兰州大学学报（医学版）2025，Vol.51Issue(4)：48-58,11.DOI:10.13885/j.issn.1000-2812.2025.04.007
HBV相关ACLF人工肝治疗新预后模型的建立及评估
Development and validation of a novel prognostic model for patients with acute-on-chronic liver failure undergoing artificial liver support therapy
摘要
Abstract
Objective To investigate the risk factors associated with abiotic artificial liver therapy that impact the 90-day prognosis of patients suffering from hepatitis B virus(HBV)infection-associated chronic acute liver failure(ACLF),and to establish a prognostic model while evaluating the diagnostic efficacy of the newly developed model.Methods A total of 168 patients with acute-on-chronic liver failure undergoing artificial liver support system were enrolled in our study conducted at the First Hospital of Lanzhou University between January 2017 and October 2023.The patients were divided into a training set comprising 117 cases and a test set of 51.The related clinical data were collected.According to the 90-day survival status of patients in the training set,they were divided into a survival group with 51 patients and a death group with 66.Single factor analysis and multivariate Logistic regression were used to analyze the risk factors of HBV-ACLF patients treated by artificial liver support system and a novel predictive model established.The calibration curve,mean square error score,mean absolute error score,and Brier score were used to evaluate the accuracy of the new predictive model,and the receiver operating characteristic(ROC)curve of the new predictive model was drawn and compared with liver failure prognostic predictive models such as model for end-stage liver disease(MELD)score,MELD-Na score,albumin-bilirubinc(ALBI)score,and COSSH-ACLF Ⅱ score to evaluate the diagnostic efficacy of the new predictive model.Results The results showed that platelet count(PLT),neutrophil/lymphocyte ratio(NLR),and Child-Turcotte-Pugh score were independent risk factors for short-term prognosis(P<0.05).A novel predictive model,i.e.Child-Turcotte-Pugh-NLR-PLT(CNP),was construct-ed using Logistic regression,and ROC curves were plotted.The results revealed that the area under the curve(AUC)for the CNP model,PLT,NLR,and Child-Turcotte-Pugh were 0.805,0.731,0.653,and 0.654 respec-tively.The AUC of the CNP model was significantly higher than those of the individual influencing factors.Furthermore,compared to MELD,MELD-Na,ALBI,and COSSH-ACLF Ⅱ scores,the CNP score demon-strated superior diagnostic performance,with an AUC of 0.805 compared to 0.632,0.625,0.693,0.626 respec-tively,and similar results were obtained in the external dataset.According to the Diagnosis and Treatment Guidelines for Liver Failure(2018),all patients were divided into 3 subgroups:early stage,middle stage,and late stage of liver failure.The ROC curves were drawn for different liver failure subgroups.The results showed that the AUC were 0.810 in early stage of liver failure,0.752 in middle stage and 0.793 inlate stage,and CNP model still had certain diagnostic value in different stages of liver failure.Conclusions Child-Turcotte-Pugh,NLR,PLT were found to be independent risk factors for high mortality in HBV-ACLF patients treated with artificial liver support system.We built a new model containing Child-Turcotte-Pugh,NLR,PLT,namely CNP,which showed potentially better prediction performance than MELD,MELD-Na,ALBI,and COSSH-ACLFⅡscore for short-term outcomes,and this could aid physicians in making optimal therapeutic decisions.关键词
乙型肝炎病毒/慢加急性肝衰竭/人工肝/预后/预测模型Key words
hepatitis B virus/acute-on-chronic liver failure/artificial liver/prognosis/predictive model分类
医药卫生引用本文复制引用
彭雪彬,李泽阳,尤泰福,次旦央宗,刘雄雄,毛小荣..HBV相关ACLF人工肝治疗新预后模型的建立及评估[J].兰州大学学报（医学版）,2025,51(4):48-58,11.基金项目
北京肝胆相照公益基金会资助项目(RGGJJ-2021-004) （RGGJJ-2021-004）
兰州大学第一医院院内基金资助项目(ZX-62000002-2021-226) （ZX-62000002-2021-226）
甘肃省教育科技创新基金资助项目(2022B-007) （2022B-007）
甘肃省感染肝病临床研究中心基金资助项目(21R7RA392) （21R7RA392）
甘肃省科技厅重点研发计划资助项目(24YFFA037) （24YFFA037）
