Powering up 3D Patient-derived Organoids: An Integrated, Multidisciplinary CRISPR Organoid Platform for Target and Drug Discovery in Oncology

Abstract: A major reason why candidate drug projects fail is due to lack of efficacy in clinical trials. The ability to translate pre-clinical evidence for a drug target into patient benefit is a critical factor for success. However, conventional in vitro 2D cell line models, commonly employed for early target identification and validation, face challenges in terms of translatability as they do not fully recapitulate the complexity of the disease of interest.

Patient-derived organoids have emerged as highly tractable models of human cancers that possess many key disease features lacking from 2D cell line models, including tissue architecture, multicellular composition, and genetic, transcriptomic and epigenetic profiles that closely match the disease of origin. However, in order to fully realise the promise of organoids in early drug discovery screens, there are several outstanding challenges to overcome, including high heterogeneity of organoid cultures, lack of reproducibility, limited scalability and resource-intensive workflows. 

Here we describe how we addressed these gaps and how we are using CRISPR in organoids to discover disease-relevant targets in cancer. Specifically, we have built an integrated CRISPR organoid screening platform complemented with AI-based image analysis and statistical modelling. By employing the use of automated liquid-handling, high-throughput lentiviral CRISPR library production, organoid-specific microwell plates and machine-learning based image analysis we can systematically interrogate the impact of different gene knockouts and drug treatments on organoid growth and viability at scale. With the ability to capture >1000s of individual organoid measurements per plate and perform robust statistical analyses, our objective is to improve the reproducibility and scalability of organoids for target identification and validation.

By applying this platform to oncology projects at AstraZeneca, we address questions related to target safety by determining therapeutic index between patient-matched cancer and healthy organoids. Additionally, we seek to identify novel targets or combination strategies to combat drug resistance by integrating CRISPR knockout in organoids with the treatment of clinically relevant inhibitors. Embracing the use of more complex disease models at this stage of the pipeline instils early confidence in target translatability and allows for better refinement of in vivo strategies, increasing the likelihood of project success and ultimately improving therapies for patients.