ΔBSIJ: A quantitative marker for early detection of medication-related osteonecrosis of the jaw in patients with prostate cancer receiving bone-modifying agents

doi:10.21203/rs.3.rs-5795022/v1

Objective: Medication-related osteonecrosis of the jaw (MRONJ) is a severe complication of bone-modifying agent (BMA) therapy in patients with prostate cancer and bone metastasis. This study aimed to assess the effectiveness of the temporal changes in jaw-specific bone scan index (ΔBSIJ) as quantitative markers for early prediction of MRONJ in patients with prostate cancer receiving BMA therapy.

Methods: This retrospective study included 33 patients with prostate cancer with bone metastases who underwent bone scintigraphy before and after BMA initiation. BSIJ was measured using BONENAVI software, and the difference between pre- and post-BMA BSIJ values was considered ΔBSIJ. Statistical analyses, including paired t-test, receiver operating characteristic (ROC) curve analysis, and Kaplan–Meier survival estimate, were employed to assess the predictive value of ΔBSIJ for MRONJ.

Results: Of the 33 patients, 10 developed MRONJ during a median follow-up period of 29 months. ΔBSIJ was significantly higher in the MRONJ group than in the non-MRONJ group (0.05 vs. –0.04, p = 0.002). ROC analysis revealed the highest area under the curve (AUC = 0.823) for ΔBSIJ compared with the pre- and post-BMA BSIJ values. A ΔBSIJ cutoff of 0.039 predicted MRONJ with 60% sensitivity and 91% specificity. Patients with ΔBSIJ ≥ 0.039 exhibited significantly shorter MRONJ-free survival than those with ΔBSIJ < 0.039 (median: 18.4 months vs. not reached, p < 0.001).

Conclusion: ΔBSIJ is a novel and clinically useful quantitative marker for the early detection of MRONJ in patients with prostate cancer receiving BMA therapy. This study highlights the potential of leveraging functional imaging and temporal changes in BSIJ to improve MRONJ management.

ΔBSIJ

Medication-related osteonecrosis of the jaw

Bone scintigraphy

Prostate cancer

Bone-modifying agents

Prostate cancer is a common malignancy in men, and a significant proportion of patients develop bone metastases during the course of the disease. Bone-modifying agents (BMAs), including bisphosphonates and denosumab, are frequently used to reduce skeletal-related events in patients with bone metastasis. However, medication-related osteonecrosis of the jaw (MRONJ), a severe condition that profoundly impacts quality of life (QOL) of the patient and interferes with ongoing cancer treatment, is a well-recognized adverse event of BMA therapy [1, 2].

Early detection of MRONJ is crucial to mitigate its progression and associated complications. Once clinically apparent, invasive management strategies, including surgical intervention, are often necessary for MRONJ, which is associated with substantial morbidity [3]. Conventionally, MRONJ diagnosis relies on imaging modalities such as radiographs, computed tomography (CT), and magnetic resonance imaging (MRI) [4]. However, these structural imaging techniques primarily detect anatomical changes but not early metabolic alterations in the bone [5].

Bone scintigraphy, a functional imaging modality widely used in oncology, offers the unique advantage of detecting early metabolic and pathophysiological changes in the bone. Furthermore, bone scintigraphy detects subtle abnormalities in bone turnover before the appearance of structural changes, unlike conventional imaging, rendering it a promising tool for early MRONJ detection [5–8]. Additionally, bone scintigraphy is routinely performed in patients with prostate cancer to detect bone metastasis and for follow-up, making it feasible to repurpose this modality for MRONJ screening without additional imaging.

Although several studies have investigated the role of bone scintigraphy in MRONJ prediction, most lack a robust quantitative framework for evaluating disease progression. The bone scan index (BSI), a semi-automated quantitative metric derived from bone scintigraphy, has proven useful in assessing skeletal metastases and prognostication in prostate cancer. However, its application in MRONJ prediction remains underexplored. Furthermore, although quantitative imaging biomarkers have been incorporated in limited studies [5, 8], evaluations of the temporal changes using these indices remain sparse.

