Introduction

Osteoporosis is a common skeletal disease characterized by low bone mineral density (BMD) and poor bone quality that reduces bone strength and increases the risk of fractures.1 It is a global public health concern that affects more than 75 million people in the USA, Europe, and Japan, resulting in more than 8.9 million fractures annually worldwide.2 In the USA, there are about 44 million people with osteoporosis or low bone mass (osteopenia) who are at increased risk for fracture.3

A white woman over the age of 50 years has a 50% lifetime risk of experiencing a fragility fracture.4 Fractures of the spine and hip are associated with disability, loss of independence, and increased risk of death. It is estimated that about 50% of patients with hip fractures will never again walk unassisted and that about a quarter will need long-term medical care.5 The increase in mortality 5 years after a hip fracture or clinical spine fracture (one that is clinically apparent, representing about one-third of the total number of vertebral fractures) is approximately 20%.6 The burden of osteoporosis on health-care delivery systems is high: in the USA, osteoporotic fractures are associated with over 432,000 hospitalizations, almost 2.5 million medical office visits, and 180,000 nursing home admissions each year.4 The cost of fractures was nearly $17 billion in the USA in 2005,7 and was estimated to be €31.7 billion in Europe in 2000.8

The WHO has identified osteoporosis as a major public health concern, due to its high prevalence and the serious consequences of osteoporotic fractures.2 At a particularly high risk for osteoporosis are postmenopausal, estrogen-deficient women, who typically have a high rate of bone remodeling with bone resorption exceeding bone formation. The result is bone loss, skeletal fragility, and an increased risk of fractures.9 Excellent methods are now available for measuring BMD in clinical practice,10 assessing fracture risk,11 and treating appropriate patients with pharmacological agents to reduce their fracture risk.12 Therapeutic agents for the treatment of postmenopausal osteoporosis act by correcting the imbalance between bone resorption and formation. 'Conventional' drugs (Table 1) currently approved for the treatment of osteoporosis are generally placed in one of two broad categories: antiresorptive or osteoanabolic. Antiresorptive drugs primarily act by reducing bone resorption, while osteoanabolic drugs primarily act by increasing bone formation (Figure 1). However, due to the 'coupling' of bone resorption and formation (Box 1), the changes in resorption and formation are generally in the same direction, albeit with variations in the time of onset, magnitude, and duration of these changes.13 Antiresorptive drugs include bisphosphonates (such as alendronate, risedronate, ibandronate, and zoledronate), raloxifene, and salmon calcitonin. Parathyroid hormone (teriparatide [PTH1–34] or full-length PTH1–84) given as a daily subcutaneous injection is osteoanabolic. Strontium ranelate might have properties that are both antiresorptive and osteoanabolic.14

Table 1 Conventional drugs for the treatment of postmenopausal osteoporosis*
Full size table
Figure 1: A conceptual view of bone remodeling on the surface of trabecular bone in health, disease, and in response to drug treatment.
figure 1

a | In a healthy premenopausal woman, bone resorption and formation are approximately equal, with stable bone mass and good bone strength. b | In a woman with untreated postmenopausal osteoporosis, bone remodeling is accelerated with an imbalance of bone resorption greater than bone formation, resulting in bone resorption pits that are more numerous and larger than in the premenopausal woman. There is thinning of the trabecular struts with loss of bone strength and increased risk of fractures. c | During treatment with an antiresorptive drug such as alendronate, the bone remodeling rate is reduced, with fewer and smaller bone resorption pits. Mineralization of bone is increased, with stabilized or increased bone density, improved bone strength, and reduced fracture risk. Trabecular and cortical thickness and microarchitecture are unchanged. d | During treatment with an osteoanabolic drug such as teriparatide, the bone remodeling rate is increased, with numerous large bone resorption pits. However, bone formation exceeds bone resorption, with some new bone sometimes exceeding the amount of bone removed during bone resorption and new bone in some locations where there has been no bone resorption. Bone strength in improved, trabecular and cortical thickness might increase, lost bone microarchitecture might be restored, and fracture risk is reduced.

