The bone remodeling unit (BRU) comprises a well-choreographed sequence of events, during which osteoclasts resorb bone during a period of about 3 weeks, creating resorption cavities that are collectively termed the remodeling space. Resorption is followed by osteoblast activation and formation of osteoid, which fill the cavities in a period of about 3 months. Primary and secondary mineralization follow ( Fig. 1 ). When active matrix synthesis is finished, osteoblasts become embedded in the matrix and function as osteocytes ( Fig. 2 ). These cells remain active in bone remodeling by maintaining connections to the bone surface, the BRU, and to other osteocytes via an extensive canalicular network. Fluid flows through this network and is believed to be able to induce signaling that allows osteocytes to function as mechanoreceptors that direct remodeling to areas that require repair and are important regulators of mineralization. Bone remodeling is an active and dynamic process and interventions that limit resorption (antiresorptive therapy) or augment formation (anabolic therapy) allow for effective treatment of bone loss and low bone mass (LBM).

The sequence of bone remodeling in healthy individuals. Remodeling is initiated when osteoclasts are activated, resorb bone, and create resorption cavities. Resorption is followed by osteoblast activation and formation of osteoid, which fills in the resorption cavity. Primary and secondary mineralization follow and continue for years. The active and dynamic process of bone remodeling is the target for pharmaceutical agents that affect bone resorption and formation. Ob, osteoblast; Oc, osteoclast.

Marker expression osteoblast to osteocyte ontogeny. Osteocytes are likely descendants of matrix-producing osteoblasts. Some osteoblasts embed in the matrix and extend projections that develop into the canalicular system. This system is critical for osteocyte communication with bone cells and for mechanoreceptor function. The signals that are important for osteoblast and osteocyte development are shown as well as secretory products of osteocytes including the Wnt inhibitor sclerostin, which is an exclusive product of osteocytes.

( From Bonewald L. The amazing osteocyte. J Bone Miner Res 2011;26:231; with permission.)

Treatment with antiresorptive agents eventually leads to a decrease in osteoblast function. The initial increase in bone mass resulting from the use of this type of therapy occurs because of inhibition of bone resorption by osteoclasts while osteoblasts continue to function and fill in the remodeling space. This uncoupling of formation and resorption is time limited (1–2 years) before osteoblast function declines and increases in bone mass begin to slow. Subsequent increases in bone mass are largely related to an increase in mineralization density, a result of reduced bone turnover and aging bone. The only known anabolic agent, teriparatide (TPTD), is limited to 24 months in the United States and 18 months in Europe because of the development of osteosarcoma in an animal toxicology study that resulted in discontinuation of the clinical trial after a mean treatment duration of 19 months. Since its release in 2002, TPTD does not seem to be associated with the development of osteosarcoma in humans. There have been case reports of osteosarcoma occurring in patients treated with teriparatide. Based on the numbers of patients treated with TPTD and the background incidence of osteosarcoma in the population, more than 2 cases would be expected.

Research is currently focusing on drugs that target the remodeling cycle by affecting osteoblasts, osteoclasts, and osteocytes, and/or molecules that control signaling pathways important for cell function and gene transcription. The role of serotonin in control of bone mass is emerging. This article reviews the types and modes of action of therapies that are in development for the treatment of patients with LBM.

Antiresorptive agents

Bisphosphonates are the most frequently used antiresorptive agents. These drugs bind avidly to hydroxyapatite and work by inhibiting farnesyl pyrophosphate synthase, an enzyme in the mevalonate pathway. This pathway is important for protein prenylation, the attachment of lipids to proteins, which is critical for cytoskeletal organization in osteoclasts. Inhibition of the pathway disrupts cytoskeletal structure, which prevents osteoclasts from being able to form a ruffled border on the bone during the remodeling process, generate a proton gradient, and resorb mineral and matrix.

