Abstract 5632: PBMC-humanized mouse model for the assessment of cytokine release syndrome caused by checkpoint and bispecific immunotherapy

Immunotherapeutic antibodies and cell therapies have proven to be highly effective cancer therapy for solid tumors, leukemia and lymphomas. The immunotherapy acts in part by stimulating and redirecting the immune system to attack cancer cells, and cytokines can be released during the process. As a consequence cytokine release syndrome (CRS) is a common adverse effect caused by immunotherapy which could be very severe in patients. The animal models and in vitro human PBMC assays presently in use unfortunately can9t reliably predict the CRS in patients. To address the gap between pre-clinical testing and clinical trials we have developed a rapid, sensitive and reproducible humanized mouse model for quantitating CRS. Immunodeficient NSG™ mice were irradiated and injected with human PBMCs intravenously. In general PBMC-engrafted mice had 10-20% human immune cells (hCD45+) in blood with ~70% of the hCD45+ cells being CD3+ T cells and ~20% being CD56+ NK cells. PBMC-engrafted mice were used for treatment within 6 days. We demonstrated that a number of human cytokines including IFN-γ, IL-2, IL-4, IL-6, IL-10 and TNF-α were elevated in blood in the humanized mice treated with monoclonal antibodies and bispecifics including anti-CD3, anti-CD28, anti-PD-1 and BiTE molecules. The cytokine release was dependent on the dose and time of treatment, and PBMCs from every human donor tested were capable of responding to produce cytokines. Notably the amount of cytokines produced varied from donor to donor in >40 different PBMCs tested so far and cytokine levels of >10-fold difference were frequently observed, indicating that the animal model could reveal individual differences in human donors. As a result human donors could be divided into three types as high, medium or low responders. PBMC-engrafted mice implanted with tumors could also be used. Tumor-bearing PBMC-humanized model would be required to evaluate molecules such as BiTE since CRS activity was most evident when both targets for BiTE molecules were present during the assay. PBMC-engrafted NSG™ mice were used for CRS assessment prior to the development of acute graft-vs-host disease (GVHD) in all of the experiments. We were able to reproduce the data using a NSG™ strain [NSG-(Kb Db)null (IAnull)] doubly deficient in murine MHC class I and II molecules, or double knockout (KO) mice. The double KO mice are known to engraft human PBMCs in a manner very similar to NSG™ and have a significantly delayed onset of GVHD [Brehm et al., FASEB J. 33, 3137 (2019)]. Results from the double KO mice further validated cytokine release was not GvHD related and the NSG™ mice is a robust model for CRS assessment of immunotherapy in vivo. We have described a PBMC-humanized model using NSG™ mice that offers rapid assessment of potential risks of therapeutic agents in causing CRS in vivo. The model could differentiate individual human differences based on the cytokine release. Citation Format: Danying Cai, Jing Jiao, Chunting Ye, Hongyuan Yang, Mingshan Cheng, Michael A. Brehm, Dale L. Greiner, Leonard D. Shultz, James G. Keck. PBMC-humanized mouse model for the assessment of cytokine release syndrome caused by checkpoint and bispecific immunotherapy [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5632.