Aging and bone loss: new insights for the clinician

Abstract

It is well known that the underlying mechanisms of osteoporosis in older adults are different than those associated with estrogen deprivation. Age-related bone loss involves a gradual and progressive decline, which is also seen in men. Markedly increased bone resorption leads to the initial fall in bone mineral density. With increasing age, there is also a significant reduction in bone formation. This is mostly due to a shift from osteoblastogenesis to predominant adipogenesis in the bone marrow, which also has a lipotoxic effect that affects matrix formation and mineralization. We review new evidence on the pathophysiology of age-related bone loss with emphasis upon the mechanism of action of current osteoporosis treatments. New potential treatments are also considered, including therapeutic approaches to osteoporosis in the elderly that focus on the pathophysiology and potential reversal of adipogenic shift in bone.

Introduction

As a result of the aging process, the bone deteriorates in composition, structure and function, which predisposes to osteoporosis. Osteoporosis is defined as deterioration in bone mass and micro-architecture, with increasing risk to fragility fractures [Raisz and Rodan, 2003]. Owing to the close relationship between the aging process of bone and the pathogenesis of osteoporosis, research on the mechanisms of age-related bone loss has increased significantly in recent years involving a combination of basic, clinical, observational and translational studies.

Bone is a dynamic organ that serves mechanical and homeostatic functions. It undergoes a continual self-regeneration process called remodeling. Remodeling removes old bone and replaces it with new bone. This regenerative process occurs in distinct areas of bone known as bone metabolic units (BMUs) [Riggs et al. 2002]. Within each BMU bone formation by osteoblasts and bone resorption by osteoclasts is coupled tightly in a delicate balance to maintain bone mass and strength to resist deformity. With aging this balance shifts in a negative direction, favoring greater bone resorption and less bone formation. This combination of bone mass deficiency and reduction in strength ultimately results in osteoporosis and fractures.

Aging in combination with intrinsic and extrinsic factors accelerates the decline in bone mass that predisposes to fractures. Intrinsic factors include genetics, peak bone mass accrual in youth, alterations in cellular components, hormonal, biochemical and vasculature status. Extrinsic factors include nutrition, physical activity, comorbid medical conditions and drugs. In this article we review the mechanisms of age-related bone deterioration and their impact on the pathogenesis of osteoporosis. In addition, current and future therapeutic approaches focused on the correction of mechanisms associated with aging bone will also be outlined.

Bone remodeling in aging bone

Remodeling is continuous and coordinated cycle of removal of old bone by osteoclasts followed by the deposition of new bone by osteoblasts in response to micro damage and variable mechanical loadings. Bone remodeling is a continuous process throughout life. In the first three decades of life, bone turnover is coupled tightly to maintain a steady state between bone resorption and bone formation. Although there are variances in turnover rates, peak bone mass and size is achieved around the age of 15–20 years in women and later in men [Raisz and Seeman, 2001]. After this, long before sex steroids deficiency occurs, bone loss becomes evident [Slemenda et al. 1996]. After reaching the peak of bone mass, bone turnover continues at a slower rate as suggested by a rapid decline in biochemical measures of bone remodeling with the predominance of bone resorption over bone formation [Raisz and Seeman, 2001]. Later in life, menopause in women significantly increases bone resorption over formation due to low levels of estrogens thus inducing accelerated bone loss. In contrast to the mechanisms of bone loss during menopause, which have been studied extensively, the triggers of an age-related transition from a steady state to one of negative net bone loss (both in women and men) remain poorly understood.

At the bone surface level, age-related bone loss is the consequence of two simultaneous but opposing processes: subperiosteal apposition, which takes place on the outside of the bone, and endosteal bone resorption, which takes place on the inside of the bone. With increasing age, bone remodeling is reduced leading to a negative bone balance at individual BMU sites. After the fourth decade of life, there is a reduction in the formation of periosteal bone and at the same time there is increasing number of remodeling units within endosteal bone resulting in a linear increase in endosteal bone resorption in both sexes. The overall consequences of these age-related changes are cortical thinning, increased cortical porosity, thinning of the trabeculae and loss of trabecular connectivity, all of which reduce bone quality and consequently bone strength [Rosen et al. 1994].

The above opposing processes are consistent with longitudinal and cross-sectional studies which showed a relatively slow rate of decline in areal bone mineral density (aBMD) in both sexes beginning at age 40 and continuing throughout the adult life [Khosla and Riggs, 2005]. Large decreases in lumbar spine volumetric BMD (vBMD) secondary to predominant vertebral trabecular bone loss beginning in the third decade and linear decrease in cortical vBMD in the wrist were also demonstrated in both sexes with advancing age [Riggs et al. 2004]. The changes were greater in women than men, owing to accelerated bone loss in the menopausal stage.