Consequently, in this study, we utilized BSI to develop a quantitative framework for MRONJ prediction. We hypothesized that changes in the jaw-specific BSI (BSIJ) before and after BMA initiation, quantified as ΔBSIJ, could serve as an early indicator of MRONJ risk. Leveraging a novel and semi-automated approach, this study represents a significant step toward improving the early detection and management of MRONJ in patients with prostate cancer receiving BMA therapy.

Patients

This single-institution retrospective study involved patients with prostate cancer and bone metastases with a history of receiving BMAs and had evaluable bone scintigraphy images taken before and after BMA administration. Patients for enrollment in this study were screen as follows: First, 1,281 patients with prostate cancer who underwent bone scintigraphy at our hospital between 2008 and 2023 were identified. Of these, 209 patients had positive findings for bone metastases, and among them, 94 patients had received BMAs. Of these, 33 patients who underwent bone scintigraphy both before and after the initiation of BMA treatment were included in the study.

MRONJ was diagnosed by an oral and maxillofacial surgeon based on clinical examination, radiographic findings from X-ray and CT imaging, and patient history. Patients diagnosed with MRONJ at any point during follow-up were classified into the MRONJ group, while those without MRONJ were placed in the non-MRONJ group.

The study protocol was approved by the Institutional Review Board of Dokkyo Medical University (approval number: R-71-6J). The research plan was communicated through in-hospital announcements and the hospital's website, with an opt-out option for patients who did not wish to participate. Individual informed consent was deemed unnecessary owing to the retrospective nature of the study.

BSIJ measurement

BSIJ was measured using anterior and posterior whole-body bone scintigraphy images. For patients who underwent multiple bone scintigraphy sessions, the scan obtained immediately before BMA administration was selected as pre-BMA bone scintigraphy, and the scan obtained just before the onset of MRONJ and after BMA administration was selected as post-BMA bone scintigraphy.

The cranial region with the upper border at the bottom of the nasal cavity and the lower border at the lower end of the mandible was considered the region of interest (ROI) for BSIJ measurement (Figure 1). The BSI was measured using the BONENAVI_® (version 2; PDRadiopharma Inc., Tokyo, Japan). This software utilizes an artificial neural network (ANN) based on a multicenter database of Japanese patients to automatically detect abnormal accumulation areas, calculate the regional BSI (rBSI) as the ratio of the abnormal accumulation area to the total skeletal area, and assign an ANN score ranging from 0 to 1 for each region. The details of this platform have been described previously [9, 10]. Although the total-body BSI only includes regions with an ANN score of ≥0.5, this study included all regions within the ROI for BSIJ, regardless of their ANN scores. The BSIJ was calculated as the sum of all the rBSI values within the ROI for the BSIJ. A urologist (T. Kokubun) blinded for the clinical information manually set the ROI for BSIJ in the BONENAVI_® software and calculated BSIJ for all cases.

As described above, the ROI for BSIJ measurement was manually set based on anatomically well-defined landmarks. As the rBSI calculation was fully automated using the BONENAVI software, the measurement process was entirely objective and did not involve any subjective judgment. Therefore, the involvement of a radiologist or nuclear medicine physician was not essential for this study.

