Aortic stenosis (AS) is a common valve disorder in an ageing population in western countries, and women, with longer life expectancy, comprise a substantial percentage of elderly patients with AS. Compared with men, women exhibit distinctive characteristics at the level of stenotic valve leaflets and subsequent compensatory responses of the left ventricle to chronic pressure overload, and in clinical presentation, consequences and response to intervention. Randomised controlled trials of transcatheter aortic valve implantation have yielded new evidence of sex differences in both short-term and long-term outcomes after intervention. A comprehensive knowledge of the existing evidence may inform our understanding of gender differences during assessment and treatment of patients with AS. In this paper, we review the available evidence regarding sex differences in AS in terms of symptoms, clinical presentation, anatomical differences and pathophysiological progression, management and outcomes after aortic valve replacement. Implications for further research are suggested.
- aortic stenosis
- prosthetic heart valves
- valvular heart disease
- cardiac imaging and diagnostics
Statistics from Altmetric.com
- aortic stenosis
- prosthetic heart valves
- valvular heart disease
- cardiac imaging and diagnostics
Aortic stenosis (AS) is a common valvular heart disease in the ageing population in western countries. The prevalence of AS increases rapidly over the age of 65 years, reaching 1.3% between 65 years and 74 years, 2.8% after 75 years and 9.8% from 80 years to 89 years.1 Although epidemiological studies have indicated a similar prevalence of AS between men and women, sex differences have been detected in pathophysiological progression, flow gradient patterns, clinical presentation and prognosis. Surgical aortic valve replacement (SAVR) is a class I recommendation for symptomatic patients with severe AS,2 but gender differences in referral and outcomes have been observed. Randomised controlled trials to assess the safety and efficacy of transcatheter aortic valve implantation (TAVI) have provided new insights regarding gender differences in the outcomes of patients with severe AS.3 4 This paper reviews the characteristics of AS in women, and sex differences in the pathophysiological adaption and treatment outcomes in AS. These observations indicate a need for further research.
Aortic valve (AV) morphology
Sex differences have been observed in terms of valvular anatomy, calcification and fibrosis in patients with AS, impacting the haemodynamic consequences. A normal AV has three thin cusps and an area indexed for body surface area (BSA) of 1.8±0.2 cm2/m2. Women have a smaller AV annulus and shorter distance from annulus to coronary ostia, corresponding to their smaller BSA. The prevalence of rheumatic valve disease has substantially decreased in developed countries, and degenerative AS secondary to leaflet calcification and fibrosis is presently the most common aetiology of AS.1 Congenital bicuspid aortic valve (BAV) accounts for 0.5%–2% of all causes of AS5 with an estimated prevalence threefold to fourfold more common in men than in women. Patients with BAV develop AV dysfunction 10–15 years earlier than patients with tricuspid AV. However, women with BAV are more likely to develop AS than men, regardless of age of presentation (≤50 years: 75% vs 34%; 51–60 years: 78% vs 37%; 61–70 years: 94% vs 70%). Men ≤70 years with BAV more often present with aortic regurgitation (AR), although the frequency of development of AR in men with BAV decreases with age.6 For tricuspid AV, aortic valve calcification (AVC) evaluated by multidetector CT (MDCT) had good association with haemodynamic severity of AS assessed by Doppler echocardiography (AV area calculated by the continuity equation and mean gradient by the modified Bernoulli equation) in both sexes, but women displayed lower AVC load than men for the same haemodynamic severity, even after adjusting for the smaller BSA or aortic annulus area.7 Thus, the odds of high-AVC load were much greater in men (OR 5.07; p<0.0001). Meanwhile, for a similar AVC load, AS was more severe in women; this remained significant when AVC load was indexed to BSA.
Although it is accepted that the relationship between AVC load and AS severity differs between the sexes, the mechanism remains unclear. Hormone status has been implicated; for example, testosterone has been associated with promoting the calcification of vascular smooth muscle culture.8 Valvular fibrosis may be another important contributor to the development of AS in women. Compared with men, women were prone to have higher valvular fibrosis score and a greater proportion of dense connective tissue in surgically resected AV tissue after adjustment for age, body mass index, aortic annulus diameter and AVC density.9 Thus, the pathogenesis of stenotic valve leaflets appears to involve a predominance of valvular calcification in men versus predominance of valvular fibrosis in women for a given mean gradient or indexed AV area by Doppler echocardiography (figure 1).
