Hot Flash! More From the HERS Trial
Abstract & Commentary
Synopsis: Postmenopausal women with cardiac histories who are still flushing do better with hormone replacement therapy.
Source: Hlatky MA, et al. JAMA. 2002;287:591-597.
The Heart and Estrogen/progestin Replacement Study (HERS) trial enrolled postmenopausal women younger than 80 years who had coronary artery disease (CAD) into a randomized, double-blind, placebo-controlled trial. After randomization, the women were given 0.625 mg of conjugated equine estrogens and 2.5 mg of medroxprogesterone (Prempro®) or placebo and followed for up to 4 years. Hlatky and colleagues defined postmenopausal in 4 ways: age 55 years or older and no natural menses for at least 5 years; no natural menses for at least 1 year and a serum follicle-stimulating hormone (FSH) level greater than 40 mIU/mL; documented bilateral oophorectomy; or reported bilateral oophorectomy and FSH level greater than 40 mIU/mL and estradiol level less than 25 pc/mL. CAD was defined as previous myocardial infarction (MI), greater than 50% luminal narrowing on coronary angiography, or previous coronary revascularization. They excluded patients with MIs or revascularizations within the last 6 months, with hysterectomies, with contraindications to hormone replacement therapy (HRT), with use of HRT within the last 3 months, with life-threatening diseases, or who could not return for follow-up visits.
The 2763 women (1380 randomized to HRT) completed a quality-of-life questionnaire that included a physical function assessment, an energy/fatigue scale, a mental health inventory, and anxiety and depression scales at baseline, 4 months, 1 year, 2 years, 3 years, and 4 years. The data were analyzed for just the first 3 years. The analysis was by intention-to-treat.
At baseline, the 2 groups did not differ significantly. The average age was 66.6 years. They were predominantly white (89%). More than half of them had a smoking history, hypertension, prior MI, and prior revascularization. Despite this, 76% rated their health as good-to-excellent. Sixteen percent were still experiencing flushing at least some of the time. Their physical function assessment was consistent with a moderate amount of physical activity.
After 3 years of follow-up, the HRT group had a larger decline in physical activity and energy/fatigue compared to the control group. Both of these findings were statistically significant; the actual numerical scores dropped 17% vs. 12% and 8% vs. 5%, respectively. Both experimental and control groups showed minor declines in mental health that were not significantly significant. Patients assigned to HRT were less depressed at 3 years.
Hlatky et al looked at the data based on whether the patients were still flushing. As expected, the flushing patients were younger (63.0 vs 67.3 years) and had been postmenopausal for a shorter time (13.7 vs 18.7 years). The prevalence of flushing declined significantly among HRT patients, dropping from 16.7% to 5.8% after 3 years. Among the control group, the prevalence dropped from 14.7% to 11.2%. At baseline, physical activity, energy/fatigue, mental health, and depression scores were all significantly worse in the flushing patients than the nonflushing patients. Over the 3 years of the study, the decline in physical function and energy/fatigue among flushing patients was not significantly different between the HRT group and the control group, but HRT was associated with improvement in mental health and depression scores. On the other hand, women who were not flushing at baseline and who received HRT had significant declines in physical functioning and energy/fatigue, but no significant change in mental health or depression.
Subgroup analysis yielded these findings. Chest pain and lower level of education lowered quality-of-life scores. Older patients had lower physical activity scores, but better mental health and depression scores than younger patients. Heart failure, diabetes, and hypertension all lowered physical activity and energy/fatigue scores but had less effect on mental health and depression. Having a revascularization procedure did not affect quality-of-life. Having had an MI had no effect on physical activity or energy/fatigue but was associated with better mental health scores and less depression.
Comment by Allan Wilke, MD
In 1998, the main results of HERS were published. Briefly, postmenopausal women with cardiac disease do not do well in the first 3 years after receiving HRT; they have more thromboembolic events and gallbladder disease and, at least in the first year, more cardiac events. The primary outcomes of the study were cardiovascular. This study looks at secondary end points and attempts to answer the question, "Do they feel better?" Measuring "feeling better" involves choosing outcomes and trying to determine "quality-of-life." The outcomes Hlatky et al chose are relevant, but they could have chosen to measure other things. For instance, in some studies, estrogen use has been associated with less Alzheimer’s disease. One could reasonably suggest that Alzheimer’s disease has an adverse effect on quality-of-life and that the investigators could have measured the Mini-Mental Status scores.
It appears that the only women in this study who benefited from HRT are those who were flushing. So how many postmenopausal, cardiac patients who are flushing would you have to treat to help one woman? This is the concept of "number needed to treat" (NNT). We can use the data to answer the question. Looking at the HRT group, we see that the "pharmaceutical" cure rate is 10.9% (16.7-5.8%). If we look at the controls, we see that the "natural" cure rate is 3.5% (14.7-11.2%). If we assume that 3.5% of women in the experimental group would have had a "natural" cure, then this leaves 6.4% of women whom HRT helped. The last piece of mathematics is to convert 6.4% into a fraction, 1/15. In other words, we would have to treat 15 women with flushing to provide relief to 1, at the risk of increased thromboembolism and gallbladder disease.
There are some nonintuitive results in this study. For instance, why would having an MI be associated with no effect on physical activity or energy/fatigue, but better mental health scores and less depression? In the main study, the women on HRT had more cardiac events in the first year than the control group. I can imagine that having an MI during the study period could have a negative effect on my quality-of-life.
Dr. Wilke is Assistant Professor of Family Medicine, Medical College of Ohio, Toledo, Ohio.