Cytokine release syndrome (CRS) is a serious and potentially life-threatening complication typically associated with biological drug products. Pre-clinical testing in vitro and in vivo studies using non-human primates had failed to reliably predict CRS. To determine if bone marrow-thymus-liver (BLT) humanized mice with a fully engrafted human immune system or a CD34-humanized mouse model could predict CRS, we tested an anti-CD28 monoclonal antibody (mAb) similar to TGN1412. This TGN1412 analogue (TGN1412A) was initially tested in vitro and found to produce significant dose-dependent increases in cytokine production. For in vivo studies, adalimumab, an anti-tumor necrosis factor-alpha antibody known not to cause CRS, served as a negative control. We evaluated immune cell activation and cytokine expression in three independent experiments. In BLT humanized mice, significant increases in levels of human cytokines were identified in animals treated with anti-CD28 mAb. As expected, CD28+ cell detection was strongly reduced in the anti-CD28 treated group. Increased T cell activation was also observed. The control group did not show reductions in CD28+ T-cells and did not experience increased cytokine levels. Responses by CD34-humanized mice showed no significant differences between adalimumab and anti-CD28 treatment at doses used to test BLT-humanized mice. These results suggest that the TGN1412A produces similar results in vitro to the original TGN1412 monoclonal antibody. The BLT immune humanized mice but not the CD34 humanized mice produce both robust and specific cytokine responses and may represent a pre-clinical model to identify CRS.
Read full abstract