| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Editorials |
1 From the Department of Radiology, University of North Carolina School of Medicine, Campus Box 7510, Rm 503, Old Infirmary Bldg, Chapel Hill, NC 27599-7510. Received September 21, 1999; accepted and revision requested September 30; revision received October 5. Address reprint requests to the author (e-mail: etpisano@med.unc.edu).
Index terms: Editorials • Breast neoplasms, diagnosis • Breast neoplasms, radiography, 00.112, 00.121, 00.30 • Breast radiography, technology, 00.121 • Radiography, digital, 00.121
In September 1991, the National Cancer Institute convened a panel of breast imaging experts to determine where future research dollars for breast cancer detection and diagnosis should be invested. The overwhelming consensus of those who attended and the ultimate conclusion of the report generated at the conference was that the National Institutes of Health should place high priority on the development of digital mammography (1).
Screen-film mammography has been extremely effective and has been widely accepted as a screening modality over the past 20–30 years. Together with breast physical examination, it has been shown to reduce breast cancer mortality by 18%–30% (2,3). Traditional mammography does have substantial limitations, however. Approximately 10%–20% of palpable breast cancers are not visible on images obtained with this modality, mainly due to insufficient contrast between normal and abnormal breast tissue (4–6). In addition, only 5%–40% of lesions recommended for examination at biopsy prove to be malignant (7–9).
Digital mammography holds promise since each part of the breast imaging chain can be optimized—image acquisition, image storage, and image display. In contrast, the radiographic film used in traditional mammography must serve all three functions. In addition, digital detectors might improve breast cancer detection due to the increased efficiency of absorption of the incident x-ray photons, a linear response over a wide range of incident radiation intensities, and the low system noise. All of these properties should enhance the visibility of subtle contrast differences between tumors and normal background tissue. Digital image processing allows the manipulation of image contrast and might further improve lesion conspicuity.
There are currently at least four full-field digital mammographic systems that are undergoing clinical testing in the United States and Canada; this includes testing for the purpose of obtaining U.S. Food and Drug Administration (FDA) approval. These are the SenoScan system (Fischer Imaging, Denver, Colo), the Computed Radiography for Mammography system (Fuji Medical Systems USA, Stamford, Conn), the Senographe 2000 D system (GE Medical Systems, Milwaukee, Wis), and the Trex Digital Mammography System (Trex Medical, Danbury, Conn).
Each of the four systems is based on a different technologic approach. The spatial resolutions of the four systems range from 41 µm per pixel for the Trex system to 100 µm per pixel for the GE and Fuji systems. The contrast resolutions range from 10 bits per pixel for the Fuji system to 14 bits per pixel for the Trex and GE systems. Other systems are also in development (10,11).
Clinical Testing to Date
To my knowledge, there have been two major clinical trials funded by nonindustrial sources, to date. These are the International Digital Mammography Development Group Digital Mammography Pilot Study, and the Department of Defense Full-field Digital Mammography Screening Trial.
The International Digital Mammography Development Group study, funded by the Office of Women's Health in the Department of Health and Human Services, involved eight centers that used the Fischer, GE, and Trex full-field digital mammographic units. Two hundred women were enrolled. Patients underwent both digital and screen-film mammography. Eligible women were recruited from the population of patients scheduled for biopsy after an abnormal mammogram or after abnormal findings at physical examination, or they were randomly selected from the population with normal problem-solving, or diagnostic, mammograms. A controlled reader study is being carried out by 18 radiologists to compare the performance of digital mammography with that of screen-film mammography in this population. The Department of Defense has recently funded a follow-up study, which will involve approximately 1,000 additional patients.
The Department of Defense Digital Mammography Screening Trial is ongoing and will enroll approximately 15,000 women who present with no symptoms for screening mammography at two centers—the University of Colorado and the University of Massachusetts. Women who agree to participate will undergo both full-field digital mammography with the Senographe 2000 D system (GE Medical Systems) and screen-film mammography, the findings of which will be interpreted independently by two radiologists. Both studies are being used clinically to lead to an additional work-up, which will include biopsy.
The interim results of this study were presented at the annual meeting of the Radiological Society of North America in 1998 (12). At that time, approximately equal numbers of cancers had been missed or detected by using each modality (five of 22 cancers missed at screen-film mammography and six of 22 cancers missed at digital mammography). Furthermore, digital mammography was associated with a lower recall rate that was statistically significant (11.5% vs 14.1%) and a higher rate of positive findings at biopsy (43.3% vs 23.7%) than conventional mammography (12).
The American College of Radiology Imaging Network is currently developing a protocol for a screening trial that will involve all types of digital mammographic equipment currently undergoing clinical testing. This trial should commence some time during the year 2000.
Other tools that will become more readily available once mammography is digital might also substantially improve breast cancer detection and breast lesion characterization. Telemammography will allow off-site interpretation, supervision of examinations, and consultation, which will allow women in remote areas access to previously unavailable screening and diagnostic mammographic services. Computer-aided diagnosis will ultimately provide a second computer-generated interpretation of the findings from all examinations (1). Tomosynthesis will provide low-dose three-dimensional tomographic imaging of the breast, which will potentially give valuable information about the surface characteristics of lesions that is currently available only though additional mammographic views (13,14). Dual-energy mammography may provide information regarding the atomic number and density characteristics of the breast tissue. For example, this could potentially provide images of the breast that contain just the calcifications (15–18). All of these promising adjunct technologies are also currently undergoing development and testing.
