DDDR-26. IDENTIFICATION OF OLIG2 AS A PREDICTOR OF OSIMERTINIB’S EFFICACY AGAINST EGFRVIII+ GLIOBLASTOMA

Abstract

Abstract BACKGROUND Molecular characterization of glioblastomas by The Cancer Genome Atlas revealed molecular abnormalities in epidermal growth factor receptor (EGFR) in over 50% of glioblastomas. These included gene amplification, a large deletion in ligand binding domain (EGFRvIII), kinase duplications and gene fusions. Our laboratory has been focusing on targeting EGFRvIII in glioblastomas and we noted EGFRvIII heterogeneity. We found that growth inhibition of EGFRvIII+ glioblastoma stem cell lines D317 and D10-0171 by osimertinib was different, with D317 being more sensitive. Molecular analyses revealed that D10-0171 has 35-fold higher levels of the transcription factor OLIG2. Here we hypothesize that increased OLIG2 expression confers osimertinib resistance in EGFRvIII+ GBM. METHODS An OLIG2 knock out GSC, D10-0171 (OLIG2-KO) was generated using CRISPR. Efficacy of osimertinib in growth inhibition of D10-0171 (WT) and OLIG2-KO was examined in a subcutaneous xenograft model. Once tumors became visible, mice were treated daily with either 200 μl vehicle or 25 mg/kg osimertinib by oral gavage. RNA-seq and reverse phase protein array (RPPA) analyses were then performed on both GSC’s. RESULTS Assessment of tumor growth of both GSC’s demonstrated that OLIG2-KO cells formed tumors at a significantly lower rate than WT with a rate-based (T/C) ratio of 0.0077. Administration of osimertinib led to tumor growth inhibition in both GSCs but had a greater effect in OLIG2KO than WT, with rate-based T/C ratios of 0.099 and 0.1669, respectively. Transcriptomic analysis of OLIG2KO revealed downregulation of several pathways when compared to WT, including MYC targets V1 and V2, oxidative phosphorylation, and fatty acid metabolism, while proteomic analysis revealed downregulation of several growth-promoting pathways including ErbB and MAPK signaling. CONCLUSION Lower levels of OLIG2 increase osimertinib sensitivity of EGFRvIII+ glioblastoma. We propose a schema for selecting EGFRvIII+ patients who will respond to Osimertinib.