Denosumab / Bisphosphonates
| Type | Recommendation |
|---|---|
| When to deprescribe | |
| CBR |
Bisphosphonates We suggest deprescribing be offered to older people who:
|
| CBR |
Bisphosphonates We suggest offering a “drug holiday” to older people receiving long-term bisphosphonate treatment (5–10 years) who have no history of vertebral fractures, particularly those with a T-score ≥ –2.5, as the risk of rare but serious adverse events (e.g. atypical femoral fracture) may outweigh the benefits of continued treatment. |
| CBR |
Denosumab We suggest deprescribing be offered to older people:
|
| Ongoing treatment | |
| CBR |
Bisphosphonates We suggest continuing bisphosphonate therapy beyond 5–10 years in people with a T-score < –2.5 and/or a history of fragility fracture. For people who have discontinued bisphosphonates or are on a “drug holiday”, we suggest consideration of restarting bisphosphonates if:
|
| CBR |
Denosumab We suggest continuing denosumab in older people likely to derive a net benefit and have no significant adverse effects, even beyond 10 years, due to the increased risk of vertebral fractures after discontinuation. |
| How to deprescribe | |
| CBR |
Bisphosphonates We suggest ceasing bisphosphonates without the need for tapering. Seek expert advice for antiresorptive therapy in the context of chronic kidney and end-stage kidney disease. |
| CBR |
Denosumab We suggest transitioning to bisphosphonate therapy either 2 months before the next scheduled denosumab dose or at the time of the due dose and continuing for at least 12 months, due to the increased risk of rebound vertebral fractures following denosumab discontinuation or delayed administration. |
| Monitoring | |
| CBR |
We suggest closely monitoring for fracture risk using the Fracture Risk Assessment Tool and/or bone turnover markers, and the need for restarting therapy for osteoporosis in people not receiving therapy at a high risk of fracture, such as at least monthly for the first six months after deprescribing, followed by monitoring every six months thereafter to maintain the therapeutic relationship while working on lifestyle optimisation to reduce falls and fracture risk through multifactorial approach (e.g. environmental changes, exercise, nutrition). |
| CBR |
We suggest monitoring requirements and monitoring intervals for C-terminal telopeptide (CTX) and procollagen type 1 N propeptide (P1NP) be guided by specialists. |
| CBR |
We suggest closely monitoring for fracture risk using the Fracture Risk Assessment Tool and/or bone turnover markers, and the need for restarting therapy for osteoporosis in people not receiving therapy at a high risk of fracture, such as at least monthly for the first six months after deprescribing, followed by monitoring every six months thereafter to maintain the therapeutic relationship while working on lifestyle optimisation to reduce falls and fracture risk through multifactorial approach (e.g. environmental changes, exercise, nutrition). |
| CBR |
We suggest assessing bone mineral density (BMD) with dual-energy X-ray absorptiometry (DXA) for men and women once, after 12 months of discontinuation of therapy, if clinically appropriate and feasible. For people receiving antiresorptive treatment, we suggest monitoring BMD with DXA every two to five years when clinically appropriate, to evaluate treatment efficacy and the need for continued therapy, using the same instrument when possible. We suggest further assessment if BMD decreases by ≥ 5% at any major site or if a fracture occurs. |
| CBR |
For people receiving denosumab, we suggest the optimal timing for elective invasive dental procedures is immediately before the next scheduled dose. We suggest against withholding the next scheduled dose due to the risk of rebound fractures. We suggest a comprehensive dental review before initiating or reinitiating antiresorptive therapy due to the potential risk of medication-related osteonecrosis of the jaw. |
CBR, consensus-based recommendation
Bisphosphonates and denosumab are indicated for the treatment of postmenopausal osteoporosis to prevent fractures and associated morbidity in people with low bone mineral density (BMD) or a history of fracture [177]. In addition to pharmacological treatment, optimising bone health should include non-pharmacological strategies. These include ensuring a daily intake of at least 1,300 mg of calcium and 800 units of vitamin D, engaging in weight-bearing exercise for at least 30 minutes on most days of the week, smoking cessation for smokers, and limiting alcohol intake to two standard drinks per day [537].
