The United States Preventive Services Task Force (USPSTF) recently updated their recommendations for breast cancer screening. These recommendations were based in part on a systematic review that included 8 randomized trials evaluating mammography screening for women ≥ 39 years old. Follow-up ranged from 11-20 years. Most trials were designed to compare the effects of inviting women for screening vs. no invitation rather than to directly compare screening vs. no screening. Invitation to mammography screening was associated with decreased risk of breast cancer mortality in all age strata for women aged 39-69 years (level 2 [mid-level] evidence).
For women aged 39-49 years, the pooled risk ratio (RR) for breast cancer mortality in screening groups was 0.85 (95% CI 0.75-0.96) in 8 trials with 348,219 women. The authors estimate that 1 breast cancer death would be prevented for every 1,904 women in this age group in a screening program. Mammography was also associated with reduced breast cancer mortality in women aged 50-59 years (RR 0.86, 95% CI 0.75-0.99 in 6 trials), with 1 breast cancer death prevented for every 1,339 women recommended for screening, and in women aged 60-69 years (RR 0.68, 95% CI 0.54-0.87 in 2 trials), with 1 breast cancer death prevented for every 377 women. There was no significant difference in breast cancer mortality in women aged 70-74 years in the only trial that included this age group.
The review also examined outcomes per screening round in a cohort of 600,830 women ≥ 40 years old who had mammograms between 2000-2005. Rates of false positive mammograms per 1,000 women screened were 97.8 for ages 40-49 years, 86.6 for 50-59 years and 79 for 60-69 years. In addition, for women aged 40-49 years, 5 biopsies were performed for each cancer detected compared with 3 biopsies per cancer detected for women aged 50-59 years and 2 biopsies per cancer detected in women aged 60-69 years.
The most serious potential harm from screening is “overdiagnosis,” meaning detection of cancers that would never have been symptomatic during the life of the woman. This can lead to unnecessary treatment with surgery, chemotherapy, or radiation. The overdiagnosis rate was estimated to be 1%-10% based on increased rates of cancer detected in screening groups compared to control groups (Ann Intern Med 2009 Nov 17;151(10):727). However, there are many factors in addition to overdiagnosis that may increase the rate of cancers detected through screening (Breast Cancer Res 2005;7(6):266).
The USPSTF now recommends against routine screening for women aged 40-49 years, suggesting the decision to begin screening should be individualized based on each patient’s context and values (grade C recommendation). They recommend screening for women aged 50-74 years (grade B recommendation) and find insufficient data to make recommendations for women aged 75 years or older (grade I recommendation). USPSTF also recommends against clinicians teaching breast self-examination (grade D recommendation) and makes no recommendation about clinical breast exams (grade I recommendation). When mammography screening is done, USPSTF suggests every 2 years instead of annually (Ann Intern Med 2009 Nov 17;151(10):716). It should also be noted that the USPSTF recommendations do not apply to women who may be at increased risk for breast cancer due to factors such as genetics.
The benefits of breast cancer screening were also examined in a recent Cochrane review of 11 randomized trials (including the 8 above) with 616,327 women. There was no reduction in overall mortality associated with mammography (RR 0.99, 95% CI 0.97-1.01 in analysis of 8 trials), but there was a reduction in breast cancer mortality (RR 0.81, 95% CI 0.74-0.87 in analysis of 9 trials) (level 2 [mid-level] evidence). This reduction, however, was not statistically significant in an analysis restricted to 4 higher quality trials (RR 0.9, 95% CI 0.79-1.02). Attendance rates for scheduled mammography ranged from 60%-100% for the first mammogram and 40%-89% on subsequent screenings (Cochrane Database Syst Rev 2009 Oct 7;(4):CD001877).
