This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Armstrong, K. Articles by Weber, B. L. Search for Related Content PubMed PubMed Citation Articles by Armstrong, K. Articles by Weber, B. L.

Breast Cancer Screening for High-Risk Women: Too Little, Too Late?

Abramson Family Cancer Research InstituteUniversity of Pennsylvania Cancer CenterPhiladelphia, PA

BREAST CANCER screening is an important tool in reducing mortality from breast cancer in all women. However, screening has limitations. Even with high levels of compliance, interval cancers remain a problem, and 30% to 40% of women diagnosed with breast cancer still die of their disease. Increasing the frequency and resolution of screening in all women might reduce breast cancer mortality, but it would carry significant economic and other costs. In an era of limited health care resources, we need to carefully consider how best to balance the potential benefits of individual screening against health care and personal costs. Recognizing the importance of breast cancer risk in making screening recommendations is an important first step toward reaching this balance.

Within the past 10 years, work defining the basis of genetic susceptibility to breast cancer in general and the identification of BRCA1 and BRCA2 specifically has greatly enhanced the accuracy of breast cancer risk prediction. The subsequent diffusion of risk prediction tools, such as genetic susceptibility testing and the National Cancer Institute’s "Risk Disk," have resulted in the identification of many women at increased risk of breast cancer.¹ Although this information has implications for many areas of medical care, its potential impact is particularly great in the area of cancer screening. In theory, the ability to stratify breast cancer risk allows intensive screening regimens to be targeted to women at high risk who are most likely to benefit, with more limited screening for women at lower risk. This approach both reduces complications associated with low-risk biopsies and saves health care resources. As with most potential improvements in health care, particularly those that involve difficult choices, the challenge is translating this approach into practice.

The study by Brekelmans et al² in this issue of the Journal of Clinical Oncology offers several important insights into breast cancer screening in women at increased risk of breast cancer. This retrospective study analyzed the incidence and characteristics of screen-detected and interval breast cancers among 1,198 women who participated in a high-risk breast cancer family clinic in the Netherlands between 1978 and 1999. In this clinic, breast cancer screening recommendations included monthly self-examination, biannual clinical examination, and annual mammography, with the optional addition of magnetic resonance imaging (MRI) beginning in 1995 for women with BRCA1 or BRCA2 mutations or very dense breasts. For the purposes of this study, women were stratified into three risk groups: known carriers of a BRCA1 or BRCA2 mutation (again, beginning in 1995), high-risk women with an estimated lifetime breast cancer risk of 30% to 50%³ or belonging to a breast/ovarian cancer family without a defined mutation status, and moderate-risk women with an estimated lifetime breast cancer risk of 15% to 30%.³ Overall, with a median follow-up of 3 years, 26 cancers were detected by screening and nine interval cancers were diagnosed. As might be expected, rates of both screen-detected and interval cancers were highest among the 128 BRCA1 and BRCA2 mutation carriers, intermediate in the 621 high-risk women, and lowest among the 449 moderate-risk women. Similarly, the sensitivity of screening (defined as the ratio of screen-detected to screen-detected plus interval cancers) was lowest in BRCA1 and BRCA2 mutation carriers (56%), intermediate in the high-risk group (78%), and highest in the moderate-risk group (100%). Although only a third of invasive cancers were node-positive overall, more than half of invasive cancers diagnosed in BRCA1 and BRCA2 mutation carriers, and in women under 40, were node-positive.

What do these results tell us about breast cancer screening for women at increased risk of breast cancer? The relationship between breast cancer risk and the rate of cancer detection at screening supports many previous studies. In addition, these data also support the epidemiologic principle that screening is both most useful and most cost-effective in groups with the highest number of expected events. Most importantly, these results provide a rationale for tailoring screening recommendations to breast cancer risk. As is both intuitive and has been demonstrated, routine screening is most likely to identify cancers in women who have the highest absolute risk of breast cancer and least likely to identify cancers in women who have the lowest absolute risk. In this study, rates of cancer detection were 10-fold higher among BRCA1 and BRCA2 mutation carriers than among women at moderately increased risk of breast cancer—almost exactly paralleling the relative risk in these groups.

