Replication in genetic association studies: Comment on the editorial by Spector et al.

imab treatment. First, the authors did not specify the time points at which the reconstitution of the naive B cell subset was measured. Rouziere et al found highly mutated memory B cells in the early phase of B cell reconstitution after rituximab treatment in 2 patients with RA (3). Second, the authors did not compare reconstitution of different subsets, but focused on the most prominent subset of naive B cells. The results could be biased because shifts within the different subsets of the B cell population are not interpreted within the context of the “normal” composition in patients with RA. We believe the frequencies of naive B cells should be compared with the frequencies of memory B cells to gain more insight into the possible mechanisms underlying early B cell reconstitution. Thus, although the majority of reconstituted B cells are naive B cells, our results suggest that the reconstitution of B cells in peripheral blood is derived from 2 different, simultaneously occurring pathways: 1) proliferation and differentiation of immature B cells from the bone marrow and 2) migration of memory B cells out of secondary lymphoid organs. The identification of these 2 mechanisms of B cell reconstitution is consistent with the findings from a study of baboons showing incomplete depletion of CD20 B cells in the lymph nodes after rituximab treatment (4). Whether tissue depletion of B cells is complete after rituximab treatment in humans is unknown, but the lower expression of CD20 on B cells from bone marrow and lymph nodes versus peripheral blood indicates that further study is clearly needed (5). Dr. Teng’s work was supported by a grant from the Dutch Organization of Scientific Research.