Screening Mammography: Does Overdiagnosis Overshadow Prevention of Advanced Breast Cancer?

Screening to lower cancer mortality should enable earlier detection of malignancies destined to be fatal while also facilitating early treatment of screen-detected cancers. U.S. investigators analyzed three decades of federal data to assess the long-term effects of screening mammography. Breast cancer rates from 1976 through 1978 (when mammography was uncommon) were used to estimate baseline incidence; data from 2006 through 2008 were used to estimate current incidence. To minimize confounding effects of menopausal hormone therapy, the transitory increase in incident breast cancers from 1990 through 2005 was not included. Models for determining the excess in screen-detected early-stage breast cancer as well as the reduction in diagnoses of late-stage cancer included the “best-guess” estimate, in which the underlying incidence of breast cancer was assumed to rise by 0.25% annually (the known percent change in women who were younger than 40).

Incidence of early-stage breast cancer rose from 112 (baseline) to 234 (current) cases per 100,000 women. During the same 30-year interval, incidence of late-stage disease declined by 8 cases per 100,000. Overdiagnosis (i.e., identification of tumors destined not to progress to advanced disease) attributable to screening mammography affected an estimated 1.3 million women (including >70,000 women in 2008 alone, when overdiagnosis accounted for 31% of tumors identified in women 40 and older). During the study period, breast cancer mortality fell by 28% among women 40 and older and by 42% in women younger than 40, a group in which screening was not prevalent.