Hendrik Jelle Teertstra MD, Presenter: Grant, Hologic, Inc, Bedford, MA

Claudette Elisabeth Loo MD, Abstract Co-Author: Nothing to Disclose

Angelique Schlief, Abstract Co-Author: Nothing to Disclose

Saar Muller, Abstract Co-Author: Nothing to Disclose

Emiel Rutgers, Abstract Co-Author: Nothing to Disclose

Kenneth G.A. Gilhuijs PhD, Abstract Co-Author: Nothing to Disclose

To compare clinical decision making with breast tomosynthesis (tomo) and digital mammography (DM) in women referred to our outpatient breast clinic because of an abnormal screening mammogram, symptoms or for second opinion.

95 symptomatic subjects (age range 28-92, median 53) underwent DM and tomo after informed consent. Tomosynthesis was not used for clinical decision-making. The BIRADS classification for breast density and lesion characterization was assessed independently for both techniques by different radiologists, all experienced in breast imaging (all tomo images were scored by the first author). Results were compared to definitive diagnosis (pathology, clinical outcome) using ROC analysis and Wilcoxon signed ranks tests.

Breast-density scores were significantly lower for tomo than for DM (N=93, BIRADS 1 through 4: 20, 36, 25, 12 vs. 13, 24, 33, 23, p<0.001). In 73 breasts, no abnormalities were found by either technique. The ability to detect and classify malignant lesions was not significantly different for tomo than for DM (AUC = 0.90 vs. 0.95, p=0.28). In 117 breasts, 120 lesions were found. 43 Lesions were carcinomas (19/43 were BIRADS-6 (second opinion)). Of the remaining 24, 2 cases had mammography scores 1-3 ( 1, 3 vs. 1, 4 for tomo). Four cases had tomo scores 1-3 (1, 1, 3, 3 vs. 1, 4, 4, 4 for mammography). 74 patients had a benign lesion. In some cases tomo image quality was reduced as a result of patient motion (scantime tomo is 10 seconds). Without further workup ( ultrasound, FNAB), 4/24 carcinomas could have been missed using tomo alone, 2 using mammography alone and 2 using both techniques.

Tomosynthesis yields significantly lower breast density scores than digital mammography. No difference in the ability to discriminate between normal/benign findings and malignant lesions was seen between tomosynthesis and digital mammography.

Tomosynthesis is promising for lesion detection and lesion characterisation because of reduced superposition , in symptomatic patients however, the role of tomosynthesis is not yet established.

Teertstra, H, Loo, C, Schlief, A, Muller, S, Rutgers, E, Gilhuijs, K, Comparison of Breast Tomosynthesis and Digital Mammography in Symptomatic Patients Referred to an Outpatient Breast Clinic.  Radiological Society of North America 2007 Scientific Assembly and Annual Meeting, November 25 - November 30, 2007 ,Chicago IL. http://archive.rsna.org/2007/5012769.html