MRONJ Incidence at Low-Dose Antiresorptive Therapy for Osteoporosis

What the evidence says

In osteoporosis patients on antiresorptive therapy, measured MRONJ incidence is low. One review gives a prevalence of approximately 0.01% to 0.03%. A scoping review of 24 osteoporosis studies (498,443 participants) found a mean incidence of 0.58%, and 0.60% when restricted to studies with more than 500 participants. The sources do not reconcile the gap between these figures. Both are descriptive, and the prevalence estimate lacks a clear denominator.

Relative to non-users, one propensity-score-matched cohort of 44,900 patients older than 70 in Korea found that long-term bisphosphonate use increases MRONJ risk. The adjusted hazard ratio was 4.09 (95% CI 2.75-6.09), or 3 cases per 1000 patients, at moderate certainty. A separate statement in the same review says the risk is only marginally higher than in the general population, without giving a magnitude. In the scoping review, the 0.58% figure had no comparison group. So the direction of excess risk is supported by one matched cohort, while the size of any excess against an untreated baseline is thinly documented.

Route of administration and dose change the magnitude. Oral bisphosphonate users show an incidence of 0.05%, against a range of 1.6–14.8% (mean 7%) in cancer patients on intravenous bisphosphonates. In the same passage, cancer patients on denosumab show 1.8%. The source does not state the indication for the 0.05% group. In postmenopausal women with osteoporosis, the denosumab exposure-adjusted rate in FREEDOM Extension was 5.2 per 10,000 person-years. In cancer patients, a meta-analysis of RCTs found denosumab increases MRONJ risk compared with zoledronic acid: RR 1.448 at 1 year, 1.448 at 2 years and 1.402 at 3 years. That corresponds to 3 to 4 more cases per 1,000 patients. These trials were all funded by the denosumab manufacturer, and the population is not osteoporosis.

Local and patient factors modify risk. Periodontal disease is associated with a pooled relative risk of 3.02 (95% CI 1.99–4.56) in patients on antiresorptives. That analysis did not stratify by dose or indication, so it does not map directly onto low-dose osteoporosis therapy. A single-center study identified smoking, glucocorticoid therapy, type 2 diabetes and rheumatoid arthritis as risk factors, but reported no magnitudes. In case series of MRONJ around dental implants, 67% of implants were in the mandible, and 84% of functioning implants with MRONJ were mandibular. Average antiresorptive duration before onset was 4.8 years.

Some expected outcomes have thin evidence. Lamina dura thickening increases with long-term high-dose intravenous exposure, but showed no measured association with MRONJ (p = 0.39). Root canal treatment in patients on oral bisphosphonates shows no measured MRONJ trigger in a scoping review. That evidence is mostly case reports and small series. No source here gives a controlled estimate of MRONJ incidence at low-dose therapy against an untreated osteoporosis group.

Where the evidence splits

The matched cohort reports an adjusted hazard ratio of 4.09 for long-term bisphosphonate use. Another statement in the same review says the risk is only marginally higher than in the general population. The sources on this page do not explain the split. The cohort was limited to patients older than 70 in Korea, with heterogeneous bisphosphonate types and dosing. The "marginally higher" claim gives no comparison figure.

Absolute incidence also varies: 0.01% to 0.03%, 0.05%, 0.21%, 0.58% and 0.60%. The claims note heterogeneity in diagnostic criteria, follow-up and sampling across the scoping review's studies. They do not say which design differences produce which figure. The split is unexplained by the sources on this page.

Thresholds and dose-response

Treatment duration beyond four years is the one threshold stated. For osteoporosis patients on low-dose antiresorptives, MRONJ incidence can reach up to 0.2% when duration exceeds four years. A second source gives 0.21% for oral bisphosphonate exposure longer than four years. The first is cited background literature, and the second depends on the quality of a cited reference. Neither gives the incidence at shorter durations.

Dose and route also separate incidence levels. Oral bisphosphonate users show 0.05%, against 1.6–14.8% for intravenous use in cancer patients. No cutoff dose is given.

Open empirical questions

• What is the MRONJ incidence in osteoporosis patients on low-dose antiresorptives compared with matched untreated osteoporosis patients, outside the older Korean population?
• How does incidence change by treatment duration below and above four years, using original cohort data rather than cited background figures?
• Does the effect of denosumab versus bisphosphonates seen in cancer-dose trials hold at osteoporosis dosing?
• Do the periodontal disease and systemic risk factors change MRONJ risk specifically at osteoporosis doses, and by how much?