Haemodynamic progression of AS
Sex differences in the pathogenesis of AS could contribute to differences in progression. In the Simvastatin Ezetimibe in Aortic Stenosis (SEAS) study, 1611 asymptomatic patients with mild to moderate AS, including 39% women, were prospectively followed for 4 years.10 The annualised increase in mean transvalvular gradient did not differ between sexes (women 4±5 mm Hg/year vs men 4±6 mm Hg/year, p=0.15). Comparable haemodynamic progression of AS between sexes was also found in a study by Tastet et al (n=323, women 30%).11 However, the slopes of correlation between mean gradient progression and valve calcification were steeper in women than in men, suggesting that calcification had a more significant impact on the haemodynamic progress of AS in women. In a smaller study that included 50 (25%) women, women had a slightly greater mean transvalvular gradient progression versus men after 3.2±1.2 years of follow-up (women 5±5 mm Hg/year vs men 3±3 mm Hg/year, p=0.04).12 Presently, there is insufficient evidence to conclude that gender has an impact on the haemodynamic progression of AS.
Sex and flow gradient patterns
Sex differences in flow gradient patterns have been observed among patients with severe AS by Doppler-derived valve area. Among patients with preserved ejection fraction, the subgroup with normal flow/low gradient (≥35 mL/m2 and <40 mm Hg) included a higher prevalence of women (58% vs 27%–42% in the other subgroups). This subgroup had the best outcome.13
Compared with men, women have smaller inner aortic root dimensions, and therefore higher pressure recovery, influencing grading of the AS, and more obesity, which may introduce an error in classification of low flow when flow is indexed to BSA.
Left ventricular remodelling in AS
The pathophysiological response of the left ventricle (LV) to chronic pressure overload in AS may differ in men and women. Early M-mode and two-dimensional transthoracic echocardiography data suggested that women developed greater LV wall thickness given a similar severity of AS.14 15 However, in a recent study, women and men had similar LV mass (LVM) and prevalence of LV hypertrophy for similar AS severity, regardless whether LVM was indexed for BSA or height1.7.16 However, with cardiac MRI (CMRI), women often exhibited lower LVM with a smaller LV cavity size, whereas men developed a larger LV cavity and greater LV wall thickness and mass for similar AS severity and comorbidities.17 18 Treibel et al observed that women with AS presented more often with normal geometry and concentric remodelling pattern of the LV, while men developed concentric hypertrophy or eccentric hypertrophy.19 LV myocardial fibrosis by CMR is an objective marker of LV decompensation and contributes to cardiac events.20 Two patterns of fibrosis are typical: focal fibrosis is irreversible, occurs late in the natural history of a disease and can be identified by late gadolinium enhancement (LGE), while diffuse fibrosis occurs earlier, is reversible and can be quantified by extracellular volume (ECV) fraction by T1 mapping.21 Compared with women, equal or higher LGE and ECV have been observed in men with severe AS.17 19 In contrast, another study of 249 patients with less severe AS (mild to moderate AS in 70%), women presented with greater ECV fraction and similar LGE across the whole range of AS severity.18 There was not a significant relationship between LVM index and ECV in women, but was in men. Women appear to have a greater degree of diffuse fibrosis during chronic pressure overload. These differing findings in the pattern of LV fibrosis between sexes may be due to differences in baseline characteristics including age, severity of AS and concomitant comorbidities. However, there is unequivocal evidence that the degree of myocardial fibrosis differs between sexes. Myocardial deformation may be readily evaluated by speckle tracking echocardiography (STE) and is associated with the severity of myocardial fibrosis defined by LGE.22 STE opens a new window to assess LV remodelling in AS and may help to characterise sex differences.
Symptom development and presentation
Patients with AS are typically asymptomatic for many years as stenosis progresses, and symptom onset, an indicator of LV decompensation and poor prognosis in severe AS, is a trigger for AVR according to guidelines.2 Women were likely to report shortness of breath and more often presented in New York Heart Association functional class III/IV with poor mobility and short 6 min walking distance. Men were likely to have chest pain, which may be partly due to a higher incidence of concurrent coronary heart disease.19 23 Besides symptoms, the baseline profiles between men and women also differed in most studies on severe AS (table 1). For the same AV area and haemodynamic impairment, women were older, more often frail, with higher Society of Thoracic Surgeon Score (STS) and higher prevalence of hypertension, chronic obstructive pulmonary disease and anaemia. Men more often had atherosclerotic comorbidities including coronary artery disease and a history of smoking. These baseline differences may contribute to mortality differences. Moreover, the development of symptom onset in asymptomatic AS patients differed. A prospective study by Singh et al 24 found that, given a similar severity of AS at baseline as defined by indexed aortic valve area (AVA), women had earlier onset of symptoms. Symptoms developed in 21.8% of men and 43.9% of women during 1-year follow-up (p=0.004). However, it is challenging to attribute symptoms to AS, especially in the elderly, and for those unable to exert themselves because of comorbidities. Thus, relying entirely on symptom onset to guide the timing of AVR is problematic for either sex.
