P790 Microbiota profile in pediatric IBD: correlations with phenotype and disease activity

Abstract

Background: Many studies have shown that in active IBD there is a dysbiosis, which could be a cause for a disturbed epithelial barrier function. The pediatric IBD patient, especially at onset of disease, offers a unique opportunity to investigate pathogenetic and particularly microbiological aspects of the disease. Method(s): Children with IBD were enrolled in the period 2013-2014, both patients at diagnosis (treatment naive) and during follow-up. Stool samples were collected and immediately frozen. The micro-biota composition was analyzed through amplification of the V3-V4 regions of the 16S rRNA gene and subsequently sequencing on Il-lumina MiSeq using the 300 bp paired-end protocol. Microbiolog-ical data were correlated to clinical (disease type, phenotype, activity) and laboratory parameters (fecal calprotectin, inflammatory markers). In a subgroup of Crohn's disease (CD) patients microbiota was analyzed before and after a course of exclusive enteral nutrition (EEN). Result(s): In the study period 16 IBD patients were enrolled (median age 12.5 years), 9 with Ulcerative Colitis (UC), 7 with CD. A total of 28 samples was collected and analyzed. The total number of sequences written was 348681 and the total number of input sequences was 94760274. After filtration the median sequence length was 318 bp (range: 214-29867 bp). Comparison between UC and CD samples revealed significance in phylum level diversity for the Firmicutes phylum (p=0.046), more represented in CD than UC. No statistical difference was found comparing active UC versus inactive UC and active CD versus inactive CD at phylum level. At species level, Fae-calibacterium prausnitzii was found to be increased in patients with active CD compared to patients in remission (p=0.02). The relative abundance was compared between samples collected before starting EEN (group pre-EEN, 3 subjects, median age 10.9 years) and post EEN and with the relative abundance of another group of CD patients (4 subjects median age 14.2 years) that had completed an EEN course from at least one year. After 8 weeks of EEN an increase in the relative abundance of Firmicutes (65.9% versus 75%) and a decrease in Proteobacteria (11.8% versus 2.9%) was observed, although this did not reach the statistical significance at the Wilcoxon rank sum test. Conclusion(s): The apparent increase in Firmicutes in CD subjects compared to UC subjects is an interesting result, worth of further exploration. The increased representation of F. prausnitzii in patients with active CD minimize the role of the decrease in the abundance of F. prausnitzii as a possible etiological factor in CD. EEN lead to a significant change in intestinal microbiota that seem to reverse in CD patients after some months from EEN discontinuation.