Pathophysiology of Aortic Valve Stenosis: Is It Both Fibrocalcific and Sex Specific?

Abstract

Our understanding of the fundamental biology and identification of efficacious therapeutic targets in aortic valve stenosis has lagged far behind the fields of atherosclerosis and heart failure. In this review, we highlight the most clinically relevant problems facing men and women with fibrocalcific aortic valve stenosis, discuss the fundamental biology underlying valve calcification and fibrosis, and identify key molecular points of intersection with sex hormone signaling.

Perspectives from the Population

Surgery for aortic valve stenosis is the most commonly performed procedure for heart valve disease in developed nations (86, 111). Age is the most significant risk factor leading to aortic valve stenosis, with those above 65 at greatest risk of progressing (134) from simple valvular sclerosis (i.e., a stiffened valve) to severe stenosis (i.e., functional impairment of valvular opening) (152). As the population of patients over the age of 65 has progressively grown (>10,000 people turning 65 yr of age every day in the U.S. alone), heart valve replacement has grown to be the second most common thoracic surgical procedure performed worldwide. Although there have been exponential improvements in the durability of bioprosthetic and mechanical valves in the past 30 yr, our understanding of the biology of valve disease—as well as perspectives on how to harness it therapeutically—have only begun to mature over the past decade.

Prevailing Biological Paradigms in Valvular Disease

For decades, valve calcification was considered to be the result of the passive accumulation of calcium associated with progressive “wear-and-tear” of the valve (25), resulting in the formation of amorphous calcium deposits on the aortic side of the valve. Seminal studies by Mohler et al. (124) and Rajamannan and colleagues (153, 155), however, reported evidence of bone matrix in aortic valves from a subset of patients with end-stage valve disease. This observation spurred nearly two decades of investigation testing the hypothesis that valve calcification proceeds by an active process similar to that observed in bone (53, 63, 99). As will be discussed in detail in ensuing sections, there are robust data suggesting that valve calcification is an active, regulated process that can proceed by both osteogenic and non-osteogenic processes (180, 185). Although it is clear that inflammation, lipoprotein profiles, and matrix remodeling are key factors initiating and modulating osteogenic signaling in stenotic valves (99, 124), the impact of sex as a key modulator of valvular phenotypes has only recently emerged and remains poorly understood (FIGURE 1).

FIGURE 1.

Differences in aortic valve phenotypes in males and females with FCAVS

Differences in aortic valve phenotypes (top) and ventricular phenotypes (bottom) in males and females with FCAVS. Note that, for any given level of severity of valvular stenosis (e.g., mild, moderate, or severe), women have lower valvular calcium levels and lower valvular calcium density compared with their male counterparts. Importantly, these changes persist after normalizing for body size (since women generally have a smaller valvular apparatus in proportion to body size), suggesting that fibrosis may a greater contributor to valvular dysfunction in women compared with men. Interestingly, ventricular adaptations to chronic, progressive increases in afterload can also differ between men and women. Although men generally develop a fibrotic, dilated cardiomyopathy in response to severe FCAVS, women often exhibit more concentric and hypertrophic changes in the myocardium with ventricular dilatation and failure occurring in later stages of the disease. Collectively, the biological underpinnings of these sex differences in the aortic valve and ventricle remain largely unexplored in FCAVS.

Calcific vs. Fibrocalcific Valve Stenosis: Is It Worth Discerning?

Although non-invasive imaging (20) and histopathological analyses show that massive accumulation of valve calcification is a near ubiquitous finding in patients with end-stage aortic valve stenosis, recent work suggests that extracellular matrix accumulation may also play a critical role in the development of valvular dysfunction (99). Perhaps more importantly, our ensuing sections highlight a growing body of evidence that there is a fundamental difference in valvular calcium burden between men and women with the same degree of valvular stenosis (172). This not only has profound implications for our understanding of mechanisms that contribute to impairment of valve function in a variety of patient populations but is also likely to shape our view of “druggable” molecular targets that aim to slow progression of valvular dysfunction. Collectively, we propose that men and women with aortic valve stenosis exist on a spectrum of fibrocalcific disease and consequently will refer to the disease process as fibrocalcific aortic valve stenosis (FCAVS) from this point forward.

With these tenets in hand, our ensuing sections aim to highlight key clinical, surgical, and biological observations suggesting that sex is a critical modulator of pathological processes that contribute to the development of FCAVS, and conclude with the implications of these observations for future research in valve disease.

