Denosumab-Induced Medication-Related Osteonecrosis of the Jaw (DRONJ): A 5-Year Retrospective Cohort Study

Denosumab (Dmab) has been suggested as a rst-line therapy for osteoporotic patients. However, a standardized protocol for the prevention of Dmab induced medication-related osteonecrosis of the jaw (MRONJ) has not yet been established. Thus, we investigated the factors that can affect Dmab induced MRONJ (DRONJ) to elucidate the relationship between invasive dental treatment and Dmab administration in patients who underwent Dmab and invasive dental treatment (especially tooth extraction) between October 2016 and March 2020. Four of the 98 patients developed MRONJ before and after tooth extraction. Two out of 4 patients developed MRONJ regardless of invasive treatment after Dmab administration and proceeded with extraction, and one patient developed DRONJ after Dmab administration and extraction. The other patient underwent a tooth extraction without osteoporosis treatment, and spontaneous DRONJ developed after Dmab administration. All MRONJ/DRONJ cases reported in this study show that MRONJ/DRONJ can develop as chronic inammation without invasive dental treatment, therefore, implementing preventive dental treatment before initiating Dmab treatment is necessary to reduce the likelihood of DRONJ. that can affect DRONJ to elucidate the relationship between invasive dental treatment and Dmab administration and report several cases of DRONJ at our institution.


Introduction
Bisphosphonates (BP) and the receptor activator of nuclear factor-κB (RANK) ligand (RANKL) inhibitor denosumab (Dmab) are the most common antiresorptive agents used in the treatment of osteoporosis [1][2][3] . Although Dmab has chie y been used as a second-line therapy after BP, it was recently suggested as a rst-line therapy for osteoporotic patients with a moderate or high risk of fracture 4 . However, Dmab must also be evaluated in light of the issues surrounding the use of bone resorption inhibitors, including BP. Particularly, medication-related osteonecrosis of the jaw (MRONJ) is one of the most controversial side effects of bone resorption inhibitors 5,6 .
MRONJ is de ned based on three criteria: 1) history of ongoing or prior treatment with antiresorptive or antiangiogenic agents; 2) exposure of the jaw bone or exposed bone that can be probed through an intra-or extra-oral stula in the maxillofacial region for >8 weeks; and 3) no history of radiotherapy or obvious metastases to the jaw 7 . Since BP-related osteonecrosis of the jaw (BRONJ) was rst reported in 2003, it continues to be of interest to dental clinicians and researchers 8,9 . However, due to indipendent reports of Dmab-related bone necrosis, the American Association of Oral and Maxillofacial Surgeons (AAOMS) revised the term BRONJ to MRONJ in 2014 7 .
Various hypotheses have been put forth to explain why bone necrosis occurring after the administration of antiresorptive agents such as BP or Dmab is limited to the jaws. However, there is insu cient evidence regarding the exact pathogenesis of MRONJ 10 . The most important systemic risk factor for MRONJ is administration of a powerful anti-resorptive agent, such as a nitrogen-containing bisphosphonate or a RANKL inhibitor, and local risk factors include dentoalveolar surgery (especially tooth extraction), dentures, and existing in ammatory dental diseases (e.g., periodontal disease) 7 .
In 2009, based on sparse data, the AAOMS recommended a 3-month break from oral BPs before and after invasive dental procedures for osteoporotic patients who have used oral BPs for >4 years with one exception. That being that the removal of acutely-infected teeth in patients taking oral BPs should proceed immediately under robust antibiotic coverage. However, the AAOMS also found no evidence that interrupting BP therapy alters the risk of ONJ in patients after tooth extraction 7,11 . However, given the lack of standard duration of the drug break for MRONJ prevention in osteoporotic patients with a history of Dmab therapy, there is a need for retrospective clinical cohort studies to develop prevention guidelines and protocols for MRONJ in these patients. Therefore, we investigated patients diagnosed with osteoporosis who were administered Dmab (Prolia ®) at the Department of Endocrinology at Severance Hospital, and underwent invasive dental treatment (tooth extraction) at the Department of Advanced General Dentistry and Oral and Maxillofacial Surgery at Yonsei University Dental Hospital. Accordingly, the aim of this study was to investigate the factors that can affect DRONJ to elucidate the relationship between invasive dental treatment and Dmab administration and report several cases of DRONJ at our institution.

Ethics statement
This study protocol was approved by the Institutional Review Board (IRB) of the Yonsei University Dental Hospital (approval number: 2-2020-0071). Deidenti ed participant data were used and written informed consent was waived because the design of the study was retrospective study. This study was performed in accordance with the Declaration of Helsinki. therapy for diseases other than osteoporosis (e.g., hypercalcemia of malignancy, solid cancers, bone metastases, giant cell neoplasm, or multiple myeloma), and a history of Dmab or invasive dental therapy at a different institution, preventing accurate evaluation.
Overall, 98 (189 teeth) of the 159 patients initially screened were included in the study, after excluding 61 patients (108 teeth).

