Let those biostatisticians slug it out down at the National Cancer Institute – I’m getting my yearly mammograms anyway. Then again, the way things are going, should I?
Last week, the latest panel of experts took a crack at sorting out the decades-old mess about mammograms: Do these X-rays really save lives? Or do they just make us feel like we’re doing something, no matter how mystical, to reduce the risk of dying from breast cancer?
The panel, called the P.D.Q. screening and prevention editorial board, is an independent group of specialists convened by the cancer institute. Its job is to sort out complicated data – the mammogram debate surely fits the bill – and put it up on NCI’s web site (www.cancer.gov).
Its specific mission was to review controversial assertions published last fall as a letter in the British journal Lancet by Danish researchers, Ole Olson and Peter C. Goetzsche, who looked at data from the seven key mammography studies.
The Danes concluded, and last week the P.D.Q. panel concurred, that there is insufficient evidence to prove that mammograms prevent deaths from breast cancer, according to Donald A. Berry, a panel member and chair of the biostatistics department at M.D. Anderson Cancer Center in Houston.
Among the flaws? Assessment of cause of death may not have been done in a “blinded” way in some research, that is, without knowing whether a woman was in the screening group or not. Some researchers did not exclude women with pre-existing cancer at equal rates in mammogram and control groups and other “egregious “mistakes, says Berry.
According to the Danish researchers, two of the seven studies were so flawed as to be completely useless – the New York and Edinburgh studies, both of which found a benefit to mammography. Three other studies – the Two-County Study in Sweden, the Stockholm study and the Goteborg study, all of which found a benefit to mammography – were deemed of poor quality.
So far, that’s five studies, all showing a benefit, all deemed unreliable.
The two studies that the Danes thought were done properly were the Malmo study and the Canadian study. The Malmo study found a slight benefit to mammography. One part of the Canadian study found no benefit at all, and the part that focused on younger women found more deaths in the screening group, though this could have been due to chance.
Needless to say, the P.D.Q. panel’s interpretation of the Danes’ interpretation of the seven key studies is not universally accepted.
Dr. Daniel B. Kopans, director of the breast imaging division at Massachusetts General Hospital and a staunch advocate of mammography calls the Danish analysis and P.D.Q.’s acceptance of it ” a travesty…It’s a mistake to use the nonscience from the Danish study to dissuade women from potentially life-saving screening.”
The five studies the Danes rejected were actually “among the best done,” says Kopans, and the Canadian study, which the Danes applauded, was poorly done. “These guys don’t know what they’re talking about.”
This morass is both big news and same old, same old.
It’s big news because invasive breast cancer is estimated to have hit 192,000 women last year to have killed 40,200. It’s also important because women have fought hard to get mammograms covered by insurance and because many women believe (or have been led to believe) that if they just get mammograms, they’ll either be magically protected against breast cancer, or at least will have a better shot at surviving it if we do. (The first is obviously untrue, the second is what’s up for grabs.)
Among the continuing believers in mammography are the American Cancer Society and, with more ambivalence, the National Cancer Institute.
Dr. Robert C. Young, president of the American Cancer Society, stresses that the death rate from breast cancer has been declining in recent years. (It’s not clear, however, whether this is due to wider use of mammography or more aggressive treatments.)
“With widespread use of mammography,” he argues, the presence of cancer in lymph nodes at the time of initial surgery “has gone from 60 percent in the 1980s to less than 30 percent in the 1990s. And the size of breast cancer lesions on average has gone from 3 centimeters in the 1980s to 2 centimeters now.” Furthermore, he says, the real benefit of mammography may not be seen for another 15 to 20 years.
Dr. Peter Greenwald, director of cancer prevention at the National Cancer Institute, acknowledges that the institute’s recommendations “have not been totally consistent.” But he is sticking by current NCI recommendations: a mammogram every one to two years for women, starting in their 40s, and starting earlier, for women at high risk whose doctors recommend it.
But this whole mess smacks of same old, same old, too. The seven studies now at issue are the same seven studies that researchers have been fighting about for years, particularly with regard to younger women.
At a consensus conference called by NCI in January, 1997, that panel of experts concluded that women in their 40s should make up their own minds about mammography because the survival benefit to younger women is so small.
That turned out to be a politically incorrect decision. The American Cancer Society disagreed vehemently. In February, the National Cancer Advisory Board, which oversees the NCI, got into the act. Then Congress, which oversees the NCI’s budget, weighed in, making it clear that if the NCI did not recommend mammograms for younger women, its budget could be cut, recalls Berry.
Surprise, surprise, the NCI then reversed itself, recommending mammograms for younger women.
All of this might be considered academic squabbling were it not for one thing. Many women dutifully have mammograms in part because they take a “Why not? It can’t hurt” attitude.
But mammography can hurt (and not just the compression during the X-ray). One outcome of mammography can be more testing, some of it painful, most of it, anxiety provoking. The good news is that most suspicious mammograms turn out to be false alarms, but to be sure of this, women put themselves through biopsies and other tests.
Mammograms aren’t exactly foolproof in the other direction, either. In younger women, mammograms miss 25 percent of cancers; in women 50 and over, mammograms miss 15 percent of cancers. That’s bad, too, obviously, because women may delay treatment.
Women’s health advocates have long held a more nuanced view of mammography than lay women, arguing that while mammograms may detect breast cancer earlier than clinical (manual) exams, this is not really early detection, merely earlier.
The latest mess “confirms what many of us have suspected,” says Fran Visco, president of the Washington,D.C.-based National Breast Cancer Coalition. “The evidence behind screening mammography is poor, certainly for women under 50. For too long, we feel, mammography has taken up too much space in the world of breast cancer…we need to look more at how to prevent the disease, how to detect it early, at access to quality care for all women….those are the big issues.”
“Mammography is not the answer to the breast cancer epidemic,” the coalition added in a prepared statement last week. ” In any age group, mortality reduction associated with mammography is less than 50 percent….Although it may be difficult to accept, it is vital that women know the truth about breast cancer screening and the false sense of security it provides.”
Cindy Pearson, executive director of the Washington, D.C.-based National Women’s Health Network, agrees. “It’s better that we know this and are confused than we are kept in the dark like in the old days.
So, ladies, where does this leave us? Among other things, in the same boat as men, who are often urged to have PSA testing for prostate cancer without understanding that this test may lead to more and more invasive interventions, some of which may be damaging or of little survival value.
As for mammograms, the P.D.Q. panel is writing up its conclusions now and expects to finalize them in March. At that point, the NCI may or may not decide to alter its current recommendations.
None of this uncertainty is going to go away. It would probably help if everybody (the media included), obsessed a bit less about breast cancer, terrifying though it can be, and a bit more about statistically greater threats, like heart disease, for which there is ample evidence that improvements in exercise and diet can reduce risk.
Ultimately, the answer is not an endless series of expert panels endlessly re-hashing old data on an obviously imperfect test. The real answer is genuine prevention and cures that work.
Meanwhile, I’m keeping that mammogram appointment.