摘要
目的探讨甲状腺结节射频消融术后复发的影响因素,构建列线图预测模型,以期指导临床诊治。方法选取2019年1月至2021年1月郑州大学第二附属医院收治的398例甲状腺结节作为研究对象,其中16例于生化复发前接受内分泌治疗,12例无术后2年随访信息,均予以剔除,经筛选后259例纳入训练集用于建立模型,111例纳入验证集用于验证模型,采用COX比例风险回归方程确定甲状腺结节射频消融术后复发影响因素,采用R软件可视化处理获得列线图预测模型,行内外部验证。结果术后2年,甲状腺结节射频消融术后复发率为15.14%(56/370);COX比例风险回归方程显示,体质量指数(BMI)(OR:6.873)、热休克蛋白70(HSP70)(OR:5.380)、弹性应变率比值(SR)(OR:3.872)、结节成分(OR:5.880)、结节内部血流(OR:6.944)、结节钙化(OR:3.764)是甲状腺结节射频消融术后复发影响因素(P<0.05);基于COX比例风险回归方程结果构建甲状腺结节射频消融术后复发列线图预测模型,该模型在训练集、验证集中曲线下面积(AUC)分别为0.856、0.874,且其预测效能与实际吻合较理想。结论基于SR、BMI、HSP70、结节成分、结节内部血流、结节钙化构建列线图预测模型对甲状腺结节射频消融术后复发具有良好区分度、精准度,有助于实现个体化预测,帮助临床医师识别高风险人群,确定合理防治措施。
Objective To explore the factors affecting the recurrence of thyroid nodules after radiofrequency ablation,and to construct a nomographic prediction model to guide clinical diagnosis and treatment.Methods A total of 398 patients with thyroid nodules admitted to the Second Affiliated Hospital of Zhengzhou University from January2019 to January 2021 were selected as the research objects.Among them,16 cases receiving endocrine therapy before biochemical recurrence and 12 cases having no follow-up information after two years of surgery were excluded.After screening,259 cases were included in the training set for model construction,and 111 cases were included in the validation set for model validation.The COX proportional hazards regression equation was used to determine the factors affecting the recurrence of thyroid nodules after radiofrequency ablation.The R software visualization processing was used to obtain a nomogram prediction model for internal and external validation.Results At 2 years after surgery,the recurrence rate of thyroid nodules after radiofrequency ablation was 15.14%(56/370).COX proportional hazards regression equation showed that body mass index(BMI)(OR:6.873),heat shock protein 70(HSP70)(OR:5.380),strain ratio(SR)(OR:3.872),nodule composition(OR:5.880),nodule internal blood flow(OR:6.944),nodule calcification(OR:3.764)were factors affecting the recurrence of radiofrequency ablation of thyroid nodules(P<0.05).Based on the results of the COX proportional hazards regression equation,a nomogram prediction model for the recurrence of thyroid nodules after radiofrequency ablation was constructed.The area under the curve(AUC) of this model in the training set and validation set was 0.856 and 0.874,respectively,and its predictive performance was in good agreement with the actual data.Conclusion The nomogram prediction model constructed based on SR,BMI,HSP70,nodule composition,internal nodule blood flow,and nodule calcification has good differentiation and accuracy for recurrence of thyroid nodule after
作者
张晓光
李娜
董峰
喻红霞
ZHANG Xiao-guang;LI Na;DONG Feng;YU Hong-xia(Department of Ultrasound,the Second Affiliated Hospital of Zhengzhou University,Zhengzhou 450014,Henan,CHINA;Department of Ultrasound,the First Affiliated Hospital of Zhengzhou University,Zhengzhou 450052,Henan,CHINA;Department of General Surgery,the Second Affiliated Hospital of Zhengzhou University,Zhengzhou 450014,Henan,CHINA)
出处
《海南医学》
CAS
2024年第5期609-614,共6页
Hainan Medical Journal
