Bisphosphonates for the prevention and treatment of osteoporosis
BMJ 2015; 351 doi: https://doi.org/10.1136/bmj.h3783 (Published 02 September 2015) Cite this as: BMJ 2015;351:h3783
Treating osteoporosis with bisphophonates infographicClick here to see an interactive graphic, showing patient assessment, treatment pathways and decisions at each stage.
- 1Division of Endocrinology, Diabetes, Nutrition, and Metabolism, Mayo Clinic, Rochester, MN 55905, USA
- Correspondence to: K Kennel kennel.kurt{at}mayo.edu
Abstract
Osteoporosis is a systemic skeletal disorder characterized by bone loss, which leads to impaired bone strength and an increased risk of fractures. Two million fractures are attributed to osteoporosis annually in the United States and they are associated with serious morbidity and mortality. Bisphosphonates reduce the risk of fracture by suppressing bone resorption and increasing bone strength, and they have been widely used for the prevention and treatment of osteoporosis. However, the use of these drugs for the management of osteoporosis remains a clinical challenge. There are several important considerations including appropriate patient selection, pretreatment evaluation, potential adverse effects, patient preferences, and adherence. This review will discuss the evidence informing the clinical strategy for using bisphosphonates in patients with osteoporosis and those at high risk of fracture, focusing on the benefits and risks of treatment. We will also consider issues related to the monitoring and duration of treatment.
Introduction
Osteoporosis is a skeletal disorder characterized by loss of bone mass, reduced bone mineral density, and deterioration of bone microarchitecture, which lead to bone fragility. According to data from the National Health and Nutrition Examination Survey III, 10 million people in the United States have osteoporosis and 43 million have osteopenia (low bone mass).1
Two million fractures are attributed to osteoporosis annually in the US, resulting in more than 432 000 admissions to hospital, almost 2.5 million medical office visits, and about 180 000 admissions to nursing homes.2 Similarly, more than 3.7 million osteoporotic fractures in men and women occur in Europe, with a direct cost of €31.7bn ($35bn; £22.5bn).3 Fractures may be followed by full recovery or they may result in chronic pain, disability, and death.4 5
During the period 1996-2006, which coincided with the introduction of bisphosphonate therapy for osteoporosis, the reported incidence of hip fractures declined substantially in the US,6 although many other secular changes could have driven this decrease.
Accordingly, a recent Canadian study found no association between the prescription of drugs for osteoporosis and the rate of hip fracture, thus questioning real world efficacy perhaps due to inappropriate patient targeting.7 Bisphosphonates that are currently approved by the Food and Drug Administration for the prevention or treatment (or both) of osteoporosis include alendronate, ibandronate, risedronate, and zoledronic acid. This review will discuss the rationale for treatment with bisphosphonates in patients with osteoporosis and osteopenia, focusing on the associated benefits and risks. It will also review approaches to management, such as treatment monitoring and duration of therapy.
Sources and selection criteria
References were identified through searches of publications listed by PubMed, Medline, Embase, Central, and World of Science from the year 2004 onwards. Our search terms included “osteoporosis”, “bisphosphonic acid derivative”, and names of individual bisphosphonates approved for treatment of osteoporosis in the US and Europe. References were also identified from relevant review articles and through searches of the authors’ files. Although etidronate, clodronate, and tiludronate are also used, we focused on the most widely prescribed bisphosphonates—alendronate, zoledronic acid, ibandronate, and risedronate. We included randomized and observational studies published in English, prioritizing studies with low risk of bias as well as reviews, meta-analyses, and clinical guidelines for osteoporosis and bisphosphonates. Finally, we searched the clinicaltrials.gov database for ongoing trials related to the use of bisphosphonates in osteoporosis.
Pharmacology and mechanism of action
Bisphosphonates are used orally or intravenously. The nitrogen containing oral bisphosphonates—alendronate, risedronate, and ibandronate—each have an absorption of about 0.7% under ideal conditions.8 After intravenous administration of bisphosphonates skeletal uptake is rapid and most of the drug that is not excreted in the urine within 24 hours is bound to bone. Bisphosphonates bind at the bone mineral surface, where they potently inhibit osteoclast mediated bone resorption (fig 1⇓), and subsequently embed in the bone, being released only during subsequent resorption.9
Fig 1 Mechanism of action of nitrogen containing bisphosphonates. (A) Nitrogen containing bisphosphonates selectively inhibit farnesyl pyrophosphate synthase (FPPS) within osteoclasts, which disrupts the HMG CoA-reductase pathway. (B) Osteoclast endocytosis of bisphosphonate from the bone surface leads to inhibition of FPPS and osteoclast apoptosis. HMG-CoA=3-hydroxy-3-methylglutaryl coenzyme A
In contrast to other antiresorptive agents, bisphosphonates with the greatest binding affinity to bone (zoledronic acid > alendronate > ibandronate > risedronate) may persist in bone, and patients continue to be exposed to the pharmacologic effects of these drugs several years after discontinuation.8 10 11
The extent of the effect of bisphosphonates on bone turnover, as assessed by biochemical markers of bone resorption, depends on the potency of the drug, its pharmacokinetics, the dosing regimen, and the characteristics of the bone disease being treated.9 12 Intravenous administration of a single 5 mg dose of zoledronic acid decreases bone resorption for as long as five years.13 Daily oral administration of bisphosphonates suppresses bone resorption to a nadir as quickly as three months after the start of treatment,8 and treatment is associated with an increase in bone mineral density (BMD) as early as three months after initiation.14
All bisphosphonates rapidly reduce bone resorption, which leads to decreased bone formation because resorption and formation are coupled. Within three to six months equilibrium is reached at a lower rate of bone turnover. The end result is preservation or improvement in bone mass and microarchitecture, which leads to a reduced risk of fracture as early as six months after administration.15 16 Each bisphosphonate has a unique pharmacologic profile, which could result in clinically meaningful differences in the speed of onset and offset of effect at various skeletal sites.12
Comparative effectiveness
No head to head studies have compared the effects of bisphosphonates on reducing the risk of fractures, so it is not possible to conclude that one is superior to another.17 However, meta-anlyses have assessed the effectiveness of bisphosphonates as a group.
