| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Lakeside Hospital Metairie, LA 70001
Memorial Medical Center New Orleans, LA 70115
In the recent article explaining the use of step-oblique mammography for the evaluation of a potential abnormality visible on multiple images in only one projection [1], it is the opinion of the authors that this is an "unwise" setting for the use of sonography. We disagree with this opinion and believe that sonography can and frequently should be the modality of first choice in this circumstance.
A sonogram obtained by an experienced sonologist and targeted to the line of projection of the suspected abnormality would certainly have found most if not all of the cancers that were identified by the step-oblique technique in the article. If additional mammographic views are then needed, observing the character and distribution of adjacent tissue will aid in selecting the best mammographic projections to reveal the lesion's margin characteristics. Sonographic examination will also allow the assessment of associated physical findings, and the possibility of satellite lesions and multifocality, and help plan for any subsequently needed interventional procedure(s).
The concern that isoechoic lesions would be missed by performing sonography first is moot for two compelling reasons. First, truly isoechoic breast malignancies, which are visible mammographically, palpable, or both, are rare. Even when truly isoechoic, breast malignancies are uncommonly occult when scanned by an experienced sonologist. Second, for all patients in whom no sonographic explanation for the mammographic finding is found, additional views would then be performed either confirming the artifactual nature of the initially detected density or verifying its validity as a true abnormality. The overall effect of the sonogram-first approach not only for abnormalities seen in only one view but also for all suspect breast masses is at the very least a significant decrease in the number of additional mammographic views or examinations required to arrive at an anatomic diagnosis with no increased risk of missed lesions.
One of the fundamental issues raised by the article's position, as we see it, is what constitutes a breast sonogram. Is it the acquisition of multiple images taken at predetermined intervals to "cover" the area of interest with those images subsequently "read" by a radiologist or is it a scan performed by the interpreting radiologist while correlating the sonographic, mammographic, and physical findings in real time? We submit that if the former is the case in a given practice situation, then the step-oblique method may be the appropriate choice. However, in the latter case, we believe sonography is not only appropriate as the first diagnostic modality of choice but oft times the only additional modality needed.
An argument can be made that the sonography-first technique results in an unacceptable increase in the cost of the diagnostic workup. This is an understandable concern, but we would challenge this notion for several reasons. First, when sonography reveals a worrisome mass for which there is no historical explanation, additional mammography in large part becomes superfluous as there is no way it will obviate the need for biopsy. This represents a potential cost savings, not to mention the patient's usual appreciation when additional mammographic views can be avoided. Second, when a clearly identifiable lesion such as a cyst is defined, no additional mammographic evaluation is warranted, again saving the expense of additional mammographic views.
Similarly, when a sonographically benign-appearing solid lesion is discovered, or not infrequently, several identical sonographically benign lesions, such as in the case of multiple fibroadenomas, performing additional mammographic images will not usually alter the treatment of that patient. Finally, in spite of "negative" step-oblique images, how often is the need for a correlative sonogram to ensure the artifactual nature of a perceived summation shadow truly obviated? A series of 69 patients is after all a small sample size to conclude accuracy sufficient enough to place the well-being of our patients in the balance. How many cases have all of us worked up in which tailored diagnostic mammographic images failed to define a density seen in only one or more views only to have sonography define a true lesion as either benign or malignant? Worse yet, without that corroborative sonogram false-negative findings and conclusions could be reached and result in an undue, potentially harmful, delay in diagnosis.
In conclusion, we believe that step-oblique mammographic imaging is an interesting and, in selected cases, potentially helpful technique, but it is naive and misleading to suggest that the use of sonography in the evaluation of densities seen only in one projection or in the setting of suspected summation artifact is inappropriate or "unwise."
References
University of California at San Francisco Medical Center and
University of California at San Francisco School of Medicine San Francisco, CA
94143-1667
Baptist Medical Center—Beaches Drs. Mori, Bean, and Brooks,
P.A. Jacksonville, FL 32216
Joyce Eisenberg Keefer Breast Center John Wayne Cancer Institute St.
Johns Medical Center Santa Monica, CA 90404
Brigham and Women's Hospital and Harvard Medical School Boston, MA
02115
We are grateful to Drs. Wells and Rupley for the opportunity to expand on several important issues concerning the complementary nature of diagnostic mammography and breast sonography, issues that were beyond the scope of discussion in our recently published article on the efficacy of step-oblique mammography [1].
