Calcific Aortopathy in Response to Aging and Injury

Authors: Dwight A. Towler, Cecilia M. Giachelli, Marta Scatena, Yin Tintut, Linda L. Demer

Abstract

The arterial vasculature is the second most frequently calcified structure in the human body after the skeleton. Calcification of the aorta and aortic valves occurs in most individuals in westernized societies with advancing age, with abdominal aortic calcification generally preceding ascending thoracic aortic disease. In cardiac valves and the thoracic aorta, however, calcification often arises earlier in common disease contexts characterized by metabolic, mechanical, or inflammatory injury (eg, metabolic syndrome, chronic kidney disease, irradiation). In these settings, calcification frequently involves the arterial media as a histoanatomic feature, and is associated with accelerated neurocognitive decline and increased cardiovascular mortality, reflecting a form of precocious aging. The term arteriosclerosis was coined nearly 2 centuries ago to describe the calcium-mediated hardening of the aorta and conduit arteries observed at autopsy with aging. However, much of our understanding of the causes, characterization, and consequences of aortic calcium deposition has emerged only within the past decade. Features of disease biology, including engagement of innate immunity, senescence (inflammaging), and ectopic activation of osteogenic mechanisms, are consistently revealed. In this article, we briefly review the burgeoning literature, highlighting recent advances in clinical and discovery science with translational implications. Given the current trajectory, after 2 centuries of disease recognition, the next decade of innovation promises meaningful progress toward effective medical treatments to prevent and treat the clinical consequences of calcific aortopathy.

Aortic calcification is prevalent, observed in 50% to 80% of individuals >40 years of age in westernized societies. Remarkable histoanatomic predilection is exhibited, with intimal atherosclerotic calcification primarily noted and markedly enriched in the abdominal aorta and iliofemoral segments with advanced age and inflammation. In aging human populations, abdominal aortic calcification precedes ascending thoracic aortic calcification, attributable perhaps to their distinct developmental origins in the neuroectoderm and mesoderm, and their different flow and oscillatory shear stress patterns. In cardiac valves and the thoracic aorta, calcification arises in many disease contexts, frequently involves the medial layer, and conveys considerable neurocognitive and mortality risk. The resulting aortic stiffness adds risk for cognitive decline even absent overt microemboli. Calcific aortic valve disease (CAVD), primarily a disease of advanced age, is also strongly associated with hypertension, bicuspid aortic valve, and other disorders associated with aortic calcification. Lobstein first coined the term “arteriosclerosis” in 1829 to denote hardening of aortic and arterial tissues due to calcification, along with the term “osteoporosis” to indicate the reciprocal loss of skeletal mineralization with advanced age. However, of the ≈2500 articles on aortic calcification or CAVD archived in PubMed, >2/3 have been published within the past decade. We briefly review this literature, citing primarily articles published since 2022, and point to some of the many enlightening advances in clinical and discovery science addressing calcific aortopathy.

Histoanatomic Classification

As originally described in Lobstein’s autopsies, Virchow’s histopathology, and Mönckeberg’s radiographs, aortic calcification exhibits considerable histoanatomic variation. It is easily identified in tissue sections by classic histochemistry (Figure 1). The anatomic distribution is relevant not only to disease biology and consequences but also to quantitative interpretation of noninvasive clinical imaging. In arteries, intimal calcification has long been recognized as an independent risk factor for cardiovascular disease morbidity and mortality, being positively correlated with atherosclerotic plaque burden and major adverse cardiovascular events. By comparison, medial calcification increases vascular stiffness, causing increased pulse wave velocity and pulse pressure, thereby contributing to hypertension, myocardial ischemia, left ventricular hypertrophy, and heart failure (discussed in the following). In the aortic valve, calcification is associated with additional clinical settings and consequences (Table 1). Patients with chronic kidney disease (CKD), especially end-stage disease, or type 2 diabetes have particularly high burdens of medial calcification, with “pipe-like” arterial stiffening giving rise to noncompressible peripheral arteries on ankle-brachial indices, mediating peripheral arterial disease with its considerable morbidity, including claudication and amputation.

Imaging

On ordinary radiography, the earliest site of aortic calcification is abdominal, preceding both thoracic and coronary calcification, and this finding is reported to predict cardiovascular events as accurately as coronary artery calcification.^1^ Computed tomography (CT) scans, based on X-ray attenuation, correspond with calcium content and are widely used for quantifying aortic calcification. Fully automated analysis of clinical abdominal CT scans and abdominal plain films^2^ has been developed using machine learning approaches such as radiomics and deep-learning segmentation, which may differentiate characteristics of calcium deposits.

^18^F-NaF positron emission tomography (PET) has been used for decades to detect bone metastases; due to pioneering work of investigators at the University of Edinburgh,^3^ it is now used to label the bone mineral in cardiovascular calcification. Fluoride ions bind covalently to the surface of hydroxyapatite, replacing hydroxyl groups to form fluorapatite. The PET signal intensity is proportional to the mineral surface area, unlike the CT X-ray attenuation signal, which instead corresponds to mineral mass or content. Aortic and aortic valve calcification detected by ^18^F-NaF PET portend negative cardiovascular outcomes, outperforming the revised Framingham Risk Score in predicting ischemic stroke.^3^ In patients who received ^18^F-NaF PET scans for assessment of bone metastases, aortic tracer uptake was associated with major adverse cardiovascular events.^4^ Uptake in the outer wall has been associated with aortic disease or death.^5^

Because surface area is greater in clusters of small deposits than in single large deposits of the same volume, CT and PET images require distinct interpretations. Mismatch—where PET results are positive and CT results are negative—may represent clusters with density below the CT threshold. Because such clusters coalesce with progression and become detectable on CT, they have been viewed as identifying future sites of calcification. Mismatch may also be caused by the resolution difference between CT and PET, smoothing algorithms, spillover, or partial-volume effects.