Statistical analysis

Clinical parameters were expressed as N (%) or median (interquartile range). The differences between MRONJ and non-MRONJ groups were determined using the Fisher's exact test and Mann–Whitney U tests. The BSIJ obtained from pre-BMA and post-BMA bone scintigraphy was defined as the pre-BMA BSIJ and post-BMA BSIJ, respectively. The difference between pre- and post-BMA BSIJ was defined as ΔBSIJ (Figure 2). Paired t-tests were used to compare the pre- and post-BMA BSIJ scores within individual cases. The predictive ability of pre-BMA BSIJ, post-BMA BSIJ, and ΔBSIJ for MRONJ onset was assessed using receiver operating characteristic (ROC) analysis. The area under the curve (AUC) was calculated, and the cut-off value for maximizing the predictability of MRONJ was determined. The predictive performance was evaluated and compared based on the AUC, sensitivity, specificity, positive predictive value, and negative predictive value. Kaplan–Meier method was employed to estimate the MRONJ-free survival, defined as the time from the initiation of BMA therapy to the onset of MRONJ. The log-rank test was used to determine the significance. All statistical analyses were performed using JMP PRO, (version 17; SAS Institute, Cary, NC, USA). Statistical significance was set at p < 0.05.

Patient characteristics

The median follow-up period after BMA initiation in 33 patients was 29 (interquartile range: 21–49) months. Of these, MRONJ occurred only in 10 patients during follow-up. Table 1 presents patient characteristics. The median age was 69 years (64–75 years) and the median prostate-specific antigen (PSA) level at diagnosis was 257.5 ng/mL (20–1015 ng/mL), with 29 patients (88%) classified into a Gleason grade group of ≥4. Androgen deprivation therapy involved gonadotropin-releasing hormone agonists and antagonists in 18 (55%) and 15 (45%) patients, respectively. Fourteen (42%) patients had undergone tooth extraction owing to dental conditions. No new diagnoses of tooth decay or periodontitis were made during the follow-up period, and no cases exhibited worsening of these conditions requiring additional treatment. None of the patients in this study had received radiotherapy to the jaw. BMAs denosumab and zoledronic acid were used in 28 (85%) and 5 (15%) patients, respectively. Patients in the MRONJ group were older than those in the non-MRONJ group; however, no differences in other clinical factors were observed between the two groups.

The interval from pre-BMA bone scintigraphy to BMA administration was 1.6 (0.8–4.3) months, and the interval from BMA administration to post-BMA bone scintigraphy was 13.0 (7.6–23.2) months. These intervals did not differ significantly between the MRONJ and non-MRONJ groups. Furthermore, the interval from BMA administration to MRONJ onset was 17.9 (15.9–21.5) months in the MRONJ group, whereas the interval from post-BMA bone scintigraphy to MRONJ onset was 7.6 (0.9–9.2) months.

BSIJ before and after BMA administration

Figure 3 shows BSIJ progression in a representative case wherein the patient developed MRONJ. In this case, the pre-BMA BSIJ score 1 month before BMA administration was 0.00; however, it increased to 0.031 and 0.068 at 6 and 11 months, respectively. The patient developed MRONJ in the 12th month. Therefore, an increase in the BSIJ score after BMA administration may indicate a higher risk of MRONJ onset. Consequently, we examined changes in BSIJ score before and after BMA administration in both groups. In the MRONJ group, the post-BMA BSIJ score was significantly higher than the pre-BMA BSIJ score (p = 0.026). Conversely, in the non-MRONJ group, the pre-BMA BSIJ score was significantly higher than the post-BMA BSIJ score (p = 0.006; Figure 4).

To assess the influence of jawbone metastases on ΔBSIJ, we conducted a subgroup analysis comparing MRONJ-positive and MRONJ-negative patients with and without jawbone metastases. The results demonstrated that in MRONJ-negative patients without jawbone metastases, the median (IQR) ΔBSIJ was –0.012 (–0.015 to 0.023), whereas in MRONJ-negative patients with jawbone metastases, the median (IQR) ΔBSIJ was –0.045 (–0.09 to –0.02). In contrast, MRONJ-positive patients without jawbone metastases exhibited a median (IQR) ΔBSIJ of 0.05 (–0.02 to 0.09), and MRONJ-positive patients with jawbone metastases had a median (IQR) ΔBSIJ of 0.09 (0 to 0.17). Despite the small sample size, these findings suggest that ΔBSIJ tends to be negative in non-MRONJ cases and positive in MRONJ cases, regardless of the presence of jawbone metastases. This indicates that the influence of MRONJ on ΔBSIJ is more pronounced than that of jawbone metastases, supporting its utility as a predictive marker for MRONJ development.

