Approach to the Patient: Pharmacological Therapies for Fracture Risk Reduction in Adults With Osteogenesis Imperfecta

Abstract

Context

Osteogenesis imperfecta (OI) is a genetic disorder characterized by increased bone fragility largely caused by defects in structure, synthesis, or post-translational processing of type I collagen. The effectiveness of medications used for fracture reduction in adults with OI is understudied and practice recommendations are not well established. Drugs currently used to improve skeletal health in OI were initially developed to treat osteoporosis. Oral and intravenous bisphosphonates have been shown to improve bone mineral density (BMD) in adults with OI and are commonly used; however, conclusive data confirming fracture protection are lacking. Similarly, teriparatide appears to increase BMD, an effect that seems to be limited to individuals with type I OI. The role of denosumab, abaloparatide, romosozumab, and estradiol/testosterone in adult OI have not been systematically studied. Anti-sclerostin agents and transforming growth factor-beta antagonists are under investigation in clinical trials.

Objective

This review summarizes current knowledge on pharmacologic treatment options for reducing fracture risk in adults with OI.

Methods

A PubMed online database search of all study types published in the English language using the terms “osteogenesis imperfecta,” “OI,” and “brittle bone disease” was performed in June 2022. Articles screened were restricted to adults. Additional sources were identified through manual searches of reference lists.

Conclusion

Fracture rates are elevated in adults with OI. Although clinical trial data are limited, bisphosphonates and teriparatide may be useful in improving BMD. Further research is needed to develop medications for adults with OI that will lead to definite fracture rate reduction.

Case Presentation

A 60-year-old woman with type I osteogenesis imperfecta (OI) was referred to a metabolic bone clinic for a consultation after a recent fracture. She reported multiple long-bone fractures during childhood but has not had any major fractures as an adult. On physical examination, she was noted to have blue sclera but had no features to suggest dentinogenesis imperfecta, hearing loss, long bone deformities, kyphoscoliosis, or clinically apparent cardiac valvular lesion or pulmonary disease.

She reached menopause at 51 years of age and used estrogen replacement from ages 51 to 57. She was treated with weekly alendronate starting at 53 years of age for a total of 6 years. The patient's bone mineral density (BMD) T-scores prior to starting alendronate were −3.8 in the spine and −2.7 in the total hip. Six years of alendronate treatment led to a 7.4% increase in patient's lumbar spine and 2.5% increase in hip BMD. Unfortunately, the patient sustained a low-trauma left scapular fracture 6 months after stopping alendronate. She would like to discuss available osteoprotective therapies for OI.

Background

OI is a rare genetic condition that has an estimated prevalence of one in 10,000 to 20,000 births (1). However, genomic sequencing studies suggest that OI may be more prevalent than previously reported (2-4). In adults, autosomal dominant forms of OI (ie, OI types I, III, and IV) are most prevalent (∼90% of all OI) and are caused by pathogenic variants in COL1A1 and COL1A2, which encode for α1 and α2 chains of type I collagen, respectively. Variants that cause haploinsufficiency of COL1A1 or COL1A2 lead to reduced quantity of structurally normal type I collagen and the milder form of the disorder, OI type I (ie, nondeforming OI with blue sclera). Missense variants that typically affect the glycine residues of the triple helical domains of type I collagen have a dominant negative effect and result in the production of collagen with abnormal structure and function. Such variants lead to the more severe forms of the disorders, OI type III (ie, progressively deforming OI) and OI type IV (ie, common variable OI). Less common autosomal recessive types of OI are the result of pathogenic variants in genes that encode for proteins involved in collagen folding, posttranslational modification or processing, bone mineralization, or osteoblast function (5).

OI can result in a wide spectrum of clinical severity, ranging from mild bone fragility to severe forms that present with numerous fractures, bone deformities, short stature, and physical disabilities. The clinical expression of OI is highly variable even in individuals with an identical collagen mutation; this can cause phenotypic variation within families that share the same mutation. The causes of skeletal fragility in OI are likely multifactorial. They include a reduction in the biomechanical properties of the bone matrix, reduced bone mass and bone size, and defects in bone microarchitecture. Remodeling abnormalities, such as increased turnover and resorption, are recognized in children with OI. Despite scant histological data, remodeling abnormalities are likely present in adults with OI as well (6). Long-bone deformities, such as bowed femurs, are common in moderate to severe OI. They may be additional causes for fractures as bowing drastically increases forces in the bones.

Since type I collagen is present in multiple tissues, OI can affect dental health, pulmonary function, the cardiovascular system, joints, and muscle strength in addition to skeletal manifestations. Thus, a multidisciplinary approach is essential in providing comprehensive care for individuals with OI.

Fractures Rates are Elevated in Adults With OI

Individuals with OI have increased skeletal fragility and high fracture risk throughout their lifetimes. Fracture rates are highest during childhood. The risks decrease in midlife while remaining higher than that of non-OI individuals. Fracture rates then rise sharply when OI patients enter older adulthood. Furthermore, women with OI have higher fracture risks in pregnancy and the postpartum period (7).

