| Background:Sensitive and specific detection of liver cirrhosis is an urgent need for optimal individualized management of disease activity. Substantial studies have identified circulation miRNAs as biomarkers for diverse diseases including chronic liver diseases. In this study, we investigated the plasma miRNA signature to serve as a potential diagnostic biomarker for silent liver cirrhosis.Methods:A multistage, case-control study was designed to evaluate plasma miRNAs as candidate biomarkers in patients with liver cirrhosis. In the primary exploration phase, eighty plasma samples, each with 723 microRNAs, were screened using human miRNA microarrays based on the data described in our recently published study. Of the 80 samples,25 had CHB-related cirrhosis and 22 had CHB, together with the remaining 33 as healthy controls. A biomarker confirmation analysis was subsequently performed with qPCR assay to refine the number of serum miRNAs in the liver cirrhosis. This analysis was carried out in 2 phases:(a) the training phase, in which plasma samples from 20 CHB-related patients,9 CHB patients and 12 healthy controls were tested for initial evaluation of six candidates discovered; then 70 CHB-related cirrhosis,23 CHB, and 48 healthy individuals were tested for evaluation of the two selected miRNAs, and (b) the validation phase, in which plasma samples from 13 CHB-related silent cirrhosis,25 CHB, and 6 healthy controls (all of them had liver biopsy) were tested for definitive evaluation; as well as 47 non-CHB-related cirrhosis,7 non-CHB-related chronic liver disease and 38 healthy individuals were finally tested for model application.Results:1. The expression profiles of the human miRNAs were determined by a microarray containing probes for 723 human miRNAs between the CHB-related cirrhosis and control groups. A list of six differentially produced miRNAs (miR-106b, miR-122, miR-144, miR-181d, miR-181b, and miR-584) was constructed as candidates for further investigation via qPCR. Three miRNAs, including miR-106b, miR-122, and miR-144, with significantly lower expression in CHB-related cirrhosis group than in the CHB group (fold change=0.06-0.11; P< 10-8). By contrast, miR-181d and miR-181b was significantly up-regulated (fold change=13.0-14.4; P<1.5×10-6) and miR-584 was down-regulated (fold change=0.1; P<1.5×10-6) in CHB-related cirrhosis group compared with the healthy group.2. The six candidate miRNAs selected from the microarray analysis were primary confirmed with qPCR in the training phase. miR-106b and miR-181b had significantly different expression between the CHB-related cirrhosis and control groups. Low expression of miR-106b (P<0.001, fold change=0.354) and high expression of miR-181b (P<0.001, fold change=8.276) were observed in patients with CHB-related cirrhosis compared with those in the control group.3. Receiver operating characteristic (ROC) curves were constructed to obtain the relation between the sensitivity and specificity of these two miRNAs at varying cutoff levels. The AUC for miR-106b and-181b were 0.715 (95% CI, 0.641-0.789; sensitivity=0.804, specificity=0.522) and 0.833 (95% CI, 0.775-0.892; sensitivity=0.678, specificity=0.870), respectively.4. The formula of the classifier is as follow:Risk score=1.763 x miR181b-14.622x miR106b-0.367. The ROC curve of the classifier has an AUC of 0.882 (95% CI,0.834,0.929; sensitivity=0.856, specificity=0.750).5. The AUC of the classifier was 0.774 (95% CI,0.589-0.959; sensitivity=0.615, specificity= 0.935; Fig,4A) in the cohort of early stage of cirrhosis, much higher than the AUC of single clinical indicators (including total bilirubin, albumin, alanine aminotransferase, prothrombin time, international normalized ratio and imaging), indicating the improved diagnostic performance of the classifier compared with clinical indicators.6. The predicted AUC of the logit model was 0.915 (95% CI,0.854-0.976; sensitivity=0.787, specificity=0.932; Fig.5A) in non-CHB-related cirrhosis cohort, which verified that the two miRNAs are not etiology-specific.7. Further, no significant difference of miR-106b and -181b levels was observed in patients with different grades of necorinflammation, and no significant causal relationship was observed between miRNAs expressions and all the other clinical indicators including total bilirubin, albumin, alanine aminotransferase, prothrombin time, and international normalized ratio. All the results indicated that the level change of these two miRNAs was not related with the liver necorinflammation, liver function, and distruction of the liver cells.Conclusion:Our study demonstrated that the combined detection of miR-106b and miR-181b has a considerable clinical value to diagnose patients with liver cirrhosis, especially those at early stage. |
PDF Full Text Request