Loading...
The premise of screening mammography is that early detection and treatment prevent late and incurable presentations of breast cancer. If this premise is true, then a rise in incidence of early-stage cancers attributed to mammography should be followed by fewer advanced cancers. Conversely, if early detection is not followed by correspondingly fewer late-presenting cancers, overdiagnosis has occurred: That is, mammography is detecting cancers that are not destined to become clinically important during a patient's lifetime.
To estimate the extent of overdiagnosis, researchers used various U.S. databases to examine breast cancer trends between 1976 (just before screening mammography became widespread) and 2008. During that interval, the annual incidence of early-stage cancers (ductal carcinoma in situ and localized disease) doubled, from 112 to 234 per 100,000 women — an increase of 122 cases per 100,000 women. However, the annual incidence of late-stage cancers during that interval decreased minimally, from 102 to 94 per 100,000 women — a decrease of only 8 per 100,000 women. The excess of early-stage cases identified annually over late-stage cases avoided annually (112 vs. 8 per 100,000 women) suggests substantial overdiagnosis occurred.
Bleyer A and Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med 2012 Nov 22; 367:1998. (http://dx.doi.org/10.1056/NEJMoa1206809)
Comment
Few dispute that early detection carries some risk for overdiagnosis; the question is, “how much?” These authors estimate that more than 1 million U.S. women have been overdiagnosed (and presumably treated unnecessarily) during the past 30 years. They also conclude that most of the improvement in breast cancer–related mortality during those years is attributable to better treatment and not to screening. The authors do acknowledge that their analysis relies on imperfect data — a point their detractors likely will emphasize.