Recent studies have indicated that bone shows auto-fluorescence under an appropriate fluorescence lamp. The aim of this preliminary study was to compare the success rates of the established tetracycline fluorescence-guided bone surgery with auto-fluorescence-guided bone surgery in the treatment of medication-related osteonecrosis of the jaw (MRONJ). Forty patients suffering from MRONJ were referred for surgical treatment and were divided randomly into two groups: auto-fluorescence ( n = 20) or tetracycline fluorescence ( n = 20) guided bone surgery. The primary endpoint was treatment success, defined as the absence of exposed bone at 8 weeks after surgery. Secondary outcomes assessed were mucosal integrity, signs of infection, pain, and loss of sensitivity; these were evaluated descriptively at 10 days, 8 weeks, 6 months, and 1 year after surgery. At 8 weeks postoperative, 18/20 patients (90%) in the auto-fluorescence group and 17/20 patients (85%) in the tetracycline fluorescence group showed mucosal integrity ( P > 0.05). At the last follow-up, 94% in the auto-fluorescence group and 89% in the tetracycline fluorescence group presented complete mucosal coverage with no exposed bone, infection, or pain ( P > 0.05). There was no significant difference between the two techniques for any of the secondary outcomes ( P > 0.05). The results of this preliminary study show that auto-fluorescence-guided bone surgery has comparable success rates to the established tetracycline fluorescence-guided bone surgery.
The primary goal of surgical therapy for medication-related osteonecrosis of the jaw (MRONJ) should always be the removal of as much bone as necessary and as little as possible. Irrespective of the approach selected (minimally invasive or resective), the delineation between necrotic and viable bone is the crucial step and is a major challenge in this procedure. Indeed, the complete removal of necrotic bone is essential, as otherwise the risk of disease recurrence or progression remains.
Prior to the introduction of tetracycline fluorescence-guided bone removal, surgical experience and the surgeon’s subjective impression, supported by various imprecise imaging modalities, were the only parameters available for distinguishing between healthy and diseased bone tissue. In this respect, tetracycline fluorescence-guided bone surgery has had an important impact on the surgical management of MRONJ, as it successfully addresses this shortcoming. Providing an objective and reproducible therapeutic approach, this technique enables the transitions between necrotic and non-necrotic bone to be defined during the surgical procedure.
Recent studies have indicated that the VELscope system (VELscope fluorescence lamp; LED Dental, White Rock, British Columbia, Canada) induces an auto-fluorescence from vital bone (but not from necrotic bone) leading to similar bone fluorescence findings without the preceding tetracycline labelling. In a preliminary investigation, promising results were found for this technique with regard to success rates and verification of the complete removal of the necrotic bone by histological work-up. Because this surgical approach is easy to apply, is reproducible, and does not rely on the subjective impression of the surgeon, it might be used to standardize surgical MRONJ therapy and lead to an improvement in treatment.
Therefore, the purpose of this study was to examine the success rate of auto-fluorescence-guided bone surgery and to compare this technique with the valuable method of tetracycline fluorescence-guided bone surgery in MRONJ patients. Given that tetracycline-induced bone fluorescence is beneficial for the surgical treatment of MRONJ, the investigators hypothesized that the auto-fluorescence of bone would be similarly useful. The specific aims of this preliminary study were to compare the two intervention groups in terms of (1) postoperative mucosal integrity and absence of bone exposure, (2) signs of infection, (3) pain scores, and (4) loss of sensitivity (numbness) after the fluorescence-guided bone surgery.
Materials and methods
This study followed the Declaration of Helsinki concerning medical protocol and ethics and the CONSORT guidelines for reporting clinical trials. After the approval of the institutional ethics committee had been obtained, informed consent was acquired from all patients. To address the research purpose, the investigators designed and implemented an open-label, parallel-group, randomized, feasibility clinical trial. Over a period of 12 months, the study population was prospectively referred for the treatment of MRONJ and divided randomly into two study groups: treatment with either (1) auto-fluorescence-guided bone surgery (AF group; intervention group) or (2) tetracycline fluorescence-guided bone surgery (TF group; control group).
