Digital Mammography Hits the Tipping Point

Twitter icon
Facebook icon
LinkedIn icon
e-mail icon
Google icon

hiit040403.jpgFilm mammography has been the workhorse for more than 35 years to examine women for breast cancer. Of the 33.5 million mammographic studies completed each year, approximately 70 percent of those are obtained for screening purposes. At this time, the vast majority are acquired on analog imaging equipment because, according to the National Cancer Institute (NCI), only 8 percent of breast imaging centers currently use digital machines despite the first FDA clearance in 2000. But that is changing.

Five letters seem to have brought full-field digital mammography (FFDM) to the tipping point — that moment writer Malcolm Gladwell describes as when a trend crosses a threshold, tips and begins to spread like wildfire. While “wildfire” may be a bit strong to describe the adoption of medical technology, DMIST has served to validate the positives of FFDM: dramatically improved image quality with lower radiation dose, higher technologist and patient satisfaction, improved interventional efficiency and the benefits of image manipulation. DMIST has reassured many in the radiology community that this technology that has been gaining strength in smaller clinical studies now has the numbers of the masses behind it. That last push facilities needed to acquire it has arrived.

In October 2001, the Digital Mammographic Imaging Screening Trial (DMIST) began accruing more than 40,000 women study subjects who had no signs of breast cancer in 33 sites across the United States. The research protocol, sponsored by the National Cancer Institute (NCI) and coordinated by the American College of Radiology Imaging Network (ACRIN) was designed to compare film mammography with scans that were acquired on digital mammography units. The results, that were published in October 2005 in the New England Journal of Medicine, suggested three categories of women who appear to benefit from the use of digital mammography:


  • Those under the age of 50, no matter the level of breast tissue density involved
  • Those of any age with heterogeneously (very dense) or extremely dense breast tissue
  • Pre- or peri-menopausal women of any age

The results of the DMIST study suggest that digital mammography may be better at detecting breast cancer than traditional film mammography for these specific categories of women. Approximately 65 percent of the women in DMIST fit into one of these three sub-sets — with 40 percent of those women classified as having dense breast tissue. During the course of the study, 335 women were diagnosed with breast cancer. 

In terms of sensitivity, DMIST found that both digital and film mammography had the ability to tell if cancer is present about 70 percent of the time in the overall study population using conventional methods for measuring sensitivity in a breast cancer screening trial. However, for women with dense breasts, sensitivity was rated at only 55 percent for film mammography versus 70 percent for digital mammography.

The initial results of the DMIST study have certainly stimulated discussion of digital mammography technology among the women’s health and radiology communities and may encourage additional breast centers to purchase digital systems. Consumers, largely women, are more aware of the DMIST results and anecdotally many mammography centers say they are getting more requests for digital mammograms from a more educated public. According to the FDA Mammography Quality Standards Act (MQSA) website in February, of the 8,874 certified mammography facilities, 865 of them were using a total of 1,210 accredited FFDM units.

Personnel in these centers are quite enthusiastic in describing the adoption of this technology and report many affirmative experiences while offering cautionary and smart advice for those who choose to follow.


Positive aspects of digital adoption


The basic advantages of digital mammography as compared to film mammography include the improved ease of image management and storage, combined with a lower radiation dose that does not compromise diagnostic accuracy.

Steven Poplack, MD, co-director of the breast imaging center at Dartmouth-Hitchcock Medical Center in Lebanon, N.H., explains that with their Hologic Selenia system they have experienced better penetration through the dense breast giving them improved contrast within the image through use of digital technology. “When you look at a dense heterogeneously or extremely dense breast, digital to film screen side by side, the images of the digital just ‘wow’ you in terms of their quality and contrast. And I’m a believer that we actually see microcalcifications better than we do with film screen, primarily because of the contrast resolution.”

Janet Storella, MD, medical director of GCM-Suburban Imaging in Bethesda, Md., describes a busy screening and diagnostic practice where they complete between 13,000 and 14,000 mammograms per year in addition to all modalities of breast imaging and intervention. They installed their first GE Senographe FFDM unit in May 2004 and have found the images to be superior in all patients. They’ve also appreciated the throughput advantages as they’ve transitioned to a digital practice.

