Introduction
One of the risk factors for osteonecrosis of the jaw (ONJ) is the use of antiresorptive agents for osteoporosis and metastatic bone tumors, and medication-related ONJ is termed MRONJ [
1]. MRONJ includes bisphosphonate (BP)-related ONJ (BRONJ) and denosumab (Dmab)-related ONJ (DRONJ). The American Association of Oral and Maxillofacial Surgeons (AAOMS) stated that the risk for MRONJ among osteoporosis patients with BPs was 20–50 per 100,000, and the risk for MRONJ among osteoporosis patients with Dmab was 40–300 per 100,000 [
1]. In addition, the risk for MRONJ among cancer patients with high-dose BP ranges from 0 to 18,000 per 100,000, and the risk for MRONJ among cancer patients with high-dose Dmab ranges from 0 to 6900 per 100,000. The incidence of ONJ is greatest in the cancer patient population, in which high-dose BPs/Dmab are used at frequent intervals, and the incidence of ONJ in the osteoporosis patient population has been estimated to be marginally higher than the incidence in the general population [
2]. However, there are few publications on the incidence ratio of ONJ among medications (high-dose BPs/Dmab, low-dose BPs/Dmab, and no antiresorptive agents).
We started a population-based study entitled “KureDREAMS” (Kure Data-based Results and Evidence Assisted by a Multiprofession Study) in 2015, which aims to prevent osteoporosis, fracture, and osteoporosis-related disorders in collaboration with multiple professions, including physicians, dentists, medical staff, medical societies, and local government in Kure city. Kure city is a regional core city with a population of approximately 210,000 located in Hiroshima, Japan. The percentage of the elderly population aged 65 years and older was 35.4% as of 2020. We launched the project to prevent osteoporosis and fractures among elderly individuals in 2014, and at the same time, medical and dental liaisons were strengthened to prevent and detect ONJ. The ONJ registry was launched in 2015, and the number/trend of ONJ occurrences has been investigated since then.
There are three core hospitals with oral surgeons in this city, and almost all ONJ cases are referred to those three hospitals. In addition, since residents in Japan are obliged to take out medical insurance, the medical claims data are administered by the city and can be examined to ascertain the medical practices that received medication in the area [
3,
4]. Using this claims database and the ONJ registry data, we determined the number of ONJ occurrences in each year and the number of patients prescribed antiresorptive agents (high- and low-dose BPs/Dmab). The purpose of this study was to estimate the ONJ incidence and trend among patients who used antiresorptive agents by dosage and people who did not use these agents in Kure and its trend from 2016 to 2020.
Discussion
We established the registration system for ONJ in 2015 and consecutively enrolled 98 eligible ONJ patients between April 2016 and March 2021 at three core hospitals in Kure city. The incidence of high-dose MRONJ was 2305.8 (range, 1775 to 3049) per 100,000 over the 5-year period, that of low-dose MRONJ was 132.5 (range, 88 to 234) per 100,000, and that of ONJ without antiresorptive agents was the lowest at 5.1 (range, 0 to 8.7) per 100,000. The strength of this study was in determining the incidence of ONJ by the dose of each antiresorptive agent based on the ONJ registry, in which precise diagnoses were made by ONJ specialists for residents insured by two kinds of insurance systems.
The results of a nationwide population-based study in Japan on ONJ were reported by Ishimaru et al. [
7] in 2022. The prevalence was 0.06%, and the incidence rate was 22.9 per 100,000 person-years among patients with osteoporosis; the prevalence was 1.47%, and the incidence rate was 1231.7 per 100,000 person-years among patients with cancer [
7]. Their and our reported ONJ incidences appeared to be roughly consistent with the results presented by the International Task Force and an AAOMS position paper published in 2022 [
1,
2]. However, the ONJ incidence in our investigation is likely to be higher than that reported by Ishimaru et al. [
7] for both high- and low-dose antiresorptive agents. There were several reasons for these differences, one of which would be the difference in the way ONJ cases were extracted: their data were based entirely on the claims database [
7], whereas we directly counted those cases diagnosed by dentists using the ONJ registry. Our study might provide an almost complete count of ONJ cases that occurred over a 5-year period, while the incidence by Ishimaru et al. [
7] may be an overestimate/underestimate because the survey was based on the claims database, in which the definitions of ONJ differ between each hospital. In addition, the nationwide population-based study included only patients who had newly begun using antiresorptive drugs, while our study did not take into account the duration of medication use, so the proportion of patients who took the drug for longer periods might be higher in our study. Taking antiresorptive drugs for longer periods (> 4 years) was reported to be a risk factor for ONJ occurrence [
1,
8,
9], which may be one of the reasons for the higher incidence of ONJ in our study.
