P-193 A Distinct Microbiome Signature Characterizes Patients with Penetrating and Fistulizing Crohnʼs Disease

Abstract

Intestinal dysbiosis resulting in an inappropriate activation of intestinal mucosal immune responses leads to Crohn's disease. While studies have been done regarding the nature of intestinal microbiota in Crohn's disease, what role the intestinal microbiota play a role in initiating, maintaining, and determining the phenotype of Crohn's disease (CD) remains largely understudied. We collected mucosal samples from macroscopically non-inflamed and inflamed sections of resected colon and small intestine (ileum) from 181 CD and 34 non-CD control patients. For 31 of the CD patients, we obtained samples from both the inflamed and non-inflamed regions of the colon and ileum. For each mucosal tissue sample, we performed 16S rRNA sequencing (MiSeq) of the V1 and V2 regions of the intestinal bacteria. Using QIIME and R, we determined operational taxonomic units (OTUs), mean bacterial species diversity (α-diversity), and diversity between individuals with similar disease phenotypes (β-diversity). Differential OTU abundance was identified using DESeq2 (P ≤ 0.01) and differential α-diversity by Kruskal-Wallis tests across disease status, disease behavior, and inflammation status. Bacterial species diversity (α-diversity) declined, but not significantly, when comparing all samples from CD patients to non-IBD controls. Inflammation status had no impact on α-diversity in either colon or ileum. CD patients with penetrating phenotype (including perianal fistulizing disease) had significant lower α-diversity in their colon and trended lower in the ileum compared to non-IBD controls. But, α-diversity in patients without penetrating disease was not significantly different than non-IBD controls. Several genera from the Firmicutes phylum showed increased abundance (Dialister, Dorea, Lactobacillus, Caloramator) and several other Firmicutes and one genus (Collinsella) in the Actinobacteria phylum showed decreased abundance in the non-inflamed colon of CD patients compared to non-IBD patients. When similarly considering the ileum, we only found increases in abundance in one genus (Bifidobacterium) in the Actinobacteria phylum and multiple genera (Streptococcus, Blautia, Oscillospira) from the Firmicutes phylum. CD patients with the penetrating disease showed increased abundance from the Bacteroides (Parabacteroides, Bacteroides genera) and Firmicutes (Streptococcus genus) phylum, and decreased abundance of bacteria from the Firmicutes (Streptococcus, Clostridium, VeillonellaI, Coprococcus, Blautia genera) and Proteobacteria (Haemophilus, Halomonas genera) phylum compared to patients with a non-penetrating and non-fistulizing disease. There was also increased abundance of bacteria from the Vagococcus genus (Firmicutes) in the inflamed ileum of CD patients compared to non-inflamed ileum. Finally, we measured the β-diversity within matched inflamed and uninflamed CD ileal samples using weighted Unifrac. Using this metric, we found that samples clustered better based on patient of origin than on inflammation status. Our study highlights yet to be defined microbial associations with Crohn's disease phenotype. We show that reduction in α-diversity appears to be primarily driven by patients with a penetrating disease phenotype, but not by inflammation. The majority of the shifts seen between penetrating and non-penetrating disease phenotypes seem to heavily involve bacteria from the Firmicutes phylum. Prospective postoperative follow of these patients is underway, which will determine whether this distinct microbiome signature associated with CD phenotype at the time of surgical resection can determine disease recurrence.