FRI0331 Identification and validation of potential metabolomic biomarkers for reliable diagnosis of behcet's disease using gas chromatography with time-of-flight mass spectrometry. (15th June 2017)

• Record Type: Journal Article
• Authors: Ahn, JK; Hwang, J; Cha, H-S; Koh, E-M
• Is Part Of: Annals of the rheumatic diseases. Volume 76(2017)Supplement 2
• Page Start: 612
• Page End: 613
• Publication Date: 2017-06-15
• Subjects: Rheumatism -- Periodicals; 616.723005
• DOI: 10.1136/annrheumdis-2017-eular.3104
• Languages: English
• ISSNs: 0003-4967

Abstract

Abstract : Background: Although many diagnostic criteria have been developed and revised by experts in the field, diagnosing Behcet's disease (BD) is still complicated and challenging. Metabolomic studies can lead to a better understanding of pathophysiological processes and may uncover new diagnostic markers for classification of disease subgroups and assessment of disease activity in rheumatic diseases. No metabolomic studies on serum have been attempted to improve the diagnosis and to identify potential biomarkers of BD. Objectives: The purposes of this study were to investigate distinctive metabolic changes in serum samples of BD patients and to identify metabolic candidate biomarkers for reliable diagnosis of BD using the metabolomics platform. Methods: Metabolomic profiling of 90 serum samples from 45 BD patients and 45 healthy controls (HC) was performed via gas chromatography with time-of-flight mass spectrometry (GC/TOF-MS) with multivariate statistical analyses. Thirty-five patients with BD and age- and sex-matched 35 HCs were randomly selected to form the discovery set, and the validation set is composed of 10 patients with BD and 10 HCs. Results: We identified a total of 104 metabolites in the serum samples of the discovery set. To maximise the discrimination between groups BD and HCs, we applied supervised partial least squared-discrimination analysis (PLS-DA). A PLS-DA score plot of the discovery set indicated that the cluster of BD patients is well separated from HC clusters in component 1. The variation values of the PLS-DA model are R 2 X of 0.246, R 2 Y of 0.913, and Q 2 of 0.852, respectively, indicating strong explanation and prediction capabilities of the PLS-DA model. The PLS-DA models did not show clear and consistent discrimination between groups according to drug administration, indicating that drugs used in patients with BD had a negligible effect on the distinct metabolic profiles discriminating BD patients from HCs. A panel of five metabolic biomarkers, namely, decanoic acid, fructose, tagatose, linoleic acid, and oleic acid were selected based on their variable importance on projection values and adequately validated as putative biomarkers of BD (sensitivity 100%, specificity 97.1%, area under the curve 0.993) in the discovery set and validation set. Principal component analysis showed clear discrimination of BD and HC groups by the five metabolic biomarkers in the validation set. Conclusions: This is the first report on detection of a characteristic metabolic profile of BD and identifying and validating potential metabolite biomarkers in serum from BD patients using GC/TOF-MS. Further studies including a larger number of patients with BD would be warranted to validate the metabolite profiles for clinical use and to ascertain the feasibility of metabolomic analysis in BD. Disclosure of Interest: None declared