Cancer-Specific Exons as Pan-Tumor CAR-T Targets in Pediatric Solid and Brain Tumor

Abstract Chimeric antigen receptor T-cell (CAR-T) therapy has transformed the management of pediatric B-cell malignancies but has so far failed to replicate this success in solid and central nervous system tumors. The central bottleneck is not effector cell engineering but antigen identification: the spectrum of reliably tumor-restricted surface antigens in children is narrow, and every clinical candidate explored to date, including GD2, B7-H3, HER2, and IL-13Rα2, is shared to some degree with normal tissue. Cancer-specific exons (CSEs), defined as exonic sequences whose inclusion or whose gene-level expression generates a proteoform largely restricted to malignant cells, have recently emerged as a conceptually distinct reservoir of pan-tumor targets. The landmark pan-cancer analysis by Shaw et al of 1,532 pediatric tumor transcriptomes formalized this concept and yielded an interactive resource of 157 surfaceome and matrisome genes including validated candidates such as the EDB domain of fibronectin 1 and COL11A1 suitable for redirection of CAR-T therapy. This review critically synthesizes the biology, discovery pipeline, and engineering implications of CSE-directed CAR-T therapy for pediatric solid and brain tumors, compares this strategy with existing tumor-associated antigen platforms, and appraises the spatiotemporal, proteomic, and manufacturing hurdles that must be resolved before clinical translation. Clinically, CSE targets offer a route toward safer pan-tumor immunotherapy that is biologically rational for a patient population characterized by low mutational burden, immunologically “cold” microenvironments, and a pressing need for curative alternatives to multimodal cytotoxic therapy.