Microbiome as a Potential Diagnostic and Predictive Biomarker in Severe Alcoholic Hepatitis

Soon Sun Kim; Jung Woo Eun; Hyo Jung Cho; Do Seon Song; Chang Wook Kim; Young Seok Kim; Sung Won Lee; Yoon-Keun Kim; Jinho Yang; Jinhee Choi; Hyung Joon Yim; Jae Youn Cheong


Aliment Pharmacol Ther. 2021;53(4):540-551. 

In This Article

Abstract and Introduction


Background: Severe alcoholic hepatitis (AH) is the most aggressive form of alcohol-related liver disease with high mortality. The microbiome is an emerging therapeutic target in alcohol-related liver disease.

Aims: To investigate the microbiome composition in patients with severe AH, and to determine microbiome recovery after rifaximin treatment in gut bacteria and bacteria derived-extracellular vesicles.

Methods: We enrolled 24 patients with severe AH and 24 healthy controls. Additional faecal samples were collected after 4 weeks in 8 patients with severe AH who completed rifaximin treatment. Treatment response was defined based on Lille score model after 7 days of treatment. Metagenomic profiling was performed using 16S ribosomal RNA amplicon sequencing.

Results: Faecal microbiomes of patients with severe AH had lower alpha diversity and higher beta diversity than those of healthy controls in both gut bacteria and extracellular vesicles. Bacilli, Lactobacillales and Veillonella were significantly increased in the gut bacteria of patients with severe AH, and Veillonella, Veillonella parvula group and Lactobacillales were significantly increased in the extracellular vesicles of patients with severe AH. Eubacterium_g23, Oscillibacter and Clostridiales decreased in the gut bacteria of patients with severe AH, and Eubacterium_g23, Oscillibacter and Christensenellaceae decreased in the extracellular vesicles of patients with severe AH. After rifaximin treatment, 17 taxa in the gut bacteria and 23 taxa in extracellular vesicles were significantly restored in patients with severe AH. In common, Veillonella and Veillonella parvula group increased in patients with severe AH and decreased after rifaximin treatment, and Prevotella and Prevotellaceae decreased in patients with severe AH and increased after rifaximin treatment. Treatment non-responders showed a significantly lower abundance of Prevotella at baseline than did treatment responders.

Conclusion: Dysbiosis was confirmed in severe AH but was alleviated by rifaximin treatment. Taxa associated with severe AH can be candidate biomarkers or therapeutic targets.


Alcohol-related liver disease is a leading cause of morbidity and mortality, accounting for 0.9% of all global deaths and 47.9% of all liver cirrhosis-attributable deaths.[1] Severe alcoholic hepatitis (SAH) is a serious form of acute decompensation of alcoholic liver disease, which is characterised by the rapid onset of jaundice, malaise, anorexia, tender hepatomegaly and features of systemic inflammatory response syndrome.[2] The Maddrey discriminant function (MDF) formula was developed in 1977, and patients with MDF >32 showed high 28-day mortality (>50%) and benefited from treatment with glucocorticoids.[3,4]

The gut-liver axis is a major pathway for alcoholic liver disease development and progression. The liver plays an essential role in the modulation of gut microbiota and its effects through multiple routes, including bile acid production and enterohepatic circulation, as well as by modulating the responsiveness to gut bacterial end products and nutrients received via the portal vein.[5] Patients with SAH show dysbiosis of the gut microbiome, characterised by an increase in streptococci, enterobacteria, bifidobacteria and Actinobacteria, and a decrease in Atopobium, Akkermansia muciniphila and Bacteroidetes.[6–8] In a previous study, transfer of dysbiotic human faecal microbiota into germ-free or conventionally raised mice conferred increased susceptibility to alcoholic liver disease.[6] Furthermore, oral supplementation of Akkermansia muciniphila ameliorated alcoholic liver disease in both an acute and chronic experimental setting.[7] This accumulating evidence of an important role of dysbiosis in alcoholic-related liver disease has motivated the development of new treatment approaches, including diet modification, antibiotics, prebiotics, probiotics and faecal microbiota transplantation, targeting modulation of the gut microbiota.[9]

The gut microbiota can secrete different types of extracellular vesicles (EVs), including outer membrane vesicles, shedding vesicles and apoptotic bodies.[10,11] EVs mainly comprise lipids, proteins, nucleic acids and metabolites, and their primary role is to transport these active biomolecules to cells over long distances.[12,13] Microbe-derived EVs have emerged as a potential link to revealing the mechanism underlying the interaction of gut microbial communities with human health. Nevertheless, the microbiome profiles in EVs of patients with SAH have not been elucidated to date.

Rifaximin is a non-absorbable antibiotic derived from the macrolide rifamycin, and has a function of the selective decontamination of gut bacteria.[14] Accordingly, rifaximin has been suggested to have a beneficial effect on the gut microbiota composition in liver disease. However, it is still unclear that to what extent rifaximin could modulate the intestinal microbiome profile and microbiota diversity in different clinical situations. A recent study demonstrated that rifaximin prevents ethanol-induced liver injury in obese KK-A y mice through modulation of the small intestinal microbiota signature, suggesting a favourable effect on gut dysbiosis in patients with alcoholic hepatitis.[15] Currently, a prospective, randomised, multi-centre trial (NCT02485106) is ongoing to examine the benefit of rifaximin as an adjuvant to corticosteroids/pentoxifylline in patients with SAH.

Therefore, we aimed to investigate the microbiome composition in patients with SAH, and to determine the degree of microbiome recovery after rifaximin treatment in gut bacteria and bacterial-derived EVs. Furthermore, we explored the role of selected SAH-specific taxa as predictive markers for disease severity or treatment response.