IMS members react to the recent JAMA paper
15 October, 2013
Menopausal hormone therapy continues in clinical use but questions remain regarding its risks and benefits for chronic disease prevention. A total of 27,347 postmenopausal women aged 50–79 years were enrolled at 40 US centers to the Women's Health Initiative trial . Women with an intact uterus received conjugated equine estrogens (CEE; 0.625 mg/day) plus medroxyprogesterone acetate (MPA; 2.5 mg/day) (n = 8506) or placebo (n = 8102). Women with prior hysterectomy received CEE alone (0.625 mg/day) (n = 5310) or placebo (n = 5429). The intervention lasted a median of 5.6 years in CEE + MPA trial and 7.2 years in the CEE-alone trial with 13 years of cumulative follow-up until September 30, 2010. During the CEE + MPA intervention phase, the numbers of coronary heart disease (CHD) cases were 196 for CEE + MPA vs. 159 for placebo (hazard ratio (HR) 1.18; 95% confidence interval (CI) 0.95–1.45) and 206 vs. 155, respectively, for invasive breast cancer (HR 1.24; 95% CI 1.01–1.53). Other risks included increased stroke, pulmonary embolism, dementia (in women aged ≥ 65 years), gallbladder disease, and urinary incontinence; benefits included decreased hip fractures, diabetes, and vasomotor symptoms. Most risks and benefits dissipated post-intervention, although some elevation in breast cancer risk persisted during cumulative follow-up (434 cases for CEE + MPA vs. 323 for placebo; HR 1.28; 95% CI 1.11–1.48). The risks and benefits were more balanced during the CEE-alone intervention with 204 CHD cases for CEE-alone vs. 222 cases for placebo (HR 0.94; 95% CI 0.78–1.14) and 104 vs. 135, respectively, for invasive breast cancer (HR 0.79; 95% CI 0.61–1.02); cumulatively, there were 168 vs. 216, respectively, cases of breast cancer diagnosed (HR 0.79; 95% CI 0.65–0.97). Results for other outcomes were similar to CEE + MPA. Neither regimen affected all-cause mortality. For CEE-alone, younger women (aged 50–59 years) had more favorable results for all-cause mortality, myocardial infarction, and the global index. Absolute risks of adverse events (measured by the global index) per 10,000 women annually taking CEE + MPA ranged from 12 excess cases for ages of 50–59 years to 38 for ages of 70–79 years; for women taking CEE-alone, from 19 fewer cases for ages of 50–59 years to 51 excess cases for ages of 70–79 years. Quality-of-life outcomes had mixed results in both trials. In conclusion, findings from the intervention and extended post-intervention follow-up of the two WHI hormone therapy trials do not support use of this therapy for chronic disease prevention, although it is appropriate for symptom management in some women.
There are several key points to the newest WHI manuscript , the largest compendium of WHI data published in one place to date. Importantly, the WHI investigators issue the following warning on page 1354 under 'Statistical Analysis': 'The p values do not adjust for multiple outcomes, sequential monitoring, or multiple subgroup comparisons due to the large number of tests conducted; therefore, the p values should be interpreted cautiously.' As Figures 2 and 4 show, even in the absence of the foregoing appropriate adjustments, almost no adverse outcomes are statistically significant, and, as the original publications from the WHI have shown for breast cancer, stroke and all of the other adverse outcomes, that, when adjusted for multiple outcomes and sequential monitoring according to the a-prioridefined statistic, all of these outcomes are not statistically significantly different between hormone therapy (HT) and placebo in both the CEE + MPA and CEE trials [2, 3]. The one exception is the venous thromboembolic event outcome in the CEE + MPA trial. In other words, all of the adverse outcomes from the WHI HT trials are not statistically significantly different between HT and placebo (except for the one exception noted previously) when analyzed across all women randomized to these trials (average age 64 years and average time from menopause > 10 years). Further, this newest WHI publication again validates that not only are all of the adverse outcomes non-significant, but they are similar to the types and magnitude of adverse events reported with other commonly used medications, that is, rare (< 1/1000 additional events per year of therapy) [2, 3].