Full size image

Despite the availability of these therapeutic agents, osteoporosis remains a disease that is underdiagnosed15 and undertreated,16 with those patients for whom treatment is started often failing to take it correctly or for long enough to achieve the expected benefit.17 This 'treatment gap'—the difference between the number of patients who could benefit from treatment and those who actually receive it18—has created the need for better strategies to reduce the global burden of osteoporotic fractures.

One potential approach to improving osteoporosis care is the development of pharmacological agents with improved therapeutic profiles (such as better risk:benefit ratio or more-convenient dosing regimens), with the hope of increasing adherence to therapy and improving clinical outcomes compared with traditional choices. As osteoporosis is a disorder of bone remodeling, advances in our understanding of the molecular regulators and mediators of this process have led to new opportunities to modulate bone remodeling and increase bone strength. The possibility of partially or totally uncoupling bone resorption and formation might offer particular benefits in managing osteoporosis. However, there are many challenges in translating basic science concepts to clinically useful drugs: these include the difficulties in recruiting clinical trial participants for large placebo-controlled trials when effective and inexpensive generic drugs are widely available; ethical concerns in randomizing some high-risk patients to a placebo group; and the high cost of drug development in a health-care delivery environment with limited financial resources and competing priorities.

This Review provides an overview of the bone remodeling process, promising new targets for therapeutic intervention (Table 2), and data supporting the development of drugs that interact with these targets. While this article focuses on data from women with postmenopausal osteoporosis, it is likely that the principles and mechanisms also apply to aging men.

Table 2 Novel targets for intervention in postmenopausal osteoporosis*
Full size table

Bone remodeling

Bone remodeling occurs through the coordinated activity of three types of cells: osteoclasts, osteoblasts, and osteocytes, as described below.

Osteoclasts

Osteoclasts, originating from hematopoietic stem cells, are multinucleated cells that remove bone in discrete packets on the surface of trabecular bone and in Haversian systems (osteons) of cortical bone. Osteoclast differentiation, activity, and survival are regulated by receptor activator of nuclear factor κB ligand (RANKL), a cytokine expressed by osteoblasts and other cell types. Macrophage-colony stimulating factor (M-CSF), a cytokine that is also expressed by osteoblasts, seems to have a permissive role for the action of RANKL on osteoclast functions. When RANKL activates its receptor, RANK, located on the cell surface of pre-osteoclasts and mature osteoclasts, there is an increase in osteoclast formation, activity, and survival (Figure 2). Osteoprotegerin (OPG) is a soluble decoy receptor produced by osteoblasts that binds to RANKL, preventing it from activating RANK; therefore, it is the balance between RANKL and OPG that is the ultimate determinant of the magnitude of bone resorption. Osteoblast expression of RANKL and OPG is upregulated or downregulated by numerous growth factors, hormones, cytokines, and drugs.19

Figure 2: RANKL–RANK–OPG regulatory pathway and osteoclastic bone resorption.
figure 2

RANKL expressed by osteoblast lineage cells binds to RANK on the surface of pre-osteoclasts and mature osteoclasts, resulting in increased bone resorption via an increase in osteoclast differentiation, activity, and survival. OPG is a 'decoy receptor', also produced by osteoblasts, that binds to RANKL, preventing RANKL binding to RANK and thereby inhibiting osteoclastic bone resorption. It is the balance of RANKL and OPG that determines the ultimate rate of bone resorption. Postmenopausal osteoporosis is associated with a high rate of bone remodeling due to an excess of RANKL over OPG. Denosumab is a fully human monoclonal antibody against RANKL that binds RANKL, much the same as endogenous OPG, resulting in a decrease in the rate of bone resorption. Demineralization is the result of acidification of the resorption lacuna due to secretion of H+ ions by osteoclasts. Cathepsin K degrades type I collagen through cleavage of the N-terminal region and the triple helical structure of collagen molecules. Odanacatib decreases bone resorption by inhibiting cathepsin K. Abbreviations: OPG, osteoprotegerin; RANK, receptor activator of nuclear factor κB; RANKL, receptor activator of nuclear factor κB ligand.

