Two patients, I'll call them Sara and Janine, both learned that they had ductal carcinoma in situ (D.C.I.S.), often referred to as Stage 0 breast cancer. Both underwent lumpectomies in their early 50s. Told that worrisome cells were found, both had another surgery. Then a third surgery was recommended.
Sara decided she would rather live with the risk. Janine had the opposite response. "Let's get this all out," she said.
Each is certain she made the best decision, but can both of them be right?
Each year about 60,000 American women are told that they have D.C.I.S., a cluster of cancer cells currently limited to the milk duct, but which may eventually spread. It used to be rare, but widespread mammographyhas led to the discovery of more of these cancers. The logic of "catching it early" now has a twist: Despite the removal of thousands of D.C.I.S. lesions each year, there hasn't been a drop in the incidence of invasive breast cancers. Some argue D.C.I.S. should not even be called cancer.
Now there's a sense that some women with D.C.I.S. probably derive little benefit from treatment. But while treatment may not be helpful for some, we still don't know enough to confidently say to any individual woman, "Your D.C.I.S. will never progress." We are thus left treating virtually everyone.
Last week a study added to the confusion. The authors used a cancer incidence database to look at the chances of dying or having a recurrence among more than 100,000 women given a diagnosis of and treated for D.C.I.S. The primary finding, and one that can't be overemphasized, is that the chances of dying from D.C.I.S. are quite low, about 3.3 percent at 20 years. Nevertheless, for some groups the risk was more than twice as high.
The most misleading suggestion from the media coverage would be that the study establishes the safety of not getting treatment. It does not.
The study does not compare treatment to no treatment because everyone was treated, as surgery is part of the standard of care. The opposite interpretation would be equally valid: Low mortality rates might instead show that treatment has been working.