Abstract Introduction Patient-derived organoids (PDOs) have emerged over the last decade as representative in vitro replicas of tumor biology, bringing increased accuracy in predicting clinical outcomes based on pre-clinical results. Large and deeply characterized organoid biobanks, covering broad varieties in indications and genetic backgrounds, have been generated that allow large scale (combination) drug testing, reliable biomarker predictions, and improved drug repurposing and patient stratification studies. However, relatively long timelines and high expenses hold the organoids back to live up to their promise. To overcome these hurdles, we have optimized processes, and here present a largescale organoid panel drug screening platform consisting of over 50 organoid models that predicts drug responses across at least seven indications in a robust, but also fast and cost-efficient manner. Methods Colorectal, breast, lung, pancreatic, ovarian, cervix and melanoma organoid models were selected for subtype and driver mutation variety, banked in large batches, and preserved in an assay-ready format to reduce lead-in time and organoid expansion phase. Drug responses of the full organoid panel were characterized with standard of care compounds (cisplatin, gemcitabine and paclitaxel) and AMG510 targeted therapy (KRAS G12C inhibitor, sotorasib) by measuring ATP levels after a 5-day exposure using a 9-dose curve. Experiments were executed using automated liquid handling equipment to standardize procedures and increase consistency. Control variability (CV), Z-factors and assay windows were calculated to assess assay performance, and IC50 values and area under the curves (AUC) were calculated to express drug sensitivity. Experiments were repeated in multiple labs to test for assay robustness. Results The largescale panel screening platform exhibits a high average assay performance, with high Z-factors (>0.6), low intra- and interplate variability (CV <15%) and reproducible IC50 outputs. Data report of the full panel was prepared within 6 weeks of project kick-off; shortening timelines compared to regular organoid screens over 6-fold, from months to weeks. Using IC50 values and AUC, drug sensitivity testing distinguished sensitive from partially and insensitive models, and models with specific genetic features (KRAS G12C vs G12X) behaved as expected when a AMG510 therapy was tested. Conclusion and Discussion Assay-ready organoid technology-based largescale panel screening as presented here has resulted in efficient timeline and cost reductions and will therefore further harness the large potential of PDO technology in the drug development pipeline. Largescale organoid panel screening enables identification of drug synergy by testing drug combinations. With the majority of included models also being available as PDX models, it will facilitate a seamless translation into subsequent in vivo studies. Moreover, combining panel screens with advanced biomarker analysis will allow early patient stratification, further aiding clinical development of novel oncology drugs. Citation Format: Liza Wijler, Annelot Staes, Linda van Seeters, Lama Alhaj Hasan, Bram Herpers, Leo Price, Mariusz Madej, Michiel Fokkelman, Xiaoxi Xu, Marrit Putker. Largescale organoid panel drug screening to short-track clinically-relevant output [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr PR016.
Read full abstract