The content of the Wells and Rupley letter primarily involves opinion. This reply however, relies much more heavily on evidence-based data supported by citation of articles in the peerreview literature. For this reason, despite the several arguments advanced to the contrary, we remain firmly convinced that diagnostic mammography rather than sonography should be done first in the workup of lesions detected on only one standard mammographic projection, both in the very narrow context of using step-oblique mammography (lesion seen on more than one image but in only one projection) as well as for the remaining (>99%) of one-view-only lesions detected on screening mammography.
Let us start by considering the issue of lesion palpability, which is raised several times in the letter from Wells and Rupley. The ability of a sonologist to correlate in real time the imaging and physical findings of a confusing case is not in dispute; this is one of several important reasons why sonography should play a major role in the imaging of palpable breast masses. However, none of the 69 lesions reported in the step-oblique mammography article were palpable, even in retrospect, completely negating the potential advantage of sonography in the workup of one-view-only cases that benefit from step-oblique mammography. What about the broader issue of the initial workup of a one-view-only lesion detected at screening mammography? More than 80% of these lesions prove to represent summation artifacts, and very few of the remaining "real" lesions are palpable, even in retrospect [2]. So again, the potential added value of sonography is negated.
Strongly arguing for the initial use of mammography rather than for sonography for all one-view-only lesions is the inability to exclude the possibility of a real lesion when no comparable sonographic findings are identified (requiring additional mammography, as acknowledged by Wells and Rupley) and the already confirmed accuracy of mammography in reliably differentiating between a summation artifact and a real lesion, accuracy that is based on a series of 2023 consecutive cases, not just the 69 cases reported in our article on the step-oblique view [1, 2]. Thus, first-use mammography will definitively stop the workup in most cases (summation artifact), and first-use sonography cannot.
We agree with Wells and Rupley that truly isoechoic, mammographically visible, noncalcified breast cancer is encountered infrequently, but this statement is valid only in the context of second-look targeted sonography once the three-dimensional location of the cancer is known. One should not perform targeted sonography for a lesion that is visible on only one mammographic projection (our many years of hands-on experience have convinced us of the inaccuracy of targeting to the "line of mammographic projection," as suggested by Wells and Rupley, simply because it is very difficult to duplicate on sonography the specific positioning and compression of the breast that was used previously on mammography). Furthermore, even if a sonographic lesion is identified in this context, additional mammography will still be needed to show whether the sonographic lesion is truly the same as the mammographic finding initially visible on only one projection. All too often in this context, sonography reveals a different lesion (almost always benign), leaving the mammographic finding unevaluated and still potentially malignant. This is another reason why sonography should not be the first examination in such workups.
Wells and Rupley also make the observation that they have identified true lesions on sonography in cases in which tailored diagnostic mammographic images failed to define a density seen on only one view. We do not doubt the truth of their statement, but we question the completeness of their tailored mammographic examinations. In fact, the major reason we wrote our article on step-oblique mammography was to provide radiologists with a reliable new method to solve what in the past has been a vexing mammographic problem [1].
Finally, Wells and Rupley address the issue of the increased induced costs that would accompany sonography as the first examination in the workup of one-view-only lesions. They acknowledge that this issue raises "understandable concern," but argue that sonography will actually result in reduced costs by obviating diagnostic mammography in many cases by depicting definitively benign (e.g., simple cyst) or clearly suspicious findings. Unfortunately, their argument is false (as already discussed) because even if first-use sonography does reveal such lesions, additional mammography will still be needed to show whether the sonographic lesions are truly the same as the mammographic findings initially visible on only one projection. Because diagnostic mammography will not be obviated, all that is left to consider is the "understandable concern" that first-use sonography will substantially increase the cost of workup.
In summary, both additional mammography and sonography contribute substantially to the workup of findings revealed on screening mammography. The decision of which technology to apply first will depend principally on the specific imaging features of the case, based on which approach is more likely to complete the workup. In the clinical setting in which there is doubt about whether a nonpalpable mammographically detected finding represents a summation artifact or a real lesion, mammography alone should be used to make the distinction. Sonography plays an important role in the further evaluation of those lesions that are shown to be real.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