Clinical and Surgical Pathophysiology

Impact of Sex on the Natural History of FCAVS: Insights from Clinical Cardiology

Auscultation of a prominent mid-systolic heart murmur during routine physical screening remains the primary gateway to further testing for definitive diagnosis of FCAVS. This auscultation “signature” is generally robust enough to allow for echocardiographic diagnosis of FCAVS in relatively early stages of the disease (i.e., mild-to-moderate stenosis), which has allowed for accumulation of several key insights into the natural history of FCAVS, and recent emergence of insights into the impact of sex on the progression of FCAVS. Doppler-echocardiography directly shows the valve and its movements (68), with the caveat that ultrasound does not accurately distinguish between the relative contributions of fibrosis or calcification to valvular dysfunction. Doppler-echocardiography also evaluates the hemodynamics of the aortic valve, with signs that measure the cardiac overload (pressure overload measured by the velocity through the valve, trans-valvular pressure gradient, or ventriculo-valvular impedance) (49) and the severity of the lesion (aortic valve area) (137, 167). Guidelines establish that mean gradient of ≥40 mmHg and valve area of <1.0 cm² are markers of severe AS (19, 133). However, a growing subset of patients present with a small valve area but a low transvalvular pressure gradient (71). These apparently discordant patients with “low-gradient severe FCAVS”—who often have severely compromised left ventricular function—have been reported alternatively as representing severe (109) or generally moderate FCAVS (174), leaving the real nature of severe low-gradient FCAVS somewhat uncertain. To resolve this uncertainty, the concept of measuring the severity of the calcification load, mechanistic to FCAVS (117, 159), has taken hold to quantify the severity of the aortic valve stenosis (48). This method has revealed that these “low-gradient severe FCAVS” patients represent a heterogeneous group (40) and also provided new insights into the natural history of FCAVS and showed several stringent differences between genders (3).

First, hemodynamic progression of FCAVS is unrelenting and over time leads to severe valvular dysfunction. Patients with FCAVS incur highly variable rates of hemodynamic progression, with rates of progression ranging from 3 to 19 mmHg per year (72, 127, 141). Generally, predictors of aggressive progression of FCAVS in term of transvalvular gradient include presence of moderate disease (peak velocity of >3.0 m/s) (63, 140), higher valve calcium at baseline (159), poor functional status score, older age (131), renal insufficiency (16, 187), and metabolic syndrome (16, 87, 149). Although male sex has been touted as a predictor of rapid progression of FCAVS, numerous prospective and retrospective studies have also reported that sex does not predict accelerated progression of FCAVS (46, 87, 140, 142, 147, 159). Few data are available on the link between progression of valvular calcification and hemodynamic progression (or viable biomarkers to predict progression of either) (132), leaving doubt as to whether sex is in fact linked to faster progression.

Second, the natural history of progression of valvular calcification in patients with FCAVS is poorly studied. Valvular calcium is strongly associated with the severity of valvular dysfunction in patients with FCAVS (45). Importantly, the rate of accrual of valve calcium in patients with FCAVS is closely associated with the baseline level of valvular calcium. In the general population, patients who are in the upper tercile of CT-measured valvular calcium (e.g., >300 AU) (118, 149) are at greatest risk of rapid progression of valvular calcification, whereas patients in the lower terciles were not likely to have significant increases in valve calcium. In patients who have reached the point of established aortic stenosis and continue to be monitored medically, total AVC progresses at a rate close to 200 AU/yr (132), with a significant correlation between calcium and hemodynamic progression. Female sex was associated with borderline lower AVC progression, but hemodynamic progression was not different between men and women. Hence, these data raise the issue of whether, despite potentially lower calcification progression, women incur similar hemodynamic progression through a more fibrotic process than men (132).

The final aspect suggesting a difference between men and women with FCAVS is based on cross-sectional data in patients with AS. Overall, all studies found a strong link between calcium load and hemodynamic dysfunction (48, 117). There is a growing body of data, however, suggesting that women are more likely to have severe hemodynamic obstruction with lower levels of valve calcium than men (3, 163). Importantly, the degree of CT-measured calcification is strongly associated with valve weight in both sexes, suggesting that expansion of the calcified lesion was the primary driver of valvular mass in these subjects (156). One may attribute the lower absolute calcification load to the smaller body size, since it is expected that smaller hearts with smaller valves may contain less calcium. This issue has led to the concept of AVC density, whereby the absolute AVC load is normalized to the size of the aortic annulus (3, 40). However, even using this size-normalized measure of calcification, it is clear that women incur more severe AS for the same AVC density (3) (FIGURE 1). Hence, the thresholds of AVC density representing a severe calcification load are different in men and women based on the link between calcification and hemodynamic dysfunction (40). Importantly, these differences are not related to hemodynamic errors due to Doppler evaluation (39) and are confirmed when the link between AVC density load and survival after diagnosis is examined. Indeed, women incur excess mortality for lower AVC densities than men (41). Hence, there is coherence of anatomic and computed tomographic studies, of physiological and outcome studies to note that the pattern of aortic valve calcification in men and women is different (although less impressive but still significant in women after accounting for their smaller size), suggesting that, in women, there must be an added component of fibrosis contributing to the hemodynamic alteration associated with FCAVS. Although detailed histopathological/calcification studies have not been conducted to date, this observation raises the intriguing possibility of a fundamental difference in which valvular calcium and/or extracellular matrix influence valve function differently in women with FCAVS.

