AUTHOR COMMENTARIES - 2008

Dr. Rossouw is the chief of the Women's Health Initiative, which was established in 1991 by the National Heart, Lung, and Blood Institute at the National Institutes of Health "to address the most common causes of death, disability and poor quality of life in postmenopausal women —cardiovascular disease, cancer, and osteoporosis."

This month, ScienceWatch.com talks with Dr. Rossouw about this groundbreaking paper.

Please tell us about the aims and design of the trial described in the paper.

The Women's Health Initiative (WHI) randomized trial of estrogen plus progestin versus placebo was designed to answer the question of whether hormone therapy given to postmenopausal women aged 50-79 years and with an intact uterus would prevent coronary heart disease. Breast cancer was monitored as a major potential adverse outcome, and we also monitored several other important conditions to assess whether the overall benefits of hormone therapy outweighed the risks.

The hormones selected for study were conjugated equine estrogen at a dose of 0.625 mg/day plus medroxyprogesterone acetate at a dose of 2.5 mg/day, since these are the most widely used postmenopausal hormones in the US, and prior data from non-randomized observational studies had suggested that they might reduce risk of coronary heart disease. The trial was conducted in 40 clinical centers across the US, and the published report was on behalf of the entire WHI investigator group. The WHI is funded by the National Heart, Lung, and Blood Institute.

What were the findings of this study?

The trial was stopped early—after 5.2 years of data had accumulated rather than the planned 8.5 years—on the recommendation of the independent Data and Safety Monitoring Board. The immediate reason for stopping was that the risk of breast cancer on estrogen plus progestin became significantly elevated, and crossed the pre-specified monitoring boundary for safety. Additionally, the risks of coronary heart disease, stroke, and venous thromboembolism were significantly elevated, and these risks were not counterbalanced by the reduced risks of hip fracture and colorectal cancer.

Hence, the primary hypothesis—that hormone therapy would reduce the risk of coronary heart disease—was not borne out and in fact the result was opposite to what was anticipated. Also, the balance of benefits and risks was in an unfavorable direction.

The implications of these findings were that this hormone therapy should not be used for prevention of coronary heart disease, and indeed, in view of the many adverse effects, it was not suitable for prevention of chronic disease generally.

How were these results received by the community?

The publication of these findings gained widespread attention in the media and in scientific circles. This resulted in a marked shift in attitudes towards this hormone therapy, with prescriptions dropping by some two-thirds in the months following publication (and remaining at those or lower levels since then). The Food and Drug Administration changed the package insert of all hormone therapies to include warnings about not using them for prevention of coronary heart disease, and relegating them to second-line treatment status for prevention of osteoporosis. The major indication for oral hormone therapy is now for the short term, at the lowest doses possible, for the relief of hot flashes and night sweats.

Cardiologists and internists who deal with long-term prevention of chronic diseases readily accepted the findings, and now focus on proven and effective non-hormonal strategies for prevention. Gynecologists had more difficulty with dealing with the findings because there are few effective alternatives for the treatment of hot flashes and night sweats.

Has the balance of the risks and benefits of hormone use in postmenopausal women been explored further since the publication of this paper? If so, have the findings differed from those in this paper?

The original paper has withstood the test of time, and indeed a parallel study in the UK (stopped within a few years of starting after WHI published its findings) using the same estrogen and progestin has published data that are very consistent with our findings.

Estrogen alone (without a progestin) has a somewhat more favorable profile. In 2004 the WHI investigators published the results of the trial of estrogen alone. This, too, was stopped early (after 6.8 years) but in this instance because of an increased risk of stroke, with no benefit for heart disease. Like estrogen plus progestin, estrogen alone also increased the risk of venous thromboembolism and decreased hip fractures, but unlike estrogen plus progestin, it did not have any effect on breast cancer or colorectal cancer. The increased risk of stroke and venous thromboembolism makes estrogen alone unsuitable for long-term prevention.

Since the publication of the primary results of the hormone trials, several secondary papers from both trials have explored the data in more detail, but the overall conclusions and recommendations have not changed. The risk of coronary heart disease due to hormones may be modified by age or years since menopause at time of initiation, such that there is no increased risk in women closer to the menopause compared to the clearly increased risk in women more distant from the menopause. On the other hand, the increased risk of stroke due to hormones does not depend on years since menopause, and neither does the increased risk of breast cancer on estrogen plus progestin.

The findings of no increased risk of coronary heart disease in women initiating hormones within 10 years of the menopause are somewhat reassuring for those women considering short-term hormone therapy for the relief of hot flashes and night sweats. In any case, at those younger ages women are at very low risk of coronary heart disease. However, the data provide no reassurance that initiation at a younger age, if continued for decades into older ages, will result in continued freedom of risk from coronary heart disease. Given this uncertainty, and together with the many off-target effects of long-term hormone therapy (such as breast cancer, stroke, venous thromboembolism), it is safe to predict that hormone therapy for prevention of coronary heart disease will not regain favor.

We have recently published the results of the first three years of follow-up after stopping estrogen plus progestin, and the results show that there was no increased risk of coronary heart disease, stroke, or venous thromboembolism after stopping. There was still some increased risk of breast cancer in the three-year period, but the trend towards ever-increasing risk towards the end of the original trial period was no longer present.

Overall cancer risk appeared to increase after stopping hormones, perhaps due to a carry-over effect of the trial exposure. The in-trial benefits for hip fracture and colorectal cancer did not carry over into the first three years after stopping.

What other research questions are the Women's Health Initiative working on?

We will continue to follow the study cohorts to gain further insights into what happens when the hormone therapy is stopped. We are doing several laboratory-based studies of risk markers, genomics, and proteomics to see if we can identify women who might be at increased risk if they take hormone therapy. One such study suggests that it might be useful to measure the blood cholesterol levels prior to hormone therapy—elevated cholesterol levels themselves confer risk of coronary heart disease, and that risk appears to be amplified further by hormone therapy.

Where do you see this work going in five to ten years?

The WHI has several other components, including trials of low-fat dietary pattern and calcium and vitamin D, together with a large observational study of women who were not enrolled in the trials. We would like to follow the entire cohort of more than 100,000 women for decades to study the long-term effects of the interventions. More importantly, many of these women are now entering their 80s and 90s and form an extremely valuable resource for studying the predictors of healthy aging, or on the other hand predictors of frailty, loss of cognitive function, and congestive heart failure. These and other conditions affecting aging women will consume an increasing proportion of the nation's health budget, and it would be important to identify factors that might help us avoid them.

Jacques Rossouw, M.D.
Women's Health Initiative Branch
Division of Prevention and Population Studies
NHLBI
Bethesda, MD, USA

Keywords: Jacques Rossouw, women's health initiative, whi, hormone therapy, estrogen, progestin, healthy postmenopausal women, randomized controlled trial, risks, benefits, elevated breast cancer risk, coronary heart disease, stroke, venous thromboembolism.