ABSTRACT
BACKGROUND: Dysbiotic gut bacteria engage in the development and progression of severe alcoholic hepatitis (SAH). We aimed to characterize bacterial communities associated with clinical events (CE), identify significant bacteria linked to CE, and define bacterial relationships associated with specific CE and outcomes at baseline and after treatment in SAH. METHODS: We performed 16-s rRNA sequencing on stool samples (n=38) collected at admission and the last follow-up within 90 days in SAH patients (n=26; 12 corticosteroids; 14 granulocyte colony-stimulating factor, [G-CSF]). Validated pipelines were used to plot bacterial communities, profile functional metabolism, and identify significant taxa and functional metabolites. Conet/NetworkX® was utilized to identify significant non-random patterns of bacterial co-presence and mutual exclusion for clinical events. RESULTS: All the patients were males with median discriminant function (DF) 64, Child-Turcotte-Pugh (CTP) 12, and model for end-stage liver disease (MELD) score 25.5. At admission, 27%, 42%, and 58% had acute kidney injury (AKI), hepatic encephalopathy (HE), and infections respectively; 38.5% died at end of follow-up. Specific bacterial families were associated with HE, sepsis, disease severity, and death. Lachnobacterium and Catenibacterium were associated with HE, and Pediococcus with death after steroid treatment. Change from Enterococcus (promotes AH) to Barnesiella (inhibits E. faecium) was significant after G-CSF. Phenylpropanoid-biosynthesis (innate-immunity) and glycerophospholipid-metabolism (cellular-integrity) pathways in those without infections and the death, respectively, were upregulated. Mutual interactions between Enterococcus cecorum, Acinetobacter schindleri, and Mitsuokella correlated with admission AKI. CONCLUSIONS: Specific gut microbiota, their interactions, and metabolites are associated with complications of SAH and treatment outcomes. Microbiota-based precision medicine as adjuvant treatment may be a new therapeutic area.
