| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Viewpoint |
1 From the Department of Radiology, University of Alabama at Birmingham, 619 S 19th St, JT 316NW, Birmingham, AL 35233. Received January 30, 1999; accepted February 5. Address reprint requests to the author (e-mail: Evarubin@aol.com).
Index terms: Breast neoplasms, diagnosis, 00.30 • Breast radiography, utilization, 00.30
These therefore represent the results of the enthusiasts, the pioneers in this field who with evangelical style have developed screening and by obsessional attention to minor abnormalities on mammograms have reduced the mortality from breast cancer.I. Fentiman (1)
Screening mammography has more than its share of critics. A search for recent articles on the subject that can be downloaded as full text (ie, those published in the major medical and surgical journals) turns up scores of articles expressing skepticism regarding the benefits and effectiveness of screening. Waves of promotion and criticism of screening mammography buffet the lay public, creating quite understandable confusion and uncertainty.
One of the areas that comes in for the most criticism, both from our colleagues and from our patients or potential patients, is the high rate of false-positive examinations, and the cost, both psychologic and financial, of resolving them. Most of what appears abnormal on a mammogram is not breast cancer. To be a good mammographer requires spending much of one's time separating the wheat from the chaff, determining which abnormalities are sufficiently suspicious as to warrant intervention. Reasonable performance in mammography suggests that approximately 10% of mammograms should be interpreted as having an abnormality indicating the need for a work-up study. Of these, at least 20% will require needle or open biopsy for definitive diagnosis. Lower false-positive (or callback) rates appear to be strongly associated with higher interval cancer rates (2).
As stated by Moskowitz (2), "The state of the art in mammography requires that more attention, not less, be paid to the signs of minimal breast cancer." Therein lies the rub, for the signs of minimal breast cancer overlap those of benignity. How then does the mammographer operate in the face of competing goals—on the one hand, to maximize detection of potentially curable breast cancers and thereby decrease breast cancer mortality and morbidity and, on the other hand, to do as little harm as possible to the vast majority of women who do not have breast cancer?
Surveillance mammography has become a well-accepted means of dealing with the probably benign lesion detected at mammography (3–5). Ten to 15 years ago, short-term follow-up examinations were often recommended at very short intervals, most often 3 or 4 months, with a doubling of the interval for the next follow-up examination. As mammography and interpreters of mammograms matured, the short-term follow-up interval lengthened. Two studies in Radiology, one from the University of California at San Francisco (UCSF) screening program published in 1991 (6) and one from South America published in 1992 (7) established the 6-month follow-up as the standard. In both these studies, nonpalpable probably benign lesions in patients placed in a surveillance protocol had a less than 2% likelihood of proving to be malignant. This suggested that follow-up was a safe and cost-effective alternative to biopsy.
There is little doubt that mammographic surveillance for appropriately selected lesions is a safe and cost-effective alternative to biopsy. The question, I believe, is not whether biopsy should be performed of probably benign lesions rather than follow-up, but whether surveillance need be performed at 6-month intervals rather than 12. Unfortunately, surveillance mammography has become synonymous with 6-month follow-up, and moreover, many radiologists perform 6-month follow-up over an entire period of surveillance, usually 3–5 years, rather than only once as currently proposed by Dr Sickles from UCSF (6).
From the standpoint of the patient, 6-month follow-up is certainly problematic. In most practices, the information regarding the mammographic examination is conveyed to the patient by the referring physician rather than the interpreter of the mammogram, such that the patient receives results colored by their personal physician's attitude toward mammographic abnormalities, an attitude which is often overly cautious. The default mode for many referring physicians, when they are not comfortable with mammography results and recommendations, is referral to surgery. The default mode for many surgeons in the same situation is biopsy. Although the percentage of patients who went on to immediate surgery in the Sickles study (6) was low (0.4%), I do not believe this is a representative number.
