Breast Tomosynthesis — Experts Discuss Their Current Use of the 3D Mammography Tool
By Kathy Hardy
Vol. 14 No. 6 P. 26
Digital breast tomosynthesis (DBT), when introduced in early 2011 as an FDA-approved technology for use in both breast cancer screening and diagnosis, presented as a promising technology but with cautions regarding optimum use and capabilities. Two years and several studies later, more women are selecting this imaging option, also known as 3D digital mammography, while breast imagers continue to weigh the pros and cons.
With the approval of Hologic’s Selenia Dimensions system (Dimensions 3D), US hospitals and imaging facilities could offer women tomosynthesis along with conventional 2D digital mammography. According to Hologic, this system has been commercially available outside the United States since 2008, so the company has had time to work on optimizing hardware and software applications as well as image processing algorithms.
DBT should be used as an adjunct to conventional 2D breast imaging, not as a replacement, according to the FDA. With that, many breast imagers believe in the technology combination, one even predicting it could become an industry standard. “We’re finding more cancers with tomosynthesis, and we have a reduced recall rate,” says Jaime Geisel, MD, an assistant professor of diagnostic radiology at the Yale University School of Medicine. “If we can reduce false-positives with mammography and tomosynthesis, then maybe we have a better screening test. The new gold standard could be 2D with 3D.”
Geisel was the principal investigator among a team of researchers at the university’s Smilow Cancer Hospital in New Haven, Connecticut, which recently announced results of a study on adding DBT to screening mammography. The retrospective review of 14,684 women who received screening mammograms at Yale from August 2011 through July 2012 showed an 11% increase in the rate of breast cancer detection among women who underwent both 2D and 3D screening. Geisel adds that tomosynthesis appeared to be more valuable in women with dense breast tissue, as 54% of the women whose cancer was found with tomosynthesis had dense breast tissue.
Findings from this study related to women with dense breast tissue dovetail with the findings of another group of Yale researchers, which assessed the potential contribution of tomosynthesis to detecting breast cancer in high-risk patients, according to Geisel. Researchers in this study concluded that adding tomosynthesis to 2D mammography may improve the visualization of invasive cancers presenting as noncalcified lesions in high-risk women but also recommended further research in this area.
A majority of the screening mammograms performed at Yale now include tomosynthesis, Geisel says, with the exception being women with breast implants and those with especially large breasts, who are already receiving extra views. The facility was an early adopter of this technology, and currently has three units and plans to add two more. Incorporating this technology into Yale’s breast imaging procedures has changed how Geisel reads mammograms, she says.
“Before, I would need to take multiple views to localize possible cancers before sending the patient to ultrasound,” Geisel explains. “Now I get localization without the extra pictures.”
She also sees the addition of DBT as a way to reduce the number of patient recalls and false-positives. “We’re not calling back women as much,” she says. “We need fewer pictures, which mean less radiation and less stress. This is good for women.”
According to Julian Marshall, senior director of global product management for Hologic’s breast imaging group, Dimensions 3D remains the only product approved by the FDA for 3D mammography. Other companies, including GE Healthcare and Siemens Healthcare, are developing 3D mammography systems but do not yet have FDA approval to market in the United States.
“The first half of 2013 has been very fruitful in terms of peer-reviewed publications of studies using our technology,” he says. “The Oslo Tomosynthesis Screening Trial publication came out in January 2013, and the results were very strong. Based on over 12,600 screening exams, the addition of 3D imaging to the conventional 2D exam resulted in a 40% increase in the detection of invasive cancers; a 27% increase in detection of all cancers, invasive and in situ combined; and a 15% reduction in false-positive rates.”
He also refers to a recent study out of Italy, the STORM trial (Screening Tomosynthesis or Mammography), that showed similar results. For breast imagers, studies such as these are stepping-stones for advancing methods of screening for and diagnosing breast cancer. Some, however, progress more cautiously, continually looking for more information as to the benefits of this evolving technology.
Still Defining When to Use It
“This is more evidence of the potential usefulness of tomosynthesis,” says Carol Lee, MD, chair of the ACR Breast Imaging Communications Committee and a diagnostic radiologist at New York’s Memorial Sloan-Kettering Cancer Center. “It is important that we get this information and continue to refine which patients are best served by this type of screening.
“We need to see what happens in future years,” she continues, “to see if the same benefits of tomosynthesis hold up year after year. Also, does using this technology increase or decrease interval cancers? That’s something we need to answer.”
Following the announcement of the initial Oslo study findings, the ACR and the Society of Breast Imaging issued a statement saying, “While the study results are promising, they do not provide adequate information to define the role of tomosynthesis in clinical practice.
“Although the cancer detection rate was higher when tomosynthesis was added to mammography alone, it is not known if an equal incremental benefit will be realized in a second screening round,” the statement continues. “This small study does not supply statistical information regarding subgroups or women that might benefit or might not benefit from adding tomosynthesis. How the technology will affect screening accuracy among women of different ages, risk profiles, and parenchymal density is uncertain.”
