Is There a Downside to Elderly Women Undergoing Screening Mammography?

Whether the potential benefits of screening mammography outweigh the harms for elderly women is unknown. Randomized controlled trials of screening mammography either did not include women of age 70 years or older or included too few to provide meaningful results (1,2). The findings reported by Welch and Fisher (3) in this issue of the Journal provide evidence of a potential harm in mammography screening in elderly women. These authors found that additional downstream testing following screening mammography occurs in a considerable number of women of age 65 years or older enrolled in Medicare, the majority of whom do not have breast cancer. For every 1000 women screened, there were 83 abnormal mammogram results, 98 additional tests were performed, and nine cancers were detected. For most women, additional tests included diagnostic mammography (61 women) or breast ultrasound (23 women). Breast biopsy rates were somewhat less common than diagnostic imaging tests with 25 excisional biopsies for every 1000 women screened. The rates of additional diagnostic testing and positive predictive value of mammography were similar to those reported for elderly women by others (4,5). Does downstream testing matter? Each woman differs in her willingness to undergo the possible consequences of a falsepositive mammogram (i.e., psychological stress, invasive follow- up procedures, or morbidity from tests), given the very small probability that she will be one of the women who averts a breast cancer death as a result of routine screening. While some women may easily tolerate the additional tests that are recommended following an abnormal screening result, others may feel considerable psychological stress. Elderly women who are bothered by medical tests, visits to doctors, or the discomfort of undergoing mammography or those who experience substantial anxiety waiting for test results and are willing to accept a small risk of breast cancer might rationally defer screening. Downstream testing would seem particularly important to those elderly women who have little chance of benefiting from screening mammography because existing comorbid conditions make their chance of dying of diseases other than breast cancer much higher. Elderly women with three or more comorbid conditions (i.e., hypertension, diabetes, arthritis, history of myocardial infarction, stroke, respiratory disease, or other types of cancer) are 20-fold more likely to die of a cause other than breast cancer within 3 years, regardless of the stage of the breast cancer at diagnosis (6). Downstream testing also adds to the cost of screening mammography. As much as a third of the costs of a screening program arises from the evaluation of screen-detected abnormalities in women without breast cancer (4,7). Welch and Fisher (3) also evaluated the amount of time it takes for downstream testing to be initiated in elderly women. Additional imaging generally occurred within 20 days of an abnormal screening result, and breast biopsies generally occcurred within 30 days. A short time period of 20–30 days to initiate a diagnostic evaluation is unlikely to have an impact on stage of disease at detection or choice of treatment. However, the waiting time may actually feel considerably longer for women wondering whether or not they have breast cancer and may provoke much anxiety and worry (8,9). Educating elderly women who request or are offered screening mammography that most elderly women (92% of women of ages 65–69 years and 86% of women of age 70 years or older) (3) who have an abnormal screening result do not have cancer may alleviate some of the anxiety and worry. For 40% of women, diagnostic testing occurred at approximately 6 months after the initial screening examination. This likely corresponds to a 6-month follow-up mammographic examination, an examination frequently used to further evaluate probably benign abnormalities (10). Recommending a 6-month follow-up examination to further evaluate probably benign abnormalities leaves a large number of women with diagnostic uncertainty for an extended period of time. Additional research is needed to determine whether the very low yield of small invasive cancers and ductal carcinoma in situ (DCIS) (10) for 6-month follow-up examinations outweighs the costs and anxiety that these examinations may provoke. One potential harm of screening elderly women not addressed in the study by Welch and Fisher (3) is the identification of large numbers of cases of clinically insignificant lesions. Screening mammography tends to discover early cancers that may never have produced symptoms. The best example of this is DCIS. The incidence of DCIS increases with age (11), with 25% of mammographically detected cancers being DCIS among elderly women (5). The natural history of DCIS is unknown (in particular, the natural history of small mammographically detected lesions). Given that the natural history of DCIS is unknown but that some lesions will progress to invasive cancer, the vast majority of DCIS lesions are treated by some form of surgery (11). Identifying and treating DCIS lesions in elderly women are unlikely to have an impact on life expectancy, since, if DCIS progresses, it does so slowly, and the risk of death from DCIS progressing to invasive breast cancer is very low (12). Thus, it was concerning to see the prevalence of mammography use in the eldest women studied (3). Up to 26% of women older than age 80 years obtained screening mammography (3), despite its unproven benefit and likely marginal impact on life expectancy (13). An 80-year-old woman has an average life expectancy of 9 years, and an 85-year-old woman has an average life expectancy of only 6 years (14). For women aged 80 years and older who undergo screening mammography, early detection will increase the rate of surgical treatment of clinically insignificant lesions, with little hope of having an impact on overall mortality, given their short life expectancy and high risk of death from cardiovascular disease. Decreasing the false-positive rate would help reduce the psychological and economic costs of screening mammography. Physicians need to inform elderly women that, if they choose to undergo screening mammography, about 8% will have an abnormal result that will require additional evaluation but that the vast majority (86%–92%) of these abnormal results do not represent cancer (3). Furthermore, about 25% of elderly women will have at least one abnormal result if they are screened regularly over a 10-year period (4,15). Elderly women who request or are offered screening mammography should be informed of the likely consequences of undergoing screening and that the benefits for women of age 70 years or older are unproven.