These elevated fracture rates were observed in the 1980s and further supported by a recent Danish population-based study (8). In this study, Folkestad and colleagues demonstrated that the fracture risk in both men and women with OI during midlife (age 20-54 years) was on average 6-fold higher than in the general population (9). Postmenopausal women (ages 55 years and older) with OI were at especially high fracture risk compared to non-OI individuals (incidence rate ratio 8.0, 95% CI, 5.6-11.4). A similar increase in the fracture incidence ratio was observed for males ages 45 years and older. For a variety of reasons, these fracture risk estimates are probably conservative, particularly for the more severe forms of OI.

The high fracture risk in adults with OI underscores the need for effective treatment options and evidence-based clinical practice recommendations. Here, our aim is to summarize the literature on available fracture protective pharmacological options for adults with OI and to suggest potential strategies and limitations for their use.

Therapies to Improve Skeletal Health in OI

All drug therapies currently used to improve skeletal health in OI were initially developed to treat osteoporosis. They are intended to affect bone remodeling favorably by reducing bone resorption, increasing bone formation, or both. Although these approaches do not address the fundamental pathogenetic mechanism in OI, they have been repurposed for OI with the hypothesis that beneficial bone remodeling effects would likely improve bone mass, structure, and strength. Despite the recent new understanding of the genetic causes of OI, therapies that address the genetic, cellular, and signaling abnormalities in OI are not yet available. It is also important to note that almost all therapeutic effectiveness data are based on individuals with the most common collagen-related forms of OI, types I, III, and IV.

Bisphosphonates

Efficacy of bisphosphonate therapy in adults with OI

Children with OI have accelerated bone turnover and excessive osteoclastic activity, and bisphosphonate therapy has been shown to improve bone micro- and macrostructure, bone mineral density, and quality of life (10, 11). Furthermore, some studies have shown probable reduction in fracture rates and scoliosis progression in pediatric patients (12-18). Although remodeling abnormalities are poorly characterized in adults with OI of any clinical type, the antiresorptive properties of bisphosphonate may still confer skeletal benefits for adults with OI. In fact, as we summarize later, bone density appears to be improved with bisphosphonate therapy. Although this effect is likely to be beneficial, the effects of these drugs on fracture incidence cannot be confidently derived from the available data, primarily because clinical trials lack adequate power to detect an effect on fracture risk and/or the study designs lack adequate controls.

Outcomes of 8 studies involving bisphosphonate therapy in adults with OI are summarized in Table 1. Studies that included randomized and quasi-randomized controlled trials with low risk for bias were summarized in an updated Cochrane review (18). However, only 2 of the 14 studies in the Cochrane review included adult participants (18-20). In 1 randomized, placebo-controlled trial, 38 participants with OI were treated with intravenous neridronate or placebo. The average baseline areal BMD (aBMD) T-scores in the lumbar spine and total hip were −3.5 and −2.6, respectively. Almost all enrolled individuals in the study had OI type I. After 1 year of therapy, aBMD increased by an average of 4.6% in the lumbar spine and by 3.9% in the hip, with an additional increase of 3.9% in the lumbar spine and 1.5% in the hip by 24 months of therapy (19). The second study randomized participants to daily doses of oral alendronate or placebo. The average baseline aBMD T-score was −3.8 in the lumbar spine and −2.3 in the total hip. After 36 months of therapy, participants treated with alendronate experienced a 10% average increase in the spine and a 3.3% increase in the hip (20). Of note, both randomized controlled trials were underpowered for detecting fracture risk reduction (19, 20).

Table 1.

Summary of studies examining effects of bisphosphonate therapy on spine and hip bone mineral density and fracture rates

Citation	Drug	Study design	Treatment duration	Patient N	OI types	Age range (years)	Spine BMD increase	Hip BMD increase	Fracture reduction
Adami et al (19)	IV NER q3 months	Double blinded, placebo controlled	24 months	46	I, III, IV	21-50	Yes	Yes	Yes
Chevrel et al (20)	Oral ALN 10 mg daily	Double blinded, placebo controlled	36 months	64	I, IV	Mean (SD): 36 (12)	Yes	Yes	No
Shapiro et al (23)	IV PAM 1.5 mg/kg q3-4 months, oral ALN 70 mg weekly, oral RISE 35 mg weekly.	Open label, prospective observational	Mean 52 months	90	I, III, IV	17-68	Mild OI IV: yes Oral: Yes Severe OI IV: Yes Oral: No	Mild OI IV: No Oral: Yes Severe OI IV: Yes Oral: No	Mild OI: No Severe OI: IV: yes Oral: No
Bradbury et al (26)	Oral RISE 35 mg weekly	Open label, prospective observational	24 months	32	I	18-76	Yes	No	No
O'Sullivan et al (75)	IV ZOL, oral ALN, oral RISE	Retrospective case series	2.5 years	10	I, III, IV	18-53	Yes	No	No
Xu et al (24)	IV ZOL 5 mg yearly, oral ALN 70 mg weekly	Open label, prospective, randomized	24 months	60	I, III, IV	Mean (SD): ZOL 35.5 (14.4) ALN 30.5(10.9)	IV: Yes Oral: yes	IV: Yes Oral: yes	IV: Yes Oral: yes
Viapiana et al (22)	IV NER 2 mg/kg (max 100 g) q 3 months	Open label, prospective, observational	36 months	114	I, III, IV	20.9-71.5	Yes	Yes	No
Pavon de Paz et al (21)	IV ZOL 4 mg q 6 months	Open label, prospective, observational	5.5 years	20	I, III, IV	Mean (SD): 35.2 (11.5)	Yes	Yes	No