The KureDREAMS project on ONJ prevention in Kure city, which began in 2015, has led to early detection of ONJ [
10] and might be one of the reasons for the high incidence rate of ONJ in Kure city. The central committee for the adequate treatment of osteoporosis (A-TOP) research group has recommended a forum to share information about ONJ among medical professionals, dentists, and patients [
11]. The Kure City Dental Association and the three major oral surgery practices in Kure city work closely together, as well as among dentists, making it easy to detect cases of ONJ early. In addition, the Kure City Dental Association regularly disseminates information on MRONJ and osteoporosis through workshops and the media as medical and dental liaisons. However, further investigations are needed to determine whether such collaboration will truly lead to the early detection of ONJ in the future. The ONJ registry started in Germany when cases of low-dose BRONJ were reported in 2004 [
12], and medical-dental liaisons and patient education began [
13]. However, there were no investigations showing a trend in ONJ incidence, and to the best of our knowledge, the results of our study are the first report of secular trends after the launch of the ONJ registry. The KureDREAMS project on ONJ prevention in Kure city, as in other countries and in the future, is expected to detect earlier stages of MRONJ and to reduce its incidence rate.
Osteoporosis patients and people without antiresorptive agents showed an upward trend in ONJ incidence, which may be due to the increased interest in ONJ since the ONJ registry was launched in Kure city. On the other hand, the incidence of ONJ from 2017 to 2020 showed a downward trend among cancer patients in Kure city, although this was not statistically significant; the incidence peaked in 2017 and declined from 3048.8 per 100,000 to 1775.1 per 100,000. One potential interpretation is that medical-dental liaisons have progressed because of the introduction of the “perioperative oral function management fee,” which began in 2012 to reduce complications after cancer surgeries. Thus, the incidence of MRONJ among osteoporosis patients would also decrease in the future, as medical and dental liaisons have been strengthened to prevent and detect ONJ since 2015.
To the best of our knowledge, the incidence of ONJ in the population without antiresorptive agents based on the registration system adopted in our study was the first report worldwide and was higher than we expected. Since the reason for ONJ without antiresorptive agents was mostly periapical periodontitis in our study (data not shown), continuous bacterial infection of the jaw by inadequate oral hygiene management was considered to be the major cause in these cases, as previous studies have shown [
14]. Poor hygiene management factors in Kure city included the following: patient factors (lack of interest in oral care), dental factors (worsening of dental disease due to delay in tooth extractions), and geographical factors (Kure city contains many small islands whose residents have limited means of transportation). In Japan, the percentage of edentulous individuals is low among countries worldwide because of the custom of trying to preserve teeth [
15]. The number of extracted teeth is reported to be a risk factor for ONJ occurrence [
7]; however, the cause of ONJ would be continuous bacterial infection of the jaw preceding tooth extraction and not tooth extraction itself. Therefore, we consider striving to preserve as many healthy teeth as possible through regular oral hygiene maintenance to be important.
The International Task Force on Osteonecrosis of the Jaw mentioned that the incidence of osteoporosis-related ONJ was only slightly higher than that observed in the general population, but it did not indicate actual ONJ incidence in the general population [
2]. ONJ in the population that does not take antiresorptive agents is rare, but the number of reported cases has increased in recent years [
16]. However, systematic reviews of ONJ related to nonantiresorptive medications were performed among cancer patients receiving nonantiresorptive agents, not community residents, as we performed in this study [
17]. In our study, the ONJ incidence in the population without antiresorptive agents was 5.1 per 100,000, and the incidence of osteoporosis-related ONJ was approximately 24 times higher than that of ONJ in the population. This was also the first report to simultaneously compare the incidence of ONJ among cancer and osteoporosis patients and the general population, clearly demonstrating the extent to which the administration of antiresorptive agents increases the risk of developing ONJ. This study suggested that ONJ prevention is necessary when using antiresorptive agents, and it would be interesting to see how much the incidence of ONJ can be reduced through collaboration for ONJ prevention in Kure city in the future.
