基于超声特征构建甲状腺微小乳头状癌中央区淋巴结转移的预测模型研究  

Prediction model of central lymph node metastasis in papillary thyroid microcarcinoma based on ultrasonic features

作　　者：吴晓恬 朱鹏飞 葛舒 姜倩[1] WU Xiao-tian;ZHU Peng-fei;GE Shu;JIANG Qian(Department of Ultrasound,Nantong Cancer Hospital,Nantong 226001,Jiangsu,China)

机构地区：[1]南通市肿瘤医院超声科,江苏南通226001

出　　处：《川北医学院学报》2022年第12期1626-1630,共5页Journal of North Sichuan Medical College

摘　　要：目的:探讨甲状腺微小乳头状癌(PTMC)中央区淋巴结转移(CLNM)的超声特征,建立CLNM概率预测的列线图模型并验证模型的诊断效能。方法:回顾性分析120例PTCM患者的临床资料。根据是否发生CLNM情况将患者分为转移组(n=32)和非转移组(n=88)。收集患者超声特征进行多因素Logistic回归分析,确立PTMC发生CLNM的独立风险因素,并根据分析结果构建PTCM患者发生CLNM的列线图模型,应用受试者工作特征(ROC)曲线评价模型的效能。结果:120例患者中,32例发生CLNM,发生率为26.67%。两组结节直径、病灶数目、边缘、内部回声、微钙化、纵横比、血流信号、肿瘤与被膜接触对比,差异有统计学意义(P<0.05)。多因素分析结果显示:结节直径5~10 mm(OR=3.714)、有微钙化(OR=5.414)、纵横比≥1(OR=5.966)、血流信号丰富(OR=5.392)、肿瘤与被膜接触(OR=3.885)是PTMC发生CLNM的独立风险因素(P<0.05)。基于超声特征PTMC发生CLNM的列线图预测模型ROC曲线下AUC为0.850(95%CI为0.774~0.909),敏感度为75.00%,特异度为85.23%。结论:结节直径5~10 mm、有微钙化、纵横比≥1、血流信号丰富、肿瘤与被膜接触是PTMC发生CLNM的独立风险因素。基于上述风险构建的列线图具有较好的预测性能。Objective:To investigate the ultrasonographic characteristics of central lymph node metastasis(CLNM)in papillary thyroid microcarcinoma(PTMC),and to establish a line graph model for predicting CLNM probability and verify the diagnostic efficiency of the model.Methods:The clinical data of 120 patients with PTCM were retrospectively analyzed.Patients were divided into transfer group(n=32)and non-transfer group(n=88)according to whether CLNM occurred.The ultrasonic characteristics of patients were collected,and the independent risk factors of CLNM in PTMC patients were analyzed by multivariate Logistic regression analysis.The bar chart model of CLNM in PTCM patients based on ultrasonic characteristics was constructed according to the results of Logistic regression analysis.The receiver operating characteristic(ROC)curve was used to evaluate the predictive efficiency of the model.Results:Among 120 patients,CLNM occurred in 32 cases,with the incidence of 26.67%.There were statistically significant differences in nodule diameter,lesion number,edge,internal echo,microcalcification,aspect ratio,blood flow signal,tumor and capsule contact between the two groups(P<0.05).Multivariate Logistic regression analysis showed that nodule diameter 5~10 mm(OR=3.714),microcalcification(OR=5.414),aspect ratio≥1(OR=5.966),rich blood flow signal(OR=5.392),and tumor contact with the capsule(OR=3.885)were independent risk factors for PTMC CLNM(P<0.05).The AUC under ROC curve of the prediction model for CLNM of PTMC based on ultrasonic characteristics was 0.850(95%CI was 0.774~0.909),the sensitivity was 75.00%,and the specificity was 85.23%.Conclusion:The independent risk factors for PTMC CLNM are nodule diameter of 5~10 mm,microcalcification,aspect ratio1,abundant blood flow signals,and tumor-membrane contact.The line chart based on the above risks has good prediction performance.

关 键 词：超声特征 甲状腺微小乳头状癌 淋巴结转移 预测模型 

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