Hemodynamic and Clinical Consequences

Biomechanics and End-Organ Damage

Aortic stiffness, resulting from calcification,^6^ fibrosis, crosslinking, and hypertrophy, contributes to cardiac and renal dysfunction as well as cognitive decline independently of hypertension. At the tissue level, stiffness also promotes cellular senescence^7^ and osteogenic differentiation,^8^ creating a positive feedback loop between stiffness and calcification. Cardiac dysfunction results from loss of ventriculo-aortic coupling (the Windkessel effect). Normally, with this effect, the proximal aorta distends and recoils, reducing afterload and cardiac work during systole while perfusing the coronaries and systemic vasculature during diastole. Stiffening leads to increased cardiac work and pulse pressure, causing hypertrophy, coronary insufficiency, heart failure with preserved ejection fraction,^9,10^ and subclinical damage to low-resistance end organs^11^ (Figure 2).

Chung et al^13^ were the first to show that aortic valve calcification portends future cognitive impairment. In support of this, subsequent studies showed that aortic calcification is associated with dementia in women^14^ and aortic stiffness is associated with evidence of barotrauma.^15^ Aortic stiffness also tracks renal dysfunction, which impairs phosphate excretion, calcium metabolism, and endocrine regulation by PTH (parathyroid hormone), vitamin D, and FGF23 (fibroblast growth factor 23), creating a “perfect storm” for aortic calcification and a self-reinforcing cycle of cardiovascular disease (Figure 3).

Understanding how calcification affects the mechanical integrity of the aorta is central to understanding aortic aneurysm rupture and aortic dissection, in which rupture occurs through delamination of layers of fibers. The fundamental biomechanical principles are independent of size. In general, rupture stress increases with the intensity of the applied pulsatile forces and the intensity of rupture (von Mises) stresses, and decreases with the strength of the anisotropic, viscoelastic material properties of the surrounding extracellular matrix (ECM) and the bonding properties at the interface with the calcium deposit. Rupture stress depends on the geometry of the deposits and their orientation relative to applied force, and is amplified by proximity to a lumen, lipid pool, and other calcium deposits. In general, larger deposits introduce greater and more widespread rupture stresses, but, when smaller deposits arise in thin fibrous caps, their proximity to both the lumen and the lipid pool generally amplifies the risk. Regardless of size, a given deposit generally has opposing effects at its different surfaces; it may increase rupture stress at the edges facing the axis of applied extensile force while decreasing rupture stress at edges perpendicular to that axis. Thus, both microcalcifications and macrocalcifications increase rupture risk in some parts of the surrounding matrix and decrease risk in other parts.^16^ Some investigators have proposed specific values for deposit size and force thresholds in the context of individual lesion morphology, but these are limited by nongeneralizability and the unknown values of key measures, such as the bond strength and the anisotropic material properties of the surrounding tissue components, which vary among plaques, among patients, by location, and over time in the pulsatile cycle as well as with plaque growth or regression.

Calcific Aortic Valve Disease

CAVD is the most common cardiac valve disease in the Western world, with profound hemodynamic consequences, increasing myocardial afterload and preload dependence. Hypertension, metabolic syndrome, CKD, and hyperphosphatemia are major risk factors. Prognosis is poor,^17^ with the only treatment being valve replacement, and calcification often destroying bioprosthetic valves. CAVD impairs flexibility, motion, and hemodynamics while increasing shear stress.^18^ In mice, calcific deposits arise almost solely in the sinuses of Valsalva. In humans, they also arise on the aortic, but not the ventricular, face of leaflets. Although nodular fibrocalcification of the leaflets is considered central, the valve leaflet hinge and sinus wall are affected first, and loss of their compliance increases transvalvular resistance. In mice, the leaflets show thickening and little to no calcification, but the robust hinge and sinus calcification contribute to valve dysfunction even without leaflet calcification.^19^ Several in vitro and in vivo models of valvular calcification have been established (Tables 2 and 3).

Central Unifying Features: Inflammaging and Innate Immunity, Osteodifferentiation, Matrix, Cells, and Genes

Soft Tissue Calcification, Inflammaging, and Innate Immunity

As a unifying concept, soft tissue calcification may be a generalized response to multiple conditions that include infectious and noxious (foreign body) processes. At those sites, mineralization may have evolved to serve as a means to control spread of pathogenic processes by walling them off, entrained to proinflammatory cues elicited by pathogen-associated or tissue damage–associated molecular patterns. Oxylipid-rich lipoprotein particles and advanced glycation end-products accumulating in tissues with chronic metabolic disease, other noninfectious inflammatory or injury states, or advanced age may resemble the antigens produced by leukocyte oxidation of lipid components of bacteria, parasites, and foreign bodies, resulting in triggering of the final common pathway of innate immune responses. The mechanisms sensing and triggering cellular damage–associated molecular programs also organize osteogenic calcification responses in the aorta.^20^

Osteochondrogenic Differentiation and the Pivotal Transcription Factor RUNX2

In all the previously mentioned conditions, osteochondrogenic gene regulatory programs (Figure 4), first delineated in bone, participate in the robust progression of matrix mineralization. A wide range of oxidative/inflammatory, mechanical, metabolic, and endocrine stresses trigger artery wall cells, such as vascular smooth muscle cells (VSMCs), endothelial cells, pericytes, adventitial myofibroblasts, and possibly mesenchymal stem cells, to undergo differentiation to an osteochondrogenic phenotype. This may occur through dedifferentiation, redifferentiation, or transdifferentiation. As in bone, high biomechanical stress in injured lung or aorta play important instructive roles for osteogenic deposition. In the following sections, we discuss pathogenesis, emphasizing the contributions of osteogenic mechanisms that arise ectopically in response to inflammatory cues that accrue with aging, dysmetabolic states, or injury to drive aortic mineralization.

