Oral Micronized Progesterone Cardiovascular Impact: Long-Term Evidence

What the long-term data actually measure

The Prometrium label states that 'the safety and efficacy of estrogen plus progestogen for the prevention of cardiovascular disorders has not been established' [1]. That sentence frames everything below. The long-term studies that included oral micronized progesterone measured cholesterol fractions, inflammatory proteins, carotid artery wall thickness, coronary calcium and venous clot risk. None of them tested heart attack or stroke as its main endpoint.

The label also rules the drug out for women with active deep vein thrombosis or pulmonary embolism, a history of either, active arterial thromboembolic disease such as stroke or myocardial infarction, or a history of these [1].

PEPI: the clearest long-term lipid comparison

The Postmenopausal Estrogen/Progestin Interventions trial randomized 875 healthy postmenopausal women aged 45 to 64 to one of five arms for 3 years: placebo, conjugated equine estrogen 0.625 mg daily alone, CEE plus cyclic medroxyprogesterone acetate 10 mg for 12 days a month, CEE plus continuous MPA 2.5 mg daily, or CEE plus cyclic micronized progesterone 200 mg for 12 days a month [2].

Mean HDL cholesterol change separated the arms into three statistically distinct groups [2]:

• Placebo: -0.03 mmol/L, about 1.2 mg/dL, a fall [2]
• CEE alone: +0.14 mmol/L, the largest rise [2]
• CEE plus cyclic micronized progesterone 200 mg: +0.11 mmol/L [2]
• CEE plus either MPA regimen: +0.03 to +0.04 mmol/L [2]

Micronized progesterone preserved most of the HDL rise that estrogen produced on its own, while the MPA regimens preserved a much smaller part of it. Two design details matter when this trial is quoted. The estrogen was conjugated equine estrogen, never estradiol, and the placebo group took placebo pills with no hormone at all [2].

One further PEPI measurement cuts against a tidy story. In 365 participants, all four active hormone preparations raised C-reactive protein and lowered soluble E-selectin, with no differences among the treatment arms. Relative to placebo, and pooling the active arms, final CRP concentrations were 85 percent higher and E-selectin 18 percent lower, while von Willebrand factor and coagulation factor VIIIc were unaffected [3]. On that inflammatory marker, micronized progesterone behaved like the other hormones rather than better than them. The authors noted that PEPI was not designed to assess clinical endpoints [3].

KEEPS: four years of artery imaging with oral progesterone

KEEPS randomized 727 healthy menopausal women aged 42 to 58 who were 6 to 36 months past their last menstrual period and had a coronary artery calcium score under 50. Treatment was oral conjugated equine estrogen 0.45 mg daily, or transdermal 17-beta-estradiol 50 mcg daily, each combined with oral progesterone 200 mg for 12 days a month, or placebo, for 48 months [4].

The primary endpoint was annual change in carotid artery intima-media thickness. Mean increases were about 0.007 mm per year and similar in every group. The share of women whose coronary calcium score rose did not differ significantly between groups, and neither estrogen preparation changed blood pressure. Oral CEE improved LDL and HDL cholesterol and raised C-reactive protein and sex hormone-binding globulin; transdermal estradiol reduced insulin resistance. Serious adverse events did not differ by treatment [4]. Because both active arms included oral progesterone, KEEPS cannot separate the effect of progesterone from that of the estrogen, and the authors state that power to compare clinical events was insufficient [4].

Why the Women's Health Initiative numbers are not Prometrium's numbers

WHI assigned 8,506 women to conjugated equine estrogen 0.625 mg plus medroxyprogesterone acetate 2.5 mg daily and 8,102 to placebo [5]. The progestin was MPA, not micronized progesterone. Across the trial the label reports relative risks of 1.18 for coronary heart disease (95 percent CI 0.95 to 1.45), 1.37 for stroke, 1.87 for deep vein thrombosis, 1.98 for pulmonary embolism and 1.24 for invasive breast cancer [1][5].

The label says that only the 0.625 mg CEE plus 2.5 mg MPA combination was studied in that trial, and that 'the relevance of the WHI findings regarding adverse cardiovascular events to lower CE plus other MPA doses, other routes of administration, or other estrogen plus progestogen products is not known' [1].

Does the timing of starting therapy matter?

