The prevalence of calcified aortic stenosis (AS) is increasing, probably as a consequence of the worldwide aging population. Additionally, a further exponential increase in the elderly population demographics is expected by 2050, which further increases the impact of AS
[1][2]. The Tromsø Study assessed the prevalence, incidence, prognosis and progression of aortic stenosis in the general population
[3]. The incidence of AS in the study was 4.9‰/year from 1994 to 2008
[3]. Notably, the overall public health burden of AS is escalating rapidly due to the increasing lifespan and prevalence of risk factors
[4][5][6]. However, according to the current guidelines, no effective medical treatment is available
[7]. The only treatment of AS represents surgical or transcatheter aortic valve replacement with an exponentially growing number of procedures, hence bearing a considerable clinical and economic burden
[1]. Accordingly, there is an unmet need for further medical treatment options capable of slowing disease progression. A comprehensive understanding of the initiation and progression pathways of aortic valve calcification is critical. Nonetheless, a significant proportion of elderly people have no significant AS, indicating that both aging and nonaging pathways are involved in the pathomechanism of the disease
[8][9][10][11][12][13]. Previously, AS was thought to be a passive, degenerative process involving age-related, replicative cellular senescence
[11]. Now it is understood to be a complex process with active elements, including endothelial injury, chronic inflammation, fibrosis, lipid deposition, matrix remodeling and calcium deposition
[8][9][10][11][12][13]. Current routine clinical diagnostic tools can identify only the later stages of the disease when calcification is already present
[7]. Future molecular and imaging diagnostic tools could help to identify the early stages, initiating the preclinical stage of the disease prior to irreversible macroscopic and later-stage valve calcification
[8][13].
2. The Structure of Aortic Valve
The normal structure of the aortic valve is avascular with three semilunar cusps and is part of the aortic root connecting the heart to the systematic circulation
[14][15]. All components of the aortic root, including the annulus, sinuses of Valsalva, sinotubular junction, interleaflet triangles, commissures and the three aortic valve leaflets, interact with each other to maintain optimal coronary perfusion and unidirectional laminar blood flow through the vascular system
[14][16]. The three aortic valve leaflets (or cusps) are named according to the name of the coronary artery derived from the cusp. The right and left coronary cusps are named after the right and left coronary arteries, while the cusp without a deriving coronary artery is determined as a noncoronary cusp. The cusps maintain unidirectional forward blood flow from the left ventricle to the aorta, meanwhile they must be compliant to open and simultaneously must be able to resist the high-pressure environment of the ventricular systole
[14][17]. Previous biomechanical studies, comparing aortic valve anatomical structure-induced mechanical strain alterations, showed that the most optimal anatomical configuration is the trileaflet aortic valve
[15][18][19]. The leaflet strain increases from the base of the valve to the tip, both in the case of tricuspid and bicuspid aortic valves
[15]. However, the geometry of the bicuspid aortic valve increases the overall mechanical stress, mainly at the commissure
[15]. The site of increased mechanical stress usually represents the initiating location of aortic valve degeneration. The bicuspid aortic valve is a congenital alteration as a result of the fusion of two cusps. The prevalence of bicuspid aortic valves is 0.5–2% worldwide, with a male predominance
[20][21]. According to the location of fusion and the presence of fibrous raphe, several morphological types can be distinguished. The most commonly used Sievers classification differentiates three main types according to the number of raphes. There is no raphe in the case of a type 0 bicuspid aortic valve. Only one raphe is present in the case of the most common type 1, and two raphes in the case of type 2 bicuspid aortic valves
[22][23]. The most common right and left coronary cusp fusion is defined as a coronary cusp fusion with a prevalence of 80%. The mixed types of fusions of right and noncoronary cusp fusions or the left and noncoronary cusp fusions are less common (17% and 2%, respectively), however, both of them represent a higher risk factor for the development of aortic stenosis compared to the common coronary cusp fusion.