In terms of the effect of aging on periosteal bone formation, the increasing levels of endosteal bone loss are concomitant with steady levels of periosteal apposition somewhat compensating for the loss of bone mass. Therefore, cortical bone loss is less in men than in women because periosteal bone formation is greater and is independent of endosteal bone resorption. Bone loss reflects the net result of all of the periosteal bone formed during aging minus all of the bone irreversibly removed from the endosteal surface [Seeman, 2002], a process that seems to be independent of hormones and closely related to potential age-related mechanisms.

In terms of vBMD in the hip, a study by Center and colleagues in 852 women and 635 men (60 years and older) without fractures reported an age-related decline in vBMD in the hip [Center et al. 2004]. In addition, vBMD was more sensitive than areal BMD in older men and similar to that in women, in whom sensitivity was similar for both areal (73%) and estimated volumetric (78%) BMD cutoffs. The authors conclude that men and women have hip fractures at the same estimated femoral neck vBMD suggesting that vBMD can provide a useful single measure that could be used in both men and women.

Mechanisms of age-related bone loss

Secondary hyperparathyroidism

Both calcium and vitamin D deficiency can contribute to secondary hyperparathyroidism [Lips, 2001]. Vitamin D deficiency is prevalent in the older population irrespective of latitude [Lips, 2007]. A low serum 25(OH)D concentration leads to a small decrease in serum 1,25-(OH)₂D and calcium absorption which then stimulates an increase in parathyroid hormone (PTH) secretion. In addition vitamin D is required for osteoblastogenesis and bone formation [Duque and Troen, 2008]. The increased serum PTH subsequently increase osteoclastic activity and bone resorption, resulting in primarily cortical bone loss [Lips, 2001]. A chronic negative calcium balance state can also occur independently of vitamin D as a result of age-related reduced intestinal calcium absorption [Eastell et al. 1991] associated with reduced dietary intake. This deficiency, when not adequately compensated through dietary means or calcium supplements, contributes to physiological secondary hyperparathyroidism.

With age, a number of other factors can also cause an increasing PTH levels. Common factors include impaired renal function, the use of loop diuretics such as furosemide and estrogen deficiency. In women, there is some suppression of PTH secretion during the rapid phase of bone loss in early postmenopausal period. In the later stage however there is gradually increasing PTH secretion which increases bone turnover [Ledger et al. 1995].

PTH secretion also increases in aging men, similar to what is seen in aging women [Khosla et al. 2008a; Khosla, 2010]. Normal circulating gonadal sex steroid levels in aging men may help to protect against bone resorption promoted by increased PTH levels. Thus it has been more difficult to demonstrate a direct role for PTH in causation of age-related bone loss in men [Kennel et al. 2003].

Gonadal sex steroid deficiency

It is well known that sex steroids have significant effects on skeletal health. The cessation of ovarian function associated with reduced estrogen levels at menopause is the start of rapid bone loss in women. During the menopause transition, serum 17b- estradiol levels decrease by 85–90% and serum estrone levels decrease by 65–75% from mean premenopausal levels [Khosla et al. 1997]. In fact, there may be a threshold level of serum bioavailable (non-sex hormone binding globulin [non-SHBG]-bound) estradiol below 11 pg/ml and below 11 pg/ml at which trabecular and cortical bone loss occurs, respectively [Khosla et al. 2005]. This phase of accelerated bone loss may persist for up to 10 years after menopause in most women.

The mechanisms of estrogen deficiency related bone loss are multiple and their relative importance in the pathogenesis of this process remains poorly understood [McCauley et al. 2003]. In general, effect of estrogen deficiency on bone is the result of loss of restraint and control estrogen has over mediators of bone resorption. Usually, estrogen may inhibit osteoclast formation and activity by increasing the production of osteoprotegerin (OPG), or transforming growth factor β (TGF-β) [Hofbauer et al. 1999; Hughes et al. 1996]. OPG is a soluble decoy receptor for receptor activator of nuclear factor kappa-B ligand (RANKL) and TGF-β induces osteoclast apoptosis [Lundberg et al. 2001]. In vitro and in vivo studies have also shown that estrogen suppresses RANKL production by osteoblastic cells and T and B lymphocytes [Eghbali-Fatourechi et al. 2003; Clowes et al. 2005]. Estrogen also directly stimulates apoptosis of osteoclast precursor cells, and decreases osteoclast precursor differentiation by blocking RANKL/macrophage colony-stimulating factor (M-CSF)-induced activator protein-1-dependent transcription by reducing c-jun activity [Lundberg et al. 1999; Mitnick et al. 2001]. Indirectly, estrogen may suppress the production of bone-resorbing cytokines such as interleukin (IL)-1, IL-6, TNF-α, M-CSF and prostaglandins [Charatcharoenwitthaya et al. 2007]. Finally, estrogen is also capable of inhibiting the activity of mature osteoclasts by direct, receptor-mediated mechanisms [Oursler et al. 1994]. In addition to changes to estrogen levels, a reduction in ovarian inhibin B across the menopause transition and perimenopausal elevated follicle-stimulating hormone (FSH) also increase bone turnover [Perrien et al. 2006].