Bone resorption and formation are tightly coupled. As described earlier, inhibition of resorption eventually results in inhibition of formation. An agent that inhibits bone resorption but allows bone formation to continue would, therefore, have a greater effect on bone mass and bone quality than currently available agents. This uncoupling of formation and resorption occurs in mice and humans with an osteopetrosis phenotype characterized by the absence of the chloride 7 channel, a membrane complex that is required to generate an acid milieu in the resorption lacunae. There is ongoing bone formation despite the absence of functional osteoclasts and reduced bone resorption.

Cathepsin K Inhibitors

Osteoclasts are specialized cells that effect bone resorption. Bone resorption requires dissolution of the mineral components and removal of organic bone matrix. Demineralization requires acid secretion by osteoclasts into resorption lacunae, whereas matrix degradation is accomplished by cysteine proteases including cathepsins ( Fig. 3 ). Eleven cathepsins have been identified. Cathepsin K, a cysteine protease produced by osteoclasts, is the most important because it has potent collagenase activity in acidic environments such as the resorption lacunae. Pycnodysostosis is a rare disease characterized by high bone mass (HBM) acroosteolysis of distal phalanxes, short stature, and cranial deformities. It results from a genetic mutation resulting in loss of function of the cathepsin K gene. Elimination of cathepsin K in osteoclasts results in inhibition of bone resorption. Drugs that inhibit cathepsin K are suggested to have less effect on osteoclast-osteoblast interaction because they do not result in osteoclast apoptosis as do bisphosphonates. Human cathepsin K inhibitors have been shown to prevent bone loss in ovariectomized mice without blunting the anabolic action of parathyroid hormone (PTH). Odanacatib inactivates the proteolytic activity of cathepsin K, is selective for cathepsin K, and does not result in accumulation of abnormal collagen in fibroblasts as seen with other cathepsin inhibitors. Cathepsin inhibitors other than odanacatib seem less specific for cathepsin K and have been associated with skin reactions. This finding has resulted in suspension of drug development of all cathepsin K except for odanacatib. Inhibition of cathepsin K in humans by this orally bioavailable agent is being evaluated in several ongoing trials. The 24-month results of a randomized, controlled trial (RCT) of 399 postmenopausal women with T-scores between −2.0 and −3.5 at the lumbar spine or hip sites has been reported. Four doses of odanacatib given as a weekly oral dose were evaluated and showed dose-dependent increases in bone density of the spine (+5.7%), total hip (+4.1%), femoral neck (+4.7%), and radius (+2.9%). Although urine N -telopeptide of type I collagen, a marker of bone resorption, declined 52%, bone-specific alkaline phosphatase, a marker of bone formation, declined only 13% (decline with placebo was −3%). This finding suggests less inhibition of bone formation than is found with current antiresorptive therapies. The drug was generally safe and well tolerated with no dose-related trends in adverse events including rashes. An extension study to 36 months enrolled 189 patients and compared the 50-mg weekly dose with placebo. Lumbar spine density increases from baseline and from year 2 to year 3 were 7.9% and 2.3%, whereas total hip increased 5.8% and 2.4% respectively ( Fig. 4 ). For those continuing therapy, markers of bone formation remained near baseline. After discontinuation of odanacatib, bone density declined at all sites and at year 3 lumbar spine density was only +1.4% more than the baseline value at study initiation, total hip density was −0.5% less than baseline. This decline was more rapid than occurs after discontinuation of bisphosphonate therapy. Markers of bone resorption increased to more than 50% more than baseline after discontinuation of odanacatib but returned to baseline by month 36 ( Fig. 5 ). Markers of bone formation also increased. This drug, like the recently released denosumab, has what could be termed a rapid resolution of effect, whereas bisphosphonates have a more prolonged resolution of effect. Rapid resolution of effect might be preferable in some clinical settings including the prevention of a blunting effect of the first therapy when subsequent treatment is initiated, if long term side effects are a concern, or in women of childbearing potential. However, interruptions in therapy and poor compliance may have a more rapid loss of bone mass and fracture effect. The fracture trial results for odanacatib are expected in 2012 ( clinicaltrails.gov NCT00529373).

Maturation of precursors to osteoclasts and the intracellular pathways in osteoclast function. Odanacatib inhibits the formation of cathepsin K and reduces matrix degradation.

