Hearts of Stone: Calcific Aortic Stenosis and Antiresorptive Agents for Osteoporosis

Linda L. Demer, MD, PhD; Yin Tintut, PhD

Disclosures

Circulation. 2021;143(25):2428-2430. 

The body has developed many failsafe systems to prevent calcification in soft tissues, probably because even tiny amounts of mineral in the wrong places, such as the cardiac valves, can be fatal. Natural inhibitors of calcification include pyrophosphate, fetuin, osteoprotegerin, and MGP (matrix γ-carboxyglutamic acid protein). It comes as a surprise, then, that cardiovascular calcification is so prevalent, even when most other soft tissues are spared. One possibility is that this pathological process derives from a purposeful response to chronic infection and inflammation that evolved in preantibiotic eras from a need to destroy invading parasites and microbes that resist cellular immune mechanisms, such as Toxoplasma gondii or Mycobacterium tuberculosis. This tertiary immune strategy is production of new bone tissue to form a shell that surrounds and confines an infectious organism. A nearly identical process occurs in response to cancer, abscesses, and foreign bodies, such as surgical sponges. This last-ditch response to many circumstances seems to have long-standing, chronic inflammation as its common feature. In many cases in the context of modern human cardiovascular systems, the invading "organisms" might be lipoprotein particles rather than parasites, and the results may be life-threatening rather than protective. In the aorta, calcification impairs the Windkessel phenomenon, reducing cardiac pump function. In valves, it tethers leaflet excursion, leading to functional obstruction.

Although cardiovascular calcification has been observed for centuries, the study of its regulatory mechanisms was delayed until about 1990 because of the widespread belief that it was a passive, degenerative process of aging. The findings of bone matrix vesicles, as well as hydroxyapatite mineral and bone morphogenetic protein[1] in atherosclerotic plaque, followed by demonstration of osteogenic differentiation of vascular cells, raised the possibility that mineralization in arteries is regulated by the same factors that control physiological mineralization of skeletal bone. Subsequent work has identified several regulatory factors that include hormones (eg, parathyroid hormone), enzymes (eg, tissue-nonspecific alkaline phosphatase), transcription factors (eg, Runt-related transcription factor 2), and matrix components (eg, osteocalcin).[2] The inhibitory factors found in soft tissues may be overwhelmed by pathological factors and conditions, such as modified lipoproteins, uremia, activated immune cells, cytokines, and inorganic phosphate. As reported by Davaine et al,[3] 65% of atherosclerotic lesions from femoral endarterectomy specimens show fully formed bone tissue, including trabeculae, osteocytes, and marrow. It is also not an unusual finding in calcific aortic valve disease.[4] As an aside, because bone is defined by its location in the skeleton, the terms osteoid metaplasia or bone-like tissue are often used in the context of the cardiovascular system, even when the tissue is indistinguishable from skeletal bone at all levels from molecular to histological architecture.

Seminal work in translating these concepts to the clinical realm has come from an investigative team at the University of Edinburgh. On the basis of the fact that sodium fluoride (the same agent used in dental care) has been used for decades in clinical bone scan imaging to detect metastatic bone formation or destruction, they showed that a positron-emitting form of sodium fluoride labeled with fluoride-18 (18F-NaF) could identify coronary calcification on positron emission tomographic imaging and that lesions positive for tracer uptake were the "culprit lesions" responsible for ischemia or acute coronary syndrome in patients.[5]

Several epidemiological studies show a correlation between cardiovascular calcification and osteoporosis independent of aging. The coexistence of atherosclerotic calcification with osteoporosis in patients suggests that they share common etiologic factors (eg, hyperlipidemia). Side-by-side in vitro studies showed that factors (eg, modified lipoproteins) promoting mineralization in vascular cells have the opposite effect on mineralization in skeletal bone cells. Similarly, inflammatory cytokines (tumor necrosis factor-α and interleukins) that promote mineralization in vasculature also induce bone resorption. Together, these findings raise the possibilities that therapies for osteoporosis may affect cardiovascular calcification and vice versa. In other words, osteoporosis inhibitors could be hypothesized to either improve or worsen vascular calcification. It is interesting that the study hypothesis did not explicitly raise the latter possibility, addressing only a potential benefit rather than a potential adverse effect.