The following inclusion criteria were applied: (1) exposed osteonecrosis of the jaw, defined as the long-standing (more than 8 weeks) transmucosal exposure of necrotic bone in the jaw with a drug history positive for anti-resorptive treatment (in accordance with the American Association of Oral and Maxillofacial Surgeons (AAOMS) ); (2) a history of anti-resorptive drug treatment (bisphosphonates and/or denosumab) in the absence of radiotherapy to the head and neck region (in accordance with the AAOMS ). Exclusion criteria were a history of head and neck irradiation, metastatic bone disease of the maxillofacial region, and contradictions to surgery under general anaesthesia.
The primary endpoint of this feasibility study was to determine any superiority or consistency of the auto-fluorescence-guided bone surgery group over the tetracycline fluorescence-guided bone surgery group in terms of the success rate. Success was defined as the absence of a MRONJ site after surgery, specified as the maintenance of full mucosal coverage (mucosal integrity) at 8 weeks after surgery. Secondary endpoints were to identify between-study group differences in (1) mucosal integrity at the remaining measurement time points, (2) loss of sensitivity (numbness) of the alveolar nerve (Vincent sign), (3) subjective pain, and (4) signs of infection at 10 days (T1), 8 weeks (T2), 6 months (T3), and 1 year (T4) after surgery. Patient demographics and baseline characteristics were also recorded for confirmation of group comparability.
Fluorescence-guided bone surgery
All surgical procedures were performed by one of two board-certified and specialized oral and maxillofacial surgeons (C.P., B.H.M.) under general anaesthesia using nasal intubation. Surgery was performed under sterile conditions following a standardized operation protocol. All patients in the TF group received 100 mg doxycycline twice a day for at least 7 days preoperatively. Patients in the AF group received ampicillin/sulbactam 2000 mg/1000 mg before the operation; however, no preoperative tetracycline labelling was performed. Patients who reported a history of hypersensitivity to penicillin or a penicillin allergy received clindamycin 600 mg three times a day instead. All patients included in this study were then treated surgically by means of the fluorescence-guided bone surgery technique, as described previously by this study group, using the VELscope Vx system (LED Dental) ( Fig. 1 ) to induce and visualize fluorescence of the jawbone.
Briefly, after surgical bone exposure, bone fluorescence showed viable bone with a bright greenish fluorescence, whereas necrotic bone areas showed no or only very pale fluorescence in both the AF group ( Fig. 2 ) and TF group ( Fig. 3 ). Specimens were harvested from the pale bone areas and were confirmed histopathologically as necrotic bone for all patients treated. Bone surgery was performed until a bright fluorescence of the affected bone was observed (for the AF group ( Fig. 2 ) and for the TF group ( Fig. 3 ), as described in previous studies ); i.e. all necrotic bone had been removed and bone fluorescence regained. Histopathological work-up performed postoperatively confirmed fluorescent areas as viable bone in all cases treated. There was no apparently observable difference between the auto-fluorescence and the tetracycline fluorescence groups.
A tension-free wound closure was achieved using mucoperiosteal flaps. All patients remained in hospital for 4 days after the operation. Patients received routine postoperative instructions and the same postoperative analgesic drug therapy. Antibiotic treatment involved the administration of ampicillin (or clindamycin instead as described above)/sulbactam 2000 mg/1000 mg intravenously while in hospital and then orally for a further 6 days after discharge from the hospital.
All measurements were acquired at five specific time points: preoperative (T0) and at 10 days (T1), 8 weeks (T2), 6 months (T3), and 1 year (T4) after surgery. The primary outcome variable, namely mucosal integrity after surgery, was defined as the maintenance of full mucosal coverage after the operation, without signs of residual infection or exposed bone at the time of the evaluation. Infection was thoroughly evaluated by assessing swelling, redness, bleeding on probing, and any purulent discharge. Pain was assessed with a ‘yes’ or ‘no’ score. Trigeminal sensitivity of the mandibular nerve (numbness) was examined and recorded following a standardized protocol (Vincent sign). Furthermore, all patients were examined for signs of sinusitis and were checked for oro-antral fistulae in the case of upper jaw lesions.
