The WHI’s HT clinical trial enrolled women at more than 40 different centers nationwide and included upwards of 250 researchers. Since 1998, Marcia L. Stefanick, Stanford University Professor of Medicine at the Stanford Prevention Research Center, has led the collaboration, chairing both the Steering and Executive Committees of the WHI. Stefanick, 52, received her bachelor’s degree in biology from the University of Pennsylvania in 1974, and her doctorate in physiology, specializing in reproductive physiology and neuroendocrinology, from Stanford in 1982. She remained at Stanford to complete a postdoctoral fellowship in cardiovascular disease prevention, and after working as a Senior Research Scientist at the Stanford Center for Research in Disease Prevention, she was appointed to the faculty of Medicine, with a courtesy appointment in Obstetrics and Gynecology, in 1997. In addition to the estrogen-progestin replacement trial, WHI includes an estrogen-only trial, plus a large diet-intervention trial involving nearly 49,000 women nationwide, a calcium and vitamin D trial of nearly 36,000 women, and an observational study of 93,000 women. Along with overseeing all of these WHI components, Dr. Stefanick directs the Stanford Clinical Centers of the NCI-funded Women’s Healthy Eating and Living (WHEL) trial of 3,100 breast-cancer survivors, and the NIAMS-funded Study of Osteoporotic Fractures of Men (MrOS), involving 6,000 men aged 65 and over. Previously, she directed Stanford’s PEPI (Postmenopausal Estrogen/Progestin Interventions) clinical site and the PEPI Safety Follow-up Study and co-directed another study, HERS (Heart and Estrogen Replacement Study), as well as several single-center trials of exercise diet, and weight loss in both men and women.

  You’ve said that in the 1980s, when you first suggested studying hormones and heart disease, people scoffed at you. What was their logic?

Well, that was 1983, when the Lipid Research Clinics reported on the observation that women who used hormones had lower all-cause mortality. But two years later, the New England Journal of Medicine published two papers side-by-side: the Framingham Heart Study said that estrogen users had twice as much heart disease, and the Nurses Health Study said that estrogen users had half as much heart disease. It seemed like a very interesting controversy. Then everybody just swung over and took it as a truth that hormones prevented heart disease.

  What prompted this transition?

Several observational studies came out with the same observation. And rather than recognizing that estrogen users may be getting less heart disease because they’re a healthier group and that it has nothing to do with hormones, people started assuming it was a cause-and-effect relationship. Then animal researchers started weighing in with interpretations of their studies to support that hypothesis. Now we’re swinging back to a new phase, where people are interpreting studies the other way. It’s a case of very interesting biases, arising from many sources.

  You worked on PEPI, which was the first hormone-replacement clinical trial. What did that tell us?

PEPI was an interesting study because it was the first women-only trial that NHLBI ever funded. I remember when we were designing it, someone from NHLBI told us that we’d better do a really good study because this was the only chance we’d ever get to look at hormones and heart disease.

Selected High-Impact Papers by Marcia L. Stefnick Published since 1991 (Ranked by total citations) Rank Paper Total Citations 1 J.R. Rossouw, et al. (Writing Group for the Women’s Health Initiative "Risks and benefits of estrogen plus progestin in healthy postmenopausal women – Principal results from the Women’s Health Initiative randomized controlled trial," JAMA-J. Amer. Med. Assoc., 288(3): 321-3, 2002. 961 2 V.T. Miller, et al. (Writing Group for the PEPI Trial), "Effects of estrogen or estrogen/progestin regimens on heart-disease risk factors in postmenopausal women – The Postmenopausal Estrogen/Progestin Interventions trial," JAMA-J. Amer. Med. Assoc., 273(3): 199-208, 1995. 831 3 P.D. Wood, et al., "The effects on plasma lipoproteins of a prudent weight-reducing diet, with or without exercise, in overweight men and women," New Engl. J. Med., 325(7): 461-6, 1991. 281 4 M. Cushman, et al., "Effect of postmenopausal hormones on inflammation-sensitive proteins – The Postmenopausal Estrogen/Progestin Intervention (PEPI) Study," Circulation, 100(7): 717-22, 1999.  225 SOURCE: Thomson ISI Web of Science

In PEPI, we were looking at heart-disease risk-factor endpoints, but not heart disease itself. We were looking at lipoprotein metabolism—primarily HDL but also LDL and triglycerides—and then carbohydrate metabolism, with insulin as the primary endpoint, and also glucose. We were also looking at systolic blood pressure and a few other variables. PEPI was published in 1995 and basically came up with two risk factors, HDL and LDL, going in the direction of supporting the idea that estrogen was beneficial for heart disease. The publicity, however, pretty much ignored the fact that triglycerides were elevated, which is bad, and that C-reactive protein went up as well, which is also bad. Then the HERS study came along, and that was really the first red flag. And it was a big red flag.

  What exactly did HERS tell you, and, if it was such a red flag, why didn’t it undermine some of the HT optimism?

HERS, published in 1998, was a study for women with established heart disease, and it only examined estrogen plus progestin. What happened in HERS was surprising. First of all there was no benefit of estrogen plus progestin after 4.1 years. The even greater surprise was that there were significantly more heart attacks in the first year with women on estrogen and progestin, primarily in the first four to eight months. This is why people put the spin on it that estrogen and progestin cause blood clots in the beginning, so they increase harm in the early phase, but down the road you still get benefits from improving the lipids. That was the interpretation: it provided benefit attributed to the effect on lipids. But the interesting issue is that HERS only published the one-year lipids, and we’ve never seen the four-year lipid data. And when you actually study the year-by-year analysis, you see that relatively few women actually completed testing between years four and five, which is when this great turnaround was supposed to have occurred.