Bone-modifying agents, including bisphosphonates and denosumab, are also used in the management of bone metastases but should not be relied upon solely for pain relief, as their analgesic effect is modest [538].
Mineral and bone disorders are common complications of chronic kidney disease (CKD), particularly in individuals with stage 3a CKD or more advanced disease (eGFR < 60 mL/min/1.73m²), due to disruptions in mineral metabolism [528]. Management becomes increasingly complex in advanced CKD (stages 4-5), where specialist guidance on antiresorptive therapy is crucial.
Refer to the narrative evidence summary, the GRADE Summary of Findings table in the guidelines, and the Technical Report for a complete presentation of the deprescribing evidence based on the GRADE framework (including other factors considered in developing the recommendations).
1. Mineral and bone disorders
The evidence supporting bisphosphonate use in advanced CKD is limited. As bisphosphonates are renally cleared, they are contraindicated in severe CKD due to potential nephrotoxicity, which has been reported particularly with pamidronate and zoledronic acid, whereas ibandronate appears to have a safer renal profile [539]. Unlike bisphosphonates, denosumab is not renally cleared and can be used in CKD. However, its use in advanced CKD requires caution due to the increased risk of hypocalcaemia (discussed further below).
2. Fracture prevention
2a) Bisphosphonates
Oral bisphosphonates are considered first-line treatment for most individuals at high risk of fractures. However, they are contraindicated in people with gastrointestinal conditions such as achalasia, Barrett’s oesophagus, or oesophageal scleroderma; certain gastric bypass procedures (e.g. Roux-en-Y); and CKD with an eGFR < 30 mL/min/1.73m². While bisphosphonates are well-documented in preventing fractures in postmenopausal women with osteoporosis, it remains unclear which subgroups with varying fracture risk derive the most benefit from treatment [540]. Additionally, a 2023 systematic review found that most osteoporosis guidelines provide limited guidance for healthcare providers, patients, and caregivers in making individualised deprescribing decisions for bisphosphonates [541].
-
Duration of therapy
-
Guidelines typically recommend reviewing bisphosphonate therapy after three to five years to assess the need for continued treatment [542]. A recent meta-analysis found that among postmenopausal women with osteoporosis, the estimated time to prevent one nonvertebral fracture per 100 women was 12.4 months (absolute risk reduction [ARR] = 0.010), while the time to prevent one hip fracture was 20.3 months (ARR = 0.005) and 12.1 months for clinical vertebral fractures (ARR = 0.005) [543]. However, this analysis excluded trials involving individuals at higher absolute fracture risk (e.g. secondary prevention of osteoporosis). For patients with limited life expectancy (< 12 months) receiving bisphosphonates for primary osteoporosis prevention and at low risk of future fractures, deprescribing may be appropriate.
-
Adverse effects and risk-benefit considerations
-
The risk of medication-related osteonecrosis of the jaw (MRONJ) and atypical femoral fractures (AFF), increases with treatment duration of bisphosphonate for osteoporosis though these events remain rare [480]. Most MRONJ cases occur in cancer patients receiving antiresorptive therapy for malignancy-related skeletal events, with incidence rates 100 times lower among those using bisphosphonates for osteoporosis [480]. Considerations on MRONJ risk are summarised in the section below.
-
There is no consensus on whether bisphosphonates should be discontinued following MRONJ in people with cancer who may benefit from pain control or skeletal event prevention. However, temporary discontinuation may be reasonable until resolution or stable improvement is observed.
-
For AFF, the risk increases with more than five years of bisphosphonate use, but the absolute risk remains low (3.5 to 50 cases per 100,000 person-years) and declines after discontinuation [544, 545]. If an AFF occurs, bisphosphonate therapy should be ceased. Conversely, osteoporotic hip fractures are associated with a three-fold increase in 12-month mortality in older people [546], making it essential to balance these risks against the benefits of fracture prevention when considering continued treatment. As with all therapeutic decisions, management should be individualised.
