Nancy Ibrahim MD, Presenter: Nothing to Disclose

Debra Somers Copit MD, Abstract Co-Author: Scientific Advisory Board, Hologic, Inc

Caroline Ling MD, Abstract Co-Author: Nothing to Disclose

Meghan Boros MD, Abstract Co-Author: Nothing to Disclose

Although there is increasing data that tomosynthesis reduces call-back rates and increases cancer detection, there is less information regarding specific abnormalities for which patients are assigned a BI-RADS 0. We have observed that distortion is more commonly seen with tomosynthesis. Having a screening population undergo routine digital mammography (DM)  with and without tomosynthesis, we were able to compare these groups to determine how often they were recalled for possible disortion. Furthermore, we sought to determine the significance of distortion in these patients in terms of diagnosing invasive carcinoma vs radial scars.

We performed a HIPAA compliant retrospective review of all screening studies (DM and DM with tomo) performed at our institution between 7/1/2011 and 12/31/2013. All screening studies assigned BI-RADS 0 were included. Data collected included lesion type for which the patient was recalled, BI-RADS category at diagnostic imaging and final pathology. Statistical analysis assessed differences in recall rates, lesion type and final pathology between the two groups.

  45,850 patients had screening mammography; 24,159 receiving tomosynthesis. Recall rate was significantly decreased from 9.9% in the DM group to 8.5% in the tomo group (95%CI, -1.9 to -.09%, p<.0001). Recall rate for distortion was significantly increased from 3.0% in the DM group to 6.3% in the tomo group (95%CI, 2.0 to 4.6%, p<.0001). The percentage of patients who had suspicious findings (BI-RADS 4/5) on subsequent diagnostic exams was similar in the 2 groups, DM (n=17, 31%) and tomo (n=43, 35%). The percentage of patients diagnosed with cancer was similar in the 2 groups, DM (n=5, 33%) and tomo (n=12, 30%). There was a nonsignificant difference in the percentage of patients diagnosed with radial scar, DM (n=1, 7%) and tomo (n=13, 32%) (95%CI, 4.3 to 42%, p=.086).

The percentage of patients called back for distortion was significantly increased in the tomosynthesis group compared to the DM group. This resulted from an increase in radial scars in the tomosynthesis group, rather than an increase in cancer detection.

Increasing data suggest cancer detection rate is increased in patients umdergoing digital breast tomosynthesis but the additional cancers diagnosed may not be solely attributed to increased detection of distortion.

Ibrahim, N, Copit, D, Ling, C, Boros, M, Distortion and Its Significance at Screening with Digital Breast Tomosynthesis: Do We Call Back More Patients for Distortion After Screening with Tomosynthesis?.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14000711.html