摘要
目的:探索术前超声图像特征结合临床指标对甲状腺乳头状癌(PTC)术后颈部淋巴结转移风险的预测价值。方法:回顾性收集2019年1月至2020年12月在温州医科大学附属第一医院接受二次手术并经病理证实的PTC颈部淋巴结转移患者43例,作为转移组。同时,收集我院2013年1月至2014年12月收治的89例自首次PTC手术起随访至少8年,未出现颈部淋巴结复发的患者作为未转移组。按照7:3的比例随机抽样将患者分为训练集(n=92)和验证集(n=40),在训练集中通过单因素和多因素Logistic回归分析,构建预测模型并绘制列线图使模型可视化,使用验证集对模型进行验证,通过ROC曲线和决策曲线分析(DCA)对模型效能和临床价值进行评估。结果:转移组与未转移组患者间手术方法、术中淋巴结转移、甲状腺结节部位、回声特性、甲状腺结节最大径线、超声提示淋巴结肿大、淋巴结长径/短径比值以及淋巴结血流特征差异均有统计学意义(均P<0.05)。多因素Logistic回归分析显示手术方式(全切vs.部分切除,OR=0.332,95%CI=0.134~0.828,P=0.018)、术中淋巴结转移(转移vs.未转移,OR=2.978,95%CI=1.112~7.953,P=0.030)、甲状腺结节部位(中部vs.下极,OR=6.624,95%CI=1.986~22.095,P=0.002)及甲状腺结节最大径线(OR=1.064,95%CI=1.002~1.130,P=0.042)为术后颈部淋巴结转移的独立影响因素。该模型在训练集中的AUC为0.872(95%CI=0.795~0.949),在验证集中的AUC为0.863(95%CI=0.749~0.976),DCA显示阈值概率在0.1~0.7时该模型预测术后PTC颈部淋巴结转移发生的风险净收益高。结论:本研究基于术前超声图像特征及临床病理指标构建的预测模型可有效预测PTC术后颈部淋巴结转移的风险,对于临床决策有一定的参考价值。
Objective:To explore the predictive value of preoperative ultrasonographic features combined with clinical indices for the risk of cervical lymph node metastasis after surgery for papillary thyroid carcinoma(PTC).Methods:A retrospective collection of 43 patients with cervical lymph node metastasis of PTC confirmed by pathology after a second surgery at the First Affiliated Hospital of Wenzhou Medical University from January 2019 to December 2020 was included in as the metastasis group.Meanwhile,89 patients who had undergone initial PTC surgery from January 2013 to December 2014 and were followed up for at least 8 years without recurrence of cervical lymph nodes were collected as the non-metastasis group.Patients were randomly sampled in a 7:3 ratio into a training set(n=92)and a validation set(n=40).In the training set,a predictive model was constructed through univariate and multivariate logistic regression analyses and visualized with nomograms.The model was validated using the validation set,and its efficacy and clinical value were assessed by ROC curve and decision curve analysis(DCA).Results:Statistically significant differences were observed between the metastasis and nonmetastasis groups in terms of surgical methods,intraoperative lymph node metastasis,thyroid nodule location,echogenicity,maximum diameter of thyroid nodules,ultrasonographic indication of lymph node enlargement,lymph node long-to-short axis ratio,and lymph node blood flow characteristics(all P<0.05).Multivariate logistic regression analysis identified the method of surgery(OR=0.332,95%CI=0.134-0.828,P=0.018),intraoperative lymph node metastasis(OR=2.978,95%CI=1.112-7.953,P=0.03),thyroid nodule location(middle vs.lower pole,OR=6.624,95%CI=1.986-22.095,P=0.002)and maximum diameter of thyroid nodules(OR=1.064,95%CI=1.002-1.130,P=0.042)as independent predictors of cervical lymph node metastasis after surgery.The model exhibited an area under the curve(AUC)of 0.872(95%CI=0.795-0.949)in the training set and an AUC of 0.863(95%CI=0.749-0.976)in the validation set,with DCA showing a net benefit in predicting the risk of postoperative PTC cervical lymph node metastasis at a threshold probability of 0.1 to 0.7.Conclusion:Based on preoperative ultrasonographic characteristics and clinical pathological indices,the constructed model effectively forecasts the risk of cervical lymph node metastasis after PTC surgery,offering valuable guidance for clinical decisionmaking.
作者
何磊
熊炳
罗银丽
陈顺平
HE Lei;XIONG Bing;LUO Yinli;CHEN Shunping(Department of Ultrasound,the First Affiliated Hospital of Wenzhou Medical University,Wenzhou 325015,China)
出处
《温州医科大学学报》
CAS
2024年第6期458-464,共7页
Journal of Wenzhou Medical University
关键词
甲状腺乳头状癌
术后颈部淋巴结转移
预测模型
列线图
papillary thyroid carcinoma
postoperative cervical
lymph node metastasis
predictive model
nomogram
