Published On January 15, 2014
THE NAME GIVEN TO a certain breast condition—ductal carcinoma in situ—contains a word almost everyone finds terrifying: carcinoma. At an NIH conference on the diagnosis and management of DCIS in 2009, a proposal was made to remove that word from the DCIS name in hopes of eliminating fear and stemming a trend of overtreatment. Though no action was taken, the name remains a point of contention among clinicians. The clue to the problem lies in the rest of the name: In situ, Latin for “in place,” implies that the abnormal cells are staying put—at least for the time being. Indeed, in perhaps 85% of cases, they do; the issue is identifying the remaining 15% that will go on to become invasive.
Barry Kramer director of the National Cancer Institute’s Division of Cancer Prevention, finds it worthy of discussion. “I don’t call it cancer,” he says. “When we call it carcinoma or stage zero cancer, that gives the wrong impression about its average behavior and what we do know about its biology. When you put a frightening term like cancer or carcinoma on a lesion that, on average, doesn’t go on to invade normal tissue, then that can prompt women who are justifiably frightened of the word cancer to have therapy that’s every bit as aggressive as if they had a true invasive cancer.”
But Paul Goss, director of breast cancer research at Massachusetts General Hospital, opposes the name change. “It’s stage zero, but it is cancer,” he says. And Seema Khan, a professor of breast surgery at Northwestern University in Chicago, argues that it’s not what the condition is called that raises difficult issues, but rather the uncertainty about how to manage treatment. Until the standard therapy shifts, she says, removing cancer from the DCIS name would only increase confusion. She asks her patients to refer to DCIS by its acronym and encourages them to take the time to explain to friends and family that it’s an abnormality that doesn’t spread or cause death. “If they internalize that distinction, it will defuse the weight of the word cancer. Until we find a better label that captures the variation of DCIS and is consistent with the treatment, I’m not sure it matters all that much whether it has the word cancer in it or not.”
Shelley Hwang, professor of surgery at Duke University in Durham, N.C., meanwhile, has found that words do make a difference: When she and her colleagues presented nearly 400 healthy women with fictitious DCIS scenarios and three treatment options—surgery, medication or watchful waiting—47% of them chose surgery when DCIS was described as a “noninvasive breast cancer.” But when the description was changed to “abnormal cells,” fewer than a third of the women said they’d opt to have the lesion removed.
It’s a scenario that’s becoming more common. A generation ago, an oncologist might have gone years without encountering DCIS. In the 1970s, DCIS represented fewer than 2% of breast cancer diagnoses, and the condition was usually discovered only after a breast lump, discharge or other symptom raised the alarm. But widespread mammography, able to detect smaller and smaller lesions, has led to a sevenfold increase in the number of new cases of DCIS. Today it accounts for about one in five breast cancer diagnoses, with some 64,000 women diagnosed with the condition last year.
That surge presents a quandary for physicians—and for their patients—because screening’s ability to find these precancerous lesions has outstripped knowledge about how to classify DCIS and how to treat it, says Khan. “We’re probably finding some DCIS lesions we don’t need to find, and then we’re faced with the conundrum of what to do,” she says.
Complicating the situation further is DCIS’s similarity to what might be considered a pre-precancerous condition—atypical ductal hyperplasia. The line between the two is “open to controversy,” according to Dennis Sgroi, director of breast pathology at MGH. Yet a pathologist’s decision about what to call the lesion may make a world of difference to a patient, because ADH gets treated much less aggressively than DCIS, which is handled pretty much like early stage breast cancer. The abnormal cells are cut out through a lumpectomy or mastectomy—the latter only if DCIS is diffuse throughout the breast (and a double mastectomy only if it’s in both breasts). Then the patient may receive radiation or the drug tamoxifen.
