Photo by Theresa Naujack
For millions of American women over 40, a yearly mammogram is as routine a ritual as changing the batteries in their smoke detectors. And it has an equally compelling rationale: "Do it and you may save your life."
The validity of this recommendation has come under fire, however, in light of a reassessment of seven large studies evaluating mammography.
The critique, published last fall by two Danish scientists in the medical journal The Lancet, drew agreement in January from the Physicians' Data Query board, a panel of experts that writes information for the National Cancer Institute's Internet site. That group also concluded that several of the studies are too scientifically flawed to be cited in support mammography's lifesaving benefits.
NCI has not changed its recommendations, which say that women, beginning in their 40s, should have mammograms every one to two years because the technique results in a significant reduction in breast-cancer mortality.
But the controversy, and overwhelming media coverage of the topic, has left many women and their doctors asking questions of their own. Does mammography save lives? Is the technique itself dangerous, and does it lead to unnecessary follow-up procedures?
To answer some of these questions and encourage discussion, researchers in the Public Health Sciences Division's Outcomes Affinity Group hosted a panel discussion April 4 entitled, "Why is there a debate over the value of mammography?"
One of the interesting aspects of the debate, said PHS investigator Dr. Scott Ramsey, a session organizer, is that such a vigorous controversy has erupted in the absence of any new data.
"These reevaluations were based on clinical trials conducted over the last 30 years, not on any new studies," he said.
Panelists included Drs. Ruth Etzioni and Noel Weiss, also of PHS, Drs. Ben Anderson, Julie Gralow and Connie Lehman of the University of Washington and Seattle Cancer Care Alliance, Dr. Bruce Porter of First Hill Diagnostic Imaging and Dr. Stephen Taplin of Group Health Cooperative Center for Health Studies. PHS investigator Dr. Nicole Urban served as moderator.
Urban commented on some of the inherent problems associated with evaluating the earlier studies.
"As a cost-effectiveness analyst," she said, "I pay particular attention to control groups. These studies did not have comparable control groups, with each study making different comparisons. That is the fundamental reason why some of the trials reach different conclusions. They were testing different hypotheses."
Although most panelists acknowledged problems with previous studies, many said their professional experience leads them to believe strongly that mammography saves lives by virtue of its ability to detect tumors early.
Lehman, a radiologist who specializes in breast imaging, said that it's clear that earlier tumor detection improves a woman's chance for surviving breast cancer.
"We know that mammography isn't perfect, but currently it's the best tool we have," she said. "Screening is an undertaking where we test the many to find the few. There will be false positives and false negatives, and we need to educate women and their doctors about the risks associated with both. But if we tell women not to have mammograms, then what are we going to recommend to them?"
Anderson, a breast surgeon, said that because it has become increasingly clear that breast cancer is a many-faceted disease, some women clearly benefit from the earlier detection afforded by mammography.
"Whether early diagnosis makes a difference or not, it depends on the cancer," he said. "Some cancers are so slow-growing that we probably don't need to detect them because they won't do much harm. Other tumors are so aggressive that there's almost nothing we can do to stop them.
"But the intermediate class of tumors are the ones where treatment does help, and earlier diagnosis clearly is beneficial in those cases."
Much of the debate about mammography's utility focuses on women under age 40, a group in which the technique, which is essentially a breast X-ray, often fails to detect tumors because younger women have denser breasts than older women. Gralow said that in young women, mammography can fail to detect about 25 percent of cancers.
Ultrasound more effective
For these women, ultrasound, which uses sound waves to detect solid masses in the breast tissue, may provide more effective screening, said Porter, an imaging expert who specializes in development of new diagnostic tools.
"I'd recommend studies to look at high-resolution ultrasound to screen women with dense breasts," he said. "Breast ultrasound has improved significantly in recent years. But reimbursement rates for this technique are so poor that most facilities can't afford the best equipment. We need to invest in good equipment and then train ourselves to use it properly."
Panelists had few concerns about radiation exposure from mammography.
A bigger safety issue is whether the technique leads to unnecessary, invasive or expensive procedures that carry risks of their own. In particular, Urban said, mammography can pick up small, non-invasive tumors known as ductal carcinoma in situ for which aggressive treatment may be unnecessary in many cases.
"Comedo-type DCIS, which is believed by some to be likely to progress if untreated, is often found by mammography but unlikely to be discovered by clinical breast exam," she said.
"A recent analysis of gene-expression profiles in breast tumors suggests that these tumors may behave more like invasive disease than like normal breast tissue. More evidence of this type is needed to enable clinicians to judge which mammography-detected lesions require aggressive treatment."
Harder to quantify is the emotional cost brought on by anxiety over false-positives and their associated treatment. Still, Taplin said a false-positive diagnosis does not appear to deter women from continuing to be screened by mammography.
Panel members were asked to look toward the future and suggest analyses that might better answer questions about mammography.
Etzioni, a biostatistician, urged the design of separate studies to answer two fundamental questions: does screening advance diagnosis of tumors that have a bad outcome, and is treatment of such tumors at an earlier stage effective? She also said that randomized clinical trials may not be the only method for obtaining such data.
"Although clinical trials are the gold standard in terms of evidence for efficacy, there is a vast body of other work on breast-cancer screening," she said. "This includes some well-designed observational studies as well as pathological and genetic studies.
"Our conclusions about the efficacy of early detection in general and mammography in particular should also take into account results of these studies."