Ultrasonographic alterations associated with the dilatation of mammary ducts: Feature analysis and BI-RADS assessment

Objective: The purpose of this study was to analyse the lesion characteristics and the patterns of dilated ducts on ultrasonography (US) to determine the appropriateness of the Breast Imaging Reporting and Data System (BI-RADS) categories. Materials and methods: From July 2001 to June 2006, 172 consecutive pathologically proved lesions with dilated ducts on US were reviewed retrospectively. All the lesions were classified into four types according to their US features, and in combination with the size, location, margins and number of lesions, the corresponding positive predictive values (PPVs) were obtained. Results: Of the 172 lesions, 55 (32%) were classified as type I, 68 (40%) as type II, 14 (8%) as type III and 35 (20%) as type IV. The PPVs for malignancy were 9% for type I, 13% for type II, 43% for type III and 17% for type IV. There was a significantly higher frequency of malignancy among type III lesions than among type I (43% vs 9%, respectively, P=0.002; Χ² test) or type II lesions (43% vs 13%, respectively, P=0.009; Χ² test). Lesions with a nonsubareolar location and noncircumscribed margins had a high probability of malignancy (P<0.001 and P=0.03, respectively). Conclusion: The four types of US classifications used in our study establish reliable references for the dilated duct patterns when stratified according to BI-RADS categories, and they clarify the indications for biopsy of these lesions.