Screening Mammography Results in Overdiagnosis of Small Cancers

Screening mammographies do not result in significant changes in deaths due to breast cancer or in the diagnosis of large breast cancers, but they are associated with an increase in diagnosis of small cancers, according to a study published in JAMA Internal Medicine.

Researchers from Harvard University in Cambridge, MA and the Dartmouth Institute for Health Policy and Clinical Practice in Hanover NH, examined the association between screening mammography and the incidence of breast cancer, as well as tumor size and related mortality among women who received a screening mammogram in the previous two years.

The ecological study looked at 16 million women, aged 40 and older, from 547 counties reporting to Surveillance, Epidemiology, and End Results cancer registries during the year 2000, and incidence and mortality were calculated for each county and adjusted to the US population.

The researchers found that 53,207 of the women were diagnosed with breast cancer that year and the women were followed for 10 years. The researchers found a positive correlation between the extent of screening and breast cancer incidence, but not with breast cancer mortality. “An absolute increase of 10 percentage points in the extent of screening was accompanied by 16 percent more breast cancer diagnoses, but no significant change in breast cancer deaths,” the authors wrote. They found that more screening was strongly associated with an increased incidence of small breast cancers (2 centimeters or smaller) but not for cancers larger than 2 centimeters. “An increase of 10 percentage points in screening was associated with a 25 percent increase of the incidence in small breast cancers and a 7 percent increase in the incidence of large breast cancers.”

The authors of the study noted that despite their conclusion that the balance of benefits and harms is likely to be most favorable when screening is directed to those at high risk, they found two features that they called “troubling.” First, although the screening was not associated with a reduction in large breast cancers, they could not tell if this was because the screening did not catch true breast cancers before they grew in size, or if the reduction was because they were concealed by an increase of large overdiagnoses. The second feature was the observed mortality. While screening was not related to reduced mortality during the follow-up period, the researchers wrote that the observed mortality from breast cancer “may be too rare and too noisy to reliably detect the 20 percent reduction at 13 years of follow-up that was estimated in a comprehensive meta-analysis of screening mammography trials.”

In a related commentary, the authors encourage physicians to be cautious about interpreting ecological analyses. “It is well known…that ecological studies provide no information as to whether the people who were actually exposed to the intervention were the same people who developed the disease, whether the exposure or the onset of disease came first, or whether there are other explanations for the observed association,” Joann G. Elmore, MD, MPH, of the University of Washington, Seattle, and Ruth Etzioni, PhD, of the Fred Hutchinson Cancer Research Center, Seattle, wrote.