The population without antiresorptive agents in this study mainly consisted of those without osteoporosis and partly included untreated osteoporosis patients and treated osteoporosis patients with medications except antiresorptive agents, of whom approximately 6.5% used active vitamin D metabolite medication (oral alfacalcidol, eldecalcitol, and calcitriol) only. Active vitamin D is prescribed for osteoporosis, chronic kidney disease, hypoparathyroidism, and other disorders of vitamin D metabolism in Japan. Although other indications besides osteoporosis might be included in the population who used vitamin D metabolites, we considered that it would be mostly osteoporosis patients. Two previous large observational databases suggested that osteoporosis itself, rather than antiresorptive agents, is a risk factor for ONJ [
18,
19], and thus, we performed an additional subanalysis in the population with osteoporosis. The ONJ incidence rate was approximately 30 times higher among osteoporosis patients using low-dose antiresorptive agents than among osteoporosis patients using vitamin D metabolites (Table
5). An AAOMS position paper in 2022 stated that the risk of ONJ among patients with Dmab was almost an order of magnitude higher than that for patients with BPs [
1]. However, no difference in the ONJ incidence was found between patients with BP and Dmab in Kure (Table
5). Our study selected osteoporosis patients who used vitamin D metabolites as the control group, whereas a previous study selected osteoporosis patients who used SERMs as controls [
18]. The review organized by the European Calcified Tissue Society (ECTS) stated that there is weak evidence for the role of vitamin D deficiency in MRONJ development [
16]. Active vitamin D medication might have prevented ONJ development in our study. No ONJ cases developed among osteoporosis patients treated with TPTD, ROMO, or SERMs.
In our study, the number of prescriptions for high-dose Dmab increased, while the number of prescriptions for high-dose BPs decreased. There are potential reasons for the increased number of high-dose Dmab prescriptions: (1) there have been reports showing that denosumab is significantly more effective against skeletal-related events than zoledronate [
20]; (2) denosumab is able to be used for cancer patients with severe renal dysfunction; and (3) denosumab is injected subcutaneously, which reduces the burden on the patients compared to intravenous zoledronic acid. For the above potential reasons, Japanese surgeons and physicians have tended to prefer high-dose Dmab over high-dose BPs for cancer patients in recent years. In contrast, since an AAOMS position paper in 2022 reported that the risk for MRONJ among cancer patients exposed to high-dose Dmab is comparable to the risk of MRONJ among cancer patients exposed to high-dose BP [
1], the trend in the number of prescriptions for high-dose Dmab/BP would not have affected the trend in ONJ incidence in Kure city.
There were several limitations in our study. The first was that ONJ cases in our study were identified at three facilities in Kure city; thus, ONJ cases diagnosed at other facilities were not counted. However, only these three hospitals have oral surgeons in this city, and patients with potential ONJ are referred to these three hospitals, even if the patients first visit a dental clinic or another hospital. Second, our study was based on local claims data and the ONJ registry in Kure city, which might not be representative of the incidence and overall trend in Japan. Third, since patients who had at least one prescription during the fiscal year were defined as “having a prescription,” this might lead to an underestimation of the MRONJ incidence due to the larger denominator value. Fourth, it is known that MRONJ in the osteoporosis population emerges after more than 3 or 4 years of BP exposure [
1], and MRONJ incidence increases with time of exposure to BPs. The database included individual data that were unavailable, and thus, it was difficult to analyze individual information considering the duration of BP use, which was a limitation in our investigation.
In conclusion, we investigated the incidence and trend of ONJ by dose of antiresorptive agents over a 5-year period in Kure city. The annual incidence of high-dose MRONJ was 2305.8 per 100,000, that of low-dose MRONJ was 132.5 per 100,000 and that of ONJ without antiresorptive agents was 5.1 per 100,000, which was higher than those reported in previous studies. The ONJ incidence ratio was 23.6 among the osteoporosis patients using low-dose antiresorptive agents and 420.6 among the cancer patients using high-dose antiresorptive agents compared with the population that did not use antiresorptive agents. This was the first report to simultaneously assess the incidence of ONJ with antiresorptive agents based on the ONJ registry. Overall, the MRONJ incidence increased from 2016 to 2020, but the incidence of high-dose MRONJ seemed to decrease, although the difference was not statistically significant. This fact-finding investigation began with efforts to prevent the onset of ONJ, and we should continue this work for the early detection and prevention of ONJ in the future.
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.