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 Google Scholar Google Scholar Articles by Bassett, L. W. Articles by Rimer, B. K. Search for Related Content PubMed PubMed Citation Articles by Bassett, L. W. Articles by Rimer, B. K. Social Bookmarking                            What's this?

Risk of Risk-Based Mammography Screening, Ages 40 to 49

American College of Radiology Task Force on Breast CancerAmerican College of Radiology, Reston, VA

To the Editor: The article by Gail and Rimer¹ describes a method for individual risk assessment for women ages 40 to 49 in order to reduce the number of women being screened in this age group. Unfortunately, their calculations fail to recognize that the vast majority of women who develop breast cancer while in their 40s have none of the risk factors that are outlined by the authors.²

In addition, their assumption that mammographic screening is justified for a woman aged 40 to 49 only if her risk is at least that of a 50-year-old woman with no risk factors is misleading. The screening controversy did not develop as a question of whether the absolute benefit from screening some women aged 40 to 49 is equal to that for screening older women, but rather whether there is a substantial relative benefit for screened versus unscreened women in their 40s. Everyone knew from the onset that because breast cancer incidence rises with age, the absolute benefit from screening younger women would be less than for their older counterparts. On the basis of a reduction in the relative risk of dying from breast cancer found in randomized trials, the American Cancer Society,³ the National Cancer Institute,⁴ and the American College of Radiology⁵ have all endorsed screening mammography beginning at age 40.

The authors base their estimate of benefit only on meta-analyses because screening trials were not intended to be retrospectively segregated into age subgroups. Even when combined in the early years of follow-up, these trials lacked any statistical power to permit analysis of women aged 40 to 49 as a separate group.⁶ The authors concede a 16% mortality reduction from a meta-analysis of all eight randomized clinical screening trials on longer-term follow-up, but they ignore a 26% statistically significant mortality reduction⁷ if the meta-analysis excludes data from the National Breast Screening Study of Canada, which had major flaws in its design and implementation.⁸ The authors fail to point out that for women in their 40s, the five Swedish trials combined show a statistically significant 29% reduction in deaths from breast cancer,^5,7 and two of the Swedish trials individually showed statistically significant mortality reductions of 45% and 35%.^5,7 These reductions are actually greater than the 34% mortality reduction achieved for women age 50 and over in the Swedish Two County trial⁹ and probably would have been substantially higher if women had been screened annually rather than every 18 to 24 months.¹⁰ Obviously, when equal-size populations of different ages are compared, the absolute benefit from screening younger women will be lower because of the lower breast cancer incidence. However, because of the relatively larger number of younger women in the U.S. population, 20% of all breast cancer deaths and 33% of all years of life expectancy lost are due to breast cancers found among women in their 40s.¹¹

The arguments used by Gail and Rimer¹ for screening only high-risk women in their 40s could be applied to women at any age when compared with women who are older because the overall risk of developing breast cancer increases steadily with increasing age. It would be wonderful if the group of women who will develop breast cancer could be accurately identified so that only they would need to participate in screening and if women who will not develop these cancers could be identified equally well so that any "unnecessary" costs and adverse consequences could be eliminated. The problem is that neither of these groups is, as yet, identifiable.

The majority of breast cancers occur among women who have none of the risk factors outlined in this latest model. For example, only approximately 30% of women diagnosed with breast cancer while in their 40s, over the last 10 years at the Massachusetts General Hospital, had a family history of breast cancer—70% did not. Far fewer women had a first-degree relative with breast cancer. Even a smaller percentage of these women fell into other Gail/Rimer high-risk categories. Even the data they reference, but do not describe, suggest that 60% of women in their 40s who develop breast cancer would be overlooked using their model. Furthermore, there are no data that prove that screening only these high-risk women will lead to decreases in deaths, as have been shown from screening all women.

Additionally, the authors' statement that there is a "low rate" of developing breast cancer among women in their 40s is misleading. The fact is that approximately 1.5 women per 1,000 develop breast cancer in their 40s compared with 2.5 women in their 50s and 3.5 women in their 60s. We suspect that most women would see little difference between a one to two per 1,000 risk versus a two to three per 1,000 risk. Moreover, there are more than twice as many women diagnosed with invasive breast cancer while in their 40s than diagnosed with cancer of the uterine cervix at all ages combined.¹²

The authors' concern over the false-positive rate for women aged 40 to 49 does not reflect the facts. The false-positive rate based on abnormal screening examinations (FP1) is virtually the same regardless of age,¹³ ranging between 6% and 8%. The only variables are the cancer detection rate that gradually increases with age and the false-positive rates for biopsy recommendation (FP2) and performance (FP3) that decrease steadily with age and reflect the prior probability of breast cancer in the population. The same arguments made for excluding women aged 40 to 49 from screening because of their false-positive rate could be made for excluding women aged 50 to 59 relative to those age 60 and over, etc.

