Prognostic Evaluation Of A Prediction Model Based On Neutrophil/Lymphocyte Ratio And Red Blood Cell Distribution On HBV-related Acute-on-chronic Liver Failure

Objective:Acute-on chronic liver failure?ACLF?is a serious of clinical syndromes that occur on the basis of chronic liver disease.The rapid development of the disease often leads to single or multiple organ failure with high mortality.In the diagnosis and treatment of ACLF,the prognostic evaluation is carried out throughout the process,especially the early prognostic evaluation.The current commonly used ACLF prognostic evaluation models in China have some limitations because of the etiology and population.For the Chinese population,the most common cause of ACLF is hepatitis B virus infection.Systemic inflammatory response is closely related to the development and prognosis of liver failure.This study aimed to establish a new model combing the inflammatory indicators including neutrophil/lymphocyte ratio?NLR?and red blood cell distribution width?RDW?with several hematological testing indicators to assess the prognosis of patients with hepatitis B virus-related acute-on-chronic liver failure?HBV-ACLF?.Methods:1.421 patients with HBV-ACLF in Affiliated Hospital of Southwest Medical University from January 2014 to October 2018,56 patients in Neijiang Second People's Hospital from January 2017 to February 2019 and100 patients in Yibin Second People's Hospital from January 2017 to February2019,totally 577 patients'clinical data and laboratory testing indicators were collected.In which,554 patients with MELD score between 9 and 40 were recruited as a derivation cohort to establish the new model.Univariate and multivariate COX regression analysis were used to identify the independent risk factors associated with the prognosis of patients with HBV-ACLF.COX regression was used to establish the predictive death model,and the accuracy of the MELD score,MELD-Na,Child-Turcotte Pugh?CTP?score and new model in predicting the outcome of HBV-ACLF patients was compared by receiver operating characteristic analysis?ROC?.2.ROC curves were adapted to validate the new model in predicting 90-day prognosis in patients with HBV-ACLF in three hospitals,respectively.Medcalc-15.10 software was employed for Data analyses.Results:1.The univariate regression analysis shows that RDW?NLR?INR?TBIL?Cr were significantly elevated in non-survivors?P<0.05?.Multivariate cox regression analysis indicates RDW?HR=1.076?95%CI=1.042¹.110??NLR?HR=1.027,95%CI=1.012¹.042??INR?HR=1.266,95%CI=1.106¹.449??TBIL?HR=1.004,95%CI=1.003¹.005??Cr?HR=1.005,95%CI=1.002¹.007?were risk factors for 90-day death in HBV-ACLF patients?P<0.05?.The risk assessment model is COX_RNTIC=0.073×RDW+0.027×NLR+0.004×TBIL+0.236×INR+0.005×Cr?P<0.001?,with a with a cut-off value of 3.59?sensitivity:78.48%,specificity:84.86%?.The area under the receiver operating characteristics curve?AUC?of the RNTIC was[0.864,95%CI?0.837–0.903?],better than the predictive value of MELD score[0.737,95%CI?0.698⁰.773?]?MELD-Na[0.686,95%CI?0.646⁰.725?]?CTP[0.670,95%CI?0.629⁰.709?]?P<0.001?2.In the validation cohort,RNTIC also performed a better prediction value than the others in three hospitals.The AUC of RNTIC,MELD,MELD-Na and CTP were[0.732,95%CI?0.687⁰.774?],[0.714,95%CI?0.668⁰.757?]and[0.703,95%CI?0.657⁰.747?]in the Affiliated Hospital of Southwest Medical University;[0.870,95%CI?0.789⁰.928?],[0.763,95%CI?0.669⁰.841?],[0.766,95%CI?0.672⁰.843?]and[0.663,95%CI?0.564⁰.754?]in the Second People's Hospital of Yibin;[0.877,95%CI?0.758⁰.951?],[0.747,95%CI?0.608⁰.856?],[0.725,95%CI?0.601⁰.851?]and[0.674,95%CI?0.531⁰.796?]in the First People's Hospital of Neijiang,respectively.Conclusion:1.In HBV-ACLF patients,compared with the survival group,the NLR and RDW values of inflammation markers in the death group were significantly increased.2.NLR and RDW values were positively correlated with MELD scores in patients with HBV-ACLF and NLR,RDW were independent risk factors for short-term death in HBV-ACLF patients in multivariate regression analysis,indicating that NLR and RDW may be related to the prognosis of the disease.3.COX multivariate regression analysis showed that NLR,RDW,Cr,TBIL,and INR were independent risk factors for 90-day death in HBV-ACLF patients.Therefore,the RNTIC model was established to predict the short-term prognosis of HBV-ACLF patients.4.The RNTIC model showed excellent predictive performance both in the derivation and validation cohorts,which proved that it could be used to evaluate the early death of HBV-ACLF patients and to evaluate the survival potential of conservative treatment in patients at an early stage,providing more support for doctors to decide if patients need liver transplantation.