MRONJ onset prediction using ΔBSIJ

The association among pre-BMA BSIJ, post-BMA BSIJ, ΔBSIJ, and MRONJ onset is presented in Table 2. Notably, no significant difference was observed in the pre-BMA BSIJ score between the MRONJ and non-MRONJ groups. Although the post-BMA BSIJ score tended to be higher in the MRONJ group, the difference was not statistically significant. However, ΔBSIJ was significantly higher in the MRONJ group (0.05, interquartile range: –0.01 to 0.11) than in the non-MRONJ group (–0.04, interquartile range: –0.11 to 0.00) (p = 0.002).

Next, we performed ROC analysis to evaluate the predictive performance of pre-BMA BSIJ, post-BMA BSIJ, and ΔBSIJ for MRONJ onset (Figure 5). The AUC was the highest for ΔBSIJ at 0.823, followed by post-BMA BSIJ at 0.713 and pre-BMA BSIJ at 0.596. The cutoff value of 0.039, corresponding to the point with the highest sensitivity ˗ (1 – specificity), was determined based on the ROC curve for ΔBSIJ. Of the 8 patients with ΔBSIJ ≥ 0.039, 6 developed MRONJ, whereas of the 25 patients with ΔBSIJ < 0.039, only 4 developed MRONJ (p = 0.004; Table 3). Similarly, the cutoff values for the pre- and post-BMA BSIJ were determined to be 0.031 and 0.068, respectively, from the ROC curve, and their predictive performances were compared (Table 4). ΔBSIJ demonstrated superior specificity (91%), positive predictive value (75%), and negative predictive value (84%) than the pre- and post-BMA BSIJ, whereas sensitivity (60%) was comparable among the metrics. Furthermore, patients with ΔBSIJ ≥ 0.039 exhibited significantly shorter MRONJ-free survival, with a median survival time of 18.4 months compared with those with ΔBSIJ < 0.039, which did not reach the median survival time during the observation period (p < 0.001; Figure 6).

This study highlights ΔBSIJ as a novel and clinically valuable marker for the early prediction of MRONJ in patients with prostate cancer receiving BMA therapy. Our results demonstrate that ΔBSIJ, defined as the difference between pre- and post-BMA BSIJ values, provides superior predictive accuracy compared to static measures such as pre-BMA and post-BMA BSIJ. The MRONJ group exhibited significantly higher ΔBSIJ values than the non-MRONJ group (0.05 vs. − 0.04, p = 0.002), and ROC analysis confirmed its diagnostic utility with an AUC of 0.823. Using a ΔBSIJ cutoff of 0.039, MRONJ could be predicted with 60% sensitivity and 91% specificity, indicating a strong ability to distinguish high-risk patients. Furthermore, Kaplan–Meier analysis revealed that patients with ΔBSIJ ≥ 0.039 had significantly shorter MRONJ-free survival compared to those with ΔBSIJ < 0.039. These findings underscore the clinical importance of ΔBSIJ as a dynamic biomarker that captures temporal changes in bone metabolism, facilitating early MRONJ risk assessment before structural changes become apparent. Additionally, as bone scintigraphy is routinely performed for metastatic prostate cancer management, ΔBSIJ represents a cost-effective and practical tool that can be seamlessly integrated into clinical workflows for MRONJ monitoring without the need for additional imaging procedures.

While the sensitivity of ΔBSIJ was 60%, the specificity was remarkably high at 91%. In clinical practice, especially in patients with metastatic prostate cancer receiving BMAs, high specificity is desirable to minimize unnecessary referrals and optimize targeted early interventions. This characteristic makes ΔBSIJ a useful tool for identifying high-risk patients who require closer monitoring or early consultation with an oral and maxillofacial surgeon.