In men, traditionally it was assumed that decreased serum testosterone was responsible for age-related bone loss. However estrogen has also been found to play a dominant role in age-related bone loss in men similar to women. A combination of cross-sectional and observational studies of aging men showed better correlations between serum estradiol and BMD than testosterone and BMD at various skeletal sites [Slemenda et al. 1997; Khosla et al. 2001, 2008b; Mellstrom et al. 2008; Szulc et al. 2001]. Further studies looking at differential effects between estrogen and testosterone confirmed that estrogen deficiency was more important than testosterone deficiency in causation of bone loss in aging men [Falahati-Nini et al. 2000; Leder et al. 2003] and that the effects of estrogen on bone were independent of FSH [Sanyal et al. 2008]. More recently, a large prospective study of older men again showed a low bio-available estradiol level to be associated with significant increased fracture risk and that testosterone in the presence of high SHBG is associated with significant increased fracture risk when adjusted for estradiol levels [LeBlanc et al. 2009]. Nevertheless, testosterone contributes to reduced fracture risk in men because of its influence on increasing bone size in men during growth and development [Clarke and Khosla, 2010].

Bone marrow fat

The predominant feature of age-related bone loss is the accumulation of bone marrow fat at the expense of osteoblastogenesis [Rosen and Bouxsein, 2006]. This accumulation of marrow fat appears to be an active process independent of estrogen since it is evident during the third and fourth decade of life [Perrien et al. 2007]. Biopsy studies with animal models [Duque et al. 2009] and humans [Verma et al. 2002; Meunier et al. 1971] have consistently demonstrated a significant increase in marrow fat in aging bone. More recently MRI studies have also demonstrated an age-related increase in marrow fat [Griffith et al. 2005; Shen et al. 2007]. In addition, there is an inverse relationship between marrow fat volume and bone volume that was independent of sex and correlated with the changes seen in people with osteoporosis [Justesen et al. 2001].

Mechanistically, there appears to be a predominant differentiation of mesenchymal stem cells (MSCs) into adipocytes at the expense of osteoblasts [Rosen et al. 2009]. The differentiation of MSC into osteoblasts involves the recruitment of MSCs, release of appropriate amount of growth factors and activation of lineage-specific transcription factors [Duque, 2007; Chamberlain et al. 2007], recruitment of appropriate numbers of MSCs to achieve proper density and confluence [Zhou et al. 2008] and adequate oxygen tension and blood supply within the bone marrow [Wang et al. 2007]. Changes induced by aging can alter these conditions thus facilitating MSCs differentiation into adipocytes [Zhou et al. 2008]. Furthermore, the main lineage-specific transcription factors that direct the differentiation of MSCs are the runt-related transcription factor 2 (Runx2) for osteoblastogenesis and peroxisome proliferator-activator gamma 2 (PPARγ2) for adipogenesis [Rosen and Bouxsein, 2006; Gimble et al. 2006]. With aging, there is a predominant expression of PPARγ2 by MSCs with a concurrent decrease in Runx2 expression and therefore lower levels of osteoblast differentiation [Lecka-Czernik, 2006].

Recently, a protein of the nuclear envelope known as lamin A/C has been reported as an essential factor in the osteogenic differentiation of MSCs. Lamins are intermediate filament proteins present in the nuclear lamina and matrix and are important regulators of stem cells differentiation [Hutchinson and Worman, 2004; Pajerowski et al. 2007]. Most adult mammalian somatic cells contain three major lamins grouped into two classes: A type (A, AΔ10 and C) and B type (B1 and B2) [Li et al. 2011]. With aging there is a decrease in lamin A expression in normal osteoblasts [Duque and Rivas, 2006]. The importance of lamins in bone biology is evident from changes in bone mass seen in patients suffering from Hutchinson Gilford Progeria Syndrome (HGPS). Patients suffering from HGPS have mutations in the lamin A/C gene [Sandre-Giovannoli et al. 2003] and show major bone changes including severe osteoporosis, osteolysis, bone deformities and spontaneous fracture [Rodrigues et al. 2002]. A similar syndrome in mice caused by defects in type A lamins resulting in low levels of lamin A/C was associated with low BMD [Mounkes et al. 2003]. Mice lacking the enzyme responsible for lamin A/C processing (Zmpste24^−/−) also show accelerated bone loss and typical features of senile osteoporosis [Rivas et al. 2009].

