P793 Microbial composition in IBD may influence clinical symptoms independent of endoscopic activity
Boland K., Turpin W., Mohammadi A., Borowski K., Kabakchiev B., Smith M.I., Nguyen G., Steinhart A.H., Croitoru K., Silverberg M.
Mount Sinai Hospital, Zane Cohen Center for Digestive Diseases, Toronto, Canada
Symptoms in inflammatory bowel disease (IBD) do not consistently mirror endoscopic activity. The aim of this study is to determine if there is a relationship between mucosal microbiome composition and clinical activity.
IBD patients and healthy controls (HC) were recruited from a tertiary IBD centre on the day of colonoscopy. Clinical and patient demographic data were recorded and mucosal biopsies obtained. Quiescent IBD was defined as Mayo 0 or SES-CD 0–2, mild as Mayo 1 or SES-CD 3–5, moderate to severe as Mayo >2 or SES-CD >6. Clinical activity was defined as partial Mayo score >3 or CDAI >220. Bacterial DNA from biopsies was extracted and the V4 region of 16s rRNA sequenced by Miseq and processed using the QIIME v1.9 pipeline. Alpha diversity was calculated using Chao1 index after rarefaction at 8,500 reads per sample prior to analysis and associations addressed using parametric t-test. Principle coordinate analysis was conducted using Bray-Curtis as the beta diversity metric and significance tested using an Adonis test after 1000 permutation on the first 3 PCoA axes. Taxa analysis was assessed using Kruskal Wallis test or Spearman correlation.
225 IBD patients and 48 HC were recruited. Sigmoid biopsies from patientswith moderate to severe endoscopic disease were analysed initially. Measures of beta and alpha diversity were similar in clinically active patients (n=51) relative to clinical remission (n=32). Taxa analysis showed reduced relative abundance (RA) of Gammaproteobacteria, Coriobacteria and Fusobacteria (0.01<q<0.04) in asymptomatic relative to clinically active patients. Increasing RA of Actinobacteria micrococcaceae correlated with increasing CDAI and partial Mayo scores (p=0.001, coefficient=0.34). In contrast, microbiome composition in quiescent IBD was compared in patients with (n=12) and without clinical activity (n=65) and to HC (n=48). HC had greater Chao1 alpha diversity (q=0.003) when compared to IBD. Among IBD patients, Fusobacteriaceae RA was increased patients reporting symptoms despite endoscopic remission (q=0.02). Symptomatic patients have greater RA of Bacteroidaceae (q=0.004), Fusobacteria (q=0.004) and Actinobacteria (q=0.01) relative to clinical remission.
In IBD patients with and without mucosal healing, increasing clinical symptoms correlated with greater abundance of bacteria including Actinobacteria and Fusobacteria. The relationship between inflammation and intestinal dysbiosis remains unclear. These data suggest an association between specific taxa and clinical symptoms in active and quiescent IBD and may help to further explain why there may be poor correlation between clinical and endoscopic activity.