| 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
Dr. Berg should be congratulated on her detailed analysis of the available data to address the question of whether sonography should be used for breast cancer screening [1, 2]. Her conclusions, however, in both her "Commentary" and "Questions and Answers" should have been more clearly stated.
Sonographic screening for breast cancer should not be performed outside carefully designed scientific trials.
Radiology has been criticized for being a specialty that relies on anecdotal information. All of medicine endeavors to move beyond "I saw a case" to develop scientific evidence as the basis for practice. This is not an easy task. Even one of our major journals recently published 35 articles, and only one was under the category of evidence-based practice. When experience is all that we have to guide care, this is reasonable. However, whenever a new test is being considered for introduction to the clinical armamentarium, scientific evaluation is needed to be certain that the test will do more good than harm. Introducing a new screening test that will be used among healthy individuals requires a different level of evidence than the level of evidence required for diagnostic approaches for those who are ill. Berg nicely pointed out the issues involved but left the conclusion ambiguous. If proper scientific evaluation was not needed, then all the issues concerning screening mammography could have been avoided. Even I, as a strong advocate of breast cancer screening, have come to understand over the years and through the debates the importance of weighing scientific evidence before subjecting a healthy population to a test that may lead to morbidity without conferring any benefit.
The purist might argue that a randomized controlled trial (RCT) of sonographic screening is needed and must show statistically significantly fewer deaths in the screened group than among the controls before sonographic screening could be considered efficacious. The problem with this approach is that it might take 10–15 years to prove a benefit. This is too long to wait. The RCTs of mammographic screening, along with other data [3], have shown fairly convincingly that the size of a cancer is fairly predictive of outcome. If sonography can reveal cancers at a smaller size and an earlier stage, a benefit may be predictable. However, the problem is not this simple. Some breast cancers never kill. If sonography shows only these nonlethal cancers, women would end up undergoing unnecessary toxic treatments. Another possibility is that sonography may find only cancers that have already metastasized to other organs. Finding them earlier would make no difference in the individual's survival, but she would know about the disease sooner and have a longer period of time with all the attendant psychologic and physical distress.
The American College of Radiology Imaging Network (ACRIN) trial that Berg [1] outlined is a major step in determining whether an RCT of sonographic screening should be performed. Because none of the studies that reports finding occult cancers had the sonographic operator blinded to the mammogram, their sonographic results are subject to significant bias. Investigators continue not to understand the fact that simply promoting sonography in conjunction with mammography does not eliminate the need to find out how sonography functions as an independent screening tool. A blinded evaluation is required for scientific proof. It is not optional.
If the ACRIN study shows that sonography truly depicts mammographically and clinically occult cancers, an RCT will be needed to prove that detecting the cancers earlier will reduce deaths. Although health planners will likely need convincing, I believe that we can avoid waiting for deaths in such a trial. Because the lead time gained from screening is likely to be less than 5 years, if sonography reveals cancers at a smaller size and earlier stage among the screened women compared to the controls, but at the end of 5 years, the same total number of breast cancers is found in both groups (the screened women had their cancers detected earlier, but the cancers eventually surfaced in the control group), the efficacy of sonographic screening could be proven in a shorter time. If over the period of study sonography showed no more cancers, but the cancers that it showed were at a smaller size and earlier stage than in the control women, we could conclude that there is a benefit. Until such a trial is performed, however, the data do not support screening with sonography.
I would also point out that, although Berg's summary repeatedly suggested that the age of 50 has some biologic meaning [1], she was merely quoting articles that, inappropriately, grouped their data so that changes that take place gradually with increasing age were made to appear to change suddenly at age 50. By comparing all the women under the age of 50, dichotomously, to all the women over the age of 50, the age of 50 is imbued with false significance. The age of 50 is arbitrary and meaningless [4].
References
Breast Imaging Consultant Lutherville, MD 20193
Note.—ACRIN protocol 6666 is supported by grants from the Avon
Foundation and National Institutes of Health (CA80098) through the American
College of Radiology Imaging Network.
Further assessment of the generalizability of the performance of screening breast sonography in combination with mammography is warranted. Clearly it is not yet time to endorse widespread sonographic screening for breast cancer. However, at selected centers, if a woman chooses to pay (outside her insurance plan) for screening breast sonography, this is not a priori an unreasonable decision if she is as informed as is currently possible regarding the potential benefits and risks. Generally, full disclosure and discussion of the limitations of a test may be best performed in the context of clinical trials.