In our practice, one of the major reasons patients give for coming to us for a second opinion is a mammogram obtained elsewhere that generated a recommendation for short-term follow-up. It does not take much time talking to patients to realize that a recommendation of 6-month follow-up for a mammographic abnormality is a major source of anxiety. Distrust and doubt in the ability of the mammographer, particularly if there has not been a face-to-face discussion between mammographer and patient, are also common reactions.
The patient presented with a recommendation for 6-month follow-up has a number of choices. She can comply or not comply, seek a second opinion elsewhere, decide never to have another mammogram, or insist on resolution by means of biopsy. Recommending 6-month follow-up to patients is a good way to lose them. In the Sickles study, nearly 31% of patients underwent their initial follow-up elsewhere, or not at all. Compliance with the entire proposed follow-up protocol occurred in less than 50% of the patients.
The issue of cost is also pertinent. Mammography costs have been kept low, some would say too low, in order to encourage screening (8). While a single 6-month follow-up diagnostic mammographic examination or ultrasonographic (US) examination, or both, may not be onerous for a well-insured individual, the added global cost can be substantial. In the UCSF study (6), which focused on studies read by a single highly experienced mammographer, 11.6% of screening mammograms revealed probably benign lesions for which surveillance was recommended. If an expert recommends short-term follow-up in nearly one in every eight patients, how often do less experienced mammogram interpreters recommend it? The potential for abuse is enormous.
Data on diagnostic work-up performed within 1 year of a false-positive mammogram were published in an article by Elmore et al (9). The definition of a false-positive mammogram used by these authors was all patients in whom there was a recommendation for other than routine follow-up within the next 12 months. This definition was broad and included patients with indeterminate and suspicious results and thus likely overestimates the percentage of work-up examinations engendered by a probably benign reading. Nonetheless, the results give one pause. The 631 false-positive mammograms resulted in 439 visits to surgeons (69.6%), 162 (25.7%) visits to nonsurgeon physicians, 384 (60.9%) additional mammograms, 176 (27.9%) US examinations, and 128 (20.3%) biopsies, of which 100 were open or core biopsies and 28 were fine-needle biopsies. The authors estimated that the cost of working up false-positive results was approximately one-third the cost of performing the screening.
The utility of 6-month follow-up must also be analyzed relative to the types of lesions for which follow-up is suggested. First, what is a "nonpalpable probably benign lesion"? In the Sickles study (6), both generalized and localized abnormalities were included in this category. Generalized abnormalities were present in 27% of the patients in his study. Three groups of findings were included in this category: (a) noncalcified, well-defined solid nodules; (b) scattered or randomly clustered tiny calcifications; and (c) discrete clusters of tiny calcifications. This category describes settings in which cancer, when it occurs, is usually less readily identified. It is, however, difficult to understand what benefit is to be derived from 6-month follow-up of patients with generalized abnormalities, none looking substantially different from another, and often bilateral. It begs the question of what exactly is being followed. At 8 years of follow-up, one cancer had developed in the 253 women with multiple well-defined solid nodules, a rate no higher than that expected for screening-detected cancer over this period. We do not know whether this cancer originated from a preexistent nodule or was simply a sporadic event in a patient who happened to have multiple benign nodules.
The presumption that change is a valid indicator of malignancy in patients with generalized abnormalities is somewhat tenuous. In the screening setting, multiple noncalcified well-defined solid nodules are most likely to represent multiple fibroadenomas or a combination of fibroadenomas and cysts. (Although simple cysts were presumably excluded in the UCSF study, it is well to bear in mind that patient accrual occurred in the years from 1978 to 1987 when breast US was both technically and interpretively different from what is available today.) Since these entities are hormonally sensitive and are most often diagnosed in the pre- or perimenopausal patient, fluctuation in size and number would not be unusual.
