乳腺癌多模态超声特征与Ki-67表达的相关性  

Investigation on the correlation between multimodal ultrasound features of breast cancer and Ki-67 expression

作　　者：贺玉卿 吴梓政 齐铮琴[1] 郑晶晶[1] 郭帅 HE Yuqing;WU Zizheng;QI Zhengqin;ZHENG Jingjing;GUO Shuai(Department of Ultrasound Diagnosis,The First Hospital of Qinhuangdao,Qinhuangdao 066001,Hebei,China)

机构地区：[1]秦皇岛市第一医院超声诊断科,河北秦皇岛066001

出　　处：《皖南医学院学报》2025年第2期159-162,共4页Journal of Wannan Medical College

基　　金：河北省医学科学研究课题计划(20231893);秦皇岛市科学技术研究与发展计划项目(202301A199)。

摘　　要：目的:探讨乳腺癌多模态超声特征与Ki-67表达的相关性。方法:选取127例经病理诊断的女性乳腺癌患者为研究对象,所有患者抗肿瘤治疗前均已对乳腺癌原发病灶行常规超声、多普勒血流显像(CDFI)及剪切波弹性成像(SWE)检查,收集包括肿瘤最大径、边界、有无微钙化、纵横比、Adler血流分级及SWE特征表现等多模态超声特征。所有患者按Ki-67表达程度划为Ki-67低表达组(41例)和Ki-67高表达组(86例),采用单因素及多因素Logistic回归分析多模态超声特征与Ki-67表达情况的相关性,建立基于多模态超声参数与乳腺癌Ki-67表达情况的预测模型。结果:Ki-67低表达组与Ki-67高表达组在乳腺癌原发灶有无微钙化、Adler血流分级、有无SWE硬环征及Emax值等方面差异有统计学意义(P<0.05)。Logistic回归分析显示,原发灶有微钙化、Adler血流分级Ⅱ~Ⅲ级、有硬环征及Emax值增高是Ki-67高表达的独立危险因素。联合以上4个危险因素构建预测模型,AUC为0.786(95%CI:0.700~0.872)。结论:乳腺癌原发灶有微钙化、Adler血流分级Ⅱ~Ⅲ级、有硬环征及Emax值增高提示乳腺癌Ki-67高表达风险更高,以多模态超声特征构建的Logistic回归模型可预测乳腺癌Ki-67表达情况。Objective:To investigate the correlation between multimodal ultrasound features of breast cancer and Ki-67 expression.Methods:127 cases of female breast cancer patients diagnosed by pathology were selected as the research subjects.All patients underwent routine ultrasound,color Doppler flow imaging(CDFI),and shear wave elastography(SWE)examinations for the primary lesion of breast cancer before receiving any anti-tumor treatment.Data collection included tumor maximum diameter,margin characteristics,presence of micro-calcifications,transverse-to-longitudinal ratio,Adler blood flow grading,and SWE feature presentation as well as other multimodal ultrasound features.Then the patients were categorized into a low Ki-67 expression group(n=41)and a high Ki-67 expression group(n=86).Single-factor comparison and multivariate logistic regression analysis were applied to investigating the correlation between multimodal ultrasound features and Ki-67 expression,based on which a predictive model was established.Results:Single-factor comparison indicated significant differences between the low Ki-67 expression group and the high Ki-67 expression group in factors such as the presence of micro-calcifications in the primary lesion of breast cancer,Adler blood flow grading,presence of SWE"hard ring sign",and Emax value(P<0.05).Binary logistic regression analysis showed that the presence of micro-calcifications,Adler blood flow grading between II and III,presence of the"hard ring sign",and elevated Emax value were independent risk factors for high Ki-67 expression.A predictive model was constructed by combining the four risk factors aforementioned,and the area under the curve(AUC)was 0.786(95%CI:0.700-0.872).Conclusion:The presence of micro-calcifications,Adler blood flow grading between II and III,presence of the"hard ring sign",and elevated Emax value in the primary lesion of breast cancer suggest a higher risk of high Ki-67 expression,and a logistic probability binary classification model constructed using multimodal ultrasound f

关 键 词：乳腺癌 超声检查 剪切波弹性成像技术 LOGISTIC回归 

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]