Specific Gut Bacteria Causally Tied to Liver Cancer Risk in New Genetic Study

Liver cancer kills roughly 800,000 people worldwide every year, and the majority of those deaths come from hepatocellular carcinoma (HCC), the most common primary liver malignancy. Despite decades of research, the biological mechanisms driving HCC in people without obvious risk factors like cirrhosis or viral hepatitis remain poorly understood. A new Mendelian randomisation study published in Digestive and Liver Disease now points to a surprising upstream culprit: the community of microbes living in your gut.

Gut Microbiome Abundances Causally Linked to Hepatocellular Carcinoma

The study, led by N. Daniel, N. Papadimitriou, A. C. Chatziioannou, and colleagues, investigated whether specific gut microbiome abundances and their associated circulating metabolites are causally connected to HCC risk, not just correlated with it. Using large-scale genetic data, the researchers identified statistically significant associations between particular bacterial taxa and liver cancer outcomes. The work builds on a growing body of evidence that gut microbiota influence liver cancer through multiple biological pathways, including the modulation of host signalling and gene expression via metabolites such as short-chain fatty acids and bile acids.

The gut-liver axis is the anatomical and biochemical highway connecting these two organs. Blood from the intestines drains directly into the liver through the portal vein, meaning that anything the gut microbiome produces, including metabolites, toxins, and immune signals, reaches the liver in high concentrations before circulating anywhere else in the body. When the microbiome is disrupted, that constant delivery can shift from protective to damaging.

Why This Matters for the 830 Million People at Risk of Liver Disease

HCC is a leading cause of cancer-related mortality globally, and outcomes remain poor even with modern treatments: five-year survival rates sit below 20 percent in most countries. The burden falls disproportionately on people with chronic liver conditions, but a meaningful fraction of HCC cases arise in individuals without a clear prior diagnosis. Identifying modifiable upstream drivers, particularly gut-derived metabolites that could in principle be targeted through diet, probiotics, or specific therapeutics, would open new angles for both prevention and early detection.

The microbiome-liver connection is not a new hypothesis. Preclinical models have shown that microbial metabolites shape liver cancer progression through immune cells, and the gut microbiome composition of liver cancer patients consistently differs from that of healthy controls. What has been missing is evidence that the relationship runs in the right causal direction: that a disrupted microbiome helps cause HCC rather than simply being a downstream consequence of liver disease.

Mendelian Randomisation: Using Genetics to Establish Causation

This is precisely where the study's methodological approach earns its keep. Mendelian randomisation uses genetic variants as instrumental variables to make causal inferences from observational data. The key insight is Mendelian: genetic variants are assigned at conception, before any disease develops, so they cannot be caused by the outcome. By anchoring exposure estimates to heritable genetic differences in gut microbiome composition rather than to self-reported diet or lifestyle data, the method sidesteps two of the biggest problems in nutritional and microbiome epidemiology: confounding by correlated behaviours and reverse causation, where early undetected cancer changes the microbiome before diagnosis.

The team leveraged genome-wide association study data on gut microbiome abundances and combined those with large genetic datasets for HCC to identify which bacterial taxa and their downstream circulating metabolites showed evidence of a causal relationship with liver cancer risk. This two-sample approach allowed them to interrogate multiple microbial species and metabolite classes simultaneously without requiring a single cohort large enough to capture a relatively rare outcome like HCC.

Limitations Inherent to the Mendelian Randomisation Design

The findings are notable, but the method carries its own constraints. Mendelian randomisation can be undermined by pleiotropy, the phenomenon where a genetic variant influences multiple traits at once rather than acting solely through the exposure of interest. If a variant that shapes gut bacterial abundance also independently affects liver function through a separate pathway, the causal estimate becomes unreliable. The authors used standard sensitivity analyses, including weighted median and MR-Egger approaches, to probe for this, but pleiotropy can never be fully ruled out.

Additionally, genetic instruments for microbiome composition tend to explain a relatively small fraction of the variance in actual bacterial abundance, meaning the instruments are weaker than those available for more heritable traits like blood lipid levels. Weak instruments can bias results toward the null, so genuinely causal associations may be harder to detect. The study also cannot pinpoint exactly which specific metabolite-receptor interaction constitutes the biological mechanism, only that the statistical signal is consistent with a causal path.

Next Steps: Identifying Which Bacterial Species and Metabolites to Target

The immediate research priority is replication: independent cohorts with both deep metagenomic microbiome profiling and long follow-up periods for HCC incidence are needed to confirm which bacterial taxa and metabolite classes drive the signal. Mechanistic work in cell and animal models will be required to trace the specific molecular routes from a particular microbial product to hepatocyte transformation or immune evasion in the liver.

If the causal relationships hold up, the clinical implications could be substantial. Gut microbiome composition is modifiable through diet, antibiotics, prebiotics, and faecal microbiota transplantation, in ways that blood type or genetic ancestry are not. A confirmed bacterial target might eventually inform HCC screening protocols for people with specific microbiome profiles, or guide dietary interventions for high-risk groups. That remains a longer horizon, but this study sharpens the map considerably.