Referral patterns for AVR
Referral bias can contribute to varying patient outcomes. A retrospective report of unoperated patients with severe AS from 10 US medical centres from 2007 to 2009 included 952 patients with severe AS; 79% of the patients were symptomatic, but only 52% of the patients were referred to a cardiothoracic surgeon, and 41% underwent AVR.25 Men, younger patients and those with symptoms or with higher gradient were more often referred for SAVR. In a 2007 study, only 31% of patients with severe AS underwent SAVR; 74% women and 59% men did not undergo SAVR.26 Another analysis of a US inpatient cohort also found higher performance of SAVR in men (63% of the patients who underwent SAVR were men).27 The reasons for referral differed between sexes: women with severe AS were older, with more atypical symptoms (dyspnoea and exertional dizziness), whereas men with AS had higher prevalence of concomitant coronary artery disease, which may increase referral for intervention including coronary revascularisation. Women tended to perceive their cardiac disease as less severe than men and were more often reluctant to undergo invasive procedures.28 Differential physician referral rates between sexes have been observed, even given the same patient baseline characteristics.29 AS symptoms may be underestimated,26 and operation may not always be offered to elderly patients.30 In data from the claims of national insurance companies in 2005, women and those ≥80 years had lower rates of specialist visits, diagnostic testing and AVR.31 Nowadays, in the TAVI era, referral patterns are changing; women with severe AS are more likely to be referred for less invasive TAVI. Among most of the published TAVI studies, women accounted for about half of the study population.32 In a global TAVI registry dataset, 58% of the patients undergoing TAVI from 2007 to 2018 were women.33 It is difficult to know if this is proportional to gender differences in the prevalence of disease among the elderly. The increasing utilisation of TAVI is expected to improve the long-term outcome of women with severe AS.
Cardiovascular outcomes during the progression of AS
Prior surveillance of cardiovascular outcomes in patients with AS were confined mainly to the severe stage of AS. Recently, registry data from National Echocardiographic Database of Australia suggested that even moderate AS is associated with poor long-term survival, the distribution of the risk of mortality among the quartile groups of mean AV gradient is ‘J’ shaped and the pivotal point of increased risk of mortality was a mean AV gradient 18–20 mm Hg.34 Higher mean AV gradient, male sex and age were associated with increased risk of both all-cause and CV-related mortality during long-term follow-up. In the prospective SEAS study of initially asymptomatic patients with mild to moderate AS, patients with known coronary artery disease, diabetes mellitus and peripheral arterial disease were excluded.10 During a median follow-up of 4 years, compared with men, women had 31% lower all-cause mortality, 40% lower rate of ischaemic CV events and 50% lower stroke and coronary artery bypass grafting, independent of age and time-varying valve area. However, for AS-related events (combined AVR, hospitalisation of congestive heart failure due to AS or death from CV causes), no sex difference existed. Ischaemia-related CV events occurred more often in men, but importantly, no sex difference in AS-related events was observed. From these studies, we can deduce that sex disparity of atherosclerotic progression may to a large extent account for sex differences in long-term survival in patients with AS.
Outcomes of surgical AVR
Surgical AVR has significantly improved the outcomes of patients with severe AS, but women have appeared to benefit less than men. In a US national database of 166 809 patients who underwent SAVR between 2003 and 2014, women experienced higher in-hospital mortality (5.6% vs 4% for men, p<0.001).27 After using propensity score matching of 28 237 pairs of men and women with isolated AVR, in-hospital mortality remained higher in women (3.3% vs 2.9%, p=0.001), and mortality was consistently higher in women throughout the 12-year study period. In the currently used STS model for risk evaluation for SAVR, which was developed from 109 759 patients who underwent isolated single aortic or mitral valve surgery, female sex was associated with a higher predicted risk (HR 1.23, 95% CI 1.10 to 1.36).35 However, in several studies, no sex difference was observed in the incidence of early complications and long-term survival after SAVR, and women even had better survival and lower rate of reoperation in the oldest age quintile (≥79 years) and among those receiving a bioprosthesis.36 37 Although the data thus vary regarding sex differences in outcomes after SAVR, SAVR was performed less often in women in current clinical practice. Accumulating evidence favours the less invasive procedure, TAVI, in women.
