PDF
Citation
Print
I. Introduction
Bisphosphonates (BPs) and denosumab (Dmab) are the most commonly used drugs for treating osteoporosis1. Dmab was developed based on an understanding of the receptor activator of nuclear factor kappa B/receptor activator of nuclear factor kappa B ligand (RANK/RANKL) pathway starting in the 1990s and was approved by the U.S. Food and Drug Administration (FDA) as Prolia, and Xgeva in June and November 2010, respectively, for treatment of bone loss associated with osteoporosis and malignancy. BPs are analogues of inorganic pyrophosphate that inhibit pyrophosphate-dependent enzymes involved in bone resorption2. Dmab is a fully human monoclonal immunoglobulin G2 (IgG2) antibody that targets RANKL3. By inhibiting the interaction between RANKL and its receptor RANK, which is expressed on the surface of osteoclasts, Dmab suppresses the formation, function, and survival of osteoclasts, thereby reducing bone resorption and increasing bone density3. Although both BPs and Dmab reduce bone resorption, their mechanisms of action differ. BPs inhibit osteoclast differentiation and maturation and induce osteoclast apoptosis2, whereas Dmab, as a monoclonal antibody against RANKL, inhibits osteoclast differentiation and function3. Unlike BPs, Dmab does not bind to bone, and its effects on bone remodeling are largely diminished within 6 months after treatment cessation. In addition, BPs have low gastrointestinal absorption and require strict dosing regimens, including fasting and postural restrictions, whereas Dmab is administered as a subcutaneous injection once every 6 months, offering greater convenience and patient compliance. Compared with BPs, Dmab has been reported to be associated with a relatively lower risk of severe side effects, such as osteonecrosis of the jaw (ONJ). For these reasons, Dmab is increasingly recommended as first-line therapy for osteoporosis. Consequently, reports of Dmab-related osteonecrosis of the jaw (DRONJ) have increased in recent years.
The first case of ONJ, reported by Marx4 in 2003, was caused by pamidronate and zoledronate. It occurred in cancer patients who received high-dose of BP through intravenously. Subsequent cases were also reported in patients treated with oral BPs5. In 2007, The American Association of Oral and Maxillofacial Surgeons (AAOMS) proposed a formal definition for BP-related osteonecrosis of the jaw (BRONJ)6. Taylor et al.7 first reported DRONJ cases, and since then, DRONJ has been increasingly recognized. In 2014, the AAOMS recommended changing the terminology from BRONJ to medication-related osteonecrosis of the jaw (MRONJ)8. MRONJ is defined by following criteria: (1) Current or previous treatment with antiresorptive therapy alone or in combination with immune modulators or antiangiogenic medications, (2) exposed bone or bone that can be probed through an intraoral or extraoral fistula(e) in the maxillofacial region that has persisted for more than 8 weeks, (3) no history of radiation therapy to the jaw or metastatic disease to the jaws. This revised definition explicitly includes Dmab as a causative medication. Several studies have reported the clinical characteristics and treatment outcomes of DRONJ9-12. However, the most previous were limited to small case series and predominantly involved patients receiving high-dose Dmab for malignant disease, making it difficult to draw definitive conclusions applicable to clinical practice. In recent years, Dmab has been widely prescribed for osteoporosis, and DRONJ has become a substantial proportion of MRONJ cases13,14. Therefore, a comprehensive analysis of the clinical characteristics of DRONJ based on a larger patient population is warranted.
The objective of this study was to retrospectively analyze DRONJ cases treated at a single institution and to evaluate: (1) The clinical characteristics of DRONJ, with particular attention to differences among low-dose Dmab, high-dose Dmab, and sequential BP and Dmab therapy, (2) treatment outcomes following surgical treatments, (3) factors influencing treatment success in patients with DRONJ.
II. Patients and Methods
1. Study population
This retrospective cohort study included patients diagnosed as DRONJ who treated at the authors’ affiliated institute, from August 2019 to August 2024. The definition and staging of affected patients were defined according to 2022 AAOMS position paper for MRONJ15.
