Digital breast tomosynthesis: Is it ready for prime time?
Researchers predict FDA approval soon, but obstacles remain
For years, scientists have been talking about how digital breast tomosynthesis (DBT), a technique that enables radiologists to view three-dimensional images of the breast, is going to revolutionize breast screening. The development and approval process has taken much longer than anyone anticipated, but that hasn't dampened the enthusiasm of leading researchers, many of whom now believe that approval from the Food and Drug Administration (FDA) for at least one DBT system will come within the next year.
Excitement over the technology stems from data showing that DBT outperforms conventional mammography on almost every metric of care, according to Joseph Lo, PhD, an assistant professor of radiology at Duke Advanced Imaging Laboratories in Durham, NC, and the principal investigator for several clinical trials, using a prototype DBT system developed by Siemens Medical Solutions, which maintains its U.S. headquarters in Malvern, PA. "All the trials have been very positive," Lo says. "There has not been a single trial that has eventually shown that [DBT] is not living up to its promise."
Further, Lo emphasizes that the clarity that DBT provides, combined with the sheer amount of information that is delivered in the form of multiple pictures of the breast taken from many different angles, is a huge asset to radiologists. "If the images are done properly, they look very much like mammograms, but instead of everything being pancaked together, you get these nice big slices," says Lo. "That is why radiologists are incredibly enthusiastic about this technology."
However, even if a DBT system is approved by the FDA tomorrow, Lo acknowledges that obstacles relating to cost and workflow stand in the way of large-scale implementation, at least for the immediate future. Nonetheless, over the long term, most researchers familiar with the technology see DBT ultimately replacing conventional mammography while significantly improving quality.
Early data show great promise
Any proper history of DBT takes you directly to Daniel Kopans, MD, who heads the breast imaging division at Massachusetts General Hospital (MGH). As far back as 1978, he came up with the idea of applying tomosynthesis to X-ray mammography, but he had to wait until digital detectors were developed in the early 1990s to proceed with the technology because they were needed to enable computers to synthesize the breast images.
"We identified a General Electric [GE] detector, which was then under development for mammography, as being best suited for approach," says Kopans, noting that he then worked with colleagues to perfect and patent the technique for MGH.
Working with a grant from the Army, MGH and GE Healthcare, headquartered in Buckinghamshire, England, developed the first prototype DBT system and began testing it on volunteers. "We undertook reader studies that showed that DBT improved the conspicuity of cancers, was more accurate than standard mammography in differentiating benign lesions from malignant lesions, and reduced the false-positive rate compared to conventional two-view mammography," says Kopans. "We are now completing a 3,000-woman pilot screening program that confirms the decrease in false positives." In that study, Kopans notes that the callback rate for women undergoing FFDM was 8.1% The callback rate was reduced to 5.1% (a decrease of 37%) when the same group of participants underwent DBT.
A tomosynthesis device is not radically different from a conventional mammography machine, except that the X-ray tube moves independently while the breast is held in position in order to image it from several different angles, Kopans says. A workstation then can display the images "like the pages of a book," he says. Kopans notes that DBT enables you to see each page, individually, without the superimposition of the breast tissues. "The radiologist is not fooled by overlapping tissues, and the exact location of a lesion is never in question, as it can be with conventional mammography," Kopans says.
Time is an issue
While DBT produces many images of the breast, Lo and Kopans emphasize that the prototype systems they are working with expose the patient to no more radiation than standard, two-view mammography. Further, they point out that there is not much of a learning curve involved for radiologists because reading DBT images is very similar to reading mammograms. However, given that there are multiple images to review with DBT, there is the issue of time.
"Instead of the standard four-view mammography that is involved with screening today, [radiologists] would be looking at potentially hundreds of slices," says Lo. "It is going to take them longer, and potentially much longer."
In addition, Lo points out that the clinical trials now under way need to definitively prove that the clarity offered by DBT, which enables radiologist to see more, probably benign, lesions, does not result in all kinds of unnecessary callbacks. "It requires [radiologists] to learn when to act on something they think they see and when to let it go, but it is not like they have to learn from scratch," Lo says. "They can use their considerable mammography experience and leverage that."
Further, much of the research involving DBT is focused on methodologies that can help radiologists effectively manage the huge volume of data that DBT provides, says Laurence Clarke, PhD, who oversees research funding for projects related to imaging technology at the National Cancer Institute (NCI). "Not only are we trying to support and promote the development of hardware, but also software solutions in parallel so that there are tools to assist the radiologist in … recognizing the areas that require their attention," Clarke says. NCI grants have been awarded to three or four research groups that are focused specifically on computerized solutions to analyze the data associated with DBT, he says.
Obstacles remain
The biggest, initial obstacle to implementation of DBT technology is probably economics. Estimates of how much a DBT system will cost vary considerably, but most researchers conclude that it will cost more than a standard, full-field digital mammography (FFDM) system, and potentially quite a bit more. To address this concern, the early-makers of DBT technology, including GE, Siemens, and Bedford, MA-based Hologic, are developing systems that are upgrades to current FFDM systems so that imaging centers that may have only recently adopted FFDM systems can make the transition to DBT without totally replacing their FFDM machines.
Further, despite the extra time and expense that DBT will involve, the enhanced accuracy of DBT will offer some economic advantages as well, says Hal Kipfer, MD, section chief of breast imaging at Indianapolis-based Indiana Radiology Partners and an assistant professor of radiology at Indianapolis-based Indiana University School of Medicine. "With current methods, which include screen-film mammography and FFDM, we have to recall an average of 100 women after their screening mammogram and do roughly 10 biopsies in order to detect three to five cancers," Kipfer says. "Ideally, our screening test would allow us to identify those three to five women directly without the extra recalls and biopsies. This would not only save money, it would save women a lot of anxiety and pain."
He is doubtful that DBT will enable radiologists to achieve that level of accuracy, but he is hopeful that it will get them closer to that ideal than they are now. In addition, he points out that the technology, which makes the cancers more visible, also should help radiologists detect an additional 10-50 cancers per 100 at an earlier stage. "If DBT is as revolutionary as predicted, it may require screening centers to upgrade just to stay in business," adds Kipfer. "On the other hand, if costs are significantly higher, without proportionate increases in reimbursement, it may not be financially feasible to offer DBT." At this point, Kipfer stresses that it becomes a societal/political question: What is society willing to invest to prevent women from dying from breast cancer?
Approval may be imminent
While the first DBT systems implemented likely will consist of upgrades to FFDM machines, investigators already have built a next-generation DBT system, explains Paul Carson, PhD, the director of the Division of Basic Radiological Sciences at the University of Michigan, based in Ann Arbor, who has been working with GE on the project.
"It was built from the ground up as a purely [DBT] unit as opposed to being an add-on to a digital unit, and that was to enable the [X-ray] tube to move fast enough and far enough to cover a wide range of angles and to stop at a lot of places in a short time," Carson says. The enhanced coverage of the stand-alone DBT system should further enhance the performance of DBT technology, according to Carson, although FDA approval and market adoption of this system is probably years away.
Predictions on precisely when the FDA will finally approve the first DBT systems are difficult to pin down, but most researchers involved in the technology believe it will be soon. DBT approval has been predicted for the past few years, Lo says, "but now the rumors are really flying, and it is probably a pretty sure bet that some kind of news is coming by the end of [2008].
For years, scientists have been talking about how digital breast tomosynthesis (DBT), a technique that enables radiologists to view three-dimensional images of the breast, is going to revolutionize breast screening.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content