Longitudinal Plasma Lipidomic Changes in Chinese Patients with Nonvalvular Atrial Fibrillation During 12 Weeks of...

Introduction: Altered lipid metabolism is implicated in atrial fibrillation, but longitudinal lipidomic changes during anticoagulation remain poorly characterized. Objectives: To characterize plasma lipidomic changes during 12 weeks of edoxaban treatment and explore associated metabolic pathways in Chinese patients with nonvalvular atrial fibrillation (NVAF). Methods: This prespecified substudy of a prospective multicenter study analyzed fasting predose plasma collected at baseline and week 12 from 74 patients receiving protocol-assigned edoxaban doses (15 mg, n = 3; 30 mg, n = 23; 60 mg, n = 48). Widely targeted UPLC-MS/MS lipid profiling, with laboratory quality control (QC CV < 30%), yielded 1099 lipid features retained for statistical analysis. Differential lipids were defined by paired testing with Benjamini-Hochberg correction (|log₂FC| > 0.58 and q < 0.05). Results: The overall, 30 mg, and 60 mg analyses identified 25, 15, and 35 FDR-significant features, respectively; none was significant in the exploratory 15 mg group. Twelve features were shared across the primary analyses and increased at week 12; they comprised mainly acylcarnitines, together with docosadienoic acid and DG (21:0/16:0/0:0), and four lacked HMDB identifiers. The corresponding FDR panels yielded participant-grouped out-of-fold AUCs of 0.918 (95% CI 0.890-0.963), 0.786 (0.701-0.890), and 0.891 (0.835-0.950) for discriminating baseline from week 12, respectively. Pathway mapping identified no multi-compound enriched pathways (q < 0.05). Observed annotations represented single-metabolite signals-most notably DG (21:0/16:0/0:0) (HMDB0094301) and octanoylcarnitine (HMDB0000791)-rather than coordinated, pathway-level alterations. Conclusions: Paired analyses identified consistent longitudinal plasma lipidomic changes in patients receiving edoxaban. The 12 shared features provide focused candidates for targeted validation. Controlled studies are needed to determine whether these changes are specific to edoxaban and to establish their biological and clinical relevance.