A meta-analysis published in 2014 reported high quality evidence that bisphosphonates (alendronate, ibandronate, risedronate, and zoledronic acid) reduce fractures compared with placebo in postmenopausal women, with relative risks in the range of 0.4 to 0.6 for vertebral fractures and 0.6 to 0.8 for non-vertebral fractures.18 The effect of ibandronate on the risk of hip fracture is unclear because hip fracture was not a separately reported outcome in trials of this agent.18
Similarly, a network meta-analysis,19 which included studies with low to moderate risk of bias, found moderate to high quality evidence to support the efficacy of bisphosphonates. A network meta-analysis of eight randomized controlled trials (RCTs) that assessed the relative effectiveness of alendronate, ibandronate, risedronate, and zoledronic acid on many fracture outcomes concluded that zoledronic acid has the highest probability of achieving the greatest reduction in any fracture. However, this analysis was sponsored by and included authors employed by the manufacturer of zoledronic acid.20
Adverse effects
Figure 2⇓ details the adverse effects associated with bisphosphonates.
Fig 2 Dosing, administration, and primary adverse effects of bisphosphonates for the treatment and prevention of osteoporosis
Gastrointestinal
Gastrointestinal side effects, including esophageal irritation, dysphagia, and heartburn, are well known for all oral bisphosphonates and represent a potential barrier to tolerance and adherence. RCTs often report similar rates of upper gastrointestinal events in control and active arms (about half of patients),21 but this could be related to patient selection bias.
A meta-analysis of adverse events including numerous RCTs showed an increased risk of mild upper gastrointestinal side effects with the use of alendronate (odds ratio 1.07, 95% confidence interval 1.01 to 1.14).18 A network meta-analysis found no significant differences between alendronate, risedronate, and zoledronic acid in the discontinuation of treatment because of adverse events.22
The risk of gastrointestinal events can be decreased by ensuring proper drug administration (appropriate quantity of water and post-dosing postural positioning), so patient education is crucial.23 Patients who cannot remain upright or who have pre-existing gastrointestinal disorders, such as esophageal strictures, severe gastroesophageal reflux, or achalasia, should preferably not be treated with oral bisphosphonates.
Rare cases of esophageal cancer have been reported in patients exposed to oral bisphosphonates. Meta-analyses of observational studies with high heterogeneity and risk of bias report conflicting data,24 25 and the FDA has not concluded that patients receiving bisphosphonates have an increased risk of esophageal cancer.
Acute phase response
Intravenous administration of bisphosphonates has been associated with transient mild to moderate inflammatory symptoms, such as fever, myalgia, arthralgias, and headache, collectively known as an acute phase response. These symptoms occur in 12-42% of patients within 72 hours of the first infusion, resolve within one to four days, and are less likely to occur after subsequent infusions.26 Oral paracetamol (acetaminophen) or ibuprofen administered four hours after infusion of zoledronic acid may reduce the incidence and severity of these flu-like symptoms.27 This is an important consideration for patient adherence because patients who experience this acute phase reaction may be reluctant to continue treatment.
Hypocalcemia
Bisphosphonates reduce calcium efflux from bone by inhibiting osteoclastic bone resorption. This can cause a small, clinically unimportant, decrease in serum calcium as long as baseline serum calcium is normal.28
In the Health Outcomes and Reduced Incidence With Zoledronic Acid Once Yearly Pivotal Fracture Trial (HORIZON-PFT), 1.3% of patients receiving zoledronic acid experienced asymptomatic and transient hypocalcemia.29 A meta-analysis showed that zoledronic acid is associated with an increased risk of hypocalcemia (7.22, 1.81 to 42.70); however, 85% of patients did not need supplemental calcium.18 Patients with vitamin D deficiency are more likely to experience more severe and prolonged hypocalcemia. Therefore, serum calcium and vitamin D levels need to be assessed before bisphosphonates are started (see section on pretreatment evaluation). Severe hypocalcemia is a contraindication for continued treatment with bisphosphonates.
Nephrotoxicity
To avoid renal injury, risedronate and ibandronate are contraindicated in patients with a creatinine clearance of 30 mL/min/1.73m2 or less, whereas the threshold is less than 35 mL/min/1.73m2 for alendronate and zoledronic acid.30 With the use of these thresholds, oral bisphosphonates for osteoporosis have not been associated with adverse renal effects. Several cases of severe nephrotoxicity with intravenous zoledronic acid have been reported, and this has resulted in an FDA warning for healthcare professionals.31 Given that renal toxicity after intravenous bisphosphonates is related to the maximum drug level achieved, rather than the area under the curve for drug exposure,32 it is not surprising that severe nephrotoxicity has been reported in patients with cancer who receive high intravenous doses.