Abbreviations: ALN, alendronate; BMD, bone mineral density; IV, intravenous; NER, neridronate; OI, osteogenesis imperfecta; PAM, pamidronate; RISE, risedronate ZOL, zoledronic acid.

Five observational studies further support bisphosphonate's positive effects on aBMD in adults with OI. While fracture risk assessment is inherently limited by the study designs, some attempted to compare fracture rates during therapy to the pretreatment time period. Two observational studies used intravenous bisphosphonates in adults with OI types I, III, and IV (21, 22). The study led by Viapiana and colleagues included 114 participants treated with neridronate every 3 months for up to 3 years. In this cohort, three-fourths of patients had OI type I, and neridronate therapy improved aBMD in lumbar spine and total hip aBMD at 36 months (∼8.5% and 4% respectively) (22). The fracture rate compared to a pretreatment interval was not significantly different. Pavon de Paz and colleagues treated 20 patients with zoledronic acid every 6 months for 3 years and annually thereafter. Zoledronic acid therapy was associated with increased aBMD in lumbar spine after 1, 3, and 5 years (6.7%, 5.7%, and 9%, respectively) and increased aBMD in total hip and femoral neck at 3 years [10.1 and 11.1%, respectively (21)].

Two observational studies examined the effectiveness of both oral and intravenous (IV) bisphosphonates (23, 24). The study by Shapiro and colleagues included 90 patients, and they received an average of 52 months of treatment. Fifty-five patients were treated with IV pamidronate or an oral bisphosphate (alendronate or residronate), while the remainder of the cohort was untreated. Approximately two-thirds of patients had OI type I. In individuals with OI type I, both IV and oral bisphosphonates were associated with an annualized increase in lumbar spine aBMD, but only oral bisphosphonates were linked to a change in the total hip. In participants with OI types III and IV, IV bisphosphonate was associated with annualized increases in lumbar spine and total hip aBMD, while oral bisphosphonate had no impact at either site. Neither IV nor oral bisphosphonates were associated with significant change in aBMD at the femoral neck in any of the OI subtypes. No fracture reduction was observed in patients with OI type I. Only IV and not oral bisphosphonate was associated with fracture reduction in adults with OI types III and IV compared to historical control data from 5 years preceding treatment (23, 25). The Xu et al study included 60 patients treated with either oral alendronate or IV zoledronic acid for 2 years in a randomized but not blinded fashion. Both oral and IV treatment groups were found to have a large increase in total hip and lumbar spine aBMD (11.7-14.7% and 10.5-11.3%, respectively) and a significant decrease in fracture rate compared to a pretreatment period (24). For unclear reasons, the large increase in hip aBMD and fracture rate reduction seen in the oral bisphosphonate group were out of proportion to the effects observed in other studies. Lastly, Bradbury and colleagues’ open-label study involving 2 years of oral risedronate treatment showed aBMD improvement in the spine (3.9%) but not in the hip (26).

Both randomized controlled trials included in the aforementioned Cochrane review showed bone turnover marker reduction in treated OI adults (19, 20). Of the 2 randomized trials, the study involving neridronate demonstrated a serum C-telopeptide (sCTX) and bone alkaline phosphatase (BAP) reduction of 25% and 20%, respectively (19). The randomized alendronate study resulted in a 60% reduction in sCTX levels (20). The observational study by Bradbury et al involving residronate, which showed an aBMD increase in the spine but not hip, demonstrated a 37% procollagen 1 N-terminal propeptide (P1NP) reduction and no changes in BAP levels (26). The study by Pavón de Paz et al, which showed increases in spine and hip aBMD after zoledronic acid, reported a significant decrease in sCTX but not BAP levels (21). Lastly, the Xu et al study, which demonstrated large increases in hip and spine aBMD in association with alendronate and zoledronic acid therapies, reported significant decreases in serum alkaline phosphatase and sCTX (30–37% and 58–64%, respectively) (24). The reasons for the discrepant changes between aBMD and bone biochemical markers are unclear, and the usefulness of biochemical markers in assessing bisphosphonate treatment effectiveness in OI remains unknown. In sum, the remodeling marker changes in the treatment trials of bisphosphonates in adults with OI indicate that there is a reduction in bone resorption and overall remodeling.

