| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Commentary |
1 Department of Radiology, George Washington University, 2150 Pennsylvania Ave., NW, Washington, DC 20037.
Received September 10, 2007; accepted after revision September 11, 2007.
Address correspondence to R. F. Brem.
Keywords: breast cancer • computer-aided detection • double reading • mammography screening • oncologic imaging • study design • women's imaging
The authors of the well-designed study reported in the previous article [1] set out to compare the use of computer-aided detection (CAD) and a single radiologist interpretation of screening mammography with double reading for the improved diagnosis of breast cancer. With the increasing use of CAD, as well as the ongoing controversy about its use, this is an important goal and the authors ask a critically important question. However, it is also a challenging goal. The comparison of single reading with CAD versus double reading requires a large patient population to achieve sufficient statistical power to definitively answer the question posed. The admirable study, reported in this issue of AJR [1], is the basis on which additional studies will continue to build our knowledge base of this rapidly expanding field. Although not the primary goal, additional and excellent insights about the use of CAD as well as its integration into clinical practice result from this study.
The authors [1] have used sophisticated and extensive statistics. They report the recall rate, biopsy rate, positive predictive value, and increased cancer detection rates for a single reader with CAD as compared with the double reading of mammograms. In this study, 6,381 consecutive screening mammograms were included in which a total of 18 cancers were present, 15 diagnosed and three false-negative cancers that were found on imaging 1 year later. The literature reports between 3 and 8 cancers per 1,000 screening mammograms [2], although the prevalence of breast cancer in the population reported in the study being discussed is 2.03 per thousand for the primary reader and 2.35 for all three "readers,"—that is, the primary reader, the primary reader with CAD, or the primary reader along with the second human reader [1]. The authors ascribe this difference to the fact that these are predominantly incidence cancers and not prevalence cancers because most patients had prior mammograms and suggest that this difference likely reflects the excellent care afforded patients seen in the investigators' clinical practice. Nevertheless, the result of the study of the 6,381 screening mammograms is 18 cancers. This is the primary limitation of this study. The authors note that the small study size does not allow for a statistical comparison. In fact, they state that ".... the number of cases was too small to determine if the CAD or if the human reader was superior,...." and therefore the question that this study set out to investigate could not be definitely answered. The authors understand the limitations of their study; they state that:
Limitations to this study may be the sample size and the inability to show statistical significance between the human second reader and the CAD reader. To show statistical significance, one would need a difference of at least six cases between the groups, and any fewer has a power of zero.
It would, of course, be optimal if this study allowed a statistically significant comparison. The authors insightfully point out that more than 24,000 screening mammograms would be needed to achieve the necessary patient population. A future, larger study will ultimately need to be done to evaluate the difference in cancer detection using double reading of mammograms versus single reading with the use of CAD. Nevertheless, it is important that this study is published for several reasons. First, it is an excellent example of a properly designed study with the clear ability to answer the question raised given a sufficiently large study population. The lack of bias among the three interpreting groups—the first human reader, human reader with CAD, and double reading of mammograms—is evident in the study design. Second, the appropriate study size for the definitive study can now be calculated as a consequence of the findings reported in this study. We in the field of breast imaging now await future studies so that we will be able to answer the question: Is single reading with CAD equivalent to, better than, or worse than double reading of mammograms?
The authors [1] describe the experience and expertise of the interpreting radiologists. The investigators and readers in the study are expert mammographers with extensive experience in breast imaging. Even so, it is also noteworthy that two of the three false-negative mammograms were identified with CAD but dismissed by the interpreting radiologists. The utilization of CAD for the optimal diagnosis of breast cancer requires further definition and training. CAD has the ability to detect not only additional cancers, but also cancers that were dismissed by not one but two experienced mammographers. However, we need to learn how to better utilize and integrate this important technology into our practices to further harness the potential of CAD for the improved diagnosis of breast cancer and the benefit of our patients. The maximum benefit of CAD is achieved only by a collaboration of the expertise of the radiologist along with the proper integration of CAD in clinical practice.
The conclusion of these investigators is that there is no difference between cancer detection using either single interpretation with CAD or double reading of mammograms. If that is the case, then we should expect a 5–15% improvement in breast cancer diagnosis with CAD because that is the well-established improvement in cancer detection when comparing single with double reading [2]. A second human reader is an enormous drain on the critical shortage of mammographers, and the limited reimbursement for mammography does not support the routine double reading of mammograms. The use of the equivalent approach—that is, CAD—would then result in the goal of improved breast cancer detection.
Although CAD does not detect all cancers and marks nonmalignant locations, at least 13 studies have been published in peer-reviewed journals that evaluate the use of CAD to assist radiologists in detecting breast cancers earlier by identifying lesions that may be overlooked [3–16]. Of these 13 studies, nine prospectively evaluated the use of CAD, three were retrospective studies with cancers missed in clinical practices that could have been detected with the use of CAD, and one was a retrospective study to compare single reading with CAD to double reading without CAD [15]. All 13 studies showed an increase in cancer detection rate, an improvement in radiologist sensitivity, or a potential reduction in missed cancers with the use of CAD. Even the recent controversial study by Fenton et al. [13] showed an increase in breast cancer detection, although they chose to report it as a decrease in accuracy due to the increase in the number of patients recalled. Clearly, you need to recall more patients in order to increase the number of cancers diagnosed.
The authors of the study reported in this issue of AJR [1] state that their results are comparable to those reported by Fenton et al. [13]. If so, then the use of CAD should result in an increase in the detection of early breast cancers. However, when the authors compare their findings with those of Fenton et al. they state that "...both studies showed a lack of improvement in cancer detection with the use of CAD." In fact, the study by Fenton et al. did show an improvement in cancer detection with CAD, although the study reported in this issue of AJR [1] did not.
The authors [1] insightfully state that one of the interesting and important findings of this study is that the cases called back by the human second reader and by CAD are different, again indicating the complementary nature of mammographic interpretation by humans and CAD. We should further study the differences in the cases called back by human readers and abnormal areas detected with CAD to use both approaches to optimize the detection and diagnosis of breast cancer so that patients can be the ultimate benefactors.
In summary, the investigators designed an excellent study [1], resulting in findings that future studies will continue to build on. Additional, larger studies should follow the design of this study to result in the definitive comparison of single mammographic interpretation with CAD and double reading. In addition, the authors render excellent insights about the operations of a clinical breast practice as well as the interesting finding that cancers identified by human readers and by CAD are different. It would be beneficial to learn how to best harness the information from CAD, use it in clinical practice, and optimize the diagnosis and care of the woman with breast cancer.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  F. J. Gilbert, S. M. Astley, M. G.C. Gillan, O. F. Agbaje, M. G. Wallis, J. James, C. R.M. Boggis, S. W. Duffy, and the CADET II Group Single Reading with Computer-Aided Detection for Screening Mammography N. Engl. J. Med., October 16, 2008; 359(16): 1675 - 1684. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
