Objective To study the value of preoperative breast magnetic resonance imaging features in predicting inva-sive breast cancer with intraductal component.Methods Patients with pure invasive breast cancer(IBC)and invasive breast cancer with intraductal component(IBC-IC)proved by pathology after surgery were enrolled retrospectively.The MRI features based on breast imaging reporting and data system(BI-RADS),as well as the presence and types of early enhance-ment adjacent to the lesion were analyzed.Continuous variables between two groups were compared using the independent t-test,while categorical variables were compared using the chi-square test or Fisher's exact test when appropriate.Significant features in the univariate analysis were included in a multivariate logistic-regression model to identify the most important predictors of IBC-IC.Receiver operating characteristic(ROC)curve was used to evaluate the performance of the model.Results 176 cases with 132 pure IBC and 44(25.0％)IBC-IC were recruited in the study.Univariate analysis revealed that the MRI features of non-mass enhancement or mass with non-mass enhancement(x2=40.028,P＜0.001),irregular or speculated mass margin(x2=6.328,P=0.012),the presence and types of early enhancement adjacent to the lesion(x2=37.140,P＜0.001)were predictors of IBC-IC.Multivariate logistic regression analysis shows that the presence of focal early enhancement adjacent to the lesion(OR=5.903,P=0.002),the presence of nodular early enhancement adjacent to the lesion(OR=3.157,P=0.035),the presence of regional early enhancement adjacent to the lesion(OR=6.117,P=0.018)and non-mass enhancement or mass with non-mass enhancement(OR=8.718,P=0.01)were the independent pre-dictors of IBC-IC,and the AUC of the prediction model was 0.864.Conclusion Preoperative MRI features,particularly the presence of early enhancement adjacent to the lesion,can serve as predictors for invasive breast cancer with intraductal component.
Breast tumorMagnetic resonance imagingInvasive breast cancer with intraductal component
NETL
NSTL
万方数据
