Introduction: While initial findings from the Women's Health Initiative (WHI) did not show cardiovascular benefits of hormone therapy in menopausal women, further analysis of this data and findings from other studies suggest that a subset of women who begin treatment early (ie, at a younger age or closer to the onset of menopause) may benefit from hormone therapy. Thus, the randomized, controlled Kronos Early Estrogen Prevention Study (KEEPS) was designed to investigate the effects of early initiation of oral or transdermal hormone therapy on progression of atherosclerosis in women who were within 36 months of their last menses.
Methods: The study involved 727 healthy menopausal women (age, 42 to 58 years) who were between 6 and 36 months from their last menses. None of the women had a history of cardiovascular disease at baseline, and all had a coronary artery calcium (CAC) score of <50 Agatston units. None of the women took estrogen or lipid-lowering therapy within 90 days of study entry.
The women were randomized to receive 0.45 mg/day oral conjugated equine estrogens (n=230) or 50 mcg/day transdermal 17β-estradiol (n=222) in addition to 200 mg of oral progesterone for 12 days per month, or placebo for 48 months (n=275). Atherosclerosis progression was measured as annual change in carotid artery intima–media thickness (CIMT) by ultrasonography (primary outcome) and CAC score.
Results: At 48 months, CIMT was available for 580 women (79.8%) of whom 464 women (63.8%) were still taking the study medication. During the 48-month follow-up, CIMT increased at a similar rate (0.0076 mm/year; range 0.0072-0.0080 mm/year) in all 3 treatment groups, with no statistically significant difference between the groups (Table). CAC scores were available for 570 women (78.4%), and showed that a similar proportion of women experienced increases (range, 17.4%-21.0%; Table).
Oral conjugated equine estrogen use was linked to significant improvement in low and high-density lipoprotein cholesterol levels, and significant increases in C-reactive protein and sex hormone-binding globulin (SHBG) compared to women who took placebo. (The P-value of this association was not specified, but was noted as significant by the study authors.) Transdermal 17β-estradiol was linked to significant decreases in Homeostasis Model Assessment of Insulin Resistance (HOMA-IR) scores compared with placebo. Blood pressure and interleukin-6 levels did not differ among the treatment arms.
Both forms of hormone therapy were linked to relief of hot flashes, which were reported by 28.3% of the placebo group compared with 4.2% and 7.4% of the oral estrogen and transdermal estradiol groups, respectively (P<0.001).
Conclusion: Menopause hormone therapy had no effect on progression of atherosclerosis as determined by arterial imaging in this study.
CAC, coronary artery calcium; CIMT, carotid artery intima–media thickness
aChange between baseline and year 4 as annual slope.
No statistically significant differences were found between the treatment arms for either measure.
Harman SM, Black DM, Naftolin F, et al. Arterial imaging outcomes and cardiovascular risk factors in recently menopausal women: a randomized trial. Ann Intern Med. 2014;161(4):249-260.
Commentary by Tamara L Wexler MD, PhD
Tamara L. Wexler, MD, PhD, is an endocrinologist specializing in reproductive and neuroendocrinology, and an Attending in Medicine, Massachusetts General Hospital, Boston, MA.
Cardiovascular disease (CVD) remains the primary cause of death in women, and any interventions that can impact CVD risk are of great interest. Given conflicting results from observational studies and the Women's Health Initiative data, the KEEPS trial (Kronos Early Estrogen Prevention Study) was initiated.
The KEEPS trial investigated markers of cardiovascular health in relatively young (mean age 52.7 years) and healthy women within 3 years of last menses (mean 1.4 years; the cohort was 6-36 months from last menses, with laboratory evidence of menopause). The trial excluded women with evidence of CVD, as assessed by history, comorbidities, and baseline CAC measurement.
Subjects were randomized to 1 of 3 arms: oral estrogen + cyclic progesterone, transdermal estrogen + cyclic progesterone, or placebo. The primary endpoint was CIMT, and secondary endpoints were CAC and other markers of cardiac health (low-density lipoprotein [LDL], high-density lipoprotein [HDL], total cholesterol, triglyceride; high-sensitivity C-reactive protein [hsCRP], interleukin-6, SHBG, HOMA-IR). CIMT and CAC are considered independent measures of atherosclerosis and predictors of cardiac events. CIMT was measured at baseline and every year thereafter. CAC was measured from coronary artery CT imaging at baseline and 48 months (and reported as a binary outcome). Serum markers were measured at baseline and 12, 36, and 48 months.
There was no statistically significant effect on CIMT or CAC. Hormone replacement therapy (HRT) did have a positive impact on vasomotor symptoms, and on some of the serum biomarkers associated with cardiovascular risk: LDL decreased and HDL increased in the oral HRT group (though hsCRP also increased); total cholesterol and HOMA-IR decreased in the transdermal HRT group. There were more subjects on HRT who were diagnosed with breast cancer (a serious adverse event) during the study duration (3 on oral HRT, 2 on transdermal HRT, 1 on placebo), but the number was very small and thus did not allow for evaluation of statistical significance. There was no significant difference in adverse events other than a higher incidence of vaginal bleeding in pooled HRT groups.
This multicenter randomized, controlled, 3-armed clinical trial was carefully double-blinded, but does have a number of limitations, which help to inform our application of the results. Some limitations (mentioned below) are related to the nature of the selected cohort and endpoints. In addition, while approximately one-third of the cohort discontinued treatment by the 48-month time point, analysis was conducted by intention to treat; thus, measurements in each category included subjects no longer on the intervention. For example, at 48 months, only 80% of the participants who had CIMT measured remained on treatment (81% of those on oral HRT, 74% of those on transdermal HRT, and 84% of those on placebo). Of the subjects included in the study, 20% of subjects had been on HRT, though had discontinued at least 90 days prior to enrollment. The percentage of prior postmenopausal HRT users was higher in the oral HRT group (26%) than in the transdermal HRT (19%) and placebo (19%) groups. A small percentage (5.4%) also was noted to begin non-study prescription HRT during the study.
Context and Conclusions
While the review by Hodis and Mack concludes that HRT used shortly after menopause may be protective against coronary heart disease, this multicenter randomized controlled trial (whose authors include Dr. Hodis) did not show evidence of protection, but found that up to 4 years of HRT within the first years after menopause does not reduce progression of atherosclerosis as measured by CIMT or CAC. The apparent difference may be due to several key factors, including cohort characteristics, HRT dose, outcome measured, and study duration. The KEEPS cohort is relatively young and healthy (including mean BMI 26), and specifically selected for low cardiovascular risk. A lower dose of estrogen was used than in prior studies such as those reviewed in Hodis and Mack) overview, in part reflecting the relatively recent nature of this study (recruited 2005-2008.) A higher dose of estrogen, or longer duration of use, may have a greater impact on atherosclerosis measurements. All interventions were of combined hormone therapy (estrogen plus progestin), versus the mix of interventions included in the analysis. And, of course, the KEEPS outcome measures are CIMT and other biomarkers, versus the hard cardiac endpoints measured in a number of other studies reviewed by Hodis and Mack).
In summary, based on the results of their study, Harman and colleagues conclude that HRT does not have a significant effect on progression of CIMT or CAC in recently menopausal women, though they do draw a distinction between these biomarkers and cardiovascular event rates. Results of all studies should be considered in light of the type and dose of hormone therapy, and the characteristics of the cohort studied. Taken in this context, this study represents good evidence that low-dose estrogen and cyclic progestin, initiated within 3 years of menopause, do not impact CIMT and CAC progression over 4 years, though they may impact other biomarkers of cardiovascular risk and help alleviate vasomotor symptoms.