钆塞酸二钠增强MRI列线图模型预测小肝癌微血管侵犯的价值  

The value of Gd-EOB-DTPA-enhanced MRI nomogram model in predicting microvascular invasion of small solitary hepatocellular carcinoma

作　　者：田雅琪 胡亚彬[2] 彭琪琪 苏晓慧[2] 王晓琳 陈静静[2] TIAN Yaqi;HU Yabin;PENG Qiqi;SU Xiaohui;WANG Xiaolin;CHEN Jingjing(School of Medicine,Qingdao University,Qingdao 266100,China;Department of Radiology,Affiliated Hospital of Qingdao University,Qingdao 266016,China)

机构地区：[1]青岛大学医学部,青岛266100 [2]青岛大学附属医院放射科,青岛266016

出　　处：《磁共振成像》2021年第10期57-60,65,共5页Chinese Journal of Magnetic Resonance Imaging

基　　金：青岛市民生科技计划项目(17-3-3-22-nsh)。

摘　　要：目的探讨钆塞酸二钠(gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid,Gd-EOB-DTPA)增强MRI列线图模型预测孤立性小肝癌(≤3 cm)微血管侵犯(microvascular invasion,MVI)的价值。材料与方法回顾性分析青岛大学附属医院2017年1月1日至2020年6月30日病理证实的149例单发肝细胞癌(≤3 cm)患者,术前1个月内进行了Gd-EOB-DTPA增强MRI检查。根据病理学诊断的微血管侵犯情况将病例分为MVI阳性组及MVI阴性组,收集术前临床资料、Gd-EOB-DTPA增强MRI资料及病理资料。将所收集资料进行单因素分析,计数资料使用χ^(2)检验或Fisher精确检验,计量资料使用Mann-Whitney U检验,基于单因素分析有意义(P<0.05)的术前资料进行多因素Logistic回归分析,根据所得结果将预测MVI的独立危险因素建立列线图预测模型,并评价其预测效能。结果将MVI阳性组(45例)及阴性组(104例)单因素分析有意义资料纳入Logistic回归分析,结果显示假包膜不完整(OR=5.951,95%CI:2.303~15.372,P<0.001)、瘤周强化(OR=3.834,95%CI:1.301~11.299,P=0.015)、肝胆期瘤周低信号(OR=9.721,95%CI:2.486~38.019,P=0.001)及较低肿瘤ADC值(低于0.990×10^(-3)mm^(2)/s)是预测孤立性小肝癌MVI的独立危险因素,ROC曲线下面积为0.844。根据多因素结果建立列线图预测模型,显示肝胆期瘤周低信号是预测孤立性小肝癌MVI的最大贡献因子(100分),其次为假包膜完整程度、瘤周强化、肿瘤ADC值;该模型C指数为0.844,具有较高的校准度。结论Gd-EOB-DTPA增强MRI检查列线图模型可以有效预测孤立性小肝癌MVI的发生。Objective:To explore the value of the nomogram based on gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid(Gd-EOB-DTPA)enhanced MRI of the microvascular invasion(MVI)in small solitary hepatocellular carcinoma(≤3 cm).Materials and Methods:In this study,we retrospectively analyzed 149 patients with small solitary hepatocellular carcinoma(≤3 cm)who underwent preoperative Gd-EOB-DTPA-enhanced MRI between January 2017 and June 2020.All patients were grouped into MVI-negative and MVI-positive cases according to the histopathological diagnosis.Then,preoperative clinical,Gd-EOB-DTPA-enhanced MRI and pathological data were collected.Univariate and multivariate Logistic regression analyses were used to analyze the relevant risk factors for MVI,and then a nomogram incorporated preoperative data was developed by multivariate Logistic regression.Finally,the prediction efficiency of the model is evaluated.Results:Analysis of the MVI-positive group(45 cases)and the MVI-negative group(104 cases)showed peritumoral hypointensity on HBP(OR=5.951,95%CI:2.303—15.372,P<0.001),peritumor enhancement(OR=3.834,95%CI:1.301—11.299,P=0.015),peritumoral pseudocapsule incomplete(OR=9.721,95%CI:2.486—38.019,P=0.001),the lower ADC values(<0.990×10^(-3)mm^(2)/s)were independent predictors of MVI of small solitary hepatocellular carcinoma.The area under the ROC curve was 0.844.A nomogram showed that peritumoral hypointensity on HBP was the largest contributing factor(100 points)in predicting MVI of small solitary hepatocellular carcinoma,and the others were peritumor enhancement,peritumoral pseudocapsule incomplete and lower ADC values.The C-index of the nomogram was 0.844,and the model was highly predictive of MVI.Conclusions:A nomogram of Gd-EOB-DTPA-enhanced MRI can effectively predict microvascular invasion of solitary small hepatocellular carcinoma.

关 键 词：肝细胞癌 微血管侵犯 钆塞酸二钠 磁共振成像 列线图 

分 类 号：R445.2[医药卫生—影像医学与核医学] R735.7[医药卫生—诊断学]