A secondary analysis pooled both WHI hormone trials (10,739 women with prior hysterectomy randomized to conjugated equine estrogen or placebo, and the 16,608-woman CEE-plus-MPA trial) and looked at coronary heart disease by years since menopause [6]. Neither trial used micronized progesterone. The hazard ratio for coronary heart disease was 0.76 (95 percent CI 0.50 to 1.16) for women less than 10 years past menopause, 1.10 (0.84 to 1.45) at 10 to 19 years, and 1.28 (1.03 to 1.58) at 20 or more years, with a trend P value of 0.02. Stroke risk was raised overall (hazard ratio 1.32, 95 percent CI 1.12 to 1.56) and did not vary significantly by age or time since menopause [6].

ELITE tested the same question with imaging in 643 women stratified as less than 6 years or 10 or more years past menopause. They received oral 17-beta-estradiol 1 mg daily, with progesterone 45 mg vaginal gel for 10 days of each 30-day cycle in women with a uterus, or placebo. In the early group carotid intima-media thickness rose 0.0078 mm per year on placebo versus 0.0044 mm per year on estradiol (P=0.008). In the late group the rates were 0.0088 and 0.0100 mm per year (P=0.29) [7]. The progesterone in ELITE was a vaginal gel, not Prometrium capsules.

Blood clots: progesterone compared with the estrogen route

ESTHER was a multicenter case-control study in France of 271 women aged 45 to 70 with a first idiopathic venous thromboembolism and 610 matched controls. Compared with nonusers, the odds ratios were 4.2 (95 percent CI 1.5 to 11.6) for oral estrogen, 0.9 (0.4 to 2.1) for transdermal estrogen, 0.7 (0.3 to 1.9) for micronized progesterone and 3.9 (1.5 to 10.0) for norpregnane derivatives [9]. The comparison group was nonusers rather than another progestin, and case-control designs carry confounding.

Whatever that observational signal means, the label still contraindicates Prometrium in women with active or prior deep vein thrombosis or pulmonary embolism [1]. It also says that, if feasible, Prometrium should be discontinued at least 4 to 6 weeks before surgery associated with increased thromboembolism risk, or during prolonged immobilization [10].

Animal and cell studies

Minshall and colleagues (Journal of Clinical Endocrinology and Metabolism, 1998) treated ovariectomized rhesus monkeys for 1, 2 or 4 weeks with estradiol, natural progesterone and/or therapeutic-level MPA. Coronary vascular muscle cells from ovariectomized animals were hyperreactive to serotonin and to a thromboxane mimetic. The hyperreactive calcium responses were abolished by estradiol and/or progesterone, but cells from animals given estradiol with MPA stayed hyperreactive [11].

Simoncini and colleagues (Endocrinology, 2004) reported that in human endothelial cells progesterone increased nitric oxide synthesis and potentiated estradiol signaling, whereas MPA did not and impaired estradiol signaling, with similar findings in aorta from ovariectomized rats [12]. These are laboratory and animal findings and do not establish an outcome in women.

What the label says about cardiovascular safety

• Fluid retention: progesterone may cause some degree of fluid retention, and women with cardiac or renal dysfunction warrant careful observation [1].
• The label says to discontinue Prometrium immediately if a pulmonary embolism, deep vein thrombosis, stroke or myocardial infarction occurs or is suspected [10].
• Postmarketing voluntary reports include hypertension, hypotension, tachycardia and circulatory collapse; the label states that frequency cannot always be reliably estimated from such reports [1].

The gap that remains

A Danish open-label randomized trial did count clinical events in recently postmenopausal women, but not with progesterone. It randomized 1,006 women aged 45 to 58 to hormone therapy or no treatment. Women with a uterus received triphasic estradiol plus norethisterone acetate; women who had undergone hysterectomy received estradiol 2 mg alone. After 10 years the composite of death, heart failure admission and myocardial infarction occurred in 16 treated women versus 33 controls (hazard ratio 0.48, 95 percent CI 0.26 to 0.87, P=0.015) [13]. The progestin was norethisterone, so the result cannot be assigned to micronized progesterone.

None of the oral micronized progesterone studies above counted heart attacks and strokes as their outcome. KEEPS and PEPI provide imaging and risk-factor data [4][2]. In PEPI, micronized progesterone preserved more of the estrogen-associated HDL increase than the MPA regimens. KEEPS found similar carotid-wall changes across its treatment groups, but it did not compare oral progesterone with MPA [2][4].