The Canadian National Breast Screening Study, published in the Sept. 3 issue of Annals of Internal Medicine, concluded annual mammograms for women age 40 to 49 probably do not help reduce deaths from breast cancer. The researchers said there was a small possibility the screenings could reduce breast cancer deaths by up to 20 percent.
Lead investigator Anthony B. Miller, emeritus professor at the University of Toronto, told United Press International, "We have not found a benefit from such screening."
In contrast, another report released Monday from the United States Preventive Services Task Force, convened by the U.S. Public Health Service, concluded a woman who starts annual mammography in her 40s probably is about 15 percent less likely to die of breast cancer. The study recommends mammograms every one to two years for women age 40 and older. The USPSTF study combined results of eight trials involving a half-million women.
Annual mammograms for women 50 and older are routine practice and not particularly controversial.
Miller, commenting on the USPSTF report, said: "They have produced a number of summary statements that I don't think that one can argue with, but it's the interpretation as to whether it's the mammography or not which is where we differ. They have probably not adequately considered the explanations we provided for the difference in benefit."
Miller's Canadian study reported it included 50,000 Canadian women at 15 sites. Women who received four or five annual mammograms, starting in their 40s, showed the same death rate from breast cancer as did women who did not receive annual mammograms, according to results. After about 14 years of follow-up, the annual mammogram group had 105 deaths from breast cancer compared to 108 among those women who had received only "usual care."
The Canadian volunteers had no history of breast cancer diagnosis. They received a physical breast examination and were instructed in self-exam techniques. The women then were randomly divided into two equal-sized groups, the study said.
One group received subsequent annual exams, including mammograms, a physical breast examination and instruction and evaluation on breast self-exams for five years. Women in the other group were not invited for further visits and were expected to receive their usual care through Canada's universal health system.
The Canadian study concludes: "Four or five annual screenings with mammography, breast physical examination and breast self-examination had not reduced breast cancer mortality compared with usual community care after a single breast physical examination and instruction in breast self-examination."
The annual mammography group during follow-up had 592 cases of invasive breast cancer compared with 552 in the "usual care" group. Annual mammography revealed 71 cases of breast cancer in situ, tumors that usually do not become invasive, compared to 29 found in the usual care group.
Stephen A. Feig, a consultant to the Canadian trial during the 1980s, was critical. He told UPI: "They developed the trial without any input from radiologists or radiophysicists or radiological technologists. As a result of that, as well as poor equipment that they had and poor quality control, the technical quality of the mammograms was extremely poor. ... Their positioning was very, very poor. They weren't doing the proper mammographic views."
Feig, professor of radiology at Mount Sinai School of Medicine in New York, director of breast imaging at Mount Sinai Hospital and a member of the American College of Radiology breast cancer task force, said he resigned from the study because his advice was not followed.
Feig criticized the study design, which included giving a physical breast exam to every woman prior to randomization.
"No other trial has ever done that. No other trial had excess late stage breast cancers in the screening group. No other trial did physical exams before women were placed into screening or control groups," he said. "We believe what happened, is that (at) the screening centers, some well intentioned clerk, they knew the woman had a breast mass, they would say 'Well, you go into the screening group.'"
If such selection occurred, it would increase the numbers of deaths in the annual mammogram group and make it appear there was less benefit to getting annual mammograms than actually existed.
D. David Dershaw, a radiologist and director of breast imaging at Memorial Sloan-Kettering Cancer Center in New York, believes the results of the Canadian study are not valid.
"At the beginning of the study there were 24 women who had clinically palpable advanced breast cancer and among those 24 women, 19 were put into the mammography arm (group) and five of those women were put into the control arm (usual care group). ... That, in fact, managed to skew the study against mammography. It has led people to discard the study because it had a fatal flaw," Dershaw said.
Miller defended the quality of the mammograms and vigorously denied speculation about failure to randomize.
"It's not true. There was no assignment at randomization of women with breast cancer. There were equal numbers (in both groups) in whom the nurse examiners found an abnormality on initial assessment. ... It's a complete misinterpretation of the data. ... There was no assignment of women with cancer. None at all."
Steven N. Goodman, a biostatistician at the Johns Hopkins Stanley Kimmel Cancer Center in Baltimore, told UPI, "The Canadian results should not be viewed in isolation, but rather in combination with the other studies."
David Thomas, an epidemiologist at the Fred Hutchinson Cancer Research Center in Seattle, told UPI, "I think they've adequately addressed all of the criticisms and shown that even if some of the criticisms are valid that the problems of the study are not of such a magnitude that they could explain these negative results."
Connie Lehman, associate professor of radiology at University of Washington Medical Center and director of breast imaging at the Seattle Cancer Care Alliance, told UPI, "We really need to look at this from a multidisciplinary, multifactorial approach and ... think about other methods that might be more helpful in younger women with dense breast tissue."
(Reported by Joe Grossman, UPI Science News, Santa Cruz, Calif.)