From cardiovascular protection to cardiovascular risk: the metamorphosis of menopausal hormone therapy

Viewpoint

On her 50th birthday, every woman in Australia receives a birthday card from the Government with an invitation for a screening mammogram. Not only does this put the spectre of breast cancer foremost in her mind but it ignores the fact that cardiovascular disease, and not breast cancer, is her most dangerous enemy. It would seem more appropriate for her to receive an invitation for fasting cholesterol or glucose. The Australian Bureau of Statistics lists cardiovascular disease, cerebrovascular disease, and dementia as the three most common causes of death in postmenopausal women, with breast cancer coming a long way fourth (Australian Bureau of Statistics 2016). Similarly, in the last 10 years, cardiovascular disease ranks as the highest burden of disease in the Australian population (Australian Institute of Health and Welfare 2023). Postmenopausal cardiovascular disease has become the most dangerous enemy and that which no one fears (Brown 2000).

The evidence for the protective effect of endogenous oestrogen against cardiovascular disease comes from data showing that early menopause is associated with a greater incidence of cardiovascular disease, and the earlier the menopause the greater the incidence (Li et al. 2021). It is also supported by the fact that women are relatively protected from cardiovascular events compared with men until the age of menopause, after which women gradually overtake men. There are multiple biological mechanisms by which oestrogen may favourably influence cardiovascular risk (Mendelsohn and Karas 1999). Oestrogen increases hepatic low-density lipoprotein (LDL) receptors leading to decreased circulating LDL and decreased entry and retention in the endothelium, as well as decreasing lipoprotein a (Lp(a)) a potent risk factor for cardiovascular disease. Conversely, oestrogen decreases hepatic lipase and increases apo A1, increasing circulating high-density lipoprotein (HDL) and thereby increasing reverse cholesterol transport from the epithelium. Triglycerides are increased by oral but not transdermal oestrogen, although given the data on oestrogen replacement and cardiovascular risk, this mechanism of increase doesn’t seem to confer risk. Oestrogen also acts to induce relaxation of smooth muscle by rapid non-genomic effects, through both endothelium-dependent and endothelium-independent pathways, by nitric oxide and cyclic GMP, as well as by genomic effects including increased expression of prostacyclin synthase and nitric oxide synthase.

The results from the Nurses’ Health Study published in 1997 are, therefore, not surprising, showing a 37% reduction (RR (95% CI) 0.63 (0.56–0.70)) in deaths from cardiovascular disease in women who took postmenopausal hormone therapy, even if they harboured increased cardiovascular risk factors like smoking, high cholesterol, high blood pressure, diabetes, parental history of premature myocardial infarction, and obesity (RR (95%CI), 0.51 (0.45–0.57)) (Grodstein et al. 1997). Against that, and perhaps surprisingly, the study did not show an increase in death from breast cancer with hormone therapy, even in those deemed at higher risk because of family history. Therefore, clinicians were comfortable that the balance was in favour of prescribing menopausal hormone therapy (MHT). That is, until the publication of the results of the Women’s Health Initiative (WHI) study in 2002 (Rossouw et al. 2002).

The WHI, a randomised controlled trial, reported both an increase in breast cancer and an increase in cardiovascular disease in women who were randomised to continuous combined conjugate oestrogen (CEE) plus medroxyprogesterone acetate (MPA) compared with placebo (Rossouw et al. 2002). Initially, the statement was made that the ‘increased risks for cardiovascular disease and invasive breast cancer were present across age strata. Results are likely to be generally applicable to healthy women in this age range.’ Not surprisingly, this publication and the media hype surrounding it caused a seismic shift in the attitude of both clinicians and their patients away from MHT. Two years later, the results of the oestrogen-only arm of the WHI were published, showing neither an increase in cardiovascular disease nor in breast cancer (Anderson et al. 2004). Three years later again, the WHI investigators published a more nuanced analysis of their data which revealed the difference between the oestrogen-only and the oestrogen plus progestogen (E + P) MHT, and the difference between starting early after menopause and starting later (Rossouw et al. 2007). There are many reasons to criticise the WHI study but it has highlighted two valuable concepts (Langer 2017). First, it has highlighted the role of progestogen in the risk of both cardiovascular disease and breast cancer, and, secondly, the benefit of an early start to MHT.

