A week ago, the American Roentgen Ray Society (ARRS) held its annual meeting in a virtual format. Researchers from the University of Massachusetts and Yale University presented a paper in which they studied the screening of women using a combination approach: digital breast tomosynthesis (DBT) and automated breast ultrasound (ABUS). While recognizing that DBT decreases screening recall rates and provides more information than full-field digital mammography (FFDM), they also pointed out that ABUS can detect an additional 4.6 cancers per 1,000 women screened compared to DBT. However, they were concerned about a potential increase in false positives.

The University of Massachusetts has a ten-year history of using both DBT and ABUS, and the current study examined 12,649 women with either heterogeneously or extremely dense breasts during the 2019 calendar year. Approximately one-third of the women underwent DBT only, while the other two-thirds had both DBT and ABUS conducted on the same day. The BI-RADS classification system was used by the radiologists to categorize both the DBT and ABUS images as either normal or abnormal.

The table at left (adapted from AuntMinnie.com) summarizes their findings. What the researchers discovered was that by adding ABUS to DBT the rate of abnormal findings was halved, from 14% to 6.7%, with the difference being highly significant statistically. Interestingly, this reduction occurred despite the finding that ABUS detected an additional 2.9% of suspicious findings not seen by DBT alone. First author Cameron Thomson commented: “These findings are important as this potentially permits women with dense breast tissue the opportunity to benefit from additional cancer detection without the fear of increased adverse, abnormal interpretation rates when opting for supplemental screening with whole-breast ultrasound.”

There are three problems when using separate DBT and ABUS systems in a screening environment: (1) it takes up to 30 minutes for both systems compared to just 10 minutes for DBT-only; (2) the breast is compressed differently, making co-registration of the DBT and ABUS images extremely difficult; and (3) there is the added expense of acquiring two systems.

As we have suggested in the past, the solution to these problems could be the integration of DBT and ABUS in a single device. A group working with Siemens recently published a paper in which they presented clinical data for a prototype system (see co-registered images at right, © European Radiology). However, their ABUS coverage of the breast was just 80% of the DBT coverage, compared with CapeRay’s patented Aceso system that has achieved more than 90% coverage.