基于多参数MRI定量特征和临床病理特征预测乳腺癌腋窝淋巴结转移风险  

A multiparametric MRI and clinicopathological features-based model to predict the risk of axillary lymph node metastasis of breast cancer

作　　者：李小苑 杨志企[2] 张胜 朱志强[2] 凌文峰[2] 程凤燕[2] 廖玉婷 陈小凤 LI Xiaoyuan;YANG Zhiqi;ZHANG Sheng;ZHU Zhiqiang;LING Wenfeng;CHENG Fengyan;LIAO Yuting;CHEN Xiaofeng(Department of Breast Surgery,Meizhou People's Hospital,Meizhou,Guangdong Province 514031,China;Department of Radiology,Meizhou People's Hospital,Meizhou,Guangdong Province 514031,China;GE Healthcare,Guangzhou 510623,China)

机构地区：[1]梅州市人民医院乳腺外科,广东梅州514031 [2]梅州市人民医院放射科,广东梅州514031 [3]GE医疗,广东广州510623

出　　处：《实用放射学杂志》2023年第8期1262-1265,1288,共5页Journal of Practical Radiology

基　　金：广东省医学科研基金项目(B2021280);梅州市社会发展科技计划项目(2020B105);梅州市人民医院培育项目(PY-C2020016,PY-C2020022,PY-C2022011)。

摘　　要：目的基于动态对比增强磁共振成像(DCE-MRI)参数、表观扩散系数(ADC)与临床病理参数预测乳腺癌腋窝淋巴结转移(ALNM)风险。方法收集168例女性乳腺癌患者资料,根据病理结果分为ALNM阴性组和阳性组。采用t检验、χ2检验或Kruskal-WallisH检验比较2组间DCE-MRI参数[容量转移常数(K trans)、速率常数(Kep)、血管外细胞外间隙容积分数(Ve)、对比剂流入浓度增强率(W-in)、对比剂流出浓度衰减率(W-out)、达峰时间(TTP)]、ADC值和临床病理特征的差异。采用多因素逻辑回归筛选ALNM的独立危险因素并建立预测模型。应用受试者工作特征(ROC)曲线评估模型的曲线下面积(AUC)。结果ALNM阳性组和ALNM阴性组患者间长径、短径、脉管浸润、T分期、Ki-67、K trans和Ve差异均有统计学意义(P<0.05)。多因素回归分析示脉管浸润、长径、Ve与ALNM独立相关。基于联合脉管浸润、长径、Ve的模型诊断ALNM的AUC最高(0.690),基于脉管浸润模型诊断ALNM的AUC最低(0.570)。结论基于DCE-MRI参数Ve、脉管浸润、长径构建的联合模型可以有效预测乳腺癌ALNM,可以作为独立危险因子。Objective To develop a model to predict the risk of axillary lymph node metastasis(ALNM)in patients with breast cancer based on dynamic contrast-enhanced magnetic resonance imaging(DCE-MRI)parameters,apparent diffusion coefficient(ADC),and clinicopathological features.Methods A total of 168 breast cancer cases were collected,then divided into the ALNM-negative and ALNM-positive groups according to the pathological results.The t-test and χ^(2) test(or Kruskal-Wallis H)were used to compare DCE-MRI parameters[K^(tans),K_(ep),V_(e),W-in,W-out,time to peck(TTP)],ADC value,and clinicopathologic features between ALNM-negative and ALNM-positive groups.The independent risk factors for ALNM were selected using multivariate logistic regression analysis and the prediction models were developed.The receiver operating characteristic(ROC)curve was used to evaluate the area under the curve(AUC)of the models.Results There were statistically significant differences in maximum tumor diameter,minimum tumor diameter,lymphovascular invasion,T stage,Ki-67,K^(trans),and V_(e) between ALNM-negative and ALNM-positive groups(P<0.05).Multivariate logistic regression analysis showed that lymphovascular invasion,maximum tumor diameter,and V were independently associated with ALNM.The model based on the combination of lymphovascular invasion,maximum tumor diameter,and V_(e) showed the highest AUC(0.69o),while the model based on lymphovascular invasion showed the lowest AUC(0.570).Conclusion The model based on the combination of Ve,lymphovascular invasion,and maximum tumor diameter,may serve as a potential predictor marker for the prediction of the risk of ALNM in breast cancer patients.

关 键 词：乳腺癌 腋窝淋巴结转移 磁共振成像 

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