Variables
Participants' age, sex, oral condition, underlying disease, BP administration history, type of drug administered after tooth extraction, period from drug cessation to extraction, period from extraction to drug initiation, location of extraction (maxillary/mandibular/multiple) and presence or absence of MRONJ/DRONJ were investigated retrospectively.
For patients with MRONJ/DRONJ, the number of Dmab administrations, date of extraction, date of diagnosis, location, staging, associated local factors, and the type of MRONJ/DRONJ treatment (conservative/surgical) were additionally investigated. The prescription record was referred for information on Dmab administration, and electronic medical records were referred for information on the invasive dental treatment performed.

Statistical analysis
Patient demographics are expressed as n (%) and mean ± standard deviation. The χ 2 test was used to compare proportions across levels of categorical variables. Due to the low incidence of MRONJ/DRONJ, 40 cases were randomly sampled to obtain stable results. To analyze possible associations between affecting factors with the development of MRONJ/DRONJ, we used a two-tailed Fisher's exact test. A two-tailed P-value <0.05 was considered statistically signi cant for all analyses. All statistical tests were performed using SPSS statistical software (SPSS for Windows, version 25; SPSS Inc., Chicago, IL, U.S.A).

Baseline characteristics
The baseline characteristics of the 98 participants surveyed in this study, including the average age, sex, and the oral condition of the patients before tooth extraction are presented in Table 1. The participants' age ranged from 36 to 91 years, with an average age of 70.5±10.3 years and the highest percentage of people in their 70s. There were 87 women (88.8%). Residual roots, dental caries and other endodontic lesions were the main reasons for tooth extraction, followed by periodontitis, fracture/crack, and impacted/supernumerary tooth. Two patients ( ve teeth) underwent dental extraction due to osteonecrosis of the jaw. Fracture / Crack 12 (11.8) / 13 (6.9) Impacted / Supernumerary tooth 6 (5.9) / 9 (4.8) Values are n (%), mean (range), as indicated.
In BP + Dmab, 16 (34.0%) of the 47 patients received Dmab after tooth extraction, 4 (8.5%) with BP, 6 (12.8%) with Serm, 1 (2.1%) with Teriparatide, and 20 patients (42.6%) who did not receive any drugs. The mean period from drug cessation to extraction was 6.9 ± 4.8, 32 patients (68.1%) within 6 months, 12 patients (25.5%) of 7-13 months, and 3 patients (6.4%) of 14-21 months. The mean period from tooth extraction to drug initiation was 1.9 ± 1.  Case description of MRONJ/DRONJ patients Four patients were diagnosed with osteoporosis in this study population. Following BP and Dmab therapy, MRONJ/DRONJ developed before extraction in cases 1 and 2, and after extraction in case 3. In case 4, the patient had no history of osteoporosis treatment at the time of extraction, and developed DRONJ afterwards, following Dmab administration 1 year after extraction (Fig. 1). Table 4 summarizes the MRONJ/DRONJ clinical cases for cases 1-4. tooth mobility due to localized chronic advanced periodontitis, and extraction was planned to be performed after ≥3 to 6 months. Conservative treatment was performed before extraction, but MRONJ was diagnosed in the right mandibular region 6 months after the last Dmab administration. The drug was changed to Calcitriol [0.25 mcg/ Soft Cap], and extraction and sequestrectomy were performed after lesion localization ( Supplementary Fig. S1 online).

Case 2
A 76-year-old woman had stable angina pectoris and coronary artery disease, apart from osteoporosis, was treated with endodontics and prosthesis for a crown and root fracture that had occurred approximately 10 years prior. From January 2007 to May 2017, she was administered alendronate and ibandronate, and in August 2017, she was administered zoledronic acid (5 mg/100mg). One year later, Dmab was administered once. Six months thereafter, MRONJ occurred in the left maxillary region in the presence of localized chronic advanced periodontitis, and bone necrosis was observed up to the mesial root site from the rst premolar to the second molar in the radiographic image. Approximately 4 weeks later, tooth extraction and sequestrectomy of the affected area were performed, followed by prosthetic rehabilitation of the missing teeth ( Supplementary Fig. S2 online).