Multiple peripheral papillomas may also manifest as multiple nodules on the mammogram and are considered risk indicators for breast cancer development. The carcinomas that develop in these patients are for the most part low grade and prognostically favorable, and there are no data to suggest that an examination 6 months after an initial abnormal screen would enhance their diagnosis. Multiple papillomas may be segmental or diffuse in distribution and may even be bilateral. In most cases, some of the nodules are not well circumscribed and US will often reveal intracystic and intraductal solid nodules rather than well-circumscribed homogeneous solid masses. With current US techniques, these patients would likely be placed in an intervention rather than a surveillance category.
Patients with metastatic disease or lymphoma involving the breast may have multiple discrete benign-appearing nodules at mammography. Again, however, these would not be expected to appear benign at modern US. And, while 6-month follow-up might very well show change in these instances, this is not likely to benefit the patient in whom a diagnosis was not made at the outset.
Six-month follow-up in the setting of patients with multiple calcifications in discrete clusters or scattered or random clusters is also difficult to support in the context of what is known regarding the natural history of the benign conditions most often responsible for these findings. The calcifications in patients with benign proliferative changes in the breast, sclerosing adenosis, and so forth, are not necessarily stable over time and minor differences in positioning on different mammographic views may change their appearance from examination to examination. Such patients might well develop calcifications sufficiently suspicious that biopsy is indicated, but this is no more likely to occur at an initial 6-month follow-up than at a subsequent screening examination. The results of the Sickles study (6) support the idea that this is a low-risk situation. Only one of 619 patients with this finding was found to have carcinoma at 8 years of follow-up; a rate that is actually much lower than would be expected.
The remaining 73% of probably benign lesions in the Sickles study were localized abnormalities, and the situation in these cases is quite different. The first and largest group in the Sickles study comprised patients with clusters of tiny round or oval calcifications. The definition in the American College of Radiology lexicon (10) for round calcifications suggests that when smaller than 1 mm these are usually calcifications in the acini of lobules and when smaller than 0.5 mm, the designation "punctate" may be used. It is not known exactly what was considered tiny in the UCSF study, which antedated the lexicon, but a cluster of predominantly punctate calcifications was likely present in most of these cases. If biopsy is performed, most of these prove to be benign conditions associated with proliferative changes in lobules; rarely, a biopsy engendered by this finding will result in a diagnosis of atypical hyperplasia or ductal carcinoma in situ, usually low grade.
One must question whether there is actual risk in following a cluster of benign-appearing calcifications that is in a focus of atypical hyperplasia or low-grade ductal carcinoma in situ, and whether given the long natural history of these processes, one would anticipate seeing a change in 6 months that would be diagnostically helpful. Stability or decrease in number of calcifications does not reliably exclude malignancy, nor does increase in number necessarily indicate malignancy (11). Clustered calcifications must be categorized when found. Either they appear benign and the patient can continue routine annual screening, or they appear indeterminate or malignant, in which case intervention is indicated. The UCSF data showed a low risk of breast cancer in patients with foci of "tiny round" calcifications—only one in 1,234 cases (6).
The risk of cancer in patients with focal areas of asymmetric fibroglandular density (14.1% of probably benign lesions) was higher than for focal calcifications, but still low (two in 448 cases) (6). This is an interesting observation in view of the fact that this is the finding most likely to produce a false-negative result (12). It would be of some interest to know what types of cancers occurred in these patients, particularly whether they were invasive lobular in type and, again, whether state-of-the-art US might not have prevented these errors.
Of the 17 cancers that occurred in the originally published UCSF series of 3,184 probably benign lesions, 12 (70.6%) occurred in the 589 patients with nonpalpable, noncalcified, well-defined solid nodules. Thus, more than two-thirds of the cancers occurred in a group that composed only 18.5% of the patients. Although the risk for the entire population of probably benign lesions was 0.5%, this group had a risk of 20 per 1,000. This is substantially higher than expected in the average population. This is one group in which, I believe, the alternative of biopsy should be discussed with the patient. The overall risk of 2.0% (12 cancers in 589 patients) is still low enough that surveillance will be considered a reasonable option by many patients. But, one must bear in mind that whatever the level of risk, the potential dire consequences of a wrong decision are always going to be greater for the patient than for the physician.
