Detection Dilemma

The last time a panel of the National Cancer Institute (NCI) evaluated whether mammograms save lives, the group debated angrily until 2:30 in the morning, saw the resignation of a member who refused to "have my name associated with" its conclusions and received hysterical phone calls accusing them of trying to kill women. The cause of the 1997 furor? The judgment that "breast-cancer mortality is no lower in women who [received mammograms in their 40s] than in controls." That was five years ago. Last week another NCI panel waded into the mammogram wars and arrived at an even more radical position: not only does mammography not prolong the lives of younger women, but there is no persuasive evidence that mammograms for women over 50 make a difference. But this time the scientists were "essentially unanimous," says biostatistician and panel member Donald Berry of M.D. Anderson Cancer Center in Houston. There has not even been any hate mail (yet). And that is as remarkable as the conclusion itself. Says panel member Dr. Russell Harris of the University of North Carolina, "Some groups are now recognizing that mammography is not going to solve the problem of breast cancer."

That doesn't mean women will stop demanding mammograms or doctors will stop recommending them. "I don't believe for a second that we should have people thinking that mammography is not a good thing for women to have in terms of screening," says Dr. Dennis Slamon of UCLA, a developer of the new breast-cancer drug herceptin who believes that mammograms can help with early detection. It's not hard to understand the reluctance to let go of the hope mammograms once offered. But the value of mammograms for younger women has long been in doubt, partly because the X-rays do not peer through the dense breast tissue of premenopausal women very well. And even "early detection" might not be early enough: a tumor detectable by mammogram contains at least 10 million cells. If even one has broken off and entered the bloodstream, it could be seeding distant sites with metastasis. "Women can have tiny tumors and already-disseminated disease," says Karla Kerlikowske of the University of California, San Francisco.

Even as doubts grew about the value of mammography for younger women, data for older women looked good. Those 50 and older who had annual mammograms seemed to have a mortality rate 30 percent lower than women of that age who did not. But now even that beacon of hope has dimmed. The 10-person panel, called PDQ (for Physician Data Query), concluded that the studies supporting that rosy conclusion are flawed. That was also the bottom line of a paper last October in the medical journal The Lancet. One problem: in many of the studies, women volunteered to have or to forgo the procedure, rather than being assigned at random. That means the women who had mammograms might be healthier, or more plugged into the medical system, skewing the pro-mammogram results.

The NCI panel did not conclude that mammograms are worthless but, rather, that there is no evidence they help--a subtle but important difference. "I think there is still probably a small benefit to mammograms for some women," says Harris. Which women? Those with "Goldilocks" tumors, growing not too fast but not too slow. The reason is that slow-growing tumors pose so little danger that, with or without mammography, they are usually curable, typically by lumpectomy, whenever a physician detects them. So skipping regular mammograms would hardly budge survival rates. At the other extreme, aggressive breast tumors are so fast-growing that catching them early may still be too late. In this case, mammograms also make little difference to survival. (Women in their 40s have much faster-growing and more dangerous tumors than older women, for reasons that remain unclear.) Only tumors in the middle, which make up maybe 15 to 20 percent of breast cancers, might benefit from early detection by mammography.

NCI is not yet changing its advice that women begin mammograms in their 40s. But women who have put it off can stop feeling guilty; as long as they have regular checkups, it's unlikely they're putting their health at risk. As for those who get peace of mind from regular mammograms, the downside is small. Mammography uses such low radiation that it poses almost no risk of triggering cancer in healthy cells. The greater risk is detecting something that is not cancer and putting a woman through unnecessary tests and surgery. The greatest risk of all, though, may be the false sense of security mammograms provide millions of women every year.