Csig-35. Modelling Heterogeneity in Glioblastoma Recurrence to Identify Mechanisms of Therapeutic Resistance

INTRODUCTION Glioblastoma (GBM) is an incurable disease; nearly every tumor recurs within the irradiated tumor bed following standard of care treatment. A multitude of studies have identified therapeutic targets using in vitro and orthotopic xenograft models, however clinical trials derived from even the most promising preclinical data demonstrated no improvement in progression-free or overall survival. GBM demonstrates marked intratumoral heterogeneity and subclonal evolution of treatment-resistant cells is theorized to potentiate tumor recurrence. OBJECTIVE We present a novel in vitro recurrence model to study variation in tumor evolution and therapy resistance, to identify heterogeneous resistance mechanisms and serve as a more faithful preclinical resistance model. METHODS Four patient-derived immortalized glioblastoma (IDH-wild type) spheroid lines underwent fractionated radiation of 2Gy daily for five days (truncated standard of care protocol in humans). These lines (as well as their respective primary, non-radiated counterparts) were maintained in parallel culture until radiated cells overcame senescence (3-6 weeks) and began to proliferate. Paired primary and recurrent lines were compared using western blot, proliferation assay, colony formation assay, radiation response assay, PCR and RNA sequencing analysis. RESULTS Analysis demonstrated variable changes in each spheroid line between primary and recurrence, at both the protein and transcriptional levels. Some recurrent spheroids demonstrated increased survival and proliferation following radiation compared to primary spheroids while others did not. Longitudinally, recurrent cells that were allowed to grow for several weeks following treatment demonstrated the largest changes in gene transcription. CONCLUSION Varying populations of GBM cells demonstrate heterogenous radiation-induced changes at the transcriptomic, protein expression, and phenotypic levels, thus multi-agent therapy is likely indicated to cover these diverse changes at the time of recurrence. Longer timeframe following treatment allowed transformation of gene expression profile, suggesting that preclinical resistance studies should divert to utilization of a longitudinal model to better recapitulate recurrence.