We strongly support providing women with information that can permit them to make informed decisions. However, not only is the author's use of risk assessment to determine which women in their 40s need to be screened based on false premises, but it is also far too complex to be of practical value. Thus it will only promote misinformation and confusion.

REFERENCES

1. Gail M, Rimer B: Risk-based recommendations for mammographic screening for women in their forties. J Clin Oncol 16:3105-3114, 1998[Abstract/Free Full Text]

2. Seidman H, Stellman SD, Mushinski MH: A different perspective on breast cancer risk factors: Some implications of nonattributable risk. Cancer 32:301-313, 1982

3. Leitch A, Dodd GD, Costanza M, et al: American Cancer Society guidelines for early detection of breast cancer update 1997. CA Cancer J Clin 47:150-153, 1997[Abstract]

4. National Cancer Institute: Statement from the National Cancer Institute on the National Cancer Advisory Board Recommendation on Mammography, March 27, 1997. Bethesda, MD, NCI, 1997

5. Feig SD, D'Orsi DJ Hendrick RE, et al: American College of Radiology guidelines for breast cancer screening. AJR Am J Roentgenol 171:29-33, 1998[Free Full Text]

6. Kopans DB, Halpern E, Hulka CA: Statistical power in breast cancer screening trials and mortality reduction among women 40-49 years of age with particular emphasis on the National Breast Screening Study of Canada. Cancer 74:1196-1203, 1994[Medline]

7. Hendrick RE, Smith RA, Ruttledge JH, et al: Benefit of screening mammography in women ages 40-49: A new meta-analysis of randomized controlled trials. Monogr Natl Cancer Inst 22:87-92, 1997

8. Kopans DB, Feig SA: The Canadian National Breast Screening Study: A critical review. AJR Am J Roentgenol 161:755-760, 1963[Abstract/Free Full Text]

9. Tabar L, Fagerberg G, Chen HH, et al: Effect of breast cancer screening by age: New results from the Swedish Two County Trial. Cancer 75:2507-2517, 1995[Medline]

10. Feig SA: Increased benefit from shorter screening mammography interval for women ages 40-49 years. Cancer 80:2035-2039, 1997[Medline]

11. Shapiro S, Venet W, Strax P, et al: Periodic Screening for Breast Cancer by the Health Insurance Plan Project and Its Sequellae, 1953. Baltimore, MD, Johns Hopkins University Press, 1988, pp 12-13

12. American Cancer Society: American Cancer Society Facts and Figures 1998. Atlanta, GA: American Cancer Society.

13. Kopans DB, Moore RH, McCarthy KA, et al: Biasing the interpretation of mammography screening data by age grouping: Nothing changes abruptly at age 50. Breast J 4:139-145, 1998

National Cancer Institute, Bethesda, MD

In Reply: We appreciate the opportunity to respond to the letter by Bassett et al. Before we do, we wish to state that this letter and our article¹ reflect our views and not the position of the National Cancer Institute (NCI). The NCI position on mammography remains as stated by the National Cancer Advisory Board in 1997.²

Bassett et al mischaracterize many points in our article.¹ Nowhere does that article state as a goal "to reduce the number of women being screened." Rather, the purpose of the article was to indicate how individualized assessment of breast cancer risk and a comparison of risk with that of a 50-year-old benchmark woman without risk factors can be used to facilitate a decision-making and counseling process. In many situations, such a comparison might encourage an otherwise reluctant woman in her 40s to begin screening.

Bassett et al say that we misleadingly state "there is a `low rate' of developing breast cancer among women in their 40s." In fact, we wrote: "What causes many observers to hesitate to recommend universal screening of women in their 40s is the relatively low chance of developing breast cancer in that age group and the high proportion of false-positive mammographic results." Our Table 1 quantitates the extent to which breast cancer incidence rates for women in their 40s are low, relative to a 50-year-old woman.

Bassett et al state: "The authors' concern over the false-positive rate for women aged 40 to 49 does not reflect the facts." In fact, we argued at length (see Discussion and Appendix) that women in their 40s with predicted incidence rates equal to that of a 50-year-old woman without risk factors would have an almost identical positive predictive value (one minus the "false-positive rate") as the 50-year-old woman, even if mammography is a little less sensitive in younger women. Unselected younger women do have a lower prevalence of breast cancer on routine screening,³ however, and hence will experience a lower positive predictive value and higher false-positive rate, as mentioned by Bassett et al.

