Comparing Digital Breast Tomosynthesis to Digital Mammography: What a Long-Term Study Reveals

In what may be the largest study of screening digital breast tomosynthesis (DBT) performed by a single institution over a 10-year period, researchers found that DBT offered enhanced breast cancer detection, a lower recall rate and an 11 percent lower rate of advanced cancer presentation in contrast to digital mammography (DM) screening.

For the retrospective study, recently published in Radiology, researchers compared 10 years of DBT screening and three years of DM screening in a total of 272,938 breast cancer screening exams performed over a 13-year period at Yale University. There were 35,544 DM screenings, 109,478 DBT screenings with full-field DM, and 127,916 DBT screenings with synthetic mammography, according to the study. Of the 1,407 cases of true-positive breast cancer, the study authors noted that 1,265 were detected with DBT and 142 were diagnosed with DM.

Digital breast tomosynthesis had a breast cancer detection rate of 5.3 per 1000 in comparison to 4.0 per 1000 for DM, according to the study authors. The researchers also noted a significantly lower proportion of advanced cancer presentation with DBT screening (32.6 percent) in contrast to DM screening (43.6 percent). The study findings also revealed a lower recall rate for DBT (7.2 percent) as opposed to DM (10.6 percent).

“Digital breast tomosynthesis (DBT) is emerging as an improved option for breast imaging. … This allows for viewing the breast tissue in thin sections that help reduce the confounding effects of overlapping tissue, allowing better identification, localization, and characterization of lesions,” wrote lead study author Liane Elizabeth Philpotts, M.D., who is affiliated with the Department of Radiology and Biomedical Imaging at the Yale School of Medicine at Yale University in New Haven, Ct., and colleagues.

The study authors noted a fair amount of similarities between DBT and DPM including size of invasive cancers, rate of malignant architectural distortion and tumor grade. For the DM cohort, 29 percent had low-grade tumors, 51 percent had moderate tumors and 20 percent had high-grade tumors, according to the study authors. In comparison, the researchers noted that 26.5 percent of tumors in the DBT group were low-grade tumors, 57.2 percent were moderate, and 16.1 percent were high-grade tumors.

Noting that they defined interval cancers as those occurring within a year if a normal examination, the researchers pointed out similar rates of interval cancers between DBT (0.2 per 1000) and DM (0.14 per 1000). They suggested that adjunctive ultrasound screening may have been a factor in this finding.

“Our institution has a robust screening (ultrasound) program and since 2009, most of our patients with dense breasts have undergone this adjuvant screening. This likely filters out many invasive cancers that may otherwise manifest as interval cancers,” posited Philpotts and colleagues.

However, the researchers also found that those who had DBT had more malignancies manifesting as masses (40 percent vs. 26.8 percent), fewer cases involving asymmetries (20.2 percent vs. 28.9 percent) and fewer malignancies manifesting with calcifications (27.9 percent vs. 37.3 percent) in comparison to those with DM screening.

“Conspicuity of calcifications at synthetic mammography compared with DM is similar. However, because calcifications often represent in situ disease and DBT trends toward depicting more invasive disease, the lower proportion of malignant calcifications is not unexpected,” added Philpotts and colleagues.

(Editor’s note: For related content, see “What New Research Reveals About Digital Mammography and Digital Breast Tomosynthesis,” “Mammography Study Finds No Additional Benefit with DBT in Women with Elevated Breast Cancer Risk” and “Mammography Study: AI Improves Breast Cancer Detection and Reduces Reading Time with DBT.”)

In regard to study limitations, the authors noted the retrospective design and varying intervals between previous screening exams in patients with breast cancer. They also pointed out possible breast density classification variation between the DM and DBT groups due to the advent of the fifth edition of BI-RADS in 2013. Given that the data came from one tertiary academic center, the researchers cautioned against extrapolation of the study findings to broader populations.