The inclusion criteria were as follows: (1) Patients with MRONJ after Dmab administration for osteoporosis treatment (low-dose Dmab, 60 mg/6 months; Prolia) or cancer treatment (high-dose Dmab, 120 mg/4 weeks; Xgeva), (2) history of Dmab treatment or switch to Dmab after BPs administration (alendronate, ibandronate, pamidronate, risedronate, or zoledronate), (3) minimum 3-months follow-up after conservative or surgical treatments.
The exclusion criteria were as follows: (1) History of radiation therapy to the oral and maxillofacial area, (2) patients whose medication history is not accurate or accessible, (3) patients who refuse treatment or were not followed up for more than 3 months after treatments.
2. Data collection
Patients were divided into 3 groups: Group 1, low-dose Dmab; Group 2, transition from BP to low-dose Dmab (BP to Dmab); Group 3, high-dose Dmab. Clinical and demographic data collected for each patient include age, sex, comorbidities, Dmab or BP therapy (dosage, duration), symptom at first visited, MRONJ stage, preceding event, location, the type of MRONJ treatment (conservative vs. surgical), number of surgical intervention(s), healing states, and number of recurrences were investigated retrospectively.
3. Evaluation of treatment outcomes
1) Primary outcomes: clinical healing
We investigated soft tissue healing status of individual patients at 3 months postoperatively. Definition of the outcomes after the surgery was divided into 3 categories according to the following clinical findings at least 3 months after conservative or surgical treatment; “complete”, “partial,” or “no healing”, which modified from a previous literature16.
① Complete healing: Absence of exposed necrotic bone or bone that can be probed through fistula, absence of purulent discharge, absence of edema or pain, and complete mucosal coverage at the surgical site.
② Partial healing: Clinical evidence of MRONJ stage improvement compared to the stage of the initial diagnosis, but did not show complete mucosal coverage.
③ No healing: Unchanged or worsening of stage compared to the stage of initial diagnosis.
The “treatment success” was defined as the combined proportion of “complete healing” and “partial healing”.
4. Statistical analysis
Patient demographics are expressed as n (%) and mean± standard deviation. To analyze the difference between the groups, continuous variables such as duration of drug administration were investigated by the ANOVA followed by Tukey’s post-hoc test. To compare the categorical variables, such as the MRONJ stage, Fisher’s exact test or chi-square test were used. Multiple logistic regression analysis was used to determine factors independently associated with the dependent variable—success rate of DRONJ treatment. Covariates in this model were age, dosage and duration of Dmab, location of the lesion, stage of DRONJ and treatment methods; 95% confidence intervals (CIs) for estimated odds ratio (OR) approximating the relative risk are also given. Statistical significance was set at P<0.05. Statistical tests were performed using SPSS statistical software (IBM SPSS Statistics for Windows ver. 29.0; IBM).
5. Ethics statement
This study protocol was approved by the Kyungpook National University Dental Hospital Institutional Review Board (approval No. KNUDH-2024-11-04-00). In written informed consent from the patients were waived, because of the retrospective nature of the study and use of de-identified participant data. This study was performed in accordance with the Declaration of Helsinki.
III. Results
After screening patients with MRONJ, 193 patients with DRONJ were initially identified as eligible for this retrospective study. However, 15 patients refused treatment or were lost to follow-up and were subsequently excluded. A total of 178 patients with DRONJ were ultimately included in the analysis.
1. Demographics and clinical characteristics
Baseline characteristics of the study population, including age, sex, and comorbidities, are summarized in Table 1. Majority of DRONJ cases (92.7%) occurred in patients treated for osteoporosis; Group 1 (low-dose Dmab, n=60; 33.7%), Group 2 (transition from BP to low-dose Dmab, n=105; 59.0%), Group 3 (high-dose Dmab, n=13; 7.3%). The mean age was 77.4±6.7 years in Group 1, 76.6±7.2 years in Group 2, and 63.5±8.9 years in Group 3 (ANOVA, P<0.001). Patients in Group 3 were significantly younger than those in the osteoporosis groups (Groups 1 and 2). The mean duration of Dmab administration before MRONJ development was 23.2±15.3 months in Group 1, 16.0±11.3 months in Group 2, and 27.9±18.8 months in Group 3. In Group 2, BPs had been administered for a mean duration of 56.7±46.4 months prior to Dmab treatment.
Overall, tooth extraction was the most common precipitating factor, accounting for 81 cases (45.5%), followed by periodontal disease (25 cases, 14.0%), peri-implantitis (22 cases, 12.4%), denture-related trauma (17 cases, 9.6%), implant surgery (11 cases, 6.2%), apical lesions (4 cases, 2.2%), and dentoalveolar surgery (2 cases, 1.1%). Idiopathic cases accounted for 16 patients (9.0%). Tooth extraction was the most frequent preceding event in all three groups, with no statistically significant difference among groups (P=0.656). Among the 178 patients, 31 (18.0%) were diagnosed with stage 1 MRONJ, 108 (60.7%) with stage 2, and 38 (21.3%) with stage 3. There were no statistically significant differences in MRONJ stage distribution among the groups (P=0.552).
Regarding lesion location, the mandible was the most commonly affected site overall (107, 60.1%), followed by the maxilla (62 cases, 34.8%) and involvement of both jaws (9 cases, 5.1%), with no significant intergroup difference (P=0.261). The mandible was the predominant site in all three groups, with prevalence of 56.7%, 59.0%, and 84.6%, respectively. The posterior mandible was the most frequently affected subsite, accounting for 100 cases (49.5%). Single-site involvement was observed in 158 patients (88.7%), whereas 20 patients (11.3%) exhibited multiple-site involvement. No significant differences were observed among groups regarding the number of involved sites (P=0.099).(Table 2)
2. Treatment outcomes
A total of 32 patients (18.0%) received conservative treatment, including antibiotic therapy, chlorhexidine gargling, incision and drainage, or minimal removal of necrotic bone. Surgical treatment, such as sequestrectomy and/or surgical curettage, was performed in 146 patients. The mean number of surgical interventions was highest in Group 3. However, there were no statistically significant differences among groups in treatment modality (P=0.657, chi-square test) or in the mean number of surgical procedures (P=0.231, ANOVA).(Table 3)
Treatment outcomes were evaluated on all 178 patients. Complete healing was achieved in 138 patients (77.5%), partial healing in 7 patients (3.9%), no healing in 10 patients (5.6%), and recurrence occurred in 23 patients (12.9%). Complete healing was more frequently observed in osteoporosis patients (Group 1, 80.0%; Group 2, 79.0%) than in cancer patients treated with high-dose Dmab (Group 3, 53.8%). The overall recurrence rate was 12.9% (Group 1, 11.7%; Group 2, 12.4%; Group 3, 23.1%). Although differences in individual outcome categories were not statistically significant among groups (P=0.102), treatment success rates differed significantly. Treatment success was observed in 85.0% of Group 1 and 82.8% of Group 2 patients, compared with 53.8% in Group 3 (P=0.027).(Table 4)
When outcomes were stratified by treatment modality, complete healing was achieved in 78.8% of patients who underwent surgical treatment and in 71.9% of those who received conservative treatment, with no statistically significant difference between the two approaches (P=0.657).(Table 5)
3. Factors influencing treatment success after DRONJ treatment
To identify factors associated with treatment success following DRONJ management, multiple logistic regression analysis was performed. Dmab dosage was identified as a significant independent predictor of treatment success (OR, 5.13; 95% CI, 1.19-22.14; P=0.028), whereas age, duration of Dmab therapy, lesion location, MRONJ stage, and treatment modality were not significant predictors. These findings indicate that patients treated with low-dose Dmab for osteoporosis (Groups 1 and 2) had significantly higher treatment success rates than those treated with high-dose Dmab for malignancy.(Table 6)
IV. Discussion
This study aims to clarify the clinical characteristics, possible risk factors, treatment modalities and outcomes of patients with DRONJ. An additional objective was to identify differences in clinical characteristics, risk factors, and treatment outcomes according to prior BP use in order to evaluate the impact of BP exposure on the development and prognosis of DRONJ. Although there were no significant differences were observed among groups with respect to DRONJ symptoms, disease stage, treatment modality, or overall treatment outcomes, significant differences were identified in the time to onset of DRONJ.
