Introduction
Medication-related osteonecrosis of the jaw (MRONJ) causes destructive necrotic invasion and extensive bone defects in the jaw. After the removal of necrotic tissue, it becomes difficult to plan oral reconstruction using implants and prosthetics due to the defect of the jawbone. To address this issue, surgical treatment for MRONJ has been accompanied by the use of adjuvant therapy to support bone regeneration, and treatment methods have been established. Recently, studies have reported the application of recombinant human bone morphogenetic protein-2 (rhBMP-2) to surgical sites to promote bone regeneration.
As a subgroup of the transforming growth factor-β (TGF-β) family, bone morphogenetic proteins (BMPs) are known to play a direct role in bone remodeling by regulating osteoblast differentiation and osteoclast activity, while also possessing strong osteoinductive and osteoconductive properties. Wikesjö et al. [1] reported that bone formation was significantly enhanced in bone defect areas treated with rhBMP-2 and an absorbable collagen sponge (ACS). In another study, Min et al. [2] assessed bone healing using the radiographic index, which is the ratio of the radiographic density between the defect and surrounding areas. They reported that the experimental group (rhBMP-2/ACS) showed an 11.4% higher value compared to the control group (non-rhBMP-2/ACS).
However, due to its hydrophilicity, rhBMP-2 rapidly diffuses into surrounding tissues from the defect area, making sustained action at the target site difficult. Using rhBMP-2 in combination with optimal carriers such as ACS or a fibrin matrix helps to prevent diffusion into surrounding tissues due to its hydrophilicity and allows for more prolonged and stable application at the bone defect sites in MRONJ patients.
In this study, we retrospectively analyzed data over a 10- year period for 45 of the 67 patients who visited our hospital since 2016, were diagnosed with MRONJ, and received surgical treatment, with follow-up observations for more than 6 months. We assessed the efficacy of rhBMP-2 at bone defect sites by measuring the volumetric changes in bone using facial computed tomography (CT) immediately after surgery and 6 months postoperatively. There are very few reports on the bone formation rate when BMP is applied in patients with osteomyelitis. Therefore, this study aims to analyze the bone regeneration rate and the factors influencing bone regeneration 6 months postoperatively in patients diagnosed with MRONJ at our department of oral and maxillofacial surgery, who underwent decortication and sequestrectomy, and had rhBMP-2 and ACS applied simultaneously to the bone defect site.
Unlike earlier studies that mostly involved animal models, mixed patient groups, or short-term assessments, this study provides a detailed volumetric evaluation of bone regeneration in a carefully selected group of stage 3 MRONJ patients, using facial CT data collected over at least 6 months. All surgical procedures were performed by one experienced surgeon to reduce procedural differences. Bone defect volume was measured using a three-axis alignment technique to account for the irregular shapes of the lesions, and various clinical and anatomical factors were analyzed using thorough multivariable statistical methods. With these approaches, the study offers a more accurate and clinically meaningful assessment of rh- BMP-2 therapy in MRONJ, clearly distinguishing its findings from those of previous research.
Materials and Methods
1. Patients
This retrospective study was conducted on patients diagnosed with MRONJ at the Department of Oral and Maxillofacial Surgery, Sun Hospital, Daejeon, between March 2016 and December 2024. The study was approved by the Institutional Review Board of the Daejeon Sun Hospital (no. DSH-인-25-03). Informed consent was obtained from all participants. Among patients diagnosed with MRONJ, only those classified as stage 3 and requiring decortication and sequestrectomy were included, whereas patients with stage 2 or lower were excluded. All included patients underwent decortication and sequestrectomy, and an ACS was applied as a carrier for rhBMP-2 to promote bone regeneration. Among the 67 patients who underwent decortication and sequestrectomy, only 45 patients who were followed up for up to 6 months postoperatively and underwent facial CT were selected as the study population (45 patients; average age 79.44 years). Since 2016, decortication and sequestrectomy followed by application of rhBMP-2 has been performed in the majority of patients at this institution. Consequently, establishing a control group was not feasible. All surgeries were performed by a single experienced oral and maxillofacial surgeon.
2. Material and methods
In this study, after performing decortication and sequestrectomy at the lesion site of patients with MRONJ, rhBMP-2 was applied and sutured using an ACS as a carrier. The volume of the lesion was measured through CT analysis immediately after surgery and 6 months postoperatively, and the bone regeneration rate was assessed accordingly. Based on these results, the statistical significance of each independent variable on the recovery rate of the alveolar bone at the defect site was analyzed.
