Intravenous zoledronate for osteoporosis: less might be more

Abstract

Annual administration of 5 mg intravenous zoledronate is moderately effective in reducing fracture risk in older adults, decreasing the relative risk of clinical fracture by 33%. However, almost 10 years after its approval for use in clinical practice there remain very substantial uncertainties about the optimal treatment regimen, that is, the lowest dose and/or longest dosing interval that is efficacious. Several pieces of clinical research suggest that the current recommendation for annual administration of 5 mg zoledronate might represent overtreatment. Clinical trials to clarify the optimal use of zoledronate for reduction of fracture risk should be undertaken.

Introduction

When considering the management of fracture risk, practitioners and their patients have several pharmacological options for which there is evidence of efficacy from randomized controlled trials [Khosla, 2009]. Bisphosphonates remain to the fore among these treatments. Within the group of bisphosphonates that are in clinical use and have been demonstrated to reduce the risk of both vertebral and nonvertebral fractures, zoledronate offers a point of difference in being administered by intravenous infusion and exhibiting prolonged antiresorptive activity after each dose. This review will summarize existing clinical trial data for the utility of zoledronate in the reduction of fracture risk, contend that the optimal dose and dosing interval is unclear, and consider the possibility that the drug might be administered less frequently and/or at lower doses than is currently recommended, without loss of efficacy: that is, less might be more.

Phase III zoledronate clinical trial programme and the status quo

Two industry-sponsored trials were conducted in the zoledronate phase III clinical trial programme [Black et al. 2007; Lyles et al. 2007]. In the first, 7665 postmenopausal women with either low bone density or mild bone loss and prevalent vertebral fracture were randomly allocated to three annual intravenous administrations of 5 mg zoledronate or placebo [Black et al. 2007]. Over 3 years, zoledronate reduced the incidence of clinical fractures (all nontraumatic fractures except those of digits and face) from 12.8% to 8.4%, a relative risk reduction of 33%, with a number-needed-to-treat (NNT) of 23. In the second, event-driven trial, 2127 men and women who had recently suffered a hip fracture were allocated to annual intravenous 5 mg zoledronate or placebo. During a median follow up of 1.9 years, zoledronate had reduced the incidence of clinical fractures from 13.9% to 8.6%, a relative risk reduction of 35%, with an NNT of 19. These trials reported relative risk reductions that ranged from 25% to 27% for nonvertebral fractures, 30% to 41% for hip fractures and 46% to 77% for vertebral fractures [Black et al. 2007; Lyles et al. 2007].

In 2007, zoledronate was approved for treatment of postmenopausal osteoporosis in both the USA and Europe, as a once-yearly infusion. In 2009, approval was granted by the US Food and Drug Administration for use of intravenous 5 mg zoledronate every 2 years for prevention of osteoporosis. This decision was based on a 2-year industry-sponsored clinical trial which reported that a single 5 mg baseline dose had similar effects to those of two annual 5 mg doses on the surrogate outcome, bone density [McClung et al. 2009]. A 2-year randomized, placebo-controlled trial of annual 5 mg zoledronate in men with osteoporosis reported a 67% reduction in the relative risk of vertebral fracture, but lacked power to assess other fracture types [Boonen et al. 2012]. Data from patients who were receiving glucocorticoids were limited to bone-density endpoints [Reid et al. 2009].

The results of the phase III trials established the annual administration of 5 mg zoledronate as the standard treatment regimen. Subsequently, two extension protocols of the larger phase III trial have been reported [Black et al. 2012, 2015]. In neither protocol was fracture incidence the primary endpoint. In the first, 1233 participants who completed 3 years of active treatment in the original protocol were randomized to continue annual 5 mg zoledronate infusions or placebo for 3 years. The relatively small number of incident fractures limited the interpretation of the data, but the hazard ratio for clinical fractures was 1.04. In the second extension, 190 participants who completed 6 years of annual zoledronate therapy were randomized to continue annual 5 mg zoledronate infusions or placebo for 3 years. Too few women had clinical fractures (n = 19) to permit inferences about fracture risk.

