Prediction of MRI-Based Peritumoral Edema of Lymph Node Metastasis Burden in Patient with Invasive Breast Cancer
10.3969/j.issn.1005-5185.2025.01.010
- VernacularTitle:基于MRI的瘤周水肿特征对浸润性乳腺癌淋巴结转移负荷的预测价值
- Author:
Hongbing LUO
1
;
Zhe CHEN
;
Qianqian XIAO
;
Jing REN
;
Peng ZHOU
Author Information
1. 四川省肿瘤医院·研究所 四川省肿瘤临床医学研究中心 四川省癌症防治中心 电子科技大学附属肿瘤医院影像科,四川 成都 610041
- Publication Type:Journal Article
- Keywords:
Breast neoplasms;
Lymph node metastasis;
Magnetic resonance imaging;
Edema;
Diagnosis,differential;
Forecasting
- From:
Chinese Journal of Medical Imaging
2025;33(1):55-62
- CountryChina
- Language:Chinese
-
Abstract:
Purpose To investigate the predictive value of MRI-based peritumoral edema(PE)of lymph node(LN)metastasis burden in patient with invasive breast cancer.Materials and Methods A total of 213 patients in Sichuan Cancer Hospital with invasive breast cancer confirmed by pathology after surgery were enrolled retrospectively from September 2017 to February 2019.Based on the status and number of LN metastases proved by axillary surgery pathology,all patients were divided into high-burden LN metastasis group(n=47 cases,total number of metastatic LN greater than 2)and low-burden LN metastasis group(n=166 cases,total number of metastatic LN less than or equal to 2).T2WI features,including PE type and PE degree,were analyzed.On DCE-MRI,MRI features of breast cancer were analyzed based on BI-RADS.The predictive value of T2WI features and MRI features for LN metastasis burden was assessed by univariate analysis,and the statistically significant features identified by univariate analysis were further subjected to multivariate Logistic regression analysis to establish a prediction model.The model's performance for LN metastasis burden prediction was evaluated via receiver operating characteristic curve,and the predictive metrics were calculated based on the Youden index.Results The rate of high-burden LN metastasis was 22.1%(47/213).The results of univariate analysis showed that PE degree(OR=18.70,P＜0.001),PE type(OR=16.00,P＜0.001),tumor maximum diameter(OR=1.40,P=0.025),and tumor minimum diameter(OR=2.01,P=0.003)were predictive features for high-burden LN metastasis.The multivariate regression analysis showed that the peritumoral edema features,including PE degree(OR=8.02,P＜0.001)and PE type(OR=5.53,P=0.001)were independent predictive features for high-burden LN metastasis.The area under the curve of the final predictive model was 0.842.The diagnostic sensitivity was 0.766,the specificity was 0.861,the positive predictive value was 0.610,and the negative predictive value was 0.929.Conclusion The PE on preoperative MRI have good predictive performance for LN metastasis burden prediction in patient with invasive breast cancer,particularly for low-burden LN metastasis.