The response to the hemodynamic alterations of AS may also extend to the ventricular status in chronic pressure overload in men and women (28, 29). Women tend to have higher left ventricular systolic function at any given level of valvular stenosis (as measured by aortic valve area) (172) (FIGURE 1). Although this may be in part due to a lower incidence of concomitant coronary artery disease in women with FCAVS, studies of subendocardial biopsies from patients with FCAVS appear to suggest that women have less interstitial myocardial fibrosis (181), which has been strongly associated with left ventricular dysfunction in a variety of conditions (77, 93, 184). Although the directional response would be different in the valve (with more fibrosis) and the ventricle (with less fibrosis), the fact that different responses are observed in women vs. men supports a line of investigation that would uncover the biological pathways leading to such site-specific differences.

Sex and Surgery for FCAVS: Peri- and Post-Operative Insights

At present, valve replacement surgery or transcatheter deployment of a tissue valve remains the only treatment option for patients with severe FCAVS (143) (18). Despite the high and increasing use of trans-catheter procedures, undertreatment of AS remains a considerable issue (109), particularly in elderly patients due to excessive comorbidity, emphasizing the urgent necessity to uncover the biological pathways to valvular calcification and develop medical treatment preventing severe AS. Although there are generally low mortality rates for both procedures (30 day mortality rate of 2–5%) (170) (96), several reports suggest complications following surgical or transcatheter valve replacement can occur at a much higher rate (FIGURE 2). These complications include transfusion for bleeding (up to 50%), prolonged post-operative ventilation (up to 25%), pneumonia (up to 10%) (144), and paravalvular leaks (exceedingly rare with surgical valve replacement but up to 50% with transcatheter approaches, with these outcomes being dependent on the device and approach) (74). Critically, these complication rates are sufficiently high to allow for detection of sex differences in these patient populations.

FIGURE 2.

Risk of peri- and post-operative complications

Risk of peri- and post-operative complications in surgical aortic valve replacement (SAVR) and transcatheter aortic valve replacement (TAVR) procedures in men and women. Note that, compared with men, women are at greater risk for stroke, transfusion, bleeding complications, and vascular complications following both SAVR and TAVR procedures. Although paravalvular leak remains an exceptionally rare occurrence following SAVR procedures, men appear to have an increased risk of paravalvular leak following TAVR, which may be related to the extent and distribution of valve cusp calcification. In line with the increased risk of multiple complications, women tend to have higher in-hospital mortality compared with men following SAVR, but this difference in survival between the sexes does not exist at 1 year. Interestingly, this difference in sex-related survival is reversed in patients in TAVR, where in-hospital mortality is comparable between men and women, but women tend to have superior survival at 1 yr. Although reports of the magnitude of such sex differences in humans are variable, these data nevertheless suggest that future clinical, histopathological, and biological studies aimed at understanding the role of sex in post-operative complications and survival are warranted.

The most common complication following valve replacement surgery is blood transfusion for bleeding/anemia (67), with several studies reporting that women have greater rates of anemia, bleeding, or blood transfusion compared with men following open surgery (160, 186). Although there are a number of studies reporting sex differences for other post-procedural complications (64), these are generally balanced by negative findings or effect sizes that are of questionable clinical significance. An emerging finding with some consistency, however, is the observation that men tend to have higher rates of paravalvular leak following transcatheter valve replacement (1, 175, 176), which could relate to the overall mass or heterogeneous distribution of calcification in the valve cusps of males (42, 85). Thus larger systematic analyses of key complications, following valve replacement procedures appear warranted, would be most robust if performed in a prospective manner and should be essential to driving future studies aimed at elucidating the biological underpinnings of any sex differences uncovered. Given the already broad scope of this review, we will not cover the complex and multifactorial interactions between hemostatic cascades, wound healing, and sex hormones.