The notion of the WHI being classed as a primary prevention study speaks to the question of what is meant by primary prevention of cardiovascular disease. From a vascular biologist’s point of view, it is prevention at the stage of small fatty streaks to prevent progression to atherosclerotic plaques. From a cardiologist’s point of view, it is the prevention of progression of atherosclerotic plaques to clinical coronary events. In fact, the results of the WHI study were predicted by the late Tom Clarkson, based on his elegant studies in non-human primates, showing that immediate institution of oestrogen treatment after ovariectomy prevented atherosclerotic plaque development whereas a delay in treatment did not (Mikkola and Clarkson 2002). The mechanisms for the disparity between the effects of early and late oestrogen therapy are detailed in the review by Mendelssohn – the acquisition of atherosclerotic disease gradually leads to loss of oestrogen receptors and the development of unstable plaques (Mendelsohn and Karas 2005).

It is, therefore, useful to think that prescribing MHT has ‘traffic rules’ – when to start, how to proceed, and when to stop.

The data from the WHI has led to the notion of the ‘window of opportunity’, with the widely promulgated notion that MHT should be started within the 10 years after the last menstrual period or not at all. This notion is based on cardiovascular benefit/risk of starting MHT and not on the efficacy of MHT for menopausal symptoms. It is based on several sources of evidence – Clarkson’s primate data, showing that early oestrogen treatment reduced atherosclerotic plaque whereas later intervention did not (Mikkola and Clarkson 2002), reanalysis of the Nurses’ Health Study, showing that cardiovascular events reduced in early starters of oestrogen (RR 0.66 (0.54–0.80)) or oestrogen + progestogen (RR 0.72 (0.56–0.92)) but not in late starters (Grodstein et al. 2006), the WHI study, showing the association with cardiovascular risk only in later age of initiation of MHT and only with E + P (Rossouw et al. 2007), and Boardman’s Cochrane Review, showing a reduction in major cardiovascular events (RR 0.52 (0.29–0.96)) in women (n = 9629) starting MHT within 10 years of menopause (Boardman et al. 2015).

However, should this notion be thought of as a window that has shut or as a timing disadvantage to be overcome? What if your patient has missed the 10-year ‘window’ and still has menopausal symptoms? There are many such patients. Bothersome menopausal symptoms can last well into the 60s and even 70s (Gartoulla et al. 2015). In an opinion piece, JoAnn Manson, herself one of the WHI investigators, stated ‘reluctance to treat menopausal symptoms has derailed and fragmented the clinical care of midlife women, creating a large and unnecessary burden of suffering’ (Manson and Kaunitz 2016). Which brings us to the question of how to proceed when faced with a symptomatic patient who is more than 10 years past menopause, or who already has elevated cardiovascular risk, or who carries comorbidities. For this patient, one must keep in mind that the treatment is for menopausal symptoms and not for prevention of cardiovascular disease and that one might need to modify therapy, particularly transdermal versus oral oestrogen therapy. There are pros and cons of both. Oral therapy has beneficial effects on lipids, increasing HDL and lowering LDL, but it does increase thrombotic risk and stroke risk in older women. Transdermal oestrogen is preferable for women with migraine or malabsorption and has no increased risk of thrombosis. However, transdermal oestrogen therapy is lipid neutral and, moreover, all the data on cardiovascular benefit reside in studies using oral, not transdermal, oestrogen therapy (Grodstein et al. 1996; Schierbeck et al. 2012; Boardman et al. 2015; Davis and Baber 2022).

There is evidence that is relevant to our decision to initiate MHT in older women for symptom relief, including studies conducted in women with cardiovascular disease. A data linkage study from Finland reported cardiovascular deaths in women starting MHT between the years 1994 and 2009, at known ages (n = 310,305), showing the younger the women had started MHT, the greater the decreases in the cardiac death (P < 0.05 for trend), upholding the early start/greater benefit hypothesis (Savolainen-Peltonen et al. 2016). Ninety percent of women used oral oestrogen plus progestogen and 70% in a sequential formulation. Even those who initiated MHT in their 80s had a reduced cardiac mortality compared with age-matched background population standard mortality rate (SMR). The HERS study recruited women with established cardiovascular disease and randomised participants to placebo or oral continuous combined CEE and MPA (Hulley et al. 1998; Grady et al. 2002). Despite using a regimen associated with increased thrombotic risk, there was no overall increase in cardiovascular events in the 6.8 years of the studies (Grady et al. 2002). Perhaps the most daring and ethically challenging study of the oestrogen effect in established cardiovascular disease comes from Rosano et al. (1993). After withdrawal of all anti-angina medications including nitrates, women with established angina, arterial stenosis, and a reproducibly positive stress test were given sublingual oestradiol or placebo before an exercise stress test. Oestradiol reduced chest pain, reduced ST segment depression, increased time to ST depression, and increased total exercise time. Given the rapidity of these responses, oestradiol must be acting via non-genomic effects on vasodilatation. These data may make one more comfortable about prescribing MHT for women with cardiovascular risk or disease but in the setting of established disease perhaps avoiding MHT carrying thrombotic risk.

