CD20- and CD19-BsAb induce concentration-dependent killing of B cells and expansion of T cells. Lymph node–derived lymphocytes were incubated without (w/o) or with anti-CD3 antibody (CD3-Ab), 4 different concentrations (C1-C4) of CD19-BsAb (n = 33 biologically independent samples), or CD20-BsAb (n = 29 biologically independent samples). Cells were harvested after 7 days and analyzed by using quantitative flow cytometry. Shown are the percentages based on the absolute numbers of viable B cells (A, B) or the x-fold expansion based on the absolute numbers of viable T cells (D, E) normalized to w/o. P values were calculated between w/o and every other condition using the two-sided Wilcoxon’s test. The mean reduction of viable B cells (C) or the mean x-fold expansion of T cells (F) across all concentrations was correlated for the treatment with CD19-BsAb (y-axis) and CD20-BsAb (x-axis). Spearman’s correlation coefficients (R) are given. (G) Histogram showing the mean percentage of viable B cells of CD19-BsAb– or CD20-BsAb–treated (purple) and untreated/control (gray) samples. Solid lines represent corresponding density curves. Dashed vertical line at 76% indicates the point of intersection of both density curves. (H) Bars represent the mean percentage of viable B cells per patient after exposure to BsAbs across all concentrations. Green or gray bars highlight responders or nonresponders, respectively. (I) Shown is the frequency of T cells at baseline for both response groups. If not otherwise indicated, P values were calculated by using the two-sided Wilcoxon’s test and adjusted according to the Benjamini-Hochberg procedure. ***≙P ≤ .001; **≙P ≤ .01; *≙P ≤ .05; missing asterisks indicate P > .05.