Pancreatic ductal adenocarcinoma (PDAC) is a highly heterogeneous cancer with poor prognosis and limited therapeutic options. Bulk transcriptomic profiling has identified two major gene expression–based molecular subtypes: classical/progenitor and basal-like/squamous/quasimesenchymal. These subtypes differ in biological characteristics, differentiation status, drug sensitivity, and clinical outcomes. Advances in single-cell and spatial transcriptomics have further revealed intermediate/hybrid states, as well as distinct subtypes within the tumor microenvironment. These technologies have also uncovered intratumoral heterogeneity, tumor–stroma interactions, and spatially organized transcriptional programs that further shape subtype identity and plasticity, which can shift over time or under therapeutic pressure. In parallel, metabolomic analyses have revealed distinct metabolic subtypes that align with molecular subtypes and highlight subtype-specific metabolic rewiring and vulnerabilities. Furthermore, recent deep learning approaches applied to histopathology allow for high-resolution, morphology-based subtype prediction using Hematoxylin & Eosin-stained slides, providing practical and potentially scalable diagnostic tools. These multi-layered insights are reshaping PDAC taxonomy and enhancing our understanding of how transcriptional, metabolic, and spatial features together define tumor behavior and therapeutic response. This review discusses molecular (transcriptomic), metabolic, and histological subtyping approaches for PDAC, with the aim of enabling practical, cost-effective diagnosis and personalized medicine. By integrating data from recent experimental and clinical studies, we aim to provide a comprehensive and accessible overview of PDAC subtype heterogeneity, which may help guide future subtype-informed therapeutic strategies.
扫码关注我们
求助内容:
应助结果提醒方式:
