AAOMS position paper on medication-related osteonecrosis of the jaw, 2022 update: the case definition, the four stages, the risk by drug and route, and why the drug holiday question has no answer

What it says#

The 2022 update is the fourth AAOMS position paper on this condition, after 2007, 2009 and 2014. A working group of three members of the association's oncologic and reconstructive surgery committee and three authors of the 2014 paper convened remotely in the fall of 2020 to appraise the literature. Its stated purpose is to update the risk estimates, to compare the risks and benefits of the medications so the treating physician, dentist, specialist and patient can decide, and to guide differential diagnosis, prevention and management by stage.

The case definition is unchanged from 2014 and has three elements, all required: current or previous treatment with antiresorptive therapy alone or in combination with immune modulators or antiangiogenic medications; exposed bone, or bone that can be probed through an intraoral or extraoral fistula in the maxillofacial region, that has persisted for more than eight weeks; and no history of radiation therapy to the jaws or metastatic disease to the jaws. The staging system is also unchanged, and the paper says so explicitly: it decided to maintain the classification with no modifications. Patients at risk are asymptomatic with no apparent necrotic bone after IV or oral antiresorptive therapy. Stage 0 is the nonexposed bone variant, no clinical evidence of necrotic bone but nonspecific symptoms or findings such as odontalgia with no odontogenic cause, dull aching bone pain, altered neurosensory function, loose teeth not explained by periodontal disease, swelling, alveolar bone loss, osteosclerosis or thickened lamina dura; progression to Stage 1 has been reported in up to 50% of Stage 0 patients, which is why the paper keeps it as a potential precursor. Stage 1 is exposed and necrotic bone, or a fistula that probes to bone, in an asymptomatic patient with no infection or inflammation. Stage 2 is the same with infection or inflammation, and the patient is symptomatic. Stage 3 adds necrosis extending beyond the alveolar bone, pathologic fracture, extraoral fistula, oral antral or oral-nasal communication, or osteolysis reaching the inferior border of the mandible or the sinus floor.

The risk numbers are reported by indication and by drug rather than as one figure. Among osteoporosis patients, placebo groups in clinical trials run 0% to 0.02%, and patients treated with bisphosphonates 0.02% to 0.05%: the paper puts IV zoledronate at or under 0.02%, no more than 2 per 10,000, and oral bisphosphonates at or under 0.05%, no more than 5 per 10,000. Denosumab carries a wider range, 0.04% to 0.3%, with 0.3% after 10 years of follow-up, almost an order of magnitude higher than bisphosphonates. Romosozumab sits at 0.03% to 0.05%, close to alendronate. Among cancer patients, placebo groups run 0% to 0.7%; zoledronate clusters under 5%, with a reported range of 0% to 18% explained by follow-up lengths from one to ten years, and a risk 2 to 10 times higher than placebo; denosumab runs 0% to 6.9%, most studies under 5%. Duration matters: among 5,723 cancer patients on zoledronate or denosumab, the risk was 0.5% and 0.8% at one year, 1.0% and 1.8% at two years, and 1.3% and 1.8% at three years.

For a dentist the extraction number is the one that governs. Among osteoporosis patients exposed to bisphosphonates, current estimates for the risk of MRONJ after tooth extraction range from 0% to 0.15%; on denosumab it was 1%. Among cancer patients exposed to bisphosphonates, the range is 1.6% to 14.8%, and the paper notes the estimates cluster between 1% and 5%, similar to osteoradionecrosis after extraction in irradiated patients. Extraction is cited as the predisposing event in 62% to 82% of MRONJ cases, though the paper is careful that pre-existing inflammatory dental disease confounds that association and that extraction may expose MRONJ rather than precipitate it. Implant risk after denosumab in osteoporosis patients was 0.5%; risk for other dentoalveolar operations is unknown. MRONJ appears in the mandible in 75% of cases, the maxilla in 25% and both in 4.5%, and denture wearing roughly doubled the risk in one cancer cohort.

