EXTH-79. Functional Precision Medicine In Glioblastoma: Advancing Drug Screening With Patient-Derived 3D Cultures And Advanced Imaging Analyses

Abstract BACKGROUND Glioblastoma (GBM) is the most lethal primary brain tumor, with a median survival of 14.6 months. Resistance to the trimodal standard—resection, radiotherapy, and temozolomide—demands novel, patient-tailored strategies. We, therefore, built a two-tier platform that couples routine 2-D monolayers with physiologically faithful 3-D organotypic cultures derived from fresh surgical specimens. METHODS Four BBB-permeable, FDA-approved inhibitors were selected: three tyrosine kinase blockers (osimertinib, crizotinib, imatinib) and the multi-target anti-angiogenic agent pazopanib. IC₅₀ values were established in LN229 and GBM8401 lines by MTS assays. Mechanistic profiling used BrdU incorporation, cell-cycle/Sub-G1 flow cytometry, Annexin-V/PI staining, ROS quantification, and western blotting of HO-1/Nrf2. Seven patient tumors, spanning glioblastoma multiforme (GBM, n = 5), high-grade glioma (HGG, n = 1), and low-grade glioma (LGG, n = 1), were expanded in parallel in 2-D and 3-D formats. Tissue clearing, light-sheet microscopy, and electron microscopy (EM) provided spatial readouts of Ki-67 and ultrastructure. RESULTS (1) All agents reduced viability dose-dependently and triggered ROS-linked apoptosis. (2) Three specimens were falsely “sensitive” in 2-D yet refractory in 3-D, underscoring flat-culture bias. (3) Drug efficacy proved patient-specific: in 3-D, imatinib was most potent in Patients 1, 3, 5, 6; crizotinib in Patients 2; osimertinib in Patients 3 and 5; and pazopanib in Patients 4 and 7. (4) Total-apoptosis assays mapped peak caspase activity to imatinib in two GBM samples, whereas Ki67 profiling revealed distinct anti-proliferative winners per case. (5) 3-D light-sheet imaging confirmed deep drug penetration and proliferation arrest, exemplified by imatinib in Patient 1 and osimertinib in Patient 5, validating functional readouts and revealing engagement with vascular and hypoxia niches. CONCLUSION Integrating mechanistic assays with patient-matched organoids unmasks actionable kinase dependencies and eliminates misleading 2-D signals. This scalable workflow refines preclinical drug ranking, supports adaptive trial enrollment, and could anchor statewide precision-oncology programs aimed at extending survival for patients with GBM and related gliomas.