Clinical outcome and volumetric 3D analysis of biofluorescence imaging system guided surgery for Medication-Related Osteonecrosis of the Jaw(MRONJ)

Abstract

Background

This study aims to evaluate the clinical outcome of biofluorescent imaging system (BIS) guided MRONJ surgery through analyzing 3D volumetric changes in CBCT data of bone structure.

Methods

BIS-guided surgery for MRONJ surgery was performed by grinding red-fluorescent area from remained residual bone detected by Qray-pen. CBCT data was collected preoperatively, postoperatively, and at last follow-up more than 3 months of each patient. Healing ratio per month was calculated through measuring difference between postoperative defect and last follow-up defect volume by analyzing CBCT.

Results

With mean follow-up period of 7.8 ± 4.8 months, twelve MRONJ surgeries of 11 patients (2 males, 9 females, mean 72.8 ± 13.7 years old) showed bone healing tendency without recurrence in average healing ratio of 5.49 ± 2.22% per month, with complete soft tissue closure. It was observed that the healing ratio per month did not show statistically significant differences based on location of MRONJ, size of post-operative defect volume, nor duration of bisphosphonate intake before surgery.

Conclusions

Based on three-dimensional bone analysis, efficacy of BIS-based MRONJ surgery was proved through the recovery pattern of underlying bone structure and no recurrence sign showing.

Clinical trial registration

Not applicable.

Background

Medication-related Osteonecrosis of the Jaw (MRONJ) is a lesion that occurs due to the side effects of antiresorptive and antiangiogenic agents. The American Association of Oral and Maxillofacial Surgeons (AAOMS) classifies MRONJ into stages based on symptoms and objective signs of the lesion and regularly updates treatment guidelines through position paper [1]. However, there is still considerable debate between conservative and surgical treatments, as each has its own advantages and disadvantages. While conservative methods can help reduce symptoms, achieving mucosal healing and complete resolution of the lesion remains challenging. Surgical methods aim for rapid healing but may require extensive resection, which can potentially reduce a patient’s quality of life and still carries the risk of recurrence [2]. Therefore, research continues into methods to achieve complete resolution of lesions through surgical approaches that minimize removal to only the necrotic bone area and reduce the recurrence rate. Techniques such as dynamic navigation are also being studied to remove lesions while avoiding major anatomical structures [3].

Among surgical methods, there is growing interest in accurately identifying lesions requiring removal intraoperatively using fluorescence [4–7]. Studies confirming efficacy and high success rates of such approaches have been published, using tools like VELscope (LED Apteryx, Akron, OH, USA; Mectron Spa, Carasco, Italy) which can differentiate necrotic bone and normal bone based on differences in fluorescence intensity [7, 8]. Nonetheless, VELscope necessitated supplementary medications to induce fluorescence variance and couldn’t pinpoint the affected area unless there were alterations in bone minerals. Among various types of biofluorescence imaging systems (BIS), the Quantitative Light-induced Fluorescence (QLF) system can examine the red fluorescence excitation of porphyrins, which appear as metabolic byproducts of late-stage oral colonies. Bone can be categorized into normal (non-red), hypo-red, and hyper-red groups: non-red fluorescence indicates sclerotic and lamellar bone tissue, hyper-red fluorescence indicates infection due to bacterial invasion and osteolysis, and hypo-red fluorescence indicates predominantly granulation tissue with inflammation, absent bone matrix, and bacterial colonies [5]. This classification allows for the precise removal of hypo-red and hyper-red fluorescent areas, which show bacterial invasion and signs of bone matrix destruction [5, 9]. However, including studies with BIS device, most research dealing with fluorescence-guided surgery has focused solely on mucosal closure and symptom resolution, without quantitative evaluation of bone changes in lesions or demonstration of recovery patterns. In this study, we aim to present the results regarding the stability of BIS-guided sequestrectomy by analyzing volume and shape changes of MRONJ lesions based on CT data. We assume that bone tissue showing non-red fluorescence—presumed to be normal bone—when left behind by BIS-guided surgery, would follow the normal bone healing process without MRONJ recurrence.

