Pancreatic ductal adenocarcinoma (PDA) is a malignancy of the exocrine pancreas with the worst prognosis among all solid tumours, and soon to become the second leading cause of cancer-related deaths. A more comprehensive understanding of the molecular mechanisms underlying this disease is crucial to the development of diagnostic tools as well as to the identification of more effective therapies. High-frequency mutations in PDA occur in "undruggable" genes, and molecular subtyping based on bulk transcriptome analysis does not yet nominate valid therapeutic intervention strategies. Genome-wide sequencing studies have also demonstrated a considerable intra- and inter-patient's genetic heterogeneity, which further complicate this dire scenario. More than in other malignancies, functionalization of the PDA genome and preclinical modelling at the individual patient level appear necessary to substantially improve survival rates for pancreatic cancer patients. Traditional human PDA models, including monolayer cell cultures and patient-derived xenografts, have certainly led to valuable biological insights in the past years. However, those model systems suffer from several limitations that have contributed to the lack of concordance between preclinical and clinical studies for PDA. Pancreatic ductal organoids have recently emerged as a reliable culture system to establish models from both normal and neoplastic pancreatic tissues. Pancreatic organoid cultures can be efficiently generated from small tissue biopsies, which opens up the possibility of longitudinal studies in individual patients. A proof-of-concept study has demonstrated that patient-derived PDA organoids are able to predict responses to conventional chemotherapy. The use of this three-dimensional culture system has already improved our understanding of PDA biology and promises to implement precision oncology by enabling the alignment of preclinical and clinical platforms to guide therapeutic intervention in PDA.
Keywords: PDA; organoids; precision oncology; preclinical models.