The impact on clinical practice of the data from the zoledronate phase III clinical trial extensions is uncertain. Recent European guidelines did not discuss the 6-year data [Kanis et al. 2013]. An expert review emphasized the benefit of treatment beyond 3 years for morphometric (asymptomatic) vertebral fractures and recommended continuing treatment indefinitely for ‘high-risk’ patients and for at least 5 years, with the potential to continue subsequently, for those at ‘moderate’ risk [McClung et al. 2013]. In contrast, Scottish guidelines recommend an interval of at least 3 years without treatment after an initial course of three annual 5 mg zoledronate administrations, after which fracture risk may be reassessed to determine the need for further therapy [SIGN, 2015].

Evidence that less might be more

The possibility that doses of zoledronate lower than 5 mg and/or less frequent administration than annually might be effective osteoporosis treatment was raised by the results of the phase II clinical trial, published in 2002 [Reid et al. 2002]. The study only lasted 1 year. Postmenopausal women were randomized to receive placebo or one of six treatments with intravenous zoledronate. The active treatment groups included total zoledronate doses of 1 mg, 2 mg and 4 mg, with dosing intervals including 3 months, 6 months and 12 months. The effects of each zoledronate treatment on the surrogate markers of bone health, bone turnover markers and bone density were superior to those of placebo, and indistinguishable from each other. No follow-up data were published. Thus, the trial failed to identify the optimal dose of zoledronate for investigation of its antifracture efficacy, and failed to determine the optimal dosing interval. A 5 mg annual dose was chosen for the phase III programme, even though that dose was not studied in the phase II trial, and had not been compared with the 4 mg annual dose.

Subsequently, several clinical trials with surrogate endpoints (bone density and/or bone turnover markers) have assessed the duration of action of single doses of intravenous zoledronate of between 1 mg and 5 mg. In the first, 50 postmenopausal women were randomized to receive a single 5 mg dose of zoledronate or placebo. After 5 years of follow up, zoledronate stably reduced bone turnover markers by at least 40%, and increased bone density in the axial skeleton by at least 4–5% [Grey et al. 2012]. Recently, in a randomized trial conducted in frail elderly women, a single dose of 5 mg zoledronate produced increases in bone density and reduction in bone turnover markers that were sustained during 2 years of follow up, without evidence of treatment offset [Greenspan et al. 2015]. To investigate the effects of doses of zoledronate lower than 5 mg, 180 postmenopausal women were randomized to receive single baseline doses of placebo, or 1 mg, 2.5 mg or 5 mg zoledronate. After 1 year, each dose of zoledronate produced similar improvements in bone density and decreases in bone turnover [Grey et al. 2012]. After 2 years, there had been some offset of the effects of the 1 mg dose, but the 2.5 mg dose produced comparable effects to those of the 5 mg dose [Grey et al. 2014].

The evidence for very prolonged antiresorptive actions of zoledronate is also apparent in a study population other than postmenopausal women. In 43 men with HIV infection randomized to two annual doses of placebo or 4 mg intravenous zoledronate, increases in bone density and decreases in markers of bone turnover were sustained for at least 5 years after drug administration, without evidence of offset [Bolland et al. 2012].

Collectively, these data suggest that annual administration of a 5 mg dose of zoledronate may be more treatment than necessary to achieve a reduction in fracture risk. There are currently no clinical trials that have tested the antifracture efficacy of dosing intervals longer than 1 year and/or lower doses than 5 mg of zoledronate. However, post-hoc subgroup analyses of the zoledronate phase III trial data support the notion that treatment with 5 mg every 3 years might reduce fracture risk as effectively as 5 mg administered annually. Among the participants in the two zoledronate phase III clinical trials, 1367 received only a single baseline infusion of the drug. During an average of 1.5 years of follow up, the relative risk reduction for any clinical fracture was 32% in those treated with a single infusion, and 34% in those treated with three annual infusions [Reid et al. 2013]. Post-hoc analyses of subgroups based on treatment adherence should be interpreted cautiously, although the usual concern about such analyses is that treatment efficacy is inflated in the subgroup with greater treatment adherence [Yusuf et al. 1991], which is not the case in this instance.