Methods

The Institutional Review Board of Seoul National University Bundang Hospital approved this study (B-2406-905-102). This retrospective study included 12 MRONJ patients who visited the Oral and Maxillofacial Surgery Department of Seoul National University Bundang Hospital from September 2022 to March 2024 with chief complaints of MRONJ-related discomfort. The inclusion criteria for this study were (1) the use of Qray-pen during MRONJ sequestrectomy (patients with osteoradionecrosis were excluded); (2) postoperative biopsy result confirming the osteomyelitis or actionmycosis as the final diagnosis result; (3) presence of CBCT results within 2 weeks after surgery as immediately postoperative CBCT, and more than 3 months after the surgery. The exclusion criteria were (1) new implant fixtures not previously present found in lesion at follow-up CBCT, which inhibits exact calculation of bone volume; (2) no postoperative or follow-up CBCT data after surgical treatment; (3) no history of chemotherapy or radial therapy. The following information was collected from the medical records and radiographic records: gender, age, MRONJ location, drug history (mostly for antiresorptive or antiangiogenic drugs, such as bisphosphonate and denosumab), medical history, smoking habit, dental history. Written informed consent was obtained from the patients.

The surgical procedure was performed by a skilled oral and maxillofacial surgeon (J.K.Ku) following the principles of minimally invasive surgery, accompanied by the Qray-pen (AIOBIO, Seoul, Republic of Korea). After MRONJ was diagnosed, medications such as anti-resorptive agents were discontinued. However, surgery was performed without intentionally implementing an additional preoperative drug holiday. The anti-resorptive agent was discontinued only until the complete healing of the soft tissue following MRONJ surgery. Mucoperiosteal flap was elevated at sites with clinical symptoms with necrotic change, and positive findings on bone scan or CBCT. We thoroughly debrided fibrotic bone, visibly necrotic bone, softened bone, and surrounding inflamed tissue (Fig. 1A) Using Qray-pen, we examined the remaining bone tissue and areas exhibiting hyper-red fluorescence were visually confirmed (Fig. 1B), and removal was carried out using a 2.0 mm round bur and bone rongeur (Fig. 1C-F) [5, 9]. After confirming the complete removal of the area showing hyper-red fluorescence through the Qray-pen (Fig. 1G, H), primary closure was achieved with adaptive periosteal releasing incision (Fig. 1J). Postoperatively, amoxicillin 1 g bid, non-steroidal anti-inflammatory drugs (naproxen, 500 mg bid), and 0.12% chlorhexidine gargle were prescribed for one week each. Patients were instructed to resume MRONJ-inducing medications according to the scheduled plan after confirmation of soft tissue healing (Fig. 1L).

Fig. 1.

Intraoperative images on Qray-pen device; A. Necrotic bone and surrounding inflamed tissue found after mucoperiosteal flap elevation. B. Hyper-red/hypo-red fluorescent bone tissue found under Qray-pen image. C,E View after additional removal of hyper-red fluorescent bone area. D,F. Hyper-red fluorescent bone tissue still remaining under Qray-pen image. G. View after complete removal of hyper-red fluorescence area. H. Confirmed complete removal of the hyper-red fluorescent area. I. Initial clinical photo J. Clinical photo after suture K. Post-operative clinical photo after 2 weeks L. Post-operative clinical photo after 2 months

