OP020 The clinical determinants affect gut microbial profile of inflammatory bowel disease patients
P. Juillerat1*, B. Yilmaz2, R. Wiest1,2, G. Rogler3, A.J. Macpherson1
1Clinic for Visceral Surgery and Medicine, Gastroenterology, Bern, Switzerland, 2University of Bern, Department of Biomedical Research, Bern, Switzerland, 3University Hospital Zurich, Division of Gastroenterology and Hepatology, Zurich, Switzerland
Alterations in gut microbial community of IBD patients still present inconsistency among the published studies and importantly did not completely allow for the unique identification of microbial signatures of CD and UC. We aimed molecularly to profile the intestinal microbiota of phenotypically and genotypically well-characterised Swiss IBD cohort (SIBDC) patients as well as newly recruited IBD patients in Bern as a local replica set including non-IBD subjects.
We deeply characterised the microbiota of UC and CD from 941 biopsy samples of 346 patients from the SIBDCS Cohort and 1254 of biopsy samples of total 397 IBD patients and non-IBD subjects from the Bern cohort, thus building one of the largest cohorts covering sequence data generated. The microbiota composition at the site of biopsy was determined by 16S Amplicon sequencing on the Iontorrent platform. Data were analysed using the QIIME pipeline and correlated with the extensive long-term longitudinal clinical data (patients’ disease trajectory) of the Swiss IBD cohort study.
In PCoA plots, CD and UC disease groups clustered into two distinct groups mostly characterised by altered bacterial composition and lower diversity in CD patients compared with UC patients and non-IBD subjects. Significant differences in taxa representation between the disease groups of each cohort were identified using multivariate association with linear models. This microbial “fingerprint” based on the Swiss IBD cohort and Bern cohort data could then be used in a machine learning algorithm process to correctly predict the CD and UC disease groups with more than 83% success rate. Disease status, disease location/behaviour, and stool consistency were the critically important variables that have effects on shaping the gut microbiota of IBD patients, as assessed in both cohort. Steroids and Anti-TNF agents’ responses, as well of surgery also induced several taxonomic changes, but not the response to thiopurines, methotrexate or 5-ASA.
Our findings revealed that CD and UC are two distinct intestinal disorders at the microbiome level, which could be differentiate based on the microbial profile. A loss of beneficial microorganisms is more associated with CD. However, the observed bacterial dysbiosis in IBD patients is not only associated with disease status itself, it is also directly linked to several clinical parameters associated with the disease trajectory of patients.