RUNX2 (runt-related transcription factor 2; also known as PEBP2αA [polyomavirus enhancer binding protein 2αA] or CBFA1 [core-binding factor A1]) has been shown by multiple laboratories to be central to vertebrate skeleton mineralization through osteochondrogenic gene programs. As in the skeleton, osteochondrogenic features of aortic mineralization also depend on RUNX2 actions in aortic VSMCs. RUNX2 is prodigiously regulated by isoform variation through alternative promoter utilization, posttranslational modifications, subnuclear localization, and protein–protein interactions to control osteogenic gene expression. The mechanisms whereby inflammation activates RUNX2 in VSMCs have previously focused on upstream osteogenic morphogens of the BMP and WNT gene families, elicited in response to TNF (tumor necrosis factor), IL-1 (interleukin-1), and other inflammatory signals. In a recent CAVD study, RUNX2 was reported to be more directly recruited by inflammatory cues to promote osteogenic mineralization in aortic valves through STAT5 (signal transducer and activator of transcription 5), induced by the prototypic inflammatory transcription factor NF-κB (nuclear factor-κB).^21^ TERT (telomerase reverse transcriptase)—a key regulator of aging phenotypes—was shown to form a noncanonical complex with STAT5 to drive RUNX2 expression.^21^

ECM Metabolism in Osteogenic Programming

In vitro, ECM and its mechanical properties are known to regulate vascular and valvular cell calcification. In vivo, analysis of atherosclerotic lesions by small angle X-ray scattering shows that 3-dimensional collagen fibril orientation varies with proximity to calcium deposits. Orientation near deposits is mainly longitudinal or radial, whereas that far from deposits is circumferential. Because fibril orientation affects anisotropic mechanical strength of the matrix, such features may predict lesion vulnerability to rupture.^22^ In the earliest stages of aortic calcification, mineral is often seen lining the elastin fibers. Elastin appears to have greater proclivity for calcification than collagen, based on CaCl2 treatment of tissue specimens made predominantly elastin-rich or collagen-rich by collagenase or elastase treatment, respectively.^23^ Crosslinking among matrix proteins also affects aortic calcification. Conditional inactivation of lysyl oxidase, a crosslinker, was recently found to reduce aortic calcification in mice.^24^ Similarly, in vitro, VSMCs deficient in lysyl oxidase had increased expression of contractile markers and decreased expression of the osteogenic protein RUNX2.^24^ In addition, enzymatic matrix degradation may contribute to aortic calcification. MMP-3 (matrix metalloproteinase-3) levels are increased in aortic calcification in mice and in phosphate-treated VSMC cultures, and mice deficient in VSMC-specific MMP-3 had reduced aortic medial calcification,^25^ an important finding given the role of matrix strength in delamination in aortic dissection.

Cellular Origins and Lineage Decisions

Previous genetic cell lineage tracing and reporter gene studies identified 3 major cell types undergoing osteochondrogenic phenotypic change in cardiovascular VSMCs, endothelial cells undergoing endothelial-to-mesenchymal transition, and activated valvular interstitial cells (VICs). Microvascular pericytes also mineralize in vitro and differentiate into osteoblastic cells in embryonic endochondral bone formation, and adventitial fibroblasts may contribute indirectly to medial calcification through release of factors found in conditioned media to induce VSMC calcification.^26^ Several in vitro and in vivo models of vascular calcification have been established (Tables 2 and 3). In response to inflammatory lipids and oxidative stress, VSMCs recapitulate key features of osteoblast-mediated embryonic bone formation, following a choreographed sequence of osteogenic gene expression^27^ involving BMP-2/4, RUNX2, ALP, AKP2, TNAP, SPP1, BGLAP, and COL1A1, ultimately producing mineralized tissue. Cardiac VICs also undergo osteochondrogenic differentiation in a recapitulation of embryonic endochondral bone formation. In the valve, however, the sequence involves myofibroblastic marker expression followed by chondrogenic differentiation, which involves SOX9 (SRY-box transcription factor 9) and production of collagen types II and X, eventually forming bone or cartilage.^28^

Single-cell and spatial transcriptomic and proteomic analyses have recently confirmed and advanced understanding of osteochondrogenic lineage decisions in aortic calcification. Single-molecule spatial transcriptomics identified a VSMC subtype enriched in the cocaine- and amphetamine-regulated transcript prepropeptide, colocalizing with calcification in thoracic aortic aneurysms, and the prepropeptide induced osteochondrogenesis in human VSMCs in vitro.^29^ Unbiased single-cell sequencing and pseudotime trajectory have shown a heterogeneous mix of VICs, endothelial cell subtypes, and mesenchymal transition as contributors to CAVD. A multiomics approach identified a disease-driver phenotype of VICs (CD44^high^CD29^+^CD59^+^CD73^+^CD45^low^) with the potential for osteogenic differentiation in vitro as a key regulator of human CAVD,^30^ resembling many features of a previously described circulating osteoprogenitor.^31^ A unique population of stromal cells was found in calcified human bicuspid aortic valves by single-cell RNA sequencing, arising from macrophage-to-mesenchymal transition of macrophages positive for the mannose receptor CD206.^32^

Genetic Origins

Three autosomal recessive disorders lead to aortic arterial calcification due to deficiency of CD73, pseudoxanthoma elasticum, and generalized arterial calcification of infancy. The first, also known as hereditary arterial and articular multiple calcification syndrome, is a rare, adult-onset condition caused by NT5E gene variations. A key manifestation is extensive calcification of the subrenal aorta and its branches in the lower extremities. The latter 2 conditions are caused by sequence variations in ABCC6 or ENPP1. They were previously believed to have distinct phenotypes; however, recent studies indicate considerable overlap, suggesting that they reflect 2 ends of a clinical spectrum. Both result from insufficiency of the potent calcification inhibitor inorganic pyrophosphate, which is generated when ABCC6 (ATP-binding cassette subfamily C member 6), a transmembrane transporter, drives cellular efflux of ATP (adenosine triphosphate), which is hydrolyzed by ENPP1 (ectonucleotide pyrophosphatase/phosphodiesterase 1) to produce inorganic pyrophosphate. One distinction between the conditions is that inhibition of ALP (alkaline phosphatase) reduces arterial calcification in Abcc6-null but not Enpp1-null mice, pointing to differences in pathogenic signaling.^33^ Inorganic pyrophosphate levels have little correlation with the severity of calcification, and, although ABCC6 interacts with lipids, lipid profiles in patients with pseudoxanthoma elasticum are minimally different from those of controls.^34^ Evidence suggests that immune modulation through the adenosine receptor may be a key contributor to the biology of extracellular ATP/inorganic pyrophosphate metabolism relevant to arterial mineralization.^20^

