基于肝脾多模态超声构建肝硬化食管胃底静脉曲张破裂出血风险Nomogram无创预测模型

Establishment of a Nomogram noninvasive prediction model for the risk of esophageal and gastric varices bleeding in liver cirrhosis based on multimodal ultrasound of liver and spleen

目的探讨肝脾血流动力学及肝脾硬度和实验室检查结果对肝硬化食管胃底静脉曲张程度的预测价值,筛选独立危险因素,通过列线图得到定量模型。方法选择河北医科大学第二医院符合标准的乙型肝炎肝硬化患者128例纳入研究,根据内镜检查结果,将无静脉曲张(G0)和轻度静脉曲张(G1)的患者分为低风险组;把中度(G2)和重度(G3)静脉曲张的患者分为高风险组。获得2组身高、体重、体重指数、门静脉直径、门静脉流速、肝动脉收缩期流速、舒张期流速、阻力指数、脾脏厚度、脾门处脾静脉直径、脾静脉流速、肝脾硬度值、白细胞、红细胞、血红蛋白、血小板、天冬氨酸转氨酶、丙氨酸转氨酶、白蛋白、总胆红素、肌酐等数据。通过单因素分析和多因素Logistic回归分析筛选出预测食管胃底静脉曲张程度的独立危险因素,利用受试者工作特征曲线(receiver operating characteristic,ROC)分析各因素及复合指标的诊断效能,通过列线图得到定量模型。结果单因素分析结果显示2组门脉内径、门脉流速、肝硬度、脾硬度、红细胞计数比较差异有统计学意义(P<0.05),进一步的多因素Logistic回归分析结果显示门脉流速、肝硬度、脾硬度是高危食管胃底静脉曲张的独立危险因素(P<0.05),ROC曲线分析显示各因素具有较高的敏感度和特异度,尤其脾硬度指标曲线下面积高达0.92。结论基于肝脾多模态超声可以对乙型肝炎肝硬化食管胃底静脉曲张破裂出血风险进行定量预测,脾硬度指标具有较高的诊断效能。

Objective To explore the value of hemodynamics of liver and spleen,stiffness of liver and spleen and laboratory test results in predicting the degree of esophageal and gastric varices in hepatitis B-induced liver cirrhosis patients,to screen the independent risk factors and to obtain the quantitative nomogram model.Methods In total,128 patients with hepatitis B-induced liver cirrhosis who met the criteria were included in the Second Hospital of Hebei Medical University.According to the results of endoscopy,patients without varicose veins(G0)and patients with mild varicose veins(G1)were selected as low-risk group,and patients with moderate(G2)and severe(G3)varicose veins as high-risk group.The data including height,weight,body mass index(BMI),portal vein diameter,portal vein velocity,hepatic artery systolic velocity,diastolic velocity,resistance index,spleen thickness,splenic vein diameter at the splenic hilum,splenic vein velocity,liver and spleen stiffness values,white blood cells,red blood cells(RBC),hemoglobin,platelets,aspartate aminotransferase,alanine aminotransferase,albumin,total bilirubin and creatinine were collected from the two groups.Independent risk factors for predicting the degree of esophageal and gastric varices were screened through univariate analysis and multivariate logistic regression analysis.Receiver operating characteristic(ROC)curves were used to analyze the diagnostic efficacy of each factor and composite index,and a quantitative model was obtained through a nomogram.Results The results of univariate analysis showed that there was a significant difference in portal vein diameter,portal vein velocity,liver stiffness,spleen stiffness,and RBC count between the two groups(P<0.05).Further multivariate logistic regression analysis showed that portal vein velocity,liver stiffness,and spleen stiffness were independent risk factors for high-risk esophageal and gastric varices(P<0.05).ROC curve analysis showed that each factor had high sensitivity and specificity,especially the area under ROC curve of the spleen stiffness was as high as 0.92.Conclusion Multimodal ultrasound of liver and spleen can quantitatively predict the risk of esophageal and gastric varices bleeding in patients with hepatitis B-induced liver cirrhosis,the index of spleen stiffness has high diagnostic efficiency.