| Chapter I Screening of differential expressed circulating mi RNAs during the pregression of hepatic fibrosis in patients with Chronic HBV infectionBackground & Aim This study aims to identify differential expressed serum/plasma mi RNAs in Chronic hepatitis B liver fibrosis stages, to discover non-invasive diagnostic serum marker and lay the foundation of diagnosis model for hepatitis b liver fibrosis.Methods:In this study, we enrolled cohorts of Chronic hepatitis B virus infection patients who phsical check up at Shanghai First People's Hospital affiliated Shanghai Jiaotong University School of Medcine from July,2012 to Dec,2013 and meet the meet the standards,totally 50 cases which consists of 10 cases in Liver fibrosis Scheuer stage S0-S4.We extract serum/plasma micro RNAs,and use Agilent Corporation Human mi RNA chip v 18.0(Affymetrix Human v18.0)to hybrid and scan mi RNA chips and extract the data.After applying Quantile normalization method for data standardization, we mainly use the independent sample t-test, one-way ANOVA to analysis differentially expressed mi RNAs in groups between S0- S4 groups,groups between the non-significant liver fibrosis and significant liver fibrosis(S0-1 /S2-4) and groups between mild and severe liver fibrosis(S0-1 / S3-4).Results:The results of microarray analysis showed 140 detectable micro RNAs with fold expression change ?2 and P value <0.01 were identified between the group S1-4 and group S0.Between non-significant fibrosis group(S2-4) and significant fibrosis(S2-4),there are 11 mi RNAs express change more than 2 times and compare severe fibrosis group(S3-4) with mild fibrosis group(S0-1), the number of differentially expressed mi RNAs up to 7.Interestingly,we found hsa-mi R-1225-3p,hsv1-mi R-H7-3p,hsa-mi R-1238-3p,hsa-mi R-3162-3p, hsa-mi R-4721 express significant difference between non-significant fibrosis group(S2-4) and significant fibrosis(S2-4),the fold change multiples are at least 3-5 times.Meanwhile, they also express differently between mild and severe liver fibrosis or in S0- S4 groups.For these reasons above,we choose these mi RNAs as candidate micro RNAs for further testing by quantitative reverse transcriptase polymerase chain reaction(q RT-PCR).Conclusions: Using serum/plasma micro RNA chip technology, we found differential expression of mi RNAs is closely associate with hepatic fibrosis staging.This study will facilitate the development and application of non-invasive biomarker for earlier diagnosis of hepatic fibrosis.Chapter II Validation of differentially expressed micro RNAs, foundation and validation of diagnostic model for Chronic hepatitis B liver fibrosisBackground & Aim This part of the study aimed to further validate the 5mi RNA candidates from part I and logistic regression method is applied to optimize the combination of micro RNAs model and estimate the risk of being diagnosed with significant liver fibrosis.We evaluate the diagnostic efficiency of the and further compare it with other prediction models(APRI score, Forn 's score, FIB- 4 score, s index) in predicting diagnostic efficiency.Methods:The candidates and patients enrolled this part are also from Shanghai First People's Hospital affiliated Shanghai Jiaotong University School of Medcine from July,2012 to Dec,2014 with the same criteria as in in part I,totally 83 cases consists of 10 CHB liver fibrosis S0 patients,15 S1 patients,13 S2 patients,10 S3 patients,10 S4 patients,13 CHB Compensatory Cirrhosis patients and 12 CHB Decompensatory Cirrhosis patients and normal control group of 20 people.In this phase,we validate 5 mi RNA candidates by Taqman quantitative reverse transcriptase polymerase chain reaction(q RT-PCR),and after get their relative expression we use independent sample t-test analysis to analyze whether they are differentially expressed between non-significant fibrosis group(S0-2) and significant fibrosis(S2-4), liver fibrosis and cirrhosis group, compensated cirrhosis and decompensated cirrhosis group,and further screen target mi RNA.Meanwhile,we use logistic regression model to optimize the combined diagnosis model,and estimate the diagnostic value by drawing the ROC curve and calculating AUROC.Furthermore,we compare the model with other prediction models(APRI score, Forn 's score, FIB- 4 score, s index) about the predicting diagnostic efficiency.Results: Five micro RNAs candidates express significantly different between significant liver fibrosis group and non-significant liver fibrosis group(p < 0.05),the raised multiples of hsa-mi R-1225-3p, hsa-mi R-1238-3p,hsa-mi R-3162-3p,hsa-mi R-4721,hsv1-mi R-H7-3p is 10, 16, 4.85, 4.7, 10.23 times,but they fail to distinguish liver fibrosis and cirrhosis group, compensated cirrhosis and decompensated cirrhosis group. We establish the receiver-operating Characteristic Curve(ROC Curve)to judge the diagnosis ability of each mi RNA, the Area Under the ROC Curve(AUC) we calculated of hsa-mi R-3162-3p, hsa-mi R-1225-3p,hsa-mi R-1238-3p,hsa-mi R-4721,hsv1-mi R-H7-3p were 0.899, 0.852, 0.656, 0.650,0.650.It is obvious that hsa-mi R-3162-3p and hsa-mi R-1225-3p are significantly effective markers to diagnose significant liver fibrosis,while hsa-mi R-1238-3p,hsa-mi R-4721,hsv1-mi R-H7-3p are not.Furthermore we use logistic regression model to optimize the combined diagnosis model:Mi R-5=1.115*mi R1225+0.157*mi R1238+6.256*mi R3162+0.072*mi R4721-0.831 mi RH7-2.729,and found AUROC of significant liver fibrosis is 0.909 which is better than better than the Forns index, index of APRI, FIB- 4 score, S index prediction model and single candidate micro RNA.Conclusions: Using Taqman q RT-PCR technology, we found hsa-mi R-1225-3p,hsa-mi R-1238-3p, hsa-mi R-3162-3p, hsv1-mi R-H7-3p, hsa-mi R-4721 express significantly different in the non-significant fibrosis/significant liver fibrosis,which makes it possible to become the non-invasive molecular marker in liver fibrosis.The diagnosis model(MIR-5)built by logistic regression can efficiently distinguish the group of significant and non-significant liver fibrosis with high sensitivity and specificity, and is superior to other noninvasive diagnosis model(Forns index, index of APRI, ratings and FIB- 4 S index) and single candidate micro RNA. |
PDF Full Text Request