首页|期刊导航|遵义医科大学学报|BRAFV600E突变联合超声特征对cN0期甲状腺乳头状癌患者颈淋巴结转移的预测价值

BRAFV600E突变联合超声特征对cN0期甲状腺乳头状癌患者颈淋巴结转移的预测价值OA

The predictive value of BRAFV600E mutation combined with ultrasound features for cervical lymph node metastasis in cN0 PTC patients

中文摘要英文摘要

目的 探讨术前BRAFV600E突变联合超声特征对临床淋巴结阴性(cN0)甲状腺乳头状癌(PTC)患者颈部淋巴结转移(LNM)风险的预测价值,以期术前准确识别病理淋巴结转移(PN+)并实施预防性淋巴结清扫,同时识别病理淋巴结阴性(PN0)患者,为个体化手术决策提供依据.方法 回顾性分析2023年9月至2025年9月于铜仁市人民医院术前诊断为cN0 PTC并行甲状腺癌手术患者,根据术后病理淋巴结结果,分为PN0组及PN+组,分别统计两组患者临床基本资料、术前超声特征及BRAFV600E突变结果.采用单因素Logistic回归筛选出P<0.05的变量,进一步行向前逐步法多因素Logistic回归确定术后为PN+状态的危险因素,并构建预测模型,通过ROC曲线评估其预测效能,并通过Bootstrap自抽样法对模型进行内部验证.结果 共纳入200例cN0 PTC患者,其中PN0组(131例,65.5%)、PN+组(69例,34.5%).单因素 Lo-gistic回归分析表明,两组在性别、年龄、身高、体质量、位置左叶、低回声、混合回声、肿瘤最大径、微钙化、边缘模糊和BRAFV600E突变方面比较,差异有统计学意义(P<0.05).向前逐步多因素Logistic回归分析进一步表明,年龄(OR=0.938 6,95%CI:0.896 3~0.982 9,P=0.007 1)、肿瘤最大径(OR=1.470 8,95%CI:1.000 1~2.163 2,P=0.05)、微钙化(OR=5.939 3,95%CI:2.349 9~15.011 3,P=0.000 2)、边缘模糊(OR=2.623 5,95%CI:1.061 1~6.486 2,P=0.036 8)和BRAFv600E突变(OR=2.342 5,95%CI:1.016 0~5.401 1,P=0.045 8)均是 cN0 PTC 患者发生 LNM 的预测因子.训练集ROC曲线下面积=0.851(95%CI:0.797~0.904),验证集ROC曲线下面积达到0.835(95%CI:0.715~0.948),其校准曲线平均绝对误差为0.013,显示了该模型有较好的拟合度.DCA曲线结果显示,在这个列线图预测模型中,当个体的阈值概率在0~0.88范围时,该模型提供的净收益比"每个人都干预"或"没有人干预"的策略更多.结论 BRAFV600E突变联合超声特征对cN0 PTC患者LNM预测模型,能够较好识别cN0 PTC患者中的PN+和PN0患者,为PTC的治疗方案及预后提供更多的依据,并优化手术策略并降低过度治疗风险.

Objective Exploring the predictive value of preoperative BRAFv600E mutation combined with ultra-sound features for the risk of cervical lymph node metastasis(LNM)in patients with clinical node negative(cN0)papillary thyroid carcinoma(PTC),in order to accurately identify pathological node positive(PN+)to perform preventive lymph node dissection before surgery,while identifying pathological node negative(PN0)pa-tients to reduce unnecessary procedures,providing a basis for individualized surgical decision-making.Methods A retrospective analysis was conducted on patients diagnosed with cN0 PTC undergoing concurrent thyroid cancer surgery at Tongren People's Hospital from September 2023 to September 2025.Based on postoperative pathologi-cal lymph node results,the patients were divided into PN0 and PN+groups,and their clinical basic data,pre-operative ultrasound characteristics,and BRAFV600E mutation results were statistically analyzed.Univariate logistic regression was used to screen variables with P<0.05,and then forward stepwise multivariate logistic regression was applied to identify the independent risk factors for postoperative PN+status,and a prediction model was constructed.The predictive performance was evaluated by ROC curve,and conduct internal validation of the model using Bootstrap self-sampling method.Results A total of 200 cN0 PTC patients were included,including the PN0(131 cases,65.5%)and the PN+(69 cases,34.5%).Univariate logistic regression analysis showed that there were statistically significant differences between the two groups in terms of gender,age,height,weight,location left lobe,hypoechogenicity,mixed echogenicity,maximum tumor diameter,microcalci-fications,blurred edges,and BRAFV600E mutation(P<0.05).The forward stepwise multivariate logistic regres-sion showed that age(OR=0.938 6,95%CI:0.896 3-0.982 9,P=0.007 1),maximum tumor diameter(OR=1.470 8,95%CI:1.000 1-2.163 2,P=0.05),microcalcification(OR=5.939 3,95%CI:2.349 9-15.011 3,P=0.000 2),blurred edges(OR=2.623 5,95%CI:1.061 1-6.486 2,P=0.036 8),and BRAFV600E mutation(OR=2.342 5,95%CI:1.016 0-5.401 1,P=0.045 8)were all independent predictive factors for LNM in cN0 PTC patients.The area under the ROC curve of the training set is 0.851(95%CI:0.797-0.904),and the area under the ROC curve of the validation set was 0.835(95%CI:0.715-0.948),and the average absolute error of its calibration curve was 0.013,indicating that the model has a good fit.The DCA curve results show that in this nomogram model,when the threshold probability of individuals is in the range of 0-0.88,the model provides more net benefits than the strategies of"everyone intervenes"or"no one intervenes".Conclusion The BRAFV600E mutation combined with ultrasound features can effectively identify PN+and PN0 patients in cN0 PTC patients,providing more basis for the treatment plan and prognosis of PTC,optimi-zing surgical strategies,and reducing the risk of overtreatment.

郑浪;秦刚

遵义医科大学研究生院,贵州遵义 563006铜仁市人民医院甲状腺乳腺外科,贵州铜仁 554300

医药卫生

甲状腺乳头状癌淋巴结转移超声特征BRAFV600E突变临床淋巴结阴性预测价值

papillary thyroid carcinomalymph node metastasisultrasonic featuresBRAFV600E mutationclinical lymph node negativitypredictive value

《遵义医科大学学报》 2026 (4)

414-424,11

评论