Osteonecrosis of the jaw
Osteonecrosis of the jaw (ONJ) has emerged as a rare complication of bisphosphonate use and can be associated with substantial morbidity. Most cases have been reported in patients with cancer who receive high doses of intravenous bisphosphonates for the prevention of skeletal complications of cancer.33
A recent systematic review reported an incidence of ONJ among bisphosphonate users ranging from 0.028% to 4.3% and refrained from further analysis due to multiple identified limitations.34 A study in Sweden based on surveys from oral surgery and dental clinics with the advantage of complete health registers reported an annual incidence of 0.067% with the use of oral bisphosphonates.35
The HORIZON-PFT prospectively found two cases of ONJ, one in each arm (placebo and zoledronic acid).29 No other cases were identified from four additional RCTs,36 37 38 39 yielding an overall incidence of less than one in 14 200 patient treatment years when zoledronic acid is used for osteoporosis.40 A recent study from Taiwan showed that the risk for ONJ was similar in patients with osteoporosis treated with bisphosphonates and those not treated with bisphosphonates.41
The risk of ONJ seems to increase when treatment is for longer than three years.42 Oral bone manipulating surgery and dental extractions are considered the most important risk factors. Other factors that might increase the risk for ONJ include poor oral health, poor fitting of dental appliances, diabetes, and glucocorticoids.42 On the basis of expert opinion, it is recommended that oral health be optimized before prescribing bisphosphonates, so that the need for subsequent dental extractions is minimized.43 Moreover, the International Task Force on ONJ suggests withholding bisphosphonates before and after invasive dental procedures until soft tissue has healed, particularly for patients with other risk factors for ONJ.44
Atypical femur fractures
Atypical femur fractures are a subset of subtrochanteric and shaft fractures that constitute only 4-10% of all femur fractures.6 Major features by which they are distinguished include location, minimal or no trauma, transverse or short oblique fracture line, absence of comminution, and medial spike with complete fracture.45
A pooled analysis of several phase III clinical trials of bisphosphonates found no increased incidence of these fractures but was limited by study population size, study duration, and access to radiographs.46 The predominant evidence for this association comes from observational studies with various definitions of atypical femur fracture, so results need to be interpreted with caution.
A recent meta-analysis of five case-control studies and six cohort studies with several limitations found an overall pooled risk ratio of 1.70 (1.22 to 2.37).47 A study of more than one million women reported that the incidence rate of atypical femur fractures increased from 1.78 per 100 000 in women receiving bisphosphonates for less than two years to more than 100 per 100 000 in those receiving bisphosphonates for eight years, suggesting that longer bisphosphonates use increases the risk.48 A recent nationwide cohort analysis using the current definition of atypical femur fracture45 found an annual absolute risk of 11 fractures per 10 000 person years of use and a 70% lower risk for each year since last use.49 A population based, nested case-control study reported a significantly increased risk of subtrochanteric or femoral shaft fracture (adjusted odds ratio 2.74, 1.25 to 6.02) in women with more than five years of bisphosphonate use, although the absolute risk of these fractures remained low compared with the risk for typical hip fractures among untreated people in the same period.50 Furthermore, the study found a decreased incidence of typical and much more prevalent osteoporotic hip fractures (adjusted odds ratio 0.76, 0.63 to 0.93) in those treated with bisphosphonates for five years or more.50
Similarly, increasing rates of subtrochanteric and shaft fractures in the US between 1996 and 2007 should be considered in light of declining rates of hip fracture during this same period. It was estimated that, for every 100 or so fewer typical femoral neck or intertrochanteric fractures, there was one more subtrochanteric fragility fracture.51
In conclusion atypical femur fractures are rare but may be associated with the long term use of bisphosphonates.
Suppression of bone formation and resorption
The “oversuppression of bone turnover” is a theoretically attractive but unproved hypothesis. Moreover, the level of bone resorption that constitutes oversuppression has not been estimated.9 It has been suggested that prolonged suppression of bone formation and resorption as a result of the use of bisphosphonates may have negative skeletal effects, with daily bone microdamage accumulating and bones becoming too brittle.52
A case series reported on eight patients who sustained non-spinal atraumatic fractures after three to eight years of alendronate therapy, raising the concern that prolonged bisphosphonate therapy might increase susceptibility to fractures.53 Furthermore, because fracture healing requires callus remodeling and the coupled activity of osteoclasts and osteoblasts, it has been suggested that fractures occurring in patients being treated with bisphosphonates may exhibit impaired healing. This notion has led some clinicians to recommend delaying treatment with bisphosphonates in patients in hospital for hip fracture, thus adding to the challenge of starting adequate anti-fracture treatment in patients with established osteoporosis. However, retrospective and secondary analyses of data from large clinical trials,54 and results from smaller studies,55 56 have found no evidence that bisphosphonates impair fracture healing.
A recent meta-analysis including these studies and four more found no clinically detectable delay to fracture healing.57 Unlike atypical femur fractures and ONJ, data supporting “oversuppression of bone turnover” as a clinically relevant adverse effect of bisphosphonates has not accumulated with further time and inquiry.
Atrial fibrillation
In the HORIZON-PFT the risk of atrial fibrillation as a serious adverse event was significantly higher (P<0.001) in the zoledronic acid arm (1.3%) than in the placebo arm (0.5%).29 This finding was not confirmed in the HORIZON Recurrent Fracture Trial (HORIZON-RFT), despite the participants being older and at greater cardiovascular risk.36 A meta-analysis, acknowledging the heterogeneity of the existing evidence found no association between bisphosphonates and atrial fibrillation overall, but a higher risk of atrial fibrillation as a serious adverse event.58 However, after reviewing data on 19 687 patients treated with bisphosphonates and 18 358 patients taking placebo, the FDA found no clear association between exposure to bisphosphonates and serious or non-serious atrial fibrillation.59
Orbital inflammatory syndromes
Although this is not well known, bisphosphonates have been associated with uveitis, conjunctivitis, episcleritis, and scleritis.60 A prospective study reported a 1.1% incidence of uveitis after intravenous zoledronic acid, which is thought to be part of the acute phase reaction.61
Musculoskeletal problems
Reports of severe musculoskeletal pain associated with bisphosphonates resulted in FDA communications of Postmarketing Drug Safety Information updates to patients and providers,62 yet no further data on this observation have been forthcoming.