MRONJ diagnosis has traditionally relied on structural imaging modalities, including radiography, CT, MRI, and cone-beam CT (CBCT) [4]. However, these techniques have certain limitations, particularly for early-stage detection. Radiography and panoramic imaging are widely accessible and offer valuable initial assessments; however, their sensitivity for detecting early bone changes is limited. CT and CBCT provide high-resolution imaging of bone structures, allowing detailed assessment of cortical erosion and trabecular density changes. These modalities are particularly effective for assessing advanced stages of MRONJ but exhibit limited utility in detecting early metabolic or vascular changes [11]. Although MRI is highly sensitive to soft tissue and marrow involvement, it is not always feasible for routine monitoring, particularly in patients with dental implants or extensive prosthodontics [11]. In contrast, bone scintigraphy detects subtle metabolic changes that often precede structural damage, making it an attractive option for the early diagnosis of MRONJ. Watanabe et al. demonstrated the utility of BSI for detecting abnormal tracer uptake in the jaw before the appearance of clinical symptoms [8]. However, previous studies are limited to static BSI measures that do not capture temporal changes in bone metabolism. The present study addresses this gap by introducing ΔBSIJ, offering a dynamic assessment that enhances the diagnostic power of bone scintigraphy.

This study innovatively utilized ΔBSIJ, which reflects dynamic changes in bone metabolism over time and is particularly valuable for distinguishing between patients with stable or improving bone conditions and those at risk of developing MRONJ. Static measures, such as pre-BMA BSIJ or post-BMA BSIJ, provide a snapshot of bone activity but cannot capture metabolic trends indicative of early disease progression. The ability of ΔBSIJ to capture these changes is valuable in cases where jawbone metastases coexist with the risk of MRONJ. Metastatic lesions often exhibit increased tracer uptake on bone scintigraphy, which can confound the static BSIJ measurements. However, ΔBSIJ effectively differentiates between changes associated with effective treatment of metastases (leading to decreased uptake) and those indicative of early necrosis (leading to increased uptake), ensuring accurate diagnosis and appropriate management of high-risk populations. Furthermore, patients with prostate cancer frequently develop jawbone metastases, which are more common than in patients with other malignancies. ΔBSIJ addresses these unique challenges by emphasizing temporal changes over static values. For instance, ΔBSIJ is typically negative in cases of jawbone metastases without MRONJ development owing to decreased tracer uptake following effective treatment. Conversely, in cases with MRONJ, ΔBSIJ is positive regardless of the metastases, presenting it as a robust tool for risk stratification across diverse clinical scenarios. Moreover, ΔBSIJ overcomes the limitation of region-specific thresholds for the maxilla and mandible [8], which has been highlighted as a notable challenge in previous studies using static BSIJ measures. This simplification enhances its clinical applicability, facilitating its easier implementation in routine practice without additional training or resources.

The integration of ΔBSIJ into clinical practice has significant implications for the management of patients with metastatic prostate cancer receiving BMAs. Current guidelines emphasize the importance of imaging for monitoring disease progression and treatment response, particularly in the era of advanced androgen receptor-targeting therapies [12–14]. These therapies often lower PSA levels to undetectable ranges, reducing the reliability of traditional biomarkers for disease monitoring. In such cases, imaging modalities, such as bone scintigraphy, are vital for detecting disease progression that may not be evident through PSA monitoring alone [14]. Early MRONJ detection can be achieved by incorporating ΔBSIJ into regular imaging protocols, prompting timely interventions to prevent disease progression and associated complications, thereby aligning with the principles of precision medicine, which prioritize personalized risk assessment and early detection to improve patient outcomes.

