Breast tomosynthesis for the clarification of mammographic BI-RADS 3 lesions can decrease follow-up examinations and enables immediate cancer diagnosis. (October 2018)

• Record Type: Journal Article
• Authors: Bahrs, Sonja D; Otto, Vanessa; Hattermann, Valerie; Klumpp, Bernhard; Hahn, Markus; Nikolaou, Konstantin; Siegmann-Luz, Katja
• Is Part Of: Acta radiologica. Volume 59:Number 10(2018:Dec.)
• Page Start: 1176
• Page End: 1183
• Publication Date: 2018-10
• Subjects: Mammography -- BI-RADS 3 -- digital breast tomosynthesis; Radiology, Medical -- Periodicals; Radiography, Medical -- Periodicals; Radiotherapy -- Periodicals; 616.0757
• DOI: 10.1177/0284185118756458
• Languages: English
• ISSNs: 0284-1851

Abstract

Background: The limited sensitivity of mammography in case of a high breast density often produces unclear or false-positive findings, so-called BI-RADS 3 lesions, which have to be followed up to prove benignity. Digital breast tomosynthesis (DBT) was developed to reduce such summation effects. Purpose: To evaluate the influence of an additional DBT on the management of mammographic BI-RADS 3 findings and whether DBT can decrease the time to definitive diagnosis or not. Material and Methods: We analyzed 87 patients with a mammographic non-calcified BI-RADS 3 lesion who underwent an additional DBT of the affected breast. A follow-up two-dimensional (2D) examination or a histological result of the lesion had to be available. The images were analyzed especially for the BI-RADS category and incremental diagnostic accuracy. Moreover, the inter-reader reliability and the radiation dose were evaluated. Results: The BI-RADS category has been changed by the addition of DBT: 57.1% were assessed as BI-RADS 1 or 2, 4.6% as BI-RADS 4, and only 38.3% remained as BI-RADS 3. The intraclass correlation coefficient for the three readers showed a good agreement for inter-reader reliability. No false-negative examination was found in the follow-ups. Nine lesions were biopsied (seven benign, two malignant). Both malignant lesions were suspicious in the DBT (BI-RADS 4). A significant higher glandular dose was necessary for the DBT. Conclusion: DBT has the potential to reduce the recall-rate of BI-RADS 3 lesions and to find and diagnose malignant lesions earlier than 2D mammography alone.