Besides that obvious clinical advantage, Storella notes that both technologists and patients are much happier with digital technology and workflow. “We found our patients like it better because they don’t have to wait for their films to be processed, and the technologists know immediately if they have positioned the patient correctly. And it is almost impossible to make a bad exposure, at least on our system.”

In addition to image quality, end-users appreciate the ability to manipulate the digital images with their powerful workstations. Kelly Wells, RT (R)(M), RDMS, who serves as technical director of Riverview Women’s Center in Noblesville, Ind., reports that when they transitioned from three analog to two digital mammogram units to accomplish the 10,000 mammograms they perform annually, they decided to digitize their old films using an iCAD Total Look digitizer so that the technologists and radiologists could compare digital with digital images. Additionally the digitized “priors” are sent to the acquisition station in the patient room so the technologists can review former studies before performing the current digital imaging study.

Their multi-modality Siemens MammoReport workstation is capable of magnifying, window/leveling and manipulating the digitized film images as well as the new digital captures. “With the Siemens unit, there are tabs so you can view previous images and then click back to look at the current image.” Although they also provide the old films in the reading room, the radiologists report they review the old cassette-based films only 5 to 10 percent of the time.

Wells says that they have received many positive comments from their patients who remark that their mammogram was more comfortable due to the flexible paddles that are part of this system. “And there is a feature called ‘Opcomp’ that tells the technologist when he or she has achieved enough compression so that no more is required.” This optimized compression feature compresses only as long as the breast is soft and pliable and then stops. An additional feature “SoftSpeed” adjusts the speed of compression based on resistance in the breast tissue.

Nancy Elliott, MD, FACS, who serves as director of the Montclair Breast Center in New Jersey, describes their use of a GE Senographe for the past four years for screening mammography as well as interventional procedures. With wire localizations, the acquisition monitor in the room with the patient provides images of where the questionable lesion area is. “On a regular analog system, you have to wait for the film to be developed while the patient is still in compression, and it is not comfortable. Procedures that used to take half an hour now take only five minutes.”

Another benefit Elliott conveys involves the ability to share digital images for second opinions or to provide the pathologist with additional clinical information. After they obtain the digital image, they can print it and send an original, not a photocopy of an original. “Sometimes copies of mammograms are so bad you really can’t read them,” she says. “Every time we print digital, it’s an original.” In this high-risk practice, that is a vitally important consideration.

Connie Deardorff, RN, BSN, program manager at Athens Regional Breast Health Center in Athens, Ga., explains that they purchased three Siemens Mammomat NovationDR units when they opened their new breast center in the fall of 2004. This $3.5 million breast center was funded by their hospital foundation to increase the breast cancer screening rate in their 17 county catchment area. Before the center was opened, they were screening 20 percent or fewer of the women in the screening population.

With their digital workflow, they find many benefits. Image manipulation and the flexibility it offers the radiologists is a major advantage. The center also utilizes computer assisted detection (CAD) from R2 Technologies.

One further benefit to the adoption of digital mammography is that the studies require approximately three-quarters of the radiation dose of film mammography. Poplack notes that while image quality is uppermost in his mind, an added benefit is the decrease in radiation exposure to the breast.  Although dose from either study is small (about 7 mSv for film mammography), being able to further decrease radiation is always a plus.


Pearls of wisdom from the front lines


While all of these positive aspects are definitely appreciated, there are some areas of challenge for digital mammography that those in the field want to share. The challenges include full integration of mammography within PACS  (picture archiving and communications systems), the archiving space required for these images and the low reimbursement rates for these studies, although reimbursement rates are somewhat higher for digital images, especially those coupled with CAD.

Maria Pilar Martinez, MD, chief of breast imaging at Baptist Health in Miami (Florida) describes their large mammography practice — that performs 50,000 studies per year system-wide — where they have purchased Siemens digital units to perform screening, diagnostic and interventional procedures, including stereotactic biopsy. They have had a very positive experience on the whole with improved images, workflow enhancements and they appreciate their multimodality MammoReport Softcopy Reporting diagnostic quality workstation with integrated CAD. As a multi-modality workstation, they can view ultrasound, MR images as well as mammograms on the diagnostic quality monitor.