Most malignant masses will demonstrate growth over time, and morphologic changes suggestive of malignancy often become more apparent as a mass increases in size. Thus, the indicators that a mass may indeed be malignant on follow-up examination are somewhat more reliable than with foci of calcification. However, the issue of 6-month follow-up with masses must also be analyzed in terms of what we know about volume doubling times of breast cancers. At a doubling time of 120 days, a tumor initially 5 mm in diameter would have attained a diameter of 7 mm at 6 months and 10 mm at 12 months; at a doubling time of 240 days, tumor diameter would be 6 mm at 6 months and 7 mm at 12 months. The small increases expected at 6 months could also be produced by minor changes in positioning or degree of compression, making distinction between measurement error and true growth difficult.
Still, it is true that a combination of short doubling time (60 days or less) and larger initial diameter might rarely lead to a diagnostically helpful change at 6 month follow-up. Doubling times are often shorter in younger women. It is for precisely this reason that I believe consideration should be given to biopsy, preferably percutaneous, for solitary noncalcified solid masses over 1 cm in diameter at initial examination in women being screened in their 40s, as opposed to 6 month follow-up.
Support for elimination of the 6-month follow-up examination during the course of a mammographic surveillance protocol is also provided by the small numbers of cancers that were actually identified at 6 months in the UCSF study (6). In the original group, there were only two; in the updated series, six. Thus, over 7,000 women had, at some expense, an additional anxiety-provoking examination to identify a cancer 6 months earlier than it would otherwise have been found in six women, with little or no detriment to those women. The benefit-to-risk ratio is certainly questionable. One also wonders whether high-quality US scans obtained at the outset in the five patients whose cancers were manifest as either mass or density would not have allowed detection on the initial study in any case. Or, how many cancers would have been found if a group of 7,500 women without previously identified abnormality had undergone a 6-month screening, or a second reading at the time of the initial screening?
It is important to remember that stability of a solitary solid mass at the time of a 6-month follow-up cannot be used to reassure a patient that the mass is benign. It is only after several years of surveillance that one can be reasonably certain of benignity (13). Whether patients are followed up at 6- or 12-month intervals, they must be informed that it is the length of surveillance that increases confidence in benignity, not the initial follow-up examination. Six-month follow-up is also not a medicolegal panacea, although many consider it to be. Choosing the wrong lesion for 6-month follow-up (eg, a spiculated mass or malignant-appearing calcifications) provides no medicolegal protection.
The following alternate protocol is proposed. Women with generalized abnormalities, all of similar nature, should have an adequate imaging work-up after their initial abnormal examination. If biopsy is not indicated, they should be followed up annually thereafter.
Patients with localized calcifications, which are benign-appearing on adequate work-up mammography, including magnification views in orthogonal planes, should be followed up annually.
Patients with asymmetric densities with echogenic areas of similar shape and size identified in the appropriate location on work-up US scans should be followed up annually. If a corresponding echogenic focus is not identified, these patients should be given the options described for patients with solitary noncalcified solid nodules as follows: Patients with solitary noncalcified solid nodules should be given the option of mammographic follow-up at 12 months or percutaneous biopsy. The younger the woman, the larger the mass, and the more the mass deviates from the typical US appearance of fibroadenoma (14,15), the more she should be encouraged to verify benignity with percutaneous biopsy. A woman who requests 6-month follow-up because she believes it would make her less anxious should undergo repeat US at 6 months.
To reiterate, a period of 12-month mammographic (and/or US) surveillance for at least 3 years is an alternative to biopsy in the majority of women with probably benign lesions diagnosed after adequate work-up imaging. Encouraging 6-month follow-up for such lesions is unnecessarily anxiety-provoking, costly, and fans the flames of screening skeptics, with little demonstrable benefit to those who comply.
