Yet another good friend told me over the weekend how she had narrowly (in her estimation) escaped death: she had had a mammogram a few months ago, a lump had been detected and deemed suspicious, surgery was scheduled, the lump was removed and found to be malignant. She is now starting the standard course of radiation, and thanks to the mammogram, she was telling me, her life has been saved.

Maybe. But maybe not. Since she can't re-run her life and not have the mammogram, seeing what would have happened if the lump had not been found and her cancer treated, we'll never know.

The trouble with mammograms is that they not only have a very high rate of false positives (detecting a mass that turns out to be benign, not breast cancer)—according to the American Cancer Society, by the time a woman has 10 mammograms she will also have a 50 percent chance of being told, wrongly, that one is suspicious—but also an unknown rate of true positives (the radiologist finds a mass, and it’s found to be malignant) that, if left alone, would not have posed any threat to a woman’s health or life. In today’s issue of the Archives of Internal Medicine, scientists are reporting a study (the journal has made it available for no charge, so read it yourself and print it out for your doctor) that strongly suggests that some of the cancers detected by mammography would have vanished on their own had they not been detected and treated.

For the study, scientists led by statistician Per-Henrik Zahl of the Norwegian Institute of Public Health examined breast cancer rates among 119,472 women age 50 to 64 who had three screening mammograms between 1996 and 2001. They then counted breast cancers among a control group of 109,784 women who were not screened. Not surprisingly, breast cancer rates were higher among screened women than not-screened women. After 6 years, all the women were invited to undergo a mammogram.

Here’s where the surprise came in. Even at the 6-year screening, the incidence of invasive breast cancer was 22 percent higher in the previously-screened group (1,909 vs. 1,564 per 100,000 women) than the control group. Because the incidence of breast cancer among women in the control group (who were no different in terms of their cancer risk than the screened women) was always less than that of the screened group, write the scientists, “it appears that some breast cancers detected by repeated mammographic screening would not persist to be detectable by a single mammogram at the end of six years. This raises the possibility that the natural course of some screen-detected invasive breast cancers is to spontaneously regress. . . . Although many clinicians may be skeptical of the idea, the excess incidence associated with repeated mammography demands that spontaneous regression be considered carefully.”

Cancer that goes away without treatment? The scientists found only 32 reported cases of spontaneous regression of invasive breast cancer, a tiny number for a relatively common disease. But, they point out, “the fact that documented observations are rare does not mean that regression rarely occurs. It may instead reflect the fact that these cancers are rarely allowed to follow their natural course.”

And that is what remains such a mystery: What is the natural course of a cancer, breast or otherwise? With prostate cancer, oncologists have come around to the idea that “watchful waiting” is appropriate for many men: the prostate tumor can sit there safely until the man dies of something else. An oncologist in favor of watchful waiting for breast cancer is as rare as hens’ teeth, and I can’t think of any woman who would agree to that course of (non) treatment.

But it’s a fascinating question, whether some significant fraction of the cancers doctors are finding with more and more screening actually pose any threat or would go away on their own. Clearly, however, mammograms are responsible for “the detection and treatment of cancers that would otherwise regress,” the scientists write.

The value of mammograms—specifically, whether they save lives—has been controversial more ways than you can count. (But see this, this and this.) In an accompanying editorial, Robert M. Kaplan of UCLA and Franz Porzsolt of the University of Ulm, Germany, argue that “despite the appeal of early detection of breast cancer, uncertainty about the value of mammography continues.”

That’s putting it mildly—but only if you include among the "uncertain" scientists who actually evaluate the data rather than women who are convinced that a mammogram saved their life and clinicians who dare not question the value of one of the only tools they have (and can you imagine the lawsuit if a doctor told a woman not to bother with a mammogram, and she developed cancer that was discovered only in a late, incurable stage?). But as Kaplan and Porzsolt point out, “We know from autopsy studies that a significant number of women die without knowing that they had breast cancer. . . . If the spontaneous remission hypothesis is credible, it should cause a major re-evaluation in the approach to breast cancer research and treatment.”

 I'm not holding my breath.