Ⅰ. INTRODUCTION
Bisphosphonate (BP) induced avascular necrosis of the jaw was first reported by Marx in 20031), but of the various bisphosphonates used to treat osteoporosis, only those containing nitrogen cause avascular necrosis2,3). This disease has been named "bisphosphonate-related osteonecrosis of jaw (MRONJ)" and many studies have been undertaken to determine its cause, but, no definite pathogenesis has been elucidated.
In 2014, American Association of Oral and Maxillofacial Surgeons (AAOMS) reported that antiangiogenic and antiresorptive agents also caused avascular necrosis of the jaw, and proposed the disease be called medication-related osteonecrosis (MRONJ)4). In addition, MRONJ disease prevalence was suggested for different medication routes of bisphosphonate, that is, a prevalence of 0.1% ~ 0.21% when it was administered orally for ≥ 4 years and a prevalence of 0.017% when administered intravenously for ≥ 6 years4). Research on MRONJ is being actively conducted in Korea. In 2015, the Korean Society of Oral and Maxillofacial Surgery (KAOMS) presented MRONJ treatment guidelines based on results of a joint study conducted with the Korean Society of Bone and Mineral Research (KSBMR). This study found the prevalence of MRONJ was 0.04% in Korean patients, and its prevalence in patients with any malignancy disease was 1.86%5). However, published prevalence was based on data collected from only limited number of medical institutions.
The diagnostic criteria of MRONJ are as follows. (1) Present or past use of an antiangiogenic agent or antiresorptive agent, (2) Exposure of the jaw in the intraoral environment or probing of the jaw through an internal or external oral fistula of duration > 8 weeks, and (3) no prior history of radiation therapy and no metastatic disease. The MRONJ staging system was modified in the 2014 AAOMS paper. Changes were as follows; 1) An at risk stage was added - patients treated with an antiangiogenic or an antiresorptive agent with no osteonecrosis. 2) Stage 0 - nonspecific symptoms and radiological changes but no clinical evidence of bone necrosis. 3) Stage 1 - necrotic bone present or probed through a fistula but without symptoms or evidence of infection. 4) Stage 2 - exposed necrotic bone or probed bone necrosis through a fistula accompanied by symptoms and evidence of infection (Fig. 5). 5) Stage 3 - necrosis progression beyond alveolar bone, necrosis with associated pathological fracture, necrosis associated extraoral, oroantral, or oronasal fistula, or necrosis progression to the mandibular inferior border or maxillary sinus floor (Fig. 1).
The treatment recommended by 2014 AAOMS paper for MRONJ is conservative and non-invasive if possible, except for stage 3. The ultimate goal of treatment is to educate the patient regarding the disease, control pain and infection, and prevent further disease progression, which should be achieved if possible with minimal intervention. Treatment strategies according to MRONJ stages are as follows: 1) at risk - no treatment is required, but the patient is educated about the disease. 2) Stage 0 - manage systemic conditions using analgesics or antibiotics. 3) Stage 1 - prescribe an antibacterial mouth rinse and perform follow-up checks quarterly. Educate the patient about the disease and reevaluate antiangiogenic or antiresorptive agent taking. 4) Stage 2 – Per oral analgesics and antibiotics should be used to treat symptoms, and debridement is performed to relieve soft tissue irritation and control infection. 5) Stage 3 - Analgesics and antibiotics are administered and surgical debridement or resection is performed to relieve pain and infection. In addition to this MRONJ staging strategy, specific treatment strategies were also introduced in the 2014 AAOMS position paper to prevent MRONJ occurrence4). The position paper recommended to treat oral inflammation before using antiangiogenic or antiresorptive agents. MRONJ risk should be considered if antiangiogenic or antiresorptive agents are taken for > 4 years, also jaw injury must be minimized. And if surgical treatment is needed, drug-free period (drug holiday) be instituted before surgery. The drug holiday of peroral medication of an antiangiogenic agent or antiresorptive agent is given for about 2 months. However, evidence on the efficacies of intravenous medications is lacking. Nevertheless, it is recommended nonsurgical treatment be considered prior to surgical treatment in patients with MRONJ4).
