Over the past 30 years, screening mammography has not only identified more early-stage breast cancer, it decreased the incidence of late-stage disease by more than a third, according to a study published online in the journal Cancer.
Dr. Mark Helvie, of the University of Michigan, and colleagues conducted their research to address the question raised by Dr. Archie Bleyer and Dr. H. Gilbert Welch in the New England Journal of Medicine of whether screening mammography meets the basic requirement of decreasing late-stage disease (November 22, 2012, Vol. 367:21, pp. 1998-2005).
Bleyer and Welch used Surveillance, Epidemiology, and End Results (SEER) Program breast cancer incidence data, which included both screened and unscreened women, to demonstrate a 69% increase in localized disease but only an 8% decrease in late-stage disease over a 30-year interval, Helvie's team wrote (Cancer, May 19, 2014).
But Bleyer and Welch did not adjust stage-specific incidence results for the underlying temporal trend of increasing breast cancer that existed before the introduction of widespread screening mammography in the U.S. in the mid-1980s. And to get a correct sense of screening mammography's efficacy, it's crucial to take those temporal trends into consideration, according to Helvie and colleagues.
"Although these underlying temporal trends absent screening mammography may appear small, their impact on expected future incidence over many decades is profound," the group wrote. "Over 30 years, a 1% annual increase results in an incidence increase of 33%, and a 2% annual increase results in an overall incidence increase of 78%."
Adequate adjustments
Drawing on SEER registry data, Helvie's group projected baseline breast cancer incidence rates for 1977-1979 over each stage of disease (early, late, and total invasive) to the period of 2007-2009, using five different annual percentage change (APC) elements taken from the following sources:
- Historic female premammographic screening data from the Connecticut Tumor Registry from 1940 to 1982
- SEER data for women 40 years or older from 1977 to 1982
- SEER data for men 40 years or older from 1977 to 2009 (included because they represented the longest-duration unique SEER population that had never been recommended for screening)
- SEER data for women younger than 40 years from 1977 to1984
- U.K. female cancer registry data from 1975 to 1987
The team adjusted the baseline values for 1977-1979 for APC of 0.5%, 1%, 1.3%, and 2% to estimate an incidence rate for each disease stage (if the underlying APC trend continued for 30 years).
"We compared the projected values with observed values in 2007 through 2009 and reported the change in the number of cases per 100,000 women as well as the percentage change," Helvie and colleagues wrote. "We grouped [ductal carcinoma in situ (DCIS)] and localized breast cancer as early-stage disease and regional and distant breast cancer as late-stage disease. Finally, we compared the projected incidence versus the observed incidence for each APC for invasive breast cancer alone and for all breast cancer."
| Cancers per 100,000 women ≥ 40 years, unadjusted for underlying temporal trends | ||||
| 1977-1979 | 2007-2009 | Absolute change | Percent change | |
| Early stage (both DCIS and local) | 112.9 | 239.6 | 126.8 | 112% |
| Late stage (both regional and distant) | 103.3 | 94.9 | -8.4 | -8% |
| Total invasive cancer | 209.9 | 276.2 | 66.3 | 32% |
| Total breast cancer | 216.2 | 334.6 | 118.4 | 55% |
"When unadjusted for underlying temporal trend increase, these rates demonstrated a marked increase in early-stage disease ... with an associated small decrease in late-stage disease," Helvie and colleagues wrote. "Localized disease was found to increase by 74.7 cases per 100,000 women or 70%, whereas late-stage disease decreased by 8.4 cases per 100,000 women or 8%."
But when the stage-specific rates were adjusted for temporal trends -- specifically at the APC rate of 1.3%, which Helvie's group considers a central estimate -- late-stage breast cancer incidence decreased by 37% over the past 30 years, while early-stage breast cancer detection increased by 48%, according to the authors. In fact, the results of the study may underestimate screening mammography's positive effect, as SEER's breast cancer stage categories include both screened and unscreened women and changes in methods of diagnosis and stage determination, the team wrote.
Because the U.S. lacks a national screening program, mammography screening is actually underused, the group concluded.
"There may be the potential to reduce the incidence of late-stage disease and invasive cancer further with the increased use of screening mammography," Helvie and colleagues wrote.
Preaching to the choir
In a statement released on June 10, the American College of Radiology (ACR) and the Society of Breast Imaging (SBI) lauded the study.
"The findings confirm what breast cancer experts have long known -- that widespread mammography screening has positively impacted the lives of women nationwide," said Dr. Barbara Monsees, chair of ACR's Commission on Breast Imaging.
Early detection allows for early treatment, which is good news for women with breast cancer, said Dr. Murray Rebner, president of the SBI.
"This latest study is another testament to the fact that mammography helps save lives," he said.
