Regular screening mammography reduces mortality from breast cancer by at least 50% in women 50 to 69 years of age, not by one fifth as estimated by the Canadian Task Force on Preventive Health Care, in the opinion of one public health researcher speaking here at the Union for International Cancer Control World Cancer Congress 2012.
Wilbur Deck, MD, from the National Institute for Excellence in Health and Social Services in Quebec City, Quebec, Canada, took the Canadian Task Force to task for their estimation of the mortality benefit from regular mammography, which he feels is “clearly biased downward” for women 50 to 69 years of age.
In their update of screening recommendations published in 2011, the Canadian Task Force recommends routine screening every 2 to 3 years for women 50 to 69 years of age on the basis of “moderate quality” evidence.
This recommendation is graded as “weak.”
“The task force did not do the actual review of the screening studies. They farmed out what I think should be their core competency, which is to analyze the data. It was the group [to whom the review was farmed out] that estimated that the reduction in mortality from breast cancer screening was 21%,” Dr. Deck told Medscape Medical News. “Even though a lot of people in those studies were never actually screened, the task force never challenged that number.”
As Dr. Deck explained to delegates, there is a large discrepancy between estimates of mortality reduction from mammography, ranging from virtually no benefit to as much as 35%.
Evidence From RCTs
By far the most common evidence used to arrive at these mortality estimates comes from randomized controlled trials (RCTs).
“RCTs show about a 20% mortality reduction [with mammography], but it’s important to point out that this estimate is unadjusted for several major factors,” Dr. Deck explained.
The first major factor is that some of the studies were initiated in the 1960s, when the technology used to screen for breast cancer was far less advanced than the dedicated mammography equipment used today.
A second is what Dr. Deck called “low contrast” in these screening studies.
“When you do an RCT, you compare those who had a trial intervention to a control group that doesn’t get the intervention; in most RCTs, that is approximately true,” he said.
However, in breast cancer screening studies, “it is not approximately true because probably less than 50% of participants are getting regular screening,” Dr. Deck noted. People in the control group can also decide they might benefit from mammography and go out and get screened, thereby diluting the control group.
“This reduces the contrast [between the 2 groups] and, obviously, it reduces how effective mammography will appear to be,” Dr. Deck noted.
The third problem is that breast cancer screening only has an impact over the very long term.
In simulations he himself has done, Dr. Deck calculated that the degree of mortality reduction is underestimated by about half if early mortality data are included in estimates and the data are only followed for about 15 years.
At 10 years, recurrence rates after the initial diagnosis of breast cancer are still high, at approximately 45%, even if a woman has been diagnosed with a small (2 to 3 cm) tumor and only 1 positive node.
In contrast, if a woman has a tumor of 1 to 2 cm and no positive nodes, “the recurrence rate at 10 years is only 17%, so it’s pretty important to detect cancer early,” Dr. Deck said. Indeed, if stage distribution and prognosis at various stages are used to estimate the mortality reduction from breast cancer related to mammography, “you see reductions of between 50% and 60%,” he added.
Similarly, a review of all case–control studies of women who received regular screening and those who did not found a reduction of about 50% in breast cancer mortality with regular screening.
We may be overdoing it in terms of underpromising for mammography.
“It’s good to underpromise and overdeliver, but we may be overdoing it in terms of underpromising for mammography,” Dr. Deck said. “If you look at RCTs and adjust for participation, contrast, and timing, you arrive at a mortality reduction of 50% to 60%; if you look at stage-distribution estimates, it’s about 50% to 60%; and if you look at case–control studies, it’s about 50%. These estimates don’t take into account probable improvements in mortality based on technological advances. I think groups like the Canadian Task Force are probably doing a disservice to people who are considering whether or not mammography might be useful for them,” he explained.
Need to Consider Other Morbidities
Session chair Jean Rousseau, PhD, from the Public Health Institute of Canada, emphasized that older women can have other morbidities that need to be taken into account when considering mammography.
Physicians need to start tailoring screening recommendations to older women on the basis of the presence or absence of potentially important morbidities, he explained. “There are a lot of women in the target age group who are still very healthy and who could benefit from mammography — there is no doubt about this — but women in this age group are more diverse than we tend to think,” Dr. Rousseau told Medscape Medical News. “You have to ask if mammography is the most appropriate strategy for women who have important morbidities and we need to consider women more on an individual basis with respect to their needs.”
Dr. Deck reports being involved with a breast cancer screening program in Quebec. Dr. Rousseau has disclosed no relevant financial relationships.
Union for International Cancer Control (UICC) World Cancer Congress 2012. Presented August 29, 2012.
– Pam Harrison0