Unifying functional experimental systems with clinical predictive technologies for personalized care in PDAC

Pancreatic ductal adenocarcinoma (PDAC) remains a clinical challenge, projected to become the second leading cause of cancer-related death by 2030. Traditional research has historically relied on established PDAC-derived cell lines to establish 2D in vitro models or in vivo xenografts for drug discovery. These approaches, however, fail to encompass the complexity of PDAC and few studies have been translated into game-changing treatments for patients with pancreatic cancer. This review advocates for a paradigm shift through the integration of complementary modeling strategies. We provide a comprehensive evaluation of more sophisticated in vitro systems, such as 3D models (spheroids and organoids), co-cultures and microfluidics, which offer scalable platforms for high-throughput screening and mechanistic analysis. Likewise, we emphasize the indispensable role of in vivo models, including genetically engineered mouse models (GEMMs) for studying stepwise progression and immune evasion, and patient-derived xenografts (PDXs) for capturing human-specific drug sensitivities and intratumoral heterogeneity. Beyond functional models, we explore the clinical utility of transcriptomic profiling to define PDAC subtypes, facilitating real-time patient stratification in landmark clinical trials. We further examine the role of liquid biopsy as a dynamic tool for monitoring minimal residual disease and the emergence of treatment resistance. Finally, we discuss the future of pharmacogenomics, the “tri-genomic” influence of the microbiome, and the integration of multi-omics through artificial intelligence to optimize individual therapeutic outcomes. By unifying these diverse predictive paradigms, we aim to bridge the gap between preclinical discovery and precision oncology, fostering more effective and individualized treatment strategies for all PDAC patients.