Scientific Session 03 — Breast - Digital Breast TomosythesisMonday, May 1, 2017
2862. Pathologic Outcomes of Architectural Distortion on Digital 2D Versus Tomosynthesis Mammography
Bahl M*, Lamb L, Lehman C. Massachusetts General Hospital, Boston, MA
Address correspondence to M. Bahl (email@example.com)
Objective: Recent research demonstrates that digital breast tomosynthesis (DBT) improves the detection of architectural distortion over conventional digital mammography (DM); however, the risk of malignancy associated with architectural distortion detected on DBT has not been well studied. The purpose of this study is to compare the risk of malignancy associated with architectural distortion detected on DM versus DBT.
Materials and Methods: The study cohort was comprised of patients with architectural distortion on mammography at a single large tertiary academic medical center from September 2007 to February 2011 (DM group, before DBT integration) and from January 2013 to June 2016 (DBT group, after DBT integration). Medical records were reviewed for imaging findings, biopsy pathology results, and surgical outcomes.
Results: Over the 3.5-year period before DBT integration, architectural distortion considered to be suspicious for or highly suggestive of malignancy was present in 121 of 166,661 (0.07%) mammographic examinations (DM group). By contrast, over the 3.5-year period after DBT integration, architectural distortion was present in 274 of 202,438 (0.14%) mammographic examinations (DBT group) (p < 0.001). The positive predictive value (PPV) of architectural distortion for malignancy (invasive carcinoma or ductal carcinoma in situ) was significantly higher in the DM group than the DBT group (73.6% [89/121] vs 50.4% [138/274], p < 0.001). The majority of malignancies were invasive rather than in situ carcinomas (95.5% invasive in the DM group and 91.3% invasive in the DBT group). The most common nonmalignant finding on pathology was a radial scar or complex sclerosing lesion in both the DM and DBT groups, but that finding was less common in the DM group than the DBT group (11.6% vs 32.8%, p < 0.001). In the DM group, architectural distortion with or without a sonographic correlate represented malignancy at similar rates (77.1% vs 63.2%, p = 0.20). In the DBT group, however, architectural distortion without a sonographic correlate was less likely to represent malignancy than architectural distortion with a sonographic correlate (29.2% vs 65.8%, p < 0.001).
Conclusion: Architectural distortion is a relatively rare finding on DM and DBT but is more commonly detected on DBT than DM. The risk of malignancy is significantly lower when detected on DBT than on DM (50.4% vs 73.6%, p < 0.001) but still warrants biopsy. Architectural distortion detected on DBT is less likely to represent malignancy if there is no sonographic correlate; however, negative findings on ultrasound do not obviate biopsy because the risk of malignancy in this setting is nearly 30%.