Background and AimHepatocellular carcinoma (HCC) is one of the most deadly tumors. Transarterial chemoembolization (TACE) is effective for unresectable HCC. In recent years, miRNAs have been proposed as novel diagnostic and prognostic tools for HCC. This study aimed to identify whether microRNAs (miRNAs) can serve as biomarkers to reliably predict outcome before HCC patients are treated with TACE.MethodsEleven miRNAs (miR-, miR-19a, miR-101-3p, miR-199a-5p, miR-200a, miR-21, miR-214, miR-221, miR-222, miR-223 and miR-, -5p) were quantified by quantitative real-time PCR (qRT-PCR) in 136 HCC patients’ serum before they received TACE therapy. Univariate and multivariate analysis were used to identify the prognostic value of clinical parameters and miRNAs. Area under the receiver operating characteristic curve (AUC) was used to evaluate the prediction potency.ResultsThe levels of some miRNAs were dramatically associated with clinicopathologic features regarding Child-Puge class, AFP, tumor size and satellite nodules. Univariate analysis revealed that miR-200a, miR-21, miR-122 and miR-224-5p were significantly associated with patients’ survival. Multivariate analysis demonstrated that AFP, satellite nodules and miR-200a were the independent prognostic factors associated with survival in this cohort (p = 0.000, 0.001, 0.000, respectively). The probability of the prognostic accuracy of miR-200a was 81.64% (74.47% specificity and 88.76% sensitivity), which was higher than the classifier established by combination of AFP and satellite nodules (76.87% probability, 70.21% specificity and 69.66% sensitivity). Furthermore, the combination of AFP, satellite nodules and miR-200a demonstrated as a classifier for HCC prognosis, yielding a ROC curve area of 88.19% (93.62% specificity and 68.54% sensitivity).ConclusionsOur study indicated that serum miR-200a may prognosticate disease outcome in HCC patients with TACE therapy. Therefore, miR-200a can potentially guide individualized treatment for HCC patients with a high risk of TACE treatment failures.