Because of the preliminary ‘proof-of-concept’ character of this study, the sample size estimation was disclaimed. In order to estimate an effect size for this explorative study, 40 patients ( n = 20 each group) were recruited over a time period of 12 months and followed up for 12 months. The primary endpoint of the study was to show the superiority or the consistency of AF treatment over TF treatment in terms of mucosal integrity at T2 (8 weeks after surgery); this was analyzed using the χ 2 test. Risk differences were calculated for possible future power estimations of a confirmatory phase III study. In accordance with the preliminary character of the study and the consequent small numbers of cases, the application of multifactorial regression models (considering covariance such as medication type or application time) was renounced. However, this might be undertaken in a subsequent confirmatory phase III study.
In addition to the descriptive statistics, the study groups were compared with each other using the two-sided independent samples t -test for differences in means. The significance level was set at 5%. The primary endpoint of the study was mucosal integrity at T2 (see the Study variables section for the definition). All other statistical tests were secondary endpoints of the study and were performed in an exploratory manner. All secondary outcomes were evaluated descriptively, and descriptive P -values were reported together with 95% confidence intervals for the corresponding effects. Missing values were not imputed. Due to the uncertainty of the parameter estimates merging from this proof-of-concept study and also in order to account for expected dropouts, it was decided to recruit 40 patients (a total of 20 in each group). All statistical analyses were performed using IBM SPSS Statistics version 20.0 (IBM Corp., Armonk, NY, USA).
A total of 40 patients, 26 female and 14 male with a mean age of 71.80 years (standard deviation ±9.4 years), were randomized in this study. All detailed patient-specific clinical and demographic characteristics are listed in Table 1 . The two treatment groups were similar with respect to baseline characteristics ( P > 0.05). Three of the 40 randomized patients (one in the TF group and two in the AF group) died after the primary endpoint at T2 (8 weeks after the operation), but before 6 months after surgery (T3). Another patient (in the TF group) died between the follow-up at 6 months (T3) and the follow-up at 1 year after surgery (T4). Two patients in the AF group failed to attend the 1-year follow-up (T4).
|Tetracycline fluorescence ( n = 20)||Auto-fluorescence ( n = 20)||Total ( n = 40)|
|Age, years||67.3 ± 9.5||71.0 ± 8.8||71.80 ± 9.4|
|Metastasis other location|
|Type of drug application|
|IV + PO||1||1||2|
|Interval of application|
|Period of application, months||47.3 ± 24.6||47.0 ± 30.8||47.10 ± 27.7|
Thirty-four of the study patients (85%) were suffering from an underlying malignant disease, whereas the anti-resorptive treatment had been prescribed for osteoporosis in the remaining six patients (15%). Of the 40 study patients, 32 (80%) had been treated with nitrogen-containing bisphosphonates. None of the patients had a history of treatment with denosumab alone; however, eight patients (20%) reported a subsequent or alternating intake of bisphosphonates and denosumab. The most common anti-resorptive drugs used by the cohort were zoledronate ( n = 17; 43%) for the bisphosphonate and denosumab ( n = 4; 10%) for the denosumab application. The mean duration of intake of the anti-resorptive drugs was 47.3 ± 24.6 (range 16–94) months for the TF group and 47.0 ± 30.8 (range 6–122) months for the AF group. Group-specific distributions are given in Table 1 .
All lesion-specific baseline characteristics and their distributions in the two intervention groups are given in Table 2 . There was no statistically significant difference between the two groups for lesion-specific baseline characteristics ( P > 0.05). The 40 patients had 51 MRONJ lesions; 33 of the lesions (65%) were located in the mandible and 18 (35%) were located in the maxilla. The majority of the lesions were classified as being stage II ( n = 41; 80%); six lesions (12%) were graded as stage III and four lesions (8%) as grade I according to the AAOMS classifications 2009 and 2014. Of the 51 MRONJ lesions evaluated, 24 (47%; TF n = 16, AF n = 8) were colonized with actinomycetes.