The aortic valve has an aortic and a ventricular surface side due to its spatial location within the aortic root
[10]. Histologically, both sides are covered by valvular endothelial cells (VEC) to ensure a nonthrombogenic surface layer and to regulate inflammatory reactions
[10]. Valvular endothelial cells represent a barrier on the surface of the valve between the tissue and blood, similar to vascular endothelial cells. However, these cells have a different phenotype compared to vascular endothelial cells in terms of barrier function, proliferative potential and sheer stress response. Diffusion of oxygen and nutrients through the valvular endothelial layer to the inside of the valve is critical, as the valve is avascular compared to the vascular system’s blood supply through the vasa vasorum. Furthermore, the arrangement of VEC is perpendicular to the direction of blood flow and not parallel with it, compared to the arrangement of vascular endothelial cells
[24]. Additionally, the VECs are mechanosensitive that respond to mechanical stimuli
[25].
Each aortic cusp consists of three layers: aortic fibrous layer, spongy layer and ventricular layer. The ventricular layer is composed of radially aligned elastin fibers to promote cusp motion, while the aortic fibrous layer consists of circumferentially aligned collagen fibers
[10][15]. The proteoglycan-rich spongiosa layer can be found between them, encompassing mainly glycosaminoglycans to offer lubrication from shear forces
[10][15]. The valvular interstitial cells (VIC) are quiescent fibroblast-like cells found throughout the three layers, producing and repairing the extracellular matrix over time
[10][15]. Although VICs are considered to be a fibroblast-like population, they have substantial phenotypic plasticity
[10]. The interplay between the cellular and extracellular matrix components of the aortic valve forms an integrated response to the mechanical effects of different hemodynamic situations to maintain normal aortic valve functions
[10][15].
3. Pathomechanism of Aortic Valve Calcification: Senescence and Steps of Calcification
3.1. The Role of Cellular Senescence in Aortic Valve Calcification
Cellular senescence is a feature of somatic cells defined by a nondividing, irreversible cell cycle arrest state due to intrinsic and/or extrinsic factors
[26]. The loss of replicative capacity is a consequence of replication-related telomere shortening or mechanical and metabolic stressors leading to deoxyribonucleic acid (DNA) damage, mitochondrial dysfunction and the accumulation of reactive oxygen or nitrogen species
[26][27][28][29]. It is worth noting that cellular senescence is different from cellular quiescence, which is an adaptive response to the nutrient environment resulting in reversible cell cycle arrest
[26]. Telomeres are nucleoprotein complexes at the cap of the chromosomes with tandem repeats of DNA and a six-protein complex called shelterin
[30]. This cap of the chromosome is largely double-stranded, however, it ends in a short single-strand, resulting in protection and replication difficulties at the end of the chromosome
[30]. The shelterin protein complex binds to the double- and single-stranded telomere DNA and protects it from unwanted degradation
[30]. Without shelterin proteins, the ends of chromosomes could be misrecognized by the DNA damage response and repair machinery as double-strand breaks, which require repair. It is known that DNA polymerase is unable to fully replicate chromosome ends, consequently, it shortens with each replication cell cycle due to DNA loss
[31]. The progressive telomere shortening results in a critical telomere length with each cell division and the somatic cell becomes senescent, also termed as a replicative senescent cell
[32]. To counteract replication difficulties, the telomerase enzyme can synthesize new telomere repeats at chromosome ends, however, it is silent in most somatic tissues and is only expressed in germline cells and a subset of proliferating somatic progenitor cells
[30][32][33]. Thus, normal somatic cells become senescent when their telomeres reach the threshold length.
