Should We Recommend Screening Mammography for Women Aged 40 to 49?

Introduction

In evaluating the controversy over regular screening mammography for women 40 to 49 years old, it is important to remember that the goal of screening is to reduce mortality from breast cancer. As we will discuss below, the benefit of screening is clear for women age 50 years and over, but evidence is less compelling for women 40 to 49 years old. If screening is beneficial for women over 50 years old, why would it not confer the same benefit for women 40 to 49 years of age?

In this article, we review the efficacy of screening mammography by age, based on evidence from randomized controlled trials, and discuss possible explanations for the differential effect of screening mammography in younger vs older women. We also explore the potential risks of performing widespread screening mammography among a population of young women when the incidence of disease is low.

Basis of the Dispute

In 1987, the National Cancer Institute (NCI), along with the American Cancer Society, recommended screening mammography for women in their 40s despite the lack of evidence showing that such screening reduces breast cancer mortality in this age group. At the same time, several prominent organizations (the American College of Physicians, American Academy of Family Medicine, US Preventive Task Force, Canadian Task Force) did not recommend screening mammography for women in their 40s because of the lack of evidence showing that mammography lowers breast cancer mortality among these women.

In February 1993, the NCI convened an international workshop to reassess whether more recent screening mammography trials demonstrated the efficacy of mammography among women 40 to 49 years old. Based on this workshop, the NCI withdrew its recommendation for screening women in their 40s since there was still no scientific evidence that screening mammography decreases breast cancer mortality among these women [1]. Instead, the NCI advised women in their 40s to discuss with their physician the risks and potential benefits of screening mammography.

At present, many scientific organizations and most countries in the world who conduct screening mammography do not recommend screening mammography for younger women, but there are still groups in the United States that advocate such screening (Table 1). This disagreement among organizations may be due, in part, to differences in requirements for recommending screening. Organizations that do not recommend screening mammography among women in their 40s maintain that the goal of screening mammography is to reduce the number of deaths from breast cancer. They further argue that proof that mammography can detect small breast tumors is insufficient grounds for recommending routine screening in this age group. Rather, there must be scientific evidence to show that detection of breast cancer by mammography reduces breast cancer mortality before making such a recommendation.

Data From Randomized Controlled Trials

Randomized controlled trials are the most unbiased means to assess whether a screening test reduces the likelihood of death in a person who has the disease, and thus, are considered the gold standard when evaluating the efficacy of a screening test. Table 2 summarizes data from eight randomized screening mammography trials. Among women 40 to 49 years old, four of the eight trials reported a nonsignificant increase in breast cancer mortality after 7 to 9 years from the initiation of screening, whereas four found a nonsignificant decrease, indicating a lack of statistically significant benefit or harm from screening mammography. This is in contrast to the data for women age 50 and older, which showed a reduction in breast cancer mortality among all studies (Table 3).

Several meta-analyses have combined data from the randomized controlled trials in order to quantify the overall impact of screening mammography among women 40 to 49 years old. The first meta-analysis by Elwood et al of published data from six of the randomized controlled trials of screening mammography found no reduction in breast cancer mortality in women 40 to 49 years old after 7 years from the initiation of screening [2].

A more recent meta-analysis, which combined data from all eight randomized controlled screening mammography trials, revealed a nonsignificant, 2% increase in breast cancer mortality after 7 to 9 years from the initiation of screening [3]. After 10 to 12 years from the initiation of screening, combined results showed a nonsignificant, 17% (95% confidence interval, -35% to +6% reduction in breast cancer mortality [3].

Pooled data from the five Swedish trials and results from the HIP trial have also suggested a benefit from screening mammography in younger women that does not occur until after 10 years from the initiation of screening [4-6]. A meta-analysis by Smart et al [7] eliminated the Canadian data [8] and included unpublished results from the Gothenborg and two-county Swedish trials. This meta-analysis contained several errors [9];when these errors were taken into account, the meta-analysis yielded results similar to those previously published [5], that is, that there is a delayed benefit, although statistically not significant, of screening mammography in younger women.

What Accounts for the Discrepancy in Screening Efficacy by Age?

It is unclear why the efficacy of screening mammography in women age 40 to 49 years varies with the length of time from the initiation of screening. The findings are most consistent with two possible explanations:

1. The reduction in breast cancer mortality noted after 10 to 12 years from the initiation of screening occurs because, in women who start screening between 40 and 49 years old, breast cancer is diagnosed when the women are 50 years or older, an age at which mammography is known to be efficacious.
2. A proportion of indolent tumors, if detected early among women in their 40s, will result in a delayed reduction in breast cancer mortality.

Continuation of Screening After Age 50--In the HIP study, 85% of breast cancers in women who started screening between 40 and 49 years old were diagnosed between ages 45 and 54, when most women would have become menopausal [10]. Likewise, the majority of women in the Edinburgh and Malmo trials, which also showed a trend toward a delayed benefit of mammography after 10 to 12 years from the initiation of screening, were also probably age 50 or older and postmenopausal when their breast cancer was diagnosed, since the youngest age of women at the start of screening was 45 years old [11,12]. Computer modeling of the Swedish breast cancer screening trial data estimated that most (70%) of the small observed decrease in breast cancer mortality for women 40 to 49 years old at trial entry may be attributable to the continuation of screening after women reach 50 years of age [13].

Differences in Tumor "Aggressiveness"--An alternative explanation for the variation in efficacy of screening mammography with the length of time from the initiation of screening is possible if one accepts the premise that breast cancer is a heterogeneous disease with two basic forms: a "less aggressive" form that mammography detects in time for treatment to be effective and a second form that is more rapidly fatal regardless of early mammographic detection. In women 40 to 49 years old, mammography may detect some less aggressive tumors that are more curable than similar tumors detected clinically, but because they are less aggressive, a reduction in breast cancer mortality among screened women is delayed for up to 10 to 12 years. If this is true, the reduction in mortality among screened women age 50 and older, who tend to have slower tumor growth rates [14], should also be delayed for 10 to 12 years. Yet, mammography reduces breast cancer mortality by about 25% after only 5 to 6 years from the initiation of screening in older women.

This finding suggests that, in women age 50 and older, mammography is able to detect "more aggressive" tumors, since a significant reduction in breast cancer mortality among screened women occurs after only 5 to 6 years from the initiation of screening. It further suggests that mammography detects a substantial proportion of less aggressive tumors, since the cumulative reduction in breast cancer mortality increases over time. However, among women 40 to 49 years of age, the proportion of less aggressive tumors detected by mammography is small, which may account for the marginal, delayed reduction in breast cancer mortality among screened women. This hypothesis is supported by the observation that a greater proportion of small, screening-detected tumors are associated with positive lymph nodes among women 40 to 49 years old when compared to similar screening-detected tumors in older women [15].

Summary of Meta-analysis Results--In summary, based on the results of meta-analyses, there is no reduction in breast cancer mortality among women 40 to 49 years old who undergo screening mammography for 7 to 9 years. It is important to emphasize that if screening mammography is effective in reducing breast cancer deaths among women age 40 to 49 years, the reduction in deaths does not occur for at least a decade following the initiation of screening and appears to be smaller than the reduction observed in women age 50 and older. Furthermore, since it is appears that, among women who start screening in their 40s, the majority of breast cancers are detected by mammography after they reach age 50 or older or become menopausal, it may be possible to wait to begin screening then and achieve the same mortality benefit as would occur if screening were started at age 40.