Boosting safety sensitivity and cross-HLA applications of PC-CARs for broad patient coverage

About This Grant

Abstract Despite curative potential, immunotherapies fail most patients due to fundamental barriers: knowledge of tumor antigens and their densities, HLA restriction, manufacturing complexity, and receptor specificity constraints that collectively limit therapeutic reach. We previously pioneered the use of peptide-centric (PC)-CAR T cells for targeting non-immunogenic intracellular oncoproteins, with our first PC-CAR showing early anti-tumor activity in a Phase I trial. Building on our experience and challenges, we have developed a new PC-CAR platform and applied it to PRAME, a cancer-testis antigen broadly expressed across tumor types and absent from normal tissues. Unlike current PRAME therapies restricted to high-expressing tumors and single HLA types, our PC-CAR platform overcomes these limitations through three innovations. First, our PC-CAR architecture delivers >1000-fold greater sensitivity than conventional CARs. Utilizing a highly selective scFv targeting PRAME peptide SLLQHLIGL/HLA-A*02:01, we achieve potent killing across eight cancer types, including low-expression tumors where standard 4-1BB CARs have no activity. Second, we engineered the CROSS-T system (Constant-Region-Orthotype-Switch-System). By expressing TCR constant-region receptors in cells, we eliminate endogenous TCR competition while preserving cells' innate cytotoxicity and low alloreactivity. This dual mechanism enables killing even after HLA loss, a common escape route, and enables off-the-shelf delivery. Third, our DualCAR configuration simultaneously targets peptides presented by two HLA alleles using separate scFvs tethered to each TCR constant chain. We identified additional excellent PRAME peptides that are highly recurrent and will develop new binders for these pHLAs for use in the DualCAR configuration, expanding patient eligibility to over 50% of the population. We demonstrate superior performance of our PC-CAR compared to the best-in-class clinical PRAME TCR (IMA203), with broader T cell engagement independent of CD8. T cells further enhance killing by secreting IFN- that upregulates HLA expression, driving a feed-forward loop in antigen density. Combined with ultra-sensitive detection and cross-HLA targeting, this platform potentially addresses 200,000 patients annually across multiple solid tumor types. This work transforms a widely presented target that has eluded pan-cancer applications into a broadly applicable, off-the-shelf immunotherapy. Animal models are required to evaluate in vivo anti-tumor efficacy, persistence, trafficking, and safety of engineered PC-CAR T cells within complex tumor microenvironments that are inadequately modeled in vitro. These studies are essential for establishing translational feasibility and supporting advancement toward clinical testing. We are advancing toward clinical translation with a product that can potentially redefine CAR applications in solid tumors.