Immune cells can remove age-related, replicative senescent cells and prevent their accumulation, except in the case of increased stress-induced cellular senescence and immune system dysfunction
[27]. Nonetheless, senescent cells are not inactive cells, as they secrete cytokines, chemokines and matrix metalloproteinases defined as senescence-associated secretory phenotypes (SASP). This contributes to the extracellular matrix remodeling and valvular structural changes associated with aging. The age-related structural changes in the aortic valve include increased collagen content and crosslinking, leading to increased leaflet stiffness, which is adopted for age-related physiologic changes in cardiac hemodynamics in order to maintain the normal aortic valve function
[34][35]. Apart from physiologic, age-related senescence and further pathologic cellular senescence can be induced by cellular stressors, such as excessive mechanical stress, oxidative stress, metabolic stress, and factors leading to DNA damage, also termed stress-induced premature cellular senescence
[26][27]. Chronic excessive stress over the years may result in the accumulation of pathologic senescent endothelial cells, mainly on the aortic side of the valve, where the blood flow is oscillatory. Notably, aging of the immune system, defined as immunosenescence, may result in decreased clearance and accumulation of senescent cells
[27]. However, aging rarely leads to severe aortic valve stenosis, as a significant proportion of the elderly population does not develop significant AS. In addition to aging, excessive mechanical stress, genetic factors and metabolic factors, such as high blood pressure, overweight and hypercholesterinemia, can induce and aggravate pathological cell senescence and calcification
[36][37]. Excessive mechanical stress is present in the case of increasing aortic stenosis severity when the blood flow is oscillatory on the aortic side and turbulent on the ventricular side. Oscillatory flow represents a different mechanical stress compared to the laminar flow of the normal aortic valve. Oscillatory shear stress has been shown to promote atherosclerotic plaque formation in arteries, and this is also assumed to be the case in aortic valve calcification, as calcification is typically present on the aortic side of the valve
[10]. Other than senescence, mechanical stress can lead to focal endothelial damage and denudation
[38].
Mechanical stress induced endothelial denudation and focal tissue disruption is repaired either by activated, somatic quiescent endothelial cell division or circulating endothelially progenitor cell adhesion to the damaged site
[38]. However, the turnover of these perilesional activated endothelial cells is low, and endothelial progenitor cells are needed to facilitate endothelial repair. The circulating endothelial progenitor cells are bone marrow-derived cells, which can divide into somatic valvular endothelial cells
[38]. Accumulation of senescent endothelial cells diminishes the regeneration of endothelial disruption, as these cells cannot divide. Furthermore, aging affects the regenerative capacity of endothelial progenitor cells as the release of these cells from the bone marrow decreases as age increases
[38]. Moreover, the senescence of endothelial progenitor cells increases as age increases. Matsumoto and coworkers showed enhanced apoptosis as well as increased senescence of circulating endothelial progenitor cells in patients with aortic stenosis, leading to a reduced circulating number of these cells. It is proposed that besides aging, cardiovascular risk factors, such as hypertension, diabetes, hyperlipidemia and smoking can influence the number of circulating endothelial progenitor cells and the regenerative capacity of the cardiovascular system, including the vascular system and the valves
[39][40]. Molecular markers of cellular senescence, such as beta-galactosidase and cell cycle arrest inductor P16INK4A (inhibitor of cyclin D-dependent kinases), have shown a correlation with tissue remodeling severity and degenerative changes in the aortic valve
[37].
In conclusion, senescent cells contribute to aging and aortic valve degeneration not only as a result of cell cycle exit and resistance to apoptosis, but also by secretion of pro-inflammatory cytokines, chemokines, matrix metalloproteinases and growth factors, promoting senescence in surrounding cells as a bystander effect
[41][42]. Eliminating senescent cells might represent a future therapeutic strategy. However, human applications of these potential therapies are still limited by the sparse knowledge of the basic molecular cell biology of senescence
[43].
3.2. The Two Phases of Aortic Valve Calcification: Initiation and Progression Phase
Previously, it was thought that aortic valve sclerosis was a passive process as a consequence of aging. It has already been revealed that aortic valve sclerosis is an active process with some similarities and differences compared to vascular atherosclerosis. Aortic valve sclerosis and stenosis are different stages of aortic valve calcification.