Rapid amelioration of FCAVS through valve-replacement procedures has also provided insights into the ventricular response to reductions in afterload. Importantly, women appear to have greater improvements in left ventricular function following valve replacement compared with their male counterparts (43, 128, 146). Although this may be in part due to superior left ventricular function before surgery (as noted above), it is more likely that post-operative improvements in ventricular function can be attributed to a lesser burden of currently irreversible, deleterious changes in in the ventricle (e.g., interstitial fibrosis).

Biological Bases for Clinical Phenotypes

Biological Underpinnings of Valvular Calcification and Fibrosis: Pathways and Processes

As noted in previous sections, current paradigms suggest that valve disease can progress by both osteogenic and non-osteogenic processes. It is critical to note, however, that the absence of organized bone matrix does not mean that the induction or progression of valve calcification is not a regulated process (FIGURE 3). Thus the ensuing sections will focus on several key pathways implicated in the pathogenesis of FCAVS, followed by an overview of the points of interaction between sex hormone signaling and molecular processes driving FCAVS.

FIGURE 3.

Key canonical signaling pathways contributing to aortic valve calcification and fibrosis in FCAVS

Canonical TGF-β signaling (through phosphorylation and nuclear translocation of Smad2/3) has been implicated in valvular fibrosis, induction of subsets of osteogenic genes in valve interstitial cells, and apoptosis of valve interstitial cells in human tissue and in a variety of model systems. Similarly, activation of other TGF-β superfamily signaling cascades—such as canonical bone morphgenetic protein signaling (through activation and nuclear translocation of Smad1/5/8)—has been implicated in induction of osteogenic gene expression in valve interstitial cells and increases in pro-inflammatory markers in valve interstitial cells and endothelial cells in FCAVS. Finally, activation of canonical Wnt/β-catenin signaling has been linked to induction of osteogenic and fibrogenic processes in valve interstitial cells in both human tissue and in several experimental model systems.

TGF-B signaling in FCAVS.

TGF-β signaling acts in an exquisitely context-dependent manner and is thought to play two key roles in the pathogenesis of FCAVS. First, TGF-β1 has been shown to induce apoptosis in valve interstitial cells in vitro, which is a key event in the initiation and expansion of calcified nodules (83, 103). Second, TGF-β1 is a powerful driver of extracellular matrix protein elaboration (73, 164) and matrix remodeling enzymes. Thus, in the right microenvironments, TGF-β1—produced either locally or potentially from circulating platelets—has been implicated in both the calcific and fibrotic aspects of FCAVS.

BMP signaling in FCAVS.

Upregulation of BMP2, BMP4, BMP7, and their respective canonical signaling cascades has been reported in numerous studies of diseased human valve tissue and several experimental animal models of FCAVS (10, 58, 130, 162, 191, 192, 195). Critically, genetic inactivation of BMP inhibitors results in widespread ectopic calcification in utero, including aortic valve tissue (66). Furthermore, pharmacological inhibition of BMP signaling significantly attenuates osteogenic differentiation of patient-derived valve interstitial cells exposed to mechanical stretch (148). Finally, amelioration of risk factors driving valvular calcification, such as hyperlipidemia in experimental animal models, results in parallel reductions in both canonical BMP signaling and valvular calcification (119, 120, 122).

Wnt/β-catenin signaling in FCAVS.

Multiple lines of evidence suggest that Wnt/β-catenin signaling is activated in valve tissue from patients with FCAVS, including upregulation of multiple Wnt ligands (Wnt3a, Wnt7a) (5, 6, 26, 31), Lrp receptor complex components (LRP5, LRP6, and Frizzled subunits) (5, 6, 26, 31), and nuclear localization of the activated β-catenin transcription factor complex (6, 31, 121, 153). Furthermore, numerous modulators of Wnt signaling are upregulated in the blood of patients with FCAVS (12). Although the functional significance of circulating Wnt modulators remains unclear at this point, it is evident that elevations in the Wnt inhibitor DKK1 are strongly predictive of increased mortality in patients with FCAVS (12). Recent observations from vascular tissues, however, suggest that LRP6 in vascular smooth muscle may protect against calcification through a non-canonical, arginine methylation-dependent relay (32) and align with the notion that Wnt signaling may play a significant protective role in specific cellular microenvironments.

Caspases and apoptosis in FCAVS.

Evidence of apoptosis [programmed cell death with preservation of internal/external cell membranes (24, 61, 88)] and necrosis (cell death with membrane lysis) in valves from patients and experimental animals with FCAVS has been reported by several investigators (37, 65, 83, 90, 113, 197). Although the mechanistic contributions of cellular apoptosis to the initiation and/or expansion of calcified nodules in vivo remains unclear, there is general consensus that calcific deposits that form following induction of cell death typically have a crystalline ultrastructure and lack live cells within the core of the calcified mass (99). Furthermore, a number of studies have reported that therapeutic interventions that reduce valvular interstitial cell calcification also elicit parallel reductions in markers of apoptosis (37, 65, 83).