Nutritional Contributions: Overnutrution, Microbiome, and Lipid Metabolism as Inflammatory CueS

Animal models of diet-induced metabolic syndrome, with or without clear manifestations of type 2 diabetes, first established the impact of overnutrition and obesity on aortic calcification risk. Years ago, periaortic and abdominal fat were shown to be risk factors for abdominal aortic calcification in the Framingham Heart Study. Subsequently, several epidemiologic studies in humans, including MESA (Multi-Ethnic Study of Atherosclerosis), confirmed the relationships between overnutrition-induced metabolic syndrome and aortic calcification in both aortic valve and abdominal aorta. The milieu of hyperglycemia, insulin resistance, hyperlipidemia, and oxidative stress generates lipid oxidation^35^ and glycation products^36,37^ that activate “inflammaging” and vascular osteogenic mineralization programs (Figure 4). Because obesogenic features of the metabolic syndrome are mediated by gut microbiota, several groups have discovered microbiota-derived phosphatidyl choline and other metabolites that correlate with aortic valve calcification. As direct evidence for a role of the microbiota in aortic calcification, antibiotic cocktails and vancomycin treatments in mice with vitamin D–induced aortic calcification exacerbated calcification by reducing intestinal Bacteroidetes and their production of acetate.^38^ Another gut organism, Prevotella copri, has been associated with CKD-associated vascular calcification in animals and patients,^39^ likely due to increased lipopolysaccharide levels and resulting inflammasome signals. These observations have clinically significant implications because, like weight gain,^40^ rapid weight loss in older individuals has recently been shown to increase risk for aortic calcification, particularly in women.^41^ Although conditionally independent of nutritional intake, microbiota-mediated changes in vasculopathy are understudied in this latter setting, and may contribute to the emerging bimodal response between weight status and aortic calcification risk with aging.^40^ Although effective in preclinical disease models,^42^ whether glucagon-like peptide-1 receptor agonists affect aortic calcification in humans in the context of weight loss remains to be examined in detail. The strong associations between intestinal bacterial profiles and lipoprotein risk profiles in clinical obesity may help explain the cardiovascular benefits achieved through weight wellness.

The strong connections between atherogenic lipid profiles and intimal calcification of the aorta are well appreciated. However, the lipoprotein with the strongest relationship to cardiovascular calcification, especially CAVD, is lipoprotein(a) [Lp(a)], a particle whose level is primarily entrained to genetic programs rather than dietary influence.^43^ This complex particle consists of a low-density lipoprotein particle (a lipid-rich domain with ApoB [apolipoprotein B]) attached by a disulfide bond to a tail made of the glycoprotein apolipoprotein(a) [Apo(a)], which has a variable-length string of kringle domains. This Apo(a) is encoded by the LPA gene; the gene name refers to Lp(a), rather than to LPA (lysophosphatidic acid), which appears later in this review. Proinflammatory and proatherogenic oxidized phospholipids bind to a specific kringle domain on Apo(a) and are also present in the lipid phase. Approximately 70% to 90% of Lp(a) levels are genetically influenced, and several single nucleotide polymorphisms in the LPA locus are significantly associated with Lp(a) levels. Whereas some of these single nucleotide polymorphisms are linked with a copy number variant, others are independently associated with both high and low Lp(a) levels. The assembly of Lp(a) from Apo(a) and ApoB occurs in the hepatocytes, but the pathways for Lp(a) removal from circulation are not completely understood. Larger Apo(a) isoforms are retained for longer periods in the endoplasmic reticulum, despite being folded at the same rate as smaller isoforms, and are subjected to increased degradation by the proteasome. This mechanism contributes to the general inverse correlation between the size of Apo(a) isoforms and plasma Lp(a) levels.

Large epidemiologic genetic studies have shown associations between Lp(a) levels and CAVD, and a recent meta-analysis of >40 studies^43^ showed that most support a strong relationship. The highest Lp(a) levels are associated with faster disease progression and worse outcomes.^44^ Because Lp(a) is not influenced by HMG-CoA (3-hydroxy-3-methylglutaryl coenzyme A) reductase, this may explain why several trials of lipid-lowering therapy with HMG-CoA reductase inhibitors (statins) targeting low-density lipoprotein cholesterol did not significantly reduce the progression or severity of CAVD.^45^ Population-based studies also identify Lp(a) as a central factor in aortic and peripheral artery calcification.^46^ Although more scarce in the circulation, Lp(a) binds strongly to matrix and to atherogenic oxidized phospholipids, leading to accumulation in the artery wall.

How Lp(a) promotes calcification is unclear. One factor under consideration is ATX (autotaxin), a secreted lysophospholipase D that binds to Lp(a) and is found in calcified human valve leaflets. ATX hydrolyzes extracellular lysophosphatidylcholine into LPA, which is a ligand for certain G protein–coupled receptors and may trigger osteogenic differentiation of VICs. In multipotent mesenchymal cells, LPAR1 (lysophosphatidic acid receptor 1) upregulates the expression and activity of RUNX2. A small molecule ATX inhibitor tested in mice, rabbits, and cultured primary VICs from patients with CAVD protected against aortic valve and VIC calcification and decreased the expression of genes for fibrosis and calcification in VICs.^47^ As such, targeting the LPAR1 signaling relay may prove useful in limiting the progression of CAVD once evidence of valve fibrosis has been identified by echocardiography.

