Clinical Values Of Non-invasive Examinations For Assessing Esophageal Varices And Variceal Bleeding In Patients With Viral Hepatitis-Related Cirrhosis

Background and Objective Liver cirrhosis is the late stage of various chronic liver diseases.It is commonly caused by viral hepatitis in China.Liver dysfunction and portal hypertension(PH)are the main performances,leading to a variety of complications.Gastroesophageal varices(GOV)is one of the main clinical manifestations of portal hypertension.And esophageal variceal bleeding(EVB)has the highest mortality in complications of portal hypertension.And it is also the most common reason of upper gastrointestinal bleeding for liver cirrhosis patients.Therefore,it is essential for these patients to identify PH and esophageal varices(EV)early and take prevention and treatment measures in time.At present,hepatic venous pressure gradient(HVPG)measurement and endoscopy are the gold standard for assessing PH and EV.However,both of these tests above are invasive and expensive,and most patients can not tolerate.Therefore,great mental and economic pressure is brought to patients,significantly reducing the follow-up rate.The latest guidelines indicate that non-invasive portal venous pressure assessment methods or models are one of the most important focuses in future's research.In this paper,it collected the data of the viral hepatitis-related cirrhosis patients,such as the indexes of serological tests, abdominal ultrasonography,liver transient elastography and other common non-invasive examinations.Then make a comprehensive analysis,and establish a regression model of those meaningful indicators.At last,evaluate the value of this model in assessing the risk of esophageal varices and variceal bleeding in liver cirrhosis.This study provides a non-invasive method for liver cirrhosis patients so as to improve their compliance.It can give them timely medical treatment and thus reduce the mortality of patients with cirrhosis.Methods Collect the clinical data of patients with liver cirrhosis caused by chronic hepatitis B and hepatitis C from January,2015 to January,2016 in gastroenterology department and infectious disease department of the First Affiliated Hospital of Zhengzhou University.According to endoscopy,the patients were divided into none esophageal varices,mild esophageal varices,moderate esophageal varices,severe esophageal varices.Collect general condition,sex,age,white blood cell count(WBC),platelet count(PLT),total bilirubin(TB),direct bilirubin(DB),serum albumin(A),serum globulin(G),plasma prothrombin time activity(PTA),portal vein diameter(PVD),spleen vein diameter(SVD),spleen diameter(SD),spleen thickness(ST),liver stiffness measurement(LSM),liver function clssification and other indicators of these patients.First,all patients were divided into two groups,none EV group and EV group.The above indexes were analyzed for differences between these two groups.Multivariate Logistic regression analysis was used to analyze the differences.And the value of the regression model was assessed by receptor operating characteristic curve.Then,patients with EV were divided into none EVB group and EVB group.The above indexes were also analyzed in these two groups.Multivariate Logistic regression analysis was used to analyze the differences.The value of the regression model for evaluating EVB was also assessed by receptor operating characteristic curve.Results In 142 cases of liver cirrhosis patients,there were 32 cases in none EV group and 110 cases in EV group which included 20 mild varices cases,25 moderate varices cases and 65 severe varices cases.Between the two groups,liver function classification,PVD,SVD,SD,ST,WBC,PLT,PLT/SD,LSM were statistically significant.The statistically significant indexes were analyzed by multivariate Logistic regression.ST and LSM finally entered the regression model.We obtained the equation Logit P1=0.112×LSM+0.088×ST-5.011.Then analyze the equation by receiver operating characteristic curve under which the area was 0.821.The cut-off of prediction probability was 0.790,the sensitivity and specificity of this regression model to predict EV was respectively 0.736,0.906.In the 110 cases of liver cirrhosis patients with EV,34 patients were clinical EVB for the first time and 76 patients were EV without bleeding.Liver function classification,LSM,PLT between the two groups were statistically significant.Through the multivariate logistic regression analysis of above statistically significant indexes,only LSM entered the regression model to obtain the equation Logit P2=0.163×LSM-5.164.According to the analysis of receiver operating characteristic curve for this equation,the area under the curve is 0.830.The cut-off of prediction probability was 0.400,the sensitivity was 0.735 and the specificity was 0.868.And the value of LSM was 29.2k Pa.Conclusions Logistic regression model constructed by ST and LSM can be used to evaluate the esophageal varices in patients with viral hepatitis-related cirrhosis.Logistic regression model constructed by LSM can predict the risk of esophageal variceal bleeding in patients of viral hepatitis-related cirrhosis with esophageal varices.