pT1~2N1M0期乳腺癌改良根治术后患者预后预测模型的构建  

Prognosis Prediction Model Construction for Breast Cancer Patients with Stage pT1~2N1M0 After Modified Radical Surgery

作　　者：高汉晶[1] 杨丽[1] 闫争 陈艳芳[2] GAO Hanjing;YANG Li;YAN Zheng(Radiotherapy Department of Tianjin Fourth Central Hospital,Tianjin 300074,P.R.China)

机构地区：[1]天津市第四中心医院放疗科,300074 [2]天津医科大学第二医院放疗科,300074

出　　处：《临床放射学杂志》2025年第3期429-434,共6页Journal of Clinical Radiology

摘　　要：目的探讨pT1~2N1M0期乳腺癌改良根治术后局部区域复发的危险因素并据此构建相关临床预测模型,旨在为后续的预后预测和个体化治疗特别是放射治疗方案的制定提供更多参考。方法回顾性纳入2012年1月至2022年1月于天津市第四中心医院及天津医科大学第二医院行改良根治术治疗的pT1~2N1M0期乳腺癌患者542例,根据术后随访是否发生局部区域复发进行分组。然后进行pT1~2N1M0期乳腺癌改良根治术后局部区域复发的单因素分析以及多因素分析。最后进行pT1~2N1M0期乳腺癌改良根治术后局部区域复发的风险预测模型构建及预测效能分析。结果542例患者的术后中位随访时间为60.0(1.0~80.0)个月,随访过程中发生局部区域复发40例,中位局部区域复发时间为35.0(4.0~80.0)个月,局部区域复发率为7.38%。单因素分析结果显示,年龄、病灶位置、T分期、腋窝阳性淋巴结数量、激素受体状态及辅助放射治疗靶区均可能与pT1~2N1M0期乳腺癌改良根治术后局部区域复发有关(P<0.05)。Logistic多因素分析结果显示,年龄、病灶位置、T分期、腋窝阳性淋巴结数量、激素受体状态及辅助放射治疗靶区均是pT1~2N1M0期乳腺癌改良根治术后局部区域复发风险的独立影响因素(P<0.05)。利用Logistic回归模型的独立影响因素构建列线图模型,该列线图模型的C指数为0.842。采用Bootstrap重抽样技术进行模型的内部验证,历经1000次迭代,绘制校准曲线,其平均绝对误差维持在0.017的水平。进一步通过受试者工作特征(ROC)曲线分析,利用列线图预测模型,对患者术后局部区域复发情况进行预测,得出的ROC曲线下面积为0.929。选择2023年1月至2024年6月在同样医院就诊的200例T1~2N1M0期乳腺癌改良根治术后患者的数据实施列线图模型的外部验证。结果显示,校准曲线平均绝对误差维持在0.021的水平。进一步通过ROC曲线分析,利用列Objective To explore the risk factors of local regional recurrence after modified radical resection of stage pT1-2N1M0 breast cancer,and to build relevant clinical prediction models to provide more references for subsequent prognosis prediction,individualized treatment,especially the formulation of radiotherapy protocols.Methods A total of 542 patients with stage pT1-2N1M0 breast cancer who received modified radical surgery at the Fourth Central Hospital of Tianjin and the Second Hospital of Tianjin Medical University from January 2012 to January 2022 were retrospectively included and grouped according to whether local regional recurrence occurred during postoperative follow-up.Univariate and multivariate analyses of local regional recurrence after modified radical resection of stage pT1-2N1M0 breast cancer were performed.A local regional recurrence risk prediction model after modified radical resection of stage pT1-2N1M0 breast cancer was constructed and its prediction efficacy was analyzed.Results This study included 542 patients who were followed up for a median of 60.0(1.0-80.0)months after surgery.During the follow-up process,40 cases of local regional recurrence occurred,with a median local regional recurrence time of 35.0(4.0-80.0)months and a local regional recurrence rate of 7.38%.Univariate analysis results showed that age,tumor location,T stage,number of axillary positive lymph nodes,hormone receptor status,and adjuvant radiotherapy target area may be related to local regional recurrence after modified radical resection of stage pT1-2N1M0 breast cancer(P<0.05).Logistic multivariate analysis results showed that age,tumor location,T stage,number of axillary positive lymph nodes,hormone receptor status,and adjuvant radiotherapy target area were all independent influencing factors for local regional recurrence risk after modified radical resection of stage pT1-2N1M0 breast cancer(P<0.05).A nomogram model was constructed using the independent influencing factors of the Logistic regression model,with a C-ind

关 键 词：乳腺癌 改良根治术 放疗 局部区域复发 

分 类 号：R737.9[医药卫生—肿瘤]