NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Gauthier K, Bai A, Perras C, et al. Denosumab, Raloxifene, and Zoledronic Acid for the Treatment of Postmenopausal Osteoporosis: Clinical Effectiveness and Harms [Internet]. Ottawa (ON): Canadian Agency for Drugs and Technologies in Health; 2012 Feb.
We identified six RCTs, whose results were reported in 20 published articles assessing the efficacy and safety of denosumab (one RCT), zoledronic acid (one RCT), or raloxifene (four RCTs) in the treatment of postmenopausal women with osteoporosis. No active-controlled RCTs were identified in the literature search; only placebo-controlled RCTs were retrieved. Therefore, no direct treatment comparisons were available to assess the relative efficacy of the drugs. We considered, but elected not to perform, indirect comparisons between the trials, considering substantial heterogeneity in patient populations and significant challenges limiting the use of statistical adjustments to manage heterogeneity, which were beyond the scope of this report.
The current evidence revealed that denosumab, zoledronic acid, and raloxifene were all effective in reducing the risk of vertebral fractures, both clinically- and radiographically-assessed (morphometric), after 36 months of treatment compared with placebo. Denosumab and zoledronic acid reduced the risk of multiple vertebral fractures, hip fractures, and non-vertebral fractures. The included trials only provided limited evidence regarding the effect of raloxifene at the recommended dose of 60 mg daily on these outcomes, which suggested that it may not be effective in preventing non-vertebral fractures, including hip fractures.
Pooled data for the two raloxifene doses in the MORE trial were not considered in our review since they did not meet our inclusion criteria. However, the results for these types of fractures were not significantly different from placebo, although a large proportion of the patients contributing to the results for raloxifene received twice the recommended dose.20 Such a conclusion is consistent with other literature indicating that raloxifene has not been shown to reduce non-vertebral fractures.4
Denosumab, zoledronic acid, and raloxifene were each associated with an increase in BMD compared with placebo after 36 months of treatment and up to seven years in one trial for raloxifene.5 Nevertheless, the clinical relevance of these findings remains uncertain. Although T-scores are commonly used in clinical practice for monitoring purposes, change in BMD does not always correlate with change in fracture risk, which is a complex multifactorial issue. Therefore, clinical patient outcomes such as fractures are preferred to assess the efficacy of osteoporosis agents. The evidence for health-related quality of life and hospitalizations/long-term care needs was not sufficient to assess the effect of denosumab, zoledronic acid, or raloxifene on these outcomes.
The proportion of patients who died during the trials, as well as the overall incidence of serious adverse events, was not significantly different between each active drug and placebo. Nevertheless, some specific toxicities were reported more frequently with denosumab, zoledronic acid, or raloxifene. In the included trials, denosumab was associated with a higher incidence of cellulitis, zoledronic acid with atrial fibrillation, and raloxifene with venous thromboembolism and hot flushes compared with placebo. While thromboembolic events have been established as a major concern with the use of raloxifene,4,6 additional data regarding atrial fibrillation with zoledronic acid suggest that they may have been false signals. Indeed, this finding was not observed consistently in other zoledronic acid trials,7 and in HORIZON PFT the events usually occurred more than 30 days after administration of the drug, suggesting they may not have been related to the infusion.19 A review of safety data by the FDA confirmed that, across all studies, no clear association between overall bisphosphonate exposure and the rate of serious or non-serious atrial fibrillation was observed therefore health-care professionals should not alter prescribing patterns.8
The incidence of ONJ and atypical fractures was low and not significantly different between the active drugs and placebo in the trials. Nevertheless, ONJ has been previously reported with bisphosphonates in postmenopausal osteoporosis, and more frequently with high doses of intravenous bisphosphonates in cancer patients.32 The role of bisphosphonates in the occurrence of atypical fractures cannot be ruled out, although causality remains unclear.7 Despite no evidence of renal toxicity in HORIZON PFT, zoledronic acid has been associated with renal dysfunction and therefore should not be used in patients with severe renal impairment and be used with caution in the presence of other products that could impact renal function.7 ONJ, atypical fractures, and renal dysfunction are uncommon adverse events and, as a result, the findings from the included trials should not exclude a potential relationship with the active drugs. The specific harms profiles should be considered when making a treatment decision. Additional potential benefits should also be weighted. For instance, raloxifene is also associated with a reduced incidence of invasive breast cancer,4,34 which may be desirable to some patients.
