Malignant transformation (MT) of IDH-mutant low grade glioma (LGG) to aggressive high grade tumors is an event of major clinical significance, eventually leading to death in the majority of LGG patients. We discovered that mutations in IDH promote an immunosuppressed microenvironment characterized by decreased production of STAT1-regulated chemokines and low CD8+ T cell infiltration in LGG. In malignantly transformed tumors, we identified the unexpected deletion of the IDH1 mutant allele that may drive counteracting changes to the immunosuppressed microenvironment specifically during MT. LGG that undergo treatment-induced hypermutation (HM), another route to MT, produce more high quality neoantigens. Overall in malignantly transformed tumors relative to patient-matched LGG, we found increasing numbers of T cell clones and increasing expression of genes involved in cytotoxic T cell attraction and effector function. Based on these data, we hypothesize that immunosuppression in IDH mutant LGG is reduced upon MT, driven by genetic alterations that are acquired primarily during malignant transformation. To address this hypothesis, we will quantify spatial and temporal changes in mutant IDH1-driven immunosuppression during MT (Aim 1). We have devised a novel 3-dimensional (3-D), tumor-wide approach in which we will acquire 10 spatially mapped samples per tumor representing maximal anatomy of the tumor. The full cohort will include 30 malignantly tranformed and 30 non-malignantly transformed recurrences from patients for which we have banked samples of the matching initial IDH1-mutant LGG. We will use a high-sensitivity T cell repertoire assay, cytometry by Time of Flight (CyTOF), RNAseq based deconvolution, and multiplex immunohistochemistry to map the immunologic landscape in 3-D, and determine the extent to which mutant IDH1-mediated immunosuppression is reduced during MT.
In Aim 2, we will determine how genetic alterations acquired during MT affect mutant IDH1-mediated immunosuppression. We will perform deep whole exome sequencing on samples collected in Aim 1 to map the intratumoral genomic landscape in 3-D during MT. We will test for the local influence of MT- associated genetic alterations, including high quality neoantigens in hypermutated tumors, deletion of the mutant IDH1 allele, or other genetic events on immunosuppression. Understanding which genetic events contribute to changes in immunosuppression is critical for selecting targeted therapies that could synergize with immunotherapies to prevent or delay MT. To begin to develop T cell based therapies, we will capture neoepitope-specific T cells, prioritizing those that are present tumor-wide, and determine the neoepitopes/HLAs they target and the amino acid sequences for corresponding T Cell Receptor (TCR) ?- and ?-chains. We will then test the cloned TCR for relative target specificity and activity against neoantigen- positive patient-specific tumor cells. The 3-D immuno-genomic landscapes across wide swaths of the tumor will be essential to the design of personalized therapies that have activity against the whole tumor.
PROJECT RELEVANCE The malignant transformation of slow-growing low grade brain tumors into lethal high grade tumors is a devastating event that eventually affects nearly all patients. We discovered that while the mutations in low grade tumors suppress the ability of the local immune system, the mutations that cause malignant transformation might relieve key elements of the suppression, creating an environment that could be more favorable to immune system attack, given the appropriate stimulation by immunotherapy in combination with agents targeting genetic effectors of immunosuppression. We propose to understand how and why the initially immunosuppressed state of the tumor changes during malignant transformation, and use this knowledge to improve immunotherapy, and immunotherapies combined with targeted agents, to prevent, delay or treat malignant transformation.
