Hormone Therapy in Postmenopausal Women: Did We Have it Wrong All Along?

After publication of the Women’s Health Initiative (WHI) in 2002, postmenopausal women have often been discouraged from using hormone therapy (HT) due to concerns regarding thromboembolic disease, cancer, and cardiovascular events.  But results from a recent study suggest that HT isn’t as bad as you may think.  When HT is initiated shortly after menopause, it may actually improve cardiovascular outcomes and have no impact on cancer risk.¹  Is it possible that we were wrong all these years about HT and have been denying our patients a beneficial treatment? 

In the past, HT was commonly used in postmenopausal women to relieve symptoms such as hot flashes and vaginal atrophy.  In the 1980’s and 90’s it was prescribed widely based on a belief that HT reduced the risk of cardiovascular disease, osteoporosis, and colon cancer based on observational studies.²    But the Heart and Estrogen/Progestin Replacement Study (HERS) published in 1998, demonstrated worse outcomes when used in women for the secondary prevention of coronary heart disease (CHD).³  Shortly thereafter, the findings of the WHI were published.  WHI was the largest clinical trial ever conducted in women and it consisted of two randomized, placebo controlled HT trials using progestin+estrogen combination therapy in one study and estrogen alone in the other.  The data documented a number of adverse outcomes, including an excess risk of CHD, stroke, venous thromboembolism, and breast cancer after 5 years of use.⁴

The objective of the Danish Osteoporosis Prevention (DOPS) study was to determine the benefits of HT in women who initiated therapy soon after the onset of menopause.¹  The DOPS was a prospective, multi-center, open-label, randomized, controlled trial which enrolled patients from 1990-1993 and followed them for 11 years.  The original study evaluated HT for the primary prevention for osteoporotic fractures but, in this analysis, the investigators report the long term effects of HT on cardiovascular outcomes.

Healthy white women were randomly allocated to receive HT (n=502) or no treatment (n=504).   Women were aged 45-58 years with a last menstrual bleed 3-24 months before study entry or had perimenopasual symptoms with elevated follicle-stimulating hormone (FSH) levels.  Women with a hysterectomy were included if they were aged 45-52 years and had a documented increase in FSH.  Exclusion criteria were a history of bone disease, any uncontrolled chronic disease, history of cancer or thromboembolic disease, previous or current treatment with glucocorticoids > 6 months,  use of HT within the previous 3 months, and alcohol or drug dependency.  

The treatment group received a combination of triphasic estradiol plus norethindrone or, if the woman was status post hysterectomy, estradiol alone.   The primary outcome of this analysis was a composite of death, admission to hospital for heart failure, or myocardial infarction.  The original trial was stopped after 11 years when the negative outcomes of the WHI studies were released but participants continued to be followed.   Data regarding health events was retrieved for all participants at years 10 and 16 through the Danish civil registration system and the national hospital discharge register, which are nationwide electronic databases. The authors note that records were nearly 100% complete and there was high precision of diagnoses.

Baseline characteristics were similar between the groups with the exception of age (mean age was 50 years in the control group vs. 49.5 years in the treatment group, p=0.007).   Patients had a mean BP=130/81 mm Hg, BMI=25.2kg/m², and 43% were smokers.  Women were postmenopausal for an average of 7 months at the time of enrollment in the study.  The mean duration of treatment (in those who received treatment) was 10.1 years.  Results were analyzed with an intent-to-treat analysis.   After 10 years of follow-up, 16 women in the treatment group experienced the primary composite endpoint versus 33 in the control group (HR: 0.48, 95% CI, 0.26-0.87; P=0.015).  At the 16 year follow-up, the composite outcomes of the primary endpoint remained significantly different (see Table 1).

Table 1
Select endpoints at 16 year follow-up after 11 years of randomized treatment (HT vs. No HT)

There were trends toward fewer events in the HT group with the exception of cancer, but none of the individual endpoints were statistically significant.  Subgroup analyses were performed for those participants under 50 years of age versus those over 50 and those with a hysterectomy versus an intact uterus. Women under 50 had a significantly reduced risk of the primary outcome (HR: 0.49, 95% CI: 0.28-0.87, p=0.015).  Women with a hysterectomy receiving estrogen had a decreased risk of death and breast cancer compared to the control group (HR: 0.42, 95% CI: 0.18-0.97, p=0.043), but primary endpoint was not statistically different. 

This study demonstrates that 10 years of HT initiated early after menopause reduced the risk of mortality, heart failure, and myocardial infarction (combined) without increasing the risk of cancer, venous thromboembolism, or stroke.   However, there are several study limitations that must be acknowledged.  Adherence was based on self-reports.  The study was open-label, and included only healthy, white women from Denmark.   The number of cardiovascular events in the study was small in both the treatment and control group.  Can these results be generalized to other populations, especially other ethnic groups and women at increased cardiovascular risk?  

In comparing this trial to the WHI, there are some notable differences. The WHI enrolled significantly older women (mean age=63 years), had a slightly more diverse patient population (84% white), and used a different HT formulation (conjugated equine estrogen and medroxyprogesterone).⁴   The WHI was much larger (n=16,608), blinded, placebo-controlled, and adherence was monitored by pill counts.  Cardiovascular outcome data collected was similar but the WHI included silent MI’s by comparing baseline and follow-up electrocardiograms.  The WHI contacted patients every six months and did not utilize a national database to track cardiovascular outcomes.⁵

It is interesting to note that a subanalysis of the WHI for younger women (ages 50-59 years) within 10 years of menopause also showed a trend toward lower rates of CHD events in those received HT.⁶  Moreover, other smaller studies and a meta-analysis have demonstrated positive cardiovascular benefits and reduced mortality in women who receive HT close to menopause.^7-8  Indeed, the International Menopause Society in their 2013 consensus statement acknowledges that HT may decrease all-cause mortality in women younger than 60 years of age.⁹

There are several unanswered questions about HT.   It seems that HT provides the most benefit to postmenopausal women under 60 with few cardiovascular risk factors, but the reasons are unclear.  Should there be a strict cut-off age for initiating HT?  What is the maximum amount of time that a woman should use HT?  Are certain formulations better than others?  

It is not appropriate to initiate HT in women in the early postmenopausal period for the sole purpose of improving cardiovascular outcomes.  Other evidence-based treatments such as lifestyle modifications, blood pressure control, tobacco cessation, and lipid management should be used instead.   But perhaps, it is time to change our perceptions of HT and admit that we may have been too restrictive. HT is highly effective for postmenopausal vasomotor symptoms⁹ and based on this study and other analyses it appears to have beneficial, not harmful, cardiovascular effects for patients under 60. Unfortunately, there remain strong warnings about the dangers of HT¹⁰ and many women are afraid to take it.  We shouldn’t let these misleading messages prevent post-menopausal women from receiving a beneficial treatment.

In light of this new data, what will you tell your female patients and health professional colleagues about HT?