Adverse effects

A plausible advantage of administering less treatment is a reduced rate of adverse events. Currently, limited data are available on adverse events for lower doses and/or less frequent administration than 5 mg zoledronate annually. The most common adverse effect of intravenous zoledronate is an acute phase reaction (APR) that is most likely to occur after the first dose [Reid et al. 2010]. In clinical practice, some patients who experience an APR are reluctant to receive further treatment. It is not clear whether the occurrence of an APR, or its severity, is related to the dose of zoledronate. In a small, double-blind randomized trial, the incidence of any symptom implicated in the APR was increased in each of the zoledronate groups (1 mg 63%; 2.5 mg 80%; 5 mg 80%) compared with placebo (37%) [Grey et al. 2012]. The differences between the zoledronate groups were not statistically significant. Osteonecrosis of the jaw is an infrequent but severe adverse effect of zoledronate, which occurs with increasing frequency over time in exposed patients [Khosla et al. 2007], consistent with a dose-dependent effect. It is not clear whether zoledronate is associated with atypical femoral fractures [Schilcher et al. 2014], so ongoing vigilance and research is required. As zoledronate is renally excreted, current recommendations are to avoid its use when the estimated glomerular filtration rate is below 35 ml/min. It is possible that administration of a dose lower than 5 mg might be safer in patients with renal function at or below this level, but no data are available to inform practice.

Why does it matter?

Annual administration of 5 mg zoledronate effectively reduces fracture risk, with few safety concerns. Why then should we be interested in refining the treatment to determine the optimal dosing regimen, which is the lowest dose that reduces fracture risk to a degree comparable to annual 5 mg dosing?

It is a fundamental tenet of medicine to administer the least treatment necessary to produce the desired result, since that approach is likely to produce the fewest adverse effects. It is also the most cost-effective strategy, as administering lower doses of a drug will reduce the cost of treatment per patient and potentially permit treatment of more individuals at risk of the outcome of interest. Treatment costs are critically important in health systems constrained by limited resources [Watson, 2006]. Finally, the possibility that longer dosing intervals might be as effective as a yearly administration has important implications for patient adherence to long-term treatment: patients are more likely to accept, and adhere to, less frequent treatments.

The future

Zoledronate has been in clinical use since 2007, and evidence suggesting that less treatment with zoledronate than 5 mg annually might effectively reduce fracture risk has been available for almost as long [Grey et al. 2009]. What steps have been taken to optimize the use of this treatment?

To my knowledge, only two trials examining the effects of less frequent than annual administration of zoledronate on fracture risk are currently under way. In the first, more than 2000 older postmenopausal women with mild bone loss have been allocated to either 5 mg zoledronate or placebo, administered every 18 months, with clinical fractures over 6 years as the primary endpoint (ACTRN12609000593235). In the second, 1050 early postmenopausal women have been randomized to receive placebo, 5-yearly 5 mg zoledronate or 10-yearly 5 mg zoledronate, with the primary endpoint of morphometric vertebral fractures at 10 years (ACTRN12612000270819). Secondary endpoints include vertebral fracture in the former trial and clinical fracture in the latter. Each of these trials will provide important information about the use and efficacy of infrequent zoledronate therapy.

Other trials that would be helpful include a 3–5 year randomized noninferiority trial comparing annual 5 mg zoledronate with a single baseline 5 mg dose, and a 3–5 year noninferiority trial comparing annual administration of either 1 mg zoledronate or 2.5 mg zoledronate with annual 5 mg zoledronate, each with clinical fracture as the primary outcome. Each of these trials would likely require sample sizes in excess of 6000 participants, but could be conducted using a pragmatic, open-label design.

Conclusion

Uncertainty persists as to how best to use intravenous zoledronate, an undoubtedly effective treatment for reducing fracture risk, almost 10 years after it entered clinical practice. In an era of constrained health resources, declining rates of development of new therapies for osteoporosis, and a global increase in the number of individuals at high fracture risk driven by ageing populations, we should strive to optimize the use of existing therapies. Doing so will require collaborative efforts by research groups internationally, but the investment is likely to benefit patients.