Bassett et al assert: "The majority of breast cancers occur among women who have none of the risk factors outlined in this latest model." We estimate that 68% of white women in their 40s who would develop breast cancer in the next year would have been recommended for screening under the exact age procedure outlined in our article.¹ This result follows because the screening recommendations are weighted toward older women (Table 4¹) and because the higher-risk women in each year of age are recommended for screening. Our calculations are based on the prevalence of risk factors in the population-based sample of women in the Cancer and Steroid Hormone Study,⁴ on the relative risk features of the model in Gail et al,⁵ on the ratios of age-specific rates in Table 1 of our article,¹ and on the 1990 U.S. census age distribution for white women. A summary of this calculation is presented in Table 1.

Bassett et al said that "the authors' use of risk assessment . . . is far too complex to be of value." During the development of our article,¹ the figures and text were shown to physicians who found the strategy and presentation easy to explain. It remains to be seen whether these tools will come into wider use, but the flow charts eliminate all need for calculation. It would be possible to simplify these procedures further through the development of an interactive computer program.

Bassett et al state that "The authors concede a 16% mortality reduction from a meta-analysis of all eight randomized clinical screening trials on longer-term follow-up, but ignore a 26% statistically significant mortality reduction if the meta-analysis excludes data from the National Breast Screening Study of Canada. . .". The purpose of our report was not to present another critique of the trials in the literature but to offer women and physicians additional information. Actually, we estimated a risk reduction of 17% with 95% confidence interval (13%, 29%) and cited similar estimates from three other sources, including the American Cancer Society. Moreover, we pointed out (Appendix) that there is no statistically significant evidence that this relative reduction for women in their 40s is smaller than the 24% reduction we found for women aged 50 to 74. Thus, we do not dispute that there is evidence of a relative reduction in breast cancer mortality from screening that is comparable to that observed in older women. The issue has always been whether the absolute reduction in breast cancer mortality in younger women is sufficient to overcome the expense of repeated examinations and the anxiety and inconvenience that result from mammograms that yield suspicious results requiring additional mammograms or other diagnostic studies, including biopsies.

It is to assist individual women and their physicians in assessing that trade-off that we put forward a way to compare their risk with that of a benchmark 50-year-old woman for whom there is universal agreement by medical organizations that the absolute survival benefits from screening justify the "costs" of screening. A reluctant 40-year-old woman whose risk factors put her at a risk comparable to that of the benchmark 50-year-old woman might be persuaded by arguments in our article that she stands to gain as much as the benchmark woman and that the weight of medical opinion would therefore strongly favor screening. On the other hand, a 40-year-old woman who was anxious about diagnostic procedures and who had no risk factors would be told her risk and advised that delaying screening is an option. If the woman had little anxiety about diagnostic procedures and was willing to endure some inconveniences to attain even a small reduction in absolute risk, however, the counselor should support the woman's decisions to begin screening.

We wrote: "A general recommendation for screening may be both necessary and appropriate for public health guidelines at the population level." Such a general recommendation will probably save lives and will convince most women of the potential benefits of screening. We disagree with the comment of Bassett et al, however, who imply that aids to individual counseling, such as we proposed, "will only promote misinformation and confusion." Women in their 40s have widely varying risks and concerns, and a blanket recommendation will not suit all of them. Some women will act on the basis of general recommendations. Others want more detailed information about their individual risks. Our article offers women in their 40s a means of comparing their risks of breast cancer and their likely risks and benefits from mammography to those of a benchmark 50-year-old woman for whom there is broad concensus that mammographic screening is beneficial. This strategy is consistent with a growing literature that supports the use of aids to decision making by patients and health care providers, especially in areas where there is controversy.⁶

REFERENCES

1. Gail M, Rimer B: Risk-based recommendations for mammographic screening for women in their forties. J Clin Oncol 16:3105-3114, 1998

2. National Cancer Institute: National Cancer Advisory Board recommendations for women aged 40-49. Rockville, MD, National Cancer Institute, 1997

3. Kerlikowske K, Grady D, Barclay J, et al: Likelihood ratios for modern screening mammography: Risk of breast cancer based on age and mammographic interpretation. JAMA 276:2639-2643, 1996

4. Wingo PA, Ory HW, Layde PM, et alCancer and Steroid Hormone Study Group: The evaluation of the data collection process for a multicenter, population-based, case-control design. Am J Epidemiol 128:206-217, 1988[Abstract/Free Full Text]

5. Gail MH, Brinton LA, Byar DP, et al: Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst 81:1879-1886, 1989[Abstract/Free Full Text]

6. O'Conner AM Tugwell P, Wells GA, et al: Randomized trial of a portable, self-administered decision aid for postmenopausal women considering long-term preventive hormone therapy. Med Decis Making 18:295-303, 1998[Abstract/Free Full Text]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

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 Google Scholar Google Scholar Articles by Bassett, L. W. Articles by Rimer, B. K. Search for Related Content PubMed PubMed Citation Articles by Bassett, L. W. Articles by Rimer, B. K. Social Bookmarking                            What's this?