Objective: B cell depletion with the anti-CD20 antibody rituximab has proven efficacy in patients with rheumatoid arthritis (RA). The effects on B cell homeostasis after repeated treatments and the relationship of certain B cell subsets to clinical response or relapse are currently not known.
Methods: In this open-label study, 17 patients with RA refractory to standard therapy were treated with 1 cycle of rituximab. Of these 17 patients, 11 received a second cycle of rituximab therapy. Immunophenotyping was performed before therapy and during B cell recovery.
Results: Twelve of 17 patients showed a good European League Against Rheumatism response after receiving 1 cycle of rituximab therapy. At the time of B cell recovery, the IgD+,CD27+ memory B cell subset was significantly larger (P = 0.019) in the nonresponder group. Within the group of 12 responders, 6 patients, whose disease was characterized by a significantly higher proportion of overall CD27+ memory B cells before therapy, experienced an early relapse (weeks 24-40 posttreatment). Eleven patients were re-treated, again resulting in a good clinical response. B cell reconstitution followed a similar pattern after each cycle. The early reconstitution phase was characterized by immature CD38++,IgD+,CD10+ B cells, whereas the number of naive B cells increased continuously thereafter. The number of memory B cells was still reduced at the time of the second depletion but recovered to levels similar to those following the first cycle of therapy.
Conclusion: Data derived from repeated B lymphocyte depletion with rituximab in patients with RA suggest that analysis of certain memory B cell subsets provides information on efficacy, response, and late as well as early relapse, consistent with the conclusion that targeting memory B cells is a key to its mechanism of action.