Yet for those 85% who have DCIS that will pose little threat, a mastectomy or a lumpectomy may make no sense. But so far, there’s no good way to distinguish the dangerous from the benign, says Kramer. Autopsy studies have examined the breast tissue of women who had never been diagnosed with breast cancer and found that many cases of DCIS had gone undetected. At least some of the lesions may go away on their own, though it’s not clear how or even whether that happens.
This situation echoes what’s going on with other cancers—lung and prostate, among others—in which diagnostic advances lead to earlier detection but offer little clarity about what needs to be treated and what ought to be left alone. “The behavior of a cancer is probably set very early in its course, and invasive cancers that start as DCIS probably show some of their personality in that lesion,” Khan says. At least in the case of DCIS, there’s hope that a large proportion of women and their doctors may soon have better information on which to base treatment decisions. Still, that won’t end the quandary for others, who will continue to have to weigh the risks and rewards of early, aggressive therapy.
TRADITIONALLY, PATHOLOGISTS CLASSIFY cancers by looking at a tumor’s features, such as its size, its grade (a measure of how abnormal the cells appear) and the types of proteins the cells make (such as receptors that may be activated by estrogen, progesterone or other hormones). But with DCIS, none of these factors does much to predict whether cells are likely to turn into an invasive cancer, Khan says. Working to find predictors is Thea D. Tlsty, a molecular pathologist at the University of California, San Francisco, who has pinpointed a two-gene combination that seems to predispose DCIS lesions to develop into cancer. Tlsty and a UCSF colleague, physician Karla Kerlikowske, obtained pathology samples from 70 women diagnosed with DCIS more than two decades ago and tested them for gene mutations that they suspected might make a DCIS lesion likely to progress to invasive cancer. The beauty of this approach was that the researchers knew what had happened to the women with these DCIS lesions, and when the researchers compared the gene test results with patients’ outcomes, “the results were stunning,” Tlsty says.
Every woman whose DCIS biopsy had detected the two-gene combination had gone on to develop cancer, she says. In a larger follow-up study that also considered other genetic markers as well as the size of a lesion and how it was discovered—through mammography or by a physical exam (by either a doctor or the patient discovering a lump herself)—confirmed that DCIS cells with these genetic markers were more likely than the others to become cancerous within eight years. A test for patients that also considers other factors that seem to raise the chances that a case of DCIS will become cancerous—such as method of detection—is under development. If it comes to fruition, Tlsty predicts that it would give actionable information to 50% of women with a DCIS diagnosis.
The other half of DCIS patients, though, would still be left to make tough choices in what continues to be a gray area for physicians. Says John Benson, professor of applied surgical sciences at Anglia Ruskin University in Cambridge, England: “We are overtreating the majority of patients to ensure that we’re not undertreating a minority.” And doing too much, with surgery to remove the DCIS cells, often paired with radiation, tamoxifen (a drug that raises the risk of endometrial cancer) or both, can harm patients’ quality of life and may increase the risk of death, he says. But right now there’s little evidence to support treating DCIS with less aggressive strategies, such as watchful waiting—and a randomized trial to compare such an approach with standard care would require withholding standard treatment from some participants. “We don’t know what happens to lesions that are left completely alone, because we rarely leave them in place once they’re diagnosed,” Kramer says.
Unable to know the ultimate outcome of a DCIS diagnosis, both physicians and patients are very likely to go after a lesion with every tool at their disposal. Hwang says doctors are understandably motivated in part by fear of a bad outcome. “If a patient has an abnormality and it turns into cancer, will my patient hold me liable?” asks Hwang. Doing more than may be warranted, on the other hand, though it may cause real harm, will likely have no such downside. “The patient says, ‘Thank you. You found my cancer and saved my life,’” she says.
Against that backdrop, the rate of voluntary double mastectomy among women undergoing surgery for DCIS rose 188% between 1998 and 2005, and that trend may be continuing, according to Benson, who says that in the past five years he’s seen an increase in the number of women who are asking not only to have the breast containing DCIS removed but also the other breast. In part, that may be influenced by the fact that women can now have immediate reconstructive surgery. But it’s also based on the erroneous belief that a double mastectomy means you’ll never have to worry about breast cancer again, Benson says; in fact, some breast tissue remains.