Full size image

Bone resorption occurs following polarization of the osteoclast cytoskeleton and attachment to the bone surface by means of a 'sealing zone'. This is mediated through integrins (αβ heterodimers with long extracellular and single transmembrane domains), principally αVβ3, resulting in a self-contained compartment between the bone surface and the 'ruffled border' of the undersurface of the osteoclast.20 The key role of αVβ3 has been demonstrated in a β3 knockout mouse that has a high BMD phenotype due to impaired osteoclast function.21 The ruffled border creates an acidic microenvironment that dissolves bone mineral, and produces proteases, principally cathepsin K, that degrade the protein matrix of bone (Figure 2). Apoptosis of the osteoclasts marks the end of the resorption phase and initiation of a reversal phase, followed by osteoblastic bone formation.

Osteoblasts

Osteoblasts are bone-forming cells derived from mesenchymal stem cells located in the skeletal microenvironment. Mature osteoblasts fill the void created by osteoclasts by producing the protein bone matrix—composed of type I collagen and non-collagenous proteins, including osteocalcin, osteopontin, and osteonectin—that subsequently becomes mineralized. The principal stimulus for osteoblast differentiation is activation of the canonical Wnt–β-catenin pathway (Figure 3) by Wnt signaling proteins, which bind to Frizzled receptor family members in association with low-density lipoprotein receptor-related proteins 5 and 6 (LRP5 and LRP6). Wnt signaling is inhibited by other proteins, including Dkk1 (Dickkopf-related protein 1) and sclerostin. Calcium-sensing receptors (CaSR) in the parathyroid glands have a role in maintaining extracellular calcium levels within the physiological range by controlling the production of PTH, which in turn can have direct and indirect effects on osteoblast function. Serotonin synthesized by enterochromaffin cells in the duodenum inhibits osteoblast proliferation by binding to the osteoblast 5-hydroxytryptamine receptor 1B (HTR1B) receptor and suppressing CREB, an intracellular transcription factor.22 Interestingly, serotonin produced in the brain has opposite skeletal effects, stimulating osteoblastic bone formation by binding to the HTR2C receptor in the central nervous system to decrease sympathetic tone.23 The fate of osteoblasts that have completed the bone formation phase is to die by apoptosis, become lining cells on the bone surface, or to be transformed into osteocytes buried in the bone matrix.

Figure 3: Wnt signaling.
figure 3

Activation of the canonical Wnt–β-catenin signaling pathway occurs when a Wnt ligand binds to the osteoblast transmembrane co-receptors Frizzled and either LRP5 or LRP6. This initiates a cascade of events resulting in accumulation of β-catenin in the cytoplasm and its translocation into the cell nucleus, resulting in transcription of target genes that ultimately stimulate osteoblastic bone formation. Dkk1 and sclerostin are endogenous inhibitors of Wnt signaling and, therefore, inhibitors of osteoblastic bone formation. Monoclonal antibodies that block Dkk1 or sclerostin stimulate osteoblastic bone formation. Abbreviations: Dkk1, Dickkopf-related protein 1; LRP, low-density lipoprotein receptor-related protein.

Full size image

Osteocytes

Osteocytes are by far the most numerous (over 90–95%) of the bone cell types compared with osteoblasts (4–6%) and osteoclasts (1–2%).24 The osteocyte cell bodies are encased within the bone in lucanae, with dendritic nerve-like extensions in canaliculi throughout the skeleton; these connect osteocytes to each other, to cells on the bone surface, and beyond the bone surface into the bone marrow. Osteocytes are thought to act as “mechanostats” that initiate the bone remodeling process, perhaps in response to cell death caused by bone microcracks or other conditions (such as immobilization, postmenopausal status, or glucocorticoid therapy), resulting in the release of signaling proteins or cell–cell interactions that promote osteoclast differentiation. Osteocytes also produce sclerostin, an inhibitor of Wnt signaling, thereby inhibiting osteoblastic bone formation, and fibroblast growth factor 23 (FGF23), a “phosphatonin” that increases renal phosphate loss and inhibits renal conversion of 25-hydroxyvitamin D to 1,25-dihydroxyvitamin D. Osteocytes are, therefore, key regulators of both bone resorption and formation, and might also have a role in bone mineralization.