CBCT parameters (Alphard 3030, Asahi Roentgen Ind., Kyoto, Japan) were as follows: 80 kV, 8 mA, 17 s exposure time, 154 _ 154 mm field of view, and 512 basis projections. Each scan was reconstructed and saved in a standard DICOM file with a voxel size of 0.1 mm. All images were transferred to a picture-archiving and communication system viewer (INFINITT PACS, Infinitt Healthcare, Seoul, Republic of Korea). Then, CT images were converted into the Digital Imaging and Communications in Medicine (DICOM) files, and they were imported into MIMICS Materialise’s interactive medical image control system (MIMICS V25.0, Materialise, Leuven, Belgium). Pre-operative defect, post-operative defect, and last follow up defect volume were measured using the specific threshold value (226-3071 Hounsfield units for bone demonstration) [10]. Value below 226 unit from MIMICS (-1024 ~ 225) were considered as the lesion area with lower value than normal bone’s demonstration. Through this detailed automatic differentiation, margin of lesion was carefully determined and radiopaque-like sequestrum found within lesion were included as defect volume even if it showed higher value than 226 units. Also in post-operative defect and last follow-up defect, only the bone formation examined from margin of the defect were considered as newly formed bone volume, to screen out the bone graft material remaining within the defect (Fig. 2). Lastly, three CT-datas were aligned into same 3D location using 3-Matic design software (MIS Medical and research 18.0, Materialise, Leuven, Belgium) in order to precisely compare and calculate the change of the defect’s conformity [11]. One examiner (SS) confirmed the border of the pre-operative, post-operative, and last follow-up defects in coronal, axial and sagittal views respectively (Figs. 3 and 4). Healing ratio was calculated as follows [12]:

Fig. 2.

Example of CT view for (A) preoperative, (B) postoperative, and (C) last follow up state for MRONJ defect. Right side of A, B,C shows the area of lesion with value lower than normal bone value (226U) from Mimics program

Fig. 3.

A and B. Example of aligning different CBCT STL file into same location through 3-matics, C. difference between postoperative defect(cyan) and last follow up defect(pink), D. Comparison in 3D analysis through aligning defects into same position

Fig. 4.

CBCT view of postoperative defect(violet) and last follow up defect(yellow) in same location within MIMICS program, showing (A) coronal, (B) axial, (C) sagittal, and (D) 3D reconstructed view from CT as well

IBM SPSS Statistics version 25.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis with 95% statistical significance level. Normality test was conducted for the entire population using the Shapiro-Wilk test. Then, healing ratio calculated using the above formula was divided by the total CT follow-up period to calculate healing ratio per month, and mean value was analyzed. For additional comparison, Shapiro-Wilk normality tests for each group categorized by: (1) location of lesion, (2) duration of osteoporosis-related medication intake, and (3) size of post-operative defect volume. Subsequently, as the data satisfied normality, paired t-tests were performed for each condition to determine the p-value. A p-value below 0.05 was considered statistically significant.

Results

A total of 11 patients (9 females and 2 males, mean age 72.8 ± 13.7 years) and 12 samples were included in this study. Among the 12 samples, 6 were lesions occurred in the maxilla and 6 were lesions in the mandible. As for the etiology, 5 occurred after implant placement, 3 after tooth extraction, 3 after denture use, and 1 was classified as unknown. The mean total follow-up period after surgery was 7.8 ± 4.8 months. The average preoperative defect volume, postoperative defect volume and total healed volume at last follow up were 1646.6 ± 1279.3 mm³, 1811.0 ± 1404.6 mm³, and 446.6 ± 368.2 mm³. The amount of additional volume removed, as observed on the CT volume after surgery compared to the preoperative defect volume, was an average of 179.3. Volume changes and postoperative recovery amounts for each patient were summarized in a Tables 1 and 2. Bone healing tendencies without sign of recurrence were observed in all patients after surgery, showing average healing ratio per month of 5.49 ± 2.22%, showing a normal distribution without any outliers (minimum healing ratio per month was 1.79%). Gingiva closure, a crucial criterion for recovery after osteomyelitis surgery, was confirmed in all cases with no recurrence observed. Lastly, it was observed that the healing ratio per month did not show statistically significant differences based on maxilla/mandible location, size of post-operative defect volume, or duration of bisphosphonate intake before surgery (Table 3).

Table 1.

