NMR based Synovial fluid metabolomics analysis of Reactive arthritis, Rheumatoid arthritis and Osteoarthritis

Reactive arthritis (ReA) is an inflammatory arthritis that often develops 2–4 weeks after an extraarticular infection with Chlamydia, Salmonella, Shigella, Campylobacter, and Yersinia species. Synovial fluid (SF) represents the diseased process under pathological conditions and its distinctive metabolic profiles may provide the diagnostic capacity and understanding of the disease state of reactive arthritis (ReA). The aim of this study is to identify characteristic metabolic changes that could differentiate ReA from non-ReA groups. The metabolic profiles of SF collected from ReA (n=58), rheumatoid arthritis (RA, n=21) and osteoarthritis (OA, n=20) patients were measured using NMR spectroscopy and compared using orthogonal partial least-squares discriminant analysis (OPLS-DA). Further, the receiver-operating characteristic (ROC) curve analysis was performed to establish the diagnostic potential of discriminatory metabolites. Finally, the correlation analysis of 38 metabolites between paired SF and serum from ReA patients was done using Pearson rank coefficient (r). The metabolomics profiles in synovial fluid were distinct between ReA, RA, and OA. The OPLS-DA demonstrated a distinctive metabolite profile of synovial fluid of ReA patient compare to RA RA and OA. Fourteen metabolites were obtained by VIP values of greater than 1 for both groups (ReA vs OA and ReA vs RA) and 12 and 6 of them were selected as potential biomarkers by student t-test for ReA vs OA and ReA vs RA groups. We utilized receiver operating characteristic analysis to determine the diagnostic value of each metabolite and identified in ReA compare to RA and OA as potential biomarkers, with an area under the ROC curve >0.8. A panel of six metabolites (NAG, glutamate, glycerol, isoleucine, alanine and glucose from ReA vs OA group) and two metabolites (alanine and carnitine from ReA vs RA group) was identified as a specific biomarker of ReA. Finally, the pearson correlation coefficient (r) between serum and SF ranged from -0.17 to 0.87 for each of the 38 metabolites, and 71% of the relations were significantly positive. Our approach successfully identified SF biomarkers associated with ReA patients that could serve as an efficient tool for early diagnosis of ReA and the role of these biomarker in the pathogenesis of ReA should be explored in future studies. This is the first study that establishes the correlation between metabolic profiles in SF and serum obtained from ReA patients.