Overall, screening with digital breast tomosynthesis (DBT) did not appear to detect more interval or advanced breast cancers compared with digital mammography, a comparative study suggests.
More specifically, for the 96.4% of women with nondense breasts, heterogeneously dense breasts, or extremely dense breasts who were not at high risk of breast cancer, the two screening modalities performed similarly.
But for a small proportion of women at high risk who had extremely dense breasts, the rates of advanced cancer were significantly lower with DBT than with digital mammography.
Overall, the findings indicate that most women “could undergo either screening test and have the same potential benefit of averting a death from breast cancer,” Karla Kerlikowske, MD, San Francisco Veterans Affairs Medical Center, California, told Medscape Medical News via email. The exception, she noted, is the 3.6% of women with extremely dense breasts and higher-than-average breast cancer risk.
The study was published online June 14 in JAMA.
The goal of DBT has been to improve cancer detection and avert deaths among women with dense breasts by decreasing the rates of missed, or interval, cancers. But the evidence to date has not consistently found that DBT outperforms digital mammography.
In the current study, Kerlikowske and colleagues evaluated whether one screening modality was associated with a lower risk of interval invasive and advanced breast cancer among women with dense breasts.
The researchers obtained data from five Breast Cancer Surveillance Consortium mammography registries, which included 504,427 women. The median age at the time of mammography was 58 years. Between January 2011 and December 2018, 308,141 women underwent only digital mammography, 56,939 underwent only DBT mammography, and 139,347 underwent both digital and DBT mammography.
Overall rates for detecting interval invasive cancer were 0.57 per 1000 examinations for DBT and 0.61 per 1000 examinations for digital mammography. The difference of 0.4 per 1000 exams was not significant (P = .43).
Although DBT detected significantly fewer advanced cancers overall ― 0.36 per 1000 examinations compared with 0.45 per 1000 for digital mammography ― when stratified by breast density and cancer risk, advanced cancer rates appeared significantly lower only for women with extremely dense breasts who were at high risk (0.27 vs 0.80), not for those at low to average risk (0.54 vs 0.42).
Screen-detected stage I cancer rates were significantly higher with DBT compared with digital mammography for women with scattered fibroglandular densities and for women with a high breast cancer risk (a 5-year risk of ≥1.67%).
The authors also found that the false positive recall rate for DBT was much lower than for digital mammography — 66.2 per 1000 examinations vs 83.4 per 1000 — as was the false positive short-interval follow-up recommendation — 11.2 per 1000 examinations for DBT vs 17.9 per 1000 for digital mammography.
A downside to DBT, Kerlikowske noted, is that “insurance companies don’t all reimburse for DBT, so women may not think it worth the out-of-pocket cost to pay for DBT when there is no added benefit of averting a death from breast cancer compared to digital mammography.”
In an accompanying editorial, two experts provide a slightly different interpretation for the findings.
Had the interval cancer rate had been lower among patients who underwent screening with DBT than among those who underwent digital mammography, this would have suggested that DBT detects more aggressive cancers earlier, which theoretically would result in a decrease in breast cancer mortality, argue Sarah Friedewald, MD, Northwestern University Feinberg School of Medicine, Chicago, Illinois, and Lars Grimm, MD, Duke University School of Medicine, Durham, North Carolina.
Given that DBT did not reduce the rate of interval cancers compared with digital mammography, the editorialists suggest that the interval cancer rate could still be reduced with improvements in screening technology or changes to the screening interval. By using multiple screening modalities — for instance, combining DBT and MRI — more cancers could be identified than with DBT alone, and that could potentially reduce interval cancer rates.
In a study involving 148,575 women who were screened with digital mammography, the interval cancer rate was 0.89/1000 for women who were screened annually, vs 1.45/1000 for women who were screened biennially.
One problem with an analysis based on registry data is that the demographic representation is limited, but Black women are more likely to experience delays in diagnosis, present with advanced disease, and die from breast cancer, the editorialists write.
Given the fact that DBT is available at more than 80% of mammography facilities in the US and given its ability to detect advanced breast cancer, DBT “may be an important factor to help achieve health equity for breast cancer screening,” Friedewald and Grimm write.
The study was funded by the Patient-Centered Outcomes Research Institute and the National Cancer Institute. Kerlikowske has received grants from the National Cancer Institute. Friedewald has consulted for Hologic and has received grants from Google. Grimm has consulted for Hologic and serves on a Medscape advisory board.