Although bone scintigraphy offers significant advantages for MRONJ detection, other functional imaging modalities such as fluorodeoxyglucose-positron emission tomography (FDG-PET) have also been explored [15, 16]. FDG-PET is highly sensitive in detecting metabolic activity, with promising results in identifying MRONJ lesions. However, it is impractical for routine use in most clinical settings owing to its high cost and limited availability. In contrast, bone scintigraphy is widely accessible, cost-effective, and has been integrated into the care of patients with metastatic prostate cancer. Additionally, ΔBSIJ further enhances its utility by offering a robust and affordable method for dynamic monitoring.

The BONENAVI® software [9, 10] improves the clinical utility of bone scintigraphy by automating BSI calculations, ensuring consistent and objective measurements. This also eliminates observer variability, a common limitation of qualitative imaging assessments, supporting the widespread adoption of the ΔBSIJ in clinical practice. Moreover, integration of artificial intelligence and extensive multicenter data in BONENAVI® further strengthens its reliability and reproducibility.

This study also has some limitations. It is subject to potential selection bias dowingue to its retrospective design. The timing of bone scintigraphy was determined by each attending physician rather than following a standardized protocol. Consequently, only 33 out of the 94 patients who received BMA therapy had both pre- and post-BMA imaging results available. This selection was not intentional but was based solely on imaging availability. While this reflects real-world clinical practice, it may limit the generalizability of our findings. Future prospective studies with standardized imaging protocols are needed to validate our results. Larger prospective studies are warranted to validate ΔBSIJ effectiveness across diverse populations and clinical settings. Additionally, although ΔBSIJ demonstrated significant predictive value, its accuracy can be enhanced further by integrating other clinical risk factors, such as age, diabetes or prolonged BMA use. The association between age and the development of MRONJ is supported by some studies but not consistently. While some studies have identified age as a significant predictor[17, 18], others have not found a statistically significant relationship. This discrepancy highlights the multifactorial nature of MRONJ and the necessity of considering multiple risk factors beyond age alone. Although prolonged BMA administration is a known risk factor for MRONJ [17, 19], in our study, the duration of BMA therapy did not significantly differ between the MRONJ and non-MRONJ groups. This suggests that ΔBSIJ serves as an independent predictor of MRONJ risk beyond the duration of BMA therapy. Exploring additional quantitative metrics, such as combining ΔBSIJ with other imaging biomarkers, may provide new insights into the pathophysiology and progression of MRONJ. Future research should also investigate the cost-effectiveness of implementing ΔBSIJ in routine practice and its impact on clinical outcomes. Correspondingly, ΔBSIJ can be further refined and optimized for widespread adoption in MRONJ management.

This study highlights ΔBSIJ as a novel and effective marker for the early prediction of MRONJ in patients with prostate cancer receiving BMAs. By capturing dynamic changes in jaw-specific bone metabolism, ΔBSIJ demonstrated superior predictive accuracy over static BSIJ measures, representing a robust clinically applicable tool. ΔBSIJ incorporation into routine bone scintigraphy offers a cost-effective and objective approach to enhance MRONJ risk assessment. Future research should focus on validating these findings in larger prospective cohorts and on exploring ΔBSIJ application across broader clinical contexts to optimize MRONJ management and improve patient outcomes.

Acknowledgement

The authors thank Hideki Nobira, Tomokazu Oshiki, and Masamichi Kajita from PDRadiopharma Inc., Tokyo, Japan for their invaluable support and collaboration in the planning of this study and interpretation of the results.

Conflict of interest

No potential conflicts of interest were disclosed.

Author contributions

All the authors contributed to the conception and design of this study. Hidetoshi Kokubun performed all radiographical assessments using BONENAVI_® software. Toshiki Kijima accumulated patient clinical information, including MRONJ development, and analyzed the predictive ability of BSI-related variables for MRONJ. Hidetoshi Kokubun and Toshiki Kijima wrote the first draft of the manuscript, and all authors commented on previous versions of the manuscript. All the authors have read and approved the final version of this manuscript.