Much controversy exists about AAOMS treatment strategies, because non-surgical treatment does not ensure patient quality of life (QOL). In MRONJ stage 2, surgical treatment has a better prognosis than conservative treatment 6,7,8,9). In the present study, Therefore, we undertook to provide guidance for the treatment of MRONJ by comparing and analyzing the outcomes of stage 2 MRONJ patients treated using different methods.
Ⅱ. MATERIAL and METHODS
1. Patients
This study was approved by the Institutional Review Board of Pusan National University Dental Hospital (PNUDH- 2018-034). The Institutional Review Board of Pusan National University Dental Hospital complies with the Helsinki Declaration. The subjects were patients treated for MRONJ at the Oral and Maxillofacial Surgery Department of Pusan National University Dental Hospital from January 2015 to December 2017. Diagnostic criteria and disease stages were as recommended by AAOMS, and only patients diagnosed with MRONJ stage 2 were selected. The total number of patients treated by a single operator during 2015-2017 was 25. Patients were allocated to a MRONJ group or a non- MRONJ group (jaw bone necrosis but with no history of an antiangiogenic or antiresorptive agent). The MRONJ group contained 14 subjects (1 male, 13 females) of average age 74.36 ± 7.76 years, and the non-MRONJ group contained 11 subjects (5 males, 6 females) of average age 69.64 ± 7.92 years.
2. Treatment methods
The medical records of the 25 study subjects were analyzed retrospectively. The analysis factors were; 1) sex and age, 2) antiangiogenic or antiresorptive agent medication method and duration (MRONJ group only), 3) underlying disease (non-MRONJ group only), 4) cause of osteonecrosis, 5) location of osteonecrosis, 6) treatment method, and 7) treatment outcome. Treatment methods were divided into two categories (Fig. 2. 1) “Conservative treatment” included medications, such as, antibiotics and anti-inflammatories and chlorhexidine mouth rinses. 2) “Surgical treatment” included “sequestrectomy” to extract formed sequestrum and “saucerization” to remove the suspected part if sequestrum formation was not clear. Treatment results were divided into two categories. 1) "success" was defined as the formation of new epithelium in the treated area with no clinical symptom or evidence of infection. 2) "failure" was defined as the presence of infection and clinical signs of recurrence.
3. Statistical analysis
The treatment results were classified as "successes" or "failures" by treatment methods (sequestrectomy, saucerization, conservative treatment). Data were analyzed using the Mann-Whitney t-test in the R language program version 3.3.3 (R Foundation for Statistical Computing, Vienna, Austria).
Ⅲ. RESULTS
1. Age and gender
The MRONJ group contained 14 subjects (1 male, 13 female) of average age 74.36 ± 7.76 years, and the non- MRONJ group contained 11 subjects (5 male, 6 female) of average age 69.64 ± 7.92 years. Percentages of females in the MRONJ and non-MRONJ groups were 92.86% (13 patients) and 54.55% (6 patients), respectively. Accordingly, the occurrence of MRONJ was higher in females than males (M/F ratio 1:13; Table 1).
2. Methods and durations of antiangiogenic or antiresorptive agent treatment in the MRONJ group
Mean durations of peroral (PO) or intravenous (IV) medication of antiangiogenic or antiresorptive agents in the MRONJ group were 35.5 ± 30.42 months and 44.25 ± 51.93 months, respectively. The longest PO medication period was 96 months, and the longest IV medication period was 120 months. IV medication was administered to 28.57% (4 cases) and PO medication to 71.43% (10 cases). MRONJ treatment failed in 3 patients, 2 were administered PO and 1 was administered Ⅳ medication. Mean durations of antiangiogenic or antiresorptive agent administration were 60 ± 50.91 and 36 months, respectively (Table 2).
3. Underlying disease in the non-MRONJ group
The prevalence of diabetes mellitus, cirrhosis, stomach and prostate cancer, and of no underlying disease were 36% (4 cases), 9% (1 case), 18% (2 cases) and 36% (4 cases), respectively. Underlying diseases of treatment failures were diabetes in 1, prostate cancer in 1, and no underlying disease in 1. In terms of treatment methods, 2 were treated conservatively and 1 underwent saucerization (Table 3).