Novel targets for therapeutic intervention

RANKL

Soon after its recognition as the principal regulator of osteoclastic bone resorption, RANKL was considered a target for intervention in the treatment of postmenopausal osteoporosis and other skeletal diseases associated with low bone mass. Potential strategies for reducing bone resorption by downregulating RANKL activity include the inhibition of RANKL production, stimulation of endogenous OPG, and administration of exogenous OPG, soluble RANK, or anti-RANKL antibodies. Clinical drug development has focused on denosumab, a fully human monoclonal antibody against RANKL.25 By binding to RANKL, denosumab prevents the interaction of RANKL with RANK and inhibits osteoclast differentiation, activity, and survival, thereby reducing bone resorption.

The efficacy and safety of denosumab was evaluated in the FREEDOM trial, a 3-year, phase III clinical trial in 7,868 postmenopausal women with osteoporosis randomized to receive either subcutaneous denosumab 60 mg (n = 3,902) or placebo (n = 3,906) every 6 months.26 Treatment with denosumab was associated with a statistically significant 68% reduction in the risk of new vertebral fractures compared with placebo (cumulative incidence 2.3% versus 7.2%, P <0.0001), a 40% reduction in the risk of hip fractures (0.7% versus 1.2%, P = 0.036), and a 20% reduction in the risk of nonvertebral fractures (6.5% versus 8.0%, P = 0.011). No significant differences in the total incidence of adverse or serious adverse events were observed between denosumab-treated patients and those receiving placebo:26 the overall risk of death, malignancies, infections, cardiovascular events, atrial fibrillation, stroke, hypocalcemia, and delayed fracture healing was similar between groups. No fractures of the femoral shaft occurred in the denosumab group, compared to three such fractures in the placebo group. No cases of osteonecrosis of the jaw (ONJ) were observed in the first 3 years of the FREEDOM trial; however, two adjudicated cases of ONJ were reported in patients treated with denosumab in the first 2-year extension, both of which healed completely without further complications, with one of those participants continuing to receive denosumab. Significant differences in the incidence of several skin-related conditions were seen between the groups: eczema was reported in 3.0% of denosumab-treated patients compared with 1.7% of patients in the placebo group (P <0.001), and cellulitis (when reported as a serious adverse event only) was more common with denosumab than placebo (0.3% versus <0.1%, P = 0.002). Falling and concussion were less likely to occur in those treated with denosumab (4.5% and <0.1%, respectively) than those receiving placebo (5.7% [P = 0.02] and 0.3% [P = 0.004], respectively). Denosumab was approved in June 2010 for the treatment of postmenopausal women with osteoporosis at high risk of fracture.

Cathepsin K

Potential inhibitors of cathepsin K have been screened for potency and selectivity using enzyme assays with purified recombinant human cathepsin K and other related cathepsins, such as L, B, and S.27 The investigation of compounds that inhibit cathepsin K has included peptidyl aldehydes, amides, α-keto heterocycles, aliphatic ketones, and nitriles.28 Those that have advanced the furthest are balicatib, relacatib, and odanacatib. Odanacatib is currently being investigated in a large, fully enrolled (n = 16,716), phase III randomized, double-blind, placebo-controlled clinical trial of women aged ≥65 years with postmenopausal osteoporosis.29 Other cathepsin K inhibitors of potential clinical interest include ONO-5334, the subject of a completed phase II trial in postmenopausal women with osteopenia or osteoporosis,30 VEL-0230, a phase I trial of which has been completed,31 and candidate drugs MIV-710 and MIV-711.32

A randomized, double-blind, placebo-controlled, dose-ranging phase II study evaluated the effects of odanacatib in postmenopausal women with low BMD (n = 399, mean age 64.2 ± 7.8 years).33,34 After 24 months of treatment, a progressive dose-related increase in BMD was apparent, with the 50 mg weekly dose of odanacatib resulting in a 5.5% increase at the lumbar spine and a 3.2% increase at the total hip. The urinary N-telopeptide/creatinine ratio (NTX/Cr), a marker of bone resorption, decreased by 52%, while the BSAP decreased by 13% with the 50 mg dose. The decrease in bone-specific alkaline phosphatase (BSAP) levels associated with odanacatib treatment is less than what is typically seen with other antiresorptive agents, such as bisphosphonates, suggesting a partial uncoupling of bone resorption and bone formation in favor of bone formation.35

