Biomarkers identified by serum metabolomic analysis to predict biologic treatment response in rheumatoid arthritis patients. (29th May 2019)

• Record Type: Journal Article
• Authors: Takahashi, Soshi; Saegusa, Jun; Onishi, Akira; Morinobu, Akio
• Is Part Of: Rheumatology. Volume 58:Number 12(2019)
• Page Start: 2153
• Page End: 2161
• Publication Date: 2019-05-29
• Subjects: biomarker -- biologic treatment -- TNF-α -- inhibitors -- abatacept -- rheumatoid arthritis -- metabolomics -- CE-TOFMS; Rheumatism -- Periodicals; Rheumatology -- Periodicals; 616.723005
• DOI: 10.1093/rheumatology/kez199
• Languages: English
• ISSNs: 1462-0324

Abstract

Abstract: Objectives: Biologic treatment has recently revolutionized the management of RA. Despite this success, ∼30–40% of the patients undergoing biologic treatment respond insufficiently. The aim of this study was to identify several specific reliable metabolites for predicting the response of RA patients to TNF-α inhibitors (TNFi) and abatacept (ABT), using capillary electrophoresis-time-of-flight mass spectrometry (CE-TOFMS). Methods: We collected serum from RA patients with moderate or high disease activity prior to biologic treatment, and obtained the serum metabolomic profiles of these samples using CE-TOFMS. The patients' response was determined 12 weeks after starting biologic treatment, according to the EULAR response criteria. We compared the metabolites between the response and non-response patient groups and analysed their discriminative ability. Results: Among 43 total patients, 14 of 26 patients in the TNFi group and 6 of 17 patients in the ABT group responded to the biologic treatment. Of the metabolites separated by CE-TOFMS, 196 were identified as known substances. Using an orthogonal partial least-squares discriminant analysis, we identified five metabolites as potential predictors of TNFi responders and three as predictors of ABT responders. Receiver operating characteristic analyses for multiple biomarkers revealed an area under the curve (AUC) of 0.941, with a sensitivity of 85.7% and specificity of 100% for TNFi, and an AUC of 0.985, with a sensitivity of 100% and specificity of 90.9% for ABT. Conclusion: By metabolomic analysis, we identified serum biomarkers that have a high ability to predict the response of RA patients to TNFi or ABT treatment.