Cellular senescence in FCAVS.

Several studies have implicated the induction of cellular senescence (52, 95) and increased senescent cell burden (95, 157, 158, 200) in the pathophysiology of FCAVS and other age-related cardiovascular diseases. Although we are not aware of any data experimentally implicating senescent cells in the pathogenesis of FCAVS, our recent reports that pharmacological or genetic clearance of senescent cells can attenuate calcification and improve nitric oxide signaling in the vasculature (158, 200) suggest additional research in this area is warranted.

Biological Underpinnings of Ventricular Adaptation: Pathways and Processes

Key ventricular (mal)adaptations to chronic pressure overload.

As noted in previous sections, although there are differences in the magnitude of the hypertrophic response to chronic pressure overload between sexes, both men and women with FCAVS have significant increases in ventricular hypertrophy compared with age-matched control subjects (50, 106). Even with the same degree of valvular stenosis, however, it is evident that women present with a concentrically remodeled, less dilated left ventricle, whereas men present with a dilated, eccentrically remodeled left ventricle (13, 28, 55, 80). Importantly, both humans and animals (62, 146, 181) exposed to chronic pressure overload also display sex-dependent changes in interstitial fibrosis of the myocardium, which is thought to play a significant role in the pathogenesis of both systolic and diastolic dysfunction. The ensuing sections will highlight key biological processes driving these changes, as well as the effects of sex hormones on these signaling cascades.

β-Adrenergic receptor signaling.

Increases in sympathetic nervous system activity are a near ubiquitous response to cardiac stressors. In response to acute and/or short-term stressors, β1-adrenergic receptor stimulation initially increases cardiac contractility, preserves cardiac output, and can result in adaptive remodeling and cardiomyocyte hypertrophy (108, 129). Chronic sympathetic hyperactivation and β-receptor stimulation, however, result in excessive ventricular hypertrophy, ventricular fibrosis, and, in later stages, cardiomyocyte apoptosis and progression to heart failure (91, 178). Since there are detailed reviews of the cellular cascades underlying these processes, we will focus on the relatively small number of studies evaluating the effects of sex hormones on β-adrenergic receptor signaling in experimental models of chronic ventricular pressure overload.

Perhaps the most frequently used model to recapture the hemodynamic challenges imposed on the ventricle in end-stage FCAVS is transverse aortic constriction (TAC) in rodent models, which elicits left ventricular hypertrophic and fibrogenic responses similar to those observed in humans and can be inhibited by β-receptor blockade (102). Much like humans, female mice are relatively protected from the hypertrophic and fibrotic changes elicited by TAC compared with male mice (17, 166). Although genetic inactivation of estrogen receptor α in this model does not alter the hypertrophic response to TAC in female mice, genetic inactivation of estrogen receptor β dramatically increases the hypertrophic and fibrogenic responses to TAC. Later studies demonstrated that ER-β-dependent signaling is essential for repressing apoptosis in this model (62) and is required for the phenotypic “rescue” effects of exogenous estrogen (14). Importantly, estrogen receptor signaling has been implicated in repression of p38 MAP kinase (14) and calcium-calmodulin kinase II activities (107), with both enzymes being key effector molecules of β-adrenergic receptor signaling (98).

Wnt/B-catenin signaling in ventricular maladaptation.

Emerging data suggest that Wnt/β-catenin signaling is a significant contributor to ventricular hypertrophy and fibrosis in rodent models of cardiac dysfunction. Perhaps the most compelling data implicating canonical Wnt signaling in pressure overload hypertrophy comes from the observation that genetic inactivation of Disheveled (critical for activation of β-catenin signaling) (177) or haploinsufficiency of β-catenin protein (151) results in substantial protection against TAC-induced cardiac hypertrophy. Critically, increasing soluble frizzled-receptor protein levels in failing murine hearts attenuates several components of adverse cardiac remodeling (21), which has profound implications for therapeutic harnessing of this pathway.

Although there are data implicating interactions between estrogen signaling and Wnt signaling in non-cardiovascular tissues (27, (70, 94), we are not aware of data experimentally determining direct interactions between estrogen and canonical Wnt signaling in the heart. Given the exquisite context dependence of Wnt signaling, we will only suggest that this is an interesting future area for investigation.

TGF-β signaling in ventricular maladaptation.

