摘要
目的:评估HT合并PTC患者中央淋巴结转移(central lymph node metastasis,CLNM)的预测因素。建立列线图预测PTC伴HT患者发生CLNM的可能性。方法:回顾性收集了2018年1月至2021年12月在我院接受甲状腺手术的4171例PTC患者的资料。最后,共纳入671例PTC合并HT患者。其中,468例患者组成训练组,其余203例患者组成了验证组,以验证模型的性能。预测因子选择采用LASSO回归模型,并采用多因素logistic回归分析建立预测模型,建立了预测CLNM的列线图,并进行了内部验证。结果:LASSO回归模型显示,有17个变量可能是影响CLNM发生的因素(P<0.05)。随后,多因素逻辑回归分析显示,年龄较低、结节性高回声、肿瘤大直径、肿瘤多灶性、甲状腺外扩张(extrathyroidal extension,ETE)、颈部淋巴结肿大、癌胚抗原(carcino-embryonic antigen,CEA)是发生中央区淋巴结转移(CLNM)的独立预测因素。基于独立预测因素构建PTC合并HT患者发生CLNM列线图,并进行内部验证。通过建立预测模型,训练组发生CLNM的ROC曲线下面积(AUC)为0.774(95%CI,0.725~0.824),验证组发生CLNM的ROC曲线下面积(AUC)为0.781(95%CI,0.712~0.850)。列线图对训练队列和验证队列以及合并数据集均显示出良好的校准和鉴别能力。结论:本研究构建的列线图预测模型对甲状腺乳头状癌伴桥本甲状腺炎患者发生中央区淋巴结转移(CLNM)有良好的预测作用。为临床治疗方案提供合理的参考,帮助临床医生为患者制定个性化的治疗方案。
Objective:To evaluate the predictive factors of central lymph node metastasis(CLNM)in PTC with HT patients.And establish a nomogram to predict the possibility of CLNM in PTC with HT patients.Methods:We retrospectively collected the data of 4171 PTC patients who underwent surgery at our hospital from January 2018 to December 2021.Finally,a total of 671 PTC patients with HT were enrolled.Among them,468 patients comprised the training cohort for nomogram development,while the other 203 formed the validation cohort to validate the model's performance.The least absolute shrinkage and selection operator(LASSO)regression models were used for factors selection,and the multivariate logistic regression analysis was used to establish the prediction model.A nomogram for predicting CLNM was established and internally validated.Results:LASSO regression model showed that seventeen variables may be the factors affecting the occurrence of CLNM(P<0.05).Subsequently,multivariate logistic regression analysis showed that lower age,hyperechoic in nodule(HIN),larger diameter,multifocality,extrathyroidal extension(ETE),cervical lymph nodes enlargement(ECLN)and carcino-embryonic antigen(CEA)were finalized risk factors for CLNM(P<0.05).A nomogram of CLNM in PTC with HT patients was constructed based on influencing factors and underwent internal validation.By establishing the prediction model,the area under the ROC curve(AUC)of CLNM in the training group was 0.774(95%CI,0.725~0.824),and the AUC in the validation group was 0.781(95%CI,0.712~0.850).The preoperative nomogram showed good calibration and discrimination for the training and validation cohorts,as well as for the combined data set.Conclusion:The nomogram prediction model constructed in this study has a good prediction effect on the occurrence of central regional lymph node metastasis(CLNM)in patients with papillary thyroid cancer associated with Hashimoto's thyroiditis.To provide a reasonable reference for clinical treatment plans,and to help clinicians to develop personalized tre
作者
李惠
黄玉婷
娄鹏威
王凯
马丽
LI Hui;HUANG Yuting;LOU Pengwei;WANG Kai;MA Li(Department of Endocrine,Traditional Chinese Medicine Hospital Affiliated to Xinjiang Medical University,Xinjiang Urumqi 830000,China;Department of Medical Administration,Traditional Chinese Medicine Hospital Affiliated to Xinjiang Medical University,Xinjiang Urumqi 830000,China;Department of Big Data,College of Information Engineering,Xinjiang Institute of Engineering,Xinjiang Urumqi 830000,China;Department of Medical Engineering and Technology,Xinjiang Medical University,Xinjiang Urumqi 830011,China)
出处
《现代肿瘤医学》
CAS
北大核心
2023年第12期2239-2247,共9页
Journal of Modern Oncology