Recently, an in vivo study of knockout lamin A/C mice demonstrated that the absence of lamin A/C increased the expression of MAN-1 protein which co localizes with Runx2 thus affecting its ability as an osteogenic transcription factor [Li et al. 2011]. This demonstrates that lamin A/C is required in osteoblastogenesis and bone formation in vivo.

In fact, aging per se, independently of hormonal changes, appears to contribute significantly to bone marrow adipogenesis raising the possibility that senile osteoporosis is a type of lipotoxic disease [Duque and Troen, 2008]. Indeed bone marrow adipocytes appear to exert a toxic effect on osteoblasts [Maurin et al. 2000]. Cocultures of adipocytes and osteoblasts reveal that adipocytes inhibit osteoblast activity and survival, possibly secondary to the release of adipokines and fatty acids by the increased number of adipocytes within the bone marrow [Musacchio et al. 2007]. Further evidence of the lipotoxicity of marrow adipocytes on bone comes from the observation of PPARγ induction by thiazolidenediones. The use of thiazolidenediones in diabetic patients was associated with bone loss and higher incidence of fractures [Grey, 2008]. The increasing levels of PPARγ induced by thiazolidenediones within the bone marrow not only affect bone formation, but also induce bone resorption [Lazarenko et al. 2007]. Furthermore, in a mice model, PPARγ was proven to functions as a direct regulator of osteoclastogenesis [Wan et al. 2007]. Given this observation, it was considered that blocking PPARγ could provide a new therapeutic approach for osteoporosis. Although PPARγ knockout mice showed high levels of bone formation [Akune et al. 2002], pharmacological attempt to increase bone mass by blocking PPARγ in diabetic mice was unsuccessful despite decreased marrow fat [Botolin and McCabe, 2006].

Other factors

A number of clinical studies demonstrated that body fat and bone mass were directly related [Felson et al. 1993; Lindsay et al. 1992; Glauber et al. 1995; Khosla et al. 1996]. It was further observed that serum leptin levels were increased in obesity and correlated positively with fat mass [Considine et al. 1996]. Subsequently the hormone mediating the relationship between fat mass and bone mass was demonstrated to be leptin. An in vitro study showed that leptin acted on human marrow stromal cells to enhance osteoblast differentiation and inhibited adipocyte differentiation [Thomas et al. 1999]. Further animal studies also reported a central regulatory role of leptin [Ducy et al. 2000; Takeda et al. 2002]. More recently, in a loss of function of its receptor mice study, leptin was shown to regulate bone mass accrual in vivo by acting through neuronal means [Shi et al. 2008].

Serotonin was also shown to regulate bone mass in rodents [Yadav et al. 2009]. In humans the role for circulating serotonin in regulating bone mass was recently suggested by the findings from a study of premenopausal and postmenopausal women [Mödder et al. 2010]. Serotonin levels were inversely associated with body and spine aBMD, and with femoral neck total and trabecular vBMD. Serotonin levels remained significant negative predictors of femur neck total and trabecular vBMD, as well as trabecular thickness at the radius, after adjusting for age and BMI.

Attainment of peak bone mass is another factor contributing to later age-related bone loss. Those persons who achieve a higher peak bone mass are less likely to develop osteoporosis later in life as age-related bone loss ensues, whereas those with low levels are at greater risk [Seeman, 1997]. Numerous other factors such as corticosteroids usage, diseases such as malabsorption, anorexia nervosa and idiopathic hypercalciuria, and behavioral factors such as smoking, alcohol abuse and inactivity can also contribute to fracture risk in 40% of men and 20% of women in the older population [Riggs and Melton, 1986]. Finally, although controversial still, sarcopenia, probably through reduced muscle loading on bone, may also contribute to age-related bone loss [Mödder et al. 2010; Seeman, 1997].

