Debating breast cancer

As patients become increasingly involved in medical decisions regarding disease prevention and treatment, physicians must become good communicators as well as good clinicians. The importance of risk communication has been underscored by the case of breast cancer, where misunderstandings and medical controversies have arisen over the risks and benefits of treatment and screening. Recently a study caused a stir in both the medical journals and the popular press by questioning whether mammography was beneficial at all.

Dr. Suzanne Fletcher

Dr. Suzanne Fletcher, professor of epidemiology at the Harvard School of Public Health, and Harvard Medical School professor of ambulatory care and prevention at Brigham and Women's Hospital, has written on breast cancer risk and particularly on evaluating mammography. Fletcher recently spoke with HMI World about the current controversies over screening and some of the larger issues of breast cancer risk that physicians need to communicate to women.

In 1997, you gave a presentation at a National Institutes of Health (NIH) consensus conference that was held to determine whether screening mammography should be recommended for women in their forties. The panel said the evidence was inconclusive to recommend screening, and that was a controversial decision. What is the general consensus now?

The panel concluded that you couldn’t come up with a uniform recommendation for all women in that age group, that women should be given the information and decide for themselves. And actually they’re not the only group that has come to this conclusion. The National Alliance of Breast Cancer Organizations says that evidence is insufficient to recommend for or against screening mammography for any group of women. The National Breast Cancer Coalition suggests that each woman and her physician make an individual decision based on her personal risk factors. The American Academy of Family Practice also suggests that women be advised of the risks and the benefits. So communicating risks is beginning to show up in more and more recommendations, both for breast cancer screening and in other kinds of health care activities.

But on the web sites of the National Cancer Institute (NCI)—which is part of the NIH—and the American Cancer Society (ACS), mammograms are recommended for women in their forties and beyond. It seems like women are getting mixed messages.

The NCI has come out with its own recommendations, which is unusual actually. They don’t usually make their own recommendations for a given screening test, but they do in the case of breast cancer. To make it even more complicated, NCI has something called Physician Data Query, or PDQ, on their web site. This has editorial boards of experts in particular fields for all kinds of cancer, and for different aspects of cancer. If you look at the PDQ editorial board that has written up an assessment of breast cancer screening, you’ll find that they’ve backed off their assessment that it works. They say it may work, that it may be beneficial. That’s within the same organization. The same inconsistencies are in the NIH.

Why is there so much confusion over whether mammography works?

In breast cancer there’s always been, for at least a decade if not more, a concern that these screening tests don’t seem to be working for younger women below 50 the way it works when you get older. There’s nothing magic at 50. The fact is that the older you are, up to the age of 70, the better it seems to work. We’ve begun to look beyond 70 and it looks like it does well there too.

That’s nothing new; it’s been a controversy for a long time. What’s happened is, over time researchers have been able to follow the women in trials longer and longer. If you follow the younger women long enough, the data look more like the screening is working, even if it’s working less. Those findings are controversial.

On top of the controversy about the age at which to start breast cancer screening is a brand new one. This new controversy comes from the analysis of two Danish scientists who went back and looked at all these old mammography studies, these randomized trials, and said that many of them have major problems and we shouldn’t be using them for data. They said that only three studies are good, and those three suggest that mammography has no effect. So all of a sudden mammography doesn’t work, period. And this is the major new controversy. It’s very complicated; you have to go through each of those studies and each of their criticisms and carefully look at whether or not you agree. That’s what’s going on in the medical journals right now.

Mammography has also been criticized because it is able to detect small cancers that we don’t necessarily know will become malignant. Is this a case of medical technology outpacing what we know about cancer?

You’ve brought up the other major risk that people talk about, and that’s overdiagnosis. In other words, having a screening test that’s so sensitive that it picks up little lesions that we call cancer but are not acting like cancer; they’re just sitting there. Which one of those will ultimately go somewhere and which ones would sit there the whole life? We don’t know how much of it is overdiagnosis. We’re almost sure that some of it is, but that leaves a woman in a terrible quandary, and her physician as well. Most of us would say: well, if we don’t know, let’s get it out of there.

