• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Screening Proves Itself in Breast Cancer Fight

Posted on: Thursday, 27 October 2005, 09:00 CDT

By Gina Kolata

Over the past decade, as researchers watched the death rate from breast cancer plummet, dropping by 24 percent from 1990 to 2000, the nagging question was, Why? Was it mammograms, which can find cancers early, when they may be more treatable? Or was it powerful new drugs to treat cancers after they were found? Both had been tested in clinical trials, but researchers had not known what happened in the real world, where women did not always follow medical advice and doctors did not always use the tests and drugs as carefully as in the studies. And while most expected that drugs would be working, there was a sharp dispute over whether the studies had established that mammograms had much of a benefit outside the ideal and, some said, artificial setting of a clinical trial. Now, for the first time, a new study finds that both contributed to the drop and thereby helps to resolve a debate over the benefit of mammograms that has been among the most contentious in health care. The dispute has drawn in cancer researchers and advocates over questions of how money was being spent and whether opportunities to save lives were being squandered. Some said that mammograms were unlikely to be very effective and that it was a horrendous waste of money and resources to insist that every woman in the country aged 40 and over have a mammogram every year. Others said mammograms were key without screening to find cancer early, the treatments would not be nearly so effective. "The emphasis was always on mammograms, mammograms, mammograms," while treatment was given short shrift, said Fran Visco, director of the National Breast Cancer Coalition. In fact, she said, until recently the federal government was paying for poor women to be screened but not paying to treat them when the tests found cancer. Others, like the American Cancer Society, insisted that annual mammograms were essential. The falling death rate, said Dr. Carolyn Runowicz, a gynecologic oncologist at the University of Connecticut and the cancer society's incoming president, would not have happened without increased use of mammograms to find cancers when they can be treated. The decline in deaths "is due to better screening, better use of screening," she said. The new study, published Thursday in The New England Journal of Medicine, results from an attempt by the National Cancer Institute to sort through the data. It asked seven independent research teams to explain what happened: The death rate was flat from 1975 until 1990, when it was 49.7 per 100,000 women aged 40 to 75. In 2000, it had dropped to just 38.0 per 100,000 women in that age group. In the meantime, the use of mammography in women over age 40 had soared. In 1985, about 20 percent of women had had a mammogram in the past two years. In 2000, the figure was 70 percent. At the same time, chemotherapy and hormonal therapy had come on the scene and their use had spread rapidly. The different groups came up with different estimates for the contribution of mammograms and treatment, but all agreed that both were important. The contribution of mammograms ranged from 28 percent to 65 percent of the effect, depending on the model, with treatment accounting for the rest. To develop their estimates, the groups used national data on the incidence of breast cancer and the death rate each year from breast cancer. They estimated how many women had mammograms each year and how many women with breast cancer had chemotherapy or hormonal therapy or both. They knew how old the women were when they were when they were diagnosed with cancer and the stage of their breast cancer. Then they used data from recent clinical trials to estimate what the effect of chemotherapy, or tamoxifen, or both, or neither would be. To do that, they started curve fitting. Suppose no one had mammograms. That would mean that some women would have their cancers diagnosed later. Would the mortality rates be the same? What if no one had chemotherapy or hormonal therapy? Such statistical analyses led the research groups to their conclusions, estimating the relative effects of screening and treatment as they are used in the United States, on the falling death rate from breast cancer. And that was how the groups estimated the effects of mammography, addressing the most divisive part of the dispute over the breast cancer death rates. It was a disagreement that had lasted for years and that had been a moving target. The questioning of mammograms began in the 1980s when the National Cancer Institute asked how good was the evidence that women in their 40s should have the screening test. It reached a peak in 2002 with the publication of a paper in The Lancet concluding that the benefits of mammograms for women of any age were unclear and it was entirely possible they did not save lives. The PDQ, an independent group that analyzes data for the National Cancer Institute to present to the public, said it could not longer tell women that mammograms prevented breast cancer deaths the benefits of the screening test were uncertain. The group said that the clinical trials of mammograms had not made a strong case that lives were saved. And, they said, just because a cancer is found early does not mean it is curable. To know that a screening test works, you need strong evidence from clinical trials, and the evidence from the mammography trials was just not convincing to them. In response, health organizations including the American Cancer Society, the American Society of Clinical Oncology and the American Academy of Family Physicians took out ads saying that mammograms saved lives and saying they were concerned that the debate was going to erode women's confidence in the screening test. Next, the U.S. Senate held hearings and concluded in favor of mammograms. Tommy Thompson, then secretary of the Department of Health and Human Services, announced at a press conference that mammograms saved lives. In subsequent years, said Dr. Russell Harris, a professor of medicine at the University of North Carolina and a member of the PDQ panel, he and others became convinced that mammography could work, in ideal clinical trial settings in Sweden. But he wondered what its effects were in the United States today. "The question was not whether you could make mammography work under ideal circumstances, but does it work now, in the real world right now?" Harris said. It was a question of particular importance, he said, because treatment was much better than it had been even a decade ago and still improving, which led him and others to ask whether mammography was making much of a difference in reducing the death toll. Some, like Donald Berry, the chairman of the biostatistics department at MD Anderson Cancer Center in Houston and a member of the PDQ panel, were not convinced that mammography was saving lives; they cited what they felt was the ambiguous clinical trial evidence. Berry led one of the seven teams of statisticians whose paper is being published Thursday and is the lead author of the new paper. "This is the first time that a study does it right in asking the effects of screening separate from therapy," Berry added. And now, he says, "my own view has changed a little because of the solidifying of the benefits."

He and others, including Harris, stressed, however, that mammograms have risks as well as benefits and that women should be counseled about screenings. The risks include false positives 90 percent of what looks like cancer turns out not to be. And they include the problem that some Berry estimates 30 percent of cancers that are detected and treated would not have caused a problem if left alone. Some cancers are indolent and never grow much or spread outside the breast. But since no one knows which cancers are dangerous and which are not, doctors treat them all.

Source: International Herald Tribune

More News in this Category