Mean percent change from baseline more than 3 years for bone mineral density (BMD) at the lumbar spine ( A ) and total hip ( B ) with 3 years of odanacatib and 2 years of odanacatib and 1 year of placebo.

( From Eisman J, Bone HG, Hosking DJ, et al. Odanacatib in the treatment of postmenopausal women with low bone mineral density: three-year continued therapy and resolution of effect. J Bone Miner Res 2011;26:246; with permission.)

Biochemical marker of bone resorption, urine NTx expressed as geometric mean change from baseline during 3 years of odanacatib and 2 years of odanacatib and 1 year of placebo.

( From Eisman J, Bone HG, Hosking DJ, et al. Odanacatib in the treatment of postmenopausal women with low bone mineral density: three-year continued therapy and resolution of effect. J Bone Miner Res 2011;26:247; with permission.)

Glucagonlike Peptide 2

Glucagonlike peptide is an intestinal hormone released in response to food intake. Bone remodeling has a circadian rhythm with a nocturnal increase in bone resorption. This circadian rhythm is affected by food intake. The circadian variation seen in humans, high in the morning and low in the evening, may not be an inherent mechanism, but may be the result of fasting or food intake. Glucagonlike peptide 2 (GLP-2) is a polypeptide released from the intestinal mucosa after food intake. Treatment with GLP-2 at bedtime results in a significant reduction in bone resorption that normally occurs overnight. GLP-2 does not reduce bone formation, as shown by osteocalcin levels. A 120-day phase 2 trial in 160 postmenopausal women given GLP-2 resulted in an increase in hip bone density, a reduction in the nocturnal increase in carboxy-terminal collagen crosslinks with no effect on osteocalcin. If this pattern were sustained, GLP-2 would have an advantage compared with available antiresorptive agents that decrease bone formation.

Nitrates

Both the Study of Osteoporotic Fractures (SOF) and the Canadian Multicenter Osteoporosis Study (CaMOS) showed small increases in bone mass and reduction in fractures in nitrate users. In animals and humans, increases in markers of bone formation and decreases in markers of bone resorption have been shown with nitrate therapy. A case control study in Denmark reported a 15% reduction in hip fractures in patients on nitrates. The NOVEL trial (Nitroglycerine as an Option: Value in Early Bone Loss), a 3-year RCT that compared placebo with nitrates in early postmenopausal women, showed no bone mineral density (BMD) increase. A 24-month RCT trial of once daily nitroglycerine (NTG) ointment, 15 mg/d given at night, versus placebo enrolled 243 women with BMD with a T-score between 0 and −2.0. Compared with placebo, subjects on NTG had increases in BMD ( P = .001) in the lumbar spine (6.7%), femoral neck (7.0%), and total hip (6.2%) at 24 months ( Fig. 6 ). NTG significantly ( P <.05) increased trabecular density (11.9%), cortical density (2.2%), cortical area (10.6%), cortical thickness (13.9%), periosteal circumference (7.4%), polar moment of inertia (7.3%), and polar section modulus (10.7%). Markers of bone resorption (NTx) declined, whereas a marker of bone formation (bone-specific alkaline phosphatase) increased ( Fig. 7 ). This uncoupling of bone remodeling is unlike typical antiresorptive agents. Headache was the most common side effect. This drug is inexpensive and further evaluations are ongoing.

Percent change in lumbar spine and total hip BMD by dual-energy X-ray absorptiometry over 2 years in subjects treated with nitroglycerin or placebo.

( From Jamal SA, Hamilton CJ, Eastell R, et al. Effect of nitroglycerin ointment on bone density and strength in postmenopausal women: a randomized trial. JAMA 2011;305:803; with permission.)

Percent change in bone-specific alkaline phosphatase (marker of bone formation) and urine N-telopeptide over 2 years in subjects treated with nitroglycerin or placebo.

( From Jamal SA, Hamilton CJ, Eastell R, et al. Effect of nitroglycerin ointment on bone density and strength in postmenopausal women: a randomized trial. JAMA 2011;305:805; with permission.)