Last week, the United States’ Food and Drug Administration (FDA) agency proposed a rule that would inform women of their breast density. Volpara applauds and support this effort to “help empower patients with more information when they are considering important decisions regarding their breast health care.”
The Good News
Women with dense breasts living in one of the states not requiring density notification will finally have access to the information they need to better understand their risks. First, they will be aware of their “masking risk”—fibroglandular tissue in dense breasts can “mask” (hide) tumors, making it more difficult for radiologists to detect cancer. Second, they will understand that women with dense breasts are also at a higher risk for developing breast cancer when compared to women with fatty breasts.
Under the new FDA rule, each woman in the United States who goes for breast cancer screening will be given information about her breast density, so she can have a more informed conversation with her primary care provider about worthwhile options such as supplemental screening or preventative treatment.
The Less Good News
While Volpara celebrates this effort, we also support it with some caution. Why? Because Volpara feels very strongly that if radiologists will be required to tell women their breast density, then those radiologists should be equipped to provide a more accurate assessment.
What is the problem? There are three issues at hand:
1) Subjectivity. Even the American College of Radiology has pointed to the fact that “the assessment of breast density is subjective and, therefore, variable,”[1] and a vast number of clinical publications agree. In other words, density assessment relies on human judgment and therefore has its limitations. Studies have shown that not only do radiologists often disagree when scoring breast density, but they often disagree with themselves when seeing the same mammogram some time later. Inter-reader agreement (between radiologists) can be as low as 57% while intra-reader agreement (with themselves) can be as low as 77%.
2) Image Processing. After the x-ray is taken, each mammogram is “processed,” which makes the image more “human viewable”. But there is no agreed-upon standard as to what that processing should be, so each equipment manufacturer provides one or more different types or settings of processing. The result is that the images from different systems, or version of systems, can appear different . . . and that has been shown to affect the radiologists’ perception of breast density.
3) Imaging Mode. Many x-ray machines used for mammography today have several different imaging modes. The latest models can make traditional “projection” x-rays, or create a tomosynthesis x-ray scan from which “tomosynthesis slices,” “tomosynthesis slabs,” and “synthetic images” can be created. To make things even more complicated, these derived images can be created in many combinations, and each manufacturer applies different image processing to each variant. This means that, depending on the mode of operation used, the radiologist may see a big variation in image appearance.
A recent study demonstrated that women screened with Digital Breast Tomosynthesis (DBT), or “3D mammography,” were categorized as having lower breast density when compared to women screened with traditional digital mammography. In addition, how the DBT images were processed affected the perception of breast density; doctors were more likely to report a lower breast density category when they read a mammogram consisting of DBT with synthetic images than when they read a mammogram consisting of DBT with traditional digital mammography images.
In the proposed language, the FDA states that “the proposed changes would . . . establish four categories for reporting breast tissue density in the mammography report that is provided to the patient’s referring healthcare provider.” Sadly, this perpetuates the four-category system used by the American College of Radiology. As described above, the assessment is subjective and suffers from image processing and imaging mode variations, making it less likely that women are steered to pathways that might benefit their health based on breast density.
There Is Hope
Even today, tools exist that deliver objective, consistent, and repeatable breast density measures and are not subject to the same sources of error. Volpara®Density™ software automatically assesses breast density for digital mammography and DBT—but instead of looking at the highly processed images that doctors view, it looks at the real image acquired by the x-ray machine. Best of all, it has been validated in several hundred clinical studies, more than 100 of which were peer-reviewed prior to publication.
With software that overcomes the challenges posed by subjectivity, image processing, and imaging mode, radiologists will be able to fulfill the promise of the FDA’s proposed requirement, notifying women of their breast density based on the most accurate assessment available.
We encourage you to ask your doctor if your mammography provider uses VolparaDensity.
[1] American College of Radiology. ACR Statement on Reporting Breast Density in Mammography Reports and Patient Summaries; 2017