To date, relatively few studies have reported on DRONJ. Scoletta et al.17, reported 37 patients treated between 2006 and 2007. Palla et al.18 published a study on 108 patients treated from 2003 to 2009. A retrospective multicenter study conducted in Germany reported 69 DRONJ sites in 63 patients treated between 2011 and 20179. More recently, Pautke et al.12 reported on 132 patients treated between 2011 and 2019; after excluding 45 patients with BP-related ONJ, their analysis focused on 87 patients with DRONJ. In contrast, the present study evaluated the clinical characteristics and prognosis of DRONJ in a larger cohort of 178 patients treated over a 5-year period from 2019 to 2024.
Martins et al.16 evaluated treatment outcomes in patients with MRONJ and categorized outcomes into five stages: healed, improved, stable, worsened, and relapse. Varoni et al.19 further refined this system by dividing “healed” category into short-term healing and long-term healing, resulting in 6 outcome categories: short-term healing, long-term healing, stable, improved, worsened, and recurrence. In the present study, cases demonstrating unchanged or worsened disease stage after treatment were consolidated into a single category, “no healing,” because no clinically meaningful differences were observed between these outcomes. In addition, as no significant distinction was identified between short-term healing and improved outcomes, these categories were combined and analyzed as “partial healing.” Consequently, treatment outcomes in this study were classified into four categories: complete healing, partial healing, no healing, and recurrence. Treatment success was defined as the combined proportion of cases achieving complete or partial healing.
In present study, females (86.0%) were affected more frequently than males, consistent with findings from most previous studies9,19,20. Earlier reports indicated that multiple myeloma or metastatic cancers such as breast or prostate cancer, were the most common primary disease in DRONJ, accounting for 42%-62%9,19,20. In contrast, de Oliveira et al.21 indicated that osteoporosis and osteopenia were the most frequent primary disease (47%), with only 5.9% of patients having breast cancer. In our study, an even higher proportion of patients had osteoporosis as the primary disease (92.7%). This difference likely reflects the dosing indications: high-dose Dmab (120 mg) is typically used for oncological purposes, whereas low-dose Dmab (60 mg) is indicated for osteoporosis and osteopenia. Consequently, unlike previous studies, we observed a high rate of MRONJ associated with low-dose Dmab. Several factors may explain the differences between our study and prior reports. First, increasing age is a well-established risk factor for osteoporosis, and rising life expectancy has led to a growing number of osteoporosis patients. Second, in South Korea, Dmab was covered by insurance for osteoporosis treatment beginning in 2016, leading to an increase in Dmab use and a corresponding decrease in BP use. This shift is likely contributing to the rising number of ONJ cases related to Dmab.
The mandible is at higher risk for developing MRONJ due to its denser structure and lower vascularization, which hinders the healing process17. The posterior mandible is particularly susceptible, as it is exposed to higher masticatory forces, resulting in continuous remodeling of the alveolar bone21. In agreement with these previous studies17,21, DRONJ was more frequently observed in the mandible in our cohort.
Tooth extraction (45.5%) was the most common local factor associated with MRONJ development, consistent with previous reports20,22. This finding underscores the importance of avoiding tooth extraction whenever possible in patients receiving antiresorptive therapy. Other preceding events such as periodontitis, peri-implantitis, periapical inflammation, and dentoalveolar surgery, also appear to increase the risk of MRONJ, likely due to the potential for local infection. Preventive dental care prior to initiating antiresorptive therapy and ongoing oral hygiene management during treatment are therefore critical. Previous studies have shown that the implementation of routine preventive dental measures can reduce the incidence of ONJ23,24.
