Reviews/Osteoporosis Fracture Prevention/Denosumab withdrawal rebound fracture
Denosumab Therapy · ostfrac_06 · v2
Borderline: within 4 points of the Established/Likely boundary, so a re-grade could land either side.
When people stop taking denosumab, bone breakdown speeds up again, and several studies find that spine fractures become more common soon afterwards. The strongest study, drawn from a large randomized trial and its extension, found the spine-fracture rate rose sharply after stopping, and that people who fractured often broke several vertebrae at once [1]. Larger real-world patient records [3][10] and detailed case series [21][28] point the same way, and laboratory work explains why bone resorption rebounds [12][18][25]. The panel judged the effect real and consistent because independent studies from different countries agree and a biological mechanism fits. It held back from the very top because the single best study showed the overall rate returning to the level of never-treated people rather than exceeding it [1], and because much of the supporting evidence measures bone density or blood markers rather than actual fractures.
A post hoc analysis of the FREEDOM trial and its extension found the vertebral-fracture rate rose from 1.2 to 7.1 per 100 participant-years after stopping denosumab, with a larger share of fracturing patients sustaining multiple vertebral fractures (60.7% vs 38.7%, p=0.049) [1]. A retrospective series of 797 cases found vertebral-fracture incidence of 2.2% during treatment and 10.3% after [2], a population cohort found a long-delay vertebral-fracture hazard ratio of 3.91 (1.62-9.45) [3], and a 28,606-patient claims cohort found discontinuation nearly doubled fracture risk (HR 1.92, 1.75-2.08) [10]. Case series defined the rebound-associated vertebral fracture phenomenon, including 9 women with 50 such fractures [21] and a biomechanical case report [28]. Mechanistic studies show elevated RANKL and osteoclast activity driving rebound resorption after withdrawal [12][18][25]. The contested point: The division is over what the strongest dataset actually shows: [1] reports the off-treatment vertebral-fracture rate (7.1) as similar to discontinued placebo (8.5) and as a return to the level seen in untreated participants, so the denosumab-specific signal is the raised multiplicity of fractures rather than a rate exceeding the never-treated. Much of the remaining corpus measures bone density or turnover markers rather than fractures, and the clearest fracture evidence is uncontrolled or observational.
All 30 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Osteoporosis Fracture Prevention, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The strongest dataset [1] reports the off-treatment VF rate (7.1) as 'similar to participants who received and then discontinued placebo' (8.5) and as a return 'to the level observed in untreated participants' — the denosumab-specific rebound is the elevated MULTIPLICITY of VFs, so the blunt 'causes risk increase' slightly overstates the single best source.
A large share of the included corpus ([6][7][9][11][12][14][16][18][24][25][27]) reports BMD or bone-turnover-marker surrogates, not clinical vertebral fractures, so much supporting evidence is indirect for a fracture-risk claim.
Where it remains uncertain. In the single best source the overall off-treatment rate (7.1) was similar to discontinued placebo (8.5) and returned to the untreated level, so the specific rebound signal is fracture multiplicity rather than an absolute rate increase [1]; much of the corpus reports surrogate markers [12][14][18] rather than fractures.
Aleatory
[1] shows risk concentrated in those with prior vertebral fractures and longer off-treatment follow-up, and [2] notes the risk is poorly predictable before prescription, so individual fracture occurrence remains partly random.
Epistemic
The corpus has no randomized trial with vertebral fracture as the primary endpoint comparing continued versus stopped denosumab; [1] is a post hoc analysis and [3][10] are observational.
Inferential
[10]'s HR 1.92 is a composite of any fracture, so attributing its magnitude to vertebral fractures specifically requires an inferential step the abstract does not support.
The load-bearing papers from a corpus of 30, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
RCT-derived post hoc quantification of the rise in VF rate after discontinuation.
“the proportion with multiple (>1) was larger among those who discontinued denosumab (60.7%) than placebo (38.7%; p = 0.049)”
Panel auditInternal validity: medium · scope of inference: Postmenopausal women with osteoporosis who received ≥2 denosumab doses then discontinued, especially those with prior vertebral fracture.Key limitations: post hoc / off-treatment not randomized · overall VF rate similar to placebo arm · multiple-VF signal is the differentiating finding
Dose-response: delayed re-dosing raises vertebral-fracture risk.
“For vertebral fractures, short delay had an HR of 1.48 (CI, 0.58 to 3.79) and long delay an HR of 3.91 (CI, 1.62 to 9.45).”
Panel auditInternal validity: medium · scope of inference: Patients ≥45 years initiating denosumab for osteoporosis in UK primary care; delay rather than permanent discontinuation.Key limitations: delay not true discontinuation · non-randomized dosing · wide CIs for vertebral endpoint
Large claims-database cohort confirming fracture-risk increase after discontinuation.
“LTD of denosumab was associated with a 92% increase in fracture risk (HR 1.92 [95% CI 1.75-2.08])”
Panel auditInternal validity: medium · scope of inference: Postmenopausal women ≥60 in France with ≥2 years denosumab who then discontinued ≥12 months.Key limitations: composite fracture not VF-specific · claims-data residual confounding · observational
No osteoporotic fractures were reported in the off-treatment period and BMD simply returned to baseline, showing no rebound fracture increase in this context.
Panel auditInternal validity: low · scope of inference: RA patients on glucocorticoids given only 12 months of denosumab, i.e. short-term use where rebound is less expected.Key limitations: short treatment duration · zero events / underpowered · post hoc subgroup
4 of 30 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Multiple independent primary studies with dose-response and a coherent mechanism carry the claim, but surrogate-heavy support and the return-toward-baseline nuance in [1] cap confidence at the Established floor rather than higher.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
No randomized continuation-vs-discontinuation fracture trial. The causal claim rests on a post hoc analysis [1] and observational cohorts [3][10]; no trial isolates discontinuation with vertebral fracture as the primary endpoint. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A randomized trial allocating long-term denosumab users to continuation versus discontinuation (with or without sequential bisphosphonate), powered on incident vertebral fractures.
Open question: Does stopping denosumab raise vertebral-fracture risk above the never-treated baseline, or mainly raise the multiplicity of fractures among those who fracture? Why it is still open: [1] reports the off-treatment rate (7.1) as similar to discontinued placebo (8.5) and a return to the untreated level, while the denosumab-specific signal is the raised multiple-fracture proportion (60.7% vs 38.7%).
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19