Aortic and Valvular Calcification in CKD–mineral and Bone Disorder

In patients with CKD–mineral and bone disorder (MBD), renal insufficiency leads not only to vascular calcification and bone diseases, but also to decreased glomerular filtration rate, hyperphosphatemia with secondary elevations in FGF23 and PTH, elevated levels of uremic toxins, and increased oxidative stress, as well as low klotho and 1,25-dihydroxyD levels. Patients with CKD-MBD have an excessive burden of aortic and valve calcification and a dramatically heightened risk of cardiovascular disease morbidity and mortality. Aortic medial and valve calcification, rather than coronary atherosclerosis, are highly progressive in CKD-MBD and consistent with the finding that heart failure is a much more common cause of death than myocardial infarction in these patients.^48^

Nontraditional cardiovascular risk factors, including elevated inorganic phosphate (Pi) level, vitamin deficiency, and uremic toxins, confer elevated risk of vascular calcification in CKD-MBD.^49^ Numerous studies over the past 20 years have shown that elevated Pi level directly promotes VSMC osteochondrogenic differentiation and calcification in vitro and in vivo.^50^ In addition, elevated Pi level may contribute to vascular calcification through regulation of autophagy, oxidative stress,^51^ endoplasmic reticulum stress, and inflammatory signaling.^52^ Pi mediates its effects in part through differential signaling through 2 Pi transporters, solute carrier 20A1 (SLC20A1) and solute carrier 20A2 (SLC20A2), on vascular cells. SLC20A1 promotes osteochondrogenic differentiation and matrix mineralization of human VSMCs through transport-independent signaling in a pathway involving ERK1/2 (extracellular signal regulated kinase type 1/2) with phosphorylation and activation of RUNX2. In addition, high phosphate due to SLC20A1 overexpression induces VIC calcification and apoptosis. Consistent with its protection from basal ganglion vascular calcification, SLC20A2 inhibits VSMC calcification in part through a pathway involving induction of OPG (osteoprotegerin), the soluble receptor decoy antagonist for RANKL (receptor activator of nuclear factor-κB ligand).^53^ Homodimerization and heterodimerization of SLC20A1 and SLC20A2 may have a regulatory role in phosphate signaling.

In addition, vitamin D and its metabolites contribute to disease risk in CKD-MBD, in a biphasic response that reflects indirect benefits through reduction of the secondary hyperparathyroidism seen in CKD and direct toxicity of excessive vitamin D3, induced by hyperphosphatemia and hypercalcemia that drive vascular calcification. Likewise, vitamin K deficiency as well as warfarin use have also been linked to vascular calcification on the basis of their roles in limiting γ-carboxylation of key calcification inhibitors, such as MGP (matrix γ-carboxyglutamic acid). In addition, uremic toxins such as indoxyl sulfate have been linked to aortic calcification^54^ and their ties to inflammatory signaling are under investigation.

WNT Signaling, Hyperparathyroidism, and PTH Hyporesponsiveness: Modulators and Mediators of Aortic Inflammation

The skeleton responds to homeostatic morphogenetic, metabolic, mechanical, endocrine, and inflammatory cues that control ECM mineralization and remodeling, with members of the WNT gene family playing central roles. The paracrine actions of these osteochondrogenic polypeptides and their cognate receptors have emerged as important in the biology of aortic calcification. Whereas early studies emphasized β-catenin–mediated canonical Wnt programs, intracellular signaling through noncanonical relays has been identified as pivotal. In addition to promoting RUNX2-dependent transcription, noncanonical Wnt signals activate NFAT (nuclear factor of activated T cells), AP-1 (activator protein 1), and mechanosensitive planar cell polarity transcriptional relays central to the innate immunity^55^ that characterizes arterial calcification responses. Noncanonical Wnt5a, Wnt5b, and Wnt11 ligands have a role in VIC mineralization, in a manner recapitulating key features of noncanonical Wnt signaling in aortic calcification. One key function of the Wnt receptor LRP6 (low-density lipoprotein receptor-related protein 6) in VSMCs is to limit noncanonical Wnt programs that promote aortic calcification and vascular stiffness.

PTH is the prototypic osteotropic hormone necessary for minute-to-minute control of calcium–phosphate homeostasis in vertebrates. PTH secretion is acutely increased by hyperphosphatemia and hypocalcemia, and normally decreased by hypercalcemia, vitamin D receptor agonists, and FGF23. The acute vasodilatory actions of PTH have been known since its discovery; however, sustained elevated PTH levels drive hypertension and impair regulation of tissue perfusion.^56^ As in bone, vascular PTH actions differ with sustained versus pulsatile physiologic or pharmacologic exposure; this reflects not only vascular PTH/PTH1R (parathyroid hormone 1 receptor) desensitization and hyporesponsiveness with sustained exposure but also different ligand-induced PTH1R signaling bias, as recently reviewed.^56^ VSMC-autonomous actions of PTH1R reduce osteofibrotic phenotypic modulation. However, chronically elevated PTH levels, as in primary or secondary hyperparathyroidism, increase aortic stiffness and aortic valve calcification in humans and increase overall cardiovascular risk, with mortality tracking PTH levels rather than calcium levels even in absence of end-stage CKD.^56^ In primary hyperparathyroidism, this occurs in the absence of the hyperphosphatemia that causes secondary hyperparathyroidism of CKD.^56^ A biphasic, U-shaped relationship has consistently been observed between serum intact PTH concentration and all-cause or cardiovascular mortality in patients with end-stage CKD on hemodialysis^56^; the nadir of risk occurs at a prevailing PTH concentration above the normal range of patients without uremia, highlighting the PTH signaling resistance arising in CKD due to inhibitory PTH fragments and other uremic toxins that accumulate in the absence of renal filtration, paralleling changes in serum phosphate with CKD. Upacicalcet, a novel calcium-sensing receptor agonist, reduces secondary hyperparathyroidism and aortic calcification in uremic rats.^57^ Whether modulators of calcium-sensing receptors mitigate aortic or aortic valve calcification in primary hyperparathyroidism remains unknown, although parathyroidectomy reduces cardiovascular risk.^56^ Because of the U-shaped cardiovascular mortality response and confounding effects of PTH resistance with declining renal function, biomarkers of healthy vascular PTH1R signaling tone are needed.

