摘要
目的:通过建立新的无创血清学模型,以评估其对慢性乙型肝炎(CHB)相关肝纤维化分期的预测价值,并与其他预测方法进行评估对比。方法:采用回顾性分析CHB导致肝纤维化患者的临床数据资料,根据肝活组织检查结果,将符合条件的病例分为无显著肝纤维化组(S 0/1)和显著肝纤维化组(S≥2),收集其血清学指标、影像学检查、天冬氨酸氨基转移酶/血小板比值指数(APRI)、4因子模型(FIB-4)、Forns指数等数据,运用t检验、二元Logisti c回归分析等方法建立了无创肝纤维化新模型(CF模型),并使用受试者工作特征(ROC)曲线,比较新模型与各个检验指标的预测准确性。结果:S 0/1组(34例)和S≥2组(20例)患者γ-谷氨酰转肽酶(GGT)、血小板计数(PLT)、Gamma蛋白、Beta1蛋白、Beta2蛋白、层粘蛋白(LN)、甲胎蛋白(AFP)、糖类抗原(CA199)、血清铁(Fe)、转铁蛋白(TFR)、未饱和铁结合力(UIBC)的差异比较均有统计学意义;根据回归分析结果建立[logit(CF)=-4.918+0.051×CA199+0.133×Fe]新模型,其ROC曲线下面积(AUC)为0.843,优于APRI、FIB-4评分及Forns指数评分的预测价值,差异有统计学意义(P<0.05),其AUC分别为APRI(0.635)、FIB-4(0.608)、Forns(0.650);优于声辐射脉冲成像(ARFI)、Fibroscan瞬时弹性成像的影像诊断价值,AUC分别为0.840、0.782,但差异无统计学意义(P>0.05)。结论:CF模型对于诊断显著肝纤维化(S≥2)的预测价值不仅高于APRI、FIB-4和Forns的模型评分,且与ARFI和Fibroscan影像学单独检查的效用相近,其评估价值可能更大。
Objective:A new noninvasive serological model was developed to evaluate the predictive value of Chronic hepatitis B CHB with respect to related liver fibrosis staging and to compare it with other established predictive methods.Methods:The hepatic fibrosis due to chronic hepatitis B patients were retrospectively analyzed the clinical data,according to the result of liver biopsy,eligible cases can be divided into group had no significant liver fibrosis(S 0/1)group and significant liver fibrosis(S≥2),collect the serological indexes,imaging examination,aspartate amino transferase/platelet ratio index(APRI),four factor model(FIB-4),Forns index data,using t test and binary logistic regression analysis method to establish a noninvasive liver fibrosis(CF)model,a new model.The predictive accuracy of the new model and each test indicator was compared using the Receiver Operating Characteristic(ROC)curve.Results:The difference of Gamma glutamyl transpeptidase(GGT),platelet count(PLT),Gamma,walk,beta 2,mucin layer(LN),a tire protein(AFP),carbohydrate antigen(CA199),serum iron(Fe),transferrin(TFR),unsaturated iron binding force(UIBC)betwen S 0/1 group(34 cases)and S≥2 group(20 cases)had statistical significance.In addition,according to the results of regression analysis,the new model[logit(CF)=-4.918+0.051×CA199+0.133×Fe]was established.The area under ROC curve(AUC)was 0.843,which was superior to the predictive value of APRI,FIB-4 score and Forns index score,and the difference was statistically significant(P<0.05).The AUC was APRI(0.635),FIB-4(0.608)and Forns(0.650),respectively.It was also better than the imaging diagnostic value of acoustic radiation impulse imaging(ARFI)and instantaneous elastic imaging(Fibroscan),and the AUC is ARFI(0.840)and Fibroscan(0.782),respectively,but the difference is not statistically significant(P>0.05).Conclusion:CF model has a higher predictive value for the diagnosis of significant liver fibrosis(S≥2)than the model scores of APRI,FIB-4 and Forns,and is similar to the individual studi
作者
魏鑫
徐懂
龙细雨
伍喜良
郭丽颖
贾建伟
WEI Xin;XU Dong;LONG Xi-yu;JIA Jian-wei(Graduate School,Tianjin University of Chinese Medicine,Tianjin,301617,China;不详)
出处
《中西医结合肝病杂志》
CAS
2021年第10期869-873,共5页
Chinese Journal of Integrated Traditional and Western Medicine on Liver Diseases
基金
天津市卫生健康委员会中医药重点领域科研项目(No.2020006)
天津市卫生健康委员会天津市中医药管理局中医中西医结合科研课题(No.2019132)。
