DOP13 Plasma metabolite fingerprint could discriminate inflammatory bowel disease patients from healthy subjects

Abstract

Abstract Background Inflammatory bowel diseases (IBD, Crohn’s disease [CD], ulcerative colitis [UC]) are chronic immune-mediated systemic diseases with unknown aetiology. In the past few years, experimental studies revealed associations of human serum and plasma metabolites’ with IBD with a possible discriminative power between IBD and healthy subjects. Our study focused on the utility of plasma metabolome profiling as a diagnostic marker in patients with different phenotypes. Methods IBD patients were enrolled in a double-centre prospective cohort study. Large number of healthy subjects were included to reduce methodological bias. Blood samples were obtained at baseline and stored at –20°C and plasma metabolites were measured by high performance liquid chromatography coupled with tandem-mass spectrometry (HPLC-MS/MS). Logistic regression, minimum redundancy maximal relevance (mRMR), principal component analysis (PCA), and linear discriminant analysis (LDA) were used to calculate discriminative power. Results Plasma metabolome profile of 26 IBD (13 UC and 12 CD) patients and 168 HC participants were analyzed, and 388 metabolites were measured. 33 and 58% of CD patients and 31 and 39% of UC patients had clinical and endoscopic activity based on clinical (mean pMayo: 2.8±3.4, CDAI: 56±62) and endoscopic (mean PanMayo: 7.4±13.3 and SES-CD: 7.3±8.2) indices. Low plasma level of cysteine (CV=-0.36; p<0.01), triglycerides (CV=-0.19; p<0.01) and hexosylceramide (HexCer(d18:1/23:0); CV=-0.18; p<0.01), while high level of arginine (CV=0.3; p<0.24), hexosylceramide (HexCer(d18:2/23:0); CV=0.2; p=0.02), glutamate (CV=0.19; p<0.01) and lactate (CV=0.16; p=0.01) were significantly associated with IBD. According to mRMR analysis, cysteine (p<0.001), triglyceride (p<0.01), and HexCer(d16:1/22:0); (p<0.001) ranked highest in terms of feature importance (discriminative power). Based on the 16 metabolites considered to be of the highest discriminative power, the healthy and the IBD groups were well separated (PC1=18.2%, PC2=14.7%; Figure 1.). Conclusion Based on our results, plasma metabolite levels may help to discriminate IBD and healthy subjects. Integrative analysis of different biological samples may potentially predict IBD and could promote achieving personalized medicine. Further sample collections and analysis within this study can reveal discriminative metabolite fingerprints among IBD subtypes and correlations with activity as well.