Hormone replacement therapy comes full circle

Abstract

Further evidence on the early benefits and late risks does not change advice to menopausal women

The publication of the combined treatment arm of the women's health initiative study in July 2002 led to a dramatic fall in the use of hormone replacement therapy and a revision of the package insert for all hormone therapy preparations. This clinical trial randomised 16 608 women, aged 50-79 years, to 0.625 mg conjugated equine oestrogen and 2.5 mg medroxyprogesterone acetate or placebo.¹ The study was stopped early, at an average follow-up of 5.2 years, because analysis did not find the expected benefit in preventing coronary heart disease. In addition, the global index score—which measures the balance between benefit and harm—showed that the benefits in preventing hip fracture and colorectal cancer were outweighed by the increased risk of breast cancer, stroke, and deep vein thrombosis.

This week's BMJ sees the publication of the women's international study of long duration oestrogen after menopause by Vickers and colleagues.² This study was originally powered to detect a 25% reduction in the number of cases of coronary heart disease and aimed to recruit 22 300 women for 10 years' follow-up. The UK Medical Research Council stopped recruitment after the first publication from the women's health initiative, however, and 5692 women, age 50-69 years (mean age 63), entered the study. Median follow-up time was 11.9 months. Hormone preparations and doses were similar to those in the women's health initiative, although women who had vaginal bleeding triggered unblinding, and they were offered an increase in the dose of medroxyprogesterone acetate. Women who had no uterus and were unwilling to take placebo were randomised to either oestrogen only or combined treatment.

How do the results of this study compare with those from the women's health initiative? Owing to early closure of the study, the number of events was small and only two outcomes reached statistical significance. These were cardiovascular events and venous thromboembolism. All of the 11 cardiovascular events (comprising unstable angina, non-fatal myocardial infarction, or sudden coronary death) occurred in the women taking hormones; nine were using combined therapy. Compared with placebo, hormone use increased the risk of venous thromboembolism, and this risk was higher for those using combined therapy (hazard ratio 7.36; 95% confidence interval 2.20 to 24.60).

The table shows the results from the women's health initiative studies.³ Both these two studies and the study by Vickers and colleagues showed the expected increase in thromboembolism. None, however, found the expected benefit in preventing cardiovascular disease.

Hazard ratios for various outcomes for women aged 50 to 79 years

Outcome	Hazard ratio (95% confidence interval)
	Combined oestrogen and progestogen	Oestrogen only
Stroke (mainly ischaemic)	1.41 (1.07 to 1.85)	1.39 (1.10 to 1.77)
Breast cancer (final results)	1.24 (1.01 to 1.54)	0.77 (0.59 to 1.01)
Deep vein thrombosis	1.95 (1.43 to 2.67)	1.47 (1.06 to 2.06)
Coronary heart disease (final results)	1.24 (1.00 to 1.54)	0.95 (0.70 to 1.16)
Dementia (women ≥65 years)	2.05 (1.21 to 3.48)	1.49 (0.83 to 2.66)
Gall bladder disease and procedure	1.59 (1.20 to 1.97)	1.67 (1.35 to 2.06)
Hip fracture	0.66 (0.45 to 0.98)	0.61 (0.41 to 0.91)
Total fracture	0.76 (0.69 to 0.85)	0.70 (0.63 to 0.79)
Colorectal cancer	0.63 (0.43 to 0.92)	1.08 (0.75 to 1.55)
Total mortality	0.98 (0.82 to 1.18)	1.04 (0.88 to 1.22)

As in the women's health initiative trial, most women in the study by Vickers and colleagues were over the age of 65. None of these studies had the power to look at outcomes for younger women, who are the main users of hormones for symptom relief. Ongoing discussion concerns the “timing hypothesis” or “therapeutic window of opportunity.” The proposal under debate, which comes from early primate studies, is that oestrogen may be cardioprotective if treatment is started before vasculature has been compromised. Subgroup analyses with small numbers and inadequate power in both women's health initiative hormone trials suggested a non-significant reduction in the risk of coronary heart disease in women aged 50-59 (oestrogen only trial) or in women who reached the menopause less than 10 years earlier (combined trial).⁴ Because these findings were in a similar direction, Roussouw and colleagues did a secondary analysis combining the data from both trials.⁴ They found that women who started taking hormone therapy closer to the menopause tended to have reduced risk of coronary heart disease rather than the increased risk seen in women more distant from the menopause. Again, however, this trend test did not meet the authors' criterion for statistical significance. The risk of stroke was raised regardless of years since menopause. Moreover, women aged 50-59 using combined therapy have a significantly increased risk of deep vein thrombosis.⁵

Roussouw and colleagues have asked the important question, “If early hormone use is cardioprotective, will this benefit continue?” and their comments suggest that the answer is “no.” They say, “Age-related progression of atherosclerosis is likely to continue even in the face of hormone therapy. Even if ongoing imaging trials confirm a slowing of early atherosclerosis, it would be unwise to extrapolate such findings to clinical benefit with continued use into old age.”⁴

Also, it has been calculated that, even if these were not subgroup analyses and even if the differences were significant, 1000 women each year would need to use hormone therapy to prevent one cardiovascular event.⁶

A recent surrogate outcome trial has now been published from the women's health initiative group. The coronary-artery calcium study looked at the effects of oestrogen on coronary artery calcification in women who were 50-59 years at randomisation in the oestrogen only arm of the original trial.⁷ Women taking oestrogen had a 42% reduction in coronary artery recalcification, supporting the likelihood of a potential cardioprotective effect in younger women. However, the accompanying editorial in the New England Journal of Medicine—by a cardiologist—reminds us that the vascular effects of oestrogen are complex: “It is important to emphasize that hormone replacement therapy should not be considered as a strategy to prevent cardiovascular disease and that there are proven therapies for cardiovascular disease that remain underused in women.”⁸

So postmenopausal hormone therapy has come full circle.⁹ It was originally used to treat menopausal symptoms, and now the indications for use are again hot flushes, night sweats, and vaginal dryness. It is the best treatment we have at present for these symptoms. Hot flushes and night sweats are mostly self limiting, and current advice recommends short term use with the lowest dose needed for relief of symptoms. Healthy women in early menopause are at a low absolute risk whether they take hormones or not, and they are unlikely to face substantially increased risks when using hormones for a few years.⁹

Long term use of hormone replacement therapy to prevent chronic disease is no longer recommended, because available randomised evidence shows that the negative outcomes outweigh the positive benefits. Menopausal genitourinary symptoms, however, usually last a long time and require ongoing treatment. Current low dose vaginal oestrogen preparations have minimal systemic absorption, and recent position statements support long term use of these preparations as long as distressing symptoms remain.¹⁰