CHICAGO – Screening all-comers with three-dimensional mammography increased breast cancer detection by 22.7% and reduced recall rates by 15.6% in a large observational study.
Although a 3D digital breast tomosynthesis (DBT) system was approved for breast cancer screening and diagnosis in the United States in 2011, DBT is typically a supplemental screening tool to standard 2D digital mammography.
In November 2011, however, the Hospital of the University of Pennsylvania, Philadelphia, took the plunge and began using DBT for every breast screening patient, regardless of age, cancer risk, breast density, or ability to pay.
Over a 17-month-period, DBT detected 82 cancers in 15,632 women, or 5.25 cancers per 1,000 cases, Dr. Emily F. Conant reported at the annual meeting of the Radiological Society of North America.
Dr. Emily F. Conant
This compares with 46 cancers, or 4.28 cancers per 1,000 cases, detected with conventional mammography in 10,752 women in the 12 months prior to the switch.
The difference in cancer detection rates between 2D and 3D mammography did not reach statistical significance (P = .226), probably because of the small number of cases, Dr. Conant, a radiology professor at the hospital, said at a press briefing.
Compared with conventional mammography, however, DBT significantly improved the proportion of positive screening mammograms from which cancer was diagnosed by 45% (P = .036) and increased the detection of deadly invasive cancers by a nonsignificant 31%.
When asked whether the study showed that invasive lobular cancer can be better detected with DBT, Dr. Conant replied, "Yes. Those were the most remarkable cases, because those tumors tend to be large at presentation and notoriously difficult to detect because they don’t often form the mass that an invasive ductal carcinoma does."
Press briefing moderator Dr. Debra L. Somers Copit, chief of mammography and director of the Gershon-Cohen Breast Clinic at Einstein Medical Center, Philadelphia, described the improvements in cancer detection in the real-world cohort as "groundbreaking work for this modality that we hope will pan out across multiple institutions."
Dr. Debra L. Somers Copit
Reimbursement is problematic, however, as insurers will only pay for 2D mammography images taken as part of a tomosynthesis screening exam, she said in an interview.
Still, she personally believes tomosynthesis should now be the standard of care, adding, "I can’t imagine reading a mammogram without it."
Results from the current study are comparable with data reported recently from trials in the United States (Radiology 2013;269:694-700) and Norway (Eur. Radiol. 2013;23:2061-71) that paired 2D imaging with 3D tomosynthesis, Dr. Conant observed.
When the investigators looked at the independent risk factor of breast density, DBT also did a better job of detecting cancer than conventional mammography in fatty, scattered, heterogeneous, and extremely dense breasts, and it improved recall rates for all categories except extremely dense breasts, she said.
Overall, DBT significantly reduced the proportion of women recalled for additional imaging from 10.39% to 8.77% (P = .001).
DBT is an exciting improvement over 2D mammography and more economical than breast screening with magnetic resonance imaging, but it is "not the solution to everything," Dr. Conant said. The radiation dose for the average breast is within safety limits, but twice that of a regular mammogram.
"The cost to the patient is not monetary at our site; it may be at other sites," she said. "I think right now the dose is the biggest cost."
The May 2013 approval of Hologic’s C-View 2D imaging software, which eliminates the need for additional 2D exposures by generating 2D images from 3D tomosynthesis data, might address this, Dr. Conant added.
Dr. Conant reported consulting for Hologic. Her coauthors reported no financial disclosures. Dr. Copit is on Hologic’s scientific advisory board.