Background:​ The rebound effect after denosumab discontinuation is lessened with subsequent zoledronate therapy. However, it is unclear whether this mitigation is sufficient after long-term denosumab treatment.

Objective:​ This retrospective observational study analysed bone mineral density (BMD) and bone turnover marker (BTM) changes after denosumab therapy according to treatment duration and subsequent zoledronate regimen.

Methods:​ We measured the outcomes of 282 women with postmenopausal osteoporosis who discontinued denosumab and received zoledronate 6 months later. In patients with longer denosumab therapy (≥5 years), BTMs were measured every 3 months and a second zoledronate infusion was administered if BTM levels increased by ≥2-fold. The BMD of all women was measured before denosumab therapy, at the last injection and 1 to 2 years after the first zoledronate.

Results:​ Bone loss after switching from denosumab to zoledronate was higher in patients with 10 ± 2 denosumab injections (n = 84) compared to 5 ± 2 injections (n = 144, p < 0.001 for lumbar spine and femoral neck), but there was no further increase with treatment durations of ≥15 ± 2 injections (n = 54, p = 0.35 and p = 0.20, respectively). BTMs in patients with ≥10 denosumab injections were elevated 6 months after zoledronate in some patients, but not all. Twenty-four women received a second zoledronate dose 6 months after the first one. BTMs in these patients were subsequently lower, but bone loss at both the lumbar spine and hip was comparable to that in patients with only one zoledronate dose (p = 0.37 for lumbar spine and p = 0.97 for femoral neck).

Conclusions:​ Rebound-associated bone loss reached a plateau after denosumab treatment durations of 4–6 years, irrespective of the frequency of subsequent zoledronate therapy.