Data availability

The datasets generated and/or analyzed in this study are available from the corresponding author upon reasonable request.

Study registration

N/A

Funding

None.

Murphy J, Mannion CJ. Medication-related osteonecrosis of the jaws and quality of life: review and structured analysis. Br J Oral Maxillofac Surg 2020;58:619–24.
Wei LY, et al. Prognosis of medication-related osteonecrosis of the jaws in metastatic prostate cancer patients. Oral Dis 2022;28:182–92.
Ristow O, et al. Treatment perspectives for medication-related osteonecrosis of the jaw (MRONJ). J Craniomaxillofac Surg 2015;43:290–3.
Stockmann P, et al. Panoramic radiograph, computed tomography or magnetic resonance imaging. Which imaging technique should be preferred in bisphosphonate-associated osteonecrosis of the jaw? A prospective clinical study. Clin Oral Investig 2010;14:311–7.
Yamamoto Y, et al. Quantitative bone scan imaging using BSI and BUV: an approach to evaluate ARONJ early. Ann Nucl Med 2020;34:74–9.
O’Ryan FS, et al. Intravenous bisphosphonate-related osteonecrosis of the jaw: bone scintigraphy as an early indicator. J Oral Maxillofac Surg 2009;67:1363–72.
Thomas C, et al. Bone scintigraphy predicts bisphosphonate-induced osteonecrosis of the jaw (BRONJ) in patients with metastatic castration-resistant prostate cancer (mCRPC). Clin Oral Investig 2016;20:753–8.
Watanabe S, et al. Bone scan index of the jaw: a new approach for evaluating early-stage anti-resorptive agents-related osteonecrosis. Ann Nucl Med 2017;31:201–10.
Ulmert D, et al. A novel automated platform for quantifying the extent of skeletal tumour involvement in prostate cancer patients using the Bone Scan Index. Eur Urol 2012;62:78–84.
Nakajima K, et al. Enhanced diagnostic accuracy for quantitative bone scan using an artificial neural network system: a Japanese multi-center database project. EJNMMI Res 2013;3:83.
Wongratwanich P, et al. Do various imaging modalities provide potential early detection and diagnosis of medication-related osteonecrosis of the jaw? A review. Dento Maxillo Fac Radiol 2021;50:20200417.
Bryce AH, et al. Radiographic progression with nonrising PSA in metastatic castration-resistant prostate cancer: post hoc analysis of PREVAIL. Prostate Cancer Prostatic Dis 2017;20:221–7.
Bryce AH, et al. Patterns of cancer progression of metastatic hormone-sensitive prostate cancer in the ECOG3805 CHAARTED trial. Eur Urol Oncol 2020;3:717–24.
Hara T, et al. Real-world analysis of metastatic prostate cancer demonstrates increased frequency of PSA-imaging discordance with visceral metastases and upfront ARAT/docetaxel therapy. Prostate 2023;83:1270–8.
Öztürk AE, et al. A comparison of 18 F-FDG PET/CT and 68 ga-PSMA PET/CT in detecting osteonecrosis of the jaw in a patient with prostate cancer. Clin Nucl Med 2024;49:e68–9.
Watanabe S, Nakajima K, Kinuya S. (☆)Symposium: Imaging modalities for drug-related osteonecrosis of the jaw (5), utility of bone scintigraphy and (18)F-FDG PET/CT in early detection and risk assessment of medication-related osteonecrosis of the jaw (secondary publication). Jpn Dent Sci Rev 2019;55:76–9.
Ciobanu GA, et al. Risk Factors for Medication-Related Osteonecrosis of the Jaw-A Binomial Analysis of Data of Cancer Patients from Craiova and Constanta Treated with Zoledronic Acid. J Clin Med. 2023; 12: 3747.
Kwack DW, Park SM. Prediction of medication-related osteonecrosis of the jaw (MRONJ) using automated machine learning in patients with osteoporosis associated with dental extraction and implantation: a retrospective study. J Korean Assoc Oral Maxillofac Surg. 2023; 49: 135-41.
Kwon YD, et al. A clinical retrospective study of implant as a risk factor for medication-related osteonecrosis of the jaw: surgery vs loading? Maxillofac Plast Reconstr Surg. 2023; 45: 31.