Although there is evidence that oestrogen is cardio-protective, the addition of progestogen to postmenopausal hormone therapy, essential for women with an intact uterus, opposes some of oestrogen’s cardio-protective mechanisms. The addition of MPA in comparison to oestrogen-only therapy was associated, in the WHI study, with more cardiovascular events, and in the primate studies, with less reduction in atherosclerotic plaque (Clarkson 1999; Rossouw et al. 2007). However, MPA is not the only progestogen available for MHT and progestogens differ widely in their affinity for receptors within the steroid receptor super family (Schindler et al. 2003). Although we lack the evidence from long-term head-to-head studies of different progestogens, there is abundant evidence from their differential effect on surrogate markers of cardiovascular risk to prefer progestogens without androgen or glucocorticoid receptor affinity (Shufelt and Manson 2021). Evidence suggests that progestogens without androgenic affinity, such as progesterone and dydrogesterone, do not negate the beneficial effect of oestrogen on lipids, vascular reactivity, and carbohydrate metabolism and are to be preferred (Stanczyk et al. 2013).

Are there any stopping rules for MHT? The advice from MIMS Australia, which provides prescribing advice for all prescription medicines, that is quoted on all product information, is that one should use the lowest dose of MHT for the shortest amount of time, with the goal being short-term use. It is a statement that leads many doctors to encourage their patients to cease MHT after 5 years. This statement is based on the mistaken belief that menopausal symptoms are a short-term, benign nuisance which always resolve within a few years (Gartoulla et al. 2015). The low-dose/short-term approach is also predicated by the concern over breast cancer risk, overlooking the greater threat from cardiovascular disease.

What are the cardiovascular safety data surrounding continuation of MHT? In several studies, continued use of MHT has been shown to reduce the incidence of cardiovascular events compared with placebo. A large study in Finland showed that MHT use reduced cardiovascular death and the reduction was positively related to duration of MHT use (Mikkola et al. 2015a). Again in Finland, Mikkola et al. examined the cardiac or stroke deaths occurring in women who stop MHT. The study compared the event rate in those who stopped with that in women who continued MHT, and found a significant increase in deaths in those who stopped, whether MHT was taken for less than or more than 5 years (Mikkola et al. 2015b). The putative mechanisms are many – withdrawal of non-genomic vasodilatory effects of oestrogen, reduced oestrogen-induced nitric oxide gene expression, increased sympathetic and decreased parasympathetic activity associated with the return of hot flushes, or return of palpitations and arrhythmias associated with oestrogen withdrawal. In short, it does not appear that continuation of MHT for menopausal symptoms is deleterious to cardiovascular health – quite the reverse. The advice to use MHT in the lowest dose and for the shortest amount of time is not supported by the data. There is one agreed stopping rule and that is hormone-receptor-positive cancer, although even that may be a relative contraindication depending on circumstance (Bluming 2022).

Guidelines published by the International Menopause Society recommend that there be no mandatory limit to MHT prescribing (Baber et al. 2016). The manifestations of the oestrogen deficiency of menopause are protean – vasomotor symptoms, urogenital atrophy, arthralgia, mood swings, insomnia, weight gain, sexual dysfunction, and bone loss, to mention but a few. Consideration of MHT should be part of the overall strategy, including diet, exercise, avoidance of excessive alcohol intake, and smoking cessation, to maintain midlife health and wellbeing (Baber et al. 2016). It should be used in the dose required to address these treatment goals and for as long as there is a need. Of course, one may adjust the mode of delivery and the type of MHT according to age and circumstance. From this point of view, there is much to refine in terms of the metabolic and health effects of the combinations and permutations of MHT – oestrogen alone or combined oestrogen and progestogen, androgenic or non-androgenic progestogens, continuous or cyclical therapy, oral or transdermal oestrogen delivery, intravaginal or intrauterine progestogen delivery – in order to improve efficacy, safety and acceptibility for postmenopausal women of all ages. One size does not fit all.

Finally, the WHI study, which caused this angst in the first place, has published its follow-up 15 years after the initial publication, showing that neither cardiovascular mortality nor all-cause mortality was increased by MHT (Manson et al. 2017). It is time to consider quality of life (Utian and Woods 2013).

In conclusion, the metamorphosis from perceived cardiovascular protection to perceived cardiovascular risk, a perception largely led by the WHI study, has shifted the clinician’s risk/benefit decision-making away from prescribing MHT for menopausal symptoms. A more nuanced view of the literature and the greater availability of different modes of delivery and formulations of MHT should give the clinician the knowledge and confidence to prescribe appropriate MHT for the greater benefit of midlife health and cardiovascular safety.