It takes a village, or so they say, to raise a child.
Well, it’s beginning to take a whole village – and a high-tech one at that – to sort out the risks and benefits of hormone-replacement therapy.
Luckily, there is such a village. It’s at New England Medical Center in Boston, where Drs. Nananda Col, John Wong, and Stephen Pauker in the Division of Clinical Decision Making have created a mathematical model to see how the benefits of hormone therapy – such as reduced risk of osteoporosis – stack up against the risks, including breast cancer, for a hypothetical 50-year-old woman at average risk of both diseases.
But their endeavor has its limits. Chief among them is that the estimates they feed into their computers come from observational studies, and not randomized, double-blind, placebo-controlled trials, the “gold standard” of medical research.
That means their data are derived from women who chose to take hormones at menopause.
In general, women who choose hormones are better educated and have better health habits than women who don’t, which can skew the results of observational studies, noted Dr. JoAnn Manson, an endocrinologist and chief of preventive medicine at Brigham and Women’s Hospital in Boston.
A better sense of the true risks and benefits of hormone therapy is expected in 2005, when the results of the massive, $625 million Women’s Health Initiative are in.
In that trial, which involves 27,000 women at medical centers around the country, volunteers are randomly assigned to take hormones or placebo for an average of nine years. Since the women who take hormones and those who don’t are otherwise similar, differences in subsequent disease risk are probably due to hormones – or lack of them.
But, for many women, five years is too long to wait for guidance, so we asked Col to use her model to run projections of risks and benefits 10 and 20 years from now for a 50-year old woman at average risk of breast cancer, hip fracture, uterine cancer, and coronary disease (defined as nonfatal heart attack, clogged arteries, chest pain or sudden cardiac death).
Because models only work if one plugs in certain assumptions, Col decided that, to be as representative as possible, our 50-year-old would have a total cholesterol of 239 milligrams per deciliter and systolic blood pressure (the top number on a blood pressure test) of 134 millimeters of mercury.
In addition, instead of flipping a coin to decide whether our heroine should have the health risks of a smoker or nonsmoker, and of a diabetic or nondiabetic, Col gave her 44 percent of the disease risks that go with smoking and one-eighth of those due to diabetes. Col also assumed the woman had no family history of heart disease or osteoporosis.
Lastly, because 20 percent of the population has a mother, sister or daughter with breast cancer, Col assumed our hypothetical woman had a 20 percent chance of such a family history, too. She also assumed there was a 28 percent chance that our woman had one breast biopsy and that she had her first child between ages 25 and 29.
Col then calculated the woman’s chances of getting breast cancer, uterine cancer, a hip fracture or heart disease in the next 10 and 20 years, and the odds of her dying from those diseases under different scenarios – no hormone replacement therapy, combination therapy with two hormones (estrogen and progestin) for 10 or 20 years, or estrogen alone for 10 or 20 years. (Progestin is added to hormone therapy to protect the uterus from cancer in women who have not had a hysterectomy.)
The data that Col fed into her computer was derived from several large observational studies, including the Nurses’ Health Study and a major study on breast cancer risk published recently in the Journal of the National Cancer Institute.
In these studies, women took the standard dose of estrogen (usually 0.625 milligrams a day of Premarin), and if they took progestin, it was 5 to 10 milligrams a day of Provera.
Col also fed into her model the risks and benefits associated with a newer hormonal therapy called raloxifene (Evista) instead of estrogen. (Raloxifene’s advantage is that, unlike estrogen, it can lower the risk of osteoporosis without raising the risk of breast or uterine cancer.) She did likewise with alendronate (Fosamax), a non-hormone drug that protects the bones and has no known effect on breast cancer, uterine cancer or heart disease.
The results were intriguing. Ten years from now, our hypothetical woman’s risk of getting breast cancer was clearly lower (one in 45) if she took estrogen alone than if she took estrogen and progestin (one in 37), much as two recent studies in JNCI and the Journal of the American Medical Association showed.
But what shows up even more dramatically in Col’s model is the bigger difference in risk of uterine cancer depending on whether a woman takes estrogen alone or with progestin. The risk of uterine cancer over 10 years is one in 19 if a woman takes estrogen alone, compared to one in 102 if she takes it with progestin.
That’s a strong reason, Col said, for women not to act on fears generated by the recent studies and dump combination therapy in favor of estrogen alone.
Manson, the Brigham and Women’s endocrinologist, agreed. Manson, who is one of the chief investigators in the Women’s Health Initiative, said she fears some women in that study may drop out because of concern about the risks of combination therapy. But there are two different ways to take combination therapy – a woman can take both hormones every day, or sequentially – estrogen every day and progestin only part of the month.
The former is thought to be safer – and, in fact, in the recent JNCI study, the risk of breast cancer was lower in women who used the simultaneous therapy than in the sequential therapy group, though this did not quite reach statistical significance.
Another observation for our mythical matron is that the benefits of hormone therapy in preventing hip fracture grow with time and become more pronounced after 20 years. “And if you go out to the next 30 years, it’s even bigger,” Col said.
With heart disease, the benefits appear to kick in earlier and persist as long as a woman takes hormones. Heart disease is both more common and more likely to be lethal than breast cancer.
In Col’s model, Fosamax, the osteoporosis drug, was comparable to hormone therapies in reducing the risk of hip fractures without raising the risk of breast cancer.
Raloxifene, the hormonal alternative to estrogen, lowers the risk of hip fracture without raising breast cancer risk, but does not help much with heart disease risk. (And some younger women who try raloxifene quit because, like some who take a similar drug called tamoxifen, they simply do not feel as well on it.)
In five years, if all goes well with the Women’s Health Initiative, there will be better ways to assess the risks and benefits of hormone therapy. And that study should answer other questions as well, such as whether hormone therapy reduces the risk of cognitive problems and Alzheimer’s disease and how hormones affect mood, sleep and other quality-of-life issues.
The expectation is that the benefits of hormone therapy will outweigh its risks for most women, Manson said.
Until then, we’re in the land of educated guesses. And women who would like to do less guessing can have their doctors contact the decision-making gurus at the New England Medical Center (hrt@lifespan.org).