Untargeted metabolomics analysis of Rheumatoid arthritis patient sera before and after rituximab treatment

Access full-text files




Sweeney, Shannon Renee

Journal Title

Journal ISSN

Volume Title



Background: Rheumatoid arthritis (RA) is an autoimmune disease with no known cure that affects approximately 1.3 million Americans. RA patients suffer from chronic pain and inflammation and are faced with probable disability, reduced life expectancy, and increased risk of several other diseases. In the last decade, biological therapies have revolutionized RA treatment. Although administration of a tumor necrosis factor (TNF) neutralizing agent is the first-line biological therapy, many RA patients show only partial or no clinical response to treatment. Subsequently, anti-B cell, anti-T cell, or anti-IL6 therapies can be evaluated. Streamlining of treatment protocols is necessary to improve patient outcomes. Methods: Serum was collected from 23 active, seropositive RA patients on concomitant methotrexate, at baseline and six months after treatment with rituximab. Based on the American College of Rheumatology improvement criteria, at a level of 20% (ACR20), patients were categorized as either responders or non-responders. An untargeted metabolomics approach was used to characterize the serum metabolome of patients. High resolution one-dimensional ¹H-NMR spectra were acquired using a Bruker Avance 700 MHz spectrometer. In addition, A Thermo Scientific Q Exactive Hybrid Quadrupole-Orbitrap mass spectrometer was used for UPLC-MS/MS of serum lipids. Data processing, statistical analysis, and pathway mapping were performed in MATLAB in conjunction with several metabolomics software packages including, NMRLab, MetaboLab, Chenomx, MetaboAnalyst, MetaboSearch, VANTED, Xcalibur, and Sieve. Results: Based on the ACR20 criteria, at baseline, 14 patients were characterized as responders and 9 patients were considered non-responders. Similarly, 20 patients followed-up at six months, 13 responders and 7 non-responders. Seven polar metabolites and 15 unique lipid species achieved a p-value of less than 0.05 for a two sample t-test prior to treatment with rituximab. Following rituximab therapy, five polar metabolites and 37 lipid species were statistically significant between groups. Pathway analysis of both polar and apolar metabolites revealed metabolic differences between responder and non-responders before and after treatment with rituximab. Conclusion: A clear relationship between blood metabolic profiles and clinical response to rituximab therapy suggests that ¹H-NMR and UPLC-MS/MS are promising tools for RA therapy optimization and acceleration of treatment protocols to improve patient outcomes.



LCSH Subject Headings