Zoledronate is a potent bisphosphonate that, by virtue of its long skeletal half-life, provides protection against fractures when administered annually by intravenous infusion.1, 2 It also decreases mortality by 28% after hip fracture.2 The combination of antifracture efficacy, low rates of adverse events, and the convenience of infrequent dosing make it likely that zoledronate will be used widely in the treatment of osteoporosis in the foreseeable future.
At present, however, the duration of action of a single dose of intravenous zoledronate, and therefore the optimal dosing interval, has not been rigorously defined. The phase II study of intravenous zoledronate lasted only 1 year, after which all participants received further therapy. At the conclusion of the 12-month trial period, bone turnover markers in the group that received a single 4-mg dose of zoledronate at baseline were substantially lower than those in the placebo group.3 In randomized, placebo-controlled trials of men with HIV infection4 and osteopenic postmenopausal women,5 antiresorptive activity was apparent 24 months after the second of two annual 4-mg doses of zoledronate and a single 5-mg baseline dose, respectively, suggesting a duration of action considerably beyond 12 months.
Defining the duration of action of zoledronate, and thereby the optimal dosing interval, has potentially important implications for patient care and health care costs. We performed a 3-year randomized, placebo-controlled trial of the effects on bone turnover and bone mineral density (BMD) of a single baseline dose of 5 mg of intravenous zoledronate. A prespecified interim analysis after 2 years demonstrated persisting antiresorptive activity.6 Here we report the data from the full 3-year protocol.
The study population and recruitment have been described previously.6 In brief, participants were 50 healthy women more than 5 years postmenopausal with osteopenia (BMD T-scores between −1 and −2 at either the lumbar spine or the total hip). Exclusion criteria were previous hip or spine fracture, spine or hip BMD T-score of less than −2, serum 25-hydroxyvitamin D [25(OH)D] concentration of less than 30 nmol/L, prior use of a bisphosphonate, use of estrogen in the past 12 months, oral glucocorticoid use in the past 6 months, and major systemic disease. Participants were recruited between February 2005 and March 2006 by newspaper advertisement and from a database of participants in previous clinical trials. Recruitment was undertaken in parallel with that for another protocol with similar eligibility criteria. During the study, one participant in the zoledronate group withdrew after 18 months for personal reasons. The flow of participants through the study is outlined in Fig. 1.
Participants were randomized to receive a single intravenous administration of either zoledronate 5 mg or placebo at baseline. No other study medication was administered. Treatment allocations were randomized by the study statistician using a variable block size schedule based on computer-generated (Excel 2003, Microsoft, Redmond, WA, USA) random numbers. To ensure masking, only the statistician had access to treatment allocation. The staff member who prepared the intravenous infusions had no contact with participants. All the other study personnel and subjects were blinded to treatment allocation throughout. Only the study statistician saw unblinded data, but he had no contact with participants. The study received ethical approval from the Auckland Ethics Committee and was registered with the Australian New Zealand Clinical Trials Registry (ACTRN12605000278639). All participants gave written informed consent.
The coprimary endpoints were the bone turnover markers serum procollagen type-I N-terminal propeptide (P1NP) and β-C-terminal telopeptide of type I collagen (β-CTX). Secondary endpoints were BMD values at the lumbar spine, total hip, and total body. With allowance for 5 subjects to withdraw from each group, a sample size of 50 subjects is adequate to detect differences (with 80% power at the 5% significance level overall, p < .025 for each of the coprimary endpoints) of at least 1 SD between the treatment arms. We reasoned that a smaller difference than this would be unlikely to be clinically significant, that is, associated with antifracture efficacy, based on the results of randomized, controlled trials of treatments for osteoporosis.
Bone turnover markers, serum calcium and phosphate, and plasma parathyroid hormone (PTH) were measured at 0, 3, 12, 18, 24, 30, and 36 months, as described previously.6 BMD measurements of the lumbar spine (L1–L4), total hip, and total body were performed at baseline, at 12 months, and thereafter at 6-monthly intervals by dual-energy X-ray absorptiometry (GE Prodigy, GL Lunar, Madison, WI, USA) by technicians certified by Synarc, the international company that provides bone density oversight for most international osteoporosis drug registration trials. All BMD scans were analyzed in a blinded fashion at study conclusion. The coefficients of variation (CVs) for measurement of total hip and lumbar spine BMD in our laboratory are 1.1% and 1.4%, respectively.