Demographic and examination data on patients treated with BIS-guided sequestrectomy surgery

Patient No.	Sex	Age	Location	Osteoporosis(OP), Bisphosphonate(BP) usage time	Cause for MRONJ	Preoperative defect volume (mm³)	Postoperative defect volume (Post-pre volume difference, mm³)	Last follow-up defect volume (mm³)
1	F	79	Mandible	OP, BP use < 3 years	Implant	72.73	166.09 (+ 93.31)	99.74
2	F	78	Mandible	OP, BP use > 10 years	Extraction	1146.76	1081.04 (-65.72)	488.82
3	F	80	Maxilla	OP, BP use > 20 years	Denture	428.67	299.58 (-129.09)	89.08
	F	80	Mandible	OP, BP use > 20 years	Denture	548.56	809.79 (+ 261.23)	439.22
4	F	69	Maxilla	OP, BP use > 10 years	Implant	3167.31	3738.91 (+ 571.60)	2272.08
5	F	83	Mandible	OP, BP use 1 year	Extraction	2696.53	2809.34 (+ 112.81)	2151.08
6	F	84	Maxilla	OP, BP use 3 years	Extraction	4220.74	4295.79 (+ 75.05)	3988.38
7	M	34	Maxilla	No OP history	Implant	1198.94	1296.08 (+ 97.14)	1025.83
8	F	75	Maxilla	OP, Denosumab > 10 years	Implant	1366.33	1577.74 (+ 211.41)	1140.20
9	F	75	Mandible	OP, BP use 11 years	Unknown	2829.72	3489.78 (+ 660.06)	2944.99
10	M	56	Mandible	No OP history	Denture	- *	1300.80 (*)	1163.78
11	F	76	Maxilla	OP, BP use 3 years	Implant	782.59	866.82 (+ 84.24)	568.69

* Preoperative CBCT data was not taken for patient 10

Table 2.

Volumetric analysis and healing ratio calculated after BIS-guided sequestrectomy surgery

	Mean ± SD
Follow-up period after surgery (months)	7.8 ± 4.8
Preoperative defect volume (mm3)	1646.6 ± 1279.3
Postoperative defect volume (mm3)	1811.0 ± 1404.6
Last follow-up defect volume (mm3)	1364.32 ± 1225.2
Total healed volume (mm3)	446.7 ± 368.2
Healing ratio (%)	32.47 ± 18.72
Healing ratio per month (%)	5.49 ± 2.22

Table 3.

Volumetric analysis for spontaneous bone healing after BIS-guided sequestrectomy surgery

Variables	Healing ratio per month (%)	P-value
Total (n = 12)	5.49 ± 2.22
Jaw
Maxilla (n = 6)	5.61 ± 2.79	0.864
Mandible (n = 6)	5.37 ± 1.72
Post-operative defect volume
Under 2,000 (n = 8)	5.79 ± 2.16	0.538
Over 2,000 (n = 4)	4.90 ± 2.53
Bisphosphonate(BP), denosumab usage time
Usage time < 4 years (n = 6)	5.04 ± 2.14	0.508
Usage time > 10 years (n = 6)	5.94 ± 2.39

Discussion

Through previous studies, it was confirmed that lesions displaying hyper-red and hypo-red fluorescence via the Qray system (AIOBIO, Seoul, Republic of Korea) correspond histologically to areas of osteomyelitis and inflammatory tissue requiring removal [5, 9]. Subsequently, through case studies, it was demonstrated that such BIS-based MRONJ sequestrectomy could aid in minimally invasive surgical procedures [9, 13]. In this paper, our primary goal was to confirm clinically observable outcomes such as gingival healing and symptom relief [6, 14], while also quantitatively analyzing bone changes pre- and post-surgery, as well as during the last follow-up CT, to track substantial recovery [15]. As a result, it was observed that in all 12 samples, not only gingival healing but also bone recovery equivalent to an average of 5.49% per month without recurrence was achieved. This recovery rate is similar to the average recovery rate post-cyst removal and reflects a normal recovery trajectory, akin to typical recovery rates for osteomyelitis [12]. Since quantitative measurement of MRONJ volume changes is challenging, most studies have focused on measuring the proportion of complete recovery at specific time points [15]. Beyond simply demonstrating the success of surgery and absence of recurrence, as evidence of gingival healing, this study is significant for being the first to introduce criteria and numerical values for monthly bone recovery rates.