Prevention is where the paper puts its weight. The modifiable factors it names are performing high-risk surgical procedures before therapy begins, using preoperative and postoperative antibiotics and antimicrobial mouth rinses, primarily closing extraction sites, and maintaining good oral hygiene, alongside smoking cessation and diabetes optimization; no individual strategy nor collection of strategies eliminates all risk, but these preventive procedures are recommended. Before oncologic antiresorptive therapy, if systemic conditions permit, initiation should be delayed until dental health is optimized, nonrestorable teeth and those with a poor prognosis should be extracted, and therapy should wait until the sites have mucosalized or there is adequate osseous healing. Before osteoporosis therapy the urgency is lower and optimization can happen concurrently. During therapy for cancer, avoid dentoalveolar surgery if possible, consider root retention rather than extraction, and the paper states that dental implants are contraindicated. During therapy for nonmalignant disease, there is no alteration of the operative plan for most patients.

On drug holidays the paper declines to give an answer, and says why. The practice was controversial in 2014 and remained so in 2021; the evidence to support or refute it is inconclusive because MRONJ is rare enough that randomized trials cannot power the question; a 2020 systematic review found papers with differing conclusions. The working group itself was unable to reach a consensus and was evenly split. The factors it lists for a case-by-case judgment are disease-related risk, cancer against osteoporosis, drug dosing frequency, duration of therapy, comorbidities, other medications especially chemotherapy, steroids or antiangiogenics, the degree of underlying infection or inflammation, and the extent of surgery to be performed. It does give one concrete caution: stopping denosumab causes a rebound increase in bone resorption and a risk of multilevel vertebral fractures, so if it is to be suspended, the planned dentoalveolar surgery can be completed 3 to 4 months after the last dose, when osteoclast inhibition is waning, and the drug reinstituted 6 to 8 weeks after surgery. Bone turnover markers, including CTX, are not validated for clinical decision making and are called out in the prevention table as not a useful tool to assess risk.

Who it applies to#

Any patient on, or previously on, an antiresorptive. In a general practice that is mostly the osteoporosis population, where the paper's message is reassurance with vigilance: the risk is low, elective surgery is generally not contraindicated, and it is clear the benefit of fracture prevention outweighs the risk of MRONJ development in osteoporotic patients. The cancer population, on oncologic dosing, is the one that changes a treatment plan, and the paper's own emphasis for those patients is prevention before therapy starts, root retention over extraction during it, and no implants.

It is a position paper from a specialty association, and its own framing is guidance rather than a mandate; the treating dentist and the treating physician decide together, and the paper repeatedly calls for a multidisciplinary approach. The AAPD's prophylaxis document lists MRONJ among the conditions for which prophylaxis may be considered after physician consultation, which is a separate question covered in aapd-antibiotic-prophylaxis-dental-patients-at-risk-2026.

What changed#

Against the 2014 edition, the case definition and the staging system are the same. What changed is the evidence around them. The risk estimates are updated and separated by drug, route and indication, with denosumab and romosozumab now sitting alongside bisphosphonates. Bone turnover markers, which had a place in the 2014 discussion, are described as a shift away, with no biomarker validated for decision making. The section on drug holidays moves from a qualified recommendation to an explicit non-consensus, with the working group's split reported in the text. And the paper adds a warning the earlier editions did not carry, about rebound bone resorption and vertebral fracture risk after denosumab is stopped, with a timing strategy that minimizes the length of any holiday.

Where the source is#

The link under the title opens the AAOMS's own 34-page PDF of the position paper at aaoms.org, which carries the abstract, the case definition, the staging system, the risk sections, Tables 1 and 2 and the four figures, and it is what the text above was read from. The published article is J Oral Maxillofac Surg 2022;80(5):920-943, PMID 35300956; joms.org and sciencedirect.com both refused automated access in this session, so the journal's own pagination and any post-publication corrections were not read here. The four figures in the PDF, which carry the paper's staging and treatment algorithms, are images without extractable text, so the stage-based management reported above was taken from the running text and from Table 2 rather than from the algorithms. The AAOMS press release of October 13, 2022 summarizes the update in the association's own words. This entry reports a position paper, not a treatment plan, and the treating dentist and physician decide together.