Antiresorptives: Safety Concerns—Clinical Perspective

Abstract

Antiresorptive drugs, such as amino-bisphosphonates and denosumab (Dmab), have dominated osteoporosis therapies for over 20 years. Since osteoporosis is a chronic disease, antifracture therapy could continue for the rest of a patient’s life. Phase III clinical trials for antiresorptive drugs assessed relatively small patient populations for short durations and excluded up to 80% of patients who might seek osteoporosis therapy in clinical practice. Postmarketing reports based upon millions of patient-years and long-term (>5 years) clinical administration have associated some previously unknown, rare adverse events with antiresorptive use including osteonecrosis of the jaw (ONJ) and atypical femur fractures (AFFs). In the osteoporosis patient population, who receive much lower doses of bisphosphonate (BP) or Dmab, the incidence of ONJ is estimated at 0.001% to 0.01%, which is only slightly higher than that seen in the general population. AFFs are insufficiency or fissure transverse fractures originating on the lateral cortex of the subtrochanteric or diaphyseal region of the femur becoming oblique as they progress medially when complete. Incidence rates of AFF range from 1.8/100,000 per year with a 2-year BP exposure to 113/100,000 per year with BP exposure from 8 to 9.9 years. Most recent pathogenic hypotheses of these rare events will be discussed.

Introduction

Antiresorptive drugs, such as amino-bisphosphonates and more recently denosumab (Dmab), have dominated the landscape of osteoporosis therapies for the last 2 decades. Since osteoporosis is a chronic disease, antifracture therapy could conceivably continue for the rest of a patient’s life. The phase III clinical trials for antiresorptive drugs assessed relatively small patient populations for short durations and excluded up to 80% of patients who might seek osteoporosis therapy in actual clinical practice. Observational register-based studies and postmarketing reports based upon millions of patient-years and long-term (>5 years) clinical administration have associated some previously unknown, rare adverse events with antiresorptive use—including osteonecrosis of the jaw (ONJ) and atypical femur fractures (AFFs; Khan et al. 2015; Shane et al. 2010, 2014). Clinical characteristics and pathogenic hypotheses of these rare events will be reviewed.

Antiresorptive Drugs: Bisphosphonates (BPs) and Dmab

Nitrogen-containing (amino-) BPs impair osteoclast function by inhibiting farnesyl pyrophosphate synthase (FPPS), thereby preventing the prenylation of small guanosine triphosphate hydrolase (GTPase) proteins (Rogers et al. 1999; Fleisch 1998). This results in disruption of cytoskeletal organization (Sato et al. 1991), loss of the ruffled membrane border (Sato et al. 1991), and altered vesicular trafficking (Rogers 2003). In addition, while osteoclasts naturally undergo apoptosis, this process is accelerated in osteoclasts exposed to BPs (Hughes et al. 1995). BPs bind to bone mineral and are taken up by mature osteoclasts at sites of bone resorption (Baron, Ferrari, and Russell 2011). BPs can remain bound to bone mineral for many years; those with greater binding affinities (zoledronic acid > alendronate > ibandronate > risedronate) possess longer skeletal residency (Russell et al. 2008). Consequently, after BP discontinuation, bound BP provides residual pharmacologic action for many years (Russell et al. 2008; Cremers, Pillai, and Papapoulos 2005), in contrast to other antiresorptive therapies in which activity is quickly lost after discontinuation (i.e., Dmab, estrogen, raloxifene, and calcitonin; Bone et al. 2011; Greenspan et al. 2002; Neele et al. 2002).

Although osteoblasts are not generally considered to be a target for BP inhibition, in vitro experiments provided direct evidence that inhibition of FPPS in osteoblasts may explain the blunting of bone anabolic response to parathormone (PTH) observed after chronic exposure to BPs in rats (Gasser et al. 2006). BPs reduce prenylation of Ras-related protein 1A (Rap1A), a small G-protein participating in cytoskeletal reorganization in osteoblasts and allowing PTH to activate flat bone lining cells into cuboid, collagen-synthesizing osteoblasts (Gasser et al. 2006). Due to the inefficient uptake of BPs into osteoblasts by fluid-phase endocytosis (Thompson et al. 2006), this mechanism is only observed after chronic treatment in vivo (Gasser et al. 2006).

Dmab is a fully human monoclonal antibody that inhibits bone resorption by neutralizing with high affinity and specificity the receptor activator of nuclear factor κ-B ligand, a key mediator of osteoclast differentiation, function, and survival (Lacey et al. 1998; Yasuda et al. 1998). Unlike BPs, Dmab inhibits osteoclast activity at all developmental stages, prefusion, multinucleated, and resorbing mature osteoclast (Baron, Ferrari, and Russell 2011), leading to easier access to bone remodeling compartments in cortical bone than BPs (Zebaze et al. 2014). Dmab is not incorporated into bone; therefore, its effect on bone resorption stops after treatment discontinuation (Bone et al. 2011).