Footnotes
Abbreviation: UCSF = University of California at San Francisco
See also the Viewpoint (pp 11–14 ) and Commentary (pp 19–20 ) by Sickles.
References
Related Articles
This article has been cited by other articles:
|
|
 |
|
  K. Kerlikowske, R. Smith-Bindman, L. A. Abraham, C. D. Lehman, B. C. Yankaskas, R. Ballard-Barbash, W. E. Barlow, J. H. Voeks, B. M. Geller, P. A. Carney, et al. Breast Cancer Yield for Screening Mammographic Examinations with Recommendation for Short-Interval Follow-up Radiology, March 1, 2005; 234(3): 684 - 692. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. B. Mainiero, A. Goldkamp, E. Lazarus, L. Livingston, S. L. Koelliker, B. Schepps, and W. W. Mayo-Smith Characterization of Breast Masses With Sonography: Can Biopsy of Some Solid Masses Be Deferred? J. Ultrasound Med., February 1, 2005; 24(2): 161 - 167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Leichter, R. Lederman, S. S. Buchbinder, P. Bamberger, B. Novak, and S. Fields Computerized Evaluation of Mammographic Lesions: What Diagnostic Role Does the Shape of the Individual Microcalcifications Play Compared with the Geometry of the Cluster? Am. J. Roentgenol., March 1, 2004; 182(3): 705 - 712. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Buchbinder, I. S. Leichter, R. B. Lederman, B. Novak, P. N. Bamberger, M. Sklair-Levy, G. Yarmish, and S. I. Fields Computer-aided Classification of BI-RADS Category 3 Breast Lesions Radiology, March 1, 2004; 230(3): 820 - 823. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. B. Gordon, F. A. Gagnon, and L. Lanzkowsky Solid Breast Masses Diagnosed as Fibroadenoma at Fine-Needle Aspiration Biopsy: Acceptable Rates of Growth at Long-term Follow-up Radiology, October 1, 2003; 229(1): 233 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kerlikowske, R. Smith-Bindman, and E. A. Sickles Short-Interval Follow-Up Mammography: Are We Doing the Right Thing? J Natl Cancer Inst, March 19, 2003; 95(6): 418 - 419. [Full Text] [PDF]
|
|
|
|
|
|
X. Varas, J. H. Leborgne, F. Leborgne, J. Mezzera, S. Jaumandreu, and F. Leborgne Revisiting the Mammographic Follow-Up of BI-RADS Category 3 Lesions Am. J. Roentgenol., September 1, 2002; 179(3): 691 - 695. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Rosen, J. A. Baker, and M. S. Soo Malignant Lesions Initially Subjected to Short-term Mammographic Follow-up Radiology, April 1, 2002; 223(1): 221 - 228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Vizcaíno, L. Gadea, L. Andreo, D. Salas, F. Ruiz-Perales, D. Cuevas, C. Herranz, and F. Bueno Short-term Follow-up Results in 795 Nonpalpable Probably Benign Lesions Detected at Screening Mammography Radiology, May 1, 2001; 219(2): 475 - 483. [Abstract] [Full Text]
|
|
|
|
|
|
F. M. Hall and E. A. Sickles Follow-up of Probably Benign Breast Lesions Dr Sickles responds: Radiology, October 1, 2000; 217(1): 303 - 305. [Full Text]
|
|
|
|
|
|
L. Liberman Percutaneous Imaging-Guided Core Breast Biopsy: State of the Art at the Millennium Am. J. Roentgenol., May 1, 2000; 174(5): 1191 - 1199. [Full Text] [PDF]
|
|
|
|
|
|
E. A. Sickles Commentary on Dr Rubin's Viewpoint Radiology, October 1, 1999; 213(1): 19 - 20. [Full Text]
|
|
|
|
|
|
E. Rubin Commentary on Dr Sickles's Viewpoint Radiology, October 1, 1999; 213(1): 21 - 21. [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