Although this study did not examine false-positive rates, many prior studies have shown that the probability of a false-positive test result increases as the expected rate of true-positive test results decreases in the group being screened. In one analysis of 40- to 49-year-old women, biopsy after an abnormal screening mammogram was three times more likely to detect breast cancer among women with a family history of breast cancer than among women without a family history (13% v 4%).⁴ By bringing together recent improvements in our ability to assess breast cancer risk with outcomes of screening in high-risk women, the study by Brekelmans et al² emphasizes the substantial impact of breast cancer risk information on the risk-to-benefit ratio of screening and offers a glimpse into the future of cancer screening recommendations.

Finally, Brekelmans et al² observed a substantial risk of interval cancers in BRCA1 and BRCA2 mutation carriers undergoing annual screening, suggesting that current approaches may be insufficient in this very high-risk group. The implications of high rates of interval cancer diagnosis for screening recommendations in women with BRCA1 and BRCA2 mutations depend on the reason for the interval cancers. Interval cancers may occur because mammography was unable to identify a malignancy that was present at the time of screening, either because the surrounding breast tissue was too radiodense or the tumor was insufficiently radiodense. If so, the use of higher-resolution modalities, such as MRI, may provide important benefits. However, interval cancers also may occur because highly proliferative cancers are more likely to reach the limit of clinical detection between screenings. This latter explanation is supported by the previous finding that 70% of BRCA1-associated breast cancers are grade III lesions with significantly higher mitotic counts and more nuclear pleomorphism than sporadic breast cancers from age-matched controls.⁵ Furthermore, interval cancers are more likely to be high grade in women between the ages of 40 and 49 than in women over the age of 50.⁶ If interval cancers arise as a result of the increased growth rate of BRCA1- and BRCA2-associated cancers, the greatest benefit may arise from increasing the frequency of screening, as opposed to changing to a higher-resolution modality.

Despite the intriguing results, this study provides only the first step toward the development of evidence-based guidelines for screening of BRCA1 and BRCA2 mutation carriers and other high-risk women. Rates and characteristics of screen-detected and interval cancers are surrogate end points and may not translate into differences in breast cancer mortality, although the fact that more than half these women had positive nodes is concerning. Furthermore, with a retrospective analysis, it is difficult to tease out the impact of individual components of the screening regimen, improvements in screening technologies over the period of the study (1978 to 1999), and individual patient characteristics, such as age and prior cancer diagnosis. Prospective studies of alternative screening regimens are clearly needed, bearing in mind the findings of the study by Brekelmans et al.² These data suggest that the most effective regimen for women with documented mutations in BRCA1 and BRCA2 may in fact come from the combination of more frequent screening and the use of high-resolution modalities, such as MRI.

REFERENCES

1. National Cancer Institute: Breast Cancer Risk Assessment Tool. Available at: http://cancernet.nci.nih.gov/bcra_tool.html

2. Brekelmans CTM, Seynaeve C, Bartels CCM, et al: Effectiveness of breast cancer surveillance in BRCA1/2 gene mutation carriers and women with high familial risk. J Clin Oncol 19: 924-930, 2001[Abstract/Free Full Text]

3. Claus EB, Risch N, Thompson WD: Autosomal dominant inheritance of early-onset breast cancer. Cancer 73: 643-651, 1994[Medline]

4. Kerlikowske K, Grady D, Barclay J, et al: Positive predictive value of screening mammography by age and family history of breast cancer. JAMA 270: 2444-2450, 1993[Abstract]

5. Breast Cancer Linkage Consortium: Pathology of familial breast cancer: Differences between cancers in carriers of BRCA1 and BRCA2 mutations and sporadic cases. Lancet 349: 1505-1510, 1997[Medline]

6. Feig SA: Determination of mammographic screening intervals with surrogate measures for women aged 40-49. Radiology 192: 311-314, 1994.

This article has been cited by other articles:

				E. A. Morris, L. Liberman, D. J. Ballon, M. Robson, A. F. Abramson, A. Heerdt, and D. D. Dershaw MRI of Occult Breast Carcinoma in a High-Risk Population Am. J. Roentgenol., September 1, 2003; 181(3): 619 - 626. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Armstrong, K. Articles by Weber, B. L. Search for Related Content PubMed PubMed Citation Articles by Armstrong, K. Articles by Weber, B. L.