Adding Tomosynthesis to Contrast-Enhanced Mammography Reduces False Positives 

Contrast-enhanced mammography (CEM) has emerged as an important tool for improving breast cancer detection, but questions remain about whether digital breast tomosynthesis (DBT) should be added to the exam for screening. A new study presented Tuesday shows that adding DBT can significantly reduce false-positive recalls while maintaining the high sensitivity achieved with CEM. 

For Wendie A. Berg, MD, PhD, professor of radiology at University of Pittsburgh School of Medicine (UPMC) and UPMC Magee-Women’s Hospital, the question was driven by day-to-day practice. 

“We have been using DBT routinely for screening since 2011, and we know that it improves performance compared to 2D mammography—both for detecting cancers and reducing false alarms,” she said. “When we started doing CEM, we did not want to lose those benefits.”

In clinical settings, radiologists inevitably see all available images, making it nearly impossible to isolate the value of DBT after CEM. Dr. Berg’s team designed a controlled reader study using software that locked radiologists’ interpretations step by step—first on low-energy (LE) images, then on recombined contrast-enhanced (RC) images and finally, DBT.

CEM Greatly Improves Sensitivity—DBT Helps Reduce False Positives

Across 403 enriched screening cases, five breast imagers recorded a substantial jump in cancer detection when RC images were added. Sensitivity rose from 44% on LE images alone to 79% with CEM.

“The cancers that we find because of contrast are mostly small invasive cancers, averaging 6–7 millimeters, and mostly found before they spread to lymph nodes,” Dr. Berg said. “These are exactly the cancers we want to find.”

Adding DBT did not raise sensitivity further. According to Dr. Berg, that is expected: noncalcified cancers that might be more visible on DBT typically enhance with contrast, while ductal carcinoma in situ (DCIS) that may not enhance is often identifiable on LE images through calcifications.

Where DBT made its impact was in reducing false positives. False-positive recalls increased when RC images were viewed, but DBT lowered them by 5%, helping radiologists dismiss benign findings without missing cancers.

“False alarms cause stress for patients and add cost, and it is always desirable to minimize them, as long as we don’t sacrifice cancer detection,” Dr. Berg said.

She noted that the reduction could be even greater in routine practice, where prior DBT exams often show stable benign findings.

Why DBT Still Matters With CEM

CEM highlights contrast uptake linked to tumor vascularity, while DBT provides thin tomographic slices that clarify structure. Dr. Berg emphasized that these anatomical details remain valuable. “DBT can help us recognize (and dismiss) normal lymph nodes, bends in blood vessels and skin lesions that can enhance on CEM,” she said. “DBT also shows mass margins better than we can see them on CEM.”

She also pointed to dose considerations. Adding DBT nearly doubles radiation compared with CEM alone, although exposure remains within accepted mammography limits. Extra caution is warranted in younger women and those with radiation-sensitive mutations such as BRCA or TP53.

Dr. Berg said the findings can help guide departments planning to adopt CEM. “Radiologists are very comfortable interpreting DBT and it improves our confidence,” she said. “When considering implementation strategies to add CEM to one’s practice, it is useful to know that there is a benefit to including DBT as part of the CEM examination.”

Access the presentation, “Adding Tomosynthesis to Contrast-Enhanced Mammography Reduces False Positives,” (T7-SSBR06-3) on demand at RSNA.org/MeetingCentral.