Machine-Learning-Based Clinical Biomarker Using Cell-Free DNA for Hepatocellular Carcinoma (HCC)

Area of Science:

• Oncology
• Molecular Diagnostics
• Bioinformatics

Background:

• Hepatocellular carcinoma (HCC) is a major global cause of cancer mortality.
• Current biomarkers for HCC diagnosis and prognosis lack sufficient sensitivity, particularly for early detection and treatment response prediction.

Purpose of the Study:

• To develop and validate a highly sensitive cell-free DNA (cfDNA) detection system for improved diagnosis and prognosis of HCC.
• To integrate cfDNA analysis with machine learning algorithms for enhanced diagnostic and prognostic capabilities.

Main Methods:

• Utilized a sensitive cfDNA detection system to analyze plasma biomarkers.
• Employed machine learning algorithms to develop a cfDNA score (cfD_HCC).
• Compared the diagnostic and prognostic accuracy of cfDNA markers against conventional serum enzymes.

Main Results:

• cfDNA, particularly alpha-fetoprotein (AFP) expression within cfDNA, exhibited superior diagnostic accuracy compared to traditional serum biomarkers.
• The integrated cfD_HCC score, combining total cfDNA and cfAFP-DNA levels, significantly enhanced diagnostic and prognostic performance.
• The cfD_HCC score demonstrated improved accuracy in determining HCC pathological features and predicting patient survival outcomes.

Conclusions:

• The developed cfDNA capture/analysis platform, utilizing the cfD_HCC score, represents a promising advancement for HCC diagnosis and prognosis.
• This cfDNA-based approach offers a more sensitive and accurate method for managing HCC patients compared to existing biomarkers.