EXTH-59. POTENT, SELECTIVE, AND BRAIN PENETRANT INHIBITORS OF EXTRACELLULAR DOMAIN EGFR ONCOGENIC MUTANTS EXPRESSED IN GBM DEMONSTRATE EFFICACY IN AN INTRACRANIAL PATIENT DERIVED XENOGRAFT MODEL

EGFR mutations identified in glioblastomas (GBM) occur nearly exclusively at the allosteric extracellular domain (ECD) and constitutively activate oncogenic signaling. Despite wide success in treating tumors expressing EGFR catalytic site mutants, no drug has demonstrated clinical utility against tumors expressing the extracellular domain EGFR mutants. We demonstrate that the family of ECD mutations are not only co-expressed in GBM, but that they all activate the oncogene through a similar disulfide bond-mediated receptor dimerization mechanism. This dimerization occurs independent of ligands and renders the Locked-dimer (LoDi)-EGFR insensitive to agents that target the EGFR kinase domain mutants in NSCLC. The kinase conformation induced by these ECD mutations seen in glioblastomas is both oncogenic and altered from kinase domain mutations, thus necessitating a new approach to targeting. By screening against cells expressing LoDi-EGFR mutants, we have identified the first inhibitors that potently and selectively target LoDi-EGFR mutants versus both canonical active site oncogenic mutants and wild type EGFR. Through an optimization effort, we have identified a novel family of potent and selective LoDi-EGFR mutant inhibitors that effectively penetrate the blood brain barrier (BBB) following oral dosing in preclinical studies. A leading exemplar, BDTX-GBM-001, inhibits the 5 major LoDi-EGFR mutants expressed in GBM with antiproliferative potency of ~10 nM while showing favorable selectivity versus the human kinome. When dosed orally in the intracranial GBM6 patient derived xenograft model at 50, 30, and 15 mg/kg, a dose responsive decrease in tumor growth, as well as a statistically significant increase in survival, were observed. These data support the continued evaluation of rationally designed BBB penetrant inhibitors selectively targeting the common LoDi-EGFR mutants and enable the first chance to fully test the clinical hypothesis of EGFR driver mutants in GBM.