The bone regeneration rate was calculated as the percentage of newly formed bone relative to the defect volume (or area) immediately after surgery. First, the volume of the lesion was measured using CT images taken immediately after surgery and at 6 months postoperatively. The volume of the bone defect was calculated by multiplying the maximum length measured along the Pitch, Roll, and Yaw axes on 1-mm-thick CT slices. Because the lesions had irregular shapes, exact volume calculation was difficult; therefore, bone defects were assessed as proportional values to compensate for this limitation. CT images were analyzed using Analyze software (Analyze Direct) to identify anatomical structures and to ensure consistent alignment of the lesions along the three axes (Pitch, Roll, Yaw), allowing measurements to be obtained at identical angles across both time points. Subsequently, the Viewrex 3.0 (TECHHEIM) program was used to measure the length of each axis (Pitch, Roll, Yaw) of the lesion, and the volume was calculated (Fig. 1). The axes (Pitch, Roll, Yaw) were labeled as a, b, c immediately post-surgery, and A, B, C at 6 months postoperatively. The bone regeneration proportion was then calculated based on the ratio of volume differences between the two periods. Here, a and A represent the length of the lesion along the Pitch axis, b and B along the Roll axis, and c and C along the Yaw axis, respectively.
3. Statistical analysis
After performing statistical analysis, a multiple linear regression model was used to assess the significance of the relationship between independent variables―such as the number of bony walls, age, and initial size of the alveolar bone defect ―and the dependent variable (bone regeneration proportion). IBM SPSS Statistics (IBM Corp.) was used for the analysis, and a p-value of less than 0.05 was considered statistically significant. The study included 45 patients with MRONJ to assess bone regeneration. Since no control group was established, the required sample size was determined based on the precision of the estimated mean. Assuming a 95% confidence interval (CI) with a margin of error of ± 15%, a minimum of 41 patients was required. Therefore, the inclusion of 45 patients was considered sufficient to ensure reliable estimation. The margin of error of ± 15% was selected to interpret the magnitude, trend, and effect of bone regeneration attributable to rhBMP-2. This threshold was determined in consideration of the biological variability typically observed in rhBMP-2-mediated bone regeneration, as reported in previous studies. By allowing for this margin, clinically meaningful changes and trends can be identified and interpreted, even in the presence of variability inherent to bone healing processes and patient responses.
4. Surgical protocol
After performing decortication and sequestrectomy at the MRONJ-affected site, fibrin glue (Tisseel; Baxter Healthcare Corporation) was applied to the base of the bone defect area to serve as a barrier, prevent the diffusion of BMP into surrounding tissues, and promote hemostasis. ACS (P-Stop Advance; Hyundai Bioland) was loaded with 0.25 g of rhBMP-2 (rhBMP-2; Novosis Dent) and grafted into the bone defect. Tisseel was again applied over the graft site to prevent the further diffusion of BMP. For protection of the surgical area and additional prevention of BMP diffusion into adjacent tissues, an absorbable membrane such as Xenoguide (NIBEC), or Biocover (Purgo) was applied (Figs. 2 and 3). The surgical site was closed with an absorbable suture (4-0 Vicryl Plus; Johnson & Johnson MedTech) using direct suturing, followed by primary closure.
Results
In this study, it was confirmed that the 45 MRONJ patients who underwent marginal resection and sequestrectomy had been prescribed medications including bisphosphonates (Risedronic acid, Ibandronic acid, Zoledronic acid, Alendronic acid), selective estrogen receptor modulators (SERMs), and Prolia (Table 1).
From an etiological perspective, 11 patients were affected due to tooth extraction, 9 patients were affected by implant installation or removal, and those affected by endodontic treatment for tooth preservation was 1 patient. The largest proportion was found in 24 patients who developed osteomyelitis due to bone necrosis caused by infection of unknown origin. In cases where MRONJ progressed due to implants, the bone regeneration rate was 59.81%. Cases associated with infection showed a regeneration rate of 58.55%, those related to extraction exhibited a rate of 60.57%, and cases associated with endodontic treatment demonstrated a rate of 76.67%. However, there was no statistically significant correlation between etiology and bone regeneration rate (Table 2).
The most commonly affected site of the lesion was the mandibular molar region. In the mandible, the bone regeneration rate in the molar region (60.52%) was relatively lower compared to the premolar region (63.72%) and the anterior region (74.91%). No distinct trend was observed in the maxilla (Table 3).
In terms of the average bone regeneration rate according to the number of remaining bony walls in the alveolar bone defect area, lesions with four bony walls showed the highest regeneration rate (Table 3). Lesions with one wall (33.15%), two walls (56.96%), and three walls (63.96%) had lower regeneration rates compared to those with four walls (85.19%). It was confirmed that as the number of remaining bony walls in the postoperative bone defect area increased, the bone regeneration rate also increased (Table 4).
