A BLOOD BIOMARKER panel combining DNA methylation and protein measurements showed promising performance in distinguishing hepatocellular carcinoma (HCC) from non-cancerous chronic liver disease, a new study suggests.
Candidate methylation markers were subsequently assessed alongside five protein markers in 299 blood samples.
The final model combined four DNA methylation markers with two protein markers, bringing together different types of biological information in a single assessment.
For HCC classified as Barcelona Clinic Liver Cancer stage 0–A, the panel achieved sensitivity of 82.9%.
These findings suggest that combining methylation and protein markers may capture information that is missed when individual established biomarkers are considered alone.
A BLOOD BIOMARKER panel combining DNA methylation and protein measurements showed promising performance in distinguishing hepatocellular carcinoma (HCC) from non-cancerous chronic liver disease, a new study suggests.
Developed in populations predominantly affected by hepatitis B virus (HBV), the six-marker model also demonstrated encouraging sensitivity for very early and early-stage HCC. However, the researchers stress that prospective validation is required before clinical implementation.
The findings represent a step towards blood-based detection, while leaving an important question unanswered: how well would the panel perform during routine surveillance of people who have not yet been diagnosed with cancer?
Combining Different Biological Signals
Researchers began with genome-wide DNA methylation profiling of 55 HCC tissues and matched non-HCC tissues. This process sought changes that could help distinguish cancerous from non-cancerous tissue.
Candidate methylation markers were subsequently assessed alongside five protein markers in 299 blood samples. The resulting diagnostic model was then evaluated in an independent cohort of 149 participants.
The final model combined four DNA methylation markers with two protein markers, bringing together different types of biological information in a single assessment.
Promising Diagnostic Performance
In the training dataset, the model achieved sensitivity of 91.8% and specificity of 95.4%. Corresponding figures in the validation dataset were 88.4% and 93.8%.
Sensitivity describes the proportion of cancer cases correctly identified, while specificity reflects the proportion of participants without cancer correctly classified.
For HCC classified as Barcelona Clinic Liver Cancer stage 0–A, the panel achieved sensitivity of 82.9%. In this case-control setting, it outperformed alpha-fetoprotein and des-gamma-carboxyprothrombin.
These findings suggest that combining methylation and protein markers may capture information that is missed when individual established biomarkers are considered alone.
Surveillance Studies Still Needed
The distinction between diagnostic discrimination and surveillance performance is central to interpreting the results.
A case-control study assesses whether a test separates participants with and without a known condition. Routine surveillance instead requires a test to perform reliably over time among people at risk, including those whose cancer is not yet clinically apparent.
The HBV-predominant study population also defines the setting in which the panel has been evaluated. Its performance in populations with different underlying causes of liver disease remains a question for further research.
Future studies should assess how the panel fits within surveillance pathways and whether its use leads to earlier diagnosis and better outcomes.
The authors conclude that the model has potential for blood-based HCC detection, while emphasising the need for prospective surveillance-cohort validation before adoption.
Reference
Zhuang H et al. An integrative multi-omic biomarker panel distinguishes hepatocellular carcinoma from chronic liver disease in HBV-predominant populations. Hepatology Research.2026;DOI:10.1111/hepr.70274.
