What a New Study Reveals About Whole-Breast Ultrasound Tomography, Mammography and Dense Breasts

For women with dense breasts, new study findings show the adjunctive use of whole-breast ultrasound tomography (UST) provides better breast cancer detection than mammography alone.

For the retrospective multicenter study, recently published in Radiology, researchers reviewed full-field digital mammography (FFDM) and UST data for 140 women (mean age of 56) with dense breasts. According to the study, 36 women in the cohort had breast cancer. The study authors noted that the 32 reviewing radiologists were Mammography Quality Standards Act-qualified radiologists with breast imaging experience ranging between two and 37 years.

The researchers found that that the combination of FFDM and UST yielded an area under the curve (AUC) of 60 percent for breast cancer detection in comparison to 54 percent for FFDM alone. Whole-breast ultrasound tomography does offer a couple of key advantages in breast cancer screening, according to the study authors.

“First, coronal volumetric image sequences are useful to identify and characterize lesions. They provide a better view of the fat-glandular interface where most breast cancers are located. Second, the stiffness fusion sequence provides tissue stiffness information, which can help differentiate cancer from benign masses and is not readily available with handheld US or automated breast US,” wrote lead author Mary W. Yamashita, M.D., a clinical professor of radiology and associate section chief of breast imaging at the Keck School of Medicine at the University of Southern California (USC), and colleagues.

For Breast Imaging Reporting and Data System (BI-RADS) 4 cases, the study authors noted that in contrast to FFDM alone, FFDM/UST had non-inferior specificity (82 percent vs. 88 percent) but offered superior sensitivity (37 percent vs. 30 percent). However, in these cases, the researchers pointed out a net increase of 5.5 more false positive findings per reader.

“(The) addition of UST to FFDM resulted in only a slight reduction in positive predictive value compared with FFDM alone (42.2% vs 45.4%, respectively),” cautioned Yamashita and colleagues. “Clinical studies are needed to determine whether supplemental UST can improve cancer detection without substantially increasing the number of benign biopsies.”

While the researchers saw no statistically significant difference with mean sensitivity in BI-RADS 3 cases (40 percent for FFDM/UST vs. 33 percent for FFDM alone), they found that FFDM/UST provided superior mean specificity (75 percent vs. 69 percent).

“The specificity improvement with FFDM plus UST at BI-RADS 3 assessment is encouraging. A BI-RADS 3 assessment is the most common source of false-positive findings in handheld US and automated breast us screening,” noted Yamashita and colleagues. “With improved BI-RADS 3 specificity, radiologists can more accurately characterize masses as benign (BI-RADS 2), leading to fewer BI-RADS 3 assessments.”

In an accompanying editorial, Ritse Mann, M.D., Ph.D, said the increased sensitivity with UST for BI-RADS 4 cases is in line with previous research looking at supplemental ultrasound but expressed reservation about the “poor overall sensitivity” and likelihood of increased false-positive findings.

“In my opinion, this study shows that the current implementation for UST is at best on par with handheld US and automated breast US, allowing detection of some additional cancers at the price of a substantial increase in the number of false-positive findings. It seems that UST may be a new implementation of US, but it ultimately yields old results,” maintained Dr. Mann, the chair of breast imaging at the Netherlands Cancer Institute in Amsterdam and chair of the scientific and program committees of the European Society of Breast Imaging.

(Editor’s note: For related content, see “Leading Breast Radiologists Discuss the USPSTF Breast Cancer Screening Recommendations,” “Whole Breast Ultrasound Screening: Is There Adequate Utilization in Patients at Higher Risk for Breast Cancer?” and “Breast Ultrasound Study: AI Radiomics Model May Help Predict Lymphovascular Invasion in Breast Cancer Cases.”

In regard to study limitations, the authors conceded that reviewing radiologists received abbreviated clinical histories, had no prior imaging for patients and interpreted images in a laboratory setting for the study. The researchers also pointed out that the reviewing radiologists had no prior experience with interpreting images from the breast ultrasound tomography (UST) system utilized in the study.