-
Long-term bisphosphonate use
-
The Fracture Intervention Trial Long-Term Extension (FLEX) trial found that extending alendronate therapy to 10 years did not significantly reduce the risk of nonvertebral fractures compared to a five-year regimen [547]. However, prolonged use was associated with a lower risk of clinical vertebral fractures, though it did not affect morphometric vertebral fracture risk, regardless of baseline vertebral fracture status. Additionally, a post hoc analysis of the FLEX trial found that alendronate significantly reduced nonvertebral fracture risk in women with a baseline T-score ≤ -2.5 but not in those with a T-score between -2 and -2.5 or > -2 [548]. Australian guidelines recommend continuing bisphosphonate therapy for five to 10 years in postmenopausal women and men over 50 with osteoporosis who have a T-score < -2.5 and/or a history of osteoporotic fractures [480].
2b) Denosumab
Denosumab is commonly used to treat postmenopausal osteoporosis in older women at high risk of minimal trauma fractures by suppressing bone turnover [177]. Its efficacy and safety in men with osteoporosis are less clear, with bisphosphonates typically being the first-line treatment [177]. Denosumab is administered via subcutaneous injection every six months, making it a potentially suitable option for individuals with high pill burden or compliance challenges who require long-term osteoporosis treatment [549]. It may also be considered for patients who experience treatment failure with bisphosphonate.
-
Safety considerations
-
Common side effects of denosumab include arthralgia, myalgia (ranging from transient to several months post-injection), hypercholesterolemia, cystitis, and flatulence. A significant concern is severe hypocalcaemia, particularly in individuals with advanced CKD. Among those with an eGFR < 15 mL/min/1.73 m² or on dialysis, the incidence of mild and severe hypocalcaemia is 24% and 15%, respectively [550]. Hypocalcaemia typically occurs within four weeks post-injection [551], and in severe cases, it can lead to serious complications, including severe weakness, tetany, prolonged QT interval requiring hospitalisation, or life-threatening cardiac arrhythmias [552]. In January 2024, the FDA issued a black box warning for denosumab due to the increased risk of severe hypocalcaemia in individuals with advanced CKD or on dialysis [553]. Kidney function and baseline serum calcium should be assessed before initiating treatment, as they strongly predict the risk of hypocalcaemia.
-
Treatment discontinuation and rebound effects
-
Unlike bisphosphonates, denosumab does not provide a sustained benefit after discontinuation [177]. An RCT of denosumab that included 256 postmenopausal women (mean age 59 years) reported a temporary increase in bone resorption markers after denosumab discontinuation before returning to baseline by 48 months, indicating a hyper-resorptive state, suggesting a temporary hyper-resorptive state [554]. Although BMD declined following discontinuation, it remained higher in the denosumab-treated group than in the placebo group. However, clinical outcomes such as fracture risk remain unclear. A cohort study found that delaying or discontinuing denosumab doses by ≥ 16 weeks was associated with a higher risk of vertebral fractures (HR 3.91, 95% CI 1.62-9.45) compared to on-time dosing (within four weeks of the scheduled injection) [555].
-
Duration of therapy
-
Denosumab, like bisphosphonates, has been linked to MRONJ and AFF, though these adverse effects are rare for denosumab [556]. Long-term safety data beyond 10 years are limited, but there are no formal restrictions on treatment duration [557]. The benefits of denosumab are not sustained after discontinuation, so drug holidays are not recommended. Given the increased risk of rebound vertebral fractures, ongoing treatment beyond 10 years may be considered for high-risk individuals with regular monitoring [558]. However, this decision should be discussed with the individual before initiating therapy, ensuring informed consent is adequately obtained.
-
If discontinuation is necessary, an alternative therapy should be initiated to prevent rapid BMD loss and fractures. Sequential treatment with bisphosphonates, particularly alendronate, is preferred and should commence six months after the final denosumab dose [559]. This should continue for at least 12 months to minimise the risk of vertebral fracture associated with denosumab discontinuation. However, the cumulative risk of MRONJ and AFF with prolonged use of denosumab for osteoporosis remains uncertain and future studies should investigate long-term management strategies.
