Department of Surgery

Research findings, published in JAMA Network Open and led by University of Vermont (UVM) Cancer Center researcher Brian Sprague, Ph.D., show that breast cancer screening with digital breast tomosynthesis (DBT, also known as 3D mammography) improves breast cancer screening performance among most radiologists.

Compared with conventional 2D mammography, DBT on average has higher cancer detection rates and lower recall rates for additional imaging. The research team, a collaboration across the Breast Cancer Surveillance Consortium (BCSC), looked at variation in DBT performance across radiologists, and found that while most radiologists reduced recall rates through its use, not all did.

The researchers’ evaluation suggests that patients and healthcare providers can expect improved breast cancer screening performance with DBT versus conventional 2D digital mammography, but urge that radiology facilities should monitor radiologist performance with DBT to ensure that the benefits are realized.

Sprague, who is a UVM associate professor of surgery and serves as director of the Vermont Breast Cancer Surveillance System (VBSS), says Vermont has been a leader in advancing breast cancer screening for decades. "The early adoption of DBT screening in Vermont created an opportunity to evaluate the performance of this new technology in community practice, in collaboration with other BCSC sites with early adopting facilities.

Since 1994, the VBCSS has been capturing screening and diagnostic breast imaging data, follow-up pathology and cancer registry data for all women in the state of Vermont, looking to evaluate and inform issues around breast cancer screening and diagnosis. The VBCSS is one of six core sites within the BCSC, funded by the National Cancer Institute and the Patient-Centered Outcomes Research Institute.

This study looked at 198 radiologists across 104 different facilities nationally and found that DBT was associated with an overall 15 percent decrease in recall rate and a 21 percent increase in cancer detection rate compared with digital mammography. These results are significant given that many US radiologists (close to 40 percent) have digital mammography recall rates above the expert-recommended rate of 12 percent. This results in a frequent need for diagnostic imaging and often, increased patient anxiety. In addition, only 69 percent of mammography facilities nationwide have DBT capabilities.

This new evaluation of breast cancer screening modalities suggests that expansion of DBT will help lower recall rates on mammography while continuing to maintain or increase cancer detection. Sprague acknowledges that those benefits must be weighed against the extra costs of DBT.

Research in this area is ongoing and will require further examination to ascertain if increased cancer detection results in fewer advanced stage cancers or reduced breast cancer mortality or, less likely, an increased detection of indolent or slow-growing cancers.  

“At minimum, the reduction in recall rate with DBT is a very good benefit for patients—with DBT, they are less likely to be called back for diagnostic imaging for something that turns out to be benign,” says Sprague.

In addition to his role as VBCSS director, Sprague serves as co-leader for the UVM Cancer Center’s Cancer Control and Population Health Sciences research program and principal investigator for the Vermont Mammography Registry. The VBCSS is one of the six core sites of the national Breast Cancer Surveillance Consortium funded by the National Cancer Institute (NCI) and the Patient-Centered Outcomes Research Institute (PCORI) to conduct research to improve the cancer screening process.