Dottorato in INNOVATION IN IMMUNO-MEDIATED AND HEMATOLOGICAL DISORDERS

Titolo della tesi: Development of GD2- and FAP-Targeted CAR T-Cell Strategies to overcome Tumor and Stroma Barriers in Osteosarcoma

Osteosarcoma (OS) is one of the most aggressive pediatric malignancies, characterised by limited survival rates and poor outcomes in cases of relapsed or metastatic disease. Although chimeric antigen receptor (CAR) T-cell therapy has revolutionized the treatment of hematologic malignancies, its application to solid tumors has been hampered by antigen heterogeneity and the immunosuppressive tumour microenvironment (TME). Within this hostile niche, the stromal compartment, especially cancer associated fibroblasts (CAFs), plays a pivotal role in establishing a dense stromal network that restricts immune cell infiltration, remodels the extracellular matrix, and impairs T-cell recruitment and activation. In this preclinical study, we developed a dual-target CAR T-cell strategy to overcome stromal-mediated CAR T-cells resistance and enhance anti-tumor efficacy in OS models. In particular, we engineered CAR T-cells directed against the disialoganglioside GD2, a tumor- associated antigen highly expressed in OS, and against the fibroblast activation protein (FAP), a type II transmembrane serine protease selectively expressed by activated fibroblasts and OS cells, but typically absent in most normal adult tissues. Transcriptomic analyses of RNA-seq datasets from the St. Jude Cloud–Pediatric Cancer Data Portal confirmed high expression of the GD2-synthase enzyme B4GALNT1 across pediatric solid tumors, particularly in OS, and cytometric analyses identified the disialoganglioside GD2 as a robust and clinically relevant tumor-associated antigen, expressed in over half of primary OS samples and upregulated in established cell lines. Third-generation GD2.CAR T-cells, based on a clinically available construct currently under evaluation for pediatric neuroblastoma (NCT03373097), efficiently eradicated GD2⁺ OS cells in 2D/3D co-culture models (p<0.0001). However, their activity was significantly impaired in OS 3D models containing CAFs (monitored in real-time by Incucyte live cell analysis system), considering a decreased secretion of Granzyme B, IFN-γ, IL-2, and TNF-α (p<0.0001). This resistance was not replicated by CAF-conditioned media, indicating that physical rather than soluble factors promote TME-driven down-modulation of CAR T-cell cytotoxicity. To overcome this barrier, we engineered FAP-targeted CAR T-cells. Both second- and third-generation constructs exhibited potent antigen-specific cytotoxicity and cytokine release against FAP⁺ CAFs (p<0.0001). In vivo, only third-generation FAP.CAR T-cells achieved sustained tumor control and prolonged survival, reflecting the superior persistence of the third-generation construct. Markedly, combinatorial treatment with GD2- and FAP-directed CAR T-cells synergistically eradicated 3D OS/CAF models, restoring full cytotoxic CAR T-cell activity. These findings reveal the stromal targeting as a critical step for overcoming resistance to CAR T-cell therapy in solid tumors. Dual CAR T-cell designs that combine tumor targeting with stroma targeting may define the next frontier in CAR T-cell therapy for OS and other FAP⁺/GD2⁺ cancer.