Molecular subtypes in pancreatic cancer: from academic promise to clinical reality

Pancreatic ductal adenocarcinoma (PDAC) continues to rank among the most lethal malignancies, with five-year survival rates stubbornly below 12%. Over the past decade and a half, transcriptomic analyses have consistently identified molecular subtypes - particularly the classical/progenitor (GATA6-high) and basal-like/squamous (GATA6-low) - that show clear associations with prognosis and treatment response. Yet, despite compelling biological rationale, the clinical adoption of these subtypes has been constrained by limited reproducibility, technical challenges, and practical barriers to implementation. Recent progress in multi-omics integration has deepened our understanding of subtype biology, while also exposing new layers of complexity in classification frameworks. A pivotal study demonstrated that a straightforward immunohistochemical assessment of GATA6 provides strong prognostic information in treatment-naive patients. However, it also underscored a critical challenge: therapy-induced molecular plasticity can compromise the reliability of static biomarkers. Advances in single-cell RNA sequencing and spatial transcriptomics have further clarified the cellular and microenvironmental dynamics underlying subtype heterogeneity, offering new insights on immune contexture and therapeutic stratification. In this review, we synthesize key developments in PDAC subtyping, critically examine translational hurdles, and propose a pragmatic roadmap for clinical implementation that prioritizes validated simplicity, contextual relevance, and real-world utility.