Table 1. Patient characteristics

Variables	Total (n=33)	MRONJ (n=10)	No MRONJ (n=23)	p-value
Age (years)	69 (64–75)	73 (69–79)	65 (63–74)	<0.01
BMI (kg/m²)	22.4 (19.7–25.4)	21.6 (18.3–25.7)	22.4 (20.0–25.5)	0.57
Diabetes mellitus	11 (33%)	4 (40%)	7 (30%)	0.70
PSA at diagnosis (ng/ml)	257.5 (20–1015)	127 (21.4–723)	298 (16.9–1527)	0.71
Gleason grade group				0.18
<4	3 (9%)	1 (10%)	2 (9%)
≥4	29 (88%)	9 (90%)	20 (87%)
Unknown	1 (3%)		1 (4%)
Prior tooth extraction owing to dental conditions	14 (42%)	5 (50%)	9 (39%)	0.71
Jawbone metastases	14 (42%)	2 (20%)	12 (52%)	0.13
Types of Androgen Deprivation Therapy
GnRH agonist	18 (55%)	4 (40%)	14 (61%)	0.448
GnRH antagonist	15 (45%)	6 (60%)	9 (39%)
Typres of bone modifying agents				1.00
Denosumab	28 (85%)	9 (90%)	19 (83%)
Zoledronic acid	5 (15%)	1 (10%)	4 (17%)
Interval between pre-BMA bone scan and BMA administration (months)	1.6 (0.8–4.3)	1.5 (0.8–5.1)	2.0 (0.9–3.8)	1.00
Interval between BMA administration and post-BMA bone scan (months)	13.0 (7.6–23.2)	10.3 (8.8–17.9)	14.6 (7.1–24.3)	0.57
Interval between BMA administration and MRONJ onset (months)		17.9 (15.9–21.5)
Interval between post-BMA bone scan and onset of MRONJ (months)		7.6 (0.9–9.2)

Values are presented as N (%) or median (interquartile range).

Table 2. Relationship between the development of MRONJ and BSIJ-related values

	MRONJ (n=10)	No MRONJ (n=23)	p-value
Pre-BMA BSIJ	0.10 (0.02 to 0.12)	0.08 (0.05 to 0.18)	0.265
Post-BMA BSIJ	0.12 (0.06 to 0.22)	0.05 (0.02 to 0.10)	0.084
ΔBSIJ	0.05 (–0.01 to 0.11)	–0.04 (–0.11 to 0.00)	0.002

Values are presented as median (interquartile range).

Table 3.

	MRONJ (n=10)	No MRONJ (n=23)	p-value
ΔBSIJ ≥ 0.039 (n=8)	6 (60%)	2 (9%)	0.004
ΔBSIJ < 0.039 (n=25)	4 (40%)	21 (91%)

Values are presented as N (%).

Table 4. Screening accuracy of BSIJ-related values for predicting MRONJ

	Cut-off	Sensitivity	Specificity	PPV	NPV
Pre-BMA BSIJ	0.031	70%	13%	26%	50%
Post-BMA BSIJ	0.068	80%	70%	53%	88%
ΔBSIJ	0.039	60%	91%	75%	84%

ΔBSIJ: A quantitative marker for early detection of medication-related osteonecrosis of the jaw in patients with prostate cancer receiving bone-modifying agents

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Materials and methods

Results

Discussion

Conclusion

Declarations

References

Tables

Status:

Journal Publication

Version 1