PersonalFinance and Health Today

Study Challenges Tie of Estrogen Use To Breast Cancer

Study Challenges Tie of Estrogen Use To Breast Cancer

By TARA PARKER-POPE
April 12, 2006; Page A1

For years doctors have warned that using the hormone estrogen during menopause puts a woman at higher risk for breast cancer. Now data on thousands of women suggest that the warning may have been unnecessary.

Investigators in the federally funded Women's Health Initiative found that using estrogen doesn't increase the risk of breast cancer and may even lower it. Estrogen users were 20% less likely to develop breast cancer after an average of seven years taking the drug than women taking a placebo, according to results being reported today in the Journal of the American Medical Association.

The 20% figure isn't statistically significant and by itself doesn't support the use of estrogen to prevent breast cancer. But the study's data do give much-needed reassurance to millions of women who take estrogen to treat menopause symptoms or cope with the effects of a hysterectomy.

The results also raise questions about the safety of another hormone, progestin, which many women take with estrogen. Doctors started adding progestin to the hormone mix in the 1980s to help avert a form of uterine cancer. But the new data, combined with earlier Women's Health Initiative results, suggest that progestin may be the culprit in raising breast-cancer risk.


"The breast-cancer story seems to be different from what we thought,'' says Marcia Stefanick, the Stanford University professor who has led much of the WHI research.

Like many results from the $750 million Women's Health Initiative, the new estrogen data won't escape controversy. The WHI studies have been plagued by design challenges and differing interpretations of the data. In the latest study, more than half of the 10,739 postmenopausal women in the study stopped taking their pills. Most of the women were also overweight, a factor known to influence breast-cancer risk.

The National Institutes of Health stopped the estrogen study earlier than planned, leaving the researchers tantalizingly short of firmer answers about estrogen and breast cancer. As a result, some researchers say it's a mistake to exonerate estrogen just yet.

"This is just one of these flukes because every other bit of information we have tells you that estrogen increases risk of breast cancer in postmenopausal women,'' says Malcolm Pike, a professor of preventive medicine and a longtime hormone researcher at the University of Southern California's Keck School of Medicine. "The notion that says this one study can turn over every other piece of scientific evidence we possess is nonsense."

The WHI, begun in 1991, is the largest randomized clinical trial of women's health and has studied everything from low-fat diets to vitamin D. Unlike earlier large-scale studies, it randomly assigned women to take hormones or a placebo, providing a scientific way to measure the effects of hormones-taking.

Since the 1960s, women have increasingly used hormones to cope with menopause symptoms such as hot flashes, which are triggered by fluctuating levels of a woman's natural estrogen due to aging. The hormones undoubtedly relieve symptoms for many. The question has always been how the drugs influence the risk of health problems -- particularly the two biggest killers, cancer and heart disease.

The first part of the WHI raised alarms on both counts. In that part, researchers studied 16,608 older women, half taking estrogen plus progestin and the other half taking a placebo. The study was stopped early in July 2002 when safety monitoring showed that the hormone users had more heart attacks and breast cancer. Some researchers challenged the data, citing design questions including the predominance of older women in the study. Still, the breast-cancer finding seemed solid because it was in tune with earlier studies.

After the first part of the WHI was stopped, researchers continued a parallel WHI trial, which compared women who took estrogen alone against those taking a placebo. These women all had undergone a hysterectomy, or removal of the uterus. That meant they weren't at risk for cancer of the endometrium, the lining of the uterus. Since the only reason for taking progestin was to prevent endometrial cancer, these women didn't need it.

Many scientists expected that the breast-cancer risk from estrogen alone would resemble the risk from estrogen plus progestin. That view was based on several important studies in the mid- to late 1990s in which researchers tracked the health habits of women over time without dictating any particular treatment. These observational studies linked estrogen use by itself to a higher risk of breast cancer.

Thus the surprise at the new results: Far from raising breast-cancer risk, estrogen seems to offer a slight protective effect. When an initial version of the findings was published in April 2004 in JAMA, many researchers dismissed the data as a statistical anomaly that was likely due to chance.

After WHI investigators subsequently submitted a more-detailed analysis of the breast-cancer data to JAMA, reviewers for the medical journal continued to raise questions. Hormone use changes the way the breast looks in a mammogram and could make it harder to spot tumors. The reviewers wanted to be sure they were really seeing fewer cases of cancer in the estrogen takers and wanted more details about the mammograms and breast biopsies. All the questions delayed publication until today.