The role of exercise

Aging is associated with a decline in physical activity and mechanical loading. Reduced mechanical loading exert diminished effects upon osteoblasts resulting in decreased osteoblast secretion of OPG and increased expression and secretion of RANKL, IL-1, IL-6, IL-11, and TNF-α. In turn, these compounds directly stimulate greater osteoclast formation and activity. The reduced OPG also permits greater binding of RANKL to RANK, which further facilitates increased osteoclastogenesis and resorption [Duque and Troen, 2008]. Animal studies of complete immobilization showed a striking remodeling imbalance with a rapid, although transient, increase in bone resorption followed by a sustained decrease in bone formation [Weinreb et al. 1989]. Recently, an anabolic response to exercise was shown to be related to the presence of lamin A/C [Duque et al. 2011b]. Lamin A/C haploinsufficient mice exposed to strenuous exercise demonstrated significant trabecular and cortical thinning and a reduction in osteoblasts and osteocyte numbers compared with their sedentary counterparts whilst the wild type exposed to exercise showed a significant increase in bone volume and number of bone cells. In humans, functional loading has been shown to improve bone mass [Leichter et al. 1989] and exercise training programs can prevent or reverse almost 1% of bone loss per year in both lumbar spine (LS) and femoral neck (FN) for both premenopausal and postmenopausal women [Wolff et al. 1999].

Bone loss due to decreased bone formation

Although sex steroids deficiency may contribute to age-related impairment of bone formation, reductions in key growth factors important for osteoblast differentiation/function may also contribute. Aging is associated with decreases in the amplitude and frequency of growth hormone secretion from the anterior pituitary [Marie et al. 1993] with growth hormone levels declining by up to 14% per decade in both elderly men and women [Rosen et al. 1994] leading to a decrease in hepatic production of insulin-like growth factor (IGF)-1 [Boonen et al. 1999; Pfeilschifter et al. 2000] and smaller decreases in IGF-2 [Boonen et al. 1999]. In addition to decreased systemic and local skeletal production of IGF-1 and IGF-2, growth factor binding proteins may also contribute to age-related bone loss. Higher serum IGF binding protein (IGFBP)-2 predicted lower BMD, and was associated with increased markers of bone resorption independent of age, body mass, and sex hormones [Amin et al. 2007].

Changes in key proteins such as sclerostin have also been implicated in bone formation reduction. Sclerostin (SOST), a glycoprotein primarily secreted by osteocytes is a potent inhibitor of osteoblastogenesis. SOST binds to coreceptors LRP5 and LRP6 and prevents colocalization with frizzled protein and Wnt signaling, thereby reducing osteoblastogenesis and bone formation [Kneissel, 2009]. Loss-of-function mutations of the SOST gene are associated with an autosomal-recessive disorder, sclerosteosis, which causes progressive bone overgrowth [Balemans et al. 2001], a deletion downstream of this gene, which causes reduced SOST expression, is associated with a milder form of the disorder called van Buchem disease [Balemans et al. 2002] and finally SOST-null mice have a high bone mass phenotype [Li et al. 2008]. Consistent with these observations, pharmacologic inhibition of SOST has shown significant anabolic effects. In aged ovariectomized rat model of postmenopausal osteoporosis, treatment with SOST neutralizing monoclonal antibody resulted in marked increases in bone formation on trabecular, periosteal, endocortical, and intracortical surfaces [Li et al. 2009]. In addition, the increases in bone formation induced by antisclerostin antibody are not associated with increases in bone resorption [Lane and Silverman, 2010]. SOST also appears to have a role in mediating bone responses to mechanical unloading. SOST knockout mice were resistant to bone loss induced by mechanical unloading [Lin et al. 2009]. In humans, a recent phase I study of a single dose of a SOST monoclonal antibody (AMG 785) administered to healthy men and postmenopausal women was associated with dose-related increases in the bone-formation markers procollagen type 1 N-propeptide (P1NP), bone-specific alkaline phosphatase (BAP) and osteocalcin, and with a dose-related decrease in the bone-resorption marker serum C-telopeptide (sCTx) [Padhi et al. 2011].

Cathepsin K

Normal bone resorption and remodeling critically depend upon the synthesis and secretion of cathepsin K (CTSK) by osteoclasts [Troen, 2004; Yasuda et al. 2005; Motycykova and Fisher, 2002]. Bone resorption begins when osteoclasts bind firmly to bone surfaces forming resorption pits. An acid medium is produced within these pits resulting in the dissolution of the osseous mineral component exposing the organic matrix. The matrix is then degraded by the enzymes metalloproteinases and CTSK. In fact, RANKL, which plays a critical role in osteoclast differentiation and activation, has been shown to stimulate CTSK mRNA and protein expression in human osteoclasts [Shalhoub et al. 1999]. Indeed many of the agents that have been shown to induce osteoclast formation and activation or to inhibit osteoclast activity enhance and suppress, respectively, CTSK gene expression [Troen, 2006].