The Matrix Vesicle: An Extracellular Modulator and Nidus for Crystal Formation

Matrix vesicles are nanosized, lipid-bilayer vesicles released from mineralizing cells that determine the temporospatial deposition of mineral in ECM. First discovered in bone >6 decades ago, they were found to carry calcium and phosphate and serve as a nidus for initiating hydroxyapatite crystal formation, often in conjunction with collagen and elastin fibers. In 1990, they were found unexpectedly in human aortic tissue. They are now known to carry a diverse cargo supporting osteodifferentiation and mineral formation, including phosphatases, annexin, Wnt ligands, nucleic acids, complement proteins, and cytokines. They are also now considered a subtype of extracellular vesicles (EVs), the term established in 2011 for a general mechanism in which bilayer vesicles guide paracrine communication and extracellular functions. Seminal work from Buffolo and colleagues^58^ has shown that EVs derived from both mesenchymal and myeloid lineages contribute to aortic mineralization. EVs isolated from calcified aortas of diabetic mice were found to carry IL-1β (interleukin-1β), and EV inhibition reduced expression of the procalcific factors BMP-2 (bone morphogenetic protein-2) and RUNX2.^59^ In diabetic mice, an inhibitor of EV release was found to reduce aortic calcification.^60^ Noncanonical Wnt ligands are abundant in EVs isolated from diseased human aortic valves and arteries,^61^ suggesting that targeting Wnt relays may prove useful in limiting aortic calcification, consistent with lessons learned from studying LRP6 molecular genetics in atherosclerotic humans and mice.

Inflammaging in Precocious Aging, IFN, and Genotoxic Stress

Precocious Aging

Human and murine molecular genetics have revealed the important role of compromised nuclear envelope integrity and structure in multiple aging phenotypes,^62^ including the precocious arteriosclerosis and osteoporosis phenotypes first noted by Lobstein. Hutchinson-Gilford progeria syndrome arises from sequence variations in the lamin A gene (LMNA), resulting in the accumulation of an unprocessable form of farnesylated lamin A that disrupts the nuclear envelope.^62^ In seminal work, Faleeva and colleagues^8^ identified that accumulation of progerin, a mutant lamin A protein produced by a sequence variation in LMNA, induces DNA damage with genotoxic stress responses that activate osteofibrogenic RUNX2 actions in arterial smooth muscle. The relevance of arterial calcification to the vascular aging phenotype in Hutchinson-Gilford progeria syndrome was demonstrated by combination pharmacotherapy that increased levels of key extracellular ATP metabolites (such as pyrophosphate and adenosine) to limit the RUNX2-regulated procalcific milieu; this combination therapy not only reduced arterial calcification but also prolonged the lifespan of Hutchinson-Gilford progeria syndrome mice. A key consequence of the matrix reprogramming in response to osteochondrogenic transcription factors such as RUNX2 and SOX9 is the elaboration of an ECM with mechanocrine properties favoring the mineralizing phenotype.^8^ Stiff extracellular matrices promote VSMC senescence by inducing mitochondrial and endoplasmic reticulum stress.^7^

IFN and Damage-Associated Molecular Patterns

Human and murine molecular genetic studies have converged to highlight a unifying feature of molecular pathogenesis relevant to inflammaging responses activating ectopic calcification in the aortic valve and ascending aorta. Earlier, several groups identified that type I IFN (interferon) antiviral signaling cascades, initiated by cellular damage with dysregulated extracellular and cytosolic DNA or RNA accumulation, promote proximal aortic calcification.^20^ Genetic underpinnings of Singleton-Merten syndrome, characterized by precocious aortic valve and ascending aorta calcification, pinpoint dominant sequence variations in cytosolic RNA-sensing, damage-associated pattern recognition receptors. These RLR (retinoic acid-inducible gene I-like receptor) proteins normally upregulate IFN production in response to minute levels of viral dsRNA structures. Pathogenic variants associated with Singleton-Merten syndrome activate mitochondrial antiviral signaling, leading to downstream activation of IRF3 (IFN regulatory factor 3) and RUNX2. However, damaged nuclear envelopes and damaged mitochondria elaborate higher levels of endogenous nucleic acid fragments that also activate these pathways, mimicking viral infections. TLR3 (Toll-like receptor 3) is a third component of the DNA/RNA pattern recognition receptor pathway that upregulates type I IFN expression and aortic calcification.^63^ Consistent with 2 large-scale human genotype–phenotype cohorts (Genetic Epidemiology Research on Aging and UK Biobank), loss of TLR3 function in mice protected against CAVD.^63^ Biglycan, a small proteoglycan best characterized in endochondral bone formation, was identified as a feed-forward TLR3 ligand in a signaling axis promoting CAVD.^63^ Although the cytosolic DNA-sensing adapter stimulator of interferon genes 1 is less well studied in this context, it promotes IFN-dependent VSMC senescence and calcification.^64^ In earlier work, Parra-Izquierdo and colleagues^65^ showed that small molecules that inhibit upstream RLR complexes or downstream Janus kinase components in the antiviral signal relay inhibit aortic VSMC or VIC calcification, respectively, and that sex-specific differences exist, with male VICs being more susceptible to procalcific type I IFN signaling. Whether these strategies may serve therapeutic purposes remains to be determined; however, a therapeutic window would need to be evident before they could be used to mitigate calcification without increasing risk of viral infection.

