分化型甲状腺癌患者喉前淋巴结转移影响因素分析及预测模型的构建与验证  

Influencing factors of prelaryngeal lymphnodes metastasis in patients with differentiated thyroid cancer and the construction and validation of predictive model

作　　者：王宁[1] 刘金彪[1] 乔楠 Wang Ning;Liu Jinbiao;Qiao Nan(The First Affiliated Hospital of Henan University of Science and Technology,Luoyang 471003,Henan,China)

机构地区：[1]河南科技大学第一附属医院,河南洛阳471003

出　　处：《临床心身疾病杂志》2023年第4期15-20,共6页Journal of Clinical Psychosomatic Diseases

摘　　要：目的 基于临床病理特征和肿瘤标志物,探讨分化型甲状腺癌患者喉前淋巴结转移影响因素,并构建与验证其列线图预测模型。方法 将96例甲状腺癌患者根据有无喉前淋巴结转移分为转移组与未转移组,收集两组临床病理特征、肿瘤标志物(甲状腺球蛋白、甲状腺球蛋白抗体、癌胚抗原、血管内皮生长因子)水平,采用Logistic回归分析探讨分化型甲状腺癌喉前淋巴结转移影响因素,构建列线图预测模型,并进行验证。结果 两组肿瘤位置、病灶个数、肿瘤最大径、包膜侵犯、中央区非喉前淋巴结转移情况比较差异有统计学意义(P<0.05或0.01)。两组血清甲状腺球蛋白水平比较差异无统计学意义(P>0.05);转移组血清甲状腺球蛋白抗体、癌胚抗原、血管内皮生长因子水平高于未转移组(P<0.01)。Logistic回归分析显示,腺体上1/3或峡部肿瘤、病灶多发、肿瘤最大直径、包膜侵犯、中央区非喉前淋巴结转移、血清甲状腺球蛋白抗体、癌胚抗原、血管内皮生长因子是分化型甲状腺癌患者喉前淋巴结转移的独立危险因素(P<0.01)。根据上述因素构建分化型甲状腺癌患者喉前淋巴结转移的列线图预测模型,校准曲线显示,该预测模型一致性指数为0.730,Hosmer-Lemeshow拟合优度检验(P>0.05);受试者工作特征曲线显示,该预测模型预测喉前淋巴结转移的曲线下面积为0.954(95%CI:0.901~1.000),敏感度为90.48%,特异度为93.33%。结论 腺体上1/3或峡部肿瘤、病灶多发、肿瘤最大直径、包膜侵犯、中央区非喉前淋巴结转移、血清甲状腺球蛋白抗体、癌胚抗原、血管内皮生长因子是分化型甲状腺癌患者喉前淋巴结转移的独立危险因素,根据上述因素构建的列线图预测模型具有可靠的预测价值。Objective To explore the influencing factors of prelaryngeal lymphnodes(PLN)metastasis in patients with differentiated thyroid cancer(DTC)based on clinicopathological features and tumor markers and to construct and validate nomogram prediction model.Methods Ninety-six DTC patients were divided into metastasis and non-metastasis group according to the presence or absence of PLN metastasis,clinicopathological characteristics and tumor markers[thyroglobulin(Tg),thyroglobulin antibody(TgAb),carcinoembryonic antigen(CEA),vascular endothelial growth factor(VEGF)]levels collected,influencing factors of PLN metastasis explored using Logistic regression analysis,and nomogram prediction model was constructed and validated.Results The intergroup differences in tumor location,number of lesions,maximum tumor diameter,envelope invasion,and non-PLN metastases in central area were statistically significant(P<0.05);intergroup difference in serum Tg levels was not statistically significant(P>0.05);serum TgAb,CEA,and VEGF levels were higher in metastasis than non-metastasis group(P<0.01).Logistic regression analysis showed that independent risk factors of PLN metastasis included tumor in the upper 1/3 of the gland or isthmus,multiple lesions,maximum tumor diameter,envelope invasion,non-DLN metastasis in the central region,serum TgAb,CEA,and VEGF(P<0.05).The calibration curve of nomogram prediction model constructed based on the above-mentioned factors showed that the consistency index of the prediction model was 0.730,Hosmer-Lemeshow goodness-of-fit test was(P>0.05);the Receiver Operating Characteristic(ROC)curve showed that area under the curve were 0.954(95%CI:0.901~1.000),sensitivity 90.48%,and specificity 93.33%.Conclusion Independent risk factors of PLN metastasis include tumor in the upper 1/3 of the gland or isthmus,multiple lesions,maximum tumor diameter,envelope invasion,non-DLN metastasis in the central region,serum TgAb,CEA,and VEGF,predictive value of constructed nomogram prediction model based on the above-mentioned fa

关 键 词：分化型甲状腺癌 喉前淋巴结转移 临床病理特征 肿瘤标志物 影响因素 列线图预测模型 

分 类 号：R581.04[医药卫生—内分泌]