Plasma lipidomic fingerprinting enables high-accuracy biomarker discovery for Alzheimer's disease: a targeted LC-MRM/MS approach.

Abstract

Dysregulation of lipid metabolism is increasingly recognized as a key factor in the pathogenesis of Alzheimer's disease (AD). Unfortunately, an accurate lipidomic fingerprints in AD patients' biofluids remains challenging. A comprehensive analysis of plasma samples from 26 patients with AD and 30 healthy individuals was performed using untargeted and targeted lipidomics techniques with strict lipid annotation criteria. By monitoring characteristic fragments per precursor, we achieved precise lipid characterization and quantification for approximately 270 lipid species. Multivariate statistical analysis revealed a distinct lipid profile between AD patients and controls, with 72 lipids significantly altered (FC > 1.5 or < 0.667, VIP > 1, P < 0.05). Notably, a biomarker analysis based on the multivariate exploratory receiver operating characteristic (ROC) curve identified a comprehensive panel consisting of 10 novel lipids as potential markers for AD, achieving 98.2% accuracy with a favorable auxiliary diagnostic value (area under curve of 0.995). Additionally, the higher levels of SM(d18:1/16:0), SM(d18:1/18:1), and LPE 18:0 were strongly correlated with the clinical dementia rating (CDR) and mini-mental state examination (MMSE) scores, underscoring the therapeutic potential of lipid modulation in AD. These findings reveal the intricate relationship between lipid alterations and AD pathology and emphasize the necessity for LC-MRM/MS lipidomics with rigorous criteria in the discovery of reliable biomarkers, enriching our understanding of lipid roles in neurodegenerative processes and informing future mechanistic investigations and drug target development.