Lee believes further refinement is needed for determining protocols for who should undergo DBT. She would like to see more effort to categorize which women would most likely benefit from DBT. “Which women need this?” she asks. “Should we look at certain risk categories, specific ages, and other factors like this? With more options available for breast cancer screening and diagnosis, we need to refine the most logical way of determining what’s best for each patient.”
Marshall notes that while DBT is believed to be beneficial for women of all types of breast compositions, the benefit for women with dense breasts is greater. However, whether or not this technology is offered to patients could come down to system availability, he says.
“While ideally tomosynthesis would be used for screening all women, in many cases, the decision of when to use tomosynthesis is driven by system availability,” he says. “If a facility has a limited number of 3D-capable systems, they may choose to triage women with dense breasts for screening. Some facilities with a limited number of systems have chosen to do only diagnostic exams with 3D, while others find the biggest benefit in a combined approach. Facilities with an adequate number of systems to screen all women with 3D find that to be the preferred use.”
In 2011, Elizabeth Wende Breast Care in Rochester, New York, was one of several clinical sites for Dimensions 3D. Attending radiologist Stamatia Destounis, MD, FACR, says since that time, the facility has added a second unit and is hoping to add a third because of the patient volume, particularly those at high risk of breast cancer. She says that with increased research and word of mouth among the patient population, DBT has become a frequent option at her facility.
“Two years ago, there was a lot of skepticism about tomosynthesis,” Destounis says. “There were questions about where this would fit in to the breast screening and diagnosis process. It was seen as expensive, with no reimbursement and an increase in radiation.”
Today, however, she sees great enthusiasm for DBT, with clinical research and data from facilities using the technology adding to its acceptance. Initially, she says the facility staff used tomosynthesis with 2D only for screening high-risk patients, but usage is now more widespread. When there is a question regarding findings in the 2D view, she says the radiologists can check the area with the 3D view, which has helped the facility reduce its recall rate, she says. “Now we use the technology for screening and diagnostics, and it’s available for any woman who asks for it,” she says. “We use it every day now and everyone benefits from this, not just a certain group of women.”
Clinicians don’t always rapidly adopt new medical technologies. Marshall says Hologic is pleased with tomosynthesis’ rate of adoption in the United States, especially compared with the rate at which its predecessor technology, digital mammography, was accepted. The company installed more 3D systems in the first year following approval than it did the first year following approval of its first 2D system. He attributes some of this difference to the fact that Hologic’s 2D system was designed for conversion to a 2D/3D system with a simple software upgrade. “Many of our customers were able to plan for and make the transition with little difficulty,” he says.
For some hospitals and breast imaging facilities, the addition of such new technology isn’t as easy. With no reimbursement for DBT, the cost of purchasing these systems could become “an obstacle to adoption,” Geisel says. There’s also the issue of how to incorporate this process into a practice, for example, determining how to handle workflow if the practice can afford only one machine.
Geisel notes that another potential drawback of combining 2D and 3D breast imaging is the increased radiation dose. Conducting these together means twice the dose, she says. This is more of an issue in the United States, where breast cancer screening exams are done every year, than in Europe, where screenings are done every two years, she adds.
“Something that would be good for the future of this technology is to be able to take the 3D data and reconstruct it in 2D,” she says. “This would mean less radiation for patients.”
Marshall says Hologic addresses this with a recent development to the company’s C-View image software, which can generate a 2D image from the tomosynthesis data set. “An FDA advisory panel met in October 2012 and voted to recommend C-View for approval,” he says, “and we are currently awaiting formal approval from the FDA. C-View images will provide workflow efficiencies by reducing exam times and time under compression as well as reduce dose for women who are concerned about radiation exposure.”
Hologic also is working to incorporate tomosynthesis into other aspects of breast cancer detection. Marshall says the company recently released a tomosynthesis biopsy package that allows radiologists to use the same modality for detection of abnormalities and localization for biopsy. He says this enhancement is especially useful for lesions visible only on 3D images.
Breast imaging experts agree that with changes in technology will come more studies to fine-tune the effectiveness of tomosynthesis in breast cancer screening and diagnosis as well as who may best benefit from this technology. Geisel says that more studies like those conducted at Yale also will increase awareness of this technology. “With time and more studies, DBT will become the norm,” she says.
Many also would agree that, as debate continues regarding the validity of even basic mammography, the introduction and adoption of new screening and diagnostic technology creates even more questions.
“It’s important to ask more questions and gather more data regarding new tools and not just adopt them because they’re new,” Lee says. “I’m happy to see this kind of progress, but it’s not the end of the story.”
— Kathy Hardy is freelance writer based in Phoenixville, Pennsylvania. She writes primarily on women’s imaging topics for Radiology Today.