Five years ago, Connecticut passed the first U.S. law mandating that medical providers tell women if their mammogram demonstrates they have dense breast tissue. Up to half of all women (about 40-50% by some estimates) have dense breasts, which can limit the efficacy of mammography screening by “masking” abnormalities. Currently, 19 states have enacted breast density notification laws. (You can read what we’ve written on this subject here).
The challenge with these “dense breast bills” is that there’s no evidence-based action to take for those women who are notified of their breast density. In California, the breast density notification law advises women to “decide which options are right for you.” However, there is no evidence with which to make this decision. Should women undergo additional screening if they have dense breasts? What are the harms and benefits? Has any imaging tool been shown to benefit the health of women with dense breasts and do we know if supplemental screening saves women’s lives?
Breast density is known to increase the risk of breast cancer and many doctors hold that screening prevents death (see BCAction’s screening brochure for evidence-based information about the benefits, limitations and harms of routine screening for average risk women). Four states mandate that insurance companies cover additional screening tests for women with dense breasts and federal legislation has been proposed to this effect as well. But is there evidence that additional screening saves women’s lives?
Dr. Jean Weigert is a radiologist at the Bradley Memorial Hospital & Health Center in Connecticut and presented “The Connecticut Experiment: 4 years of screening women with dense breast with bilateral ultrasound” on December 12, 2014 at the San Antonio Breast Cancer Symposium (SABCS). The study consisted of a retrospective review of the medical records of women from two medical centers in Connecticut between 2009 and 2013 in order to look at the number of screening ultrasounds performed after the breast density law was implemented and resulting cancer detection.
After the dense breast bill went into effect in Connecticut, women with dense breasts were offered an ultrasound a mammogram even if there was no finding on their mammogram. About 30% of eligible patients came in for the additional test, which is covered by insurance in Connecticut.
Each year, approximately 30,000 mammograms and 3,000 ultrasounds were performed in these two medical centers. The ultrasound screening found between three and four breast cancers per 1,000 women screened, resulting in around 11 diagnoses over the four years. However, many women underwent a biopsy from a false positive ultrasound, which has a lower positive predictive value (the likelihood that a positive test actually indicates disease) than mammography, one reason in addition to expense that it’s not widely used as a screening tool.
In the first three years of the study, only 6-8% of biopsies recommended after an ultrasound screening actually found cancer, meaning the vast majority of women (well over 90%) who were sent for a biopsy after an ultrasound endured an unnecessary invasive test and all the associated harms. By the fourth year of the study, fewer women were sent for a biopsy after their ultrasound and the positive predictive value improved to 17.2%, which still means that more than 80% of biopsies were unnecessary. In her subsequent talk (reviewed below), Dr. Jafi Lipson notes that while some might argue that doing fewer biopsies is a benefit, the reality is that many of these women are now in a “vortex of following” as they are monitored and observed. Rather than a reduction of harm, this represents merely a shift in harm from invasive procedure to observation.
Dr. Weigert over-sells the benefits of ultrasound. She concludes by suggesting ultrasound imaging is appropriate for women with a family history of breast cancer and other risk factors as well as women with dense breasts. However, this is a short-term study with no control and no long-term follow up. No mortality statistics were presented and there was no cost analysis (although in her consistent optimism, Dr. Weigert suggests she believes the cost of ultrasound is not much higher). Without additional data, including reduced mortality, the true clinical impact is unknown and no claims can be made about the benefit of ultrasound screening for women with dense breasts.
As mentioned above, Dr. Jafi Lipson from Stanford was the Discussant immediately following Dr. Weigert’s presentation. Her comments focused on the benefit to harm ratio of ultrasound and tomosynthesis or 3-D mammography, but not MRI. Benefit to harm ratio depends on risk of breast cancer and efficacy of screening tool.
Among the limits and harms of ultrasound in Dr. Weiert’s study discussed by Dr. Lipson are the number of biopsies that women undergo (“many more than from screening mammography”), the fact that most are false positives, and the lack of cost effectiveness. “Would the cancers found through ultrasound be found at the next mammogram while still small?” she asks. “Does ultrasound impact mortality?”
Dr. Lipson then contrasted ultrasound with tomosynthesis, using data from two prospective and two observational studies. While Dr. Lipson touted the 15-30% decrease in false positives and 30-50% increase in cancer detection with tomosynthesis, as with ultrasound there is no longer term follow-up and no mortality data.
There are no studies comparing tomosythesis to ultrasound. While both modalities find slightly more cancers than mammography, they both come with the harms of false positive. Dr. Lipson’s argument in favor of tomosynthesis is that it has 30-40% fewer false positives than ultrasound.
At the end of the day, there is not enough data to suggest that urging women with dense breasts to undergo supplemental screening is beneficial. Routine screening of healthy, average-risk women is not supported by evidence that such screening saves women’s lives and there is no evidence that the data is any different for women with dense breasts. It is time to question the evidence on screening as a primary strategy to reduce death from breast cancer. No screening tool can prevent breast cancer. And when it comes to saving lives, we know that what makes a difference to outcomes is access to evidence-based, affordable, high-quality and culturally competent healthcare.