Genotoxic Stress of Radiation Therapy, Another Inflammaging Process

Remarkable advances in chemotherapeutics and checkpoint immune modulators have transformed oncology care; however, for many malignancies, external beam radiation therapy remains a key component of adjuvant or neoadjuvant strategies. Among radiation-treated cancer survivors, patients with breast cancer, in whom therapy is directed at the thoracic cage, are the most common. Ionizing radiation produces reactive oxygen species that damage DNA; it also triggers damage-associated molecular pattern recognition and immune activation, leading to inflammatory cytokine release and inflammation. Inflammation is also induced at the tissue level by vascular damage and cellular necrosis. Radiation activates type I IFN–dependent antitumor immunity and RLR signaling. Human sensitivity to this genotoxic stress increases with aging. Thus, it is not surprising that external beam radiation therapy brings markedly increased risk for aortic valve and ascending aortic calcification, affecting >33% of patients following a latency period of 1 to 2 decades after treatment.^66^ Because radiation increases osteochondrogenic differentiation rapidly in cultured cells, it is not clear why such a long latency precedes clinically significant aortopathy.^66^ Treatment with angiotensin receptor blockade is associated with reduced risk for CAVD progression in older patients, raising the possibility that similar strategies may mitigate radiation-induced aortopathy. Given that ≈1 in 8 women develop breast cancer, and >50% of them receive radiation therapy, development of a preclinical model for radiation-induced aortic calcification is essential.

Potential Therapeutic Strategies for Aortic Calcification

Antiresorptive and Resorptive Therapies

Bisphosphonates, a class of antiresorptive pharmaceutical agents, are nonhydrolyzable analogs of pyrophosphate that bind to mineral surfaces. Etidronate, a first generation bisphosphonate, is used empirically for generalized arterial calcification of infancy,^67^ with anecdotal reports of potential benefit. A small pilot study of long-term etidronate therapy in patients with arterial calcification due to deficiency of CD73 showed a possible slowing of lower extremity vascular calcification.^68^ However, newer generation aminobisphosphonates and other antiresorptives were clinically ineffective for aortic and valve calcification in adults in a randomized controlled trial^69^ and a meta-analysis.^70^

Harnessing mineral-resorptive osteoclastic cells represents an exciting possibility. These cells are uncommon in vascular calcification, possibly due to low serum levels of the osteoclast differentiation factor RANKL and high levels of its soluble inhibitor OPG in vascular and valvular disease. They may be inhibited functionally by increased IL-18 (interleukin-18) released from VSMCs undergoing osteogenic differentiation. Cellular treatment with osteoclasts has reversed vascular cell calcification in vitro, but not in vivo, in part due to their dispersion away from sites of calcification. Promising new developments include that human induced pluripotent cells can be differentiated into osteoclasts^71^ and that engineered osteoclasts as well as elastin-targeted nanoparticles containing calcium sequestrant regress both heterotopic ossification^72^ and aortic calcification,^73^ respectively, in mice. The nanoparticles containing the calcium sequestrant ethylenediaminetetraacetic acid are coated with elastin antibody. Because early calcification often appears on fragmented elastin, the presumed mechanism is reduced calcium availability for mineralization along the fibers.

Interventional Therapy

Lithotripsy, a standard treatment for renal calculi and cholelithiasis using ultrasonic shock waves to disrupt mineral, is now clinically administered by catheter to disrupt calcified coronary and peripheral artery lesions. It has also been applied noninvasively for calcified aortic valves in preparation for insertion of stents or valve replacements.^74^

Lp(a) Therapeutics for Aortic and Aortic Valve Calcification

Whether Lp(a) reduction affects aortic calcification and the amount of lowering required remain unknown. Consistent with inefficacy of statins in CAVD, statins also fail to lower Lp(a) levels, and may increase them. Other agents that hold promise for reducing Lp(a) levels include PCSK9 (proprotein convertase subtilisin/kexin 9) inhibitors,^75^ antisense oligonucleotides, small molecule inhibitors, and siRNAs targeting PCSK9^76^ or Apo(a) biosynthesis.

CKD-MBD Therapeutics

Despite accumulating evidence linking nontraditional risk factors with aortic calcification and cardiovascular disease in CKD, clinical studies targeting these factors to date are equivocal, generally due to inadequate statistical power and study design features relevant to the heterogeneity of treatment effect. A systematic review of 77 randomized and prospective, nonrandomized clinical trials concluded that magnesium, sodium thiosulfate, and non–calcium-based phosphate binders probably reduce vascular calcification progression, whereas results were inconclusive for antiresorptive therapies, calcimimetics, SNF472 (myo-inositol hexaphosphate), sotatercept, and oral-activated charcoal.^77^ Based on clinical evidence, factors considered unlikely to reduce calcification progression include vitamin D, vitamin K2, statins, nicotinamide, and spironolactone. Etidronate was the only bisphosphonate consistently associated with reduced vascular calcification progression.^77^ Limitations included inability of imaging modalities to distinguish intimal from medial calcification and lack of placebo-controlled studies.^77^ Several placebo-controlled clinical studies targeting risk factors for CKD-MBD have been undertaken. The IMPROVE-CKD trial (Impact of Phosphate Reduction on Vascular End Points in Chronic Kidney Disease) showed no effect of the phosphate-lowering drug lanthanum carbonate on extensive baseline abdominal aortic calcification. Interestingly, baseline data from that study showed that 25-hydroxyvitamin D levels were not associated with calcification severity in patients with CKD-MBD with extensive calcification.^78^ In the K4Kidneys randomized controlled trial (Vitamin K Therapy to Improve Vascular Health in Patients With Chronic Kidney Disease), patients with stage 3b or 4 CKD received either 400 μg of vitamin K2 or placebo for 1 year. Vitamin K2 treatment showed no benefit in vascular stiffness or aortic calcification over placebo.^79^

Animal studies have also been used to test potential therapies for CKD-MBD. In a Cy/+ rat model, the authors showed prevention or slowing of aortic calcification with a nicotinamide adenine dinucleotide phosphate oxidase inhibitor that reduces reactive oxygen species; ferric citrate and intravenous iron to reduce phosphate; and inulin, which reduces microbiota-associated uremic toxins.^80^ In a rat model, the phosphate binder lanthanum hydroxide significantly reduced aortic calcification and inhibited NF-κB and RANKL/OPG signaling pathways.^81^ Consistent with earlier studies, BMP-7 (bone morphogenetic protein-7) reduced hyperphosphatemia, hyperparathyroidism, and vascular calcification in a rat model.^82^ Antihypertensives reduced blood pressure but failed to block aortic medial calcification in a rat model of CKD-MBD, confirming that calcification can proceed independent of blood pressure modulation.^83^ Another study using the same model showed no effect of the calcimimetic R568 on CAVD.^84^

For patients with CKD on dialysis, elevated circulating ALP level is an independent risk factor for cardiovascular mortality. ALP is ectopically expressed in VSMC in response to pathogenic stimuli and promotes calcific aortopathy. A small molecule inhibitor of ALP, SBI-425, is known to prevent medial calcification and improve survival in a murine model of CKD-MBD. However, this inhibitor did not significantly inhibit bone mineralization, indicating that a therapeutic window exists for selective intervention.

