Presented at Society for Laboratory Automation and Screening (SLAS) 2025
Pancreatic ductal adenocarcinoma (PDAC) is one of the most deadly cancers, with few effective treatments available. In particular, the highly dense and fibrous stromal environment contributes to disease progression by limiting drug penetration and increasing resistance to conventional therapies. 3D models such as patient-derived organoids (PDOs) are a powerful tool in creating clinically predictive models, presenting higher physiological relevance compared with immortalized cell lines, and the ability to recapitulate tumor heterogeneity across patients. However, while PDOs are already widely utilized in preclinical oncology studies, many traditional models lack critical aspects of the tumor microenvironment such as tissue-relevant matrix composition, stiffness, and cell-cell interactions, and are also limited in scale. The RASTRUM™ platform addresses this by enabling matrix tunability and establishment of complex 3D co-cultures, and allowing model generation at a scale amenable to drug screening.
Here, we present a 3D in vitro patient-derived PDAC model generated using an innovative RASTRUM™ platform designed to support high-throughput screening. The model was exposed to known chemotherapeutic compounds, with evaluation of novel therapeutic candidates in parallel. The model was also printed into 30x plates from a single PDO passaging step, in one working day. In brief, PDAC PDO monocultures (BMS) or co-cultures with cancer-associated fibroblasts (CAFs; BMS) were bioprinted into RASTRUM matrices in 96- and 384-well plate format using optimized conditions. PDO monocultures were exposed to chemotherapeutic compounds (9-point dose-responses) from day 3–8. Viability was characterised directly post-printing and post-exposure using luminescent (CellTiter Glo) and fluorescent imaging (Live/Dead stain) methods. Intra- and inter-plate metrics for reproducibility and assay quality were calculated using CellTiter Glo, across a subset of plates from the 30-plate series. The PDAC model showed characteristic 3D morphology and consistent viability with low variation over multiple plates. PDOs showed a toxicity response to a subset of test compounds, with high assay quality measured by Z’ factor of positive and negative controls.
Overall, we demonstrate development of an advanced PDAC model exhibiting low technical and biological variation and relevant drug responses in high-throughput. Use of RASTRUM™ enabled scale-up to 30 plates per experiment, equating to over 9000 individual PDAC model replicates. The biologically relevant culture conditions, reproducibility at scale, and compatibility with standard downstream readouts make this model suitable for integration into HTS workflows, or for further investigation of PDAC-relevant phenotypes.