COMBINATORIAL THERAPY TARGETING TRANSCIPTION AND IMMUNO-METABOLISM IN RECURRENT GLIOBLASTOMA

Abstract Recurrent glioblastomas (GBM) are notoriously difficult to treat, and in spite of aggressive chemo- and radiation therapy they inevitably recur in almost all GBM patients within 14 months following initial diagnosis. To interrogate pathways driving therapeutic resistance, we compared matched primary and recurrent IDH wild type glioblastoma samples from a cohort of patients, using RNA-Seq. Our analyses showed that pathways involved with tumor immune and metabolic reprogramming were up-regulated in recurrent GBMs compared to untreated, primary samples. Based on these findings, we tested the anti-tumor efficacy of over 20 rationally selected targeted therapeutic agents alone and in combination in a high-throughput drug screen assay designed to measure long term cell viability of primary tumorspheres from patients who underwent surgery for recurrent GBM. Top performing drug combinations from these screens were validated in vivo using patient-derived intracranial mouse models of glioma. Our data demonstrated that the combination of the transcriptional/metabolic inhibitor TG02 and the dual PI3K/mTOR inhibitor GDC-0084 significantly prolonged in vivo survival of tumor bearing mice, performing better than either drug alone. Notably, both inhibitors are undergoing human clinical trials as single agents with positive initial safety profiles and superior blood brain barrier penetrance. RNA Seq and functional assays of tumor samples from treated mice showed that only the combination treatment (TG02 + GDC-0084) significantly inhibited expression of immunomodulatory cytokines driving tumor progression, including those identified by us to be overexpressed in recurrent human GBM samples. Overall, these data suggest that combining TG02 and GDC-0084 is an effective treatment for recurrent GBMs. Furthermore, our results support the use of a translational research platform consisting of a personalized pharmaco-genomics testing, using patient-derived tumor samples towards designing more effective treatment for recurrent GBM patients.