Other than trying to describe that possibility to a woman, I don’t think we can put any probabilities on that risk yet. And by the way, that’s not just for breast cancer; we’re beginning to see this in a lot of other screening. We’re picking up a lot of little abnormalities and we don’t know how to tease apart which one of these we should be concerned about, and which ones we ought to leave alone.

With these debates happening in the medical community, what should patients be concerned about?

One challenge we’re facing is how to communicate all this to the lay public and women who are interested. What I’m trying to suggest is that you actually have to communicate far more than how well these tests work, because that’s only one piece of a complicated puzzle. It turns out that we may be arguing over the number of angels you can put on the head of a pin in terms of mammography’s benefits. Of all the women undergoing mammography screening, the number who would be "saved" by mammography is very small, whether you believe it is working or not. The problem is that the result is very important—in fact life-saving—for these few women.

Meanwhile there’s all this other information about breast cancer risk that affects more women and that they should be aware of. They should know their risk of developing breast cancer, their risk of dying from it, their risk of dying in general, as well as the benefits and risks of screening. And they should know their risk of experiencing a false positive result, which is much higher than their chance of getting breast cancer. A lot of women over time are called back to their doctor’s office, and it is usually not cancer.

Breast cancer incidence increases much faster with age than breast cancer mortality. For an average woman without breast cancer, the chance of dying from breast cancer within the next 10 years is extremely small if she is young and it rises to about 1% if she is over age 65. The risk plateaus for older women because rates of death from other causes rise sharply. Courtesy of NCI

What are the aspects of risk that women are most unaware of?

The most important breast cancer risk factor for almost all women is age, and women don’t understand the enormous effect of age on risk. Young women worry about it far more than they need to, and older women underestimate their risk. That’s the major thing.

The other thing that women don’t understand is the idea of competing risks. Basically they don’t understand the difference between developing breast cancer and dying of breast cancer. They tend to think that if they get breast cancer they are going to die. And it’s not a death sentence; most women who get breast cancer live. In fact, most women who get breast cancer live even if they were never screened. Our treatments are getting better and better but even back in the 70s most patients lived—and that was before screening.

And meanwhile, even though breast cancer becomes more common as a woman gets older, her risk of dying from something else is obviously much greater than her risk of dying of breast cancer. Women often don’t understand the idea that, as women get older, breast cancer doesn’t kill them nearly so often as heart disease does.

Is some of the fear that younger women have of breast cancer a consequence of too much communication, in the form of raising awareness and promoting screening?

Well, it is possible to take it too far. This whole idea of risk communication— we’re learning how to do it. In fact, the ACS was one of the groups that were communicating the statistic that one in eight women will develop breast cancer. But they’re now deciding they have to reconsider that message because women begin to think that it’s one in eight this year or the next few years, but that number is actually over a whole lifetime. When you look at shorter time periods, you see that for women who are young, the risk of developing breast cancer is very small in the next ten years—1.5 percent at age 40, and still only four percent at age 70. We’re learning that you have to communicate in a way that’s understandable to people.

The other thing that’s really tough, and I don’t exactly know how to do this, is that by and large the public is not particularly numerate, they don’t think in terms of numbers, especially in the case of something as emotionally important as breast cancer. I think we have to give meaningful comparisons, because just the numbers alone make it very hard to grasp.

I also sometimes worry that when we talk about numbers, we could communicate that if it’s a small number it doesn’t matter, and that’s not true. It matters enormously to the person who’s involved. But with screening, the vast majority of people will never have the illness regardless of what we’re screening for. You have to consider those people too, not just the people that you’re hoping to find with screening. The women who are unfortunate enough to have breast cancer, they’re the most important of all, there’s no question. But we can’t ignore the others.