Abstract
BACKGROUND: Ankylosing spondylitis (AS) is an autoimmune rheumatic disease mostly affecting the axial skeleton. Currently, anti-tumour necrosis factor α (anti-TNF-α) represents an effective treatment for AS that may delay the progression of the disease and alleviate the symptoms if the diagnosis can be made early. Unfortunately, effective diagnostic biomarkers for AS are still lacking; therefore, most patients with AS do not receive timely and effective treatment. The intent of this study was to determine several key metabolites as potential biomarkers of AS using metabolomic methods to facilitate the early diagnosis of AS. METHODS: First, we collected samples of plasma, urine, and ligament tissue around the hip joint from AS and control groups. The samples were examined by nuclear magnetic resonance spectrometry, and multivariate data analysis was performed to find metabolites that differed between the groups. Subsequently, according to the correlation coefficients, variable importance for the projection (VIP) and P values of the metabolites obtained in the multivariate data analysis, the most crucial metabolites were selected as potential biomarkers of AS. Finally, metabolic pathways involving the potential biomarkers were determined using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, and the metabolic pathway map was drawn. RESULTS: Forty-four patients with AS agreed to provide plasma and urine samples, and 30 provided ligament tissue samples. An equal number of volunteers were recruited for the control group. Multidimensional statistical analysis suggested significant differences between the patients with AS and control subjects, and the models exhibited good discrimination and predictive ability. A total of 20 different metabolites ultimately met the requirements for potential biomarkers. According to KEGG analysis, these marker metabolites were primarily related to fat metabolism, intestinal microbial metabolism, glucose metabolism and choline metabolism pathways, and they were also probably associated with immune regulation. CONCLUSIONS: Our work demonstrates that the potential biomarkers that were identified appeared to have diagnostic value for AS and deserve to be further investigated. In addition, this work also suggests that the metabolomic profiling approach is a promising screening tool for the diagnosis of patients with AS.