Adverse effects of bisphosphonates remain uncommon even 18 years after alendronate was introduced into clinical practice in the US. However, these effects should be a pivotal point of discussion with patients and providers considering treatment with bisphosphonates. Notably, most patients still do not receive treatment for osteoporosis after hospital discharge for hip fracture, and treatment rates dropped significantly between 2002 and 2011.63 There may be several reasons for this, possibly including a fear of adverse effects.
Pretreatment evaluation
Because of the weight of evidence supporting their clinical efficacy, bisphosphonates are considered first line therapy in the treatment of osteoporosis. Before starting treatment it is important to assess the patient with a complete history and physical examination (box). Unrecognized metabolic bone diseases other than osteoporosis, such as hyperparathyroidism, osteomalacia, or chronic kidney disease-metabolic bone disease, may also be associated with low BMD. In such cases it can be ineffective or harmful to diagnose osteoporosis solely on the basis of BMD and to start treatment with bisphosphonates.
Principal considerations for bisphosphonate therapy in osteoporosis
Considerations before starting treatment
-
Take a complete history and perform a pertinent physical evaluation
-
Assess patient factors that could increase fracture risk or affect dosing and administration of bisphosphonates
-
Assess fall risk and treat modifiable risk factors
-
Obtain bone mineral density (BMD) measurements and assess fracture risk:
-
-Check calcium, vitamin D, and renal function
-
-Consider excluding metabolic bone diseases other than osteoporosis
-
-Exclude secondary causes of osteoporosis
-
-
Counsel on the risk of osteoporosis related fractures and decide about starting bisphosphonates using clinical judgment and taking into account patient preferences
-
Assure optimal oral health
-
Advise on calcium and vitamin D intake
-
Advise on stopping smoking and avoiding excessive alcohol intake where appropriate
Considerations during treatment
-
Assess for adverse effects and patient adherence
-
Perform a clinical assessment for new fractures or new risk factors for bone loss
-
Consider vertebral imaging in patients with height loss, new back pain, or postural changes
-
Consider re-evaluating BMD after two years or more often when medically appropriate
-
Consider checking bone turnover markers:
-
-Not recommended for routine use
-
-Could be useful for assessing treatment efficacy or patient adherence if baseline data are available
-
-
Consider transition to a different osteoporosis agent in case of treatment failure:
-
-Two or more incident fractures have occurred during treatment
-
-Bone remodeling markers are not suppressed by bisphosphonates on serial measurements
-
-BMD continues to decrease
-
-
Consider a drug holiday after comprehensive risk reassessment, including femur neck BMD, as follows:
-
In patients at modest fracture risk bisphosphonates could be discontinued after 3-5 years
-
In patients with higher fracture risk, continue bisphosphonates
-
Transition to another type of osteoporosis treatment can be considered
-
Considerations after discontinuing bisphosphonates
-
Clinical assessment for new fractures or new risk factors for bone loss
-
Consider ending drug holiday. The decision to resume bisphosphonates incorporates several factors, such as:
-
-Potency of the initial bisphosphonate
-
-Duration of the drug holiday
-
-Baseline femur neck BMD
-
-Presence of other clinical risk factors for fractures
-
-Short term changes in BMD and increases in bone turnover markers (of limited value)
-
Baseline vertebral imaging could be performed according to National Osteoporosis Foundation (NOF) guidelines to assess for asymptomatic vertebral deformities in selected patients.64 Moreover, relevant tests should be obtained when the patient’s history, physical examination, or BMD (for example, low Z score) raises the suspicion of secondary causes of osteoporosis,65 especially when treatment of such could obviate the need for bisphosphonates (table 1⇓).
Selected secondary causes of osteoporosis and laboratory evaluation64
Before starting bisphosphonates it is necessary to assess for and correct hypocalcemia and vitamin D deficiency. Serum creatinine should also be measured before starting treatment to assess renal impairment. Finally, in an attempt to prevent ONJ it is important to ask about imminent plans for invasive dental procedures and assure optimal oral health.43
Whether to start
Randomized trials have shown that bisphosphonates prevent osteoporotic fractures in selected patients.66 Data from the pivotal trials showed benefit in preventing clinical fractures only in patients with previous vertebral or hip fracture or a hip BMD T score of −2.5 or less.29 36 67 68 69 70 71 72 73 74
In the absence of fragility fractures, screening BMD may be performed to detect osteoporosis in postmenopausal women.64 75 However, validation trials of screening and starting bisphosphonates that have outcomes on fractures are lacking.76
The identification of patients without osteoporosis who might still benefit from bisphosphonates has evolved to include demographic and physical characteristics, personal and family history, certain health conditions, and drugs that have been shown to influence risk of an osteoporotic fracture.77
One of several available tools,78 FRAX, is commonly used in clinical practice to calculate patients’ 10 year probability of hip and major osteoporotic fractures. However, FRAX has limitations,79 80 including underestimation of the risk of fracture in patients with recent fractures, as well as those with multiple osteoporosis related fractures, low BMD at the lumbar spine with a relatively normal femoral neck BMD, and increased fall risk. Moreover, FRAX is not intended for use in patents under 50 years and has not been validated in patients who had been previously treated for osteoporosis.
