Breast ultrasound can reliably detect nonpalpable malignancies in patients with pathologic nipple discharge who have negative results on mammography, Korean radiologists report in an article published in the Journal of Breast Cancer. They also determined that the US BI-RADS lexicon reliably predicts malignancies.
Radiologists at the Breast Cancer Clinic of Severance Hospital in Seoul conducted a retrospective study of women with pathologic nipple discharge who had negative mammograms followed by ultrasound to determine the effectiveness of breast ultrasound in detecting malignancies. They also hoped to identify specific characteristics seen on ultrasound associated with both malignant and benign lesions.
The patient cohort consisted of 65 women who, after having both exams and an ultrasound-guided biopsy, either had surgery or were followed for two years. The protocol at Severance Hospital for women with BI-RADS category 4 or 5 by mammography or ultrasound (or BI-RADS category 3 if requested by the patient or her physician) is to have an ultrasound-guided percutaneous biopsy.
For the study, two dedicated breast imagers blinded to the final pathologic diagnosis of all breast lesions reviewed the ultrasound studies in consensus. They classified all features according to the BI-RADS lexicon. Ductal involvement patterns were reviewed and classified into three groups: a single mass not associated with a duct, a mass incompletely filling a duct, and a mass completely filling a duct or extending beyond it. The radiologists also assessed ductal wall irregularity if a lesion was associated with ducts. They also recorded the existence of adjacent ductal dilation and recorded the number of involved branch ducts.
The malignancy rate was 15.1%, with breast ultrasound identifying all eight malignant lesions. These included six ductal carcinoma in situ (DCIS), one invasive ductal carcinoma, and one solid papillary carcinoma. 43 patients had surgery for 43 of the 67 lesions identified by breast ultrasound. No patients with benign lesions developed breast cancer in the ensuing 24 months.
Lead author Chae Junk Park, MD, and colleagues reported that malignant lesions more frequently displayed a round or irregular shape (75% compared to 32.2% respectively) and nonparallel orientation compared to the benign lesions (33.3% compared to 7.3% respectively). They noted that lesion size and distance from the nipple were not statistically different, nor were other radiologic findings such as margins, tissue composition or the presence of calcifications.
The small number of patients with ductal involvement did not allow the researchers to definitively say that the ductal involvement pattern was not significant between benign and malignant lesions. This also was true for masses.
The researchers did state that “we assessed whether the ductal wall shows irregularity among lesions that were associated with ducts. Though the number of masses showing irregular ductal wall changes was small, the positive predictive value of ductal wall irregularity was 16.7%, which is comparable to the malignancy rate of BI-RADS category 4. Thus, we suggest that the irregular ductal wall changes at breast ultrasound should be considered as category 4.”
Breast US identifies malignancies in women with nipple discharge who have negative mammograms. Appl Radiol.