临床-病理-影像联合模型构建乳腺癌术后复发/转移的列线图预测模型  

A nomogram prediction model of postoperative recurrence/metastasis of breast cancer based on the clinic-pathological-imaging combined model

作　　者：黄颢 朱丽华[1] 唐静[1] 余勇江 崔珠红 刘金 HUANG Hao;ZHU Li-hua;TANG Jing;YU Yong-jiang;CUI Zhu-hong;LIU Jin(Department of Breast and Thyroid Surgery,the First People's Hospital of Yibin,Sichuan Province,Yibin 644000,China)

机构地区：[1]四川省宜宾市第一人民医院乳腺甲状腺外科,四川宜宾644000

出　　处：《中国现代普通外科进展》2025年第1期34-39,共6页Chinese Journal of Current Advances in General Surgery

摘　　要：目的:探讨基于临床-病理-影像联合模型构建乳腺癌术后复发/转移的列线图预测模型的临床意义。方法:回顾性选取2019年6月—2022年6月在四川省宜宾市第一人民医院就诊的194例乳腺癌患者,采用单因素和多因素Logistic回归分析筛选乳腺癌术后复发/转移的独立预测因素,并基于独立预测因素构建模型。以7∶3(训练集∶验证集)的比例另取2022年7月—2023年2月83例乳腺癌患者作为验证集验证该模型。结果:乳腺癌术后复发/转移率为29.90%,Ki-67表达水平≥20%、肿瘤位置在内上象限及外上象限、病灶最大径≥20 mm、多个病灶及BI-RADs分级为5级均是乳腺癌术后复发/转移的独立危险因素(P<0.05);而PR阳性表达是乳腺癌术后复发/转移的独立保护因素(P<0.05);临床-病理-影像联合模型的诊断效能(AUC:0.900)较临床病理参数(AUC:0.655)及MRI参数模型(AUC:0.857)高。其基于临床-病理-影像联合模型构建的列线图预测乳腺癌术后复发/转移中,训练集中AUC为0.900(95%CI:0.859~0.942),区分度良好,最大约登值为0.710,敏感度为0.931,特异度为0.779;验证集中AUC为0.820(95%CI:0.712~0.928),区分度良好,最大约登值为0.554,敏感度为0.630,特异度为0.914;两集的校准曲线的理论值和实际值有较好的一致性,决策曲线表示乳腺癌术后复发/转移预测模型的净获益,显示具有良好的预测能力。结论:基于临床-病理-影像联合模型构建的列线图模型具有良好的预测能力、准确性和临床适用性,有助于临床医师评估乳腺癌术后复发/转移的风险。Objective:To investigate the clinical significance of constructing a nomogram based on a combined model of clinical-pathological-imaging data for predicting postoperative recurrence/metastasis of breast cancer.Methods:A retrospective study was conducted on 194 breast cancer patients who were admitted to the department of breast and thyroid surgery from June 2019 to June 2022.unvariate and multivariate Logistic regression analyses were used to screen independent predictors of postoperative recurrence/metastasis of breast cancer,and a model was constructed based on the independent predictors.Another 83 breast cancer patients from July 2022 to February 2023 were taken as the validation set to verify the model with the ratio of 7:3(training set:validation set).Results:The postoperative recurrence/metastasis rate of breast cancer was 29.90%.Ki-67 expression level≥20%,tumor location in the inner upper quadrant and outer upper quadrant,lesion size≥20 mm,multiple lesions,and BI-RADS grade of 5 were independent risk factors for postoperative recurrence/metastasis of breast cancer(P<0.05).PR positive expression was an independent protective factor for postoperative recurrence/metastasis of breast cancer(P<0.05).The diagnostic performance of the combined clinical-pathological-imaging model(AUC:0.900)was superior to that of the clinical-pathological parameters(AUC:0.655)and the MRI parameter model(AUC:0.857).In its nomogram model constructed based on a combined clinical-pathological-imaging model to predict breast cancer recurrence/metastasis after surgery,the AUC in the training set was 0.900(95%CI:0.859~0.942)with good discrimination,the maximum Yoden value was 0.710,the sensitivity was 0.931,and the specificity was 0.779,and the AUC in the validation set was 0.820(95%CI:0.712~0.928),well differentiated,with a maximum Yoden value of 0.554,a sensitivity of 0.630,and a specificity of 0.914.The theoretical and actual values of the calibration curves of the two sets were in good agreement,and the decision curve indicated t

关 键 词：乳腺肿瘤 复发转移 磁共振成像 列线图 

分 类 号：R737.9[医药卫生—肿瘤] R445.2[医药卫生—临床医学]