However, Martinez notes that the interconnectivity between all of the different systems is important, and that the image data sets require a fair amount of archiving space. Each image is approximately 30 Megabytes (MB). Since each normal screening exam requires four images, that equates to 120 MB for each patient. Diagnostic studies often require eight images or 240 MB per patient. Given that a busy center may perform between 100 and 120 mammograms per day, the storage issues requires some careful up-front consideration.

In addition, Martinez says that when a facility switches from analog to digital, the role of information technology experts is key. Network experience is vital, and a fair amount of patience is required as the transition proceeds.

Andrew Maidment, PhD, assistant professor of radiology and chief of the physics section of the department of radiology at the University of Pennsylvania in Philadelphia, says they have primarily employed their GE Senographe systems in a research setting, although they triage some of their patients through the digital units. They have found their technologists really appreciate working with the digital systems, and their throughput seems to have improved because they do not have to develop films and they can see the images to QC them immediately.

Maidment describes the units as “rocks” and says they have a very high up time and their image quality is outstanding. One problem he notes is that they lack a large detector panel, and that is a problem for imaging women with large breasts. (GE presented a large detector panel as a works-in-progress at RSNA 2005, and expects FDA clearance soon.)

In terms of their network configuration, he describes a “T” pattern for the images. One copy of the image goes to the PACS immediately, while another is sent to the physicians’ workstation. Their problem has been that if the radiologists mark the images, those annotations do not get sent to the PACS, so they are working to improve that storage issue. 

Athens Regional’s Deardorff emphasizes the need for careful consideration of naming conventions in considering the functionality of the PACS with digital mammography. “For example, we ran into problems if we used the word ‘scheduled’. The PACS thinks that means that the patient is here and the exam has been ordered. For pre-fetch functionality, they may use other terminology. I always tell people that if you’re looking at a PACS, make sure that items that are ‘scheduled’ or ‘ordered’ or ‘arrived’ are defined so there is no confusion.”

Laurie Anne Cleland, MD, who is a radiologist in private practice in Vancouver, Washington, says their workflow was initially slowed with FFDM because there is so much more information presented in the images.  “You have to adjust to how you read the exams because you see so much. That may slow you down initially, but then once you get used to it, you don’t want to go back.” Their practice has been using a Hologic Selenia system since June 2004. Now that more than a year has transpired and as they compare digital images to digital images, she is amazed at how much more quickly she can accomplish the tasks at hand.

One of the decision points that breast imaging facilities must consider is whether or not to digitize the prior film studies. Some centers do and some do not. The issue becomes the ease of comparison of prior exams for the radiologists to review. Of course, another factor includes the issue of lost films.

The primary challenges to adoption of this technology seem to center around the transition period where the learning curve comes into play. With careful pre-planning, some can be ameliorated, the experts say.


Financial realities


Just when all of the benefits of digital mammography and workflow seem ready to propel this technology forward, financial reality strikes given the current inadequate reimbursement rate for mammography. Storella believes that many small centers may not be able to afford to purchase digital systems which typically run between 1.5 and 4 times more expensive than their analog counterparts, according to the NCI. The cost of the equipment may be amortized over a long period of time, but the reimbursement rates are not significantly higher.

Reimbursement by Medicare in 2005 for film-screen mammograms was $85.65 and for digital screening mammography was $135.29 with an additional $19.71 if CAD technology is employed. (Medicare reimbursement rates reflect national average rates which are adjusted for individual locales.)

Even with an increase of 58 percent of digital over film reimbursement without CAD and 81 percent if CAD is included, given the cost of the equipment plus the usual overhead requirements that include staff salaries, it could take a center a fair amount of time to pay for the additional expenses of a digital unit. The name of the game has to include increased volume of patients with digital mammography. The authors of the DMIST trial are completing an extensive cost-effectiveness analysis at this time which they hope to complete by year’s end.


Conclusion


Digital mammography is a very promising technology for the diagnosis of breast cancer in certain classes of patients. The early adopters of these systems have found claims to be true that image quality is significantly better, throughput eventually improves, and there is a very positive response from technologists and patients alike. Transitioning from analog to digital requires thoughtful planning and careful implementation. If costs can be managed, this technology appears to greatly benefit certain populations of patients. Many women are quite savvy and well educated about the benefits of digital mammograms, so demand may drive adoption of this technology.


