SAN ANTONIO — Screening ultrasound in addition to mammography should be routine for women with dense breast tissue, according to a study from Connecticut, where the first breast density notification law was passed.
The positive predictive value of screening ultrasound rose significantly during the first 4 years the legislation was in place.
At the same time, there was a consistent cancer detection rate of 3.2 to 4.0 cases per 1000 women screened, reported lead investigator Jean Weigert, MD, a clinical radiologist and director of breast imaging at The Hospital of Central Connecticut in New Britain.
“Cancers are being found in women who return for yearly screening breast ultrasound, indicating that this test should be part of the routine screening workup in women with dense breasts,” Dr Weigert told Medscape Medical News before she presented her data here at the San Antonio Breast Cancer Symposium 2014.
“Perhaps it is time to think of a new paradigm — of utilizing screening ultrasound as part of the routine screening” for women with dense breasts, Dr Weigert said during her presentation. “Patients should have both a mammogram and ultrasound as part of their regimen.”
In their retrospective analysis, Dr Weigert and her colleagues evaluated data from two radiology practices in Connecticut from October 2009 to December 2013.
Approximately 30,000 screening mammograms were performed in each of the 4 years. In addition, there were 2706 screening ultrasounds for dense breasts in the first year, 3351 in the second year, 4128 in the third year, and 3331 in the fourth year.
The ultrasound screens detected 11 breast cancers in the first, second, and fourth years, and 13 in the third year.
Although the number of detected cases stayed steady during the 4 years, there was decrease in the rate of ultrasounds falsely classified, with the Breast Imaging Reporting Data System (BI-RADS), as being a “suspicious abnormality – biopsy should be considered” (code 4) or “highly suggestive of malignancy – appropriate action should be taken” (code 5).
Table. Cases Suspicious or Suggestive of Malignancy
Study Year BI-RADS Code 4 and 5, n Positive Predictive Value, %
First 151 7.1
Second 180 6.1
Third 178 6.1
Fourth 53 17.2
“Over the 4 years, we have learned how to better identify those breast ultrasound abnormalities that may truly be cancer,” Dr Weigert said in a statement. “The positive predictive value for mammography is about 20% to 30%. We are getting close to this with breast ultrasound now that we are more experienced.”
The cancers found during the study period were “all types in terms of histologic grade, were predominantly estrogen-receptor- and progesterone-receptor-positive, had very little sentinel lymph node positivity, had no axillary metastasis, and had very few risk factors other than dense breast tissue,” Dr Weigert reported.
There was also a progressive decrease in the size of lesions identified, with the size averaging less than 1 cm.
“This is the same type of cancer that we would want to find in screening mammography,” she said. “But we couldn’t see it on the screening mammogram; it took the ultrasound for us to find these cancers.”
It is concerning that although 40% to 50% of women are classified as having dense breast tissue, only 30% of eligible women take advantage of breast ultrasound screening, Dr Weigert noted.
She suggested this could be due to a lack of education, but might also be related to cost. Although Connecticut has legislated insurance coverage for breast ultrasound, patients still shoulder a large proportion of the cost.
In a recent study (Ann Intern Med. Published online December 9, 2014), it was demonstrated that supplemental ultrasound screening for women with dense breasts is not cost-effective; in fact, it substantially increasing costs while producing relatively few benefits, as reported by Medscape Medical News.
“I have seen the study and I find it of concern,” Dr Weigert said.
“It is true that my research did not include a cost analysis, but that is certainly something we need to be aware of,” she told Medscape Medical News. “However, finding small node-negative cancers should impact cost in a positive way. I would certainly like to have a discussion about that in the future.”
But money is not the only cost associated with breast ultrasound screening, said study discussant Jafi Lipson, MD, assistant professor of radiology in the Department of Breast Imaging at Stanford University School of Medicine in California.
The other cost of screening breast ultrasound is the high false-positive rate. “Many more biopsies are generated by screening ultrasound than by screening mammography, and most end up being false positives,” Dr Lipson pointed out.
“The positive biopsy rate from screening mammography ranges from 20% to 40%, while on screening breast ultrasound, it generally ranges from 5% to 10%,” she said.
And there are other questions raised by the currently available data on breast ultrasound, she noted.
“None of the studies have been performed with control groups and there has been no long-term follow-up, so the true clinical impact of finding these additional cancers is unknown,” Dr Lipson explained. “Specifically, would these additional cancers be detected at the next mammography screen, while still small, node-negative, and at an early stage? And does the detection of these cancers have an impact on mortality?”
Although breast density is unquestionably a problem, in that it masks the mammographic detection of lesions, perhaps ultrasound is not the best solution, Dr Lipson said.
For example, consider breast tomosynthesis, a 3D digital mammography technique.
All the current data show that the addition of breast tomosynthesis to standard mammography screening “is associated with a 15% to 30% decrease in false positives and a 30% to 50% increase in cancer detection.”
Both breast screening ultrasound and screening tomosynthesis, “have incremental cancer detection as a benefit, and they’re about similar,” she said. “But false positives are much greater for ultrasound (about 4 times greater than mammography) and much less for tomosynthesis (about 30% to 40% less than mammography).”
“To date, 20 states have enacted breast density legislation, and more are likely to do the same,” said Murray Rebner, MD, professor of diagnostic radiology at the Oakland University William Beaumont School of Medicine in Rochester, Michigan, and president of the Society of Breast Imaging.
“This study shows that, over time, it may be possible for radiologists to reduce their false positives and maintain the same cancer detection rate,” he told Medscape Medical News. “Hopefully, these results will be reproducible throughout all communities and all radiology practices.”
Dr Weigert and Dr Rebner have disclosed no relevant financial relationships.
San Antonio Breast Cancer Symposium (SABCS) 2014: Abstract S5-01. Presented December 12, 2014.