MRI形态学特征对T1-2期直肠癌淋巴结转移的预测价值  

Morphological predictors for lymph node metastasis on Magnetic Resonance Imaging in stage T1-2 rectal cancer

作　　者：李夙芸 莫梓阳 陈鑫 许伟雄[2] 梁演婷 刘再毅 LI Suyun;MO Ziyang;CHEN Xin;XU Weixiong;LIANG Yanting;LIU Zaiyi(School of Medicine,South China University of Technology,Guangzhou 510006,Guangdong,China;Department of Radiology,Guangdong Provincial People's Hospital(Guangdong Academy of Medical Sciences),Southern Medical University,Guangzhou 510080,Guangdong,China;Guangdong Provincial Key Laboratory of Artificial Intelligence in Medical Image Analysis and Application,Guangdong Provincial People's Hospital(Guangdong Academy of Medical Sciences),Guangzhou 510080,Guangdong,China;Department of Radiology,Guangzhou First People's Hospital,Guangzhou 510180,Guangdong,China;Guangdong Cardiovascular Institute,Guangzhou 510080,Guangdong,China)

机构地区：[1]华南理工大学医学院,广东广州510006 [2]南方医科大学附属广东省人民医院(广东省医学科学院)放射科,广东广州510080 [3]广东省人民医院(广东省医学科学院)广东省医学影像智能分析与应用重点实验室,广东广州510080 [4]广州市第一人民医院放射科,广东广州510180 [5]广东省心血管病研究所,广东广州510080

出　　处：《暨南大学学报（自然科学与医学版）》2023年第2期219-226,共8页Journal of Jinan University(Natural Science & Medicine Edition)

基　　金：国家杰出青年科学基金项目(81925023);国家重点研发计划项目(2021YFF1201003);广东省科技计划项目(2022B1212010011,202201020001,202201010513)。

摘　　要：目的:探讨MRI主观形态学特征预测T1-2期直肠癌治疗前直肠系膜淋巴结转移的诊断价值。方法:回顾性分析119例经病理证实为T1-2期直肠癌患者的术前临床信息及MRI资料,随机分成训练组84例和验证组35例。由两名医师独立测量和评估原发肿瘤和淋巴结MRI主观形态特征。在训练组中通过单因素及多因素Logistic回归分析(逐步后退法),筛选T1-2期直肠癌淋巴结转移的独立预测因素,构建预测模型并以受试者工作曲线下面积(area under the curve,AUC)、准确度、灵敏度和特异度评价模型的预测效能。用Delong检验比较模型间的性能。结果:119例T1-2期直肠癌患者中,28例患者淋巴结转移(23.53%)。在训练组,单因素Logistic回归分析提示肿瘤是否含黏液成分、肿瘤黏膜下强化带、淋巴结的短径、淋巴结形状、淋巴结分布、淋巴结边缘化学位移效应(CSE)、边缘轮廓、内部信号以及T2加权成像淋巴结信号均具有统计学意义(P<0.05)。Logistic多元逐步后退分析结果提示,CSE不规则(OR=5.58,95%CI:1.48~21.01,P=0.01)、CSE消失(OR=11.62,95%CI:3.63~37.22,P<0.001)、内部信号紊乱(OR=8.34,95%CI:3.10~22.47,P<0.001)是T1-2期直肠癌淋巴结转移的独立预测因素。本预测模型的AUC、准确度、敏感度、特异度分别为0.92、0.89、0.85及0.91,本模型效能显著优于欧洲胃肠和腹部放射学会指南(ESGAR)预测水平(P=0.02)。在验证组,模型的AUC、准确度、敏感度、特异度分别为0.83、0.89、0.75及0.93。结论:与ESGAR指南预测水平相比,联合淋巴结边缘CSE和淋巴结内部信号的T1-2直肠癌淋巴结转移预测模型评估治疗前淋巴结是否转移更为简便并有更好的预测性能。Objective:To investigate the diagnostic value of MRI features in predicting lymph node metastasis in stage T1-2 rectal cancer.Methods:Preoperative MRI images and clinical information of 119 patients with pathologically confirmed stage T1-2 rectal cancer were retrospectively analyzed.MRI features of the primary tumor and lymph node were measured and assessed independently by two physicians.Patients were randomly divided into a training group of 84 and a validation group of 35.Univariate and multiple logistic regression analyses were used to identify significant LNM predictive variables.Then a model was developed using the independent predictive factors and the predictive efficacy of the models was evaluated in terms of area under the curve(AUC),accuracy,sensitivity,and specificity.The Delong test was used to compare the performance between the models.Results:LNM were found in 28/199 patients(23.53%).In the training group,univariate logistic regression analysis suggested that some characteristics of the tumors(presence of mucus component and submucosal enhancing stripe status)as well as lymph nodes[their largest short diameter,shape,cluster,chemical shift effect(CSE)status,border,internal signal,and T2-weighted imaging signal]were statistically significant(P<0.05).Three independent risk factors were determined in the multiple logistic regression analysis,including irregular CSE(OR=5.58,95%CI 1.48-21.01,P=0.01),absent CSE(OR=11.62,95%CI 3.63-37.22,P<0.001),and heterogeneity internal signal(OR=8.34,95%CI 3.10-22.47,P<0.001).The AUC,accuracy,sensitivity,and specificity of the prediction model were 0.92,0.89,0.85,and 0.91 respectively.The model performance was significantly better than the European Society of Gastrointestinal and Abdominal Radiology(ESGAR)guideline's model(P=0.02).In the validation group,the AUC,accuracy,sensitivity and specificity of the model were 0.83,0.89,0.75 and 0.93,respectively.Conclusion:Compared with ESGAR model,the lymph node metastasis prediction model for stage T1-2 rectal cancer combining

关 键 词：直肠癌 淋巴结转移 磁共振成像 

分 类 号：R445.2[医药卫生—影像医学与核医学] R318[医药卫生—诊断学]