The early phase of aortic valve degeneration usually begins at the aortic side with the dysfunction of the endothelial barrier, allowing lipids from the blood to enter into the subendothelial space
[44][45][46]. Valvular endothelium disruption may occur as a result of several aging and nonaging factors, usually as a multifactorial and complex process. Age-related (replicative) and stress-induced cellular senescence processes are discussed above, and further proinflammatory and profibrotic processes are involved in the initiation phase
[8][9][10][11][12][13]. Even under physiological circumstances, the mechanical stress pattern caused by the blood flow over the years might initiate aortic valve sclerosis, affecting mainly the aortic side of the valve, usually beginning at the base of the leaflet
[47]. On the aortic side of the normal valve, within the sinuses of Valsalva, valvular endothelial cells are exposed to “sclerosis prone” oscillatory low shear stress in systole and turbulent flow vortices in diastole. Meanwhile, cells on the ventricular side of the normal aortic valve experience “less sclerosis prone” linear high-shear stress of systolic forward laminar flow
[9][48]. The laminar flow on the ventricular side becomes turbulent in the case of AS. Cheng C. and coworkers worked out a shear-stress carotid artery mouse model to examine plaque formation under low, high and low-oscillatory shear stresses
[49]. Atherosclerotic lesions evolved in the regions of low shear stress in this mouse model
[49]. In this region, the expression of proatherogenic inflammatory mediators and matrix metalloproteinase activity was higher
[49]. Expression of vascular cell adhesion molecule-1 (VCAM-1), intercellular adhesion molecule-1, vascular endothelial growth factor, C-reactive protein and proinflammatory cytokine interleukin 6 (IL-6) was upregulated in the lowered shear stress region
[49]. It is worth noting that the physiologic shear stress pattern of laminar flow over time might not result in significant AS, as a significant proportion of the elderly do not suffer from significant AS
[9]. Additional genetic and acquired factors may have additive effects on the development of significant AS
[50]. Furthermore, the congenital bicuspid aortic valve morphology represents an altered mechanical stress, as the systolic oscillatory shear stress and the ascending aorta wall shear stress are altered compared to the tricuspid aortic valve. Subsequently, the onset of aortic valve degeneration is earlier in the case of bicuspid aortic valve, the progression of the disease is more rapid and, in many cases accompanied by the dilatation of the ascending aorta
[51]. Further nonaging factors, such as radiation, have been revealed to initiate valvular endothelial injury and valvular inflammation, leading to aortic valve disease
[52].
Endothelial dysfunction and disruption in the early phase lead to lipid deposition from the blood into the subendothelial space. Moreover, lipid deposition and endothelial barrier dysfunction initiate inflammatory cells and cytokines to enter into the valvular interstitial space
[9][10][11]. These inflammatory cytokines (tumor necrosis factor-α, interleukin-6 and interleukin-8) promote endothelial to mesenchymal transformation, resulting in a new myofibroblastic cell phenotype migrating into the interstitial space
[53][54][55][56]. The endothelial to mesenchymal transition was first described in 2001 by Paranya and coworkers, however, there is still debate regarding the role of the activated myofibroblast-like VEC in the extracellular matrix regulation of the aortic valve
[10][53][54]. The membrane of a VEC encompasses the endothelial nitric oxide synthase (eNOS) to produce nitric oxide (NO), which inhibits fibrosis and calcification
[57][58]. Valvular endothelial NO increases the expression of neurogenic locus notch homolog protein 1 (NOTCH1) signaling in VIC, which inhibits regulators of osteoblast cell fate (e.g., runt-related transcription factor 2, RUNX2) and increases the expression of anticalcification factors, such as SRY-Box transcription factor 9 (SOX9) and bone morphogenic protein 2 (BMP2)