As noted above, TGF-β is a central driver of fibrogenic and matrix remodeling processes in a variety of tissues, and there is an abundance of data implicating TGF-β in myocardial fibrosis and ventricular remodeling in a variety of contexts (75) (183) (59) (47). Furthermore, numerous studies have shown that attenuation of TGF-β signaling and subsequent reductions in interstitial fibrosis and remodeling significantly improve cardiac function in animal models of pressure overload (171).

Numerous studies in rodents have reported that cardiac fibrosis is dramatically accelerated by ovariectomy and is ubiquitously associated with increases in TGF-β signaling (69). Importantly, both fibrosis and TGF-β signaling are reduced in parallel following administration of exogenous estrogen or estrogen analogs in these models (69, 145), and estrogen can repress Smad3 transcriptional activity in other cell types (112). Reciprocally, canonical TGF-β signaling molecules can also attenuate estrogen receptor signaling (188), suggesting that robust upregulation of TGF-β signaling may attenuate some of the protective effects of estrogen on cardiovascular function. In contrast to the effects of estrogen on fibrosis, androgenic molecules have been shown to augment TGF-β signaling in cardiomyocytes and cardiac fibroblasts, and orchiectomized mice are protected against TAC-induced cardiac dysfunction (126). It is evident, however, that some degree of androgenic signaling is likely to be beneficial in coordinating molecular responses to chronic pressure overload (38, 76). Although the direct therapeutic harnessing of these interactions is complicated, targeting ancillary pathways modulated by sex hormones [e.g., NO signaling (97), etc.] may prove to be a more therapeutically viable way to protect both males and females against pressure overload-induced ventricular dysfunction.

Caveats with most animal models of TAC.

It is important to note that the vast majority of studies using TAC have been performed in relatively young animals, and it is known that lifelong reductions in estrogen secondary to ovariectomy (190) result in cumulative deleterious effects in the left ventricle of aged rodents. Thus the translatability of these studies to aged humans is complicated by the diverse molecular changes that occur throughout the aging organism (e.g., epigenetic aberrations, DNA damage, etc.), and future studies in aged animals will be essential to driving the field forward.

Biological Underpinnings of Sex Differences: A Truly Tangled Web

The biological phenomena underlying sex differences in the aortic valve and left ventricle are incredibly complex, and include factors that are both intrinsic and extrinsic to the cell type of interest. In the ensuing sections, we aim to provide an overview of some of the key molecular mechanisms that may influence development of different times of disease onset as well as development of different phenotypes in males and females.

Intrinsic cell differences between sexes.

In the most reductionist of terms, each cell has a sex that is dictated by its chromosomal complement (123). In vitro, this effect can readily be investigated by the isolation and expansion of valve interstitial cell populations derived from animals of each sex. Although few such studies have been conducted to date, the limited data available suggest that cells from males and females continue to exhibit sex-specific molecular signatures in vitro that are likely to confer altered susceptibility to osteogenic and fibrogenic cell phenotypes in vivo. More specifically, seminal work by Masters and colleagues reported that valve interstitial cells derived from female swine were less likely to express molecular signatures related to inflammation, apoptosis, and cellular proliferation compared with cells derived from their male counterparts (115).

Studies probing the role of sex chromosomes in vivo are far more complicated and require the leveraging of more complex animal models such as the “four core genotypes” model. Critically, offspring in this mouse model have a functional separation between gonadal sex (ovaries vs testes) and sex chromosome complement (XX vs. XY) (11). Again, although data from this model are limited with regard to cardiovascular calcification, recent work suggests that sex chromosome complement can significantly alter blood pressure in gonadectomized mice (51, 82). Perhaps most interestingly, presence of the XX chromosomal complement resulted in greater pressor responses to angiotensin II in gonadectomized mice (82), suggesting that some maladaptive responses in postmenopausal women may not be attributable to simply the loss of estrogen but also due to biological effects encoded within the sex chromosomes themselves.

Effects of estrogen on osteogenic and fibrogenic signaling cascades.

Estrogen signaling impacts a myriad of molecular processes that are likely to play a significant role in the pathogenesis of FCAVS. As highlighted in FIGURE 4, once estrogen diffuses into the cell, it can bind to classical estrogen receptors (ERα/β), estrogen binding proteins (EBPs), or G-protein-coupled estrogen receptors (e.g., GPR30) (54, 114, 116). Depending on the target to which estrogen binds, it can subsequently initiate rapid cellular responses (e.g., non-genomic intracellular signaling cascades) or slower, transcriptionally regulated responses (e.g., nuclear/genomic signaling) (23, 116).

