摘要
目的筛选外周血液检测指标作为自身免疫性肝炎(AIH)患者肝硬化预测的关键因子,并构建无创诊断模型。方法回顾性分析苏州市第五人民医院2021年7月至2024年5月临床确诊为AIH且接受肝活检病理组织学检查的患者48例,分析其22项血液检测指标。根据肝活检病理组织学检查分为AIH患者非肝硬化组和肝硬化组,通过比较筛选出血液指标中的关键预测诊断因子,采用SPSS软件进行二元Logistic回归分析,联合关键因子构建AIH患者肝硬化预测诊断模型。再选取苏州市第五人民医院2024年6月至8月诊治的AIH患者10例作为测试集样本,进行模型验证评价。并与现有的外周血肝纤维化诊断模型包括基于4因子的纤维化指数(FIB-4)、门冬氨酸氨基转移酶-丙氨酸氨基转移酶(AAR)和门冬氨酸氨基转移酶/血小板计数比值(APRI)的效能进行对比。结果国际标准化比值(INR)、总胆红素、直接胆红素和IgA在AIH患者非肝硬化组和肝硬化组间差别均有统计学意义(P<0.05)。受试者工作特征曲线(ROC)分析显示,国际标准化比值、总胆红素、直接胆红素和IgA在AIH患者肝硬化预测诊断中均有一定的价值。二元Logistic回归分析结果显示,联合上述4项关键血液检测指标构建最佳AIH患者肝硬化预测诊断模型,其AUC为0.799(95%CI 0.642~0.955),灵敏度为81.8%,特异性为75.7%,预测诊断准确率为81.2%,测试集AIH患者的肝硬化预测诊断准确率为70.0%。FIB-4在两组间差别有统计学意义(P<0.05),其诊断AIH患者肝硬化的AUC为0.698,灵敏度为72.7%,特异性为67.6%。结论国际标准化比值、总胆红素、直接胆红素和IgA在构建AIH患者肝硬化预测诊断中具有较好的应用价值,联合上述4项血液检测指标构建的预测诊断模型有望为临床上AIH患者肝硬化的预测诊断提供理论依据。
Objective To screening blood test indicators as key factors in the prediction of cirrhosis in patients with autoimmune hepatitis(AIH) and to construct a noninvasive diagnostic model.Methods 48 patients with AIH diagnosed and treated in the Fifth People's Hospital of Suzhou from July,2021 to May,2024 were retrospectively analyzed.Liver biopsy histopathological examination and 22 items of blood tests were all performed in these subjects.According to liver biopsy histopathology,AIH patients were divided into the non-cirrhosis group and cirrhosis group.The key predictive diagnostic factors in blood indexes were screened by the comparative analysis.SPSS software was used for binary logistic regression analysis,and key factors were combined to construct the prediction and diagnosis model of cirrhosis in AIH patients.10 AIH patients diagnosed and treated in the Fifth People's Hospital of Suzhou from June to August,2024 were selected as the test set for model validation.The results were compared with existing diagnostic models based on peripheral blood for hepatic fibrosis,including 4-factor based fibrosis index(FIB-4),aspartate aminotransferase-alanine aminotransferase(AAR) and aspartate aminotransferase/platelet count ratio(APRI).Results There were significant differences in international normalized ratio(INR),total bilirubin,direct bilirubin and IgA between AIH patients with non-cirrhosis and cirrhosis(P<0.05).Receiver operating characteristic curve(ROC) analysis showed INR,total bilirubin,direct bilirubin and IgA were all valuable in the prediction and diagnosis of cirrhosis in patients with AIH.Binary logistic regression analysis showed that the optimal liver cirrhosis prediction and diagnosis model for AIH patients was constructed by combining the above four key blood test indicators.The AUC was 0.799(95% CI 0.642-0.955),with the sensitivity of 81.8% and the specificity of 75.7%,while the accuracy of prediction diagnosis was 81.2%,and the accuracy of prediction diagnosis of cirrhosis in patients with AIH was 70.0%.FIB-4 was significantly different between the two groups(P<0.05),but AAR and APRI were not(P>0.05).The AUC for FIB-4 diagnosis of cirrhosis in AIH patients was 0.698,with the sensitivity of 72.7% and the specificity of 67.6%.Conclusion INR,total bilirubin,direct bilirubin and IgA have good application values in the prediction and diagnosis of cirrhosis in patients with AIH.The prediction and diagnosis model combined with the above four blood test indexes is expected to provide a theoretical basis for the prediction and diagnosis of cirrhosis in patients with AIH.
作者
吴敏娟
宋翌琰
周芬
黄丽娜
孙艳
胥萍
WU Minjuan;SONG Yiyan;ZHOU Fen;HUANG Lina;SUN Yan;XU Ping(Department of Clinical Laboratory,the Fifth People’s Hospital of Suzhou,Infectious Disease Hospital Affiliated to Soochow University,Suzhou 215000,China;Department of Pathology,the Fifth People’s Hospital of Suzhou,Infectious Disease Hospital Affiliated to Soochow University,Suzhou 215000,China)
出处
《标记免疫分析与临床》
2024年第12期2259-2265,共7页
Labeled Immunoassays and Clinical Medicine
基金
苏州市卫生健康委员会科技计划项目(编号:LCZX202217)。
