Structural Lipidomics Uncovers C═C Location-Specific Lipid Signatures and the Response to Immune Checkpoint Inhibitor in dMMR/MSI-H Colorectal Cancer

Lipidomics offers valuable insights for cancer research. Paternò–Büchi (PB) reaction-based LC-MS/MS enables precise lipid structural resolution at the C═C location level. Although mismatch repair-deficient or microsatellite instability-high (dMMR/MSI-H) status predicts anti-PD-1 response in colorectal cancer (CRC), nearly half of the patients do not benefit, underscoring the need for better biomarkers. Here, we applied sequential LC-MS, LC-MS/MS, and LC-PB-MS/MS to profile serum lipids from 58 dMMR/MSI-H CRC patients receiving anti-PD-1 treatment. The resulting lipidomic data were subsequently explored to identify predictive biomarkers for efficacy. We identified 814 glycerophospholipids, 285 of which were resolved at the C═C location level. Comparative analyses revealed 56 differential lipids between patients achieving complete response (CR) and progressive disease (PD) before treatment and 214 after therapy (p < 0.05, VIP >1), most elevated in CR groups. Correlation and clustering patterns indicated coordinated lipid remodeling, and principal component analysis (PCA) demonstrated group separation at both the baseline and post-treatment. Key lipids associated with treatment response included PC(14:0_18:2(Δ9,Δ12)), PG(31:1), PI(39:7), PG(41:7), LPC(24:0), PE(O-42:9), PE(43:5), PC(16:0_20:5(Δ5,Δ8,Δ11,Δ14,Δ17)), LPE(18:2(Δ11,Δ14)), and PG(16:0_20:2) before treatment and PI(44:1), PI(34:1), PC(18:1(Δ11)_20:2(Δ8,Δ14)), PI(44:0), PG(42:4), PC(O-16:0/18:1(Δ10)), PC(15:0_18:1(Δ10)), PC(14:0_18:1(Δ9)), PI(16:0_18:1(Δ15)), and PC(16:0_16:1(Δ7)) after treatment. Overall, LC-PB-MS/MS enables deep structural lipidomics, uncovering lipid signatures capable of stratifying dMMR/MSI-H CRC patients by therapeutic outcome. The identified lipid markers hold promise for monitoring treatment and for developing novel therapeutic strategies.