Establishment And Clincial Application Of The Esophageal Variceal Bleeding Prediction Model In Chronic Liver Failure

Background and Aims Liver failure is the inability of the liver to perform its normalsynthetic and metabolic function as part of normal physiology when large parts of theliver become damaged beyond repair.It is a life-threatening condition that sometimesdemands urgent medical care.Most often liver failure occurs gradually and overmany years.Chronic liver failure usually occurs in the context of cirrhosis,itselfpotentially the result of many possible causes.Variceal bleeding carries a highmortality rate in patients with liver cirrhosis.Bleeding from esophageal varices is aleading cause of death in chronic liver failure.Severity of the bleeding episode,failure to control bleeding and early recurrence are indicators of poor outcome.Allthese factors depending on the severity of the underlying liver dysfunction.Increasedfrequency of hyperfibrinolytic activity was reported in patients with cirrhosis.However,the incidence,clinical presentation and the parameters related tohyperfibrinolysis remain largely unknown in these patients.Accelerated intravascularcoagulation with secondary hyperfibrinolysis has been reported in patients with liverfailure.Hyperfibrinolysis may delay primary hemostasis,thereby aggravating varicealbleeding and facilitating recurrence.This study try to find the value of the dynamic changes of plasma D-dimer,which may have correlations with variceal bleeding in cirrhotic patients.Toinvestigate the relationship between Model for End stage Liver Disease (MELD) orCTP and plasma D-dimer in the patients with chronic liver failure.Also,thecorrelation among infection,endotoxaemia and hypertibrinolysis should be found ifexists.One of the aims of this study was to ascertain whether or not the dynamicchanges of plasma D-dimer evaluated serially are independent factor relate to theclinical outcome of patients with chronic liver failure,no matter complicate varicealhaemorrhage or not.Some factors,include MELD,plasma D-dimer and the bleedinggrading score combined to predict the risk of massive hemorrhage of esophagealvarices in cirrhosis patients especially with chronic liver failure.Then,the patientwith chronic liver failure who has a high risk of massive hemorrhage of esophagealvarices would be suggested to take the therapeutic endoscopy.The death rate observed to assess the prognostic significance of using the model in high risk patientswith massive hemorrhage of esophageal varices in chronic liver failure.Methods 1.Patients with chronic liver failure was divided into two groups:bleedinggroup (group A,n=50) and non-bleeding group (Group B,n=50).Plasma D-dimerwas compared between two groups.In the group A,serial D-dimer measurementswere performed in cirrhotics bleeding from esophageal varices at baseline and duringfollow-up(3 days,8 days,15 days after the bleeding onset) and in 50 non-bleedingcirrhotic patients at baseline only.Both MELD score and CTP score were computedfor each patient according to the original formula on admission day.Area under thereceiver operating characteristic curve (ROC) and cutoff value were used to assess thevalue of MELD score for predicting the prognosis and so as the value of bleedinggrading score for predicting the bleeding from esophageal varices.Some factors likeMELD,plasma D-dimer and the bleeding grading score were integrated into a modelfor predicting the risk of massive hemorrhage of esophageal varices in patients withchronic liver failure.2.Forty-eight patients with chronic liver failure were studied and observed indifferent groups according to using the model which include MELD,plasma D-dimerand bleeding grading score.The death rate was observed within 2 weeks in differentgroups.The predictive of the model and effect of endoscopic therapy was evaluatedby the odd ratio of death.The predictive accuracy of MELD was evaluated by ROCcurve.Results 1.Patients with chronic liver failure and high levels of plasma D-dimer are athigh risk of bleeding.The clinical and biochemical parameters didn't show significantdifferences in both groups.Baseline measurements of coagulation activation andfibrinolysis were more impaired in bleeders.Average level of plasma D-dimer in thegroup A(the first one) were significantly higher than the group B.In bleeders,levelsof plasma D-dimer were persistently more abnormal in patients who died ultimately.CTP and MELD were the significant predictors of death.High levels of plasmad-dimer at the time of bleeding and infection are the dangerous factors of death.CTPand MELD were independent predictors of death when multivariate analyses wasused to assess the independent prognostic value of these variables.The cutoff value of MELD to predict death is 26.The cutoff value of D-dimer to predict death is 700ug/L.The cutoff value of bleeding grading score to predict bleeding is 10..The trendanalysis shows the changing tendency of the average level of plasma D-dimer inbleeders:the plasma D-dimer will reach high peak on the third day after bleeding andthen descend gradually until back to the original state around 2 weeks after bleeding.The model for predicting the risk of massive hemorrhage of esophageal varices inpatients with chronic liver failure is:MELD≥26 and bleeding grading score≥10and/or plasma d-dimer＞700ug/L2.CTP and plasma D-dimer showed significant differences between twogroups while the other parameters of clinical and biochemical didn't.The odd ratio ofdeath is 2.94,95% CI(1.23,7.04).Multivariate analysis disclosed no markerpredictive of death after endoscopic therapy.Conclusions Among the patients with chronic liver failure:1.Infection andendotoxin level may correlated with death.The higher fibrinolytic activity,the higherrisk of bleeding which may affected by infection.2.Cutoff value of MELD forpredicting the death is 26.3.Cutoff value of plasma D-dimer(first time from bleedingonset) for predicting the death is 700ug/L.4.The average value of plasma D-dimerreach the peak 3 days after bleeding onset,then descend gradually until back to theoriginal state around 2 weeks after bleeding.5.Cutoff value of bleeding grading scorefor predicting the variceal bleeding is 10.6.The model for predicting the risk ofmassive hemorrhage of esophageal varices in patient is:MELD≥26 and bleedinggrading score≥10 and/or plasma D-dimer＞700ug/L.7.The model which includeMELD score combined with bleeding grading score and/or plasma D-dimer shouldsucceed in predicting the prognosis of chronic liver failure as an useful predictor ofesophageal variceal bleeding.Moreover,high risk population with massivehemorrhage of esophageal varices in chronic liver failure could be predicted anddiscovered by using the model.Endoscopic therapy would be suggested and may playan important role to improve the prognosis of these patients then.