Although there is general agreement that screening women over age 50 years with mammography reduces mortality from breast cancer, there has been controversy over the effectiveness of mammography in women 40 to 49 years old. Until 1993, the most widely accepted recommendation for screening mammography included baseline mammography between ages 35 and 40, mammography every 1 to 2 years between ages 40 and 50, and annual mammograms after age 50.
In February 1993, the National Cancer Institute (NCI) sponsored the International Workshop on Screening of Breast Cancer and kicked off a vocal exchange on the merits (or lack thereof) of screening women age 40 to 49. Since that workshop, the NCI has withdrawn support for screening of women age 40 to 49, advising instead that patients discuss their individual situations with their physician. This decision has drawn fire from breast cancer advocacy groups, Congress, and various physician groups. Currently, the American Cancer Society, American College of Obstetricians and Gynecologists, and American College of Radiology, among other groups, continue to recommend screening mammography for women in their 40s.
Esserman and Kerlikowske review the results of eight randomized controlled trials and offer explanations for mammography's lack of benefit in women age 40 to 49. They discuss the potential benefits and risks of screening younger women and propose that screening commence at age 50. To strengthen the case for initiating screening at this age, the authors suggest that the delayed benefit in mortality reduction from breast cancer (seen 10 to 12 years after the initiation of screening) in the younger women may be due to detection of tumors after this cadre has reached age 50 or menopause.
Most of us have seen anecdotal cases in which screening mammography turned up unexpected cancers in young women, and we know that mammography can help individual patients. Based on some of the earlier results from studies such as that conducted by the Health Insurance Plan of New York (HIP), the medical community has been strongly encouraging women to follow the established screening guidelines. Mobile vans make mammography more accessible; October is designated breast cancer awareness month; and newspapers and magazines run full-page advertisements extolling the benefit of mammography.
It is not surprising that there is resistance to accept the results from the NCI workshop, which run counter to our expectations. Just because individual women may benefit from mammography, however, cannot be sufficient reason to recommend it to the population at large. The randomized controlled trials seem to indicate that there may be limited, if any, mortality reduction in younger women. Do we now have enough information to make a strong recommendation against screening younger women?
What Can We Infer from the Randomized Controlled Trial Data?
Randomized controlled trials represent the most reliable method for the assessment of mortality reduction from screening. Eight randomized trials of breast screening have been conducted during the past 30 years. Most of the trials showed some relative risk reduction in women age 40 to 50, but the benefit was not obvious until 10 to 12 years after the commencement of screening. The newer Canadian National Breast Screening Study (CNBSS) demonstrated increased mortality in the 40- to 49-year-old age group.
Neither the relative risk reduction observed in the randomized trials nor the increased risk shown in the CNBSS were statistically significant. Because the older trials were not specifically designed to accrue enough younger women to show a significant benefit in this group, and because data regarding the effectiveness of screening in younger women were obtained by retrospective subgroup analyses, one would have expected a low probability of yielding significant benefits . The CNBSS has been heavily criticized for flaws in randomization, patient accrual, and execution [2,3]; it has also been staunchly defended by its executors . Due to its relative short duration of follow-up so far (7 years), its conclusions are tentative at present.
To summarize the situation: Several trials (and meta-analyses) show that, statistically, it is not 95% certain whether or not screening benefits younger women, although the older studies were not designed to show a significant benefit in the 40- to 49-year old subgroup. The jury is still out on the Canadian trial.
Can We Delay Screening Younger Women Until They Reach Age 50 or Menopause?
Since it is uncertain whether screening reduces mortality in the 40- to 49- year-old group, and since any mortality reduction is not manifested for 10 to 12 years after screening, should we simply delay screening this group of women until they reach age 50? Some data regarding age at entry and age at breast cancer diagnosis are available, principally from the HIP study . Data from this study show that, among women who were age 50 or younger at the time of diagnosis of breast cancer, only a small portion of cases of breast cancer were detected by mammography alone . The majority of breast cancer cases in these women were diagnosed either by physical examination alone or a combination of physical examination and mammography.
Data regarding age at diagnosis of breast cancer are not readily available from the other trials. Again, we are in a situation in which enough information simply is not available. To make a recommendation at this point would be premature.
What Do We Really Know About Breast Cancer and Mammography?
Since much of the information regarding screening is inconclusive, let us focus on some of the accepted data. We know, from Surveillance, Epidemiology and End Results (SEER) data, that the incidence of breast cancer rises steadily with age. The younger the woman, the lower is her risk of being diagnosed with breast cancer in the next 15 to 20 years. We know that mammography has a lower sensitivity and lower positive predictive value in young women, for various reasons. As a corollary, many more younger women will have to be screened and possibly subjected to interventional diagnostic procedures for every one life extended . Yet, the younger the woman, the greater is her fear of getting breast cancer and the more inflated is her belief in the value of mammography .
One of the accepted benefits of mammography is its ability to detect small cancers; a questionable benefit is its ability to detect a larger number of ductal carcinoma in situ (DCIS) cases. Esserman and Kerlikowske discuss the issue of DCIS with regard to the heterogeneity of DCIS lesions, their varied natural history, and the outcome of treating women with DCIS. It is unknown whether detection or excision of DCIS leads to either improved quality of life or decreased breast cancer mortality.
What Can and Should We Do?
We find ourselves in the uncomfortable situation of having performed, analyzed, and heatedly argued over eight randomized controlled trials involving approximately 450,000 women (one-third of whom were 40 to 49 years old), and are still unable to offer a firm recommendation to younger women. We have angry advocacy groups, a confused population at large, and media reporting conflicting opinions. What should we do?
Since we do not have firm answers regarding screening, we cannot make any formal recommendations. We can, however, address the misguided sense of risk in younger women. A public education effort regarding age-related risks of breast cancer could be undertaken. Such an educational campaign would benefit not only the younger woman but also the older woman who has a higher risk of breast cancer but is less likely to undergo mammography screening. This information is necessary if the woman is to be a partner, with her primary-care physician, in making the decision regarding screening.
Finally, the decision to continue or cease to recommend screening of women age 40 to 49 must be carefully weighed in the context of available health-care funds. This is a national health policy issue that needs to take into account all of the competing health-care needs and costs.