A Research On The Relationship Between ABVS Imaging Features And Molecular Subtypes Of Breast Invasive Ductal Carcinoma

Object: To explore the relationship between Automated breast volume scanner(ABVS)imaging features and molecular subtypes of breast invasive ductal carcinoma to reveal more predictive imaging features and provide more accurate imaging methods for non-invasive diagnosis of breast invasive ductal carcinoma.Methods: Included in the ABVS imaging examination performed in the Ultrasound Department of the Third Xiangya Hospital of Central South University from June 2019 to June 2021 and patients diagnosed with invasive ductal carcinoma of the breast by surgery or biopsy.A total of 364 women with 364 masses were selected.ABVS imaging features include lesion size,shape,margin,internal echo intensity,retraction phenomenon,post-acoustic feature,hyperechoic halo,microcalcification,orientation,aspect ratio,whether it invades surrounding tissues,and Breast imaging reporting and date system(BI-RADS)classification;The grades were determined according to the pathological results,and the molecular subtypes were determined by immunohistochemistry.The imaging features of breast invasive ductal carcinoma were examined by ABVS,and statistical analysis was performed using Chi-square test or Fisher’s exact test,and then the univariate and multivariate logistic regression analysis was used to further analyze the relationship between ABVS imaging characteristics and molecular subtypes of breast invasive ductal carcinoma.Results:(1)There were significant differences in lesion size and histological classification among the four molecular subtypes(P<0.05),among which the lesion size of Luminal A subtype was mainly ≤2cm in diameter,and its lesion size was significantly smaller than that of other subtypes;and Luminal A had the highest incidence of low-grade breast invasive ductal carcinoma,while triple-negative type had the highest incidence of high-grade breast invasive ductal carcinoma.(2)There were significant differences among the four molecular subtypes in the five ABVS imaging features of retraction phenomenon,post-acoustic feature,hyperechoic halo,microcalcification and aspect ratio(P<0.05).(3)From the univariate logistic regression analysis,it can be concluded that the predictors of Luminal A subtype includedd retraction phenomenon(OR=2.745,P=0.004),post-acoustic shadowing(OR=2.742,P=0.020),hyperechoic halo(OR=2.379,P=0.014),non-microcalcification(OR=2.678,P=0.004),aspect ratio>1(OR=5.148,P=0.006);the predictors of Luminal B subtype included microcalcification(OR=1.591,P=0.035),aspect ratio≤1(OR=3.407,P=0.031);the predictors of HER-2amplified subtype include non-retraction phenomenon(OR=7.724,P=0.005),post-acoustic enhancement(OR=8.784,P=0.040),microcalcification(OR=1.899,P=0.039);the predictors of triple negative subtype included non-hyperechoic halo(OR=4.309,P=0.049),nonmicrocalcification(OR=2.914,P =0.007).(4)From the multivariate logistic regression analysis,it can be concluded that retraction phenomenon(OR=2.645,P=0.014),postacoustic shadowing(OR=2.115,P=0.007),hyperechoic halo(OR=3.194,P=0.008)and non-microcalcification(OR=2.309,P=0.024)were the independent predictors in the presence of Luminal A subtype;microcalcification(OR=1.568,P=0.043)and aspect ratio ≤1(OR=3.288,P=0.038)were the independent predictor in the presence of Luminal B subtype;non-retraction phenomenon(OR=7.282,P=0.007)and microcalcification(OR=2.118,P=0.019)were the independent predictor in the presence of HER-2 amplified subtype;non-hyperechoic halo(OR=5.496,P=0.023)and non-microcalcification(OR=3.402,P=0.002)were the independent predictor in the presence of triple negative subtype.Conclusion: Based on the imaging features of ABVS,there was a strong correlation with the molecular subtypes of breast invasive ductal carcinoma,and it had the potential for non-invasive diagnosis in the identification of breast invasive ductal carcinoma.