.