Agents that stimulate the osteoclast to produce increased amounts of CTSK include NFAT, TNF, IL-1, PPARΔ/β, stretching, and extracellular matrix proteins (ECM). Inhibitors of CTSK expression include estrogen, interferon-γ (IFN-γ), and OPG [Troen, 2004]. A phase I study of a CTSK inhibitor, odanacatib (ODN), showed that it was well tolerated, had a long half life, and exhibited significant and sustained suppression of bone resorption markers with weekly and daily regimen with no effects on markers of bone formation [Stoch et al. 2009]. A subsequent phase II study of postmenopausal women [Bone et al. 2010] demonstrated dose-dependent increases in BMD in all sites. The greatest increase was seen with the highest dose. Resorption markers fell in a dose-dependent manner for the first 6 months after which they increased and the difference with placebo disappeared. Bone formation markers increased with significant differences compared with placebo observed at 12 and 24 months. Recently, results of an extension of the phase II study for another year was reported [Eisman et al. 2011]. Continued treatment with 50 mg of ODN for 3 years produced significant increases from baseline and from year 2 in BMD at the spine (7.9% and 2.3%) and total hip (5.8% and 2.4%). Urine cross-linked N-telopeptide of type I collagen (NTx) remained suppressed at year 3 (−50.5%), but bone-specific alkaline phosphatase (BSAP) was relatively unchanged from baseline. Treatment discontinuation resulted in bone loss at all sites, but BMD remained at or above baseline. After ODN discontinuation at month 24, bone turnover markers increased transiently above baseline and resolved by month 36. There were similar overall adverse event rates in both treatment groups.

Finally, an important extrinsic factor causing reduced bone formation is glucocorticoids.

Glucocorticoids suppress bone formation by inhibiting Wnt/β-catenin signaling thereby impairing osteoblastogenesis, inhibit osteoblastic function directly and by inhibiting IGF-I synthesis [Canalis et al. 2007].

Osteoporosis therapy: present and future

Based on the mechanisms underlying age-related bone loss, the main goals of therapy should include the inhibition/restriction of osteoclastic activity, the enhancement of osteoblastic activity, and the regulation of bone marrow adipogenesis. In addition, contributing factors should be corrected or minimized. Currently the main classes of agents are antiresorptives, which suppress osteoclastic activity, and anabolic agents, which target osteoblasts (Table 1).

Table 1.

Pharmacological effect of osteoporosis treatments on the typical features of age-related bone loss (adapted from Duque and Troen, 2008).

Compound	Osteoblast	Adipocyte	Osteoclast
Bisphosphonates	↑ differentiation	↓ differentiation	↓ differentiation
	↑ activity		↓ activity
	↓ apoptosis		↑ apoptosis
Calcitonin			↓ activity
			↑ apoptosis
PTH	↑ activity	↓ differentiation	↑ activity
	↑ survival
	↑ differentiation
SERMs			↓ differentiation
			↓ activity
Strontium ranelate	↑ activity		↓ activity
	↑ differentiation		↓ survival
Vitamin D	↑ activity	↓ differentiation	↑ activity
	↑ differentiation	↑ trans-differentiation to osteoblasts
	↓ apoptosis
Denosumab			↓ differentiation
			↓ activity
			↑ apoptosis

PTH, parathyroid hormone; SERM, selective estrogen-receptor modulator

Antiresorptives

This class comprises bisphosphonates, hormone replacement therapy (HRT), selective estrogen- receptor modulators (SERMs), strontium ranelate, and RANKL antibody.

The nitrogen-containing bisphosphonates, such as alendronate, risedronate, ibadronate, and more recently zoledronate are first-line agents for the treatment and prevention of osteoporosis. These agents inhibit bone resorption by inducing osteoclasts apoptosis, thus reducing the number of osteoclasts in the BMU [Riggs and Melton, 1986], suppressing the capacity of osteoclasts to resorb bone by modifying their shape and nullifying their enzymatic capacity and also enhances secondary mineralization of preformed osteons [Russell, 2006]. There is also evidence suggesting that bisphosphonates (specifically alendronate) may promote osteoblast proliferation and maturation [Boonen et al. 1999], while inhibiting bone marrow adipogenesis [Duque and Rivas, 2007; Duque et al. 2009].

Most of the bisphosphonates have well-established antifracture efficacy. Alendronate has vertebral fracture reduction efficacy in postmenopausal women [Black et al. 1996], in men [Orwoll et al. 2000], in glucocorticoid-induced osteoporosis [Adachi et al. 2001] and nonvertebral fracture reduction including hip fractures [Wells et al. 2008a]. Similarly risedronate reduce vertebral fractures [Harris et al. 1999; Reginster et al. 2000], nonvertebral fractures [Wells et al. 2008b], hip fractures in osteoporotic older women [McClung et al. 2001], and glucocorticoid-induced osteoporosis [Reid et al. 2000].

