Summary There are an estimated 12,390 cases of glioblastoma (GBM) in the United States with a median survival rate of only 14.6 months despite advances in surgery, chemotherapy, and radiation therapy. The microenvironment of these tumors is often highly infiltrated with immunosuppressive cells, including tumor-associated macrophages (TAMs). Checkpoint blockade therapies rely on the infiltration of tumors with cytotoxic T cells to be effective, and therefore some GBM patients may not respond well to them. Therefore, GBM has been an attractive target for treatment with chimeric antigen receptor (CAR) T cell therapy to overcome immune evasion often observed in GBM patients by producing T cells that respond to the GBM tumor neo-antigen epidermal growth factor receptor (EGFR) variant III (EGVRvIII). Despite some advances in response, the microenvironment can play an important role in modulating the response of tumors to immune modulatory therapies. GBM patients that do not respond to checkpoint blockade therapy often have high levels of TAM infiltration, which might limit the efficacy of CAR T therapy. Previous work has shown that tumor derived extracellular vesicles (EVs) can modulate the microenvironment toward a pro-tumor, immune inhibitory environment. Specifically, they can be taken up by macrophages to drive them toward a pro-tumor, immune-suppressive phenotype. Numerous studies have outlined the ability of exosomes to directly bind T cells and block cytotoxic activity against tumors. A blood-based ‘liquid biopsy’ is ideal to determine which patients will respond to immunotherapy without the need for invasive biopsies. I therefore hypothesize that tumor derived EVs transmit cargo to the immune system regulating therapeutic response to immunotherapy and this cargo can serve as prognostic biomarkers. In the first aim, I will determine the impact of tumor EV secretion on immune infiltrate and immunotherapy efficacy by reducing EV secretion using Rab27a KO. These models will be tested in both humanized mice as well as syngeneic models to test the impact of EV secretion of CAR T cell efficacy and checkpoint blockade therapy (respectively). In the second aim I will use novel microfluidics to capture tumor EVs, measure mRNAs, test if any of these differentially expressed mRNAs are prognostic or predictive biomarkers. I will additionally compare these markers to immune infiltrate in the same patients using multispectral imaging. My proposal focuses on one major question: How EVs regulate the immune system and the tumor response to immunotherapy. Insights into EV mRNA cargo, will both identify patients that will respond to immunotherapy as well as help identify alternative treatments to overcome an immunosuppressive environment.