 

CR Mammography Waits in the Wings

In general radiology, digital x-ray images are acquired either by direct digital radiography (DR) which sends the image to a workstation within seconds or using a re-usable cassette based system (CR) where the images are captured on a cassette, then read and processed on a CR reader. Similar procedures occur with computed radiography for mammography as compared to full-field digital mammography.

While CR mammography has been adopted and well-received in Europe and Japan, currently these systems have entered the process of review and approval by the FDA in the U.S. Considering the tight regulation of mammography in general, extreme care and caution are the watchwords for adoption of new methods in screening healthy women for breast cancer. But the hope is that some systems will be available for sale this year.

The Kodak DirectView CR 850, CR 950 and CR 975 systems offer the ability to produce mammographic images with an upgrade (Kodak DirectView CR Mammography Feature) and the use of DirectView CR Mammography cassettes and new EHR-M screens. In this way, healthcare facilities can perform general radiography and mammography studies on the same CR platform. Launched in 2005, Kodak has installed hundreds of CR systems with mammography capabilities outside the U.S. Clinical trials are underway with commencement of the pre-market submission process with the FDA, as well as seeking regulatory approval from Health Canada.

Konica Minolta previewed their new REGIUS PureView Mammography System with a worldwide launch at RSNA 2005 in Chicago.  This system combines phase contrast mammography (PCM) technology with computed radiography (CR) to deliver digital mammography images at the same resolution as film screen images. Currently beginning the FDA approval process, this system has been in clinical use in Japan and Europe.

Agfa Health provides DR mammography capability in collaboration with the Siemens Mammomat NovationDR by building solutions including the addition of their workstation and other components. In the CR realm, they anticipate release of the CR 85-X mammography system in Europe in the second quarter of 2006. Currently Agfa leaders are contemplating their next steps in entering approval processes in North America.

FujiFilm Medical Systems USA reports the Fuji CR for mammography requires FDA approval and is not currently available in the U.S., although Fuji did recently receive approval of this system from Health Canada.

Sectra MicroDose Mammography is currently sold in Europe and in Asia Pacific and has entered the FDA approval process designed to certify the system for marketing in the U.S. 

Once CR mammography is approved, many institutions may select this option because it facilitates a retrofit of their existing mammographic system. But for now, the waiting game continues.



Breast Tomosynthesis on the Horizon

Tomosynthesis is an investigational technique that shows promise in reducing “call-backs” for ambiguous results created by screening mammograms. A report at the 2005 RSNA meeting revealed that about 12 percent of screening patients require additional imaging studies. However, investigators at the Dartmouth Hitchcock Medical Center in Lebanon, N.H., reported that of a cohort of 98 women with questionable results on FFDM (full field digital mammography) 49 would have been given a definitive diagnosis on the first try with tomosynthesis.

Steven Poplack, MD, co-director of the Breast Imaging Center at Dartmouth explains that tomosynthesis is a technique where three-dimensional images of the breast are created. Using a prototype add-on to their Selenia FFDM system under development by Hologic, a series of exposures are taken in an arc with a regular x-ray tube-filter combination. The data are then reconstructed into a digital data set as one-millimeter or one-and-one half millimeter thick slices, so that the physician is able to “cine or page through” those images.

General Electric Health Care has been developing a breast tomosynthesis system based on the Senographe 2000 Digital Mammography system for several years in a collaboration with Massachusetts General Hospital and others in an effort led by Daniel Kopans, MD, the director of breast imaging at Harvard and Massachusetts General Hospital in Boston. One of the challenges of 2D mammography is that tumors can be obscured and thereby go undetected. Digital breast tomosynthesis produces reconstructed 3-dimensional projection images that enable a physician to “page through” the interior of the breast without obstruction by surrounding superimposed tissue.

Siemens Medical Solutions has been collaborating with the Duke University Medical Center to develop a digital breast tomosynthesis system based on a prototype modified Mammomat NovationDR designed to acquire image projections of the breast from many different angles. The expectation is that this system will enhance diagnostic ability through improved sensitivity of breast cancer detection, as well as reduce unnecessary biopsies.

Estimates suggest that tomosynthesis may be ready for clinical use in approximately two years.