Assessment of potential off-target reactivity induced by a mAb/B-BiTE complex. (A-C) Fresh human PBMCs isolated from 5 different donors were incubated in the presence of 2.5 μg/mL B-BiTE alone (A), 7.5 μg/mL Poly-Ig with 2.5 μg/mL B-BiTE (B). Much higher concentrations of B-BiTE alone or Poly-Ig with B-BiTE were also tested as indicated (C). Cytokine production by CD8⁺ T cells, CD4⁺ T cells, and NK cells was assessed by intracellular cytokine staining. Human IgG1-based antihuman CD3 mAb (clone OKT3) was used as a positive control. Each dot represents each donor. A paired t test (2-sided) was performed for comparison, and error bars show the SD. ∗P  < .05; ∗∗P  < .01; ∗∗∗P  < .001; n.s., not significant. (D) Ten million freshly isolated human PBMCs were injected IV into NOD/Shi-scid IL-2rγ(null) mice (n = 5). After 4 days of PBMC injection, these mice were administered indicated antibodies (antibody, 30 μg/mL; B-BiTE, 10 μg/mL). A series of human cytokines that were contained in each murine serum were measured by cytokine beads assays. Each dot represents each mouse. The Welch t test (2-sided) was performed for comparison, and error bars show the SD. Each group as well as each of the time points in the group treated with Poly-Ig/B-BiTE complexes were compared. The decrease in cytokines after treatment was not assessed statistically. ∗P  < .05; ∗∗P  < .01; ∗∗∗P  < .001; ∗∗∗∗P  < .0001; n.s., not significant. (E) Antihuman CD3 mAb with the human Fc domain is able to activate T cells and NK cells in the absence of target cells, resulting in induction of off-target reactivity (left). In contrast, the mAb/B-BiTE is able to activate neither human T cells nor NK cells without target cells (right). Poly-Ig, polyclonal immunoglobulin; SD, standard deviation.