Data were analyzed using procedures of SAS (Version 9.1, SAS Institute, Inc., Cary, NC, USA). All analyses were by intention to treat. Continuous, normally distributed variables were analyzed with a mixed-models analysis of variance. Significant main effects of time and treatment and their interaction were created. Significant main or interaction effects were further explored using the method of Tukey to preserve the overall pairwise error rate. All analyses were performed on absolute changes from baseline values, but BMD data are presented as percent change from baseline for ease of interpretation. Since the dependent variable was change from baseline, the baseline level of each variable was included as a covariate in the mixed-models analysis. Model reduction and post hoc testing proceeded as for the primary analysis of absolute change data. Since this study was subject to an interim statistical examination, the method of O'Brien-Fleming was used to determine the critical p value (p < .0024) for the 3-year examination (PASS 2002, NCSS and PASS Number Cruncher Statistical Systems, Kaysville, UT, USA).
There were no significant differences between the study groups in baseline characteristics (Table 1). Prior to enrolment in the trial, 7 women in the zoledronate group had sustained a total of 13 fractures during their lifetime, whereas 14 women in the placebo group had sustained a total of 22 fractures. During the study, 9 participants in the placebo group and 5 in the zoledronate group commenced treatment with either a calcium supplement or cholecalciferol (p = .35). Two participants in the placebo group commenced treatment with estradiol during the trial, each for 21 months.
Table 1. Baseline Characteristics of Participants
Note: Data are mean (SD), except where indicated. 25(OH)D = 25-hydroxyvitamin D; P1NP = procollagen type-I N-terminal propeptide; β-CTX = β C-terminal telopeptide of type I collagen; BMD = bone mineral density. To convert 25(OH)D to µg/L, divide by 2.5.
Current smoking, n
Dietary calcium intake (mg/day)
Serum 25(OH)D (nmol/L)
Lumbar spine BMD (g/cm2)
Total-hip BMD (g/cm2)
Total-body BMD (g/cm2)
Bone turnover markers
Each of the bone turnover markers was significantly lower in the zoledronate group than in the placebo group during the study, and values in the zoledronate group were in the lower part of the normal range for premenopausal women7 (Fig. 2). At the end of the study, the resorption marker β-CTX was 44% lower in the zoledronate group [absolute difference 217 ng/L, 95% confidence interval (CI) 11–325], and the formation marker P1NP was 40% lower in the zoledronate group (absolute difference 21 µg/L, 95% CI 8–35). The between-group differences for each marker were similar at 3 years to those at 1 and 2 years (Table 2). β-CTX and P1NP in 2 and 1 participants, respectively, were in the upper half of the normal range for premenopausal women at each of the 24-, 30-, and 36-month time points. β-CTX and P1NP in 1 and 2 participants, respectively, were below the lower limit of the normal range for premenopausal women at each of the 24-, 30-, and 36-month time points.
Table 2. Between-Groups Differences in Bone Turnover Markers and Bone Mineral Density Between 1 and 3 Years
Note: Data are mean (95% CI) differences between groups in absolute values (β-CTX and P1NP) or percent change from baseline (BMD). Negative values indicate lower levels in the zoledronate group. β-CTX = serum β C-terminal telopeptide of type I collagen; P1NP = procollagen type-I N-terminal propeptide; BMD = bone mineral density; LS = lumbar spine; TH = total hip; TB = total body.