There are other unresolved issues in the use of bisphosphonate therapy in adults in OI. First, the most appropriate bisphosphonate therapy duration in adults with OI has not been ascertained. While the most appropriate duration of bisphosphonate therapy in osteoporosis has been extensively discussed (27, 28), sufficient data is not available to develop evidence-based recommendations for OI. Second, clinical trial data are limited and do not allow for an adequate comparison between IV and oral bisphosphonate. Finally, and critically, the effectiveness of bisphosphonate therapy in reducing fracture risk in adults with OI remains unsettled. Whereas some observational results suggest that fracture rates in patients treated with bisphosphonates are lower than in the pretreatment period, the 2 available randomized clinical trials were not designed to detect changes in fracture risk. In sum, while bisphosphonates appear to increase aBMD in adults with OI, the relationship between aBMD increases and fracture risk reduction in OI is unclear.

Potential complications of bisphosphonate therapy in OI

Osteonecrosis of the jaw (ONJ) and atypical fracture of the femur (AFF) are rare but serious complications of bisphosphonate therapy. At least conceptually, some factors may put individuals with OI at higher risk for these complications compared to osteoporosis patients. These include longer duration of therapy, higher frequency of bisphosphonate use, and underlying dental abnormalities from dentinogenesis imperfecta. However, the existing data do not support the presence of higher risks.

The Risk of Osteonecrosis of the Jaw in OI Adults Treated With Bisphosphonates

The proposed pathogenic mechanisms for ONJ include bone turnover suppression, reduced vascular perfusion, soft tissue trauma, chronic infection, and direct soft tissue toxicity (29). The incidence of ONJ in cancer patients treated with IV bisphosphonates can be as high as 1% to 10%. On the other hand, bisphosphonate-associated ONJ rates in patients with osteoporosis is low (0.01–0.001%) (30). Despite frequent dosing of cyclic bisphosphonate therapy in children and adults with OI, no ONJ cases have been reported in OI patients (31). This may be due to the small number of bisphosphonate-treated OI patients and the heightened ONJ awareness that prompts dental evaluations prior to and during bisphosphonate use. Although the true incidence of bisphosphonate-related ONJ in OI adults is unknown, it is unlikely to be significantly higher than that of adults treated with bisphosphonate for osteoporosis.

The Risk of Atypical Fracture of the Femur in OI Adults Treated With Bisphosphonates

The first case of AFF in an adult with OI was published in 2012. It occurred in a 75-year-old woman with OI type I who had been treated with weekly alendronate for 3 years. Radiographs revealed an atypical fracture of the right femoral shaft with thickening of the entire lateral cortex and a localized periosteal reaction at the level of the fracture. Since then, at least 4 additional case reports of AFF in OI adults after bisphosphonate use have been published (32-35). However, in the absence of control groups untreated with bisphosphonates, the link between bisphosphonates and AFF in adults with OI is unsubstantiated. In fact, a retrospective study of children with OI found that femoral fractures with atypical characteristics occurred with equal frequency in children treated and untreated with bisphosphonates (36). Similarly, Anderson et al reported that some femoral fractures with characteristics of AFF occur in adults with OI who have not been exposed to bisphosphonates and speculate that the bone deformities, collagen abnormalities, and mineralization changes that occur in OI may cause these fractures (37). Thus, adults and pediatric patients who develop AFF during bisphosphonate treatments may be due to inherent abnormality in bone quality rather than to bisphosphonate exposure.

Teriparatide

Teriparatide, a human 1-34 parathyroid hormone, has osteoanabolic and fracture protective effects in patients with osteoporosis (38, 39). A double-blind, placebo-controlled trial that included 79 patients with OI types I, III, and IV demonstrated 18 months of teriparatide treatment was associated with increases in aBMD in lumbar spine and total hip (6.1% and 2.8%, respectively). Furthermore, teriparatide therapy was associated with increase in volumetric BMD (vBMD) at the spine and an average increase in estimated vertebral strength as assessed by finite element analysis. Effects on bone remodeling markers, and the time course of the skeletal effects, were also similar to studies of teriparatide in osteoporosis. The study was not designed to assess the effect of therapy on fracture risk. Interestingly, a post hoc evaluation of OI subgroups suggested that there were positive skeletal effects in those with type I OI but not in patients with more severe forms of OI (ie, OI types III and IV). Thus, the authors of the study raised the possibility of a differential treatment response in adults with quantitative defects of collagen (OI type I) vs those with qualitative collagen defects in OI types III and IV. The reasons underlying these potential differential effects are unexplained. Because teriparatide resulted in positive effects on remodeling markers in type I but not types III/IV OI, the post-receptor signaling and resulting cellular effects of teriparatide may be blunted in type III/IV participants due to altered extracellular matrix functions. Also, the potential for transforming growth factor-beta (TGF-β)-mediated parathyroid hormone insensitivity has been hypothesized as a contributor of this observation, but rigorously powered trials are needed to test this (40).