Glucagon-like peptide-1 receptor agonists have shown promise in reducing renal and cardiovascular disease risk in patients both with and without diabetes.^85^ Whether these findings are related to effects on vascular calcification is not yet known, but recent studies suggest an inhibitory effect on vascular calcification in vitro and in vivo.^86^

Senolytic Soluble Guanylate Cyclase Activation and MicroRNA Modulators of Aging

Recognizing that cellular senescence accompanies aging, Roos et al^87^ found that 2 members of a new class of senolytic drugs—dasatinib and quercetin—significantly reduced aortic calcification and osteogenic signaling in mice without lowering lipids. Activity of epigenetic enzymes (eg, acetyltransferases and deactylases, such as SIRT1 [sirtuin 1]), which are involved in longevity pathways, also regulates the phenotypic transition of VSMCs from a contractile to a calcifying state.^88^ In another seminal and insightful study— the phase II, randomized, double-blinded, placebo-controlled efficacy trial (A Study Evaluating the Effects of Ataciguat [HMR1766] on Aortic Valve Calcification; Unique NCT02481258)—Zhang and colleagues^89^ recently demonstrated that the soluble guanylate cyclase activator ataciguat was able to slow the progression of calcification (P=0.051) in mild to moderate CAVD in humans and a murine preclinical model. Aortic osteofibrogenic programs driven by BMP-2 signaling were concomitantly reduced. Moreover, a variety of nucleic acids affect both aging phenotypes and calcification, including microRNAs (miRNAs) such as miR-29,^90^ miR-34a, miR-145, miR-204-5p, and miR-302b, as well as the long noncoding RNA produced by NEAT1, some of which are carried in EVs. These introduce the possibility of specific, targeted therapy, given the capacity of miRNAs to differentially regulate carotid atherosclerosis versus aortic valve stenosis.^61^

Summary and Future Directions

As briefly reviewed, the past decade has yielded an explosion of information on the causes, characterization, and consequences of aortic calcium deposition (Figure 5). However, whether aortic calcification is purely pathologic or has an adaptive function remains unclear. Moreover, although there are clear unifying elements, as first described in the skeleton, there is also mechanistic heterogeneity in the histoanatomic types of aortic tissue mineralization. As such, there is no single best model for in vitro and in vivo studies of vascular calcification, just as there are no best models for studying diabetes, atherosclerosis, or CKD. Different models (Tables 2 and 3) are required that best recapitulate the relevant clinical setting, depending on which disease mechanism (eg, diabetic, dyslipidemic or other metabolic, renal, genetic) and type of calcification (intimal, medial, valvular, generalized soft tissue) are under consideration. With respect to the burgeoning clinical needs of the aging population, models that integrate key features of metabolic syndrome (eg, insulin resistance, hyperlipidemia, hypertension), together with the dysmetabolic state of declining renal function, hold broad relevance. However, such models do not recapitulate important features relevant to cardiovascular calcification that arise with radiation, chemotherapy, or immune modulation administered to treat breast cancer or other malignancies, thus representing an unmet need. Moreover, it is necessary to define the settings wherein vascular calcification directly promotes adverse cardiovascular events, serves to prevent them, or manifests as a biomarker for vascular disease. The impact of specific mineral composition or crystalline phase on vascular physiology has also not been systematically examined. The negative consequences of aortic valve calcium accumulation on cardiovascular structure, function, and clinical outcomes are established. However, whether therapeutic approaches should be focused on preventing, modifying, or promoting vascular calcification as a target beyond the aortic valve needs clarification. For example, establishing whether biomechanically different types of coronary calcification exist—and whether that could explain why certain groups, such as elite endurance athletes, have greater prevalence of vascular calcification but not higher incidence of events—is paramount and requires detailed clinical investigation.

Pathomechanisms of aortic calcification consistently exhibit molecular features of ectopic calcification arising from foci of infection, foreign bodies, and malignancy, and sites of mechanical injury and hemorrhage, highlighting a fundamental connection to potentially beneficial innate immune processes. To advance the field, research should move beyond merely identifying interventions or factors that reduce vascular calcification, focusing instead on determining how changes in calcification affect clinical events and on identifying factors that can modify vascular calcification without adversely affecting skeletal calcification or ectopic calcification involved in immune responses. This may require more collaboration with bone biologists and immunologists.

The emerging mechanism-based therapeutic strategies outlined here, as well as others not captured herein, offer opportunities to validate and vet our fundamental understanding of aortic and vascular calcification. Furthermore, learning how to control and induce mineralization of soft tissues may lead to new therapies for a wide range of noncardiovascular disorders, including tissue regeneration and repair at the bone–tendon interface, fibrous nonunions, and myositis ossificans. Given the recent trajectory of innovation, the next decade promises to deliver effective medical treatments that target both the causes and consequences of aortic calcification.

Article Information

Sources of Fundings

This work was funded by the National Heart, Lung, and Blood Institute (grant HL069229 to Dr Towler, grant R35HL171342 to Drs Giachelli and Scatena, and grants HL137647 and HL151391 to Drs Demer and Tintut).

Disclosures

None.