DBT recalls fewer asymmetries, boosts ultrasound use on follow-up

A study of digital breast tomosynthesis (DBT) in clinical practice has once again shown a decrease in recall rates with the technology, but also revealed a significant difference in the distribution of the types of abnormalities recalled compared with digital mammography, according to results published online Sept. 22 in Radiology.

The study also revealed a significant change in the imaging that followed a recall from DBT screening, with more patients undergoing ultrasound alone instead of additional mammographic views, according to Ana P. Lourenco, MD, of the Alpert Medical School of Brown University and Rhode Island Hospital in Providence.

In reporting on their initial clinical experience with DBT, the authors reviewed all screening mammography exams performed without tomosynthesis from March 1, 2011, through Feb. 29, 2012, with DBT exams performed from March 1, 2012, through Feb. 28, 2013. They identified all BI-RADS category 0 exams, which indicated a need for additional imaging, and the corresponding radiology and pathology reports were reviewed.

Recall rates were pegged at 9.3 percent for digital mammography and 6.4 percent for DBT, a reduction of 31 percent. The differences in biopsy positive predictive value and cancer detection rate were not statistically significant.

Lourenco and colleagues noted the differences in the types of abnormalities recalled, with DBT resulting in a lower recall rate for asymmetries and focal asymmetries, while mammography had a lower recall rate for masses, distortions and calcifications.

“Because of its pseudo–three-dimensional nature, DBT enables better assessment of superimposed breast tissue, which results in a decrease in the number of false-positive recalls due to summation artifact from overlapping tissue and better assessment of mass margins,” wrote the authors. “Multiple studies have suggested that DBT could replace mammographic spot compression views for the evaluation of noncalcified lesions. The decrease in recalls for asymmetries and focal asymmetries with DBT may be related to improved characterization of overlapping breast tissue as a benign finding, which would reduce recalls for summation of overlapping tissue.”

The authors suggested the 3D nature of DBT also enabled the increase in recalls for distortions due to better visualization of architectural distortion and the associated spiculations amid normal adjacent tissue.

With regard to additional imaging performed after recall, use of ultrasound for diagnostic evaluation increased from 2.6 percent following mammography to 28.3 percent following DBT. This is in line with previous studies and suggests that DBT could effectively replace additional mammographic views in clinical practice, according to the authors.

“In cases where the mass margins are well seen with DBT, the patient may proceed to [ultrasound] without the need for additional mammographic views,” they wrote. “This results in improved efficiency of the diagnostic evaluation and may decrease patient anxiety.”