"It's a lot harder to publish something that goes contrary to what people believe,'' says Dr. Stefanick.


At this point, the WHI investigators don't believe women should use hormones to prevent breast cancer. Nonetheless, some additional analyses raise the possibility that estrogen may in fact offer protection. Because 54% of women had stopped taking pills by the study's end, investigators decided to also analyze data from just those women who consistently took their medications. Among these women, breast cancer risk was lowered by 33% -- a trend that is considered statistically valid. When investigators looked at specific cancers, estrogen appeared to lower risk for ductal carcinoma -- the most common type of breast cancer -- by 29%.

Why did earlier studies show opposite results? One answer may be that hormone users in earlier studies were more likely to get mammograms, says Rowan Chlebowski, a WHI investigator and breast-cancer researcher at the University of California, Los Angeles. Doctors would have found more cancer in estrogen users simply because they were looking harder.

In fact, one earlier study supports the new results. A little-noticed 2003 study in the Journal of Clinical Oncology looked at the mammograms of nearly 375,000 postmenopausal women. Women who used estrogen for five years or more were 8% less likely to be diagnosed with breast cancer compared with nonusers.

Still, several quirks in the WHI study population are likely to raise debate. Nearly half the women in the estrogen-only study were obese, so it's possible that the results are less applicable to thin women. The reason: Fat is a source of estrogen, and it may be that overweight women have so much natural estrogen from fat that adding more from pills doesn't make much difference.

Also, 41% of the women in the study had both their ovaries removed, a practice that can lower breast-cancer risk. Women in the WHI studies used Premarin, a complex mixture of estrogens derived from horse urine that may work differently than other forms of estrogen on the market.

The WHI study of estrogen-only users was supposed to continue through 2005 but officials at the National Institutes of Health stopped it a year early in February 2004. Although an independent safety monitoring board had narrowly decided that the study could continue, the NIH overruled it, citing a slight increase in stroke risk among the estrogen users.

In hindsight, the early end to the study made it harder to reach conclusions about estrogen and breast-cancer risk. Dr. Chlebowski says it would be "nicer" to have the data from the extra year. "The investigators wanted to keep the study open," he says, but the decision was the NIH's to make.

NIH director Elias Zerhouni says the agency consulted outside experts as well as the safety board and investigators. He says the data about stroke and blood-clot risks had to take precedence over the still-unclear breast-cancer data, which "was really counter to what we know about estrogens."

"We weighed the risk-benefit and decided it was time to protect the patients rather than look at potential positive data on the breast cancer side," says Dr. Zerhouni. "A trial when you're dealing with healthy people has to have a much lower tolerance for complications."

The WHI estrogen results could fuel new avenues of research into the role estrogen plays in breast cancer. Scientists already know that certain breast cancers thrive in the presence of endogenous estrogen -- the kind made by a woman's body. Medications such as tamoxifen or new aromatase inhibitors help stave off breast cancer either by blocking natural estrogen's effects or interfering with production of it. Yet doctors also sometimes use high doses of estrogen to treat breast cancer in certain patients.

How could adding estrogen to a woman's body possibly reduce her risk for breast cancer? Scientists think a woman's natural estrogen may affect breast-cancer risk differently than estrogen taken in the form of pills, patches and creams. However, researchers say the issue needs much more investigation.

The new results are also likely to make doctors take a harder look at progestin. Its use took off in the 1980s, after women were frightened away from estrogen because of reports that it dramatically raised the risk of uterine cancer. To solve the problem, doctors began prescribing progestin, which was known to blunt the effects of estrogen on the uterus, thereby preventing endometrial cancer.

But in trying to prevent one form of cancer, doctors may have inadvertently raised the risk of another. "It was an accepted dogma that estrogen was the bad guy for the breast and endometrium," says Kent Osborne, director of the breast center at the Baylor College of Medicine in Houston. "Now it's turning out that it's progestin that's bad for the breast and estrogen that's bad for the endometrium."

Recently some doctors have tried giving progestin only a few times a year rather than every month. This may be enough to protect the uterine lining from precancerous changes without exposing women's breasts to the long-term effects of progestin.

"I'd like to see a shift in this country away from using so much progestin," says Hugh S. Taylor, an associate professor at Yale University School of Medicine. "Many ob-gyns are thinking this but they are afraid to do it."

Write to Tara Parker-Pope at tara.parker-pope@wsj.com

posted by MobyGrape @ 7:54 AM