MEANWHILE, IN TODAY’S ERA of patient-centered care, there’s a push to give patients a greater role in medical decisions—and patients are very likely to choose aggressive treatment, for DCIS as well as for other cancer-related diagnoses. That tendency was documented in a 2005 study in the journal Medical Decision Making, “Cure Me Even If It Kills Me,” which showed that people who received hypothetical cancer diagnoses opted for surgery even in scenarios in which they were told that it would cause more harm than benefit.
And given that push to involve patients, physicians need to do a better job of explaining DCIS to patients, says Joann Elmore, professor of medicine at the University of Washington School of Medicine in Seattle. “We are sometimes clumsy and imprecise in our description of DCIS,” Elmore says. Many women feel they need to have it taken out as soon as possible before it spreads, she says, even though DCIS—by definition—doesn’t spread. Elmore urges patients to take their time, get multiple recommendations, make sure they understand the statistics, and figure out what’s important to them. And she thinks physicians need to try to put potential risks into perspective. For instance, they might tell patients that over the following decade, breast cancer will kill eight of 1,000 women 65 years of age who never smoked—but 25 of those women will die of coronary disease.
A further problem, Elmore says, is that even trained experts may misunderstand the risks. In a study published in 2005 in the Journal of the National Cancer Institute Monographs, Elmore and her colleagues presented 123 radiologists with vignettes about patients coming in for mammography. “These were radiologists who were actively interpreting mammograms, and their estimates of a woman’s breast cancer risk were extremely high—more than 93% of them estimated it as higher than it actually was,” Elmore says.
Fears of doing too little to treat DCIS may also obscure the very real risks of doing too much, says Kramer. “Radiation and tamoxifen are known carcinogens,” he says, and mastectomies can provoke life-altering changes, such as persistent pain, muscle weakness and loss of sensation. Tamoxifen, in addition to raising the risk of endometrial cancer, can cause strokes or deep vein thrombosis.
And doctors, too, may need to step back from their own biases about how to proceed. “The incentives are lined up for physicians to listen when patients say, ‘Just cut it out.’ I’m a surgeon, that’s what I do,” Hwang says, but that may not always be in the patient’s best interest. And while Hwang’s study on patient preferences suggests that women are less likely to select the most aggressive treatments when they understand the noninvasive nature of DCIS, rising rates of prophylactic mastectomies imply that more may need to be done.
“This is an unwanted trend that needs to be halted,” says Goss. But he’s hopeful that new genetic markers under study will allow physicians to sort DCIS lesions into those that need treatment and those that don’t. “I predict that within the next five to 10 years we will resolve this problem,” he says. “The vast majority of women with DCIS will be fine without treatment, and we’ll soon have the markers to identify who those are.”
Dossier
Final Panel Statement from the NIH State-of-the-Science Conference: Diagnosis and Management of Ductal Carcinoma in Situ (DCIS), held Sept. 22-24, 2009, in Bethesda. Summarizes the current evidence and questions regarding DCIS diagnosis and treatment.
“Overdiagnosis and Overtreatment in Cancer: An Opportunity for Improvement,” by Laura J. Esserman, Ian M. Thompson Jr. and Brian Reid, JAMA, Aug. 28, 2013. An editorial arguing that the word cancer should be removed from the DCIS name.
“Communication Between Patients and Providers and Informed Decision Making,” by Joann G. Elmore, Pamela S. Ganschow and Berta M. Geller, Journal of the National Cancer Institute Monographs, 2010. An examination of why informed decision making is difficult to carry out in cases of DCIS and suggestions for ways to communicate information about the diagnosis and treatment of DCIS.
Stay on the frontiers of medicine