Jung Min Chang MD, Presenter: Nothing to Disclose

Hye Ryoung Koo MD, Abstract Co-Author: Nothing to Disclose

Woo Kyung Moon, Abstract Co-Author: Nothing to Disclose

To determine the supplemental breast cancer detection yield of screening ultrasound (US) in women at a health screening center.

Between January and December 2008, 6096 women underwent screening mammography (MG) and US at a single health screening center. Of those women, 3731 received prevalence screens and 2365 received incidence screens. Follow up exams were available for 1950 women receiving prevalence screens. Of these 1950 women, 281 women were excluded for positive or suspicious findings on MG. Finally, 1669 women with negative MG constituted our study population. When classified by BI-RADS breast composition, 570 were found to have fatty breasts, and 1099 showed dense breasts. BI-RADS final assessments were analyzed retrospectively, and the reference standard was defined as a combination of pathology and 12-month follow-up. The cancer detection rate (yield), positive predictive value (PPV1) of recalled cases (BI-RADS category 3, 4, 5), and positive predictive value (PPV3) of biopsies following US with negative MG were calculated.

Among a total of 1669 women who underwent US with negative MG, 1159 (75.0%) were classified as BI-RADS category 1 or 2, 415 (20.0%) were classified as BI-RADS category 3, and 95 (5.0%) were classified as BI-RADS category 4 (4A [n=87], 4B [n=5], 4C [n=3]). Core biopsies were recommended and performed for 90 breast lesions, with five lesions being malignant (all BI-RADS category 4). The mean size of invasive cancer was 1.0cm±0.4 (range 0.7-1.5cm). All women with breast cancers had dense breasts. Of the 5 cancer diagnoses, 2 were invasive ductal cancer, 1 was mixed invasive ductal and lobular cancer, and 2 were ductal carcinoma in situ. The cancer detection rate was 3 (5/1669 *1000) per 1000 screens (95% CI, 1.1-7.2). The PPV1 for recall was 0.01 (5/ 510; 95% CI, 0.004-0.023), and the PPV3 for biopsies was 0.06 (5/90; 95% CI, 0.02-0.13). The number of screens needed to detect 1 cancer was 334 (95% CI, 299-371) for US with negative MG.

The addition of screening US to negative MG identified 3 cancers per 1000 screens in normal population, slightly lower than the cancer detection rate in high risk screening with much higher number of false-positives.

Our study supports the benefit of screening breast US in women, especially with dense breasts. However, adding US to MG in this group may not be appropriate without reducing false positives.

Chang, J, Koo, H, Moon, W, Detection of Breast Cancer with Addition of Screening Ultrasound: Results from a Single Health Screening Center.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13044165.html