Preoperative breast MRI combined with axillary ultrasound for the prediction of lymphovascular invasion in invasive ductal carcinoma of the breast / 中华放射学杂志

ABSTRACT

Objective:To investigate the value of preoperative breast MRI combined with axillary ultrasound in predicting lymphovascular invasion (LVI) of breast invasive ductal carcinoma.Methods:The clinical, pathological and imaging features of 160 female patients [age 25-74(49±10)years] with breast invasive ductal carcinoma from March 2014 to December 2017 in Shanxi Cancer Hospital were retrospectively analyzed. According to the LVI status determined by postoperative pathology, 160 patients were divided into LVI positive group (56 cases) and LVI negative group (104 cases). The clinical characteristics, pathological characteristics and imaging features of LVI positive group and LVI negative group were compared by the independent t test, Mann-Whitney U test or χ 2 test. Multivariate logistic regression analysis was performed to identify independent predictors for predicting LVI and construct a predictive model. The receiver operating characteristic (ROC) curve and area under the curve (AUC) was used to evaluate the discrimination of the prediction model, and the Hosmer-Lemeshow test was used to evaluate its calibration. Results:There was no significant difference in age, menopausal status, estrogen receptor, progesterone receptor, human epidermal growth factor 2, Ki67 index and molecular subtype between LVI positive group and negative group ( P>0.05). Tumor size, peritumoral edema, adjacent vessel sign, multifocality or multicentricity, peritumoral maximum-apparent diffusion coefficient (ADC), peritumour-tumour ADC ratio, MRI axillary lymph node status and ultrasound axillary lymph node status between LVI positive group and LVI negative group showed significantly statistical difference ( P<0.05). Variables with significant difference in the univariate analysis were entered into multivariate logistic regression analysis to explore predictors for LVI. Peritumoral edema (OR=3.367, 95%CI 1.382-8.201, P=0.008), multifocality or multicentricity (OR=4.026, 95%CI 1.268-12.776, P=0.018), high peritumoral-tumor ADC ratio (OR=7.321, 95%CI 2.226-24.079, P=0.001) and positive ultrasound axillary lymph node (OR=6.779, 95%CI 2.819-16.303, P<0.001) were independent predictors for predicting LVI. A logistic regression model was constructed using the above four indicators, and ROC showed AUC of this model for predicting LVI was 0.882, superior to any of the single indicator ( P<0.05); its sensitivity was 80.36% and specificity was 84.62%. Hosmer-lemeshow test showed that the prediction model had good calibration ( P=0.503). Conclusion:The combined prediction model constructed by preoperative breast MRI and axillary ultrasound could help to predict the LVI status of breast invasive ductal carcinoma.