[59]. This NOTCH1-RUNX2-SOX9-BMP2 signal route is short-defined as the NOTCH1 pathway. Aging is associated with increased oxidative stress, decreased extracellular superoxide dismutase activity and decreased eNOS activity leading to a diminished NO bioavailability and endothelial dysfunction
[57][58]. Valvular endothelial cell dysfunction and the subsequent reduced NO production promote a fibrotic process within the valve. Furthermore, the upregulated renin-angiotensin system (RAS) is also critical in aortic valve disease
[11]. Valvular endothelial cells under increased shear stress promotes TGFβ1 to activate the quiniscent VIC set in the fibrosa, spongiosa and ventricularis layers
[60]. The activation of these cells may result in myofibroblastic differentiation characterized by the expression of alpha-smooth muscle actin (αSMA). The myofibroblastic VICs secrete structural matrix proteins and matrix metalloproteinases leading to extracellular matrix remodeling, leaflet thickening and increased leaflet stiffness
[61]. Microcalcification, also defined as dystrophic calcification, in the early phase, is a result of myofibroblastic VIC death and the release of apoptotic bodies in the area of lipid deposition and inflammation
[11]. Briefly, under physiological shear stress circumstances, VEC protects VIC from myofibroblastic differentiation by reducing αSMA expression, and calcification by producing NO and increasing the expression of NOTCH signaling target genes in these cells
[59][62]. Moreover, VIC can suppress endothelial-to-mesenchymal transformation and osteogenic differentiation of VEC, emphasizing the importance of VEC and VIC interactions in valve homeostasis
[63].
In the progression phase of the disease, there is a constant remodeling of the extracellular matrix and ongoing calcification, leading to impaired leaflet opening and closing over the cardiac cycle. Previous in vitro studies revealed osteoblastic differentiation of quiniscent VIC by exposure to BMP2, RUNX2 and osteopontin
[64]. Moreover, Schloter and coworkers showed that myofibrotic differentiation may precede osteoblastic differentiation of VIC. This was demonstrated with transcriptomics, showing that VIC isolated from fibrotic areas of valves exhibited intermediate gene profiles between nondiseased and calcific regions
[65]. Osteogenic differentiation factors implicated in osteogenic cell differentiation are upregulated, including the NOTCH pathway, receptor activator of nuclear kappa B (RANK)-RANK ligand (RANKL) and osteoprotegerin (OPG) pathway
[11]. Dayawansa and coworkers proposed that progression of aortic valve stenosis promotes further mechanical stress alterations, endorsing a positive feedback loop mechanism and a vicious cycle of chronic inflammation and calcification
[9]. Once calcification develops in the valve, a constant vicious circle of calcification and valve injury is maintained. Calcific deposits in the leaflets lead to increased mechanical stress and injury-induced activation of further osteoblast differentiation
[11].
Overall, the early phase of aortic valve degeneration is dominated by inflammation, subendothelial lipoprotein oxidation, fibrosis and microcalcification, while the later phase is dominated by self-perpetuating progressive calcification. Valvular interstitial cell activation with myofibroblast differentiation is present mainly in the early phase, while the osteoblast differentiation is mainly in the later stage. Grim and coworkers showed that inflammatory macrophages may initiate a myofibroblast-to-osteogenic intermediate VIC phenotype, which may mediate the switch from fibrosis to calcification during AS progression
[66]. Nevertheless, the early phase is dominated by interleukins secreted by macrophages, whereas the later phase is dominated by the NOTCH and RANK/RANKL/OPG pathways
[11][55][56]. Focal discrete aortic valve calcification is defined as aortic sclerosis. Approximately 10–15% of patients with aortic valve sclerosis will progress to obstructive calcification in their lifetime with mild, moderate or severe stenosis
[67]. Furthermore, AS is not only a valvular disease, as the increased afterload induces adaptive hypertrophic left ventricular remodeling and may lead to heart failure. Most patients enter the healthcare system because of heart failure symptoms, hence the valve disease is diagnosed in this late phase
[68].