Multiomic Profiling Identifies High-Risk Markers in Pancreatic Precursors

Large-scale proteomic and glycoproteomic analyses have uncovered new molecular features of pancreatic cancer progression. The study, led by investigators at Johns Hopkins University, applied Clinical Proteomic Tumor Analysis Consortium (CPTAC) mass spectrometry workflows to a diverse cohort of 64 intraductal papillary mucinous neoplasms (IPMNs), 55 pancreatic cyst fluid samples, 104 pancreatic ductal adenocarcinomas (PDACs), and 76 normal pancreatic duct controls.

After initial pathology review, all tissue samples underwent proteomic and glycoproteomic profiling via data-independent acquisition mass spectrometry (DIA-MS). This effort identified over 10,000 proteins and 22,000 glycopeptides across the tissue samples.

Comparative analysis revealed 756 proteins upregulated and 438 downregulated in IPMNs relative to normal ducts. Upregulated proteins were enriched in glycan biosynthesis and mucin-type O-glycan pathways, while downregulated proteins reflected loss of normal pancreatic function. Spatial proteomics and immunolabeling further resolved the cellular origin of these signatures, distinguishing proteins expressed by neoplastic cells from those derived from the surrounding stroma. 

Focusing on disease progression, the study identified 67 proteins overexpressed in high-grade IPMNs. These included markers such as PLOD3, IRS2, and TACSTD2 (Trop-2). By including matched cyst fluids in the analysis, investigators confirmed that several of these proteins are released into the cyst contents. Because cyst fluid can be collected endoscopically, these markers could represent a minimally invasive option for clinicians to classify cysts.

Beyond protein abundance, the researchers observed widespread alterations in glycosylation, particularly involving mucins (MUC1, MUC2, MUC4, MUC5AC, and MUC6), integrins, and immune-related glycoproteins. Specific glycopeptide signatures distinguished IPMNs from normal ducts with high accuracy, achieving area under the curve values approaching 0.9 when used in combination.

Finally, reanalysis of 104 PDAC tumor samples from a previous CPTAC study using the same DIA-MS platform revealed a “PDAC-like” molecular signature in a subset of high-risk IPMNs. This suggests that precursor lesions may acquire key molecular “fingerprints” of invasive cancer before they have fully progressed.

Taken together, these findings highlight the potential of proteomics and glycoproteomics to deepen biological insight into pancreatic precursors while enabling improved risk assessment and improved patient outcomes.