FDA Trials
On June 19, 1996, the FDA published "Information for Manufacturers Seeking Marketing Clearance of Digital Mammography Systems" on its Web site (19). This document described how manufacturers were to conduct clinical trials to show agreement between screen-film mammography and digital mammography to obtain FDA approval for their devices through the 510(k) or premarket approval mechanism. Manufacturers were instructed to discuss their evaluation plans with the FDA Center for Devices and Radiological Health.
According to the FDA guidance document, clinical trials were to be designed to demonstrate that the probability of a positive digital mammogram was greater than 0.90 for patients with abnormal screen-film mammograms and that the probability of a negative digital mammogram was greater than 0.95 for patients with negative screen-film mammograms. The FDA estimated that approximately 520 women (260 with abnormal screen-film mammograms and 260 with normal screen-film mammograms) would be needed to achieve these target agreement figures. Significantly, the FDA did not initially require that the manufacturers determine the truth regarding the presence or absence of cancer, only that the interpretations of the screen-film mammograms and those of the digital mammograms agreed.
Subsequently, all four manufacturers designed the agreement studies, which were discussed extensively with the Center for Devices and Radiological Health. Two companies, Trex Medical and GE Medical Systems, completed the trials and submitted data for FDA review.
On August 17, 1998, the FDA convened an Advisory Panel to discuss the guidance document and the mechanism for review of the digital mammographic devices. The consensus at that session was that the initial FDA guidelines were flawed since the level of agreement required for digital mammography and screen-film mammography was not attainable, even when screen-film mammography was compared with itself, because of intra- and interreader variability (20–22). The large inherent reader variability for mammographic interpretation is partially explicable because of differences in reader performance, but some of the variability is attributable to minimal differences in positioning and compression that take place when two sets of images are obtained in the same woman.
Consequently, on February 8, 1999, the FDA revised its guidance document. The digital mammographic trials for FDA approval must now be based on the truth regarding breast cancer status and not on the direct agreement with findings at screen-film mammography. That is to say, methods such as receiver operating characteristic analysis must now be used to prove the substantial equivalence of this new technology to screen-film mammography. The studies designed under the prior guidance document were no longer useful for the purpose of FDA approval. After more than 2
years of effort, the manufacturers were forced to return to "square one."
Unfortunately, the process now in place for FDA approval of these devices is not proceeding quickly to the development of new studies that will ultimately allow for the marketing of these products. Consequently, it is impossible to predict when these devices will become widely available in the United States. It is understandable that federal regulators are obligated to exercise caution in the evaluation of any new imaging technology, especially one with a public profile as high as that of digital mammography.
However, it is not clear why more than a full year has passed since the August 1998 FDA Advisory Panel meeting without the release of new specific guidelines for the manufacturers. This delay in bringing products to testing for FDA approval is clearly not in keeping with the intentions of Congress. Section 205 of the FDA Modernization Act of 1997 requires the FDA, in consultation with the product sponsor, to consider the "least burdensome" means that will allow appropriate premarket development and review of a product without unnecessary delay and expense to manufacturers (23).
The rest of the world is not waiting for the FDA to grant approval to invest in this new technology. According to information available on the GE Medical Systems Web site (24), "GE Medical plans to install more than 50 of the units outside the United States by the end of 1999." According to Robert A. Uzenoff, Executive Assistant to the President of Fuji Medical Systems USA, there are approximately 200 Fuji systems that are being used for mammography worldwide (Uzenoff B, oral communication, 1999).
The women and physicians of the United States will continue to wait. Of course, the women of this country deserve the best tools available for the diagnosis of this dreaded disease. Digital mammography has multiple inherent advantages over traditional technology. The manufacturers need and deserve prompt assistance from federal regulators in designing trials that will allow approval and marketing of these devices.
Footnotes
Abbreviation: FDA = Food and Drug Administration
The University of North Carolina Department of Radiology is being paid by the Fischer Imaging Corporation of Denver, Colorado, for the author's assistance in the design and implementation of the Food and Drug Administration approval trial for the Fischer SenoScan digital mammographic system.
References
This article has been cited by other articles:
|
|
 |
|
  J. Keyrilainen, M. Fernandez, M.-L. Karjalainen-Lindsberg, P. Virkkunen, M. Leidenius, K. von Smitten, P. Sipila, S. Fiedler, H. Suhonen, P. Suortti, et al. Toward High-Contrast Breast CT at Low Radiation Dose Radiology, October 1, 2008; 249(1): 321 - 327. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Mahesh AAPM/RSNA Physics Tutorial for Residents: Digital Mammography: An Overview RadioGraphics, November 1, 2004; 24(6): 1747 - 1760. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Obenauer, K-P Hermann, and E Grabbe Dose reduction in full-field digital mammography: an anthropomorphic breast phantom study Br. J. Radiol., July 1, 2003; 76(907): 478 - 482. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. B. Cole, E. D. Pisano, E. O. Kistner, K. E. Muller, M. E. Brown, S. A. Feig, R. A. Jong, A. D. A. Maidment, M. J. Staiger, C. M. Kuzmiak, et al. Diagnostic Accuracy of Digital Mammography in Patients with Dense Breasts Who Underwent Problem-solving Mammography: Effects of Image Processing and Lesion Type Radiology, January 1, 2003; 226(1): 153 - 160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. F. Good, J. H. Sumkin, M. Ganott, L. Hardesty, B. Holbert, C. M. Johns, and A. H. Klym Detection of Masses and Clustered Microcalcifications on Data Compressed Mammograms: An Observer Performance Study Am. J. Roentgenol., December 1, 2000; 175(6): 1573 - 1576. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
