TMOD-15. TOWARD BRAIN CANCER ORGANOID-INFORMED PRECISION MEDICINE FOR GLIOBLASTOMA

Abstract BACKGROUND Improved preclinical models for glioblastoma are urgently needed to enhance the translation of research to clinical applications. Patient-Derived Tumor Organoids (PDTOs) offer a promising model that can be generated quickly. Here we explored their feasibility. METHODS Primary (n=16) and recurrent (n=2) PDTOs were established using the method described in Jacob et al. Cell. 2020 by dissecting tumours into small fragments, preserving their cytoarchitecture. Histological features of the parent tumour were compared to the matched PDTOs by a neuropathologist. To assess the molecular fidelity, comprehensive whole genome sequencing compared matched tumor tissue, PDTOs, and corresponding normal blood samples (n=11), along with RNA sequencing of matched tumor tissue and PDTOs (n=10). We assessed PDTO responses to regorafenib (n=18) and conventional TMZ (n=9), lomustine (n=9) and 10 Gy of ionizing radiation (IR) (n=6) using the automated imaging system IncuCyte S3® to evaluate whole spheroid brightfield and invasion areas following drug/IR treatment and assess viability. RESULTS We achieved a 100% establishment rate of primary GBMs using same-day procedures and an 80% rate if held under 3 days. Histological features were retained in the PDTO. Molecular overlap was greater than 90%, along with a high concordance in gene expression. PDTOs exhibited notable resistance to temozolomide and lomustine, however, showed a very modest response to IR and a moderate response to regorafenib. IncuCyte S3® analysis revealed discernible changes in whole spheroid brightfield and invasion area for regorafenib-treated PDTOs. Clinically relevant concentrations of regorafenib reduced GBM cell proliferation (Ki67 staining). Beyond this concentration, regorafenib exhibited efficacy in eliminating GBM cells. CONCLUSION PDTOs recapitulate the molecular and histological features of the original tumor and may enable the prediction of individual patient responses to therapies and guide more effective treatment selection. We are now applying our established methodology to assess a new set of promising drug candidates.