多模态MRI联合临床病理预测乳腺癌新辅助治疗后腋窝淋巴结病理完全缓解OA
Multimodal MRI combined with clinicopathological indicators to predict pathological complete response of axillary lymph nodes after neoadjuvant therapy for breast cancer
目的 结合治疗前后乳腺、腋窝淋巴结及瘤周区域的多模态MRI影像与临床病理指标,建立初始腋窝淋巴结阳性乳腺癌新辅助治疗(neoadjuvant therapy,NAT)后腋窝病理完全缓解(pathological complete response,pCR)的预测模型,并评估其效能.材料与方法 回顾性分析2022年1月至2025年8月我院收治的乳腺癌伴腋窝淋巴结转移患者,收集NAT前后MRI及临床病理资料;按8:2随机分为训练集与验证集.采用logistic回归筛选腋窝pCR独立预测因子并构建列线图,采用受试者工作特征(receiver operating characteristic,ROC)曲线、校准曲线及决策曲线分析(decision curve analysis,DCA)评估模型性能,并采用Bootstrap重抽样(1000次)进行内部验证以评估模型稳定性.结果 共纳入197例患者,腋窝pCR率为58.38%.HER-2 阳性、NAT后瘤周水肿消失、残余病灶无强化为独立预测因素(P<0.05).模型在训练集与验证集曲线下面积(area under the curve,AUC)分别为 0.895(95%CI:0.839~0.943)和 0.895(95%CI:0.773~0.984),Bootstrap 校正后的乐观校正 AUC 为0.893(95%CI:0.846~0.944),Hosmer-Lemeshow检验提示拟合良好,DCA显示在12%~90%的阈值概率区间内临床净获益显著.结论 本研究模型可有效预测初始腋窝淋巴结阳性乳腺癌NAT后腋窝pCR,判别和校准能力良好,能为腋窝个体化手术决策提供依据,减少不必要的腋窝淋巴结清扫(axillary lymph node dissection,ALND),并改善患者预后.
Objective:To establish a prediction model for axillary pathological complete response(pCR)after neoadjuvant therapy(NAT)in breast cancer patients with initially positive axillary lymph nodes,by integrating multimodal MRI features of the breast,axillary lymph nodes and peritumoral region before and after treatment,as well as clinicopathological indicators,and to evaluate its predictive performance.Materials and Methods:A retrospective analysis was performed on patients with breast cancer and axillary lymph node metastasis admitted to our hospital from January 2022 to August 2025.MR imaging data and clinicopathological information before and after NAT were collected.Patients were randomly divided into a training set and a validation set at a ratio of 8:2.Logistic regression was used to screen independent predictors of axillary pCR and construct a nomogram.The model performance was assessed using receiver operating characteristic(ROC)curves,calibration curves and decision curve analysis(DCA),and bootstrap resampling(1000 times)was performed for internal validation to evaluate model stability.Results:A total of 197 patients were included,with an axillary pCR rate of 58.38%.Positive HER-2 status,disappearance of peritumoral edema after NAT,and non-enhancement of residual lesions were identified as independent predictive factors(P<0.05).The model achieved an area under the ROC curve(AUC)of 0.895(95%CI:0.839 to 0.943)in the training set and 0.895(95%CI:0.773 to 0.984)in the validation set.The bootstrap-corrected optimism-corrected AUC was 0.893(95%CI:0.846 to 0.944).The Hosmer-Lemeshow test indicated good model fitting,and DCA showed significant net clinical benefit within a threshold probability range of 12%to 90%.Conclusions:The model established in this study can effectively predict axillary pCR after NAT in breast cancer patients with initially positive axillary lymph nodes,with favorable discrimination and calibration.It can provide evidence for individualized surgical decision-making for the axilla,reduce unnecessary axillary lymph node dissection(ALND),and improve patient prognosis.
彭雪华;王新正;黎瑞枫;陈翠美;吕蕙芯;易芹芹
龙华区人民医院医学影像科,深圳 518109龙华区人民医院医学影像科,深圳 518109龙华区人民医院医学影像科,深圳 518109龙华区人民医院医学影像科,深圳 518109龙华区人民医院医学影像科,深圳 518109深圳市人民医院放射科,深圳 518020
医药卫生
乳腺癌磁共振成像新辅助治疗腋窝淋巴结病理完全缓解
breast cancermagnetic resonance imagingneoadjuvant therapyaxillary lymph nodespathological complete response
《磁共振成像》 2026 (8)
88-97,10
评论