甲状腺乳头状癌原发灶超声特征对颈部非典型声像图转移性淋巴结的预测价值

Predictive value of the ultrasound characteristics of the primary lesion of papillary thyroid carcinoma for the cervical non-typical sonographic features of metastatic lymph nodes

目的通过分析甲状腺乳头状癌(PTC)原发灶的超声特征构建颈部非典型声像图转移性淋巴结的预测模型。方法回顾性选取2022年6月至2023年6月于苏州大学附属第一医院诊治的PTC患者411例,排除颈部典型声像图转移性淋巴结或资料不完整者,最终201例患者纳入本研究。按颈部淋巴结是否转移分为两组,分析两组患者的临床及术前超声资料,确定颈部非典型声像图转移性淋巴结的独立危险因素,并建立列线图预测模型,使用ROC曲线及校准曲线评价模型的区分度及校准度。结果201例患者中转移组52例、无转移组149例。单因素分析显示两组患者在原发灶最大径、与被膜是否接触、内部回声、形态、边缘、钙化类型方面差异有统计学意义(均P<0.05);多因素分析显示结节的最大径(OR=1.181,95%CI=1.084~1.288)、与被膜接触(OR=2.962,95%CI=1.344~6.530)、边缘模糊(OR=3.711,95%CI=1.314~10.477)、内部存在微钙化(OR=3.236,95%CI=1.427~7.342)是发生颈部非典型声像图转移性淋巴结的独立危险因素。基于独立危险因素绘制列线图模型,模型的曲线下面积(AUC)、敏感性、特异性、准确性分别为0.853、75.00%、81.90%、81.59%。结论基于原发灶最大径、与被膜是否接触、边缘及微钙化特征构建的列线图模型对PTC颈部非典型声像图淋巴结转移有较好的预测效能,可降低非典型声像图转移性淋巴结的漏诊率,从而为临床决策提供更多的影像学信息。

Objective To construct a predictive model for cervical non-typical sonographic metastatic lymph nodes based on the ultrasonic characteristics of the primary lesion of papillary thyroid carcinoma(PTC).Methods A total of 411 PTC patients treated at the First Affiliated Hospital of Soochow University from June 2022 to June 2023 were retrospectively collected.Patients with typical sonographic metastatic lymph nodes or incomplete data were excluded,resulting in the inclusion of 201 patients in this study.The patients were divided into two groups based on whether there were cervical metastatic cervical lymph nodes.The clinical and preoperative ultrasound data of the two groups were analyzed to determine the independent risk factors for cervical non-typical sonographic metastatic lymph nodes.A nomogram model was established,the ROC curve was used to determine the discriminatory ability,and the calibration curve was applied to evaluate the calibration ability.Results A total of 201 patients were included in this study,including 52 in the metastasis group and 149 in the non-metastasis group.Univariate analysis revealed significant differences between the two groups in terms of the maximum diameter of the primary tumor,contact with the capsule,internal echo,morphology,margin,and calcification type(all P<0.05).Multivariate Logistic regression analysis showed that the maximum diameter of the nodule(OR=1.181,95%CI=1.084-1.288),contact with the capsule(OR=2.962,95%CI=1.344~6.530),blurred edges(OR=3.711,95%CI=1.314-10.477),and internal microcalcifications(OR=3.236,95%CI=1.427-7.342)were independent risk factors for non-typical sonographic metastatic lymph nodes.Based on these independent risk factors,a nomogram was constructed,and the model exhibited an area under the curve(AUC)of 0.853,sensitivity of 75.00%,specificity of 81.90%,and accuracy of 81.59%.Conclusions The nomogram model based on the features of maximum diameter of the primary lesion,contact with the capsule,margin,and microcalcification has good predictive efficacy for non-typical sonographic metastatic lymph nodes in PTC.It helps to reduce the misdiagnosis rate of non-typical sonographic metastatic lymph nodes and provide more imaging information for clinical decision-making.