3. Monitoring responses
Individuals initiating antiresorptive therapy should undergo a dual-energy X-ray absorptiometry (DXA) scan within one to two years, ideally using the same densitometer for consistency [479]. Treatment is considered effective if BMD remains stable or increases compared to prior scans. A change in BMD is only statistically significant if it exceeds the least significant change (LSC) for the specific densitometer used. If the LSC is unavailable, a threshold difference of ≥ 5% has been suggested [560]. For individuals with stable or improving BMD, subsequent DXA scans may be spaced out over two to five years. However, BMD stability does not always correlate with fracture risk reduction. Some patients may require more aggressive treatment despite maintaining a stable BMD.
For those receiving therapy for at least 12 months, a significant BMD decrease (≥ LSC) or a new fragility fracture should prompt further evaluation. In such cases, adherence, malabsorption, calcium and vitamin D intake, and potential secondary causes of osteoporosis should be assessed. Bone turnover markers (e.g. C-terminal telopeptide [CTX] and procollagen type 1 N-terminal propeptide [P1NP]) may be useful in specialist settings to differentiate non-adherence from malabsorption. Suppressed bone turnover markers indicate adequate medication adherence and absorption [560]. Patients with new fragility fractures or T-scores ≤ -2.5 may require a transition to anabolic therapy.
For those who discontinue antiresorptive therapy, it is important to continue close monitoring of fracture risk and the need for restarting therapy for osteoporosis, such as monthly for the first six months. Fracture risk can be monitored using the Fracture Risk Assessment Tool [478] and/or bone turnover markers as mentioned above.
As part of a multifactorial approach to fall and fracture prevention, it is also essential to address modifiable risk factors through other strategies. These may include nutritional review, environmental modifications and participating in fall prevention exercise programs, which have been shown to significantly reduce fall-related injuries, including fractures [480].
4. Preventive measures for MRONJ in people on antiresorptive medicines
Both bisphosphonates and denosumab are associated with MRONJ, a rare but serious condition [561]. A retrospective cohort study found that comprehensive dental care significantly reduces MRONJ risk [561]. Preventative dental care includes dental evaluation before initiating antiresorptive or antiangiogenic drugs and limiting invasive dental procedures during treatment. High-risk individuals (e.g. those on high-dose therapy, receiving treatment for more than three years, or with existing MRONJ risk factors) may require more frequent dental monitoring and prophylactic treatment [562].
For individuals taking bisphosphonates for osteoporosis, there is no clear evidence supporting a "drug holiday" for the purposes of minimising MRONJ risk prior to invasive dental procedures [563].
For individuals taking denosumab for osteoporosis, it is considered appropriate to perform invasive dental procedures four weeks after the last denosumab dose [563]. Guidelines suggest invasive dental procedures in patients on denosumab should ideally be undertaken just prior to the next six-monthly injection because the in vivo effect on bone suppression will be waning [480]. However, it may be reasonable to schedule invasive dental procedures no later than six weeks before the next due denosumab administration to allow adequate healing of the extraction socket [563].
For individuals receiving bisphosphonates or denosumab for cancer or skeletal fracture prevention, decisions regarding discontinuation due to dental procedures or MRONJ development should be individualised and made in consultation with the oncology team [563].
We identified six studies (two RCTs, two before-and-after studies, and two prospective cohort studies) related to bisphosphonate deprescribing (alendronate, risedronate, pamidronate, zoledronic acid) from the systematic review and meta-analysis; however, no studies related to denosumab deprescribing [547, 564-568].