Additional studies support the positive skeletal effects of teriparatide in OI adults. An observational study included 13 postmenopausal women with OI type I who had incurred new vertebral fractures during treatment with neridronate. These patients were switched to an 18-month course of teriparatide. The study reported a mean spine aBMD increase of 3.5% but no changes in total hip aBMD. Biochemical markers of bone turnover (P1NP, BAP, sCTX) increased significantly, and no new fractures were reported during the study (41). The authors noted that the mean spine aBMD increase in this study was much lower than the expected spine aBMD increase of 9% to 10% seen in non-OI patients treated for osteoporosis (38).

A multicenter, randomized, and double-blinded study examined the effects of teriparatide vs neridronate treatment in 98 patients with type I OI (42). All patients were 25 years of age or older, with T-scores of less than −2.5 and with 2 or more lifetime fragility fractures. The mean lumbar spine aBMD change at 2 years was 5.1% in the teriparatide group and −1.6% in the neridronate group. The number of new fragility fractures in teriparatide and neridronate groups were not significantly different (16% and 27%, respectively). The change in spine aBMD in the teriparatide group was similar to that of the aforementioned study involving postmenopausal women with type I OI (41). However, the lack of spine aBMD improvement in the neridronate group was unexpected as prior studies had shown an approximately 4% aBMD increase in OI adults treated with IV bisphosphonate (19, 23). Moreover, the high fracture rates in both treatment groups were unanticipated (42).

Although effects on bone density appear to be positive, at least in type I OI, none of the studies involving teriparatide was designed to draw confident conclusions about the effect of therapy on fracture risk. A large prospective, open-label, randomized controlled trial is underway in the United Kingdom with the primary objective to investigate fracture rates after a 2-year course of teriparatide followed by a single infusion of zoledronic acid in adults with OI as compared to usual care. The study planned to include 380 participants with 190 individuals in each treatment group (43).

Denosumab

Denosumab, an anti-RANKL antibody, is an antiresorptive medication effective in increasing aBMD at the lumbar spine and hip and in reducing fracture risk in osteoporosis. The safety and effectiveness of denosumab is surprisingly understudied in the OI population. Currently, only 1 retrospective case series including 5 adult patients has been published (44). Except for 1 patient, increases in aBMD were observed in the lumbar spine and hip and no fractures were observed during a median observational period 18 months (2-54 months).

In patients with osteoporosis, denosumab withdrawal can be accompanied by excessive osteoclastic bone resorption and rapid bone loss. Since OI may be associated with increased bone remodeling in the baseline state, stopping denosumab therapy in OI could be especially problematic. In addition, the risk of ONJ or atypical femoral fracture in OI during denosumab treatment is unknown. Without additional trial data, the usefulness of denosumab in OI is unclear.

Estrogen and Raloxifene in Women

Estrogen alone or estrogen in combination with progesterone are effective in preserving aBMD and reducing fracture risk in postmenopausal women with osteoporosis (45, 46). Yet, increased risks for breast cancer and cardiovascular disease make estrogen a less desirable option for fracture prevention in most postmenopausal women (45-47). Bone remodeling and osteoclastic resorption are likely increased in postmenopausal women with OI, resulting in further impairment of bone mass, structure, and strength. The heightened fracture risk after menopause in women with OI reinforces that notion. However, data examining the skeletal sequelae of estrogen deficiency and estrogen replacement in postmenopausal OI individuals are lacking.

Raloxifene, a selective estrogen receptor modulator, increases aBMD and reduces vertebral fracture risk in women with postmenopausal osteoporosis (48). However, raloxifene is not recommended for women with severe osteoporosis due to the lack of evidence of hip and nonvertebral fracture protection. Similar to estrogen therapy, raloxifene has not been studied in women with OI.

Testosterone in Men

Because of its potential to improve both physical performance and bone strength in hypogonadal men (49, 50), testosterone has been considered a treatment option for men with OI. Nonetheless, there are no data supporting the effectiveness or safety of testosterone supplementation in OI.

Romosozumab

Romosozumab, a monoclonal antibody that inhibits sclerostin, increases bone formation and decreases bone resorption. It is an effective and Food and Drug Administration-approved treatment for postmenopausal osteoporosis (51). Currently, only 1 case report of romosozumab use in an adult with OI has been published. This case involved a 65-year-old man with OI type I who had recurrent fractures after 1 year of alendronate treatment. Thus, he was switched to romosozumab. After completing 12 months of therapy, his lumbar spine and total hip aBMD T-scores improved from −3.9 to −3.0 and from −5.6 to −3.8, respectively. The corresponding changes in spine and hip aBMD were 22% and 136.4%, respectively (52). This magnitude of aBMD increase was unanticipated and not reported when used as osteoporosis treatment. Romosozumab is potentially an effective therapy for individuals with OI, but additional studies are needed to confirm its efficacy and safety.

