Designing Safer and More Efficacious CAR T-Cells Against Solid Tumors

Abstract

The two projects described in this thesis sought to tackle two distinct aspects attributed to the general failure of Chimeric Antigen Receptor T cells (CAR Ts) against solid tumors. The first is the possibility of on-target off-tumor toxicity. Epidermal growth factor receptor (EGFR) is a promising target based on its overexpression in a variety of cancers, but its presence on healthy tissue led to severe toxicities through clinical trials of anti-EGFR CAR T cells. Due to murine EGFR (mEGFR) sharing 88% amino acid sequence with human EGFR, we decided to generate a syngeneic mouse model to replicate and help predict toxicities seen clinically. To model potential effects of these CAR T toxicities, we have developed a system using both CAR T cells and tumors derived from syngeneic C57BL/6 mice. Since all cells originate from the same genetic background, we can observe the effects of our anti-mEGFR CAR T cells under an intact immune system. These findings underscore the importance of our model predicting CAR T cell-mediated toxicities. Although still early, this system may be used as an effective model of other potential CAR T-cell toxicities against antigen targets that are comparable between human and mice. The second project sought to resolve the problem of the immunosuppressive tumor microenvironment (TME) perpetrated in pancreatic ductal adenocarcinoma (PDAC). We developed a bicistronic lentiviral vector for a CAR T cell that additionally secretes a bispecfic T-cell engaging Antibody Molecule (TEAM) denoted as CAR-TEAM. This two-target approach seeks to eliminate the surrounding Cancer Associated Fibroblasts (CAFs) that play a large role in immunosuppression of immune cells in clinical PDAC cases and increases the anti-tumor efficacy. Based on in vitro results, our CAR-TEAM system may be a clinically relevant immunotherapy for PDAC.