Preliminary Research Of The Assessment Of Indocyanine Green Clearance Test In The Prediction Of Short-term Prognosis For Acute-on-chronic Hepatitis B Liver Failure

Object:Acute-on-chronic hepatitis B liver failure is a clinical severe liver disease, its prognosis is poor. There is no unified standard that the influencing factors and the judgment of the short-term prognosis. Therefore, the main purpose of the present study is to investigate the affecting factors of short-term prognosis of acute-on-chronic hepatitis B liver failure and to analysis the correlation between the indocyanine green retention rate at15minutes (ICGR15) and King’s College Hospital (KCH), Child-Turcotte-Pugh (CTP) classification or model for end-stage liver disease (MELD) score. Then, establishing a prognosis model, compare with the accuracy of the various prognostic model for acute-on-chronic hepatitis B liver failure.Methods:1. According to the diagnostic criteria of acute-on-chronic hepatitis B liver failure, one hundred and twenty-seven patients were recruited in this study. All patients were followed up to three months after diagnosis, then were divided into survival group and death group, and collected the patients clinical data and laboratory indicators.2. In order to choose the indicators that affect the prognosis for patient, we should compare the date between survival group and death group using t test and Wilcoxon signed-rank test or chi-square test, according to the properties of various data.3. The correlation between indocyanine green retention rate at15minutes and difference models (KCH standard, CTP classification and MELD score) was investigated, respectively. The patients were divided into different groups according to the models score, and indocyanine green retention rate at15minutes were compared between different groups.4. ICGR15and MELD score were introduction into logistic regression analysis and the MELD-ICGR15model was constructed. Then the value of various model evaluation on short-term prognosis of patients with HBV-ACLF was investigated, and the forecasting precision of various models were compared.Results:1. Age, total bilirubin, cholesterol, creatinine, cholinesterase, albumin, blood sugar, total bile acid, hemoglobin, prothrombin time, ratio of international standardization, the hyponatremia, hepatorenal syndrome, hepatic encephalopathy and peritonitis are factors that influence the short-term prognosis of patients with HBV-ACLF. However, alanine amino transferase, aspertate glycine transferase, triglycerides, white blood cells, platelets, liver cirrhosis and ascites, upper gastrointestinal bleeding, HBsAg, the state of HBeAg and the load of HBV-DNA have no effect on the prognosis of patients.2. Application of artificial liver can reduce the mortality of patients with early and middle stage, but it can not reduce the mortality of patients with advanced. Antiviral therapy can reduce the mortality rate of patients, but it has no statistical significance between lamivudine treatment group and entecavir treatment group.3. In HBV-ACLF patients, there was a significant positive correlation between ICGR15and KCH (r=0.285, P<0.01), and there was a significant difference between group Ⅰ(KCH=0) and group Ⅱ (KCH=1). There was a significant positive correlation between ICGR15and CTP score(r=0.418, P<0.01), and there was a significant difference between groups (group I:CTP≤10, group Ⅱ:10<CTP≤12, group Ⅲ:CTP≥13). There was a significant positive correlation between ICGR15and MELD score (r=0.489, P<0.01), and there was a significant difference between groups (group Ⅰ: MELD≤20, group Ⅱ:20<MELD≤30, group Ⅲ:30<MELD≤40, group Ⅳ:MELD>40).4. Considerring the outcome of all patients at3months after dianosed to be dependent variable (survival=0, dead=1), MELD score and ICGR15as the independent variable, then we analysis it by binary logistic regression. The MELD score (OR=1.200,95%CI:1.088～1.324, P<0.01) and ICGR15(OR=1.107,95%CI:1.056～1.160, P<0.01) were considered as the risk factors for short-term death. ICGR15and MELD score were introduction into logistic regression analysis, and we constructed the MELD-ICGR15model that the prognosis equation is as follows:Logit (P)=0.183×MELD+0.102×ICGR15-9.499. 5. Area under the curve of various models in the short-term prognosis on HBV-ACLF patients were MELD-ICGR15model (AUC:0.858), ICGR15(AUC:0.796), MELD score (AUC:0.779), KCH criteria (AUC:0.658), CTP score (AUC:0.640), further pairwise comparison shows that the prediction accuracy of the MELD-ICGR15superior to any single model, was statistically significant (P<0.01or P<0.05), respectively.Conclusion:1. The short-term prognosis of patients with HBV-ACLF was influenced by many factors, such as biochemical indicator, coagulation parameters, complications and different treatments.2. In patients with HBV-ACLF, there was a significant positive correlation between ICGR15and the KCH criteria (CTP classification or MELD score). ICGR15was statistically significant in difference grougs, which were divided into difference grougs according to the properties of the various models.3. ICGR15and MELD score had higher predictive value for the HBV-ACLF than CTP classification and the KCH criteria. Furthermore, the MELD-ICGR15model is better than any single forecasting model in predicting the short-term prognosis of patients with HBV-ACLF.