Period Effect Of Leukocyte Derived Inflammatory Markers On Prognosis Of Patients With Acute-on-chronic Liver Failure Caused By Hepatitis B

The purpose of this study was to investigate the value of serum NLR,MLR,NMR and other leukocyte derived indicators in predicting the prognosis of chronic hepatitis B ACLF;And construct NLR related 28 day and 90 day mortality risk models to predict the mortality of chronic hepatitis B ACLF patients at 28 days and 90 days.Methods: 137 patients with chronic hepatitis B(ACLF)diagnosed in the Department of Gastroenterology of the First Affiliated Hospital of the Air Force Military Medical University from January 2016 to December 2021 were included.The diagnosis of chronic hepatitis B shall be in accordance with the Guidelines for the Prevention and Treatment of Chronic Hepatitis B(2022);ACLF diagnosis is based on the criteria of "Consensus on Management of Chronic Acute Liver Failure(2019 Edition)" of Asia Pacific Liver Research Society.The patients were divided into 28 day death group(42cases),28 day survival group(95 cases),90 day death group(61 cases)and 90 day survival group(76 cases)according to their survival status.Univariate and multivariate logistic regression analysis were used to screen the variables significantly related to prognosis;The 28 day and 90 day mortality risk models were constructed by Cox proportional hazards regression model,and the models were statistically verified;According to the Logistic regression quantitative model,draw ROC and evaluate the effectiveness of the model.Results:A total of 137 patients(111 men,accounting for 80.9%)were included,44 patients died within 28 days(31.2%),and 48 patients died within 90 days(44.7%).The NLR between the 28 day survival group and the death group was 3.6(2.4-5.8);5.8(3.4-9.4);P=0.001.MLR is 0.5(0.3-0.7)respectively;0.8(0.4-1.2);P=0.003。MELD scores were 21.4 ± 0.5;25.0±1.2;P=0.023。The NLR between the 90 day survival group and the death group was 3.4(2.3-5.5)respectively;5.4(3.2-8.1);P=0.018.MLR is 0.5(0.3-0.7)respectively;0.7(0.5-1.0);P=0.006.MELD scores were 21.4 ± 0.5;25.1±1.2;P=0.001.The ratio of leukocytes in the death group at 28 days and 90 days was generally higher than that in the survival group.From baseline to the 28 th day,the dynamic curves of NLR and NMR in the survival group showed a downward trend,while those in the death group showed an upward trend;The trend of MLR is opposite.The MELD score(P=0.000),NLR(P=0.001),NMR(P=0.072)(AIC=539.352)were finally selected as independent risk factors affecting the 28 day prognosis by single factor regression,elastic network and full subset regression;MELD score(P=0),NMR(P=0.060),MLR(P=0),Plt(P=0.002)(AIC=527.758)are independent risk factors affecting 90 day prognosis.Based on the multivariate COX regression results,the 28 day mortality risk model: MELD score HR(95% CI)=1.083(P=0.001),NLR HR(95% CI)=1.082(P=0.001),NMR HR(95% CI)=1.004(P=0.021);90 day mortality risk model MELD score HR(95%CI)=1.097(P=0.000),NMR HR(95% CI)=1.005(P=0.003),MLR HR(95% CI)=1.856(P=0.003),PltHR(95% CI)=0.99(P=0.003);The 28 day and 90 day mortality risk models were constructed(see appendix).Compare the prediction model values,and draw ROC curve jointly with MELD score,Cliff SOFA score,and AARC score: the area under the curve of the 28 day and 90 day prediction model is 0.758(P<0.05)and 0.786(P<0.05)respectively,which are larger than the area under the curve of other models,with the best prediction efficiency and high accuracy.Conclusion: NLR,a leukocyte related derivative index,may be used to predict the28 day prognosis of patients,and MLR can predict the 90 day prognosis of patients.And the increase of NLR is positively related to the risk of death.Leukocyte related derived indicators combined with MELD mortality risk prediction model may be simple and fast to predict the prognosis of patients without liver transplantation,which is helpful to make decisions on early liver transplantation for patients with poor prognosis.