Development and Validation of a Novel Circulating miRNA-Based Diagnostic Score for Early Detection of Hepatocellular Carcinoma

With the rise of liquid biopsy in oncology, circulating miRNAs have become one of the most promising noninvasive biomarkers for early detection of hepatocellular carcinoma (HCC). However, a reliable HCC-related circulating miRNA panel and corresponding diagnostic model remain to be explored.

Five large public datasets related to intact miRNA profiles in the serum or tumors of HCC patients were included and divided into training cohorts (GSE113740 and TCGA-LIHC) and validation cohorts (GSE112264, GSE113486 and GSE106817). Compared with non-cancer controls and high-risk patients, key miRNAs dysregulated in both the serum and tumors of HCC patients were identified by differential expression analysis and overlapping analysis. The corresponding diagnostic model was constructed by LASSO logistic regression and evaluated by receiver operating characteristic curves and a nomogram with calibration plot.

A distinctive panel of HCC-related circulating miRNAs, including three upregulated miRNAs (miR-184, miR-532-5p, miR-221-3p) and three downregulated miRNAs (miR-5589-5p, let-7b-3p, miR-26b-3p), were rigorously screened out, all of which displayed significant discriminability between HCC patients and controls (all P < 0.05). In addition, a reliable six-circulating miRNA-based diagnostic score was constructed and displayed robust diagnostic ability for HCC (particularly for early-stage HCC) (AUC = 0.9535, P < 0.05) compared with that of the serum α-fetoprotein test. Importantly, its efficacy was sufficiently validated in three independent datasets (AUC = 0.9780/0.9961/0.9681, all P < 0.05). Furthermore, a visual nomogram based on the diagnostic score was correspondingly established to strengthen its clinical applicability.

The six-circulating miRNA-based diagnostic score may be a reliable noninvasive biomarker for early-stage HCC screening and dynamic monitoring of postoperative recurrence.