Previous studies have reported surgical treatment success rates ranging from 68% to 91.5%, which is comparable to the results of this study, in which the combined proportion of complete and partial healing was 85%9,10,12,18. The initial reports of MRONJ had discouraged surgical intervention and suggesting that spontaneous healing might occur7. However, more recent studies have demonstrated favorable outcomes with surgical management. Hoefert et al.10, reported complete healing in 80% of patients treated with major surgery. Pautke et al.12, reported surgical success rate of 91.5% in DRONJ. Similarly, our study demonstrates that surgical treatment is effective in DRONJ, with a comparable success rate of approximately 80.9%.
The recurrence rate in this study (12.9%) was lower than that reported by Varoni et al.19 (30.4%), and Mücke et al.25 (28.7%), but higher than the 9% reported by Pautke et al.12. Direct comparison among studies remains challenging due to methodological heterogeneity, variations in follow-up duration, and the fact that recurrence may occur several months after treatment, which can bias the observed rates.
In a previous study by Pautke et al.12, logistic regression analysis was conducted to determine whether factors such as time to MRONJ onset, sex, age, underlying disease, MRONJ stage, and precipitating events influenced treatment success; no significant association were identified. This finding is similar to the results of our study. However, in the present study, Dmab dosage was significantly associated with treatment success. It should be noted, however, that direct comparisons are complicated by differences in patient populations: patients receiving high-dose Dmab for cancer (e.g., breast or prostate cancer) may have poorer general health compared to those receiving low-dose Dmab for osteoporosis.
In this study, patients who switched from BP to low-dose Dmab developed ONJ after a significantly shorter period of Dmab treatment compared to other groups (P<0.001), suggesting that prior BP use may accelerate the onset of MRONJ. Interestingly, a previous study12 reported a faster onset of MRONJ in the Dmab-only group (2.0 years) than in the BP to Dmab switch group (4.1 years). Other reports also indicated that switching from BPs to Dmab can increase the risk of MRONJ and may accelerate its onset26-28. Among cancer patients with bone metastases, the risk of MRONJ was significantly higher when switching from zoledronic acid to Dmab29. The mechanism underlying this increased risk may be explained by pharmacokinetics: Switching from BPs to Dmab induces a rapid alteration in bone remodeling, which may contribute to increased susceptibility to MRONJ. Additionally, the cumulative duration of antiresorptive therapy, including both BP and Dmab administration, may further influence MRONJ risk28. In our BP to Dmab switching group, the mean duration of prior BP therapy was 56.7±46.4 months, which may explain the accelerated onset of DRONJ. However, a previous study by Hayashida et al.11 reported that prior BP administration did not appear to affect MRONJ pathogenesis, clinical characteristics, or treatment outcomes. Further research with larger patient cohorts is needed to clarify these findings.
In the current study, apart from differences in the time to DRONJ onset, no significant differences were observed among the three groups regarding MRONJ stage, preceding event, lesion location, treatment modality (conservative vs. surgical), healing status, or recurrence. These results suggest that switching from BP to Dmab may influence the timing of DRONJ onset but does not appear to affect disease severity or prognosis.
This study provides valuable information on the clinical characteristics of DRONJ but has several limitations. As a retrospective study, some data—such as duration of medication, triggering factors, and the circumstances at the time of onset (for example, whether having drug holiday or not)—were obtained from patient history, which may be subject to recall bias. Furthermore, because the medical records were not specifically designed for this study, variations in surgical technique among oral and maxillofacial surgeons and a lack of standardized data collection may have affected the results. Future well-controlled prospective studies should further investigate the relationship between drug holiday periods, preceding events, and MRONJ onset.
V. Conclusion
Overall postoperative healing was favorable 3 months after DRONJ treatment. Tooth extraction was the most frequent preceding events associated with DRONJ development. Although the duration of Dmab administration differed significantly between the low-dose Dmab group and the BP to Dmab switching group, prior BP use appeared to accelerate the onset of DRONJ. This earlier onset in the BP to Dmab group may be attributable to the cumulative duration of antiresorptive therapy. However, prior BP exposure did not appear to influence disease severity or treatment outcomes. Dmab dosage was the primary factor affecting treatment success, whereas age, treatment duration, lesion location, and MRONJ stage were not significant predictors.
XML Download