Technology and Industry: Major study compares digital and film-based mammography

By Kathleen M. Dallessio
pdf path

Two major studies addressing breast cancer detection were published recently. One attempted to quantify the widely acknowledged role of screening mammography in the reduction of breast cancer mortality; the other addressed the issue of the relative efficacies of digital and analog mammography for breast cancer detection.

CISNET study

The Cancer Intervention and Surveillance Modeling Network (CISNET) Collaborators used 7 independent statistical models of breast cancer incidence and mortality to assess the relative and absolute contributions of screening mammography and adjuvant treatment in the nearly 25% decline in breast cancer mortality seen in the United States since 1975. 1 The various models showed that screening mammography was responsible for 28% to 65% of the reduction, with a mean reduction of 46% between the models. Adjuvant treatment was credited with the remainder of the decrease.

The DMIST study

For clinicians, however, the current question is which mammography technology-digital or analog-is going to help them find the most cancers as early as possible. The Digital Mammographic Imaging Screening Trial (DMIST) was designed to compare the sensitivity of digital mammography with that of film-based mammography for breast cancer detection. 2

This large-scale 2-year study, conducted by the American College of Radiology Imaging Network, included more than 42,000 women in the final data analysis. Initially, more than 49,500 asymptomatic women who presented for screening mammography at 33 sites in the United States and Canada underwent both film-based and digital mammography, in random order, on 1 of 5 first-generation digital mammography systems: The SenoScan from Fischer Medical (Denver, CO); the Computed Radiography System for Mammography from FUJIFILM Medical Systems USA (Stamford, CT); the Seongraphe 2000D from GE Medical Systems (Waukesha, WI); and the Digital Mammography System and the Selenia Full-Field Digital Mammography System from Hologic, Inc. (Bedford, MA).

All mammograms were interpreted independently by 2 radiologists-one read the film and the other read the digital image-using a 7-point malignancy scale and the BIRADS classification. Biopsies or aspirations of suspicious lesions were performed if either of the radiologists recommended the procedure. A participant was classified as positive for breast cancer if cancer was pathologically verified within 455 days of the initial mammogram.

Data were compared for all participants and across a variety of subgroups, including age (younger or older than 50 years), breast density (heterogeneously dense, extremely dense, or less dense), menopausal status (pre-, peri-, or postmenopausal), race (white, black, or other), and lifetime risk of breast cancer (>25% or <25%), as determined by the Gail model.

A total of 335 breast cancers were diagnosed during this study. The authors reported finding no significant difference in the overall accuracy of the 2 methods, nor did they find any significant difference in accuracy between film and analog mammography in relation to race, the patient's breast cancer risk, or the brand of system used. The authors concluded that "the overall diagnostic accuracy of digital and film mammography . . . is similar, but digital mammography is more accurate in women under the age of 50 years, women with radiographically dense breasts, and premenopausal or perimenopausal women." They also reported that cancers detected with digital mammography that were missed on film included many invasive and high-grade in situ cases.

The importance of DMIST

"I think what makes this study important is the fact that the results demonstrated that there is technology out there that can actually improve the diagnostic accuracy of mammography for women in particular subgroups," said Andy Vandergrift, National Program Director for Women's Health Care Imaging at FUJIFILM. "The subgroups that were mentioned actually represent a majority of the screening population: 65% of the 49,000+ women who participated in DMIST fell into at least one of those categories. Since it was such a large study, I think that we can at least draw the conclusion that the majority of the women in the screening group would benefit from digital mammography."

John Pavlidis, President and Chief Executive Officer of R2 Technology, Inc. (Sunnyvale, CA) agreed. "It is the most significant study of its kind," he said. "It is very comprehensive and has very credible methodology. It's the only study of this scale that really compares film-based and digital mammography. It's been long and eagerly awaited and its publication is a major milestone. It shows how a forward step in technology can help radiologists have greater accuracy in evaluating mammograms."

Georgia Hitzke, Vice President of Clinical Development at Hologic was most impressed with the findings in women with dense breasts. "Patients with dense breasts are at the highest risk for breast cancer and we've always tried to optimize technology for these women," she said. "I think there is a fallacy about postmenopausal women-that older women don't have dense breasts. DMIST showed that 40% of them do have dense breasts. So there is a large population-it's not just the younger people that we have to make sure we can diagnose; it is all ages."

The continuing role of analog mammography

This does not, however, mean that there is no longer a role for analog mammography. From a practical standpoint, there are relatively few digital mammography systems currently installed in the United States. "In terms of the installed base for digital systems, I understand that it is approximately 8%, but the growth rate for digital has been quite high," explained Pavlidis. "It is reportedly 35% to 40% per year. Our sense is that this will further accelerate the adoption of digital mammography."

Furthermore, the study found that, in some instances, analog mammography showed some lesions that digital missed. "DMIST said that digital was better in dense breasts, but film-screen was equal or maybe just a little bit better in fatty breasts," said Hitzke. "We want to be sure that people still get film-screen mammography on a good system. It doesn't make a difference which manufacturer it is, but it is important that it is a good system for dense breasts or any type of breast, because it will pick up microcalcifications and other lesions. This study just shows that digital can sometimes pick up different types of things or maybe pick them up a little bit earlier in certain populations."

"It should be stressed that film-screen mammography is still very good and it's the best screening tool we have today to diagnosis breast cancer," agreed Vandergrift. "I think it still is the gold standard; however, as new technologies, such as digital mammography, are introduced, they are typically compared with the standard. What we see here is that the new technology actually offers advantages, and that is how change comes about."

"I think the study definitely says that there will be an ongoing role for film-based mammography," added Pavlidis. "The benefits seen with digital were evident in a subset of the population. That, by no means, implies that there is no benefit for analog mammography. Interestingly enough, it is also important to note that the improvement in detection with digital over analog in the study is similar or even smaller than the improvement in detection by adding computer-aided detection (CAD) to either digital or analog mammography."

Use of CAD in mammography

"CAD was not part of the study," continued Pavlidis, "But CAD is a proven technology that increases cancer detection rates, and CAD works for both analog and digital mammography. Published rates range from a 7.4% to 19.5% increase in cancer detection. We have seen a very high adoption rate of CAD in association with either analog or digital mammography sales. We believe that the DMIST study will further accelerate the adoption rate for digital mammography."

Looking forward

"The interesting thing about this study is that a lot of the technology used is already old technology," added Hitzke. "Currently available digital technology is actually better. A lot of us thought the 2 types of systems might be equal, so I was pleasantly surprised by the results. Given the technology we have today, I think that digital will be even that much better than film."

"The study shows that we are moving in the right direction with digital," concluded Hitzke. "Full-field digital mammography is good, and the most important thing on these systems is image quality and the best detection. To me, getting the best technology is a cause worth fighting for. I am happy that something great came out of this study."

Back To Top

Technology and Industry: Major study compares digital and film-based mammography.  Appl Radiol. 

December 06, 2005
Categories:  Section

Copyright © Anderson Publishing 2016