−292 (−20, −382)
−23 (−13, −33)
5.7 (4.7, 7.4)
3.5 (2.2, 4.9)
1.4 (0.5, 2.2)
−219 (−124, −315)
−24 (−14, −33)
6.5 (4.7, 8.2)
3.9 (2.0, 5.8)
1.7 (0.9, 2.4)
−217 (−11, −325)
−21 (−8, −35)
6.8 (4.6, 9.1)
4.0 (1.8, 6.3)
2.0 (0.9, 3.0)
Bone mineral density
BMD at each of the sites assessed was higher in the zoledronate group than in the placebo group during the study (Fig. 3). At the end of the study, BMD in the zoledronate group was 6.8% (95% CI 4.6–9.1) higher at the lumbar spine, 4.0% (95% CI 1.8–6.3) higher at the total hip, and 2.0% (95% CI 0.9–3.0) higher at the total body. The between-group differences in BMD at each site were similar at 3 years to those at 1 and 2 years (Table 2).
During the first 2 years, there was evidence of mild secondary hyperparathyroidism in the zoledronate group.6 At 3 years, however, levels of serum calcium and PTH were similar in the treatment groups (mean between-groups differences 0.01 mmol/L, 95% CI −0.02–0.05, and −0.48 pmol/L, 95% CI −1.21–0.25, respectively).
No cases of osteonecrosis of the jaw, atrial fibrillation, ocular inflammation, or symptomatic hypocalcemia were observed. Four fractures occurred in 4 participants in the zoledronate group (1 each of finger, rib, forearm, and fibula) and 2 in 2 participants in the placebo group (1 each of toe and forearm, p = .68).
This study demonstrates that the antiresorptive actions of a single dose of intravenous zoledronate persist for at least 3 years in osteopenic postmenopausal women. Markers of bone turnover were about 40% lower in the zoledronate group than in the placebo group after 3 years, and there was no evidence of loss of antiresorptive activity between years 1 and 3. Throughout the study, the mean level of each turnover marker was in the lower part of the normal range for premenopausal women.7 Similarly, BMD was significantly higher in the zoledronate group at all sites assessed after 3 years, and the between-group differences at each site were similar at each assessment between years 1 and 3. The effects on BMD and bone turnover are similar in magnitude to those induced by risedronate,8 alendronate,9 and annual administration of intravenous zoledronate1, 2 in studies that demonstrated prevention of osteoporotic fractures with these agents. Although these data demonstrate that the antiresorptive actions of a single dose of zoledronate are sustained for 3 years, it is not known whether such a dosing interval reduces fracture risk. This important question should be evaluated in future clinical trials. Such trials also should evaluate mortality as an endpoint because annual dosing of zoledronate after hip fracture reduces the risk of death by 28%.2
The finding that intravenous zoledronate may be effective when administered considerably less frequently than once a year has important implications for osteoporosis therapeutics and pharmacoeconomics. Short- to medium-term compliance with oral bisphosphonate therapy is low, which is a major impediment to successful prevention of fragility fractures.10, 11 It is likely that less frequent drug administration will be more acceptable to patients and be accompanied by greater rates of adherence to treatment. Reducing the frequency of zoledronate administration to individuals at high risk of fracture also would reduce the cost of therapy per patient, which is likely to allow access to treatment for a greater number of patients.
Finally, these data suggest that maintaining bone turnover in the lower part of the normal premenopausal range, with consequent stability of bone mass, may be possible in the long term by infrequent administration of zoledronate. Thus these findings raise the possibility that prevention of early postmenopausal bone loss and the fractures that ensue might be achieved effectively, safely, and economically. Clinical trials to investigate this possibility are warranted.
AG has received consultancy fees from GSK; MB has no conflicts of interest to declare; DW has no conflicts of interest to declare; AH has no conflicts of interest to declare; GG has no conflicts of interest to declare; IR has received research funding and speaker and consultancy fees from Novartis, Merck, Procter & Gamble, and Amgen.
This study was supported by the Health Research Council of New Zealand.
AG designed the study and contributed to writing the manuscript; MB designed the study and contributed to writing the manuscript; DW recruited the participants, collected the data, and contributed to writing the manuscript; AH recruited the participants, collected the data, and contributed to writing the manuscript; GG supervised the randomization of participants, performed the statistical analyses, and contributed to writing the manuscript; IRR designed the study and contributed to writing the manuscript.