Ultrasound bested tomosynthesis for screening dense breast tissue

Key clinical point: Ultrasound outperformed tomosynthesis for the incremental detection of breast cancer in women with negative mammograms and radiologically dense breast tissue.

Major finding: Ultrasound detected about 7.1 additional cancers for every 1,000 screens (95% confidence interval, 4.2-10), compared with 4.0 additional cancers per 1,000 tomosynthesis screens (95% CI, 1.8-6.2; P = .006).

Data source: A multicenter, prospective trial comparing the two modalities in 3,231 women.

Disclosures: The University of Genoa sponsors the ASTOUND study. Dr. Tagliafico disclosed honoraria from Esaote-Philips, patents, royalties, or other intellectual property from Springer, and travel, accommodations, and expense support from Hologic and Technologic. Four coinvestigators reported financial relationships with several device and pharmaceutical companies. The senior author and the other seven coinvestigators had no disclosures.

View on the News

Results support ultrasound over tomosynthesis

Currently, 24 American states have laws requiring that women receive some level of notification about breast density with their mammography results. Dense breast tissue can hide cancer on mammography, especially when the cancer lacks calcifications, resulting in delayed diagnosis and worse outcomes. Moreover, dense breast tissue is an independent risk factor for developing breast cancer.

Because the primary goal of screening is detection of early breast cancer, ultrasound would seem the clear choice, compared with tomosynthesis. Given comparable false-positive rates in ASTOUND, the estimated cost per cancer detected would be similar or more favorable for ultrasound than tomosynthesis. Ultrasound equipment is becoming much less expensive, requires no ionizing radiation, and it is easy to guide needle biopsy of lesions seen only on ultrasound.

Preliminary results from ASTOUND are extremely important in helping to inform personalized screening choices for women with dense breasts. Guidelines on these issues are planned, but often limit recommendations to those based on evidence from randomized trials with mortality as an end point. Our knowledge of the natural history of breast cancer and results from randomized trials of mammography should inform guidelines for supplemental screening.

Dr. Wendie A. Berg is at Magee-Womens Hospital of University of Pittsburgh Medical Center. She reported serving in a consulting or advisory role with SuperSonic Imagine. These comments were taken from her accompanying editorial (J Clin Oncol. 2016 Mar 9. doi: 10.1200/JCO.2015.65.8674).




Ultrasound was about 1.8 times more sensitive than tomosynthesis for the incremental detection of breast cancer in women with radiologically dense breasts and negative two-dimensional mammography screening, according to interim results from the first prospective trial to directly compare the two modalities.

“However, future application of adjunct screening should consider that tomosynthesis detected more than 50% of the additional breast cancers in these women, and could potentially be [a] primary screening modality,” wrote Dr. Alberto Tagliafico of the University of Genoa (Italy) and his associates. The study was published online March 9 in the Journal of Clinical Oncology and presented simultaneously at the European Breast Cancer Conference.

Radiologically dense breast tissue undermines the sensitivity of mammography and is itself an independent risk factor for breast cancer. Recently, many states began requiring that women be informed of their breast density and adjunct screening measures, such as ultrasound. But estimates of the sensitivity of ultrasound have ranged from about 1.9 to 4.2 cancers for every 1,000 screens, said the researchers. This variance, combined with costs and concerns about false-positive recalls, have fueled debates about the value of adjunct measures in breast cancer screening, they added. To help clarify these issues, the multicenter ASTOUND (Adjunct Screening With Tomosynthesis or Ultrasound in Women With Mammography-Negative Dense Breasts) study compared physician-performed ultrasound and tomosynthesis results for 3,231 asymptomatic women aged 44 to 71 years, whose median age was 51 years (J Clin Oncol. 2016 Mar 9. doi: 10.1200/JCO.2015.63.4147).

In all, the researchers detected 24 additional breast cancers, 23 of which were invasive. Thus, ultrasound detected about 7.1 additional cancers for every 1,000 screens (95% confidence interval, 4.2-10), compared with 4.0 additional cancers per 1,000 screens for tomosynthesis (95% CI, 1.8-6.2; P = .006). Only one cancer was detected by tomosynthesis alone. The rate of false-positive recalls was similar for the two modalities – 53 cases for tomosynthesis, versus 63 for ultrasound (P = .26). Rates of false-positive recalls leading to biopsy also were similar. Needle biopsies usually sufficed in recalled cases, but two women underwent surgical biopsies, both of which revealed radial scars.

If the final results of ASTOUND confirm these interim data, “it could be argued that breast tomosynthesis has little value in a setting where adjunct ultrasound is frequently used for screening women with mammography-dense breasts,” said the researchers. But tomosynthesis may have a role as a primary screening modality in other setting, especially because tomosynthesis acquisitions that also provide reconstructed 2D mammography are now available, lessening concerns about unjustified radiation exposure, they added.

The “modest” number of cancers in the interim report led to relatively wide confidence intervals, the investigators noted. Biomarker data were not available for all cancers, and both prevalent and incident ultrasound data were compared with prevalent tomosynthesis data, which might bias false-positive recall results in favor of ultrasound, they added.

Next Article:

   Comments ()