Other commercially available antiresorptive drugs (estrogen, raloxifene, bazedoxifene, and calcitonin) have not been associated with ONJ or AFF when used alone. ONJ has been reported in one patient on raloxifene who has been previously treated with an oral BP (Baur et al. 2015). AFFs have been reported in patients receiving estrogen in addition to BPs (Odvina et al. 2005; Schneider 2006).

ONJ

The International Task Force on ONJ defined ONJ as “(1) exposed bone in the maxillofacial region that does not heal within 8 weeks after identification by a health care provider; (2) exposure to an antiresorptive agent; and (3) no history of radiation therapy to the craniofacial region” (pg. 5; Khan et al. 2015). There are 3 stages of ONJ, all of which include exposed bone in the oral cavity: (1) asymptomatic with no adjacent or regional soft tissue inflammation or infection; (2) associated pain, adjacent or regional soft tissue inflammatory swelling, or secondary infection; and (3) stage 2, in addition to a pathologic fracture, an extraoral fistula or oral antral fistula or radiographic evidence of osteolysis extending to the inferior border of the mandible or the floor of the maxillary sinus (Khan et al. 2015).

The incidence of ONJ in oncology patients treated with intravenous (iv) BPs ranges from 0% to 12.2% (0 to 12,222 per 100,000) person-years and with Dmab ranges from 0% to 2.3% (0 to 2,316 per 100,000) person-years (Khan et al. 2015). The incidence of ONJ has been reviewed in an integrated analysis of 3 clinical trials comparing Dmab 120 mg subcutaneously (sc) monthly to zoledronic acid 4 mg iv monthly in the prevention of skeletal-related events (Saad et al. 2012). Overall, in 5,723 patients studied over approximately 30 months, there were 89 ONJ cases: 52 in the Dmab arms (1.8%) versus 37 in the zoledronic acid–treated arm (1.3%), but this was not statistically different.

In the osteoporosis patient population, who receive much lower doses of antiresorptives (BP or Dmab), the incidence of ONJ is very low, ranging from 0.15% (150 in 100,000) to less than 0.001% (1 in 100,000) person-years of exposure, which appears to be only slightly higher than that seen in the general population (Khan et al. 2015). In the FREEDOM extension (Dmab 60 mg sc every 6 months for up to 7 or 10 years in postmenopausal women with osteoporosis), 13 adjudicated cases of ONJ were reported, leading to an incidence of 0.05% (52 in 100,000) person-years of exposure (Bone et al. 2017).

At the present time, evidence would suggest that there is a dose–response relationship between antiresorptive use and the development of ONJ (Compston 2011), but the pathogenic mechanisms have yet to be elucidated. Suppression of bone turnover may also play a role in the development of ONJ (Allen 2007). However, low bone turnover is not characteristically seen in affected tissue from ONJ patients (Lesclous et al. 2009). Finally, the ability of serum C-telopeptide (sCTX), a bone resorption biomarker, to act as a predictive marker of ONJ risk has not been explored in a large clinical trial, and no currently available clinical data support the use of an sCTX threshold as a guide to minimize the risk of BP-associated ONJ in patients receiving antiresorptive treatment (Baim and Miller 2009).

ONJ has also been described in several patients treated for cancer with antiangiogenic agents, in particular sunitinib (Hansen et al. 2006) and bevacizumab (Guarneri et al. 2010), although in these patients other risk factors were also present. BPs are known to have antiangiogenic properties (Santini et al. 2003), and it has been suggested that these may also contribute to the development of ONJ. Dental disease is a well-established risk factor for ONJ (Marx et al. 2005), implicating infection and inflammation in the pathogenetic process. A study in a rice rat model of periodontitis (development of periodontitis promoted through the administration of a high sucrose and casein diet) comparing vehicle, alendronate, and low-dose and high-dose iv zoledronic acid showed that only high-dose zoledronic acid induced ONJ-like lesions in mandibles of rice rats after 18 and 24 weeks of treatment (Aguirre et al. 2012).

Factors associated with the development of ONJ include poor oral hygiene, glucocorticoid use, and invasive dental procedures such as dental extraction (Khan et al. 2015). Therefore, it is suggested that patients should complete any invasive dental procedures before initiating an antiresorptive (BPs or Dmab) to minimize the already small risk; however, those on treatment should not delay emergency dental procedures nor discount dental implants (Khan et al. 2015). The likelihood of developing ONJ can be minimized through the implementation of prophylactic dental assessment and active dental intervention in cancer patient (Sim Ie et al. 2015).

