膵臓がん患者血清の脂質プロファイリングを強化するRP-UHPLC/MS/MS

Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal cancers, mainly due to the late diagnosis and the lack of reliable biomarkers. Lipidomics provides a promising approach for identifying disease-related alterations, but existing methods are often limited to lipid class profiles with insufficient molecular detail. Reversed-phase ultrahigh-performance liquid chromatography coupled to tandem mass spectrometry (RP-UHPLC/MS/MS) offers the possibility to determine lipids at the fatty acyl/alkyl level. Here, we address the need for a validated quantitative workflow that enables accurate and reproducible lipidomic profiling of human serum in the context of PDAC.

We developed and validated an RP-UHPLC/MS/MS method using multiple reaction monitoring, enabling the identification of 455 lipid species from 22 subclasses, with 381 species from 21 subclasses quantified. The workflow included a response factor correction for sterol esters, which markedly improved their quantification accuracy. The application to serum samples from 54 PDAC patients and 55 healthy controls yielded highly reproducible data, with clear group separation observed in both unsupervised and supervised statistical analyses. Dysregulation was most prominent in sphingolipids and phospholipids. Very long-chain saturated sphingolipids (≥C22) were downregulated, while some shorter or unsaturated chains showed mild upregulation. Phospholipid alterations were dominated by species containing polyunsaturated fatty acyls, particularly 18:2 and 20:4, with plasmalogens showing the strongest changes. These structurally resolved findings were further supported by gas chromatography - mass spectrometry analysis of fatty acid methyl esters.

This validated workflow provides comprehensive quantitative coverage across 21 lipid subclasses with the structural resolution critical for biological interpretation. The detailed mapping of sphingolipid and phospholipid dysregulation in PDAC demonstrates that only the fatty acyl level annotation reveals molecular signatures that may reflect specific enzymatic activities or pathways. The method delivers a robust platform for biomarker discovery and mechanistic studies in cancer lipidomics.