Document resource
Background While chimeric antigen receptor (CAR) T-cell therapy has revolutionized the treatment of hematologic malignancies, its application to epithelial cancers remains limited by unpredictable toxicities. Insights from CD19-directed CAR T cells are only partially informative, as epithelial antigens are embedded within organ-specific innate immune networks. To clarify mechanisms of toxicity, we developed a human-CEA (hCEA) transgenic mouse model expressing physiological levels of antigen in normal tissues, enabling comparison with antigen-negative wild-type littermates to definitively distinguish on-tumor from off-tumor CAR T-cell effects.Methods To evaluate the in vivo effects of hCEA CAR in a murine pancreatic cancer model, we retrovirally transduced Kras-p53-PDX-Cre (KPC) cell lines to stably express human CEACAM5. KPC-hCEA cells were implanted into the pancreas or liver of hCEA transgenic or wild-type mice, and then received 1–3 × 106 CD8+ CAR T cells or control transduced T cells five days post-tumor implantation. Changes in tumor size, distribution of CAR T cells, and histological tissue damage were quantified 10 days after T-cell transfer. We also validated our findings in an independent syngeneic KPC-EpCAM model.Results In this system, CAR T cells recognizing low-density antigen on cancer-free organs such as the lung and colon exhibited robust, antigen-dependent infiltration and activation. In these low-antigen, cancer-free organs, CAR T cells were accompanied by myeloid cell accumulation and release of inflammatory cytokines and chemokines. Among the myeloid populations, neutrophils showed the most dramatic increase in the antigen-expressing organs. Notably, toxicities were absent in antigen-negative hosts and were markedly reduced by neutrophil depletion in antigen-expressing hosts, without impairing CAR T-cell mediated tumor clearance.Conclusions Taken together, we identify neutrophil-driven innate amplification, rather than antigen abundance alone, as a critical determinant of toxicity and establish a platform to dissect mechanisms and evaluate interventions that broaden the therapeutic index of CAR T-cells for solid tumors.