Differential Expression of Tissue microRNAs as Diagnostic Tools for Segregating Hepatocellular Carcinoma and Cirrhosis: A Biomarker Discovery Study Using Liver Biopsy Data

Background: Hepatocellular carcinoma (HCC) is an important cause of cancer-related mortality, often arising in cirrhotic livers. Current surveillance methods, including ultrasound and serum α-fetoprotein (AFP), have limited sensitivity and specificity. Circulating microRNAs (miRNAs) have shown promise as non-invasive biomarkers, but tissue-based miRNA profiles in HCC versus cirrhosis remain underexplored. This study investigates differentially expressed miRNAs in liver biopsies to identify potential diagnostic biomarkers for HCC.Methods: miRNA expression data from liver biopsies (333 samples, including 25 HCC and 30 cirrhosis samples) were obtained from the Gene Expression Omnibus dataset GSE51429. Data preprocessing, and quality control were performed using DESeq2 and limma in R. Differential expression analysis identified HCC-associated miRNAs, followed by feature selection using LASSO and Random Forest. Diagnostic performance was evaluated via ROC curve analysis.Results: Limma analysis identified 15 differentially expressed miRNAs (FDR < 0.05), including miR-625, miR-208b, and miR-138. PCA and heatmaps confirmed distinct clustering between HCC and non-HCC samples. Twenty-four miRNAs were differentially expressed, with miR-196b, miR-138, and miR-187 showing significant fold changes. ROC analysis revealed high discriminative power for miR-215-5p (AUC = 0.947), miR-373-3p (AUC = 0.902), and miR-200b (AUC = 0.873).Conclusion: This study identifies distinct tissue miRNA signatures differentiating HCC from cirrhosis, with several miRNAs demonstrating high diagnostic accuracy. These findings support the potential of tissue miRNAs as complementary biomarkers for HCC detection in high-risk cirrhotic patients. Further validation in larger cohorts is warranted for clinical translation.