1030 Distinctive Urinary Metabolite Profiles Correlate With Microbial Composition and Are Linked With Endoscopic Post-Operative Disease Recurrence in Crohn's Disease Patients

Abstract

Endoscopic disease recurrence after ileocolonic resection and re-anastomosis in Crohn's disease (CD) patients occurs in a majority of patients. The aim of this study was to examine environmental exposures, the gut microbiome and urinary metabolites in CD patients following ileocolonic resection to determine if specific gut microbes, their metabolites or toxins or environmental factors could be linked with either relapse or remission. Methods: Biopsies were obtained from the neo-terminal ileum after ileocecal resection from CD patients in endoscopic remission (n=20) or in relapse (n=18). Microbial DNA was extracted from biopsies using the QIAamp DNA stool mini kit. Multitag pyrosequencing of the 16S gene was used to characterize the mucosa-associated microbiota, and the RDP 10 classifier used to assign taxonomic groups. Genes encoding for colibactin, C. difficile toxin B, C. perfringens alpha toxin, shiga-like toxin 1 were quantified by qPCR. Urine samples were collected and metabolites measured using 1H NMR. Peak identification was done using Chenomx NMRSuite v7.5 and multivariate analysis performed using MetaboAnalyst 2.0. An environmental questionnaire was obtained that examined early environmental exposures, family history, diet, smoking behaviours, animal exposure and current medication. Principle component analysis (PCA) and support vector machines were used to determine differences between patients with endoscopic relapse and those that remained in remission. Results: Support vector machines applied to total microbial abundance at the genus level and PCA did not show a clear separation between CD patients in relapse or remission. In addition, there were no differences in any toxin encoding genes. In contrast, patients with endoscopic disease recurrence could be separated from patients without ileal inflammation through partial least squares discriminant analysis of urinary metabolites. In addition, correlation analysis showed a positive correlation between urinary metabolites involved in essential processes such as energy production and amino acid metabolism and Bacteroidaceae and negative correlations between amino acid metabolism and Clostridia in the inflamed group. In contrast, the non-inflamed group showed positive correlations between metabolites involved in energy production and Enterobacteriaceae. The most important environmental features associated with disease relapse included smoking cigarettes and/or pipes. Age, gender, years since last surgery, domicile location, or use of immunosuppressants did not differ between CD patients suffering relapse or remaining in endoscopic remission. Conclusions: Recurrence of Crohn's disease after surgery was associated with a unique urinary metabolomics signature that could be linked with specific gut microbes. The major risk factor associated with recurrence of disease was smoking.