Jennifer F. Wells MD, Presenter: Nothing to Disclose

Regina J. Hooley MD, Abstract Co-Author: Nothing to Disclose

Madhavi Raghu MD, Abstract Co-Author: Nothing to Disclose

Melissa Angeline Durand MD, Abstract Co-Author: Nothing to Disclose

Paul H. Levesque MD, Abstract Co-Author: Nothing to Disclose

Liva Andrejeva-Wright MD, Abstract Co-Author: Nothing to Disclose

Laura Jean Horvath MD, Abstract Co-Author: Consultant, Siemens AG

Liane Elizabeth Philpotts MD, Abstract Co-Author: Nothing to Disclose

The positive predictive value (PPV) of biopsy of masses detected at screening breast ultrasound (SBUS) is low compared to masses at screening mammography. The purpose of this study was to determine if biopsy can potentially be avoided in some solid masses found at SBUS.

An IRB approved retrospective chart review was performed on 100 BI-RADS 4 lesions in 91 patients detected only on SBUS during 10/15/2009 - 9/26/2012. Two radiologists blindly reviewed the sonographic images of each lesion, recorded BI-RADS features, and assigned each lesion a final assessment score.

The average patient age was 53 years (range 32 – 84). The average lesion size was 8.8 mm (range of 3-30). 80/100(80%) lesions were identified on a prevalence screening exam. 96 lesions were aspirated or biopsied, yielding 5 invasive ductal carcinomas and 1 DCIS for a PPV3 of 6.3%. All cancers (6/6) had indistinct, angular, or irregular margins and hypoechoic or heterogeneous echotecture. 5/6 cancers had a non-parallel orientation. All malignancies were assessed as BI-RADS 4B or 4C, except for the one case of DCIS assessed as BI-RADS 3 by one reader. Of the 94 benign masses, 37(39%) were circumscribed, 56(60%) were hypoechoic, 64(68%) were oval or round, 53(56%) had a parallel orientation, and 68 (72%) had enhanced or no posterior acoustic features. The negative predictive value of an oval mass with circumscribed margins was 100%, regardless of echo pattern or posterior acoustic features. 7 retroareolar lesions were associated with a dilated duct and all were benign. 80/99 (81%) of cases were assigned a BI-RADS 2, 3, or 4A final assessment by at least one reader. If the lower BI-RADS score was used and these lesions were not biopsied, the PPV would have increased to 37.5% and one case of DCIS would have been downgraded to BI-RADS 3.

Retrospective blinded review showed many BI-RADS 4 lesions detected on SBUS have benign features and biopsy could probably have been avoided. Most cancers detected on SBUS have suspicious US features and are correctly identified by radiologists. Solid masses detected on SBUS require careful evaluation and may benefit from double reading in order to improve specificity.

Screening breast US is being more widely performed in the United States. Strategies are needed to improve overall diagnostic performance and decrease false positive biopsies.

Wells, J, Hooley, R, Raghu, M, Durand, M, Levesque, P, Andrejeva-Wright, L, Horvath, L, Philpotts, L, Reducing False Positive Biopsy Rate of Screening Ultrasound Detected Breast Masses.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14045479.html