TMIC-23. Glioblastoma subtype influences overall survival and migration of glioblastoma cells and T cells in synegeneic mouse models and are modeled with a novel Brownian Dynamics Tumor Simulator (BDTS)

Abstract Based on transcriptional expression, most studies divide glioblastoma roughly into proneural (PDGFRA), mesenchymal (NF1) and classical (EGFR) subtypes. We created genetically engineered, spontaneous glioblastoma-like tumors in mice by Sleeping Beauty (SB) transposable element. These genetically induced mouse models of different glioblastoma subtypes showed increased survival in the NRAS oncogene-driven mesenchymal-like subtype (median survival: 70.5 days) compared with the PDGFB-driven proneural-like subtype (median survival: 32 days). To investigate the difference in overall survival, the migration of glioblastoma cells and immune cell content was examined. The tumor cells in mesenchymal subtype of glioblastoma-like tumors were more infiltrative and migrated faster. The primary glioblastoma cells from mesenchymal-like mouse models and human patient-derived xenograft (PDX) lines with mesenchymal characteristics both showed significantly higher random motility coefficient (p<0.001) and cell area (p<0.001) compared with proneural-like mouse models and human patient-derived xenograft (PDX) lines (Shamsan et. al, 2022, BioRxiv). Based on these observations, we hypothesized that cytotoxic T lymphocytes (CTLs) contribute to the differential survival in mesenchymal and proneural subtypes of glioblastoma-like mice tumor models, and the difference can be computationally modeled. Endogenous CTL migration profiles were collected and the survival of adaptive immune system deficient mice with glioblastoma-like tumors of different subtypes is ongoing. It is expected that adaptive immune system deficient mice with mesenchymal glioblastoma-like tumors live shorter than wildtype mice. The Brownian Dynamics Tumor Simulator (BDTS) is a three-dimensional computation model for glioblastoma that incorporates a wide range of parameters involved in tumor growth. This computer-based simulator consists of tumor cells and T cells, in which both cell types migrate, proliferate, and undergo apoptosis. The migration profiles of cells under the microscope can help parameterize the simulator and make it more resemble in vivo tumor development. The BDTS successfully modeled tumor growth and predicted anti-migratory therapy responses in glioblastoma.