Patients with osteopenia (T score between −1.0 and −2.5) at the femoral neck or lumbar spine and a 10 year probability of a major hip fracture or osteoporosis related fracture of ≥3% or ≥20%, respectively, based on the US adapted World Health Organization algorithm, could be considered appropriate for bisphosphonates therapy.64 However, results from retrospective analyses of clinical trials regarding the correlation between FRAX based fracture probability and the anti-fracture efficacy of bisphosphonates treatment are inconsistent.81 82
Bisphosphonates are approved by the FDA for preventing osteoporosis in menopausal women,64 but it is not known whether their use decreases the risk of fractures. Patients with osteopenia who require drugs that could induce bone loss83—such as glucocorticoids,84 aromatase inhibitors,85 86 and androgen deprivation therapy86—could potentially benefit from bisphosphonates.
Other patients with osteoporosis who may benefit from bisphosphonates but whose complex comorbid conditions are not fully covered in this review include premenopausal women,87 organ transplant recipients,88 or those with disuse osteoporosis.89 In these and other less common subgroups of patients with osteoporosis, evidence supporting bisphosphonates is based on improvement in BMD rather than reduced fracture risk.
In patients who are reluctant to take bisphosphonates or who have an intermediate risk of fracture, additional data might be sought to determine whether to start bisphosphonate treatment. Although not recommended for routine use, biochemical markers that correlate with the overall rate of bone turnover, such as beta-CrossLaps (beta-CTx), are available.
For those considering bisphosphonates for prevention of bone loss that is independent of the risk of fracture, bone turnover markers could predict the rapidity of bone loss.90 New technologies for the non-invasive assessment of bone strength might be used. Trabecular bone score is an FDA approved technique that analyzes the dual energy x ray absorptiometry lumbar spine image to yield a metric that correlates with the microarchitectural structure of vertebral trabecular bone. Trabecular bone score can improve estimates of fracture risk in patients with similar BMD on dual energy x ray absorptiometry of the lumbar spine.91
Ultimately all treatment decisions require clinical judgment, consideration of unique risk factors and comorbidities, and the patient’s preferences. Focusing on an outcome that matters to patients, such as an individualized estimate of fracture risk and risk reduction, rather than on a technical term such as the T score, and use of a decision aid can help patients participate in the decision to use (or not use) bisphosphonates and potentially improve adherence (fig 3⇓).92
Fig 3 Osteoporosis encounter decision tool for a patient with a 10 year fracture risk of 43%. The decision aid conveys the benefits (fracture risk reduction) and consequences of treatment (side effects, cost, daily routine changes). Reproduced, with permission from the Knowledge Evaluation and Research Unit-Mayo Clinic (http://osteoporosisdecisionaid.mayoclinic.org)
How to start
Oral bisphosphonates
Alendronate is FDA approved for the prevention and treatment of postmenopausal osteoporosis, treatment of osteoporosis in men, and treatment of glucocorticoid induced osteoporosis in men and women.93 Alendronate was shown to reduce the annualized incidence of clinical vertebral fracture (0.82/100 person years of alendronate, 1.77/100 person years of placebo; relative risk 0.46, 0.28 to 0.75) and hip fractures (0.37/100 person years of alendronate, 0.77/100 person years of placebo; 0.49, 0.23 to 0.99) versus placebo over three years in women with a previous vertebral fracture and hip fractures (0.23/100 person years of alendronate, 0.53/100 person years of placebo; 0.44, 0.18 to 0.97) in those with a femoral neck T score of −2.5 or less.72 It also reduced the incidence of clinical vertebral fractures (13.1% alendronate, 19.6% placebo; 0.64, 0.50 to 0.82) over about four years in women with a femoral neck T score of −2.5 or less, but there was no benefit for those whose score was greater than −2.5.74
Oral ibandronate is FDA approved for the prevention and treatment of postmenopausal osteoporosis. Ibandronate daily (2.5 mg) or intermittently (12 doses of 20 mg every other day every three months) reduced the incidence of clinical vertebral fractures using life table analysis by about 50% (estimated incidence rate 5.3% (3.7% to 6.9%) in the placebo group, 2.8% (1.6% to 3.9%; relative risk 0.51; P=0.0117) in the daily group, 2.8% (1.6% to 3.9%; 0.52; P=0.0143) in the intermittent group) compared with placebo over three years in patients with a previous vertebral fracture and lumbar spine T score of less than −2.0. No reduction in the risk of non-vertebral fractures was documented however.69
Risedronate is approved for the prevention and treatment of postmenopausal osteoporosis, osteoporosis treatment in men, and for the prevention and treatment of glucocorticoid induced osteoporosis in men and women.94 In one of the larger trials, risedronate reduced the incidence of radiographic vertebral fractures (absolute risk 18.1% risedronate v 29% placebo; relative risk 0.51, 0.36 to 0.73; P=0.001) and non-vertebral fractures (absolute risk 5.2% risedronate v 8.4% placebo; relative risk 0.61, 0.39 to 0.94; P=0.02) over three years in patients with a previous vertebral fracture.68 70 In elderly women with a femoral neck T score less than −4.0 and those with a femoral neck T score less than −3.0 and a non-skeletal risk factor for hip fracture, two years of daily risedronate reduced the risk of hip fracture by 40% (absolute risk 1.9% risedronate v 3.2% placebo; relative risk, 0.6; 95% CI, 0.4 to 0.9; P=0.009).73
Once weekly regimens with alendronate and risedronate are pharmacologically equivalent to daily regimens and may improve long term adherence to treatment. The approval of these non-daily regimens was not based on data showing a reduction in the risk of fracture, but on the expectation that an equivalent increase in BMD or reduction in bone turnover mediated by identical molecular structures results in an equivalent reduction in the risk of fracture.95 However, this expectation has not been evaluated in clinical trials.