The national publicity surrounding screening with sonography and MRI raises the risk that some centers may capitalize on anxious women, offering and advertising these services for cash only, for profit, with inflated claims of performance. The quality of both sonography and MRI of the breast varies widely, depending on the equipment, the operator, and the experience of the interpreting radiologist. The patient should inquire about the number of breast sonograms the practitioner has obtained and interpreted; a minimum of 500 per year for 2 years is required for investigators in the American College of Radiology Imaging Network (ACRIN) protocol 6666 [2]. The practitioner should also have experience interpreting mammograms. ACRIN protocol 6666 requires investigators to have interpreted a minimum of 2,500 examinations per year for 2 years.
The practitioner should either personally perform (as in the ACRIN protocol 6666) or at least carefully review the sonographic examination (if performed by a technologist) while the patient is in the imaging suite; generally, the woman should expect to meet the radiologist at least briefly while a targeted sonogram is obtained if she has any question about an abnormality on clinical examination or on a mammogram that may have prompted the sonogram or about the sonogram itself. The interpreting radiologist for either sonography or MRI should be prepared to perform a biopsy of any suspicious findings. The facility should have procedures in place to follow up any abnormal results and identify cancers missed as well as generate statistics on how many biopsies were recommended and what the results of those biopsies were.
ACRIN protocol 6666 will be used to evaluate the performance of screening sonography across multiple centers in women at high risk of breast cancer who have at least heterogeneously dense parenchyma in at least one quadrant of the breast on mammography [1, 3]. The primary aim of the protocol is to compare the cancer detection rates of mammography combined with sonography to those of mammography alone. More important, mammographic and sonographic interpretations will be independent to assess the true performance of screening breast sonography. A second integrated interpretation will be performed. The chance of finding cancer not seen on mammography in high-risk women is at least two to three times greater than in women with average risk. If the results are encouraging, broader population-based studies may be warranted. In that event, it will be necessary to identify widely available human resources capable of performing high-quality breast sonography and to facilitate the development and evaluation of automated whole-breast sonographic devices.
As Kopans correctly points out, the size and grade of invasive cancers and the lymph node status are quite accurate predictors of survival [4, 5]. Therefore, broader studies of new breast cancer screening technologies with a 5-year horizon should allow accurate evaluation of the technology compared or combined with mammography. Waiting for survival as an end point does not appear justified as long as treatments are independent of method of detection. Debates about all-cause mortality are likely to continue but are also not unique to the method of detection.
No magic age exists at which mammography or sonography or any other test will be of dramatically greater benefit than the year before. National recommendations for screening are based on cost per year of life saved. At the age of 45, if results of single-center studies are maintained and three cancers are found per 1,000 women screened with sonography, preliminary calculations for ACRIN protocol 6666 (Schleinitz MD, personal communication) show that the marginal cost of annual screening breast sonography per year of life saved will be in the range of $50,000–70,000. This is similar to the cost of using seat belts, mammography, dialysis, and many other widely accepted health interventions [6].
A woman must weigh the limited evidence, risks, and potential benefits of additional screening tests beyond mammography if she is at average or only slightly elevated risk of breast cancer, has no access to clinical trials, or has relatively dense breasts. Performing highquality screening breast sonography can be challenging. If a capable practitioner and informed patient choose to proceed with caution, it is not unreasonable. We do not know the optimal timing for serial examinations, although an annual examination similar to that of mammography may be appropriate. In the interim, it is of paramount importance that well-designed trials to evaluate new screening technologies are well supported and proceed expeditiously so that we will be better positioned to make recommendations on the basis of broader scientific evidence.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  O. Graf, T. H. Helbich, G. Hopf, C. Graf, and E. A. Sickles Probably Benign Breast Masses at US: Is Follow-up an Acceptable Alternative to Biopsy? Radiology, July 1, 2007; 244(1): 87 - 93. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. S. Gerson and L. L. Fajardo Screening mammography, sonography of dense fibrocystic breast tissue Am. J. Roentgenol., December 1, 2003; 181(6): 1715 - 1715. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