Sclerostin

Modified selected lymphocyte antibody methods have been used to generate anti-sclerostin monoclonal antibodies36,37 that were demonstrated to have osteoanabolic effects in preclinical studies.36,38,39 This suggests that anti-sclerostin antibodies might be a useful treatment of osteoporosis and other skeletal disorders. At least three pharmaceutical companies (Amgen/UCB, Novartis, and Eli Lilly) are developing monoclonal antibodies against sclerostin and another (OsteoGeneX) is reportedly developing a small anti-sclerostin molecule.40 Of these, the Amgen/UCB product, AMG 785, has advanced the furthest.

A phase I randomized, double-blind, placebo-controlled, ascending single-dose study evaluated AMG 785 in healthy men and postmenopausal women. With a single subcutaneous or intravenous dose of AMG 785, a dose-dependent increase in the serum levels of bone formation markers and a dose-dependent decrease in bone resorption markers were observed. Compared with placebo, a single subcutaneous dose (0.1, 0.3, 1.0, 3.0, 5.0, or 10.0 mg/kg) of AMG 785 increased BMD at the lumbar spine and total hip in all cohorts at all time points (days 29, 57, and 85), with the exception of total hip BMD for the 5.0 mg/kg cohort at day 29, in an approximately dose-dependent manner. AMG 785 was generally well tolerated at all administered doses;41 the most commonly reported adverse effects with subcutaneous administration of either placebo or AMG 785 were injection site erythema, back pain, headache, constipation, injection site hemorrhage, arthralgia, and dizziness, all of which were considered mild and not serious. One serious adverse effect—a case of severe non-specific hepatitis—was reported in a patient receiving AMG 785, which resolved by day 26 of the study.

A fully enrolled phase II clinical trial is currently evaluating the safety and efficacy of AMG 785 in postmenopausal women with low BMD. It was recently announced—without the release of data—that this trial showed an increase in lumbar spine BMD at 12 months with AMG 785 compared to placebo, and that it compared “positively” with two active comparators, teriparatide and alendronate.42

Serotonin

Serotonin is now recognized as a regulator—perhaps an important one—of osteoblastic bone formation. It seems to have opposite effects depending on whether it is produced in the gut or brain.22 Pharmacological inhibition of gut-derived serotonin with orally administered LP533401, a small-molecule inhibitor of tryptophan hydroxylase 1, has been shown to increase bone formation, prevent bone loss, and increase bone mass in a proof-of-principle study in ovariectomized rodents.43 These findings are consistent with the observation of low plasma serotonin concentrations in individuals with high bone mass phenotypes.22,44 Inhibitors of gut-derived serotonin might, therefore, represent a new class of osteoanabolic agents for the treatment of osteoporosis.

Nitric oxide

Nitric oxide (NO) has been identified as a signaling molecule with a role in the regulation of bone formation and resorption, as well as its other important cellular functions.45 Organic nitrates, such as isosorbide dinitrate and nitroglycerin, are NO donors that are commonly used as vasodilators in the treatment of angina pectoris. These agents were associated with beneficial skeletal effects, including increased BMD,46 decreased bone resorption markers and increased bone formation markers,47 and reduced fracture risk in a nationwide case–control study.48 A 24 month randomized, placebo-controlled trial in 243 postmenopausal women with baseline lumbar spine T-scores between 0 and −2.0 showed increased BMD and decreased bone resorption when treated with nitroglycerin ointment 15 mg, applied once daily at bedtime, compared to placebo.49 However, another randomized controlled trial, conducted over 3 years in 186 postmenopausal women with baseline T-scores between 0 and −2.5, showed no BMD benefit of daily nitroglycerin ointment 22.5 mg compared to placebo.50 The reason for the discordance in findings between these two trials is unclear, but might be partly attributable to poor adherence to therapy in the latter study. Further investigation is needed to validate the optimal dose, dosing interval, and potential skeletal benefits of nitroglycerin.