Ibandronate has demonstrated efficacy against vertebral fractures in postmenopausal women and non vertebral fractures in higher-risk subgroup (femoral neck BMD T score < −3.0) [Chesnut et al. 2004]. The intravenous bisphosphonate zoledronate also has fracture reduction efficacy for vertebral, hip, and nonvertebral fractures, in both men and women [Black et al. 2007; Lyles et al. 2007] and glucocorticoid-induced osteoporosis [Reid et al. 2009].

Among these bisphosphonates, there are differences in side-effect profile, tolerability, and compliance rate. A review of these clinical considerations including treatment duration is discussed in an earlier issue [Langdahl and Harsløf, 2011] in this journal.

Denosumab

Since identified as a key molecule in mediating osteoclast development, activity, and survival [Lacey et al. 1998], the inhibition of RANKL activity has been tested as a therapeutic target for osteoporosis. Recently, a fully human monoclonal antibody to RANKL called denosumab was developed and tested. Denosumab blocks RANKL binding to RANK thus inhibiting the development and activity of osteoclasts. Denosumab administered every 3 or 6 months to postmenopausal women with low BMD over 12 months resulted in an increase in bone mineral density at the lumbar spine of 3.0–6.7%, at the total hip of 1.9–3.6%, and at the distal third of the radius of 0.4–1.3% [McClung et al. 2006]. Near-maximal reductions in mean levels of serum C-telopeptide from baseline were evident 3 days after the administration of denosumab.

Six-monthly subcutaneous injections of denosumab for 36 months was shown to reduce the risk of new radiographic vertebral fractures by 68%, reduce hip fractures by 40% and reduced the risk of nonvertebral fractures by 20% [Cummings et al. 2009]. There was no increase in the risk of cancer and infection compared with placebo; however, the major concern about the long-term use of denosumab relates to its possible effects on the immune system, since RANKL is expressed not just on bone cells but also on immune cells. Although not statistically significant there was a significant increase in rates of eczema and hospitalizations for cellulitis [Cummings et al. 2009] and more neoplasms and serious infections in the denosumab group compared with placebo [McClung et al. 2006]. These results suggest ongoing surveillance of patients receiving denosumab is prudent, particularly when the drug is used in the wider community in patients with comorbidities that might not have been included in clinical trials. Nevertheless it has several advantages over the bisphosphonates: (1) convenient biannual subcutaneous administration that could improve adherence; (2) lack of gastrointestinal side effects; (3) reversibility, because it targets RANKL and is not incorporated into the bone mineral; and (4) useful for impaired renal function because of nonelimination by the kidneys.

Other antiresorptives

The other antiresorptives such as HRT and SERMs have largely fallen out of recommendation in recent years. Although hormone therapy reduces vertebral, nonvertebral, and hip fractures, this is offset by increased risk of breast cancer and cardiovascular diseases [Mödder et al. 2010; Seeman, 1997]. Whilst raloxifene, the only SERM approved for the prevention and treatment of postmenopausal osteoporosis, only has vertebral fracture efficacy [Riggs and Melton, 1986] and is associated with increased risks of venous thromboembolic events and hot flushes. A newer SERM, lasofoxifene, showed a small advantage over raloxifene with a 24% reduction in nonvertebral fracture at 5 years (primarily peripheral fractures, however) [Becker, 2010].

Anabolics

PTH increases bone formation through several actions, including increasing commitment of MSCs to the osteoblast lineage, increasing osteoblast maturation and possibly life span, and reducing the osteocyte production of sclerostin to further stimulate bone formation. PTH stimulation of osteoblastogenesis also increases RANKL production, which then stimulates osteoclast maturation and activity, increasing bone remodeling overall; however, the overall effect is a positive formation balance [Lane and Silverman, 2010].

The anabolic effect of exogenous PTH was first reported in humans 20 years ago. Paired bone biopsies from a small group of patients receiving teriparatide by daily sc injections for 6–24 months demonstrated substantial increases in iliac trabecular bone volume, with evidence of new bone formation [Reeve et al. 1980]. Since then two anabolic agents have been approved for the treatment of osteoporosis, teriparatide, a 1–34 amino acid fragment of human recombinant PTH [PTH (1–34)] and in Europe, the full-length PTH (1–84) molecule.

Teriparatide has shown vertebral and nonvertebral fracture reduction in postmenopausal women with osteoporosis [Neer et al. 2001]. In men with osteoporosis, those who received teriparatide and who may have received follow-up antiresorptive therapy had a decreased risk of moderate and severe vertebral fractures [Kaufman et al. 2005]. Teriparatide also has efficacy in glucocorticoid-induced osteoporosis. Compared with alendronate, teriparatide induced earlier and greater gains in BMD at the lumbar spine and total hip and was more effective in preventing new vertebral fractures [Saag et al. 2007]. As for PTH(1–84) efficacy against vertebral fractures in postmenopausal women has been demonstrated [Greenspan et al. 2007].

