Digital breast tomosynthesis (DBT), touted as a major technological advance over the traditional digital mammogram (DM) it replaced, does not offer significant performance advantages or outcomes improvements for women undergoing breast cancer screening, a surprising new study from Houston Methodist finds.

In an analysis of more than 12,000 breast scans deemed suspicious (categorized by radiologists as BI-RADS 4), the researchers found comparable biopsy outcomes, malignancy rates, cancer detection rates and biopsy-derived positive predictive values for both those performed with DM and those performed with DBT, also known as 3D mammography.

"Although earlier studies have reported that digital breast tomosynthesis can detect up to 40% more cancers than DM, particularly in certain patient groups, our findings indicate that performance and sensitivity in assigning BI-RADS 4 cases — and ultimately making a diagnosis of breast cancer — are similar regardless of mammogram modality, tumor characteristic, density of breast tissue or other demographic factors," says Stephen Wong, Ph.D., the John S. Dunn Presidential Distinguished Chair in Biomedical Engineering at Houston Methodist Hospital and the study's principal investigator.

International implications

This study, recently published in the European Journal of Radiology, is likely to have the largest impact on those who currently use traditional 2-D mammograms. While health care organizations in the U.S. have already switched over to DBT as the standard of care, a large number of countries still employ traditional DM.

"The important thing is that recommended groups get a mammogram — not get a mammogram of a certain type," says Wong, who also serves as a Professor of Radiology, Pathology, Laboratory Medicine, Neurology and Neurosciences. "It should hopefully lift a weight off the health care systems' budgetary decisions — knowing that their existing technology performs equally well in this crucial cohort."

Since mammogram screenings became mainstream in the 1970s and 1980s, the overall breast cancer death rate has declined by more than 40%. The reduction in breast cancer mortality has been attributed to better, more targeted treatments and earlier detection, when the chances of long-term survival are highest.

Categorizing risk

Among the BI-RADS categories, BI-RADS 4 (suspicious findings with a recommendation for biopsy) stands out for its high degree of uncertainty, with a likelihood of malignancy that ranges from 2% to 95%. By contrast, probability of malignancy is less than 2% in BI-RADS 3 and greater than 95% in BI-RADS 5.

In the Houston Methodist study, approximately 15% of patients designated as BI-RADS 4 went on to receive a diagnosis of breast cancer, with no statistically significant differences between DM and DBT groups. There were 6,356 scans in the DM group and 5,896 in the DBT group.

""One of the early selling points for 3D mammograms was their potential to reduce false positive results," says Wong. "But subsequent studies, including this one, have not necessarily shown this to be the case."

For that reason, he notes, it remains important for patients in this category to receive needle biopsies to rule out the possibility — albeit small — of a malignant case.

Caveats

Wong emphasizes that the results of the Houston Methodist study apply specifically to BI-RADS 4. Among other groups, DBT may exhibit advantages over DM, such as providing clearer, more detailed images of breast masses, asymmetries and other anomalies than 2D mammography. 

But Wong notes that the study provides an important takeaway: a reminder that current screening technologies have their limitations.

"This study brings to the fore the pressing need to identify better tools to differentiate risk, reduce unnecessary biopsies and mitigate against the associated health care costs and anxieties they can cause," says Wong.