By Jane Hoback
Results of a Canadian study of mammography published in February have reignited the debate over the value of regular screenings and whether they save lives.
The study, published in in the British Medical Journal, compared breast cancer incidence and mortality over 25 years in 90,000 women aged 40 to 59. The women were divided into two groups: those who had mammograms and a control group that only had physical exams.
It found no difference in the death rates from breast cancer in women who underwent annual mammogram screenings and those who didn’t. The study also found 22 percent of cancers found by mammograms qualified as “over-diagnosis” – that means they were not a threat to the woman’s health and did not require treatment such as surgery, chemotherapy or radiation.
Controversy over the study is only the latest round in the debate over the benefits and risks of regular screenings, and the guidelines about the age women should start having mammograms and how often.
“There are arguments on both sides and women are confused. They have all this mixed information,” said Dr. Dev Paul, a breast oncologist with the Rocky Mountain Cancer Centers.
In Colorado in 2012, an estimated 68 percent of women aged 40 and older had a mammogram within the previous two years, and 71 percent of women 50 or older had the screening, according to the Colorado Department of Health and Environment.
The American Cancer Society estimates that 3,780 women in Colorado will be diagnosed with breast cancer this year and 530 will die from the disease.
The cancer society says the consensus of most experts is that breast cancer screening reduces the relative risk of death in women 40 to 59 years old by 15 percent.
The annual cost of mammogram screening in 2010 was estimated to be about $7.8 billion, according to a February 2014 study published in the Annals of Internal Medicine. Annual screening cost an estimated $10.1 billion per year, while biennial screening cost $2.6 billion.
Dr. Ned Calonge, president and CEO of the Colorado Trust, said the Canadian study should be considered in context with eight other studies on mammography over the past few decades.
“People are wrestling with whether we have the right answer yet,” he said.
The Canadian study, which he points out has been reported twice during its 25-year observation period, contains important information “about how much over-diagnosis there is,” Calonge said. “One of five cancers that are detected by mammography is over-diagnosed. That’s saying you do have breast cancer but it doesn’t need to be treated. It will either spontaneously remit or it’s slow-growing and would not otherwise be detected or affect the woman’s health.
“The problem is we can’t tell which ones, so we end up treating them all.”
Study protocols questioned
Critics of the study, however, question the results, arguing the report is flawed in two crucial areas: the quality of the mammograms and the randomization of the patients.
“The quality of the mammograms performed was poor,” said Dr. Wei-shin Wang, assistant professor of radiology at the University of Colorado School of Medicine.
He said the study found that only 32 percent of the cancers were detected by a mammography alone, which meant only 32 percent of the time the quality of the mammogram was acceptable without requiring additional testing.
That was “far below the 60 to 70 percent standard at the time,” Wang said.
In addition, he said the technologists performing the mammograms were not trained in proper positioning and the radiologists had no dedicated training in reading mammograms. “These factors alone can explain the results of the study. You cannot make any conclusions on the efficacy of a certain procedure if it is not done properly.”
Wang also noted that the patients in the study were not truly randomized. Each patient had a physical exam before being placed in one of two sections of the study: those who underwent mammograms and those assigned to the control group who didn’t have the screening.
Wang and other critics suspect that patients who had palpable lumps in their breasts were sent to the mammogram group in the study. “Early in the trial, a statistically large number of advanced cancers were found in the mammography arm,” Wang said. “Also, the five-year survival for breast cancer at the time in Canada was 75 percent, while the control arm of the study had a five-year survival of 90 percent. This difference can only be accounted for by shifting patients with suspicious physical exam findings from the control arm to the mammography arm of the study.”
Paul, the oncologist, said the study included women who had large lumps in their breasts called T3 lesions, which are about 5 centimeters. “Putting somebody with a huge palpable lump that’s a stage 3 breast cancer in the study obviously is going to affect the outcome and skew the survival rates significantly.”
New guidelines coming
The Canadian study will be included with others that will be evaluated by organizations such as the American Cancer Society, the U.S. Preventive Services Task Force and the American College of Radiology when they review their guidelines on mammograms. Reviews by the cancer society and the task force are underway, and updated guidelines are expected later this year.
