P26.02.A AN IN VITRO MULTIDIMENSIONAL DRUG ASSESSMENT TOOL FOR PERSONALIZED PREDICTION OF DRUG RESPONSE AND RESISTANCE IN HUMAN GLIOBLASTOMA ORGANOIDS

Abstract Glioblastoma (GBM) is a highly aggressive brain tumor treated with surgery, radiotherapy, and concomitant as well as maintenance temozolomide (TMZ). Despite treatment adherence to this so called Stupp’s regime, patient survival is typically less than 18 months, with high relapse rates. Current pre-clinical drug research is performed with cell line or patient-derived xenografts or with in vitro 2D models. These models lack inter- and intratumoral heterogeneity as well as a tumor microenvironment, thus bench to bed translational issues occur. Using patient-derived glioblastoma organoids (PDOs) the necessary cell heterogeneity can be provid-ed by growing them out of micro-dissected biopsy samples from 3 different GBM tumor sub-regions (pe-riphery, contrast-enhancing zone and tumor core). PDOs are considered in vitro models with high clinical rel-evance to evaluate the anticancer potential of novel preclinical drugs. In this project, we translated the Stupp’s regime to an in vitro multidimensional drug assessment tool using our in vitro GBM model. This tool allows us to monitor drug response over time and predict the performance of preclinical drugs with high predictability. The multidimensional drug assessment tool was initially optimized with TMZ in glioblastoma spheroids and then extended to PDOs. Our assessment included a morphological analysis, 2D & 3D confocal imaging, NGS mutational profiling, apoptotic/necrotic cell fraction assessment, cell metabolic activity assays (XTT), and su-pernatant-based cell toxicity assays (LDH). The apoptotic cell fraction was detected by looking at nuclear morphology in H&E stained tissue slices and detected using a pre-trained AI-driven image readout software. Over the duration of 14 days, GBM spheroid and organoids were treated according to the adapted in vitro Stupp’s regime and TMZ was monitored with our in vitro multidimensional drug assessment tool. Multiple concentrations and time points were evaluated to detected TMZ effect on tumor size reduction, metabolic activity, cytotoxicity, appearance of new drug-induced mutations, and extent of apoptosis and necrosis. GBM models treated with clinically relevant TMZ concentration (100 uM) were approximately 33% smaller after 10 days and metabolic activity decreased by up to 53%. Further preclinical drug candidates will be compared to TMZ performance and clinical data. To conclude, we are confident that this dynamic and multidimensional screening tool using physiologically relevant patient-derived GBM models provides high predictive power in order to assist preclinical anti-GBM drug screening and can address current translational issues persisting in drug development.