Accordingly, in this issue of Circulation, Pawade et al[6] from the University of Edinburgh assessed the effects of bone resorption inhibitors, denosumab and alendronate, on cardiac valve calcification and aortic stenosis in a double-blind, randomized controlled trial in patients with calcific aortic stenosis. Effective inhibition of bone resorption was confirmed (on the basis of the release of carboxy-terminal telopeptides of fibrillar collagens into the circulation), but no effect was seen on progression of aortic valve calcification or function even after 2 years. This result may come as a relief to patients with osteoporosis but as a disappointment to those with calcific aortic stenosis. Because inhibition of mineral resorption has the potential to increase mineral density in the valves as it does in the skeleton, it may have raised concerns that the treatment might initiate or accelerate valve disease.

This result may also lead some to wonder if cardiovascular calcification is even an osteogenic process at all, especially given a previous study showing denosumab also did not affect aortic calcification in patients with osteoporosis.[7] However, these findings are remarkably consistent with what is known about osteogenic processes in the vasculature: that bone resorptive activity is seriously impaired in the context of the vascular wall.

Most first-line therapies for osteoporosis work by interfering with the bone resorptive cells, osteoclasts, which originate from bone marrow myeloid precursors, circulating monocytes, and, potentially tissue macrophages. Denosumab, a monoclonal antibody to human RANKL (receptor activator of nuclear factor kB ligand) reduces bone loss by blocking formation of bone-resorbing osteoclasts by interfering with RANKL binding to its receptor on monocytes, which is required for differentiation into osteoclasts. Alendronate, a nitrogen-containing bisphosphonate similar to the endogenous inhibitor pyrophosphate, also acts by inducing apoptosis of osteoclasts and by binding to the bone mineral hydroxyapatite, suppressing its breakdown of hydroxyapatite. The RANKL/osteoprotegerin axis is also active in cardiovascular tissue. In vitro, RANKL induces osteoblastic differentiation of vascular smooth muscle cells indirectly via endothelial production of BMP-2 (bone morphogenetic protein-2). In preclinical studies, treatment of hyperlipidemic mice with a soluble decoy receptor for RANKL, recombinant osteoprotegerin, inhibited aortic calcification.

One innovative research direction for a natural approach to reverse cardiovascular calcification is to promote resorption by osteoclasts, essentially to induce an osteoporosis-like state within cardiovascular tissues. Direct implantation of osteoclasts at sites of subdermal elastin calcification in rats reduced elastin mineralization by almost 50% without damage to the elastin fibers.[8] However, it turns out that, in the vasculature, osteoclastic cells are rarely found. When they are present, they tend to be nonfunctional. In human atherosclerotic plaque, osteoclastic cells near calcium deposits were shown to have a dysfunctional phenotype characterized by low expression and activity of hydrolases and defective resorption.[9] When osteoclastic cells were derived from human femoral artery calcium deposits, very few were able to resorb mineral in vitro.[3] Given the evidence that osteoclastic resorption is impaired in the vasculature, if this translates to the clinical context, then inhibitors that target resorption, such as bisphosphonates, are unlikely to modulate cardiovascular calcification regardless of stage. In contrast, therapeutic approaches based on inducing mineral resorption by activating osteoclastic cells through RANKL signaling in macrophages (not in bone marrow- or spleen-derived osteoclasts) have a greater likelihood of reducing cardiovascular calcification.[10]

The mechanism for osteoclast impairment in the vasculature is unknown, but evidence points to active inhibition by vascular cells. In culture, vascular smooth muscle cells produce factors that inhibit osteoclast differentiation, including osteoprotegerin and interleukin-18, and blocking interleukin-18 restored osteoclast differentiation and production of the mineral-resorbing acid phosphatase.[11] Vascular smooth muscle cells also produce osteopontin, and exogenous osteopontin has been shown to inhibit normal macrophage-to-osteoclast differentiation,[12] instead promoting an alternative, anti-inflammatory phenotype, with impaired hydrolase activity.[13] It remains possible that when vascular calcification has metabolic, rather than inflammatory, origins, such as in chronic kidney disease, osteoporosis treatment may have different effects.[14,15]

Thus, osteoporosis treatments, such as denosumab and alendronic acid, that act by inhibiting osteoclastic activity, were found to have no effect on atherosclerotic vascular calcification, and a likely mechanism is that osteoclastic function and resorptive activity are already seriously impaired in the cardiovascular context. This may extend to other forms of cardiovascular calcification, such as calcific aortic stenosis. A key future research direction is to find ways to target proresorptive activity specifically to cardiovascular tissues.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....