In all cases, new bone formation could be observed on CBCT taken at least three months post-surgery compared to immediately after surgery, despite not discontinuing the medication that could cause MRONJ before and after the surgery. Moreover, it was possible to quantify the amount removed post-operatively, compared to the size of the original lesion after removing MRONJ lesions based on the Qray-cam device. When there was uncertainty regarding whether the area required removal during surgery, instead of relying solely on the traditional method of giving safety margin until fresh blood was observed, direct confirmation through BIS device was employed to finalize the surgical area [1, 2, 7, 15, 16]. James et al. have recently reported that necrotic bone was observed at the boundary from histologic analysis in 33.3% of cases, even when routinely applying safety margin. Additionally, other fluorescence-guided surgery using VELscope (LED Apteryx, Akron, OH, USA; Mectron Spa, Carasco, Italy) which can differentiate normal bone and necrotic bone, but not bacterial-invased bone tissue, showed 10–15% of bone exposure remaining after the surgery [16, 17]. It was also mentioned that in terms of mucosal healing, surgery using VELscope did not show significantly better results compared to the conventional method [18, 19]. Compared to such datas, results obtained through a BIS-guided minimally invasive approach without recurrence show notable outmarks.

Limitation of this study is that despite confirming that duration of preoperative antiresorptive agents use do not significantly affect bone recovery rates, there was a lack of comparison between denosumab, which has shorter half-life than bisphosphonates, due to the small sample size [2]. Additionally, bone healing and recovery from osteomyelitis could be greatly influenced by the configuration of the defect. Considering the periodicity and characteristics of bone recovery, rather than demonstrating a steady bone recovery rate, there might be a higher recovery rate per month as time passes, thus long-term data collection are needed. While a considerable number of lesions in our study formed radiolucent lesions with sequestrum, there are reports indicating slower recovery rates when sclerotic changes appear throughout the lesion [20]. Furthermore, based on this case series study, there is a need to conduct clinical research comparing the outcomes of patients who did not undergo BIS-guided surgery. Considering that the pathogenesis of MRONJ still remains unresolved, retrospectively identifying the cause of the different fluorescence patterns observed in bone tissue upon BIS application may provide insights into understanding pathogenic mechanism of MRONJ.

Conclusions

BIS-guided MRONJ surgery, which can visibly confirm pathologic tissue intraoperatively, showed complete gingiva recovery and resolution of symptoms. Moreover, through three-dimensional analysis, the absence of recurrence and the recovery pattern of underlying bone structure were confirmed in all cases. BIS-guided MRONJ surgery was confirmed to be a highly utilizable, and effective minimally invasive surgical approach for MRONJ treatment, through 3D analysis of multiple cases.

BIS-guided MRONJ surgery, which enables the intraoperative visualization of pathological tissue through three-dimensional analysis, demonstrated no recurrence and a consistent recovery pattern of the underlying bone structure in the cases studied. While these initial results are promising, the limited sample size and follow-up necessitates cautious interpretation. Further research with a larger cohort is needed to fully validate the effectiveness and utility of this minimally invasive surgical approach for the treatment of MRONJ.

Author contributions

SSH wrote the main manuscipt text, prepared Figs. 1, 2, 3 and 4; Tables 1, 2 and 3, performed all analyses, contributed to the data analysis strategy. JKK performed the assessments and all surgeries in the study. HKJ curated data. SSH . JKK, PYY contributed to the interpretation of results and drafting of the manuscript. YSK validated all analysis. HCY provided software program and resources. All authors read and approved the final manuscript.

Funding

This work was supported by the Seoul National University Bundang Hospital Research Fund grant number 14-2024-0003.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the institutional review board of Seoul National University Bundang Hospital (IRB No. B-2406-905-102), and was conducted according to the principles of the Declaration of Helsinki for research on humans, protecting the participants’ anonymity and privacy and maintaining public confidence. All participants provided written informed consent for participation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.