Researchers at Duke Cancer Institute and the Fred Hutchinson Cancer Research Center say that one in seven cases of breast cancer in the U.S. may be over-diagnosed.

In a new study, the authors claim that the public has a poor understanding of over-diagnosis and the risks associated with it, and that while advances in technologies such as mammography have helped reduce breast cancer-related deaths, they have also caused an increase in over-diagnosis.

Defined as the detection of cancer with mammography that would not be clinically evident in a woman’s remaining lifetime, over-diagnosis leads to unnecessary treatments and stress for patients. Previous studies on the frequency of these cases show different estimates that range from 0% to 54% of the time, with variability coming from diverse definitions of over-diagnosis, study methods and populations.

While health guidelines generally recommend mammography for women who are age 50 and over every two years, the authors say that better understanding over-diagnosis and its risks can help make more informative decisions on the appropriate use of the technology among this patient population.

"Overdiagnosis is defined as the screen detection of a cancer that would not have caused any symptoms or harms during the person’s remaining lifetime. We cannot know what would have happened to a woman diagnosed with breast cancer had she not undergone mammography screening in the first place, and hence we can never directly know whether a screen-detected cancer was overdiagnosed or not," lead author Marc Ryser, assistant professor in Duke’s department of population health sciences and the department of mathematics, told HCB News.

Using data from the Breast Cancer Screening Consortium, Ryser and his colleagues assessed the histories of nearly 36,000 women, including 82,677 mammograms and 718 breast cancer diagnoses. They found that 4.5% of all preclinical cancers were estimated to be non-progressive. They also studied tumor latency period, the time it takes for a preclinical cancer to develop signs and symptoms if not found with screening (the time from preclinical onset to clinical detection).

The authors then performed computer simulations that combined both types of estimates with life tables for mortality by other causes to predict the extent of over-diagnosis. In a program of biennial screening between the ages of 50 and 74, 15.5% of cancers were estimated to be over-diagnosed, with 6.1% made up of slow-growing preclinical cancers and 9.3% of progressive preclinical cancers in women who would have died from another cause unrelated to breast cancer.

Women's Health Homepage