Abstract B021: Enhancing clinical decision-making: The role of ex vivo drug sensitivity profiling in pediatric precision medicine

Abstract Despite advancements in identifying targetable driver genes in pediatric cancer, precision medicine's efficacy is hindered by inadequate models to promptly predict response to therapy. In vivo studies are costly and subjected to ethical considerations. Additionally, the establishment of patient-derived xenografts (PDXs) and organoids often takes too long to be beneficial for the patient. Thus, preclinical models for precision medicine must be rapid to guide clinical decisions, feasible to establish and reliable in predicting drug response. This study proposes ex vivo short-term drug screening of tumor material as a solution to timely determine tumor drug sensitivities, focusing on solid tumors. The ITCCP4 consortium established 350 pediatric cancer PDX models, from which 50 pan-solid models were selected for ex vivo short-term screening. In addition, PDX models were selected to build a pan-cancer organoid repository. The patient, PDX and organoid models were profiled using whole exome sequencing, low-coverage whole genome sequencing, RNA-sequencing, and methylation arrays. Compound sensitivity was determined for ex vivo short-term and organoid models (140-224 compounds). Ex vivo short-term screening is rapid, as drug sensitivity profiles were determined within 14 days after receiving the primary sample. Moreover, 37 ex vivo pan-cancer short-term screens were successfully performed (74% establishment success), demonstrating the feasibility of the model. The cohort of samples covers nine tumor types, such as neuroblastoma (36%), ewing sarcoma (19%), osteosarcoma (8%) and hepatoblastoma (5%). The protocol demonstrated high reproducibility in drug sensitivity profiling, with tumors from a single model harvested from two distinct PDXs showing a correlation of r = 0.95. As expected, drug sensitivity-based hierarchical clustering of ex vivo short-term screens show aggregation based on tumor type. Consistent with literature, we find significantly lower sensitivity for idasanutlin in TP53 mutated patients compared to TP53 wildtype patients. In addition, 5 PDX-derived neuroblastoma organoids were established (50% establishment success). Interestingly, drug sensitivities of NB ex vivo short-term screens show a strong correlation with organoid screens derived from the same patient (r = 0.72 to 0.96, N. = 4), contributing to the reliability of the model. This study demonstrates that ex vivo short-term screens are a suitable model to guide clinical decisions in pediatric solid tumors. This model is fast enough to benefit the patient, feasible to establish from solid tumors and drug sensitivities correlate with those of organoids and known genetic events. Currently, the technique is implemented in the clinic, where ex vivo short-term screens are performed on patient samples and ex vivo drug sensitivities are correlated with clinical outcome. In conclusion, incorporating ex vivo short-term screens into precision medicine programs accelerates the timely identification of effective therapeutic strategies. Citation Format: Marlinde C. Schoonbeek, Lindy Vernooij, Sarah E.M. Swaak, Vicky J.E. Amo-Addae, Jan Koster, Sander Van Hooff, Selma Eising, Marlinde T.L. Van den Boogaard, Jan J. Molenaar. Enhancing clinical decision-making: The role of ex vivo drug sensitivity profiling in pediatric precision medicine [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Advances in Pediatric Cancer Research; 2024 Sep 5-8; Toronto, Ontario, Canada. Philadelphia (PA): AACR; Cancer Res 2024;84(17 Suppl):Abstract nr B021.