Intravenous bisphosphonates
Intravenous bisphosphonates provide an alternative for patients who cannot tolerate oral bisphosphonates or who have difficulty with the administration requirements. Evidence from RCTs shows that the intravenous formulation of ibandronate improves BMD to a similar degree seen with daily oral ibandronate.96 97 Intravenous ibandronate is approved for treatment (but not prevention) of postmenopausal osteoporosis.
Zoledronic acid is approved for the prevention and treatment of osteoporosis in postmenopausal women and the treatment of osteoporosis in men. It is also approved for the prevention and treatment of osteoporosis in men and women expected to be on glucocorticoid therapy for at least 12 months.37 98
In the HORIZON-PFT trial 7765 postmenopausal women were randomized across diverse ethnic groups to yearly infusions of zoledronic acid versus placebo. Over three years of follow-up, zoledronic acid significantly reduced the incidence of clinical vertebral fractures (absolute risk 3.3% zoledronic acid, 10.9% placebo; relative risk 0.30, 0.24 to 0.38; P <0.001), hip fractures (1.4% zoledronic acid, 2.5% placebo; 0.59, 0.42 to 0.83; P=0.002), and non-vertebral fractures (8.0% zoledronic acid, 10.7% placebo; 0.75, 0.64 to 0.87; P<0.001) in patients with osteoporosis defined by prevalent vertebral fractures or femoral neck T score −2.5 or less.29
Moreover, in the HORIZON-RFT study, an annual infusion of zoledronic acid within 90 days of repair of a low trauma hip fracture in women and men significantly reduced the rate of new clinical vertebral (1.7% zoledronic acid, 3.8% placebo; 0.54, 0.32 to 0.92; P=0.02) and non-vertebral fractures (7.6% zoledronic acid, 10.7% placebo; 0.73, 0.55 to 0.98, P=0.03) and improved survival (9.6% zoledronic acid, 13.3% placebo; 0.72, 0.56 to 0.93, P=0.01).36 Therefore, zoledronic acid is also indicated for the prevention of new clinical fractures in women and men with a recent low trauma (osteoporosis related) hip fracture.
Only one RCT of men with osteoporosis was designed to assess the outcome of primary fractures. At two year follow-up, men treated with zoledronic acid had a 67% reduction in the risk of new radiologically detected vertebral fractures (1.6%) compared with men receiving placebo (4.9%)(relative risk 0.33, 0.16 to 0.70; P=0.002); however, no significant difference in clinical fractures was seen.38
Figure 2 presents the different doses and forms of administration of bisphosphonates. Table 2⇓ summarizes the FDA approved indications for bisphosphonate therapy and the agents, with evidence from FDA registration trials on the reduction in the risk of fracture. Although the primary endpoint for FDA approval of treatment for osteoporosis is a significant reduction in incident radiographic vertebral fractures, this outcome is of variable importance to providers and patients compared with clinical fractures.
Food and Drug Administration approved bisphosphonates for osteoporosis*
Adherence
As in other chronic asymptomatic diseases, non-adherence with drug treatment is a problem.99 A systematic review found that about half of patients prescribed oral bisphosphonates discontinue treatment within one year.99 It has been suggested that weekly dosing could increase adherence compared with daily dosing. The systematic review supported this suggestion—compliance at one year was higher in patients receiving weekly versus daily alendronate. However, overall patient adherence and persistence (duration of time from initiation to discontinuation) with treatment remained below that needed for optimal fracture prevention.99 100
Poor adherence to bisphosphonates could limit the effectiveness of treatment in women at risk of fractures. A meta-analysis of six studies (171 063 patients) showed that after 1-2.5 years of follow-up the increase in the risk of fractures for non-adherent patients was 28% (relative risk 1.28, 1.06 to 1.53) for hip fractures and 43% (1.43, 1.26 to 1.63) for clinical vertebral fractures.101 Another meta-analysis found about a 30% increase (odds ratio 1.29, 1.22 to 1.38) in fracture risk for non-adherent patients and 30-40% increase (1.40, 1.29 to 1.52) for non-persistent ones after 0.8-4.2 years of follow-up.102 The main reasons for non-adherence with treatment include financial hardship, fear or experience of adverse effects, concerns about drug treatments in general, and lack of perceived need for the treatment.103
Improving adherence
A systematic review found that periodic follow-up appointments with health professionals improve adherence and persistence with osteoporosis drugs,104 but few intervention strategies had a large effect size. An RCT showed that the use of a decision aid during discussions about drugs for osteoporosis significantly improved the quality of clinical decisions about treatment and may have improved adherence,92 suggesting that patients’ belief in a treatment and its expected benefits can contribute to adherence.
A recent systematic review of 20 studies found that simplification of dosing regimens, decision aids, electronic prescriptions, and patient education may help to improve adherence to or persistence with osteoporosis drugs.105
Ultimately, clinicians need to be aware that many patients who start treatment for osteoporosis will discontinue the drug, thereby limiting the treatment’s effectiveness. Effective patient-provider communication and integration of care with other healthcare providers, especially the patient’s primary care physician, are essential.
Monitoring
BMD
NOF guidelines recommend that patients on bisphosphonates should have clinical and BMD re-evaluation after two years, or more often if medically appropriate. Vertebral imaging should be performed in patients with height loss, new back pain, or postural changes during treatment.64 Bisphosphonates increase BMD in most patients, but trials of bisphosphonates have found that even when serial BMD measurements have not increased or have modestly decreased, the risk of fracture has still significantly decreased.106 107 108 109 This suggests that other mechanisms of bone strength, not related to a change in BMD, are at play. Nevertheless, the finding of a significant decrease in BMD in a patient taking bisphosphonates should prompt an evaluation for possible contributing factors, such as poor adherence to treatment, poor absorption, inadequate calcium or vitamin D intake, or development of a disorder that contributes to bone loss.
