Society: AASLD
Objective: To investigate the clinical relevance of fecal microbial dysbiosis in HCC development
Design: A case-control study. Fecal microbial composition and predicted genetic function inferred from high-throughput 16S ribosomal RNA sequencing were analyzed between HCC (n=53) and age-, gender- and body mass index-matched non-HCC subjects (n=53). The predicting performance of microbial dysbiosis index (MDI) of selected contrasting taxa was examined in the testing dataset and validation dataset. Further animal studies were conducted by transplantation of feces containing the selected contrasting taxa.
Results: A significant alteration of fecal microbial diversity and composition were found in HCC patients. HCC patients with early stage and liver cirrhosis had higher fecal microbial diversity but not richness. We identified the subset of 4 species (Ruminococus gnavus, Bifidobacteirum longum, Eubacteirum dolicum, and Rothia mucilaginosa)-based biomarker which achieved a great association with HCC development (area under the curve [AUC] 0.916). The AUC value increased up to 0.966 when combined with AFP values. This fecal microbial signature performed well in the validation cohort (AUC 0.920). The fecal microbiota signature in HCC was predicted to be associated with enhanced signaling pathways involving L-arginine biosynthesis, urea-cycle, heme biosynthesis, and nicotinamide adenine dinucleotide (NAD) salvage pathways; and reduced signaling pathways involving allantoin degradation and pyrimidine deoxyribonucleotide. Fecal microbial transplantation study demonstrated a significantly increased tumor growth in mice receiving feces containing the 4-species signature.
Conclusion: A distinct gut microbial profile was found in HCC patients. The specific fecal microbial signature may involve tumorigenesis and predict HCC development.

Background & Aims: Non-alcoholic fatty liver disease (NAFLD) is a rapidly growing cause of hepatocellular carcinoma (HCC) worldwide. Emerging evidence implies that dietary modification could modulate the pathogenesis of NAFLD-HCC. In this study, we aim to elucidate the roles of the fermentable fiber inulin and non-fermentable fiber cellulose in NAFLD-HCC progression and its underlying mechanisms through directly modulating gut microbiota and metabolites.
Methods: Two NAFLD-HCC models were established: C57BL/6 mice (2-3 weeks) were injected with Diethylnitrosamine (DEN) (25mg/kg) and fed with high fat high cholesterol diet (HFHCD) or choline deficient, high fat diet (CD-HFD) for 6 months. Meanwhile, 10% fermentable fiber inulin or non-fermentable fiber cellulose was added to respective diets. To track the in vivo incorporation of inulin into gut microbiota and metabolites, samples from mice labeled with 13C-inulin were analyzed with shotgun metagenome sequencing and non-targeted metabolomics. Human NAFLD-HCC cell lines (HKCI2 and HKCI10) and mouse NAFLD-HCC organoids were used for bio-functional study of the candidate metabolites.
Results: Inulin administration, but not cellulose, consistently suppressed NAFLD-HCC in DEN-injected mice fed HFHCD with significantly decreased tumor number and tumor load. This was confirmed in a second NAFLD-HCC mouse model induced by feeding mice with CDHFD. Metagenome sequencing combined with 13C DNA stable isotope labeling of fecal sample of mice identified the enrichment of beneficial microbes by inulin, including Bacteroides cellulosilyticus, Bacteroides thetaiotaomicron, Bacteroides caccae, Akkermansia muciniphila, etc. In vivo validation by oral gavage of these top enriched bacteria revealed that Bacteroides cellusilyticus as the most effective bacterium in inhibiting NAFLD-HCC formation with significantly reduced tumor number and tumor load. Meanwhile, the non-targeted metabolomics combined with 13C-inulin labeling identified inulin-derived bioactive metabolites pentadecanoic acid and nicotinic acid were consistently enriched in the stool and the portal vein serum of inulin-treated mice. Integrative analysis showed that these two metabolites highly correlated with Bacteroides cellusilyticus. HKCI2 and HKCI10 co-cultured with pentadecanoic acid or nicotinic acid suppressed cell viability and proliferation, and induced apoptosis (all P<0.05). Consistently, treatment with pentadecanoic acid or nicotinic acid significantly inhibited the growth of mouse NAFLD-HCC-derived organoids (all P<0.05).
Conclusion: Dietary inulin is effective in suppressing NAFLD-HCC development by enriching gut beneficial bacterium Bacteroides cellusilyticus and their associated bioactive metabolites (pentadecanoic acid and nicotinic acid).