Dual mode of action

Strontium ranelate appears to have a mixed mode of action by increasing bone formation and reducing bone resorption leading to rebalancing of bone remodeling in favor of bone formation. Mechanisms by which strontium acts include: increases osteoblast replication, differentiation, and activity [Canalis et al. 1996; Caverzasio, 2008], downregulates osteoclast differentiation and activity [Baron and Tsouderos, 2002; Takahashi et al. 2003], increases the OPG/RANKL ratio directly [Atkins et al. 2009] or via a calcium-sensing receptor [Brennan et al. 2009] and increases apoptosis of osteoclasts [Mentaverri et al. 2003].

Strontium ranelate reduced vertebral fractures [Meunier et al. 2004] and nonvertebral fractures including hip [Reginster et al. 2005] in postmenopausal osteoporosis. Trial evidence suggests that it is effective in a wide range of patient profiles, from early postmenopausal women with osteopenia to elderly women over the age of 80 years, and its antifracture efficacy is independent of baseline severity of osteoporosis, bone turnover level or the presence of clinical risk factors [Reginster et al. 2010]. Furthermore, the antifracture efficacy of strontium ranelate is sustained over 8 years [Reginster et al. 2009].

Novel approaches to osteoporosis treatment

Newer agents with novel modes of actions are under investigation in clinical trials or have shown promissory results in animal studies (Table 2).

Table 2.

Novel approaches to osteoporosis treatment.

Compound	Osteoblast	Adipocyte	Osteoclast
Cathepsin			↓ differentiation
K inhibitor			↓ activity
			↑ apoptosis
SOST antibody	↑ activity
Interferon γ	↑ activity	↓ differentiation	↑ activity (uncoupling favoring formation)
	↑ differentiation
BMP agonists	↑ activity
	↑ differentiation

BMP, bone morphogenetic protein; SOST, sclerostin.

SOST antibody

Positive bone formation results from a recent phase I study [Padhi et al. 2011] make the SOST antibody a promising therapeutic drug. At this time, however, the monoclonal antibody to SOST is in early phase II clinical trials in men and postmenopausal women with osteoporosis [ClinicalTrials.gov: NCT01101048]. The long-term safety of SOST is yet to be addressed.

Cathepsin K inhibitor (odanacatib)

At 36 months, ODN achieves increases in BMD similar to zoledronate and denosumab. However, compared with these agents, the reduction in resorption markers is less but there is also a smaller reduction in bone formation markers. What is not available yet is fracture data. The results of an ongoing randomized double-blind placebo-controlled clinical trial of 16,000 treatment-naïve postmenopausal women age 65 and above [ClinicalTrials.gov: NCT00529373] is eagerly anticipated.

New anabolic targets

Three regulatory proteins, which have also been intensely investigated as potential therapeutic targets, are bone morphogenetic proteins (BMPs), elements of the Wnt signaling pathways and IFNγ. BMPs [Canalis et al. 2003] and Wnt [Krishnan et al. 2006] induce the differentiation of mesenchymal cells toward mature osteoblasts. BMPs may also induce osteoclastogenesis by enhancing the expression of RANKL [Kaneko et al. 2000]. Despite the use of locally administered BMPs for the treatment of nonunion fractures and to enhance the formation of spinal fusions, the systemic administration of BMPs would be limited by their nonskeletal effects, mitogenicity, and short half life. In terms of neutralizing Wnt antagonist such as with Dkk-1 antibodies, preclinical models have shown an increase in BMD, trabecular bone volume, osteoblast surface, and bone formation in rodents [Grisanti et al. 2006]. Currently, there is no information on their value for the treatment of osteoporosis.

Finally, recent evidence [Duque et al. 2011a] suggests that IFNγ, a protein that is produced by MSCs in the bone microenvironment, could be used as an anabolic treatment for osteoporosis at low doses. Oophorectomized and aged mice treated with IFNγ showed a significant gain in bone mass, which was mostly dependent on bone formation. Considering that IFNγ is currently used as a treatment for other diseases such as hepatitis C and osteopetrosis, this compound could become a promissory alternative as a bone anabolic in the near future.

Conclusion

Age-related bone loss is a complex and heterogeneous disease. A combination of genetic, hormonal, biochemical, and environmental factors underlie its pathophysiology. The result is a decline in bone quantity and quality that increases fracture risk in a progressive manner. Despite greater understanding of the mechanisms of these contributing factors through clinical and animal studies, more research is needed to determine the relative contributions of each of these factors in order to improve preventative and therapeutic options. In addition, despite the availability of an armamentarium of agents, the optimal agent remains a challenge.