Along with shedding new light on the issue, two recent Radiology studies are generating debate on the effectiveness of supplemental ultrasound and MR imaging in breast cancer detection in patients at elevated risk for the disease—including those with dense breasts.
In the first study, investigators discovered that supplemental handheld screening breast ultrasound offered to women in the general population with dense breasts aided in detecting small mammographically occult breast cancers, although the overall positive predictive value was low. The other study demonstrated high accuracy achieved by a dedicated breast MR imaging system. Both studies appeared in the October 2012 issue of Radiology.
Authored by Regina Hooley, M.D., an assistant professor of diagnostic radiology at the Yale University School of Medicine, New Haven, Conn., and colleagues, the first study examined the effectiveness of a 2009 Connecticut law requiring providers to alert women with dense breasts and offer supplemental ultrasound screening.
Connecticut is one of five states including Texas, Virginia, New York and California that now have mandatory breast density notification laws. Similar bills are under consideration in many other states. In addition, a federal Breast Density Inform bill (HR 3102) has been introduced. Opponents of the mandates say the role of supplemental ultrasound as a screening modality is unproven and additional screening doesn’t improve detection enough to justify the cost or the risk of false positives that lead to unnecessary biopsies.
Experts also stress that patients must be aware that ultrasound screening is meant only as supplement to mammograms—not a replacement.
Conducted at Yale, the study retrospectively reviewed the records of 935 women with dense breasts at mammography who subsequently underwent supplemental handheld ultrasound screening and/or whole-breast ultrasound. A majority (65.7 percent) were at low risk, 15.9 percent were at intermediate risk and 9.3 percent were at high risk for breast cancer. The ultrasound examinations found 187 (20 percent) results classified as BI-RADS category 3 and 47 (5 percent) classified as BI-RADS category 4. Out of 63 biopsies and aspirations performed based on the ultrasound exam results, three were malignant, one in each risk group. All three were solid masses, smaller than 1 cm, in postmenopausal patients.
“This study shows that radiologists can provide supplemental breast screening successfully,” Dr. Hooley said. “We had a lot of patients who wanted this test, although we didn’t offer it before the law went into effect. We found we could provide it efficiently by allowing technologists to perform exams and this proved comparable to physician-performed exams. We can detect small invasive mammographically occult cancers.”
Dr. Hooley said some physicians refer only their high-risk patients—those who have risk factors in addition to dense breast tissue—for additional screening, while others routinely refer all of their patients who have dense breasts. “We’re doing a study now to trying to determine physician referral patterns,” she says.
Most women in the screening study were of average risk, and Dr. Hooley said the results could be different if only high-risk patients were studied.
Dr. Hooley’s research, originally published online in Radiology in June 2012, is generating controversy among breast imaging experts who contend that the data do not represent the established definition of “screening” and that the role density plays in the risk of breast cancer is still unclear.
In an editorial in the October 2012 issue of Radiology, Carl J. D’Orsi, M.D., and Edward A. Sickles, M.D., stress that universally accepted definitions of “screening” and “diagnostic” for breast imaging examinations are critical to ensuring valid comparisons of breast imaging technologies and assessing whether their greatest impact would be in a screening or diagnostic setting or both.
The authors also discuss the potential impact of Dr. Hooley’s research on the legal mandates concerning breast density that increasing number of states are poised to pass into law. “The lack of separation between screening and diagnostic examinations in the report of Hooley et. al, has considerable relevance as a potential indicator of future benchmark performance, given that the reported data skew the outcomes expected from true screening ultrasound to falsely appear more favorable than they really are,” the authors write. “These data may be used to affect the decisions of lawmakers who are considering similar legislation in other states. In our opinion, such government mandates are premature.”
In addition, experts caution that the false positive rate associated with supplemental ultrasound screening is exceptionally high, so much so as to limit its acceptability.
The second study evaluating the effectiveness of a dedicated breast MR system analyzed patients without regard to breast tissue density, although patients in all screening cases had risk factors for breast cancer that included breast tissue density.
Lead researcher Steven Harms, M.D., a radiologist at the Breast Center of Northwest Arkansas in Fayetteville, a clinical professor of radiology at the University of Arkansas for Medical Sciences in Little Rock, and colleagues evaluated a breast-specific MR imaging system developed by Aurora Imaging Technology, North Andover, Mass., which underwrote the study. The design, conduct, and analysis of the study were performed by American College of Radiology (ACR) Image Metrix, Philadelphia, an independent imaging contract research organization, which also oversaw the writing of the manuscript under contract to Aurora Imaging Technology.
Researchers analyzed results from 937 screening and diagnostic breast MR images from four sites using the specialized device. Biopsy data from cancer cases and one-year follow-up information from negative cases were available. The negative predictive value was 98.9 percent for diagnostic cases and 100 percent for screening cases. The false positive rate was 11.2 percent overall, but only 4.9 percent for screening cases. The analysis showed no difference in performance relative to breast density.
A remaining challenge is to reduce the cost per exam, said Dr. Harms, medical director of Aurora Imaging Technology and a stockholder in the company. He noted that the Aurora system costs about $1 million and the cost per study varies from $500 to $1,100 depending on demographics and available insurance coverage. The machines perform two studies an hour compared with four or five for a digital mammography system.
“We have to be judicious in how we use the resource, but if it were cheaper, we would recommend it all the time,” Dr. Harms said. “The interesting thing is that all of our MR studies are done the same way. We might be able to develop a screening MR study with lower cost and faster throughput.”
“One of the key roles I see for ACR Image Metrix is to help companies like Aurora evaluate their products for special attributes that, if the research is positive, can help differentiate them in the marketplace,” said study author Bruce Hillman, M.D., ACR Image Metrix’ founder and chief scientific officer and the Theodore E. Keats Professor of Radiology and Public Health Sciences at the University of Virginia in Charlottesville.
While investigators concluded that high accuracy was achieved using dedicated breast MR imaging, they cite limitations to the research, including the retrospective design of the study and a patient sample that was heavily weighted toward diagnostic, rather than screening examinations. Moreover, they are concerned about the study’s “generalizability to other settings.” As Dr. Harms noted, “MR imaging was performed at dedicated breast centers with expertise in breast imaging. However, the comparison studies were all from subspecialized radiologists in academic centers who used rigorous diagnostic criteria in a research setting. It is unlikely that these findings are totally caused by interpretation skills alone.”
To hear Radiology Senior Deputy Editor Deborah Levine, M.D., conduct a podcast discussion of “Screening US in Patients with Mammographically Dense Breasts: Initial Experience with Connecticut Public Act 09-41,” with researcher Liane E. Philpotts, M.D., go to radiology.rsna.org/content/265/1/suppl/DC1.
To access the full study by Regina J. Hooley, M.D., and colleagues, go to radiology.rsna.org/content/265/1/59.full?sid=cb83c334-0970-4535-8d3a-9e64f27db3fd.
Access the Radiology editorial on that research, “To Seek Perfection or Not? That is the Question,” at radiology.rsna.org/content/265/1/9.full.
Access Dr. Hooley’s response to the editorial and additional letters on the issue in the March 2013 issue of Radiology at radiology.rsna.org/content/266/3/997.full.
Access the study “Diagnostic Performance of a Dedicated 1.5-T Breast MR Imaging System” at radiology.rsna.org/content/265/1.toc.
For more information on the ACR Image Metrix, go to www.acrimagemetrix.com.
Join a global community of leaders in the radiologic sciences.
Continue your education with top-quality learning resources.
With grant applications increasing, the R&E Foundation needs you.