The Preliminary Analysis Of Combined Serum Four MicroRNAs As Biomarkers For The Early Diagnosis Of Primary Hepatocellular Carcinoma

ObjectiveCirculating biomarkers are benefit to early diagnosis of HCC. However, it is unsatisfactory that current tumor markers are applied to clinic. To investigate the possibility of miR-125 b, miR-223, miR-27 a, and miR-26 a in serum as a novel tumor marker for primary hepatocellular carcinoma(HCC) and the diagnosis value of combined detection of the four miRNAs in order to increase accuracy, sensitivity and specificity of screening diagnosis for HCC. MethodsNinety cases of primary HBV-related HCC patients were selected as training group, 30 of healthy volunteer and 30 of HBV patients were involved as control groups from Tianjin Medical University Cancer Institute and Hospital. HCC patients with early stage were defined by BCLC and TNM staging system, respectively. The relative expressions of serum mi R-125 b, miR-223, miR-27 a and miR-26 a were detected by qRT-PCR. Moreover, relative expression quantities of miRNAs was analyzed by 2-△△Ct.. Statistical analyses were performed by R 3.1 software. After the single marker detection of HCC, a stepwise logistic regression model to estimate the risk of being diagnosed with early stage and all stages of HCC was applied for marker panels. The relationship between the diagnostic sensitivity and specificity for established marker panels was evaluated by using receiver operating characteristic(ROC) analysis. All these analyses were performed in heath control, HBV group and non-cancer group(heathy control + HBV group) vs. HCC group, respectively. Computer-intensive bootstrap-resampling method was used to estimate the stability of the established models. Results1. The relative expression levels of the serum four miRNAs of HCC group were down regulated compared with healthy controls and HBV group which showed significant differences(P<0.0001).2. Compared HCC group with non-cancer group, the analysis of combined four miRNAs detection yielded AUC of 0.870 with 79.8% sensitivity and 81.7% specificity. Compared HCC group with healthy control, the analysis of combined four miRNAs detection yielded AUC of 0.970 with 94.0% sensitivity and 90.0% specificity. Compared HCC group with HBV group, the analysis of combined four miRNAs detection yielded AUC of 0.833 with 82.0% sensitivity and 76.7% specificity.3. Based on BCLC staging system, the analysis of combined four miRNAs detection yielded AUC of 0.843 with 81.5% sensitivity and 75.0% specificity, compared HCC group with non-cancer group. Compared HCC group with healthy control, the analysis of combined four miRNAs detection yielded AUC of 0.960 with 80.0% sensitivity and 100.0% specificity. Compared HCC group with HBV group, the analysis of combined four miRNAs detection yielded AUC of 0.812 with 70.4% sensitivity and 83.3% specificity.4. Based on TNM staging system, the analysis of combined four miRNAs detection for HCC with early stage yielded AUC of 0.920 with 92.3% sensitivity and 83.3% specificity, compared HCC group with healthy control.5. Statistical analysis tested that the established models for diagnosis of HCC were stable in non-cancer group, healthy control and HBV group. The significance of relative expression levels of serum mi RNAs was evaluated for the models. Conclusion1. The relative expression levels of the serum four miRNAs of HCC patients were down regulated compared with controls which showed significant differences.2. Combined serum miR-125 b, miR-223, miR-27 a, and miR-26 a could increase accuracy, sensitivity, and specificity of screening diagnosis for HCC.3. The relationship between relative expression levels of serum four miRNAs and the risk of HCC was evaluated by the models. It is possibility that serum miR-125 b, mi R-223, miR-27 a, and miR-26 a could be novel tumor markers in order to apply to clinic for diagnosis of HCC with early stage.