基于超声造影的预测模型对甲状腺乳头状癌颈中央区淋巴结转移的诊断价值  

Diagnostic value of a predictive model based on contrast-enhanced ultrasound for prediction of central cervical lymph node metastasis in patients with thyroid papillary carcinoma

作　　者：陈启阳 刘玉江[1] 刘金苹[1] 谭小蕖[1] 钱林学[1] 胡向东[1] Chen Qiyang;Liu Yujiang;Liu Jinping;Tan Xiaoqu;Qian Linxue;Hu Xiangdong(Department of Ultrasound Medicine,Beijing Friendship Hospital,Capital Medical University,Beijing 100050,China)

机构地区：[1]首都医科大学附属北京友谊医院超声科,100050

出　　处：《中华医学超声杂志（电子版）》2023年第4期442-448,共7页Chinese Journal of Medical Ultrasound(Electronic Edition)

基　　金：中华人民共和国工业和信息化部项目(先进医疗设备应用技术评价:甲状腺成像体系超声评价规范)(2018MND102015)。

摘　　要：目的建立基于超声造影(CEUS)的预测模型以评估甲状腺乳头状癌(PTC)患者颈部中央区淋巴结转移(CLNM)风险,并讨论该模型的诊断效能。方法回顾性分析2017年1月至2019年12月于首都医科大学附属北京友谊医院确诊为单发PTC的患者108例。根据有无CLNM将其分为CLNM组(48例)与无CLNM组(60例),比较2组间一般临床资料、常规超声以及CEUS表现之间的差异。根据多因素Logistic回归分析建立预测模型,绘制ROC曲线评价模型及模型内各变量因素的诊断效能。结果单因素分析结果显示,患者年龄、病灶最大径、是否接触被膜、增强程度以及增强均匀性在CLNM组与无CLNM组之间差异具有统计学意义(P均<0.05)。多因素Logistic回归分析显示,年龄≤42岁(OR=4.783,P=0.002)、接触被膜(OR=4.833,P=0.006)、等或高增强(OR=3.696,P=0.013)以及不均匀增强(OR=5.264,P=0.004)是CLNM的独立危险因素。建立预测模型为:Y=-4.017+1.565×年龄≤42+1.576×病灶接触被膜+1.307×等或高增强+1.661×不均匀增强。该模型ROC曲线下面积(AUC)为0.800,敏感度及特异度分别为70.8%、78.3%;模型AUC大于模型内单一变量参数的AUC(P均<0.05)。结论基于CEUS并联合常规超声及临床参数建立的预测模型,诊断效能较好,可用于评估PTC患者CLNM的风险。Objective To establish a predictive model based on contrast-enhanced ultrasound(CEUS)to evaluate the risk of central cervical lymph node metastasis(CLNM)in patients with thyroid papillary carcinoma(PTC),and to evaluate the diagnostic efficacy of the model.Methods A retrospective analysis was performed on 108 patients with single PTC who were examined by conventional ultrasonography and contrast-enhanced ultrasonography before operation at the Beijing Friendship Hospital of Capital Medical University from January 2017 to December 2019.The patients were divided into either a CLNM group or a non-CLNM group according to whether CLNM occurred,and the differences in clinical,conventional ultrasound,and CEUS characteristics between the two groups were compared.A predictive model was established according to multivariate regression analysis,and the diagnostic efficacy of the model was evaluated by receiver operating characteristic(ROC)curve analysis.Results Univariate analysis indicated significant differences between the CLNM and non-CLNM groups in age,maximum tumor diameter,contact with capsule,degree of enhancement,and uniformity of enhancement(P<0.05).Multivariate logistic regression analysis showed that age≤42(odds ratio[OR]=4.783,P=0.002),contact with capsule(OR=4.833,P=0.006),Hyper-or isoechoic enhancement(OR=3.696,P=0.013),and heterogeneous enhancement(OR=5.264,P=0.004)were independent risk factors for CLNM.ROC curve analysis showed that the area under the curve(AUC),sensitivity,and specificity of the predictive model were 0.800,70.8%,and 78.3%,respectively.The AUC of the predictive model was greater than that of any single independent risk predictor(P<0.05).Conclusion The prediction model based on CEUS combined with conventional ultrasound and clinical factors has good diagnostic efficiency and can be used for the risk assessment of CLNM in PTC patients.

关 键 词：甲状腺乳头状癌 中央区淋巴结转移 超声造影 LOGISTIC模型 

分 类 号：R736.1[医药卫生—肿瘤] R445.1[医药卫生—临床医学]