Bone turnover markers
It has been suggested that bone turnover markers should be monitored when assessing treatment efficacy and patient adherence to bisphosphonates110 111; however, data are inconsistent.112 113 Bone turnover markers predict BMD response when repeated three to six months after starting treatment.114 A reduction in bone turnover markers, independent of BMD changes, could represent evidence of a therapeutic response.115 116 However, baseline bone turnover markers and the accuracy and precision of laboratory markers vary widely.117
A major limiting factor is that the degree of suppression of bone resorption needed to optimize the reduction in fracture risk is unknown.12 With a few exceptions,115 116 in most clinical trials baseline turnover did not predict the extent of the reduction in fracture risk with bisphosphonate therapy. Owing to the lack of evidence, routine use of bone turnover markers is not recommended.
Incident fractures
In the absence of a valid test, monitoring for treatment response could include monitoring for new fractures during treatment. In the face of limited evidence, experts recently proposed criteria for treatment failure to include when two or more incident fractures have occurred during treatment, when serial measurements of bone remodeling markers are not suppressed by anti-resorptive therapy, and when BMD continues to decrease.118 If a BMD decrease truly reflects a treatment failure, clinicians could consider transition to a different drug (for example, from an oral to intravenous bisphosphonates or other osteoporosis agents). Combination therapy with a bisphosphonate and estrogen or raloxifene at the outset or after treatment failure is not recommended because the additional benefits, if any, are small and the risk of adverse effects is increased.64 Early studies that combined teriparatide with alendronate119 or zoledronic acid120 found no additional benefit or little additional benefit on BMD, respectively. Given that data on fracture risk reduction are lacking, the potential role for such treatment in select patients remains to be determined.
When to stop or not to stop
Osteoporosis is a chronic disease and, as such, how long to treat should be an a priori question when establishing the efficacy and safety of treatment.11 Even so, clinicians are faced with a lack of information when deciding whether to stop or continue treatment.
After starting bisphosphonates, changes can occur that affect the benefits versus risks of continuing treatment. These include:
-
Changes in patients’ characteristics as they get older (such as the development of sarcopenia and changes in adherence or social, environmental, or behavioral factors)
-
The biology of the underlying disease (for example, the appearance or resolution of relevant risk factors, such as a low body weight or glucocorticoid use)
-
The degree to which continued treatment modifies the risk of fracture
-
The cumulative risk of the adverse effects of therapy.
The fact that rare adverse effects, such as ONJ and atypical femur fractures, seem to be more common with long term use of bisphosphonates adds to the uncertainty regarding the optimal duration of treatment.44 45 Residual antiresorptive activity of retained bisphosphonates in bone is expected to maintain some degree of anti-fracture efficacy after discontinuation.11 121
Evidence on the effects of discontinuing bisphosphonates
Although each of these considerations has contributed to the application of “drug holidays,” the primary question that remains is—in which patients could a drug holiday entail an unacceptable risk of osteoporotic fracture?122 123
In the Fracture Intervention Trial Long-term Extension (FLEX) study, after a 10 year follow-up, no significant difference in the risk of non-vertebral fractures was seen between patients who continued (19%) or discontinued alendronate (18.9%).124 However, patients who continued taking alendronate had a significantly lower risk of a clinical vertebral fracture (2.4% for alendronate v 5.3% for placebo; relative risk 0.45, 0.24 to 0.85). A retrospective analysis of the FLEX data showed that in women without vertebral fractures at baseline, continuing alendronate reduced non-vertebral fractures only in those with femoral neck T scores −2.5 or less (0.50, 0.26 to 0.96) at the start of the drug holiday.125 However, the FDA analysis of the FLEX trial showed that the rates of clinical fractures were similar whether patients continued on alendronate or switched to placebo during the extension even for the subgroup of patients with femoral neck T scores −2.5 or less.126
Further analysis of these data found that older age and lower hip BMD at discontinuation of alendronate after four to five years of treatment predict clinical fractures during the next five years.127
In the HORIZON-PFT, women who had received zoledronic acid for three years were randomly assigned to continue for another three years or to switch to placebo. The women who continued on zoledronic acid had a lower incidence of radiographic vertebral fractures (3.0% v 6.2%; odds ratio 0.51, 0.26 to 0.95), but there were no significant differences in clinical fractures.128 Overall, these findings suggest that after three years of treatment with zoledronic acid, many patients could discontinue treatment for up to three years.
The extension of the Vertebral Efficacy with Risedronate Therapy MultiNational trial showed that one year after discontinuation of risedronate in patients who had received up to seven years of treatment the risk of new radiographic vertebral fractures was reduced in the former risedronate users compared with the former placebo users (6.5% v 11.6%; relative risk 0.54, 0.34 to 0.86).129 However, the interpretation of these data is limited by the small sample size and the lack of information on clinical vertebral fractures.
Only the HORIZON-PFT and FLEX studies used a randomized withdrawal design and in none of the extension studies was the total initial randomized population included, thereby introducing selection bias. All trials had a small sample size and were conducted in postmenopausal women, so the results may not apply to younger women or to men. In contrast to the original trials, the extension trials used BMD as the primary outcome, with fractures often assessed as exploratory endpoints or adverse events with limited statistical power. The data on long term use of bisphosphonates suggest that the use of alendronate for five years and zoledronic acid for three years may allow residual anti-fracture benefits even after these drugs are discontinued, but this may not be applicable to risedronate.
No data are available on anti-fracture effects after discontinuing ibandronate.130 In addition, it should be noted that the statistical limitations of all these retrospective analyses preclude the identification with confidence of a subgroup of patients who are more likely to benefit from treatment for longer than three to five years.
