Identifying Robust Biomarkers for the Diagnosis and Subtype Distinction of Inflammatory Bowel Disease through Comprehensive Serum Metabolomic Profiling

Abstract Background Inflammatory Bowel Disease (IBD), encompassing Crohn's disease (CD) and ulcerative colitis (UC), presents diagnostic challenges owing to overlapping clinical presentations. This study aimed to delineate specific serum metabolomic biomarkers that differentiate IBD patients from healthy controls and further discriminate between CD and UC. Methods We enrolled a total of 346 participants, including 134 with CD, 124 with UC, and 88 normal controls (NC). Serum samples and their clinical metadata were systematically collected. Untargeted profiling was performed with Gas Chromatography-Time-Of-Flight-Mass Spectrometry, and targeted profiling of bile acids and tryptophan used Liquid Chromatography-Triple Quadrupole-Mass Spectrometry. The identification of distinct metabolites and potential biomarkers of IBD patients from NC and that of CD patients from UC were achieved through extensive univariate and multivariate statistical analyses which supplemented by Receiver Operating Characteristic (ROC) curves, pathways, and network analyses. Results Distinct clustering separated IBD patients from the NC, although the CD and UC subgroups overlapped in the non-targeted profiling. Targeted metabolomics revealed elevated tryptophan and indole-3-acetic acid levels in CD and UC patients. Increased kynurenine and indole-3-propionic acid levels were unique to CD, whereas UC was characterized by decreased indole-3-acetic acid, serotonin, and acetylcholine levels. Both IBD subtypes exhibited reduced primary-to-secondary bile acid ratios compared with the NC. The ROC analysis underscored the discriminatory power of the biomarkers (AUC values: NC vs. CD = 0.9738; NC vs. UC = 0.9887; UC vs. CD = 0.7140). Pathway analysis revealed alterations in glycerolipid metabolism, markedly differentiating UC from CD. Beta-alanine, arginine, and proline metabolism were linked to IBD compared to NCs. Network analysis correlated metabolomic markers with the clinical phenotypes of IBD. Conclusion Serum metabolomic biomarkers offer promising avenues for the diagnosis and subtype differentiation of IBD. Targeted metabolomics analysis is critical for distinguishing CD from UC.