A comparison of bone regeneration rates at 6 months postoperatively based on the presence or absence of disruption within 2 weeks after surgery showed that 14 patients experienced surgical site disruption, with an average regeneration rate of 57.87%. In contrast, 31 patients had no disruption, with an average regeneration rate of 60.52%. The presence of disruption did not have a statistically significant effect on bone regeneration (Table 5).
To evaluate the effect of each independent variable (number of bony walls, age, and initial alveolar bone defect size) on the dependent variable (bone regeneration rate), a multiple linear regression analysis was conducted to determine the statistical significance. The significance test of the regression coefficients showed that the number of bony walls remaining after surgery (p ＜ 0.001) was the only factor significantly associated with the bone regeneration rate. According to the results of the multiple regression analysis using the backward elimination method, patient age and initial alveolar bone defect volume did not have a significant effect on the bone regeneration rate (Table 6). To evaluate the statistical power of the regression analysis, a post hoc power analysis was conducted using G*Power (Universität Düsseldorf). The observed effect size (f2 = 1.90; calculated from the explained variance, R2 = 0.655), a total sample size of 45 patients, three predictors in the model, and a significance level of 0.05 (one-tailed) were used as input parameters. The analysis showed that the statistical power (1-β error probability) was 1.00, indicating an extremely low probability of a type II error under the studied conditions and sufficient power to detect meaningful effects in bone regeneration outcomes.
Discussion
rhBMP-2 stimulates osteogenic differentiation by activating the Smad signaling pathway and the transcription factor Runx2, which upregulate key bone formation genes. This process enhances the expression of osteogenic markers such as alkaline phosphatase, osteocalcin, and collagen type I, leading to robust bone matrix production and mineralization. In summary, BMP-2 acts as a central regulator, directing mesenchymal stem cells toward the osteoblastic lineage and promoting efficient bone tissue formation [3].
To date, no previous study has directly compared experimental and control groups using rhBMP-2 for volume-based bone regeneration specifically in osteomyelitis models volumetric changes. Although this investigation was performed in a large animal (miniature pig) model rather than in a clinical setting, Probst et al. [4] evaluated the bone regeneration effect of customized hybrid scaffolds fabricated using CT-based threedimensional design technology in a critical-size mandibular defect model in miniature pigs. A comparison between the experimental group (hybrid scaffold combined with rhBMP-2) and the control group (hybrid scaffold alone) revealed a significant difference in new bone formation. Micro-CT analysis demonstrated that the experimental group exhibited a markedly higher bone volume/total volume ratio of 34.80 ± 4.80%, compared to 22.40 ± 9.85% in the control group. Histological evaluation also confirmed that the rhBMP-2-loaded scaffold promoted denser and more mature new bone formation.
Lee et al. [5] demonstrated that increased numbers of residual alveolar bone walls result in higher new bone formation after alveolar ridge preservation. The presence of multiple bony walls provides abundant regenerative cells, stabilizes the physical structure, and ensures strong vascularization, which together enhance both histological and clinical outcomes. Therefore, the number of remaining bone walls is a key prognostic indicator for successful bone regeneration because it not only improves cell migration and wound stability, but also optimizes the blood supply for tissue healing.
In this study, we analyzed the characteristics of a group of patients who underwent osteomyelitis surgery performed by a single, experienced oral and maxillofacial surgeon between 2016 and 2024. Among these, we selected patients in whom BMP was applied to the alveolar bone defect site. The aim was to evaluate the efficacy of BMP over a 6-month period and identify factors affecting the volume change of the defect area. Since all surgeries were performed by a single surgeon, operator variability was excluded from the study group.
Also, volumetric assessment of lesions following osteomyelitis surgery is challenged by the irregular morphology of bone defects; no standardized method has been established, thereby necessitating the development of a novel approach for volume measurement. In this study, defect volume was estimated by multiplying the maximum lengths of the bone defect along the pitch, roll, and yaw axes, and the difference between post operative day 0 and 6 months postoperatively was used to assess the amount of bone regeneration. Although a ratiobased calculation was applied to reduce measurement error, this approach remains limited in that it cannot fully capture the complex and irregular morphology of bone defects. Therefore, more sophisticated methods are required to achieve a precise assessment of bone regeneration.
Due to insufficient data from patients who underwent osteomyelitis surgery without BMP (such as cortical bone decortication and sequestrectomy) and had follow-up over 6 months, we were unable to conduct comparative research between patients who did and did not receive BMP. Future studies comparing a control group (non-rhBMP-2/ACS) with a experimental group (rhBMP-2/ACS) will be necessary.
