Margarita Louise Zuley MD, Presenter: Research grant, Hologic, Inc

Jules Henry Sumkin DO, Abstract Co-Author: Scientific Advisory Board, Hologic, Inc

Marie Adele Ganott MD, Abstract Co-Author: Nothing to Disclose

Andriy I. Bandos PhD, Abstract Co-Author: Nothing to Disclose

Amy Elizabeth Kelly MD, Abstract Co-Author: Nothing to Disclose

Amy H Lu MD, Abstract Co-Author: Nothing to Disclose

Victor J Catullo MD, Abstract Co-Author: Nothing to Disclose

Denise Chough MD, Abstract Co-Author: Nothing to Disclose

David Gur ScD, Abstract Co-Author: Nothing to Disclose

Ronald L. Perrin MD, Abstract Co-Author: Nothing to Disclose

To compare diagnostic performance of digital breast tomosynthesis (DBT) to additional diagnostic digital mammography (FFDM) views in the classification of masses, distortions and asymmetries initially identified on standard 2 view mammography, ultrasound, clinical exam or MRI.

217 consecutively accrued lesions in 182 consenting patients who underwent diagnostic FFDM and DBT under an IRB approved protocol were retrospectively reviewed by 8 breast imaging radiologists who have experience interpreting DBT. The IRB protocol under which the patients were recruited included patients presenting with a clinical symptom, for screening recall, or high risk screen monitoring. The cohort included 182 (84 %) masses, 25 (11 %) asymmetries and 10 (5 %) distortions initially detected by clinical exam (n=1)(8 %), mammography (n=173)(80 %), US (n=25)(11 %), or MRI (n=2)(1 %). Truth was based on biopsy (n=191), US showing an anechoic cyst (n=12) or 1 year normal follow up for benign stable asymmetries (n=14). Lesion location and type was provided to the readers for this study. Each lesion was interpreted twice; once with DBT and the standard mammographic views (MLO and CC) and once with all clinically obtained diagnostic additional FFDM views and standard 2 view mammography in a fully crossed and balanced design. In both modes, radiologists provided a 5 point BI-RADS score. A generalized linear model for binary data was used for analysis using a cutpoint of BI-RADS 4.

The cohort included 72 (33%) cancers and 145 (67%) benign lesions, measuring 2-93mm (mean 18mm, median 13mm). Across all cases, biopsy recommendation rate (defined as BI-RADS 4 or 5) was 58% using DBT vs 61% with additional views (p=0.25). Specificity of DBT was 55% and additional views 52% (p=0.16). Sensitivity of DBT was 86% and additional views 88% (p=0.97).

Digital breast tomosynthesis has comparable sensitivity and specificity to standard diagnostic additional mammographic views in the classification of non-calcified breast lesions.

Digital Breast Tomosynthesis may be able to replace supplemental diagnostic mammography images in the evaluation of patients suspected of having a non-calcified breast lesion.

Zuley, M, Sumkin, J, Ganott, M, Bandos, A, Kelly, A, Lu, A, Catullo, V, Chough, D, Gur, D, Perrin, R, Digital Breast Tomosynthesis vs Supplemental Diagnostic Mammography Images for the Evaluation of Noncalcified Breast Lesions.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11015838.html