Background: Screening mammography has reduced breast cancer mortality in women 50 to 74 years of age, but trials involving women older or younger than this age range have been inconclusive. The optimal screening interval (e.g., annually versus biennially) is also unclear. Nevertheless, screening commonly begins for most U.S. women at 40 years of age. Mandelblatt and colleagues reviewed 20 strategies for mammography using varying screening intervals to determine the most effective types of screening.

The Study: The authors chose strategies from the Cancer Intervention and Surveillance Modeling Network of the National Cancer Institute. These strategies were evaluated using six established models that estimate age-specific incidence of breast cancer and mortality trends among a cohort of women born in 1960. The effects of the screening strategies on detecting invasive and in-situ breast carcinomas, tumor size, stage, and growth were then predicted. Effects on mortality and life-years gained by the different strategies were studied, as were potential harms such as false-positive tests, unnecessary biopsies, and overdiagnosis.

Results: All models found that biennial screening from 50 to 69 years of age reduced mortality, maintaining an average of 81 percent (range: 67 to 99 percent) of the benefits of annual screening, and halving the number of mammograms performed. Mortality was further reduced when screening was expanded above or below this age range, but a greater benefit occurred by extending screening to 79 years of age (8 percent further mortality reduction with annual screening versus 7 percent for biennial screening) than by starting at 40 years of age (3 percent reduction with annual or biennial screening, which translated into one additional death averted per 1,000 women receiving annual screening). However, more life-years were gained if screening was started at 40 years of age (median life-years gained per 1,000 women screened annually = 33) than by extending screening to 79 years of age (median life-years gained = 24).

More false-positive results occurred with annual screening and when screening was started before 50 years of age. For instance, annual screening over a 30-year period from 40 to 69 years of age yielded 2,250 false-positive results for every 1,000 women screened, nearly twice the number that occurred with biennial screening. Risk of overdiagnosis (i.e., detecting a latent cancer that would not have clinically surfaced in a woman's lifetime) increased when screening was extended beyond 69 years of age. Biennial screening reduced this risk, but by less than one half the rate of annual screening.

Conclusion: The authors conclude that biennial mammography maintains most of the benefit of annual screening with less risk of adverse effects. Screening biennially from 50 to 69 years of age yields the greatest improvements in breast cancer mortality reduction and life-years gained. Some additional benefit could be achieved by extending the screening years.