Immediate reconstruction following jawbone resection is effective. In cases where resection is unavoidable, reconstruction is essential for addressing both esthetic and functional issues [6]. After cortical bone decortication and sequestrectomy in the jaw, bone grafting is typically performed once infection is fully resolved. However, in such cases, the degree of bone regeneration is unpredictable, and due to soft tissue shrinkage at the surgical site, the reconstruction period using implants and prosthodontics can be significantly prolonged. Using rhBMP-2 simultaneously during osteomyelitis surgery has the advantage of preventing soft tissue shrinkage and enhancing bone regeneration.
In this study, due to the hydrophilicity of rhBMP-2 and its tendency to diffuse into surrounding areas, rhBMP-2 was not used independently. Instead, it was applied using a carrier, such as an absorbable collagen matrix or fibrin matrix, and then sealed with Tisseel and a resorbable membrane. This method allowed for a gradual and sustained release into the surgical site, and primary closure was performed whenever possible. Notably, even when surgical site disruption occurred, there was no statistically significant difference in bone regeneration rates. When disruption occurred, the area was packed with Vaseline-coated gauze (Reno-tulle; Shinhan Meditech), which likely helped prevent rhBMP-2 loss due to its hydrophilic nature and protected the site from food impaction or other sources of infection. This likely contributed to the lack of significant differences in bone regeneration related to wound disruption.
In this study, MRONJ occurred most frequently in the mandibular molar region, and the bone regeneration rate was lower in this area compared to the mandibular anterior and premolar regions. Baghele [7] reported that in the maxilla, the buccinator muscle attaches near the alveolar ridge in the molar area, whereas in the mandible, it attaches at the pterygomandibular raphe of the bucco-pharyngeal fascia. Dutra et al. [8] noted that considering the attachment sites of the buccinator muscle, localized mechanical pressure is applied to the alveolar bone and gingiva. Based on these findings, it can be inferred that tensile forces caused by muscle attachments in the posterior maxilla and mandible may contribute to the lower bone regeneration rates in those regions.
In this study, due to follow-up loss in patients with maxillary lesions, statistical significance could not be demonstrated. In the maxillary anterior region, however, postoperative infection in one patient resulted in further bone resorption, with the defect volume increasing by an additional 35.88% relative to the immediate postoperative measurement. This markedly skewed the mean regeneration rate in this area to 23.96%. When this patient was excluded, the bone regeneration rate in the maxillary anterior region increased to 83.80%. Overall, the average bone regeneration rate was 59.70%, with the highest in the mandibular anterior region (74.91%) and the lowest in the maxillary anterior region (23.96%).
All patients in this study who underwent decortication and sequestrectomy were classified as stage 3 and required complete surgical removal of necrotic tissue. Various factors can influence bone regeneration, among which diabetes is known to impair bone healing due to microvascular dysfunction, increased infection risk, and changes in cell growth factor signaling pathways. Long-term corticosteroid use may also reduce osteoblast function, leading to decreased bone matrix production and cell numbers. This study has the limitation of not fully accounting for systemic health conditions and pharmacologic factors that can affect bone regeneration, such as diabetes, corticosteroid use, and medication duration. These variables may directly or indirectly impact bone healing; notably, impaired glycemic control or immunosuppressive effects, as well as metabolic alterations from long-term medication use, can serve as significant confounding factors in the interpretation of clinical outcomes. Future investigations should systematically evaluate and incorporate each patient’s medical comorbidities, medication history, and other potentially influential clinical variables into the analysis in order to enhance the accuracy and reliability of bone regeneration results.
Yon et al. [9] reported that residual adjacent bone walls in the defect site help maintain the blood clot and facilitate alveolar bone regeneration by providing a source of regenerative cells. Another study by Nibali et al. [10] emphasized that not only the number of residual bone walls but also flap design plays an important role in bone regeneration. Consistent with these findings, our study also found that the number of remaining bony walls had the most significant effect on alveolar bone regeneration. Specifically, lesions with four bony walls postoperatively showed an 85.19% regeneration rate, while those with only one wall showed a significantly lower regeneration rate of 23.61%. In the multiple linear regression analysis, the unstandardized coefficient (B) for residual bone walls was 16.225 (standard error = 2.93, t = 5.53, p ＜ 0.001), indicating a significant positive association with bone regeneration. The 95% CI for B was estimated to be 10.45–22.00.
In conclusion, in the treatment of osteomyelitis, the simultaneous application of rhBMP-2 with decortication and sequestrectomy may enhance bone defect regeneration, shorten the healing period, and lead to more predictable outcomes. Thus, the use of rhBMP-2 appears to provide favorable results for oral reconstruction in affected patients. However, further studies incorporating a control group are warranted to allow direct comparisons and to more clearly validate these findings.
