ObjectivesTo investigate potential pathogenic pathways in the synovial fluid of osteoarthritis (OA) patients at different disease stages [early vs. late, determined based on the Kellgren–Lawrence (KL) grading scale], through metabolite profiles that were performed by using gas-chromatography/time-of-flight mass spectrometry (GC/TOF MS). MethodsSynovial fluid samples were obtained from 15 patients with knee OA, divided into early- (KL grade: 1 and 2) and late-stage OA (KL grade: 3 and 4). Metabolite profiles of OA based on KL grading scale were performed using GC/TOF MS, with multivariate statistical analyses conducted by orthogonal partial least squares discriminant analysis (OPLS-DA) and hierarchical clustering analysis (HCA). ResultsA total of 114 metabolites were identified and classified into various classes, such as amino acids, sugars and sugar alcohols, fatty acids, and organic acids. Significant discrimination of metabolite profiles between the early- and late-stage OA groups was shown by OPLS-DA and HCA. Twenty-eight metabolites, including malate, ethanolamine, squalene, glycerol, myristic acid, oleic acid, lanosterol, heptadecanoic acid, and capric acid, were identified as critical metabolites for discriminating between the early- and late-OA groups by using Student's t-test, as they showed significant differences in abundance between the two OA groups. These metabolites were related to fatty acid metabolism, glycerolipid metabolism, and the tricarboxylic acid cycle. ConclusionsThese results revealed that metabolite profiles are robustly altered along the radiographic stage of knee OA. Metabolomic approaches based on GC/TOF MS could provide valuable information on the underlying pathogenic mechanisms of OA progression.