4. Causes of jaw necrosis
Types of surgical trauma that resulted in bone exposure to the oral environment and jaw necrosis were investigated. In the MRONJ group, tooth extraction accounted for 85.74% (12 cases), implantation for 7.14% (1 case), and periodontal disease for 7.14% (1 case). In the non-MRONJ group, tooth extraction accounted for 72.73% (8 cases) and periodontal diseases for 27.27% (3 cases). Treatment failures in the MRONJ group occurred in 2 patients that had undergone tooth extraction and in 2 patients with periodontitis. Treatment failures in non-MRONJ group occurred in 3 patients that had undergone tooth extraction (Table 4).
5. Location of jaw necrosis
All MRONJ group diseases were observed in the posterior area. The percentage of patients with a maxillary posterior area was 42.86% (6 cases) and the percentage with a mandibular posterior location was 57.14% (8 cases). A maxilla to mandible ratio of 3: 4. 10 patients in the non- MRONJ group had disease in the posterior area, and the other had disease in the mandibular anterior area. The percentage of patients with a maxillary posterior location was 36.36% (4 cases) and that of patients with a mandibular posterior location was 54.55% (6 cases). A maxilla to mandible ratio of 4: 7. Two patients with a mandibular posterior location and 1 patient with a maxillary posterior location experienced treatment failure; two were treated conservatively and one by saucerization (Table 5).
6. Treatment methods
Conservative treatment 14.29% (2 cases), sequestrectomy 64.29% (9 cases) and saucerization 21.43% (3 cases) were performed in the MRONJ group, and 2 treatment failures were encountered after conservative treatment and 1 after saucerization. Two patients (18.18%) in the non-MRONJ group received conservative treatment, 7 (63.64%) underwent sequestrectomy and 2 (18.18%) saucerization. Two treatment failures were occurred after conservative treatment and 1 after saucerization. Mann-Whitney t-test results for different treatment methods are summarized in Table 7. No statistical significance was observed. Table 6
7. Treatment outcomes
In the MRONJ group, treatment success was achieved in 78.57% (11 cases) and treatment failure in 21.43% (3 cases). Two treatment failures received conservative treatment and 1 underwent saucerization. The mean duration of antiangiogenic or antiresorptive agent administration was 52 ± 38.57 months, and the medication method used were peroral in 2 and intravenous in one. In the non-MRONJ group, treatment success was achieved in 72.73% (8 cases) and treatment failure in 27.27% (3 cases). Two of the three treatment failures received conservative treatment and 1 underwent saucerization. Underlying diseases were diabetes, prostate cancer, and no underlying disease (Table 8).
8. Histopathologic Finding
Histopathologic examination of the sequestra was performed. Along with necrotic bone tissue, numerous acute and chronic inflammatory cells infiltrated (Fig. 3).
Ⅳ. DISCUSSION
This retrospective study was conducted to identify an appropriate treatment method for stage 2 MRONJ. Our analysis of risk factors produced results that concurred with AAOMS position paper conclusion, and showed a high rate of morbidity and high prevalence’s of a mandibular location and a female gender4). It has been reported MRONJ is caused by pre-existing inflammatory dental disease (such as periodontal disease) and surgical trauma to alveolar bone4). The risk of MRONJ has been shown to increase when an antiangiogenic or antiresorptive agent has been administered for > 4 years. The 2014 AAOMS paper reported a mean medication duration without MRONJ of 3.5 years4). However, in the present study, mean medication duration in the MRONJ group was 38.00 ± 35.77 months. It is generally believed methods of administering antiangiogenic or antiresorptive agents influence the risk of MRONJ occurrence. However, no significant difference was observed in the present study.
