Return To Abstract Listing

Scientific Session 08 — Breast - Ultrasound

Tuesday, May 2, 2017

Abstracts 2395-3218

2649. Can Automated Breast Ultrasound Be the Key to Increased Cancer Detection and a Low Callback Rate?

Giakoumis-Spear G*,  Jeske J,  Ecanow J,  Goldstein L,  Tremper C. NorthShore University HealthSystem, Evanston, IL

Address correspondence to G. Giakoumis-Spear (

Objective: The purpose of our study was to evaluate efficacy of automated breast ultrasound (ABUS) as an adjunctive screening tool in women with dense breasts and to help establish practice guidelines for implementation and use of ABUS in the screening setting.

Materials and Methods: A retrospective review of the first 500 ABUS examinations of women with dense breasts (BI-RADS density category C or D) in our hybrid academic and private practice setting between November 1, 2015, and February 19, 2016, was performed. Invenia ABUS (GE Healthcare) scan stations and workstations were used for image acquisition, postprocessing, and interpretation of 3D ultrasound imaging. Board-certified, fellowship-trained breast imagers interpreted the ABUS examinations. All patients were asymptomatic and had a negative screening mammogram within 12 months of ABUS.

Results: Six mammographically occult, small (< 1.5 cm), node-negative, invasive breast cancers were detected on Invenia ABUS (cancer detection rate of 12/1000) during the first 4 months of ABUS implementation. Thirty-six of 500 patients were recalled for additional imaging because of ABUS findings (ABUS callback rate, 7%). Eleven recalled patients underwent image-guided biopsy; positive biopsy rate was 55%. Positive predictive value of ABUS was 6/500 (17%). Logistic regression analysis revealed that patients with prior mammographic callbacks were 3.6 times more likely to be called back because of an ABUS finding (p < 0.05). Those with prior breast cancer history were 6.1 times more likely to have cancer found on ABUS (p < 0.05).

Conclusion: Invenia ABUS is an effective adjunctive screening tool for detection of breast cancer in women with dense breast tissue, overcoming the masking effect of 2D mammography. Clinical practice factors including physician experience in performing and interpreting handheld ultrasound and interpretation of concurrent mammography prove to be integral in successful interpretation of ABUS. We anticipate that our results and analysis will help establish practice guidelines that are much needed to assist with early detection of breast cancer in women with dense breast tissue.