Overall, there are reports that discontinuation of bisphosphonates may lead to loss of BMD gains while on treatment, although BMD at some body sites remained higher than baseline. Additionally, in postmenopausal women who had a history of vertebral fractures and low femoral neck BMD, discontinuation of bisphosphonates (alendronate, risedronate, zoledronic acid) after three to five years of treatment may lead to an increased risk of vertebral fractures. However, such a significant difference was not observed in non-vertebral fractures. These results support a legacy effect of bisphosphonates up to five years after ceasing therapy. For people with a T-score ≥ –2.5 and no prior vertebral fractures, deprescribing of alendronate may be considered after five years without a significant increase in their risk of non-vertebral fracture. However, the reported outcomes in the studies included are of low and very low certainty. Several extension studies lacked a true active group for direct comparison. There is a lack of evidence for newer therapy such as romosozumab and a lack of evidence for discontinuing bisphosphonate therapy for osteoporosis treatment in men. For instance, romosozumab as a newer bisphosphonate promotes both bone formation and suppresses bone resorption. All in all, the evidence at this stage is inadequate to inform evidence-based recommendations.
Key study characteristics and results
A narrative summary of each study is provided below, highlighting key characteristics and main findings.
Black 2006 conducted a study that extended from an original placebo-controlled RCT that examined the effects of alendronate on BMD and fracture risk in postmenopausal women with low femoral neck BMD (< 0.68 g/cm2) [547]. The current study included a subset of participants who were randomised to receive alendronate in the original RCT (mean of five years of alendronate treatment). Participants were excluded from this study if their total hip BMD was less than 0.515 g/cm2 (T score < −3.5) or lower than at the baseline of the original RCT. Concomitant use of hormone therapy or raloxifene was permitted. All participants were offered oral calcium (500 mg) and vitamin D (250 units). All eligible participants were randomised to stop alendronate (n=437) or continue alendronate for another five years (n=662). At five years, there was no significant difference between the two groups in terms of mortality (OR 1.54, 95% CI 0.80, 2.94), non-vertebral fractures (OR 1.01, 95% CI 0.74, 1.37), or adverse drug events (OR 1.11, 95% CI 0.75, 1.63). However, the continuation group had significantly more favourable BMD changes (i.e. either higher gain or lower reduction in BMD) for total body (MD 1.28, 95% 1.25, 1.31), trochanter (MD 3.17, 95% 3.14, 3.20), lumbar spine (MD 3.74, 95% CI 3.71, 3.77), femoral neck (MD 1.94, 95% CI 1.91, 1.97), and total hip (MD 2.36, 95% CI 2.33, 2.39). Additionally, there was a significantly increased risk of clinical vertebral fractures among people who discontinued alendronate compared with those who continued (OR 2.24, 95% CI 1.17, 4.30).
Black 2012 conducted a study that extended from an original placebo-controlled three-year RCT that examined the effects of zoledronic acid on reducing vertebral fracture and hip fracture in postmenopausal women either with femoral neck BMD T score ≤ −2.5, or ≤ −1.5 with evidence of two or more mild vertebral fractures or one moderate vertebral fracture [564]. Concomitant use of hormone therapy, raloxifene, calcitonin, tibolone, or tamoxifen was permitted. All participants received oral calcium (1000 to 1500 mg) and vitamin D (400 to 1200 units). The current study included a subset of participants who were randomised to receive zoledronic acid in the original RCT. All eligible participants were randomised to stop zoledronic acid (n=617) or continue zoledronic acid for another three years (n=616). At three years, there was no significant change in mortality (OR 0.68, 95% CI 0.37, 1.25), adverse drug events (OR 0.95, 95% CI 0.66, 1.38), or percentage change in BMD for spine (MD 2.03, 95% CI 0.76, 3.30), femoral neck (MD 1.04, 95% CI 0.43, 1.65), and total hip (MD 1.22, 95% CI 0.75, 1.69). However, there was a significantly increased risk of clinical vertebral fractures in people who discontinued zoledronic acid compared with those who continued (OR 2.14, 95% CI 1.12, 4.09).