The treatment of MRONJ continues to be controversial. Hayashi et al. summarized the situation in an article published in 2017 and concluded no consensus has been reached regarding the most appropriate treatment6). AAMOS maintains the recommendation that MORNJ be treated as conservatively as possible in their 2014 position paper and 2015 update version4). The final goal of treatment is to improve patient understanding of the disease, control pain and infection, and prevent disease progression. Accordingly, surgical treatment should only be used when non-surgical attempts have failed, because it inevitably results in further bone exposure. This was also alluded to in the Journal of the Japanese Society for Bone and Thoracic Surgery (JSBMR)10). However, reports continue to claim that the prognosis of surgical treatment is better. Eguchi et al. concluded surgical treatment was more effective in 25 patients, and reported successful results for 89.3% of surgical treatments and of 33.3% for non-surgical treatments11). Furthermore, Hayashida et al. stated sequestrum rarely recover to normal bone and non-surgical treatment demands periodic visits, and thus, negative affects quality of life (QOL). They found healing in 76.7% of 361 stage 2 MRONJ patients that received surgical treatment, but in only 25.2% of patients treated conservatively. These results suggest that active surgical intervention is more effective for stage 2 MRONJ patients6). In addition, the prognosis of surgical treatment is better when the range of operation is wider6,7,9,12). The results of this study are like studies of Eguchi and Hayashida. Stage 2 MRONJ had a healing rate of 78.57% when treated surgically, and the non-MRONJ group had a healing rate of 72.73%. Although the number of study subjects was too small to determine the significance of this difference, we did find surgical treatment more favorable. Regarding the three treatment failures in the MRONJ group, 2 patients received conservative treatment and 1 patient underwent saucerization, and similar results were observed in the non-MRONJ group. On the other hand, conservative treatments did not result in healing and infections and jaw necrosis spread during treatment, and as a result, lesion sizes and stages increased.
For surgical treatment, we found healing state depended on surgical range, which concurs with the results of Fliefel et al.7). Treatments were deemed successful when surrounding necrotic bone containing sequestrum was reliably removed. We believe that the stimuli applied to surrounding normal jaw and soft tissue were minimized by the widespread removal of necrosis around sequestrum. However, if only sequestrum exposed in the oral cavity is removed (as recommended by AAOMS), adjacent necrotic bone remains in a state of continuous infection, interferes with soft tissue healing, and results re-exposure of necrotic jaws. As a result, patient visits frequency is increased, and QOL of patients is threatened. Therefore, we believe choosing a suitable surgical option that involves removal of sequestrum and surrounding infected bone provides better results than conservative treatment in stage 2 MRONJ patients.
Another factor affecting prognosis is the duration of antiangiogenic or antiresorptive agent medication. The effect of medication duration on the incidence of MRONJ is detailed in the 2014 AAOMS paper4). A drug-free period must be considered when dental treatment is needed for patients with osteoporosis or metastatic malignant disease. However, studies on such drug holidays for antiangiogenic or antiresorptive agents also show varying results. Hellstein et al.13) found no drug holiday was required for a BPs duration of < 2 years. But in another study it was stated that if BP duration is > 4 years, a drug holiday should be maintained to allow bone healing.14).
AAOMS suggested that a drug holiday of about 2 months is appropriate based on consideration of BP half-live4), and others have recommended drug holidays of > 4 months15). The Korean Society of Bone and Mineral Research (KSBMR (2015)) recommended a drug holiday if the following conditions are met. 1) If peroral BP medication has been administered for < 4 years, no special consideration is required. 2) If peroral BP medication has been administered for < 4 years, but other drugs, such as, steroids were co-administered, a 2 ~ 4 months drug holiday is needed. 3) If peroral BP medication has been administered for ≥ 4 years and there is no other systemic risk factor, a 2 ~ 4 months drug holiday is needed, and if the patient's systemic condition permits the drug holiday be extended until bone has completely healed. 4) If BP has been administered IV, caution is needed because of a lack of information on the topic5).
Several authors have suggested a minimum drug holiday duration of 2 months for those taking oral BPs. However, in a multicenter retrospective study, Hasegawa et al. found that drug-holiday-used and non-used groups showed similar healing after tooth extraction, and suggested rapid removal of the cause of infection in alveolar bone promoted early healing16). In the present study, no drug holiday was used, and surgical intervention was undertaken without stopping antiangiogenic or antiresorptive agents and without restoring bone healing ability. In a few numbers of patients showed lesions enlarged or recurred during follow-up. Possible factors that contributed include treatment method, gender, and the absence of a drug holiday.
The limitations of this study are as follows: 1) The number of subjects was too small to perform statistical analysis, a larger-scale study is required to obtain more reliable results. 2) No evidence was obtained on the effects of surgical site primary closure or on drug type on healing after sequestrectomy or saucerization.