Setrusumab

Setrusumab, another monoclonal antibody that inhibits sclerostin activity, was evaluated in a 12-month, randomized, double-blind, phase 2b study in adults with OI. Ninety participants were treated with 1 of 3 available doses of setrusumab. The study did not meet the primary endpoint of increasing radial trabecular vBMD measured by high-resolution peripheral computed tomography, but total radial vBMD was increased in a dose-dependent manner. At the highest dose, spine aBMD increases for type I and type III/IV OI were 8.81% and 10.38%, respectively. At the highest dose, total hip and femoral neck aBMD increased in all OI types (2.48% and 3.37% respectively). Moreover, the study showed significant increases in aBMD in the spine and total hip in all dose groups. Setrusumab injections were well tolerated with no obvious safety concerns (53).

TGF-β Inhibitors

In preclinical studies, excess TGF-β has been implicated as a key player in remodeling abnormalities observed in the more severe forms of OI (54). The inhibition of TGF-β signaling with the use of a blocking antibody was associated with improvements in bone mass, certain bone biomechanical properties, and correction of abnormal alveolar pattern of the lungs (55). A small phase 2 trial of fresolimumab, a TGF-β neutralizing antibody, was associated with decreases in bone remodeling as assessed by plasma levels of osteocalcin. In this study, a positive effect on lumbar spine aBMD was seen in participants with OI type IV but not in several type III and VIII patients (55). Larger studies are underway to further evaluate the usefulness of this approach.

Considerations for Initiating Osteoprotective Treatments in Adults With OI

For several reasons, the decision of when to begin osteoprotective medication is difficult in adults with OI. Importantly, while aBMD values are important in fracture risk estimation in osteoporosis, the predictive value of aBMD measurements in OI individuals is undefined. Due to the qualitative defect in bone material properties, conventional aBMD measurements may be less reliable in predicting fracture risk in OI. Moreover, long bones in OI are more gracile and more often bowed compared to unaffected individuals; the resulting biomechanical disadvantage may not be adequately reflected in aBMD measures. As a result, BMD thresholds for starting therapy for osteoporosis may be inappropriate when applied to patients with OI. In addition to these challenges, the presence of bone deformity or metal implants often precludes density measures at commonly used anatomical sites. Nevertheless, very low BMD (eg, T-scores ≤ −3.0) at the lumbar spine or hip probably suggest increased fracture risk and the need for drug therapy. Additional loss of bone mass in an individual with reduced baseline aBMD is also worrisome, and, if secondary causes cannot be identified and corrected, such loss in bone density should prompt consideration of pharmacological therapy.

Apart from aBMD measures, other factors might be useful in estimating fracture risk and in informing decisions to initiate pharmacological therapy in adults with OI.

• OI type. The type of OI is likely to influence the estimation of an individual's fracture risk and need for therapy. Fracture rates in adults with OI type I are generally lower than in OI types III and IV (56). Yet, while the clinical type of OI is important in understanding the natural history of patients with OI, there is considerable variability in fracture patterns within each type, and it should not be assumed that all patients of a particular type will have similar rates of fracture.

• Fracture history. Fracture history provides an important insight into future fracture rates. Generally, the incidence of fractures in OI declines after growth ceases. However, if fractures continue to occur in adulthood, or fractures begin after a period of quiescence, pharmacological therapy may be needed. In an adult with OI, a fracture in the last 2 years (unless the result of severe trauma) is commonly considered a sign of heightened fracture risk and a need for drug therapy.

• Age. Fracture risk in OI increases with age, so the benefits of therapy may be greater in an older patient; postmenopausal women are at particular risk.

• Muscle mass. Muscle mass and physical performance has been reported to be reduced in people with OI (57-60). Thus, decreased muscle mass and weakness can lead to more frequent falls and fractures risks in OI individuals.

• Falls risk. People with more severe forms of OI are frequently affected by deformities that limit mobility and may change the likelihood of, and the types of, trauma that could precipitate a fracture. Instituting pharmacological therapy may be appropriate in patients that are at higher risk of falls and therefore fractures.

• Comorbid conditions. Comorbid conditions (eg, rheumatoid arthritis) or medications that adversely affect bone (eg, glucocorticoids) can add to skeletal fragility, increase fracture risk, and increase the need for therapy.

What Drugs to Consider?

If, on the basis of the considerations here, a clinical judgement is made that the risk of future fracture is high in an adult with OI, consideration should be given to instituting pharmacological therapy. The following considerations may help to decide which of the therapeutic options may be appropriate.