Clinical practice
Given that there is limited evidence of efficacy to guide treatment decisions beyond three to five years, the decision about drug holidays needs to be based on the individual.11 123 Patients at high risk (as defined by the extension trials) with femoral neck T scores −2.5 or less might continue on treatment for an additional three to five years,131 switch from an oral to an intravenous bisphosphonate (typically zoledronic acid), or switch to an alternative drug such as denosumab or teriparatide. Those with femoral neck T scores greater than −2.5 should be clinically reassessed for factors that might preclude discontinuation of osteoporosis drugs, such as intercurrent fractures, new chronic diseases, or the use of drugs that could increase bone loss (for example, corticosteroids, aromatase inhibitors, androgen deprivation therapy). Otherwise, for patients who are not at high risk or those whose femoral neck T score is greater than −2.5, it is reasonable to discontinue bisphosphonates after three to five years.131
Owing to the limitations of the data,127 in clinical practice the decision to resume bisphosphonates for osteoporosis varies and typically incorporates multiple factors, such as the potency of the initial bisphosphonate, duration of the drug holiday, baseline and subsequent changes in BMD, increase in bone turnover markers, and presence of other clinical risk factors for fracture.
Emerging treatments
The use of bisphosphonates for treating osteoporosis is an area of active research. A multicenter RCT with estimated completion date July 2016 is examining whether teriparatide is superior to risedronate in reducing the occurrence of new vertebral fractures over 24 months of treatment (https://clinicaltrials.gov/ct2/show/study/NCT01709110).
A prospective observational study is being conducted in the US to determine the risk of atypical femur fractures after bisphosphonate therapy for more than five years. (https://clinicaltrials.gov/ct2/show/NCT02155595). An RCT in the US with estimated completion date in June 2015 will investigate the effects of the combination of testosterone replacement and alendronate on BMD and parameters of bone quality in older men with hypogonadism and low bone density (https://clinicaltrials.gov/ct2/show/NCT01460654).
Guidelines
Updated guidelines for the prevention and treatment of osteoporosis applicable to postmenopausal women and men age 50 years or more were recently released by NOF in the US.64 European guidelines for assessing and treating postmenopausal women with osteoporosis at risk of fractures were published in 2013.132 Revised in 2010, the Canadian guidelines concentrate on the assessment and management of women and men over 50 years who are at high risk of fragility fractures.133 Bisphosphonates are universally recommended as first line therapy for patients with osteoporosis and those with an increased risk of fractures (who do not have a contraindication). Responding to an increasing focus on rare adverse effects, the Endocrine Society published a position statement in 2012 emphasizing that the risk of serious complications associated with bisphosphonates is very low, particularly when the established benefits of fracture risk reduction in patients with osteoporosis or those with high risk for fractures are taken into account.66
Conclusion
Bisphosphonates have been shown to prevent bone loss and reduce the incidence of fractures in patients with osteoporosis and are therefore the most widely prescribed antiresorptive agents. In the absence of prevalent fragility fractures, BMD and tools that assess the risk of fractures serve an important role in identifying patients most likely to benefit from bisphosphonates. Although these drugs have some rare but serious adverse effects (ONJ and atypical femur fractures), in patients at high risk of fracture their benefit in preventing fractures far outweigh the risks. Although bisphosphonates differ pharmacologically, no convincing comparative effectiveness data are available to conclude superiority of any one bisphosphonate. Data to support the utility or frequency of any single method of monitoring patients after starting (or stopping) bisphosphonates are limited. Similarly, the optimal duration of treatment and the impact of drug holidays in mitigating the risks associated with long term use of bisphosphonates are both unclear. As when deciding about starting treatment, decisions about the duration of treatment, consideration for and length of a drug holiday, and subsequent drug therapy should be determined after reassessing the individual patient’s fracture risk including femoral neck BMD. Patient factors, including patient preferences, and the provider’s clinical judgment should also be taken into account. Effective communication between clinicians and patients about the use of bisphosphonates is therefore essential throughout the management of osteoporosis.
Research questions
-
How effective are bisphosphonates in preventing fractures over extended intervals of time? As a corollary, does prevention of menopausal related bone loss reduce the risk of fractures in ensuing decades? If so, do such benefits exceed potential risks?
-
How long should bisphosphonates be continued? Can novel biomarkers or bone strength analysis further inform and individualize decisions about the duration of treatment?
-
How do bisphosphonates compare with other antiresorptive drugs in the prevention of osteoporotic fracture? Which patient features (such as patterns of bone loss or parameters of bone strength) might inform the selection of bisphosphonates versus alternative antiresorptive or anabolic agents?
How patients were involved in the creation of this article
Two patients accepted an invitation to review the manuscript and registered as patient reviewers for The BMJ. They were asked to indicate which sections or information were of greatest importance and relevance to them as patients and which sections were the least useful. Alternatively, or in addition, they could write general comments about what they felt was missing or underemphasized. As a result of these patients’ input, the authors expanded the discussion of bisphosphonate therapy in the setting of osteopenia. The authors also put more emphasis on the importance of effective communication between patients and providers, shared decision making, and care coordination for the use of bisphosphonates in the management of osteoporosis.
Notes
Cite this as: BMJ 2015;351:h3783
Footnotes
-
Thanks to Patricia Erwin for her help with the literature search and Michael R Gionfriddo for his editorial assistance.
-
Contributors: SM and KAK contributed equally to the writing of this article. KAK is guarantor.
-
Competing interests: We have read and understood BMJ policy on declaration of interests and declare the following interests: none.
-
Provenance and peer review: Commissioned; externally peer reviewed.