Case 3
A 54-year-old woman had diabetes mellitus, osteoporosis, and a history of ibandronate (Bonviva [3 mg/ syringe]) treatment at a different hospital in 2016. In April 2019, Dmab was administered once. Within a month, the mandibular right second molar had a hopeless prognosis due to a fracture in the tooth, and the tooth was extracted. Three months after extraction, necrotic bone exposure (MRONJ) was found on the lingual side of the extraction site. With Dmab clearing up around October 2019, the area was observed during follow-up and treated conservatively ( Supplementary Fig. S3 online).

Case 4
A 79-year-old woman with underlying diseases such as hypertension, diabetes mellitus, and acute myocardial infarction, apart from osteoporosis, did not undergo any osteoporosis treatment before tooth extraction. With the removal of partial dentures used >10 years, the mandibular anterior teeth had severe mobility and periodontic-endodontic lesions. The teeth were extracted in July 2018. Dmab was administered twice 11 months after the extraction, and bone loss and increased sclerosis were observed in the anterior mandible in April 2020, 10 months after Dmab administration. Conservative treatment was performed by changing Dmab by raloxifene (60 mg/T) in June 2020 ( Supplementary Fig. S4 online).

Discussion
Various factors have been implicated as causes of MRONJ, but recently, it was suggested that anti-resorptive agents affect the immune function of the bone.
Tooth extraction has been reported as a predisposing factor of MRONJ in approximately 45-61% cases, but the prevalence and incidence rates of other spontaneous occurrences, which are the second largest factor and present as periodontal or periapical lesions, implants, or dentures, as well as extractions, have not been reported 12,13 . In animal studies, lesions similar to ONJ have been reported in cases of periapical in ammation and periodontitis 7,14 . However, there is still a lack of clinical research on Dmab-induced MRONJ and its incidence, and no protocol has been established yet for invasive dental treatment of the patients receiving this drug.
In this study, we reported four cases of MRONJ/DRONJ, out of 98 patients who received either only Dmab injections or with BP and who underwent tooth extractions. Although all four cases have something in common, i.e., they are cases of MRONJ/DRONJ occurring after Dmab administration, each situation is distinct. Three of the four DRONJ cases (Cases 1-3) had a history of BP treatment. As for predisposing events, Cases 1-3 can be considered tooth extraction due to periodontal disease, and case 4 as a natural occurrence due to an existing pathogen. Over time, case 1 and 2 resulted in MRONJ due to worsening of the chronic in ammation of the existing lesion after the Dmab administration, regardless of tooth extraction. This means that chronic active periodontitis may have been a signi cant risk factor in the two cases, suggesting the importance of preventive dental treatment. For the MRONJ related to an existing lesion, the prevalence rate of BP-induced MRONJ is 28.6%, but that for Dmab-induced MRONJ has not yet been reported 15 .
Several hypotheses have been raised regarding the transformation of existing chronic in ammatory lesions into MRONJ, and an M1 macrophage shift is considered the most likely explanation. Kang  In Case 3, the extraction was performed one month after the administration of a single Dmab dose following ibandronate injection. DRONJ developed three months after surgery. The lesion was classi ed as grade 2 DRONJ, and cured with a ve-month conventional treatment. This case suggests a one month time between the Dmab injection and extraction may be too short.
The patient in Case 4 had no history of BP administration, and the development of DRONJ was "non-identi ed" after two administrations of Dmab a year after extraction. This non-identi ed occurrence is peculiar in that the possibility of a pathological fracture in the jawbone during Dmab treatment cannot be ruled out. The most common causes of MRONJ are tooth extraction and obvious periodontal disease, followed by "non-identi ed" occurrences with no identi able cause 15 . This led to the consideration of systemic or local causes that could have other effects other than on the onset. Nevertheless, an MRONJ case associated with a predisposing event (i.e., "non-identi ed") is the third most common group of MRONJ cases reported in the literature, with an estimated prevalence between 16-70% 32 . Most of these non-identi ed occurrences were found to occur in relation to the lower posterior tooth region, but the lesion described in Case 4 developed anteriorly. Importantly, a non-identi ed MRONJ occurrence poses the risk of a pathological fracture of the jaw after Dmab administration.
This study provides information on the incidence of MRONJ/DRONJ in patients with Dmab but has limitations. First, it was a retrospective study using data from a single institution. In addition, bone turnover markers (BTM) were not investigated because all participants' BTMs were not measured at the same time. However, clinical data only makes it di cult to predict the occurrence of DRONJ and requires another predictive factor, therefore molecular biomarkers need to be included in future research.
In conclusion, the incidence of DRONJ after Dmab administration was 4.1%, and the timing of drug administration and release varied in our study population.
DRONJ can occur as a chronic in ammation without invasive dental treatment, therefore it is necessary to reduce its likelihood by implementing preventive dental treatment before Dmab treatment.