Guidelines issued by both the American College of Radiology and the American Cancer Society recommend that women 40 and older have a mammogram every year as long as they are in good health. There is no upper age limit. Women who are at increased risk of breast cancer because of family history or a genetic syndrome known as BRCA1 or BRCA2 should have annual mammograms beginning at age 30, according to the guidelines.
Shane Ferraro, spokesman for the American Cancer Society in Denver, said the “findings from the Canadian study are not sufficient to change the recommendation that women get regular mammograms, as they differ from most other clinical trials on mammography.” But it will be considered as part of the society’s 12-person panel review.
Guidelines issued by the preventive services task force in 2009 are still causing waves.
In contrast to recommendations by other groups, the task force does not recommend annual screenings for women beginning at age 40 who are not at increased risk. Instead, it recommends an “individualized decision to begin biennial screening according to the patient’s context and values.” It recommends screening every two years for women aged 50 to 74.
Screenings every other year “preserve most of the benefit of annual screening and cut the harms nearly in half,” the task force said. Those harms include anxiety, radiation exposure, unnecessary treatment and biopsies in women without cancer, as well as false positive results, particularly in younger women.
Risks, benefits weighed
Calonge, the former chair of the U.S. Preventive Services Task Force, said he “can’t guess” what the group will do when it issues updated guidelines. But he thinks its approach continues to be reasonable, based on the evidence of not only the Canadian study but others as well.
“I do think it will bring more clarity into how doctors should discuss the potential harms of mammography with women,” he said.
For Calonge, “over-diagnosis is the real risk. You could have treatment that you don’t need – chemotherapy, radiation, surgery. Women need to know that they have a 20 percent chance of that.”
In addition, “there are a lot of false positives. The younger a woman is, the greater the rate of false positive,” because younger women’s breast tissue is denser, making it more difficult to read mammograms.
“For women who start having mammograms in their 40s, half of them are going to have to have some additional testing for false positive mammograms. Women should know that.”
On the other hand, he said, the benefit is that mortality rates decrease up to 25 percent among women whose breast cancer is detected by mammogram.
“The medical care system should discuss the harms and the benefits with women and let them make their own informed decision,” Calonge said.
Others experts, however, say the benefits of annual mammograms outweigh the risks.
Paul, who treats women who have breast cancer or who are at high risk, sees an unqualified value in women having annual screenings beginning at age 35 to 40, depending on their risk factors.
“I see women dying every day from breast cancer,” Paul said. “I don’t think the risk is that high with the things we do nowadays. Of course there’s always the possibility that the mammogram is abnormal, you work it up, you do a biopsy and it’s benign. You’ve just wasted your time and money and you’ve stressed out about it. We acknowledge the fact that there are false positives. But I don’t think that kills anybody.
“With the women I see, they’d rather find something sooner than later.”
The bottom line for Dr. Kelly McAleese, medical director of the Women’s Imaging Center, is “detecting the breast cancers as small as possible as early as possible.
As far as the issue of over-diagnosis, she said, “the question is, are we finding breast cancers that would really have killed the patient or not? We can’t determine that. We have no way of knowing.”
The Women’s Imaging Center follows the American College of Radiology guidelines, but McAleese also said the issue of whether women aged 40 to 49 should get mammograms every year or every two years “is a reasonable debate. We’re OK with either. It’s not that when you turn 40, you’re suddenly at a higher risk. It’s a slope. It’s more common to find 48-year-olds with breast cancer than 40-year-olds.”
But McAleese and Paul also predict that more sophisticated testing methods may render the debates moot, or at least outdated.
Paul sees promise in digital tomosynthesis, a sort of 3-D mammography that provides a clearer view of the breast and can pinpoint lumps more accurately. He also thinks genetic testing, called BRCA1 or BRCA2, will be “very important” for women at higher risk of breast cancer.
McAleese calls it “risk stratification. It’s not just reading the mammography and saying I see something or I don’t. Breast cancer isn’t just one type of cancer present in one type of exam. Some can be easily seen by mammography. Some can’t and are more readily detected by ultrasounds, for example.
“It means patients bring their medical history and their family history, and then discuss with their clinicians what their best pathways are for detection as well as for prevention. We can look at their age, their risk, their habits and then use the appropriate tools.”
Editor’s note: An earlier version of this story said that Dr. Ned Calonge was a member of the U.S. Preventive Services Task Force. In fact, he was the chair.