目的对门静脉高压症患者进行影像学评分,以预测其出血风险。方法回顾性分析南京大学医学院附属鼓楼医院2008年6月至2014年6月期间收治的所有乙型肝炎门静脉高压症患者的临床和影像学相关数据,筛除了肝细胞癌患者、伴有门静脉血栓患者及自身免疫性肝炎、胰腺炎、血液疾病等导致的门静脉高压症患者。结果本研究共纳入了98例门静脉高压症患者,根据患者是否出血分为出血组(57例)和未出血组(41例)。2组患者的临床指标比较:出血组和未出血组在年龄、凝血酶原时间、血清白蛋白、血清肌酐、血清钠、血白细胞及血小板计数方面比较差异均无统计学意义(P>0.05);未出血组患者的血清总胆红素和血红蛋白高于出血组(P=0.023、P=0.000),肝功能比出血组好(P=0.039)。2组患者的影像学指标比较:门静脉高压症患者发生出血与门静脉直径、肠系膜上静脉直径、肠系膜下静脉直径、脾静脉直径、附脐静脉直径、冠状静脉直径、脾肾分流直径、胃短静脉直径、奇静脉直径及脾脏体积均无关(P>0.05),而其与胃后静脉直径(P=0.028 3)及食管曲张静脉分级(P=0.022 1)有关,即胃后静脉直径越大及食管曲张静脉分级越高,出血风险越小。将出血因素分析中P<0.20的指标纳入logistic回归分析建立预测模型,肠系膜下静脉直径、胃后静脉直径、食管曲张静脉分级及胃短静脉直径被加入到该模型中,然后这些指标进行评分,结果发现,随着评分的增加而出血的风险增加。另选取2014年7–12月期间的26例门静脉高压症患者对该模型进行验证,其结果和本研究建立的模型在一定程度上是吻合的,受试者操作特征曲线下面积为0.884 9。结论从本研究有限的数据初步研究结果来看,建立的包括肠系膜下静脉直径、食管曲张静脉分级、胃后静脉直径及胃短静脉直径的影像学评分预测模型,可以较好地预测门静脉高压症患者的出血风险,评分越高,出血的风险越大,不过还需要进一步扩大样本量及设计前瞻性研究来验证此模型。 Objective To evaluate the imaging risk of portal hypertension in order to predict the risk of bleeding. Methods The clinical and radiographic data of all patients with hepatitis B and portal hypertension who were admitted to Gulou Hospital Affiliated to Nanjing University Medical College from June 2008 to June 2014 were retrospectively analyzed. Screening patients with hepatocellular carcinoma with portal vein thrombosis Patients and autoimmune hepatitis, pancreatitis, blood diseases caused by patients with portal hypertension. Results A total of 98 patients with portal hypertension were enrolled in this study. They were divided into hemorrhage group (57 cases) and non-hemorrhage group (41 cases) according to whether hemorrhage occurred. There was no significant difference in the clinical indexes between the two groups of patients in terms of age, prothrombin time, serum albumin, serum creatinine, serum sodium, white blood cells and platelet count (P> 0.05) The serum total bilirubin and hemoglobin in patients without hemorrhage were higher than those in hemorrhage (P = 0.023, P = 0.000), and liver function was better than that in hemorrhage (P = 0.039). Two groups of patients with imaging criteria comparison: patients with portal hypertension bleeding and portal vein diameter, the diameter of the superior mesenteric vein, the diameter of the inferior mesenteric vein, the diameter of the splenic vein, the diameter of the attached umbilical vein, the diameter of the coronary vein, the diameter of the splenorenal shunt, the short gastric veins (P = 0.028 3) and esophageal varices (P = 0.022 1), which means that the larger the diameter of the posterior venous vein and the diameter of the vena cava are Esophageal varicose vein grading higher, the smaller the risk of bleeding. The index of P <0.20 in the analysis of bleeding factor was included into the logistic regression analysis to establish the prediction model. The values ​​of the diameter of inferior mesenteric vein, the diameter of inferior vena cava, the classification of esophageal varices and the diameter of short gastric vein were added to the model. Then, As a result, the risk of bleeding was increased as the score increased. This model was also validated in 26 patients with portal hypertension from July to December 2014. The results were in good agreement with the model established in this study, with an area under the operating characteristic curve of 0.884 9. Conclusions From the preliminary results of the limited data in this study, the established imaging predictive models including the size of the inferior mesenteric vein, esophageal varices, the diameter of the posterior gastric vein and the diameter of the short gastric vein can be used to predict portal hypertension Patients with bleeding risk, the higher the score, the greater the risk of bleeding, but also need to further expand the sample size and design a prospective study to validate the model.