| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Harvard Medical School Massachusetts General Hospital Avon Foundation Comprehensive Breast Evaluation Center Wang Ambulatory Care Center Boston, MA 02114
Houssami et al. [1] made several minor mistakes and one major one in their article comparing the "accuracy" of mammography and sonography among young symptomatic women. Although the reviewers were blinded, those who obtained the studies were not, and consequently, acquisition of the sonographic images was not blinded and may have led to potential biases. In addition, without long-term follow-up, the true false-negative rates for the studies are not known. Of greater importance, however, is the basic premise that they consider mammography to be diagnostic.
Although we use the term "diagnostic" mammography, anyone involved in breast imaging should be well aware that mammography is rarely diagnostic. On occasion, a palpable lesion has classic benign features on the mammogram that can, legitimately, be used to avoid further intervention (e.g., calcified fibroadenoma, hamartoma, oil cyst). This situation occurs, however, with few patients. Most of the time, a lump is either not visible on the mammogram or indeterminate. When a woman presents with a sign or symptom of breast cancer, the clinical problem needs to be addressed even if all the imaging findings are negative [2]. We and others pointed out years ago that the role of mammography in the symptomatic woman is almost always as a screening test [3, 4]. Although mammography may provide some information about the clinical problem, its role is really to screen the rest of the ipsilateral breast and the contralateral breast for unsuspected cancer.
Beyond screening, the real value of any imaging test for the symptomatic patient is limited. On occasion, imaging may prompt intervention when the clinical suspicion is low, but our experience is that this is rare. One could argue that imaging is of little value in the symptomatic woman. Many lumps prove to be cysts. It is usually less expensive to have the clinician perform a clinically guided needle aspiration than it is to do a sonographic study. Thus, if the lesion is a cyst, not only does aspiration confirm the diagnosis, but it also eliminates the cyst in one step (simple cysts do not need draining—it is just a free benefit of the diagnostic aspiration). Similarly, unless you believe that you can tell benign solid masses from malignant lesions using sonography (despite common misconceptions, this has not been proven), showing that a mass is solid only indicates that a biopsy is needed. The sonographic study does not alter care except perhaps to show that an imagingguided biopsy could be performed. The main reason to image the breast of a symptomatic woman is to screen the rest of the breast and the contralateral breast for unsuspected cancer.
Certainly the issue for women 40 years old and older is clear. Because mammographic screening reduces the death rate, all women 40 years old and older should be screened. It makes no sense for a woman to be denied mammographic screening simply because she is symptomatic. As far as evaluating the area of clinical concern, the question remains, Does any imaging test, beneficially, alter the care of the patient?
References
The University of Sydney and Westmead Hospital Westmead, NSW 2145,
Australia
University of Sydney NSW 2006, Australia
Royal North Shore Hospital NSW 2065, Australia
My coauthors and I thank Dr. Kopans for his interest in our article [1]. As we acknowledge in paragraph 1 of the Discussion section of our article, we agree with him that acquisition of the sonographic images was not blinded, although images were interpreted by radiologists in a blinded manner. We point out that our study has used the highest quality methodology to date on this topic. It is theoretically possible that a small number of cancers may not have been identified at time of evaluation (false-negatives), but in reality, none of the patients was identified as having breast cancer on our check for registration with the New South Wales Central Cancer Registry within 2 years from attending the study center. We therefore believe that our estimates of accuracy are not biased by any error in classification by the reference standard.
Kopans states that we have made a major mistake in considering mammography as a diagnostic study and that "beyond screening, the real value of any imaging test for the symptomatic patient is limited." We do not agree with his views. We have studied a clinically referred symptomatic population. Most radiologists and clinicians providing care to women with breast symptoms would agree that breast imaging is a key and critical component in the investigation of clinical breast abnormalities, and this view is supported by the letter from Dr. Hall [2]. Most radiologists would also be aware that the positive predictive value of a clinical lump is generally low, particularly when referral comes from a primary care physician or relates to a patient-identified lump. Negative imaging findings in this context would support the clinical impression that further intervention is not warranted.
In Australia, the contribution of breast imaging to the diagnosis of symptomatic women is endorsed through national recommendations [3]. A combination of the findings on breast imaging and breast clinical examination guides clinical decision making concerning further investigation and treatment of symptomatic women [4]. Our study makes no recommendation that women 40 years old and older be "denied mammographic screening," as stated in Kopans' letter. Whether the potential benefit of screening these young women outweighs possible harm is still under debate and is not the focus of our study, which "makes no inferences as to the accuracy or merits of screening mammography" [1].
Finally, we believe that there is evidence that sonography does contribute to differentiating malignant from benign solid lesions [5], and our data on the accuracy of sonography in symptomatic women [1] provide further support to this knowledge.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Strano, P. Crystal, and D. Kopans Adjunct Sonography and Not Screening in Cancer Detection Am. J. Roentgenol., August 1, 2004; 183(2): 539 - 539. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