The current evidence was limited to a total of six placebo-controlled RCTs. Although we searched for trials comparing denosumab, zoledronic acid, and raloxifene with each other, no active-controlled RCTs were identified in the literature search. Therefore, no direct treatment comparisons were available to assess the relative efficacy of the drugs. First line therapy is oral bisphosphonate. This review focused on second-line treatment options for postmenopausal women with osteoporosis. In clinical practice, denosumab, raloxifene and zoledronic acid are all used as options following an inadequate response or intolerance to oral bisphosphonates. Heterogeneity exists among the trial populations, especially with regard to the risk of fracture. Indeed, patients in the denosumab and zoledronic acid trials were older and their BMD measurement appeared slightly lower than those of the patients in the raloxifene trials. While patients in the denosumab trial were less likely to have a prevalent fracture at baseline, the corresponding proportion was the highest in the zoledronic acid study, with marked variations existing for the raloxifene trials. As a result, we considered, but elected not to perform, indirect comparisons between the trials: there were significant challenges limiting the use of statistical adjustments to manage heterogeneity, which were beyond the scope of this report. In addition, the validity of these adjustments would have been limited by the low number of studies retrieved in light of our research question.
Despite some variation in the quality of the included trials, most of them showed an acceptable degree of methodological rigour. There were, however, some quality concerns, such as insufficient reporting to allow adequate judgment on allocation sequence and concealment, blindness of patients and investigators, as well as patient withdrawals. The trials were usually representative of the Canadian population of women living with osteoporosis; except for the exclusion of patients with recent bisphosphonates experience (oral bisphosphonates are widely used in clinical practice). In addition, we could not find any trials evaluating the efficacy and harms of denosumab, zoledronic acid, and raloxifene as a second-line treatment option for postmenopausal women with intolerance or inadequate response to oral bisphosphonates, although these treatment options may be used in clinical practice in this indication. These limitations may weaken to some extent the strength of evidence regarding the effectiveness of denosumab, zoledronic acid, and raloxifene in postmenopausal women with osteoporosis.
No active-controlled RCTs were identified in the literature search; i.e., no direct treatment comparisons were available to assess the relative efficacy of the drugs. A systematic review recently published by Hopkins et al.35 attempted to address this evidence gap through an indirect comparison of nine osteoporosis drugs, including those drugs reviewed in this report. Based on the combination of effect size, and probability of being most efficacious (based on an unadjusted statistical analysis), teriparatide, zoledronic acid, and denosumab ranked highest for reducing non-vertebral and vertebral fractures.35 The authors highlighted some important limitations regarding their analysis and conclusions. They stated that statistical adjustments were important to explore due to the existence of key differences in patient characteristics across trials, which might have affected estimates of the comparative effectiveness among treatments in the analysis. These differences pertain to particularly important factors, including age, bone mineral density, and history of vertebral fracture. However, the modeling required for exploring statistical adjustments was not possible due to the low number of studies for each drug, and it is therefore unclear whether the estimates of effectiveness are robust.
In addition, the authors raised concerns regarding the results for etidronate. In their analysis, this drug had a high probability of being the most efficacious treatment, which is contrary to other evidence.35 Therefore, the results from Hopkins et al. should be interpreted with caution.
Except where otherwise noted, this work is distributed under the terms of a Creative Commons Attribution-NonCommercial- NoDerivatives 4.0 International licence (CC BY-NC-ND), a copy of which is available at http://creativecommons.org/licenses/by-nc-nd/4.0/