Return To Abstract Listing

Breast Imaging

E2590. Mammographic Evaluation of Symptomatic Women 30–39 Yrs: Chance to Screen Clinically Occult Cancers?

Chen Y1,  Chou S1,  Blaschke E2,  Specht M1,  Lehman C.1 1. Massachusetts General Hospital, Boston, MA; 2. Milwaukee Radiologists , Greenfield, WI

Address correspondence to Y. Chen (

Objective: The American College of Radiology supports either mammography (MG) or breast ultrasound (US) for the initial imaging of symptomatic women aged 30-39 years, including bilateral MG at the discretion of the radiologist. Although the negative predictive value of US approaches 100% in this clinical setting, some recommend bilateral MG as an opportunity to screen for cancer distant from the area of clinical concern. There is sparse evidence to support added value of bilateral MG in this population. In this study, we assessed the performance and added value of bilateral MG in identifying clinically occult cancers (COC) distant from the area of clinical concern that would have otherwise been undetected.

Materials and Methods: In this retrospective, institutional review board–approved study, we identified all women aged 30–39 years who presented for MG and US evaluation of an area or areas of focal clinical concern from June 1, 2006, and August 8, 2016. Each breast with an area or areas of focal clinical concern was designated as a case, yielding 4906 cases. Outcomes were determined by imaging, biopsy, and pathology through linkage with our hospital tumor registry. Performance measures were defined according to the BI-RADS atlas, fifth edition.

Results: Among the 4906 cases, 3892 (79.3%) were assessed as BI-RADS category 1 or 2, 435 (8.9%) were category 3, 564 (11.5%) were category 4, and 15 (0.3%) were category 5. A total of 445 breast biopsies were performed (411 in ipsilateral, 34 in contralateral breast), with a biopsy rate of 9.1%. The 445 biopsies diagnosed 67 cancers (positive predictive value, PPV3 = 15.1%), 16 of which were COC detected by MG only – five in the contralateral breast and 11 in the ipsilateral breast distant from the area of clinical concern – for an added cancer yield of 3.3 per 1000. The 16 COCs included eight ductal carcinoma in situ, seven invasive ductal carcinoma, and one invasive lobular carcinoma. Of the 16 patients with COC detected by MG, one carried the BRCA2 mutation, one had neurofibromatosis type 1, and three had family history of breast or ovarian cancer in first-degree relatives. Fourteen had negative genetic testing, one did not undergo genetic testing, and none had prior diagnosis of a high-risk lesion.

Conclusion: Bilateral mammographic evaluation of women aged 30–39 years with an area or areas of focal clinical concern resulted in an added cancer yield of 3.3 per 1000. The low added cancer yield may support the judicious, rather than routine, use of MG in this patient population.