That’s why we did HERS-2, which was no longer blinded and which came out in JAMA the week before the WHI paper. It extended the study by 2.7 years by asking women who had originally been assigned to the active hormone group to continue taking it with the expectation of later benefit. But we simultaneously asked women who were originally assigned to the placebo not to start hormones, because now there was good trial evidence that they would be at greater risk if they started taking them. So in this phase the message was, "If you’ve been on hormones, stay on them because you may now get a benefit, but if you were on the placebo or have not been taking hormones, don’t start, because you get early harm." That was the hypothesis being tested. What HERS-2 showed was no benefit; there was no evidence that women got any later benefit.

  How did the press and the public respond? In other words, why was your WHI paper such a huge surprise?

Well, that was kind of interesting. The first HERS paper came out in August, 1998, on the day President Clinton said that he really did have a relationship with Monica Lewinsky. HERS went from being a potentially hot story to being kind of buried. The media didn’t make much of it at all. It still shocked the medical community because, before HERS came out, the American College of Cardiologists had actually built into their guidelines the recommendation that any women with a heart attack should be put on hormones. There was already a huge belief that hormone replacement was important for secondary prevention. That’s the hypothesis HERS tested, and it came up not only with no benefit, but with early harm—with significantly more heart attacks in the first year.

  When did you first become aware in WHI that the hormones might be doing more harm than good?

We had been made aware of a higher number of heart attacks, strokes, and blood clots in women on active hormones in early 2000. The data-safety monitoring board requested that we inform the participants of this. That was subsequently communicated in a letter that went out to all women in the hormone trial, including the estrogen-only group. On that occasion, we conveyed to the women that this idea had come out of HERS, that there was harm early on, but after the first two years there may be benefit over the long term. Since most of these women were past the first two years, the message suggested that they may be past the harmful phase.

  Enough of them stayed on to make the trial meaningful?

The majority stayed on. In fact, the percent of women coming off the study pills was substantially greater before April 2000 than after. About 15% had stopped after two years; from that point on, it averaged about 4% a year. When you look at the observational studies, you see that only women who are long-term hormone users are those who like to be on hormones. The idea that they’re good for all women misses the fact that more than 50% of women who start hormones stop within a year. The majority who go on hormones prefer not to be on them. So when you look at the observational studies, not only are these women healthier but they also have potentially different responses to hormones. You get a really biased picture when you’re looking only at observational studies.

  Do you think the medical community at large has accepted the WHI findings?

There’s still an active controversy. I am still being invited to come and debate people. There is at least a community of physicians who are still convinced that hormone replacement prevents heart disease. For example, critics say that we studied women who were too old to see a benefit, because the mean age in our study is 63. We have to keep telling them that half the women are younger than 63. We actually have the largest number of women in their 50s ever studied, and the data is consistent for all ages. But they ignore this.

  What else have we learned from WHI?

Well, one of the more important results was in the breast cancer paper. Rowan Chebowski is the lead author (JAMA-J. Amer. Med. Assoc., 289[24]: 3243-53, 2003). What that paper showed was that not only do you increase breast cancer with the hormones but you do it in a much shorter period of time than previously expected from observational studies. You get a really high rate of abnormal mammograms. So this drug not only increases the risk of disease, but makes disease more difficult to detect. It obscures the mammogram, and you can’t actually see the tumor until it’s a little more advanced. So the tumors you’re seeing in the estrogen-and-progestin group are actually larger and more likely to have spread to the lymph nodes. This also contradicts the orthodox wisdom that breast cancers with estrogen and progestin are not as aggressive. That idea was completely wrong. Another important finding came out in JAMA last May (S.A. Shumaker, et al., JAMA-J. Amer. Med. Assoc., 289[20]: 2651-62, 2003). It showed that women over 65 who were taking estrogen and progestin had twice as much dementia as the women on placebo. Again, this was completely contrary to the observational studies, which gave the impression that users were protected from Alzheimer’s and dementia. So the whole hormone story was wrong.

  What happens next?

What we published in July 2002 was only the estrogen-plus-progestin arm of the study. We still have an estrogen-only trial in full swing. That won’t be published until December 2005, unless it is also stopped early. We don’t know yet what that will say. As far as new studies, or NIH funding for new studies, I think that’s a big challenge now. What would you study in the hormone area that would justify the funding? There was a big workshop at NIH last October with all the directors of the big institutes and the NIH directorate, and they addressed that question. They informed the attendees that many of the estrogen-progestin trials had been stopped and deliberations were underway about others. Mostly they were telling us that all studies got stopped simultaneously after WHI came out.

  What’s life been like after July 2002 for you?

Well, there’s been a huge interest in this issue and so my life has been kind of consumed by hormones and heart disease. The biggest focus has been menopause. Everybody is asking, what do I do about menopause? How do I take care of hot flashes and stuff like that? For me, personally I’d like to get back to exercise and weight-control studies, which is what I was focusing on in the 80s and 90s. I’m still trying to figure out the best way to prevent heart disease. It’s pretty clear it’s not hormones.