|Tetracycline fluorescence ( n = 26)||Auto-fluorescence ( n = 25)|
|Left + right||0||1|
|I + II||0||1|
The mucosal integrity after fluorescence-guided bone surgery for the TF and AF groups, at all measurement points (T0 to T4), are displayed in Table 3 . The primary endpoint of the study, namely mucosal integrity at 8 weeks (T2) after fluorescence-guided bone surgery, showed no statistically significant difference ( P > 0.05) between the AF group (integrity ‘yes’ for 18/20 patients (90%) and ‘no’ for 2/20 patients (10%)) and the TF group (integrity ‘yes’ for 17/20 patients (85%) and ‘no’ for 3/20 patients (15%)). Lesion-specific results for mucosal integrity are also displayed in Table 3 .
|T0||No||20/20 (100.0%)||26/26 (100.0%)||20/20 (100.0%)||25/25 (100.0%)|
|Yes||0/20 (0.0%)||0/26 (0.0%)||0/20 (0.0%)||0/25 (0.0%)|
|T1||No||3/20 (15.0%)||3/26 (11.5%)||1/20 (5.0%)||1/25 (4.0%)|
|Yes||17/20 (85.0%)||23/26 (88.5%)||19/20 (95.0%)||24/25 (96.0%)|
|T2||No||3/20 (15.0%)||3/26 (11.5%)||2/20 (10.0%)||2/25 (8.0%)|
|Yes||17/20 (85.0%)||23/26 (88.5%)||18/20 (90.0%)||23/25 (92.0%)|
|T3||No||3/19 (15.8%)||4/23 (17.4%)||2/18 (11.1%)||2/21 (9.5%)|
|Yes||16/19 (84.2%)||19/23 (82.6%)||16/18 (88.9%)||19/21 (90.5%)|
|T4||No||2/18 (11.1%)||2/22 (9.1%)||1/16 (6.3%)||1/19 (5.3%)|
|Yes||16/18 (88.9%)||20/22 (90.9%)||15/16 (93.8%)||18/19 (94.7%)|
a Patient- and lesion-specific results for mucosal integrity in the two study groups at all evaluated measurement points: preoperative (T0) and at 10 days (T1), 8 weeks (T2), 6 months (T3), and 1 year (T4) postoperative. Results are given as the total number of patients/lesions that were actually diagnosed with the corresponding percentage in parenthesis. There was no statistically significant difference between the two groups at any of the measurement points evaluated ( P > 0.05).
It should be noted that four of the 40 patients evaluated (TF group n = 3 (15%); AF group n = 1 (5%)) with four of the 51 lesions (TF group n = 3 (12%); AF group n = 1 (4%)) showed stage improvement and were free of pain, but still had mucosal dehiscence at 10 days after surgery (T1). For all of these patients, the lesions were extensive (AAOMS stage III). One patient at this stage (TF group) with a mandible lesion and presenting an extraoral fistula had to undergo a second surgical intervention involving fluorescence-guided bone surgery after T3 (6 months after surgery). Two patients treated conservatively after surgery (TF group n = 1; AF group n = 1) showed mucosal integrity at T2 (8 weeks after surgery) that lasted with full mucosal healing without signs of infection or pain throughout T4, 1 year after surgery. One patient presenting a lesion in the maxilla (AAOMS stage III) combined with an oro-antral fistula (TF group) rejected a second surgical intervention. Interestingly, all of these cases had received anti-resorptive medicine for a period longer than 50 months.
At the time of last follow-up (1 year after surgery), 16 of 18 (89%) remaining patients in the TF group showed full mucosal integrity without concomitant symptoms (20 of 22 (91%) lesions), whereas 15 of 16 (94%) remaining patients in the AF group (18 of 19 (95%) lesions) were free of symptoms with complete mucosal closure at T4 ( P > 0.05). All results for mucosal integrity after T2 are secondary endpoints of the study and are therefore described descriptively.
The results of all secondary endpoints evaluated (signs of infection, pain, and numbness), comparing the AF group with the TF group, are displayed in Table 4 . There was no statistically significant difference between the two groups for any of the parameters evaluated at any of the measurement points ( P > 0.05).