Bisphosphonates

Bisphosphonates are the most studied and most commonly used medication for OI. Substantial evidence supports the effectiveness of bisphosphonates in improving aBMD, but it remains unknown if the increased aBMD translates into fracture protection. Regardless, bisphosphonates should be considered the most appropriate option when initiating therapy in most adults with OI. The risks of ONJ and AFF in adults with OI adults do not appear to be higher than rates observed in patients without OI. IV bisphosphonates generally have better adherence rates and may be more effective in increasing aBMD in OI. While oral bisphosphonates are feasible alternatives based on patient preference, individuals with OI may have higher incidences of gastroesophageal reflux disease and gastrointestinal symptoms at baseline, which may be exacerbated by oral bisphosphonates (61). Furthermore, scoliosis and an inability to stay upright may interfere with oral bisphosphonate administration. The most appropriate frequency of IV bisphosphonate therapy, duration of therapy, and the usefulness of a drug holiday are undefined in OI. It is reasonable to utilize guidelines developed for osteoporosis treatment and tailor them for individuals with OI based on thoughtful clinical assessments (27, 62).

Teriparatide and Abaloparatide

Teriparatide therapy has been reported to improve bone mineral density and parameters of vertebral bone strength in adults with type I OI. Although there are not adequate data to conclude that teriparatide reduces fracture risk, teriparatide may be an attractive treatment alternative for type I OI adults with high vertebral fracture risk or intolerance to bisphosphonates. Similar to its use in osteoporosis treatment, a duration of therapy of 18 to 24 months followed by an antiresorptive agent appears prudent. The usefulness of abaloparatide has not been specifically studied in OI, but conceptually there is no reason to consider it less effective than teriparatide (63, 64).

Denosumab

Currently, no data are available to support the use of denosumab in the treatment of OI. Nevertheless, denosumab may be an option for OI adults with renal impairment, medication intolerance, or poor therapeutic response to bisphosphonate and/or teriparatide. The increase in bone resorption and resulting bone loss after denosumab discontinuation is particularly concerning in OI patients. In that situation, the use of a potent bisphosphonate (eg, zoledronic acid) would be appropriate, with careful monitoring of aBMD, and possibly biochemical markers of bone resorption, to aid in bisphosphonate dosing frequency.

Estrogen in Postmenopausal Women

Although estrogen deficiency is likely detrimental to bone health in postmenopausal women with OI, the usefulness of estrogen replacement for fracture protection in this scenario is unknown. Similar to its use in osteoporosis, estrogen is not generally recommended as primary treatment for preserving BMD in OI due to potential adverse effects. However, estrogen remains a sound option for postmenopausal OI women with concurrent vasomotor symptoms and who are at low risk for breast cancer and cardiovascular disease.

Raloxifene

Raloxifene is not recommended as a primary therapy for postmenopausal women with significant fracture risk. Also, there is no information concerning the safety or efficacy of raloxifene in OI. Nonetheless, raloxifene might be considered an alternative approach in women with OI who have low deep vein thrombosis risk, high breast cancer risk, and intolerance to other osteoprotective medications (27).

Testosterone in Men

Current Endocrine Society guidelines suggest that testosterone supplementation should be considered in men with low testosterone levels and with clinical symptoms of hypogonadism (62). In those men, testosterone results in improvements in bone structure and density (49, 50), but the effectiveness or safety of testosterone use in men that do not meet these criteria are unclear. Testosterone replacement has not been shown to reduce the risk of fractures. Although the effects of testosterone treatment on bone in OI have not been studied, gonadal function and sex hormone responsiveness appear to be normal in men with OI. Therefore, there are no a priori reasons to anticipate that testosterone treatment considerations and general guidelines should be different for men with OI.

Sclerostin Inhibition

Romosozumab is clinically available for treating severe osteoporosis. Even though no trials of romosozumab in OI are available, positive preliminary results with setrusumab suggest that inhibition of sclerostin may be useful. Off-label use of romosozumab could be considered in adults with OI who are at high risk of fracture and not candidates for bisphosphonate or teriparatide therapy.

Assessing the Response to Therapy in Adults With OI

The most appropriate way to assess osteoprotective therapy response in adults with OI is unclear. Nevertheless, several approaches are reasonable to consider.

Bone Mineral Density Measurements

An increase in aBMD is expected with successful use of any of the treatment options available. Regular (every 1-2 years) aBMD measurements during therapy should be useful with appropriate consideration of the “least significant change” required to detect BMD change with confidence (65). More frequent testing may be appropriate in some patients. A fall in BMD during therapy ought to prompt a clinical evaluation for additional factors deleterious to bone health and to ensure medication adherence. If a remediable explanation for bone loss is not apparent, a change in therapy should be considered. In a patient with declining BMD, stabilization of BMD after therapy initiation might be considered successful. Areal BMD with DXA is conventionally used to examine change, although in some patients with OI the presence of previous fracture and/or deformity may make the common measurement sites (spine, hip, and radius) difficult to assess. Recently, high-resolution peripheral quantitative computed tomography has been used in research settings to gauge the effects of interventions in OI, but there is limited experience with its use, and it is not routinely available in clinical settings.

