Abstract
BACKGROUND: Current diagnostic markers for hepatocellular carcinoma are compromised and limited by their low sensitivity and speci- ficity. In this study, circulating microRNAs were utilized as a diagnostic tool to segregate hepatocellular carcinoma patients from healthy subjects. METHODS: We analyzed 2 public datasets for differences in plasma microRNA expression profiles of hepatocellular carcinoma patients and healthy controls to identify biomarkers related to hepatocellular carcinoma. Plasma samples from hepatocellular carcinoma patients and control subjects were then collected for next-generation microRNA sequencing analysis. The differential microRNAs obtained from the above 3 parts were intersected to obtain microRNAs that were significantly different between the 2 groups. We then analyzed 58 specimens, which come from hepatocellular carcinoma and the control group, for validation through a quantitative polymerase chain reaction. The diagnostic value of these differentially expressed miRNAs was assessed by receiver operating characteristic curve analysis. RESULTS: The levels of miR-206 and miR-222 were significantly higher (P < .05) and the level of miR-126 was lower (P < .05) in patients with hepatocellular carcinoma than in healthy subjects. Receiver operating characteristic analysis established a powerful diagnostic accuracy when miR-206, miR-222, and miR-126 were combined (area under curve = 0.887), which was similar to that of the markerα-fetoprotein (area under curve = 0.889). When the microRNAs were combined with α-fetoprotein, the accuracy of hepatocellular carci- noma diagnostic potential was further improved (area under curve = 0.989). CONCLUSION: We identified 3 microRNAs significantly altered in the plasma of hepatocellular carcinoma patients and they can screen patients at risk of hepatocellular carcinoma.