Da Silva 2011 conducted a prospective cohort study that included postmenopausal women with osteoporosis who had been taking alendronate for at least five years and had their treatment discontinued (n=40) and those who had been taking alendronate for at least one year and continued treatment (n=25) [565]. All participants received vitamin D 1000 units daily during the study and those with a low calcium intake (not defined) received calcium supplementation in a dose sufficient to achieve 1000 mg/daily. At 12 months, there was no significant difference between the two groups in terms of non-vertebral fractures (OR 1.94, 95% CI 0.08, 49.40) or the proportion of participants with a clinically significant BMD loss in the femoral neck (OR 7.20, 95% CI 0.84, 61.38). However, significantly more participants who discontinued alendronate had clinically significant BMD loss in the spine (OR 10.67, 1.43, 100.39). The study considered BMD losses of ≥ 2.8% in the lumbar spine and ≥ 4.2% in the femur as clinically significant.
Eastell 2011 conducted a prospective cohort study that included 61 postmenopausal women with osteoporosis who had previously received risedronate for either two years (n=30) or seven years (n=31) in an RCT [566]. All participants enrolled in the original RCT were at least five years post-menopausal and had at least two vertebral fractures at baseline. The current study excluded those who used calcitonin, calcitriol or vitamin D supplements in the past month, anabolic steroids, estrogen, estrogen-related drugs or progestogen in the past three months, or bisphosphonates, fluoride or subcutaneous estrogen implant in the past six months. All participants received 1000 mg elemental calcium supplementation daily and those who required vitamin D supplementation received up to 500 units daily. Risedronate was discontinued in all participants. At 12 months, participants who had only two years of prior risedronate use had a significantly lower mean percentage change in BMD of the lumbar spine (MD 7.82, 95% CI 6.44, 9.20) and femoral neck (MD 4.33, 95% CI 2.90, 5.76) compared to those with seven years prior use. There was no significant difference between the two groups in terms of non-vertebral fractures (OR 0.33, 95% 0.01, 8.51) or adverse drug events (OR 1.24, 95% CI 0.45, 3.41).
Watts 2008 conducted a before-and-after study as a follow-up to an original placebo-controlled three-year RCT that examined the effects of risedronate on vertebral and nonvertebral fractures in postmenopausal women with osteoporosis [568]. Participants enrolled in the original RCT were at least five years post-menopausal and had at least two vertebral fractures at baseline or one vertebral fracture with low spinal BMD (T score ≤ −2). All participants received 1000 mg elemental calcium supplementation daily and those who required vitamin D supplementation received up to 500 units daily. In the current study, those who were randomised to receive risedronate (n=398) or placebo (n=361) in the original RCT stopped therapy. Concomitant calcium and/or vitamin D were permitted if the baseline levels were low. At 12 months, the participants who had risedronate discontinued had a significantly higher percentage change in BMD of trochanter (MD 3.08, 95% CI 2.06, 4.10), lumbar spine (MD 2.60, 95% CI 1.56, 3.64), and femoral neck (MD 2.32, 95% CI 1.40, 3.24). There was no significant difference in non-vertebral fractures between the two groups (OR 0.96, 95% CI 0.49, 1.85) but participants who had previously received risedronate for three years had a significantly reduced odds of vertebral fracture (OR 0.53, 95% CI 0.32, 0.89).
Orr-Walker 1997 conducted a before-and-after study as a follow-up to an original placebo-controlled two-year RCT (plus a one-year open-label period) that examined the effects of pamidronate on BMD and vertebral fractures in postmenopausal women with osteoporosis [567]. Participants enrolled in the original RCT had at least one vertebral fracture at baseline, and people with concurrent use of hormone replacement therapy or had treatment with sodium fluoride, calcitonin, anabolic steroids or bisphosphonate in the past six months were excluded. All enrolled participants received 1000 mg elemental calcium supplementation daily. At 12 months after pamidronate discontinuation, there was a non-significant reduction in total body BMD (-0.3 ± 0.7%, p = 0.7) compared to the original trial's inception. However, BMD at other body sites remained higher than baseline; lumbar spine (7.1 ± 1.1%, p < 0.0001) and femoral trochanter (4.5 ± 1.8%, p < 0.03), femoral neck (2.2 ± 1.3%, p not stated), ward's triangle (0.1 ± 2.5%, p not stated).
The method of deprescribing was not specified in all studies, but it appears to have involved abrupt discontinuation.