Abstract
BACKGROUND: The prognostic and pathophysiologic significance of the biliary microbiota in pancreaticobiliary malignancies is little understood. Our goal was to find malignancy-related microbiomic fingerprints in bile samples taken from patients with benign and malignant pancreaticobiliary diseases. METHODS: Bile specimens were collected from consenting patients during routine endoscopic retrograde cholangiopancreatography. We used PowerViral RNA/DNA Isolation kit to extract DNA from bile specimens. The Illumina 16S Metagenomic Sequencing Library Preparation guide was used to amplify the bacterial 16S rRNA gene and create libraries. QIIME (Quantitative Insights Into Microbial Ecology), Bioconductor phyloseq, microbiomeSeq, and mixMC packages were used for post-sequencing analysis. RESULTS: Of 46 enrolled patients, 32 patients had pancreatic cancers, 6 had cholangiocarcinoma and 1 had gallbladder cancer. Rest of the patients had benign diseases including gallstones, and acute and chronic pancreatitis. We used multivariate approach in mixMC to classify Operational Taxonomic Units (OTUs). Doing this, we found a predominance of genera Dickeya (p = 0.00008), [Eubacterium] hallii group (p = 0.0004), Bacteroides (p = 0.0006), Faecalibacterium (p = 0.006), Escherichia-Shigella (p = 0.008), and Ruminococcus 1 (p = 0.008) in bile samples from pancreaticobiliary cancers as compared to benign diseases. Additionally, bile samples from patients with pancreatic cancer exhibited a predominance of genus Rothia (p = 0.008) as compared to those with cholangiocarcinoma, whereas bile samples from patients with cholangiocarcinoma exhibited a predominance of genera Akkermansia (p = 0.031) and Achromobacter (p = 0.031) as compared to those with pancreatic cancers. CONCLUSIONS: Both benign and malignant pancreaticobiliary diseases have distinct microbiomic fingerprints. The relative abundance of OTUs in bile samples varies between patients with benign and malignant pancreaticobiliary diseases, as well as between cholangiocarcinoma and pancreatic cancer. Our data suggest that either these OTUs play a role in carcinogenesis or that benign disease-specific microenvironmental changes differ from cancer-specific microenvironmental changes, resulting to a clear separation of OTU clusters. We need more research to confirm and expand on our findings.