甲状腺乳头状癌侧颈淋巴结转移的危险因素分析及预测模型构建  

Risk factor analysisand prediction model constructionof lateral lymph node metastasis in papillary thyroid carcinoma

作　　者：李佳[1] 汤铜[1] 宋杨[1] 张磊[1] 史加宁[1] LI Jia;TANG Tong;SONG Yang;ZHANG Lei;SHI Jianing(Department of Thyroid and Breast Surgery,the Second Hospital of Anhui Medical University,Hefei 230032,China)

机构地区：[1]安徽医科大学第二附属医院甲乳外科,安徽合肥230032

出　　处：《安徽医学》2022年第7期743-747,共5页Anhui Medical Journal

基　　金：安徽省自然科学基金项目(项目编号:2008085QH423);2021年度安徽高校人文社会科学研究重点项目(项目编号:SK2021A0167)。

摘　　要：目的探讨甲状腺乳头状癌侧颈区淋巴结转移的危险因素并构建临床预测模型。方法回顾性分析2019年1月至2019年12月就诊于安徽医科大学第二附属医院的230例甲状腺乳头状癌患者的临床资料,根据是否存在侧颈区淋巴结转移将患者分为转移组(n=39例)和非转移组(n=191例)。比较两组患者的性别、年龄、单/双侧、肿瘤最大径、是否多灶、有无合并桥本甲状腺炎、是否紧贴包膜、有无钙化等指标的差异,对差异有统计学意义的指标应用多因素logistic回归(逐步回归)分析侧颈区淋巴结转移的独立危险因素。通过R软件进行临床建模并绘制列线图,并计算C(校正)指数、绘制校准曲线、受试者特征(ROC)曲线进一步评估预测模型的效能。结果两组患者性别、单/双侧、肿瘤最大径、是否多灶、是否紧贴包膜、有无钙化比较,差异有统计学意义(P<0.05);多因素logistic逐步回归分析结果显示,性别、单/双侧、肿瘤最大径为侧颈区淋巴结转移的独立危险因素。根据上述3个因素构建诺莫预测模型,模型预测甲状腺乳头状癌患者发生侧颈转移的ROC曲线下面积为0.82(95%CI:0.76~0.89);构建模型的C-index为0.82(95%CI:0.76~0.88),校正C-index为0.80;模型的校准曲线与实际曲线重合较好。结论性别、单/双侧癌、肿瘤最大径是甲状腺乳头状癌侧颈区淋巴结转移的独立危险因素;基于上述因素进行的临床建模对甲状腺乳头状癌患者发生侧颈区淋巴结转移的风险预测具有一定的价值。Objective To investigate the risk factors of lateral lymph node metastasis in papillary thyroid carcinoma and to construct a clinical prediction model.Methods A retrospective study was performed on 230 cases of papillary thyroid carcinoma from January 2019 to December 2019 in the Second Hospital of Anhui Medical University.According to the presence or absence of lateral lymph node metastasis,the patients were divided into two groups:the metastasis group(39 cases)and the non-metastatic group(191 cases).The differences in gender,age,unilateral/bilateral,maximum tumor diameter,multifocal,combined Hashimoto's thyroiditis,close to the capsule,and calcification were compared between the two groups.The independent risk factors of laterallymph nodemetastasis were analyzed by binary logistic regression(stepwise regression).The clinical modeling and nomogram were performed by R software,and the C(corrected)index was calculated,and the calibration curve and the subject characteristic curve were drawn to further evaluate the performance of the prediction model.Results The comparative analysis of general data between the two groups showed that gender,unilateral/bilateral,maximum tumor diameter,multifocal,close to the capsule,calcification were different withstatistical significance(P<0.05);multivariate Logistic Stepwise regression analysis showed that gender,unilateral/bilateral cancer,and maximum tumor diameter were independent risk factors for lateral lymph node metastasis.According to the above three factors,a Nomo prediction model was constructed.The area under the ROC curve of the model predicting lateral neck metastasis in patients with papillary thyroid carcinoma was 0.82(95%CI:0.76~0.89);the C-index of the model was 0.82(95%CI:0.76~0.88),the corrected C-index was 0.80;the calibration curve of the model coincided well with the actual curve.Conclusions Gender,unilateral/bilateral cancer,and maximum tumor diameterare independent risk factors for lateral lymph node metastasis in papillary thyroid carcinoma;the clinical mo

关 键 词：甲状腺乳头状癌 侧颈区淋巴结转移 列线图 预测模型 

分 类 号：R736.1[医药卫生—肿瘤]