二维超声联合临床病理特征对乳腺癌Ki⁃67表达水平的预测价值  

Prediction of Ki⁃67 expression in breast cancer by two⁃dimensional Ultrasound combined with clinicopathological characteristics

作　　者：常乾坤 白春强 丁智超 王伟芳[3] 刘铭函 吴文瑛[1] CHANG Qiankun;BAI Chunqiang;DING Zhichao;WANG Weifang;LIU Minghan;WU Wenying(Department of Ultrasound,the Affiliated Hospital of Chengde Medical University,South Campus,Chengde,Hebei,China,067030;Department of Radiology,Chengde,Hebei,China,067024;Department of Ultrasound,Chengde,Hebei,China,067024;Functional Department of Chengde traditional Chinese medicine hospital,Chengde,Hebei,China,067024)

机构地区：[1]承德医学院附属医院南院区超声科,河北承德067030 [2]承德市中心医院放射科,河北承德067024 [3]承德市中心医院超声科,河北承德067024 [4]承德市中医院功能科,河北承德067032

出　　处：《分子诊断与治疗杂志》2025年第1期124-127,136,共5页Journal of Molecular Diagnostics and Therapy

基　　金：河北省医学科学研究课题计划(20231364)。

摘　　要：目的探讨基于二维超声联合临床病理特征预测乳腺癌患者Ki⁃67表达水平的可行性。方法回顾性分析2017年6月至2023年3月承德医学院附属医院乳腺外科收治的624例乳腺癌患者资料。根据免疫组化结果分为Ki⁃67低表达组和高表达组,所有患者术前均接受超声检查。收集超声及临床病理特征资料,采用单因素及多因素Logistic回归分析与Ki⁃67表达相关的因素,分别建立乳腺癌原发灶超声特征、淋巴结超声特征、临床病理特征、超声联合特征及超声联合临床病理特征的预测模型,并应用受试者工作特征(ROC)曲线、曲线下面积(AUC)、De Long检验对各模型进行评价。结果单因素分析,两组在肿物最大径、生长方向、后方回声、微钙化、血供、淋巴结皮质增厚、门结构是否清晰、血流分布、病理类型、ER、PR、HER2及分子分型等因素比较差异有统计学意义(P<0.05)。多因素Logistic回归分析,乳腺癌原发灶病理类型、最大径、后方回声、血供及淋巴结血流分布与乳腺癌Ki⁃67高表达相关(P<0.05)。超声联合临床病理特征模型预测乳腺癌Ki⁃67高表达优于单一特征及超声联合特征预测模型(P<0.05)。结论基于乳腺癌原发灶超声、淋巴结超声和临床病理特征的联合模型能预测乳腺癌Ki⁃67表达水平,具有较好预测价值。Objective To explore the feasibility of predicting Ki⁃67 expression in breast cancer patients based on two⁃dimensional ultrasound combined with clinicopathological characteristics.Methods A retrospective analysis was conducted on 624 breast cancer patients treated at the Department of Breast Surgery,Chengde Medical College Hospital,from June 2017 to March 2023.Based on the immunohistochemical results,patients were divided into low Ki⁃67 expression group and high expression group.All patients underwent ultra⁃sound examination before surgery.Ultrasound and clinical pathological data were collected,and related factors were analyzed and multiple factor logistic regression analysis were used to analyze the related factors.Prediction models for ultrasonic characteristics,lymph nodes ultrasonic characteristics analysis,clinicopathological charac⁃teristics,ultrasonic joint characteristics,and ultrasonic joint clinicopathological characteristics of breast cancer were established.The models were evaluated using receiver operating characteristic(ROC)curve,area under the curve(AUC),and De Long test.Results Univariate analysis showed statistically significant differences in indicators such as maximum tumor diameter,growth direction,posterior echo,microcalcification,blood supply,thickening of lymph node cortex,and clarity of lymph node hilum structure,distribution of lymph node blood flow,pathological type,ER,PR,HER2,and molecular typing(P<0.05).Multivariate logistic regres⁃sion analysis showed that the pathological type,maximum diameter,posterior echo,blood supply and lymph node blood flow distribution of the primary breast cancer focus were significantly correlated with the high expression of Ki⁃67 in breast cancer(P<0.05).The prediction of Ki⁃67 overexpression in breast cancer using an ultrasound combined with clinicopathological feature model was better than that of single feature and ultrasound combined models(P<0.05).Conclusions A combined model based on ultrasonographic features of primary breast cancer

关 键 词：乳腺癌 超声检查 淋巴结 临床病理 Ki⁃67 

分 类 号：R445.1[医药卫生—影像医学与核医学] R737.9[医药卫生—诊断学]