Other potential targets of interest

Inhibition of Dkk1 with fully human monoclonal antibodies (such as RH2-18 or BHQ880) diminishes Dkk1 suppression of Wnt signaling, and has been shown to stimulate bone formation in animal models of estrogen-deficiency bone loss,51 inhibit bone loss in a mouse model of rheumatoid arthritis,52 and prevent the formation of osteolytic lesions in a mouse model of myeloma.53 BHQ880 is currently being studied in a phase 1b/2 clinical trial in multiple myeloma patients.54 It has a potential role in the management of postmenopausal osteoporosis, but has not been studied in this population.

CaSR antagonists (calcilytics) represent a new class of orally administered drugs designed to stimulate endogenous PTH production in a pulsatile fashion, with the goal of mimicking or perhaps even exceeding the osteoanabolic effects of daily injectable PTH.55 Achieving a rapidly reversible spike in PTH levels and obtaining the anticipated beneficial skeletal effects has proven difficult, resulting in the discontinuation of several drug development programs that initially appeared promising. One calcilytic drug, MK5442, has been studied in a dose-ranging phase II clinical trial in women with postmenopausal osteoporosis.56

Alternative approaches to improved care

The development of new therapeutic agents with novel mechanisms of action is not the only approach to improving the pharmacological treatment of patients with osteoporosis. A delayed-release formulation of the oral bisphosphonate risedronate has been developed and recently approved,57 providing the added convenience, for some patients, of taking the drug immediately after breakfast instead of following an overnight fast. Investigational delivery systems for other currently approved drugs include an oral formulation of salmon calcitonin as an alternative to intranasal administration58 and administration of PTH transdermally,59 orally,60 or by oral or nasal inhalation61 as an alternative to daily subcutaneous injection. New selective estrogen receptor modulators (SERMs) have been studied in postmenopausal women with low BMD or osteoporosis: one of these, bazedoxifene, has been combined with estrogen as the first in a novel class of drugs termed tissue selective estrogen complexes (TSECs).62,63 Other combinations of interest have focused on daily injectable PTH plus an antiresorptive agent given sequentially (before or after)64,65,66 or simultaneously (continuously or cyclically),67,68 with the goal of enhanced skeletal benefit, improved patient convenience, or lower cost than PTH monotherapy. No studies have been sufficiently powered to evaluate the antifracture efficacy of PTH–antiresorptive combinations; however, there is good evidence that PTH therapy should be followed by an antiresorptive agent to maintain the benefit achieved with PTH.65,69

Conclusions

Advances in basic science are leading to the development of novel drugs that target newly recognized molecular regulators and mediators of bone remodeling. Some of these drugs modulate bone remodeling in ways that are not achievable with traditional osteoporosis therapies, with the potential to improve clinical outcomes in patients with this disease. As more agents with a favorable balance of benefit and risk become available for use in clinical practice, physicians will be better equipped to individualize treatment according to each patient's level of risk, comorbidities, and preferences. More-convenient dosing and/or longer dosing intervals might improve adherence to therapy and be more effective in reducing the global burden of osteoporotic fractures.

Review criteria

PubMed was searched for English-language articles published since 2005 using the following search terms: “osteoporosis” and “treatment” with “new”, “emerging”, “denosumab”, “cathepsin K”, “sclerostin”, “DKK1”, “serotonin”, “calcilytic”, “nitroglycerin”, “nitric oxide”, “SERM”, and “TSEC”. When appropriate, a search for registered clinical trials was conducted at www.clinicaltrials.gov. Additional references obtained from the reference lists of review articles were examined. Abstracts and oral presentations from major scientific congresses since 2005 were considered. Press releases since 2008 from pharmaceutical companies with osteoporosis drugs in development were also searched. Targets for intervention and drugs that interact with them were included in this non-systematic review according to their potential for clinical applications over the next 5 years.