Identifying the Determinants of Immunoediting During Glioblastoma Immunotherapy

Project Summary: The most common primary malignant brain tumor, glioblastoma (GBM) remains uniformly fatal despite standard therapy including surgical resection, radiation, and chemotherapy. Immunotherapy, such as immune checkpoint blockade (ICB), has transformed standard of care across multiple solid tumors but has yet to demonstrate treatment efficacy in phase III trials for GBM. Reasons for immunotherapy resistance in GBM include its immunosuppressive tumor microenvironment (TME) and marked intra-tumoral heterogeneity. Indeed, recent studies have elucidated GBM malignant cellular heterogeneity by defining four highly plastic tumor cell states. Patients harboring a predominantly mesenchymal (MES)-like tumor state have the worst prognosis to standard therapy, with non-MES- like GBM tending to shift to the MES-like state at recurrence, reflecting treatment resistance. However, MES-like GBM has also been shown to have high HLA class I and II expression, features that are predictive of ICB response in other tumor types. Supporting the notion that MES-like GBM is particularly immunogenic, I have shown in preliminary work that, despite the opposite for standard therapy, MES-like GBM demonstrates improved survival to anti-PD-1 blockade, with patients shifting away from the MES-like state at recurrence. Moreover, although GBM responders to ICB have been shown to undergo neoantigen immunoediting during treatment, my preliminary data reveal that neoantigen burden is not associated with ICB response. This suggests that other tumor features besides neoantigen availability, such as tumor cell state, dictate neoantigen-directed tumor eradication in GBM. The overarching goal of this work is to identify the determinants of immunoediting during immunotherapy for GBM, and to elucidate whether and by what mechanisms immunotherapeutic modality shapes mechanisms of immunotherapy resistance. I hypothesize that while neoantigen immunoediting during anti-PD-1 monotherapy occurs predominantly of MES-like GBM cells, personalized neoantigen vaccination (‘NeoVax’) enables neoantigen immunoediting agnostic of tumor cell state by upregulating HLA in non-MES-like tumor cells, via IFNγ from NeoVax-induced T cells. Aim 1 will characterize the cell state and TME of tumor cells undergoing neoantigen immunoediting in a cohort of anti-PD-1 treated GBM patients. Bulk genomic analyses will be integrated with a novel single-nucleus spatial transcriptomic approach, Slide-GoTags, that concurrently enables targeted genotyping of neoantigens and TCR-sequencing. Aim 2 will elucidate the mechanisms by which NeoVax enables immunoediting of non-MES-like tumor cells during anti-PD-1 therapy, using patient samples from an ongoing clinical trial. This will entail Slide-GoTags, ex vivo assays for measuring and identifying vaccine-specific T cell clonotypes, and co- cultures involving patient-derived gliomaspheres and CRISPR-edited T cells. Elucidating the mechanistic determinants governing tumor eradication and immune escape during GBM immunotherapy has the potential to improve GBM clinical outcomes by guiding the development of future immunotherapeutic strategies.