Duodenal and pancreatic tissue microbiome profiles of PPI users and non-users

Abstract

Factors that influence the pancreas microbiome are not well understood. Regular proton pump inhibitor (PPI) use induces significant alterations in the gut microbiome, including an increase in the abundance of Streptococcus, and may be associated with pancreatic cancer risk. The aim of this study was to examine whether PPI use is associated with pancreatic and duodenal tissue microbiomes. We compared 16S rRNA microbiome profiles of normal pancreatic and duodenal tissue from 103 patients undergoing pancreatic surgery for non-malignant indications, including 34 patients on PPIs, accounting for factors including age, smoking, body mass index and the presence of main pancreatic duct dilation. Histologically normal tissue from the pancreatic head had higher alpha diversity and enrichment of Firmicutes by phylum-level analysis and Streptococcus species compared to normal pancreas body/tail tissues (16.8 % vs 8.8 %, P = .02, and 5.9 % vs 1.4 %, P = .03, respectively). Measures of beta diversity differed significantly between the pancreas and the duodenum, but in subjects with main pancreatic duct dilation, beta diversity of pancreatic head tissue was more similar to normal duodenal tissue than those without pancreatic duct dilation. Duodenal tissue of PPI users had significant enrichment of Firmicute phyla (34.7 % vs. 14.1 %, P = .01) and Streptococcus genera (19.5 % vs. 5.2 %, P = .01) compared to non-users; these differences were not evident in pancreas tissues. By multivariate analysis, PPI use was associated with alpha diversity in the duodenum, but not in the pancreas. However, some differences in pancreas tissue beta diversity were observed between PPI users and non-users. In summary, we find differences in the microbiome profiles of the pancreas head versus the pancreatic body/tail and we find PPI use is associated with alterations in duodenal and pancreatic tissue microbiome profiles.

Keywords: Duodenal microbiome; Pancreas microbiome; Pancreatic cancer; Proton pump inhibitor.

Figure 1.. The microbiome of the pancreatic head vs. the pancreas body/tail:

(A) Pancreatic relative abundance of genera and (B) phyla between pancreas head and pancreas body or tail. The data shown were filtered by frequency higher than 1%. (C) The pancreatic microbiomes between the 2 groups were analyzed for alpha-diversity measures.

Figure 2.. Beta diversity measures of the microbiome of the pancreatic head vs. the pancreas body/tail:

(A) Principal coordinates analysis (PCoA) for beta diversity by Jaccard (left) and Bray-Curtis indices (right) between duodenal samples (red dots), pancreas head samples (blue dots), and pancreas body/tail samples (green dots). (B) PCoA for beta diversity by the Jaccard (left) and Bray-Curtis indices (right) between duodenal samples (red dots), pancreas head samples with pancreatic duct dilation (blue dots), pancreas head samples without pancreatic duct dilation (green dots), and pancreas body/tail samples without pancreatic duct dilation (purple dots).

Figure 3.. Pancreatic microbiome profile of PPI users and non-users:

(A) Pancreatic relative abundance of genera and (B) phyla between PPI users and non-users. The data shown were filtered by frequency higher than 1%. (C) The pancreatic microbiomes between the 2 groups were analyzed for alpha-diversity measures and (D) beta diversity measures. PCoA microbiome beta diversity plot of Jaccard distance (D left) and Bray-Curtis dissimilarity plot (D right) between PPI non-users represented by red dots and PPI users by blue dots.