IP10 armoring enhances CLDN-18.2–directed CAR-T cell therapy in pancreatic cancer by promoting T cell infiltration and vascular remodeling, with additive benefits from IL-15 co-modification

Introduction Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant tumor with a dismal prognosis, largely due to its desmoplastic and immunosuppressive tumor microenvironment (TME) that restricts effector T cell infiltration and limits chimeric antigen receptor (CAR)-T cell therapy efficacy. Claudin-18.2 (CLDN18.2) is frequently overexpressed in PDAC and represents an attractive therapeutic target. Our previous study demonstrated that IL-15-armored CLDN18.2 CAR-T cells elicit superior antitumor activity in murine models. To further enhance therapeutic efficacy, we investigated whether chemokine modification of CLDN18.2 CAR-T cells could promote tumor infiltration and amplify antitumor activity. Methods The chemokine CXCL10 (IP10) recruits immune cells via CXCR3. We generated CLDN18.2-targeted murine CAR-T cells secreting IP10 alone (H9 CAR-IP10) or in combination with IL-15 (15×10 H9 CAR). Cell phenotypes, cytotoxic activity, and chemotactic capacity were assessed in vitro using flow cytometry, xCELLigence, and Transwell assays. An immunocompetent Panc02-CLDN18.2 subcutaneous tumor model was applied to evaluate in vivo antitumor efficacy, survival, and TME alterations. Results IP10 modification did not affect CAR expression, expansion, phenotype, exhaustion markers, or cytokine secretion. Both H9 CAR T and H9 CAR-IP10 T cells exhibited potent antigen-specific cytotoxicity. IP10-modified CAR-T cells secreted IP10 upon antigen stimulation and demonstrated enhanced CXCR3-dependent chemotactic migration. In vivo , H9 CAR-IP10 T cells enhanced tumor growth inhibition, with 3 of 5 mice surviving by day 47 (including one tumor-free), whereas all controls died by day 41. Mechanistically, IP10 promoted early intratumoral T cell infiltration and, at later stages, reduced CLDN18.2 expression and increased CD31 expression, suggesting tumor vascular remodeling, though survival benefit did not reach statistical significance (p = 0.24). Dual IP10/IL-15 modification further enhanced antitumor activity, producing pronounced survival benefit with 3 of 5 mice tumor-free at day 78 and 90% reduction in mortality risk (p = 0.0069). Discussion In conclusion, our preclinical findings demonstrate IP10 modification enhances CLDN18.2 CAR-T cell efficacy by promoting T cell infiltration and remodeling the vascular microenvironment, and dual IP10/IL-15 modification yields potent and durable antitumor benefits at low doses, offering a promising strategy that warrants further validation in orthotopic and clinically relevant models.