Zoledronate Sequential Therapy After Denosumab Discontinuation to Prevent Bone Mineral Density Reduction: A Randomized Clinical Trial

Abstract

Importance: Discontinuation of denosumab without transitioning to another antiresorptive agent results in rapid bone loss and an increased risk of fracture. Previous randomized studies reported inconsistent results regarding the efficacy of zoledronate as sequential therapy.

Objective: To investigate the use of sequential therapy with zoledronate to prevent bone loss and decreased bone mineral density (BMD) after denosumab discontinuation in the first year.

Design, setting, and participants: The Denosumab Sequential Therapy prospective, open-label, parallel-group randomized clinical trial was conducted at a referral center and 2 affiliated hospitals in Taiwan. Recruitment was conducted from April 1, 2019, to May 31, 2021, and a 2-year follow-up was planned. The trial included postmenopausal women and men aged 50 years or older who received regular denosumab treatment for at least 2 years and did not have previous exposure to other antiosteoporosis medication or meet other exclusion criteria.

Intervention: Participants were assigned via stratified randomization to 1 of 2 groups: group A received continuous denosumab treatment (60 mg twice yearly) as the positive control, whereas group ZOL received 1 dose of zoledronate (5 mg) in the first year.

Main outcomes and measures: The coprimary outcomes were BMD percentage changes in the lumbar spine (LS-BMD), total hip (TH-BMD), and femoral neck (FN-BMD), respectively. An intention-to-treat analysis was performed.

Results: This study included 101 patients (95 women [94.1%]; median age, 72.0 [IQR, 67.0-76.0] years). There were 25 patients in group A (23 women [92.0%]; median age, 74.0 [IQR, 70.0 to 78.0] years) and 76 in group ZOL (72 women [94.7%]; median age, 71.0 [IQR, 65.7 to 76.0] years). In the first year, group ZOL had a significant median decrease in LS-BMD (-0.68% [IQR, -3.22% to 2.75%]) compared with group A (1.30% [IQR, -0.68% to 5.24%]) (P = .03). No significant differences between groups A and ZOL were observed for TH-BMD (median, 1.12% [IQR, -0.06% to 2.25%] vs 0% [-1.47% to 2.15%]) (P = .24) and FN-BMD (median, 0.17% [IQR, -2.29% to 2.90%] vs 0.18% [-2.73% to 3.88%]) (P = .71). We observed a significant difference in the median LS-BMD percentage change for the ZOL subgroup with 3 or more years of denosumab treatment before enrollment (-3.20% [IQR, -7.89% to 0.68%]) compared with group A (1.30% [IQR, -0.68% to 5.24%]) (P = .003).

Conclusions and relevance: In this randomized trial of sequential therapy after denosumab discontinuation, bone loss was observed in LS-BMD in the first year among patients receiving zoledronate. A longer duration of denosumab treatment was associated with a further decrease in LS-BMD after zoledronate sequential therapy. Further randomized clinical trials and large-scale studies that investigate the strategies of sequential therapy after long-term denosumab treatment are needed.

Trial registration: ClinicalTrials.gov Identifier: NCT03868033.

Figure 1.. Flow Diagram of the First Year of the Trial

^aOne patient withdrew consent owing to noncompliance with the follow-up protocol; the other patient withdrew consent due to difficulty in drawing blood, which led to frustration after multiple attempts.

Figure 2.. Changes in Bone Mineral Density (BMD) in the First Year of the Trial

A, Compared with the positive control (group A), a significant difference in the percentage change in BMD after sequential therapy was observed in the lumbar spine for the zoledronate treatment group (group ZOL, which combines groups B, C, and D). B and C, No significant difference in the percentage change in BMD was observed for the total hip or femoral neck. D, Subgroup analysis suggested that the significant difference was associated with the duration of denosumab treatment before enrollment. A significant difference was only observed in the subgroup with 3 or more years of denosumab treatment before enrollment compared with group A and the ZOL subgroup with less than 3 years of treatment. Horizontal lines and boxes represent the median and IQR percentage changes in BMD, respectively; error bars represent 95% CIs.

Figure 3.. Serum Levels of Bone Turnover Markers

A, At the end of the first year, there was no significant difference regarding C-terminal telopeptide (CTX) serum levels between the positive control (group A) and the treatment group (group ZOL, which combines groups B, C, and D). B, Compared with group A, the ZOL subgroup with 3 or more years of denosumab treatment at baseline had a significant difference in CTX serum levels at the end of the first year. No significant difference was observed for the ZOL subgroup with a shorter duration of denosumab treatment (<3 years). C, N-terminal propeptide of type 1 collagen (P1NP) serum levels increased significantly at the end of the first year in group ZOL. D, Both subgroups had significant differences in P1NP serum levels compared with group A.