FIGURE 4.

Role of sex hormone signaling in key molecular processes that are known to interact with signaling cascades driving tissue calcification and fibrosis

Note that estrogen can bind to a number of receptor subtypes (ER) and estrogen binding proteins (EBPs) to activate both nonnuclear and nuclear/genomic signaling cascades. Although the majority of estrogen’s effects appear to activate pathways that are consistent with protection against valve calcification and fibrosis, there is a subset of pathways in which estrogen may promote tissue calcification and fibrosis. Perhaps most importantly, most of the studies conducted to date have not used model systems that replicate therapeutic strategies that would be implicated clinically, whereas there are compelling data to suggest that late-life repletion of estrogen may reduce cardiovascular calcification but increase overall morbidity and mortality. Unlike estrogen signaling, androgen signaling tends to have a closer balance between protective and deleterious molecular effects that interact with osteogenic and fibrogenic signaling implicated in FCAVS. Androgens share the complexity of activating both nonnuclear and nuclear/genomic responses following receptor activation, and also display a critical context dependence in which abnormally low androgen levels may favor deleterious effects, “;normal” levels favor cardiovascular protection, and supraphysiological levels again favor deleterious cardiovascular effects. Collectively, there is a tremendous amount of work to be conducted to understand the biological roles of normal changes in sex hormones over the course of the human lifespan in the modulation of FCAVS, as well as the viability of therapeutically targeting sex hormones or their downstream molecular effectors as a viable strategy to slow progression of FCAVS in a sex-specific manner.

Generally, physiological levels of estrogen have been shown to be essential for suppression of p53 signaling and prevention of cellular senescence in young animals (78, 79, 104, 199), suppression of nuclear factor kappa B signaling and inflammation (22, 136, 189), activation of endothelial nitric oxide signaling and increased nitric oxide bioavailability (2, 30, 36), augmentation of antioxidant expression and activity (193), reductions in NADPH oxidase activity (198), and suppression of RANK ligand signaling (138). Importantly, virtually all of these downstream molecular changes elicited by physiological levels of estrogen have been implicated in protection against progression of valvular calcification and production of excessive extracellular matrix. Importantly, the relative contributions of these mechanisms to observations that estrogen can reduce canonical TGF-β signaling (112, 125) and other specific processes driving calcification, fibrosis, and inflammation remain largely undefined, particularly in the aortic valve.

Critically, however, not all of the biological effects of estrogen may protect against progression of valvular stenosis and ventricular fibrosis and dysfunction. More specifically, estrogen has been shown to suppress PPARγ signaling in calcifying tissues (168) and adipose tissue (81) but may upregulate PPARγ signaling in the vasculature of young animals (173) and promote activation of Wnt/β-catenin signaling through both increased elaboration of Wnt ligands and the amplification of canonical β-catenin signaling (60, 92, 196). Although these processes are likely to be key mechanisms underlying estrogen's protective effects on bone, repression of PPARγ and activation of Wnt/β-catenin pathways promote osteogenesis (33, 37) and augment fibrogenesis (4) in soft tissues as well. Thus such mechanisms—as well as yet-to-be-identified changes with organismal aging—may explain some observations that estrogen replacement/repletion may have deleterious effects on cardiovascular outcomes later in life (7, 136, 193).

Effects of testosterone/androgens on osteogenic and fibrogenic signaling cascades.

Androgens such as testosterone also initiate complex, context-dependent effects that are likely to play a role in the initiation and progression of valvular and ventricular phenotypes in FCAVS. Similar to estrogen receptor signaling, testosterone and its related hormones diffuse into the cell and bind to androgen receptors, which consequently can elicit genomic and non-genomic signaling within the cell (36, 105, 179) (FIGURE 4).

The general epidemiological observation that male sex is a major risk factor may lead one to posit that testosterone exerts predominantly deleterious effects in the aortic valve, vasculature, and ventricle of aging males. Physiological levels of testosterone, however, exert an array of biological effects that make the net effect of androgen signaling difficult to predict. For example, much like estrogen, testosterone has been shown to suppress p53 signaling and development of cellular senescence (8, 139), suppress inflammation (84, 135, 169), increase eNOS activity and nitric oxide signaling (8, 100, 139, 194), suppress apoptosis (9, 101, 150, 182), and potentially reduce RANKL signaling (169). This is counterbalanced, however, by observations that testosterone can activate NADPH oxidase and increase reactive oxygen species production (34, 44), repress PPARγ signaling (56, 161), and augment Wnt/β-catenin signaling (165).