Outcomes of transcatheter AVI
Multiple studies from different countries have suggested that TAVI is of considerable benefit in women (table 2). In a randomised controlled study (CoreValve US High-Risk Pivotal Trial, n=353), TAVI-treated women experienced a lower 1-year all-cause mortality over SAVR-treated women (12.7% vs 21.8%; p=0.03).38 Moreover, women had similar or favourable 1–2 year outcomes compared with men in most of the early TAVI studies.32 39–42 In the US national 2011–2014 database (n=23 652, 49.9% women), 1-year mortality after TAVI was 21.3% in women vs 24.5% in men (p<0.001).23 A meta-analysis with longer follow-up (3.28±1.04 years) also found women had better survival after TAVI versus men.43 Nonetheless, women had higher rates of major bleeding events and vascular complications after TAVI during hospitalisation but no difference in 30-day death, stroke or cardiovascular events (figures 2 and 3).23 32 33 Advanced age and smaller vessel, annulus and LV outflow tract dimensions were probably contributors to the higher incidence of vascular complications and major bleeding in TAVI-treated women. The incidence of these complications has decreased (from 6.5% in 2012–2013 to 4.2% in 2014) due to improvements in the delivery system, which has become smaller and more flexible, as well as the prosthetic valve apparatus, with more sizes of both self-expanding and balloon-expandable TAVI valves now available.44 The mechanism for favourable survival of women with TAVI likely includes lower valvular AVC load, smaller annulus size and less patient prosthesis mismatch (PPM) incidence compared with SAVR, as well as fewer comorbidities of women at presentation.23
Patient prosthesis mismatch
After SAVR, women develop PPM at higher rates than men (58% vs 36%), with occurrence related to aortic root diameter and prosthesis type.45 PPM was associated with increased morbidity and mortality at short-term and long-term follow-up after SAVR.46 Compared with SAVR, the frequency of PPM was lower following TAVI in which no sewing ring occupies the annular space. Women have had equivalent or even less PPM than men after TAVI (PPM rates 15%–38%) although women frequently need smaller valves.45 TAVI in women and in patients with smaller BSA may favourably impact long-term outcomes.
Regression of LV remodelling after AVR
Gender differences in reverse remodelling after AVR have been studied. In one study, LV hypertrophy reversed more frequently in women than in men after SAVR (increased LV diameter by echocardiography persisted in 34% of men vs 12% of women at 3±1 days after SAVR compared with preoperatively, p<0.023), and women exhibited more pronounced regression of LVM early after surgery.15 In contrast, in another study, at 6 months after AVR, there were no sex-related differences in percentage reduction of baseline-indexed LVM after AVR by CMR (postoperatively, men 21.7%±10.1% vs women 18.4%±11.0%, p=0.121).17 Instead, the main predictor of each reverse remodelling parameter was the baseline level of that parameter, not gender. Adaption after AVR is a consequence of the relief of transvalvular pressure and LV wall stress, while LVM regression would theoretically take longer after AVR. Thus, data preclude conclusions regarding sex differences in LV remodelling after AVR, and further research is needed.
Implications for future research
Although there are encouraging data suggesting that women have better outcomes after TAVI, several considerations merit further attention. First, most TAVI-treated patients were octogenarians, and a physiological long survival trajectory of these women may contribute to these findings. However, promising evidence of favourable outcomes for women after TAVI should prompt the development of a new risk prediction model for TAVI. Non-invasive assessment of calcification load by MDCT, LV fibrosis by CMRI and LV deformation by STE offer promise as supplementary objective indicators to determine the optimal timing of AVR, as the development of symptoms can be confounded by comorbidities and difficult to ascertain in both sexes. Finally, it remains unclear whether sex differences in the outcomes of AVR are driven by differences in anatomy versus pathophysiological adaptation. Development of more suitable valve devices, delivery systems or patient-level individualised AV apparatus could reduce complications caused by anatomical factors, but if pathophysiological adaptive differences are the primary driver, better timing of intervention may be most important.
Women with AS have multiple distinctive characteristics compared with men. These include less calcification and more fibrosis of the valve leaflets, smaller annulus, lower LVM with a smaller LV cavity size, a distinctive adaptive pattern to LV overload, early-onset symptoms and less referral to AVR. Substantial evidence has demonstrated that compared with men, women have a favourable 1–2 year outcome after TAVI. A sex-specific AS evaluation strategy and a new risk prediction model for TAVI should be considered. Further studies to explore the mechanisms of sex differences in patients with AS may generate new ideas to improve disease outcomes.
Correction notice Since the online publication of this article, Figure 1 has been updated to remove the legend that was present in the figure file.
Contributors YS and PAP planned the document. YS drafted the first version. Both authors have contributed to extensive revisions. Both are responsible for the entire overall content.
Funding PAP is supported as the Betty Knight Scripps Professor of Clinical Cardiology Research. YS was supported by a scholarship from Huashan Hospital Fudan University.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Patient consent for publication Not required.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.