Article Text

Download PDFPDF
PTU-048 Correlating urinary metabolic profiles with serological and genetic determinants of inflammatory bowel disease
  1. H R T Williams1,
  2. I J Cox2,
  3. D G Walker1,
  4. H Sato1,
  5. S D Taylor-Robinson1,
  6. S E Marshall3,
  7. T R Orchard1
  1. 1Department of Gastroenterology, Imperial College London, London, UK
  2. 2Department of Imaging Sciences, Imperial College London, London, UK
  3. 3Department of Immunology, University of Dundee, Dundee, Scotland, UK


Introduction Subgroups of inflammatory bowel disease (IBD) patients have specific serological and genetic associations, reflecting abnormal immunogenetic responses to enteric microbes. We have recently shown significant differences between the urinary metabolic profiles of Crohn's disease (CD) patients, ulcerative colitis (UC) patients and healthy control individuals.1 The differentiating metabolites included several strongly influenced by gut microbial metabolism.

Methods It was hypothesised that the urinary metabolic profiles of CD patients would differ depending on possession of 1) anti-Saccharomyces cerevisiae antibodies (ASCA) and antibodies to an E coli outer membrane porin (anti-Omp C), and 2) CD-associated CARD15/NOD2 polymorphisms. It was hypothesised that profiles of UC patients would differ depending on their possession of perinuclear anti-neutrophil cytoplasmic antibodies (pANCA). Urinary metabolic profiles were acquired from 74 CD and 50 UC patients, with nuclear magnetic resonance spectroscopy. Sera were assayed for ASCA IgA and IgG, anti-Omp C (CD) and pANCA (UC) status; CD patients were genotyped for CD-associated CARD15/NOD2 polymorphisms. Multivariate analysis was carried out to investigate differences between the CD cohort depending on ASCA, anti-Omp C and CARD15/NOD2 status, and the UC cohort depending on pANCA status. This involved partial least squares discriminant analysis with orthogonal signal correction (OSC-PLS-DA), with leave-one-out cross-validation of the multivariate models created.

Results OSC-PLS-DA modelling discriminated the profiles of CD patients possessing ASCA (58% of the cohort) from those who did not, with a statistically significant Q2 value of 0.31 (sensitivity 88%, specificity 60%). The major contributory metabolites to this model included the gut microbe-related metabolites hippurate, lower in the group seropositive for ASCA, and trimethylamine-N-oxide (TMAO), higher in this group. No significant distinctions were found between CD urinary profiles analysed according to anti-Omp C or CARD15/NOD2 status, nor UC profiles when stratified by pANCA status.

Conclusion The urinary metabolic profiles of CD patients with ASCA differ significantly from those without. The involvement of hippurate and TMAO in this multivariate model suggests that gut microbial metabolism differs in CD patients possessing ASCA.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.