Atypical Femur Fracture

Atypical femur fractures (AFFs) are insufficiency or fissure transverse fractures (Ng et al. 2014) originating on the lateral cortex of the subtrochanteric or diaphyseal (shaft) region of the femur becoming oblique as they progress medially (medial spike) when complete (Shane et al. 2010, 2014). AFFs occur in femoral shaft (cortical bone), a region of high biomechanical stress, in contrast with typical osteoporotic fractures occurring in femoral neck or intertrochanteric regions (trabecular and cortical bone). AFFs initiate on the lateral cortex in a region of maximal tensile loading (Koh et al. 2011). Subtrochanteric and shaft fractures account for 4% to 10% of all femur fractures (Nieves et al. 2010); and of these, only a minority are AFFs. AFFs can be complete or incomplete, and up to two-thirds of cases are bilateral (Probyn et al. 2015). Minor features often include prodromal thigh pain (Shane et al. 2010), cortical thickening, periosteal reaction in the lateral cortex, delayed healing, comorbid conditions, and concomitant drug exposures including BPs, glucocorticoids (Giusti et al. 2011), and proton pump inhibitors (Giusti, Hamdy, and Papapoulos 2010). Presence of prodromal thigh pain should trigger X-rays of the full-length femurs and/or radioisotope bone scan to investigate for signs of AFF (Shane et al. 2010).

Concern has arisen that long-term BP use may increase the risk of these fractures through a number of putative mechanisms including a prolonged reduction in bone remodeling. Incidence rates of AFF range from 1.8 per 100,000 cases per year with a 2-year BP exposure to 113.1 per 100,000 cases per year with BP exposure from 8 to 9.9 years (Shane et al. 2014). However, only 2 AFFs (8 per 100,000 person-years) were reported with the prolonged use of Dmab (up to 7 or 10 years), a more potent antiresorptive, in treatment-naive postmenopausal women with osteoporosis: one in the long-term group during year 4 of the extension (year 7 of Dmab treatment), and one in the crossover group during year 3 of the extension (year 3 of Dmab treatment; Bone et al. 2017). In postmenopausal women with osteoporosis previously treated with oral BPs (alendronate for a mean duration of 6.2 to 6.4 years), AFF was adjudicated in 1/321 patients transitioning to zoledronic acid 5 mg iv and in 2/322 patients transitioning to Dmab over 1 year (Miller et al. 2016). Transition to a more potent antiresorptive in patients previously treated with long-term BPs could prevent clearance of accumulated BPs from bone matrix and increase the risk of AFF. This would have to be assessed in a much larger population transitioning from oral BPs to iv zoledronic acid or Dmab.

Furthermore, Schilcher, Michaelsson, and Aspenberg (2011) reported that the risk of AFF decreased 70% per year after stopping BPs, which is much quicker than the residual pharmacologic action that usually lasts for many years (Russell et al. 2008; Cremers, Pillai, and Papapoulos 2005). These data suggest that the accumulation of BPs within the bone matrix could lead to inhibition of FPPS in osteoblasts interfering with the bone anabolic action (Gasser et al. 2006). A variant p.Asp188Tyr of geranylgeranyl pyrophosphate synthase 1 gene was identified by exome analysis in 3 sisters with atypical femur fracture (Roca-Ayats et al. 2017). This mutation would be expected to severely impair enzyme activity by disrupting an Mg2+ binding site that is critical for binding of the FPPS substrate and for catalysis (Kavanagh et al. 2006). Isoprenylation of geranylgeranyl pyrophosphate (GGPP) is critical for activation of small GTPases. Indeed, GGPP/Rho signal stimulates osteoblastic proliferation.

AFFs occur also in patients who are naive to pharmacological therapy as well as other metabolic bone diseases associated with mineralization defects such as hypophosphatasia (Whyte 2009) and rickets or with high remodeling levels leading to disorganized bone matrix (woven rather than lamellar bone) such as Paget’s disease of bone (Singer 2016), all being associated with femur bowing.

The geometry of the hip and proximal femur determines in part the stresses that are experienced on the lateral aspect of the femoral cortex (Koh et al. 2011). The bilateral incidence of AFFs and similar fracture location on the contralateral femur in cases with bilateral fractures (Probyn et al. 2015) suggest a relationship between the axis of the lower extremity and risk for AFF. Asians are at higher risk of AFF and known to have different femur geometry compared to women of white race (Lo et al. 2016). These geometric factors include smaller bone size, shorter hip axis length, larger femur varus angle, and bowing, all contributing to increased tensile stresses on the lateral femur (Lo et al. 2016), and were also observed in French Canadian Caucasian women (Mahjoub et al. 2016; Morin et al. 2016).

In summary, the prolonged reduction in bone remodeling, observed in all patients receiving BPs or Dmab, alone cannot explain the pathogenesis of AFF. Bone material properties, proximal femoral geometry, and genetic predispositions might be involved individually or in combination.

While AFFs appear to be associated with BP use, this risk needs to be put into perspective. In United States from 1996 to 2007, it was estimated that “for every 100 or so reduction in typical femoral neck or intertrochanteric fractures, there was an increase of one subtrochanteric fragility fracture” (pg. 553; Wang and Bhattacharyya 2011). For high-risk individuals, the risk of an AFF is greatly overshadowed by the antifracture benefit gained from BP therapy (Brown et al. 2014).