The second important fact to realize from this newest WHI publication is that Figures 5 and 6 show that all outcomes are reduced with CEE relative to placebo in women who are 50–59 years old when randomized. Pulmonary embolism and hip fracture, which show three additional events per 10,000 women per year of therapy, are the only exceptions to the consistent reduction in outcomes. As such, the data in these figures do not support the conclusion of the manuscript that HT should not be used for chronic disease prevention. Some may argue that these are subgroup analyses and therefore should be ignored. However, if this is the case, then no statement concerning chronic disease prevention can be made based on the WHI data except for women over all ages, the majority of whom are well beyond the age in which HT is initiated. Women require HT and HT is initiated predominantly in women < 60 years of age, typically in the perimenopause or early menopause, when the data support clear benefit relative to risk, as shown in a recent 10-year randomized trial of such women . In other words, the WHI has no relevance to clinical practice as presented across the entire cohort of women; its importance is in showing a reduction of coronary heart disease, breast cancer, stroke and, most importantly, total mortality by 30% with CEE as well as CEE + MPA therapy in women who were initiated on such therapy when < 60 years of age. Based on the WHI data, it has recently been estimated that 50,000–90,000 women over the last decade have needlessly died by avoiding the use of HT . This estimate appears to be consistent with the inexplicable worsening mortality rate of women in the United States relative to men from 1992 to 1996 (pre-WHI) relative to 2002–2006 (post-WHI) .
It must be emphasized that WHI data cannot inform about chronic disease prevention since benefits and risks of HT when initiated in a 70–80-year-old women cannot be assumed to equate to the benefits and risks of women 20–30 years after initiating HT at 50–60 years of age. The contrary, however, is that which follows known preventive therapy logic in which risk is reduced initially and is followed by reduction in adverse events, such as bone fracture prevention. There is absolutely no evidence to indicate that HT operates differently from this well-known paradigm of chronic disease prevention. The fact that '…the risk reductions dissipated post-intervention', as stated on page 1366 of the newest WHI manuscript, clearly indicates that chronic disease prevention will only be appreciated with continuous HT, as seen with the prevention of bone fractures.
Howard N. Hodis
Atherosclerosis Research Unit, Division of Cardiovascular Medicine, Keck School of Medicine, University of Southern California, USA
The recent updated report of the WHI study (follow-up of 13 years) did not add anything of importance to help practitioners who care for postmenopausal women . It must be emphasized again that the methodology followed by the WHI study does not in any way reflect what is meant by good clinical practice. This is their capital error. Let it not be forgotten that all enrolled patients were treated with the same medication and dosage irrespective of their age and tolerance. Furthermore, the estrogen (CEE) and the progestogen (MPA) used were not the best choices since they may cause important side-effects that are rare with 17β-estradiol and natural progesterone, as mostly used in Europe. Past observational studies and the recent KEEPS study, much closer to good clinical practice, did not support most of the conclusions of the WHI investigators. The only important observation is that the good candidates for hormone therapy are women younger than 60 years of age or within 10 years after the menopause. There is nowadays a general consensus that CEE + MPA should not be first-choice medications and that the more one mimics physiology the better, i.e. 17β-estradiol and progesterone itself, or dydrogesterone. The parenteral route for the administration of estradiol seems to be safer. WHI investigators, with their unbalanced conclusions, have already caused much damage to women who have been deprived of appropriate and needed hormone therapy and thus of the resulting preventive effects at the level of the cardiovascular system and bones.
1. Manson JE, Chlebowski RT, Stefanick ML, et al. Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women’s Health Initiative Randomized Trials. JAMA 2013;310:1353-68.
2. Hodis HN, Mack WJ. The timing hypothesis and hormone replacement therapy: a paradigm shift in the primary prevention of coronary heart disease in women. 1. Comparison of therapeutic efficacy. J Am Geriatr Soc 2013;61:1005-10.
3. Hodis HN, Mack WJ. The timing hypothesis and hormone replacement therapy: a paradigm shift in the primary prevention of coronary heart disease in women. 2. Comparative risks. J Am Geriatr Soc 2013;61:1011-18.
4. SchierbeckLL, Rejnmark L, Tofteng CL, et al. Effect of hormone replacement therapy on cardiovascular events in recently postmenopausal women: randomised trial. BMJ 2012;345:e6409.
5. Sarrel PM, Njike VY, Vinante V, et al. The mortality toll of estrogen avoidance: an analysis of excess deaths among hysterectomized women aged 50–59 years. Am J Public Health 2013;103:1583-8
6. Kindig DA, Cheng ER. Even as mortality fell in most US counties, female mortality nonetheless rose in 42.8% of counties from 1992 to 2006. Health Affairs 2013;32:451-8.