Medication-related osteonecrosis of the jaw (MRONJ) is a debilitating oral condition characterized by the exposure of necrotic jawbone, typically occurring in patient on certain medications, such as bisphosphonates (BPs) [1]. It is precipitated by dental surgery but occasionally manifests spontaneously. While this disorder has multiple etiologies, its association with BP medication is of particular concern, especially given the widespread use of BPs for managing osteoporosis and cancer-related conditions like metastasis to the bone [2]. In the management of osteoporosis, BPs, selective estrogen receptor modulators (SERMs), and denosumab are commonly utilized as first-line pharmacological agents. These medications are frequently cited in multiple clinical guidelines pertaining to the treatment of osteoporosis [3].

One of the most significant adverse effects associated with BPs is osteonecrosis of the jaw (ONJ). Notably, ONJ related to surgical dental procedures, such as tooth extraction, periodontal treatment, and implantation, is a well-documented complication [4]. Additionally, individual comorbidities and lifestyle factors, such as steroid use, cancer, diabetes, and smoking, elevate the risk of ONJ occurrence [5].

As the aging population continues to grow, the prevalence of osteoporosis is also on the rise, leading to an increase in the number of individuals prescribed BPs. The incidence of ONJ among those taking BPs has been reported to range from 0.05% to 0.21% [6]; however, large-scale cohort studies to accurately determine this incidence rate have been lacking. In an aging population where the demand for dental treatments such as implantation is increasing, the potential for complications like ONJ poses a risk of diminished quality of life [7].

In particular, some epidemiological evidence exists regarding the increased risk of ONJ among BP users based on the duration of medication use, sex, and age. However, studies specifically examining the severity of ONJ related to medication use have been lacking. Therefore, we aimed to ascertain the incidence and risk of ONJ among patients with osteoporosis prescribed with BPs and SERMs. We provided data on the incidence rate and risk of these patients stratified by age groups, medication types, and duration of use. Additionally, we presented data on the severity outcomes of MRONJ based on dental treatments.

Among the 61,183 osteoporosis patients aged 40 years and above who were confirmed to have been prescribed osteoporosis medication, 5,537 were male and 55,646 were female, indicating a female majority. When examined by age group, there were 3,903 individuals aged 40 to 49 years, 14,945 in their 50s, 22,465 in their 60s, and 19,870 aged 70 years and above. There were 52,743 were taking BPs, 3,101 patients taking SERMs, and 5,339 were taking both types of medications (Table 3).

The occurrence of ONJ was examined based on sex, age group, and whether osteoporosis medication was administered (Table 4). The incidence of ONJ was significantly higher in the osteoporosis medication group (RR, 3.81; 95% confidence interval [CI], 2.66–5.48), particularly among females with an RR of 4.99 (95% CI, 3.21–7.74). In the 40s age group, the RR was relatively low at 1.44 and was statistically insignificant (95% CI, 0.26–7.84).

The time from the initial medication administration to the occurrence of ONJ was analyzed during the study period. In both the BP and SERM groups, the majority of cases occurred after 5 years, with 74 cases (51.0%) for the BP group and six cases (42.8%) for the SERM group (Table 5).

In the treatment of ONJ, we examined both invasive and noninvasive surgical treatment methods (Table 6). Among the group that received osteoporosis medication, invasive treatment was more common, accounting for 127 cases (71.3%). In contrast, in the group that did not receive medication, invasive treatment was relatively less common, with 20 cases (55.6%), although this difference was not statistically significant. When examined by medication type, invasive treatment was performed in 69.9% of the BP group, 85.7% of the SERM group, and 75.0% of the group receiving both medications.

MRONJ is considered a rare condition due to its low incidence rate, and its treatment and prognosis are often poor, necessitating prevention and caution. The condition involves complex prescriptions and treatments across multiple specialties, including internal medicine, orthopedics, dentistry, oral and maxillofacial surgery, and plastic surgery. While it is challenging to manage solely within one specialty, there is room for prevention if healthcare providers from any specialty take the risk of MRONJ into account and communicate appropriately with colleagues from other departments. Through this study, we confirmed the incidence rate of MRONJ among patients with osteoporosis who were taking BPs and assessed its severity based on the invasiveness of the treatment methods. While epidemiological studies on the relationship between BP use and MRONJ are somewhat known, this is the first study, to the best of our knowledge, that evaluates the severity of MRONJ based on medication use in a large-scale cohort.

The prevalence of osteoporosis is increasing and is currently known to be 18.3% worldwide [9]. Elderly women are specifically at risk. Over 90% of patients with osteoporosis are women, particularly those older than 50 years. The demand for dental treatment also increases with age. Consequently, it was hypothesized that the incidence of MRONJ would also increase with age among those taking osteoporosis medications. This study confirmed epidemiological evidence to support this. Given that over 90% of patients with osteoporosis are women and that BP prescriptions are also more common among female patients, it was observed that the incidence of MRONJ is higher in women. Specifically, when examined by sex, it was confirmed that women are more vulnerable to MRONJ while on medication than men, which contrasts with previous studies that suggested a higher incidence of MRONJ among men taking BP compared to women [10,11].

A significantly higher incidence of MRONJ was observed in the group taking osteoporosis medications compared to those who were not. The incidence rate of MRONJ was found to be 0.17%. The incidence rate of MRONJ among patients with osteoporosis varies from study to study; for instance, a study in Japan reported an incidence rate of 0.06% [12]. This finding is consistent with previously known incidence rates, lending credibility to the reliability of this sample cohort.

We aimed to examine the differences between the BP group and the SERM group as a control. Interestingly, the incidence rate of MRONJ in the SERM group was found to be at the same level as that in the BP group. While MRONJ is generally not well-known to occur in patients taking SERMs, it has been reported to occur even in those receiving SERMs alone [13,14]. However, the association between MRONJ and SERMs should be carefully interpreted. The number of MRONJ cases should be considered as it is a relatively small number compared to that of the BP group. In addition, we observed MRONJ cases only in patients with osteoporosis. This may have resulted in increased cases in the SERMs group, not in breast cancer or other SERM users. Although the results could be interpreted as an effect of past BP use in the SERM group, it suggests the need for comprehensive guidance on the risks of MRONJ for patients with osteoporosis undergoing drug treatment, including those in the SERM group.

There are several limitations in this study. First, the representativeness of the data extracted from the sample cohort was limited as it was secondary data. Although the data from the NHIS claims to cover the entire population, the 2% of the allegedly entire population uniformly extracted may still have lacked representativeness. Second, securing statistical significance was challenging due to the rarity of MRONJ. Lastly, while it would have been ideal to evaluate the severity of MRONJ through staging in actual clinical settings, this was not feasible due to poor staging and limitations in accessing medical records. Therefore, we utilized a research method based on treatment method names.

The evaluation concerning the severity and stage of MRONJ in clinical setting was as follows [15]. Stage 1 patients had exposed bone and were asymptomatic with no localized soft tissue infection. Stage 2 patients had exposed bone, pain, and regional soft tissue inflammation or infection. Stage 3 patients had exposed bone with associated pain, localized soft tissue inflammation (or secondary infection), pathological fracture, and extraoral or oral-antral fistulas. Radiographically, the bone showed osteolysis extending to the inferior mandibular border or maxillary sinus floor. However, those staging systems rarely applied to the patient, and the stage information was not available in NHIS.

Despite these limitations, we were able to utilize nationwide data accumulated over a long period in this study. Particularly, we evaluated the severity of MRONJ by using treatment methods as variables. Future research should focus on specific populations such as patients with osteoporosis and cancer and further explore the relationship between medication use and the location and severity of MRONJ.