SUMMARY/ABSTRACT Despite a tremendous effort in basic science, clinical trials, drug development, and technical advances in surgery and radiation oncology, glioblastoma remains incurable and improvements in overall survival have been marginal. While radiotherapy is still one of the most effective treatment options for glioblastoma, it cannot control the disease over time. This suggests that novel combination therapies are desperately needed to improve radiation treatment outcome for patients suffering from this disease. The studies outlined in this proposal are based on a hypothesis that is backed by our extensive preliminary data and rigorous published data in the literature. The overall hypothesis is that biomarker-based drug selection predicts synergistic lethality of combination therapies in GICs and glioblastoma bulk tumor cell populations, prevents radiation-induced GBM phenotype conversion and allows for individualized optimization of radiotherapy. The three aims of this study will address this aspect of glioma biology using an innovative tool to track GICs and their progeny, while leveraging the unique resources and expertise available at UCLA and the NIH/NCI CTEP portfolio of drugs. Aim 1 will identify compounds in the NCI CTEP portfolio that interfere with radiation-induced phenotype conversion in glioblastoma and develop biomarker profiles predictive of synergistic lethality in combination with radiation. Studies in Aim 2 will optimize combination therapies in vivo. Finally, Aim 3, will use patient avatar studies to validate biomarker-based drug selection in PDX models of glioblastoma.