Breast cancer screening: A group of Canadian and American researchers say they’ve found “significant flaws” in the studies that shaped Canada’s modern breast cancer screening guidelines.
In a new commentary, they suggest lives have been lost because of methodology errors in two Canadian trials, which found mammograms for women in their 40s did not reduce death rates from breast cancer.
Seven other trials conducted around the world, however, found the opposite: mammograms for women did reduce mortality rates in that age group.
“It is heartbreaking to know that women’s lives have been lost due to the influence of these flawed studies on screening policies,” Dr. Paula Gordon, radiology researcher and clinical professor, said in a University of British Columbia news release on Wednesday.
“We hope shedding light on these major problems will encourage policymakers to revisit current guidelines.”
The commentary paper was authored by Gordon and colleagues at the Toronto-based Sunnybrook Research Institute, the Ottawa Hospital, the University of Alberta and Harvard Medical School. It was published in the Journal of Medical Screening this week.
The 1980s study, known as the Canadian National Breast Screening Study (CNBSS), had two trial groups, but the paper says most women had a clinical breast exam prior to being placed in the groups.
They should have been placed in the trial groups randomly, but in some cases, their breast exam results influenced which group they were put in, authors claimed, citing staff interviews and whistleblower testimony among other forms of evidence.
More women who were already sick with cancer ended up in the group that received mammograms, they said. Over seven years, more of those women died than in the group with fewer sick women, supporting a study “bias” against the effectiveness of mammograms in preventing death, over routine care and physical examinations.
“Given these significant issues, the trial results are unreliable and should not be used to inform policies on breast cancer screening,” said lead author Dr. Martin Yaffe of the Sunnybrook Research Institute.
Yaffe’s research found that “even a small imbalance” in the assignment of the women who entered the CNBSS with advanced cancer would shift its results away from showing reduced death rates.
A possible reason for the imbalanced trials, authors theorized, is that a nurse with “good intentions” urged that women whose breast exams revealed lumps or other findings be placed in the trial group that would get them a mammogram right away.
Dr. Anthony Miller, who led the original CNBSS, called the commentary’s claims “false” and said there is “good evidence” of the randomization of female study participants in the 1980s. He stood by the initial trials and their findings published in 1992.
“I don’t think it’s a study at all,” said the professor emeritus at the University of Toronto Dalla Lana School of Public Health, in an interview after reading the published paper.
“It’s some people whose jobs depend on a mammography deciding once again, to attack the Canadian National Breast Screening Study because we didn’t find benefits from mammography.”
The CNBSS is the only breast screening study in the world that was able to collect risk factors for breast cancer in all women who participated, he added. He said the commentary paper would “confuse” the public.
“I do not believe mammography screening is beneficial,” he said.
“Mammography is a good tool for diagnosis if women have reasons to be concerned about their breasts, and that’s the way it should be used — as a diagnostical tool.”
Used for screening, mammograms can do “all sorts of peculiar things,” like detect lesions that would not have progressed into anything harmful, he said. Ultimately, they can lead to potentially dangerous “overtreatment.”
Concerns about the study’s findings and its randomization have been reported for years.
In 1997, however, an external investigation into the research “failed to uncover credible evidence” that random assignment of women in trials was “subverted” in any way. Even if subversion existed, said the report, it would have been minimal and “could have had only a trivial effect” on the findings.
In 2019, the Canadian National Breast Screening Study was named in the British Medical Journal’s top five research papers of the decade.
Susan Kinghorn, a medical radiation technologist who worked on the CNBSS, said she was given a list of 10 or 12 study participants at the start of the day who were to receive mammograms.
They were supposed to be random, based on a list of 20 to 24 women altogether, she told Global News, but sometimes the names of the women on her original list would change after physical breast exams.
“I was told I wasn’t to question it, I was just to go ahead and do it because ‘Mrs. Smith needed to have a mammogram,’” she said. “Let’s say there was a Mrs. Jones on the list, later in the day they would then take Mrs. Jones off of the list, and Mrs. Smith has replaced Mrs. Jones.”
Kinghorn always had a total of between 10 and 12 names, and while it didn’t “sound exactly right” she said she was a new graduate and grateful to have a job. When she asked questions, she said her concerns were dismissed and what really mattered was that the patients “got checked out.”
In the end, Kinghorn only completed four to five weeks of her three-month contract on the trial study.
As it stands, the Canadian Task Force on Preventive Health Care does not recommend routine mammograms for women between 40 and 49 unless they have pre-existing conditions or higher-than-average risk.
The Canadian Cancer Society recommends women in that age group talk to their doctors about the benefits of mammograms, while advising women between 50 and 74 to get one every two years.
The study’s influence on policy, said Yaffe, may have contributed to the avoidable deaths of more than 400 Canadian women annually.
“Screening saves lives,” said co-author Dr. Jean Seely, head of breast imaging at the Ottawa Hospital. “There is a 98-per-cent five-year survival rate for localized breast cancer when it is detected early.”
