Project Abstract The remarkable clinical activity of adoptively transferred T cells redirected with a chimeric antigen receptor (CAR) specific for the CD19 antigen and encoding co-stimulatory endodomains may change the conventional treatment of B-cell malignancies in the near future. The development of CAR-T cell-based strategies to treat patients with glioblastoma (GBM) remains challenging because this tumor shows heterogeneous expression of targetable tumor-associated antigens, contains stem-cell-like cells that are not usually mirrored by human GBM cell lines used in preclinical models and shapes pro-tumor and immune suppressive environment. In addition, the delivery of CAR-Ts in GBM requires novel approaches to ensure that T cells reach and stay within the tumor environment. We have identified chondroitin sulphate proteoglycan 4 (CSPG4) as highly relevant GBM- associated antigen for targeted immunotherapy since it is highly and strongly overexpressed in over 68% of primary GBM samples and not in normal brain. CSPG4 is highly expressed by human stem-cell like GBM cells such as GBM-derived neurospheres (GBM-NS), and its expression is induced by TNFα released by inflammatory cells surrounding the tumor. Our data indicate that CSPG4.CAR-Ts can control the growth of GBM-NS in vitro and in vivo indicating that this strategy may have significant clinical impact. However, although effective, CSPG4.CAR-Ts do not consistently eradicate GBM-NS. It is our main hypothesis that to eradicate GBM-NS we need further engineering of CSPG4.CAR-Ts to overcome checkpoint inhibition. We have also devised an injectable hydrogel that may ensure gradual and longer release of CAR-Ts injected intratumor. Information generated by the proposal will be instrumental to design the optimal CAR-Ts to eradicate GBM.