Bone Turnover Markers

The International Osteoporosis Foundation and European Calcified Tissue Society recommend using bone turnover markers P1NP and collagen type I C-terminal telopeptide (CTX) to assess osteoporosis treatment effectiveness (66). More specifically, adequate CTX suppression with bisphosphonate therapy is associated with greater aBMD increases and lower fracture risk (67). Although the use of biochemical markers has not been formally evaluated in OI, available clinical trial data in OI suggest similar bone turnover marker responses to therapeutic interventions.

Fracture Incidents

A fracture occurrence during therapy should trigger an evaluation of factors that contribute to fracture risk and consideration for alternative treatment options. For instance, an incident vertebral fracture should prompt consideration of drugs that are especially potent in increasing vertebral strength (teriparatide, abaloparatide, romosozumab). As has been suggested for osteoporosis, repeated spine x-rays may be warranted if the patients status changes in a way that suggests a new fracture (documented height loss, undiagnosed new back pain, postural change, or a finding of new vertebral deformity on chest X-ray) (68).

Goal-directed Approach

A goal-directed approach to osteoporosis therapy has been proposed (69). Major goals are to increase aBMD to above an accepted threshold (eg, T score > −2.5) and to reduce the 10-year fracture risk to less than 3% and 20% for hip fractures and major osteoporotic fractures, respectively, using the US-adapted World Health Organization absolute fracture risk estimator (FRAX). The use of a goal-directed approach is equally attractive in OI, but appropriate aBMD and FRAX thresholds have not been established for OI.

Other Considerations

Considerations for all fundamentally important fracture-reducing strategies are critical. This includes optimal nutrition, physical therapy, and activity as tolerated or feasible and minimization of fall risk during transfer, ambulation, or exercise. Calcium and vitamin D nutrition should be consistent with recommendations for adults in general with adjustment for size and weight in adults with OI (70). While the benefits of vitamin D supplementation in the general population have been questioned (71, 72), some data suggest that vitamin D deficiency is more prevalent in OI (73, 74). Thus, it is reasonable to assess vitamin D status and supplement as needed in OI individuals.

Limitations of Current Knowledge

There is a dearth of information regarding fracture protective effects of pharmaceutical interventions for adults with OI. The available information is summarized here, and many important questions remain unsettled. Most studies used bone density and remodeling marker measurements and lacked statistical power to detect carefully adjudicated fracture incidents.

Furthermore, questions about long-term treatment effectiveness and safety are unanswered due to the short duration of available studies. Almost all studies have been in young or middle-age adults, leaving unresolved concerns about treatment effectiveness in older OI adults. Lastly, inadequate information exists regarding special situations such as fracture protection in pregnancy and the peripartum period and in individuals with recessive forms of OI. Clearly, additional research is needed to fully understand and optimize osteoprotective drug therapies for adults with OI.

Back to Our Patient

At the time of initial consultation, the patient did not want to continue therapy with a bisphosphonate or another agent despite her recent scapula fracture. A few years later, multiple vertebral compression fractures were identified radiographically. The patient was referred back to the metabolic bone clinic due to her elevated risk for additional fractures. After weighing the risks and benefits of each available osteoprotective medication, she elected to begin a course of teriparatide.

Conclusions

OI is a rare genetic condition primarily caused by defects in collagen synthesis and/or function. OI has varied phenotypic manifestations involving skeletal health as well as the function of other major organ systems. Increased fracture risk is a prominent problem in adults with the disorder and merits concerted attention. To date, high-quality clinical trial data are limited, and no Food and Drug Administration-approved drug treatments are available to reduce fracture risk. Nevertheless, bisphosphonates remain as first-line therapy for adults with OI while anabolic agents represent potential alternatives for selected patients. Additional agents are under active investigation. Treatments should be tailored to the individual patient with OI and should include both pharmacological and nonpharmacological measures to reduce the risk of fractures.

Funding

The work was made possible with funding by the Rare Diseases Clinical Research Network Brittle Bone Disease Consortium (Grant #1U54AR068069–0 from the National Center for Advancing Translational Science).

Data Availability

Data sharing is not applicable to this article. No datasets were generated or analyzed for this review.

Disclosures

W.L. has nothing to disclose. B.L. reports research support from Sanofi and Genzyme. S.C.S.N. has nothing to disclose. L.N. reports consulting with Ultragenyx. F.R. reports consulting with Ibsen and Sanofi; and research support from PreciThera, Ultragenyx, and Catabasis. E.T.R. reports consulting with Alexion, Aruvant, Ascendis, Biomarin, Inozyme, Ipsen, and Ultragenyx and research support from Alexion and Ultragenyx. V.R.S. reports consulting with Aceragen and research support from Ultragenyx and Arcturus. E.O. reports consulting with Amgen, Angitia, Bayer, Biocon, NextCure, Quince Therapeutics, Radius, Sanofi, and Ultragenyx and research support from Mereo.