Importantly, studies in gonadectomized animals suggest that having “normal” levels of testosterone protects against cardiovascular dysfunction and injury (100, 139). Consequently, there are emerging data that suggest minimization of age-associated declines in testosterone protects against deleterious changes in the cardiovascular system (101, 105, 182). Once supraphysiological levels of testosterone have been attained, however, testosterone's net impact appears to shift toward deleterious effects on the cardiovascular system (105, 182). Since testosterone can be converted to estrogen by aromatase enzymes throughout the body (36, 89), the relative contributions of testosterone and estrogen under such supraphysiological conditions are difficult to dissect.

Key Unanswered Questions and Opportunities

Although there are substantial bodies of literature describing the natural history of FCAVS and outcomes following surgery, there has been surprisingly little investigation into biological mechanisms underlying key clinical and physiological differences between the sexes. Whether this is due to the fundamentally flawed perception that cardiovascular diseases in men and women initiate and progress via identical mechanisms or is due to the difficulties associated with conducting robust mechanistic work is largely irrelevant in today’s day and age. Armed with abundant data describing sex differences in FCAVS, methodological and technical advances now place us in a strong position to make significant strides in our understanding of the role of sex in FCAVS in the fields of clinical cardiology, cardiac surgery, and basic science. Perhaps most importantly, integrative physiological approaches will need to lead the way in this field to understand the complex, context-dependent effects of sex hormones in aging humans and model organisms.

Clinically, advances in non-invasive imaging modalities have allowed us to gain remarkable insights into pathophysiological changes occurring in humans in vivo. This not only includes robust evaluation of calcium mass and distribution within the aortic valve (172) but also the opportunity to evaluate fibrosis in the valvular and ventricular structures (15). Furthermore, newer imaging modalities aimed at evaluating valvular calcium deposition over time may someday provide a much more sensitive readout for disease progression and effectiveness of drug interventions (57, 110). Perhaps most importantly, leveraging such tools to gain a deeper understanding of the natural history of FCAVS in men and women may ultimately allow for novel (and potentially sex-specific) composite end points in clinical trials (35), thereby allowing for accelerated development of novel treatments to delay or prevent surgical intervention.

Surgically, combining superior risk prediction with optimization of patient management to improve postoperative outcomes remains the coin of the realm in defining destination medical centers around the world. By understanding the biology of changes in sex hormones on hemostasis, wound healing, organ recovery, and overall frailty/resilience, cardiovascular centers should not only be able to pick the right procedure for the right patient at the right time but also potentially devise preoperative interventions and/or prehabilitative strategies to optimize patient outcomes in a sex-specific manner.

Scientifically, compared with other fields such as heart failure or atherosclerosis, our understanding of mechanisms underlying FCAVS largely remains in its infancy. We now know that FCAVS is not merely “atherosclerosis of the valve,” making the application of the large body of work from the field of atherosclerosis complicated at best. Furthermore, recent insights from the clinical and surgical studies cited herein showing differences in calcium (and potentially fibrosis) between men and women at a given level of valvular stenosis suggests that different therapeutic strategies may be required to slow progression of valvular dysfunction in these populations, which represents an area ripe for biological investigation using appropriate animal models. Critically, it is now evident that appropriately aged animals should provide a critical context in which to understand the biological outcomes of hormone repletion therapy in both males and females, and careful consideration of both sex hormone levels as well as chromosomal complements will be critical to generating data that accurately translate from bench to bedside. Finally, understanding the influence of sex (both hormonal and chromosomal) on the ventricular responses to chronic, progressive left ventricular overload in appropriate animal models should be of tremendous importance in developing therapies aimed at preserving or restoring left ventricular function in patients with advanced valvular heart disease.

Conclusions

In closing, we believe that clinical, surgical, and biological data collectively suggest there are significant sex differences not only in the molecular underpinnings of FCAVS, but also in the ventricular responses to chronic pressure overload. Such sex differences appear to be evident across the spectrum from discovery to clinical observations, and consequently require the assembly of multidisciplinary teams comprised of clinicians, surgeons, and scientists to ensure the most relevant questions are interrogated with scientific rigor and cutting-edge model systems. Although several funding agencies and journals recently established guidelines and mandates for equal inclusion of both sexes in preclinical and clinical research, we propose that equality in scientific or medical advancements should not be kept by regulations or by force but instead be achieved by rewarding the pursuit of thorough understanding. Just as with other cultural and socioeconomic issues, if we do not embrace equal understanding of sex differences across the spectrum of discovery and translation and continue to pursue questions focused on the study of a single sex, such negligence will delay insights into a major biological context and ultimately delay development of efficacious treatments for both sexes.