Animal models for bisphosphonate-related osteonecrosis of the jaws - an appraisal. D Sharma1, S Hamlet1, E Petcu2, S Ivanovski1. 1School of Dentistry and ...
Oral Diseases (2013) 19, 747–754 doi:10.1111/odi.12067 © 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd All rights reserved www.wiley.com
REVIEW ARTICLE
Animal models for bisphosphonate-related osteonecrosis of the jaws - an appraisal D Sharma1, S Hamlet1, E Petcu2, S Ivanovski1 1
School of Dentistry and Oral Health, Griffith Health Institute, Griffith University, Gold Coast, Qld; 2School of Medicine, Griffith University, Gold Coast, Qld, Australia
The prolonged use of bisphosphonates has been shown to cause a condition termed ‘bisphosphonate related osteonecrosis of the jaws’ (BRONJ). BRONJ is a disease entity which has only been described relatively recently, and its multi-factorial aetiology is yet to be fully elucidated. Therefore, the treatment of BRONJ lesions remains a challenge, and animal models are necessary to assist researchers in better understanding the disease. This has led to the recent publication of a number of studies utilising a variety of animal models of BRONJ. This review outlines the factors to be considered when selecting an animal model for BRONJ and discusses the current literature in this rapidly progressing field of research. It is important to consider the applicability of a given model to the clinical condition presenting in humans, and to this end, thorough characterisation of the clinical, histological, radiographic and systemic features is necessary. The development of a clinical lesion is an important consideration in terms of choosing a relevant model, and it appears clear that surgical manipulation, generally involving tooth extraction, is necessary for successful induction of the classic ‘clinical’ lesion of BRONJ. Oral Diseases (2013) 19, 747–754 Keywords: animal models; bisphosphonates; BRONJ; osteonecrosis
Introduction Animal models are an indispensable research tool for the study of diseases affecting humans and have played a vital role in the discovery, testing and validation of preventive and therapeutic drugs for human use. Vaccines and hormone replacements are examples of two major Correspondence: Saso Ivanovski, BDSc, BDentSt, MDSc(Perio), PhD, Professor of Periodontology, School of Dentistry and Oral Health, Gold Coast Campus Griffith University, Qld, 4222, Australia. Tel.: +61 7 5678 0741, Fax: +61 7 5678 0708, E-mail: s.ivanovski@griffith.edu.au Received 19 November 2012; revised 22 December 2012; accepted 23 December 2012
classes of drugs derived through successful animal trials that are subsequently being used as part of a daily routine to effectively minimise and manage disease morbidity in humans. Over the past two decades, bisphosphonates (BPs) have become the primary class of drugs prescribed for the management of bone pathologies associated with excessive bone resorption like osteoporosis, Paget’s disease, and primary and secondary bone malignancies (Francis, 1995; Drake et al, 2008; Papapoulos, 2008). Although BPs are effective treatment agents for these conditions, their use has also been associated with significant side effects such as bisphosphonate-related osteonecrosis of the jaws, viz BRONJ (Marx, 2003; Advisory Task Force on Bisphosphonate-Related Osteonecrosis of the Jaws, American Association of Oral and Maxillofacial Surgeons, 2007; Khosla et al, 2007; Ruggiero, 2011). Following the description of BRONJ as a distinct clinical entity attributable to BP use less than a decade ago, various factors have been identified that may contribute to its occurrence (Marx, 2003; Ruggiero et al, 2009; Otto et al, 2012). However, no one aetiological biological mechanism has been shown to be specifically associated with the pathogenic process involved in the establishment and progression of the BRONJ lesion. Thus, a current focus of BRONJ research is to determine plausible etiopathogenic factors and pathways, and this requires the use of animal models. Ideally, an animal model for BRONJ should mimic the clinical description of the lesion following the administration of a regimen of BPs that is commensurate with that used in humans. Specifically, it has to fulfil the criteria of clinical exposure of the jaw bone that does not heal after 8 weeks, with a concomitant history of BP use (especially intravenous nitrogen containing BPs) and in the absence of radiotherapy (Ruggiero et al, 2009). However, with the recent clinical description of ‘nonexposed BRONJ lesions’ in humans and a lack of subjective symptomatic data from animal models, accurate diagnosis of BRONJ now also requires histological demonstration of bone and soft tissue necrosis (Fedele et al, 2010).
Animal models for BRONJ
D Sharma et al
748
Challenges in developing an animal model for BRONJ There are many factors to be considered in assessing the suitability of an animal model to be used as a research tool in the study of a particular human disease. Ideally, the animal should be similar genetically in order to mimic the initiation and progression of the disease in humans. However, this is not always possible as the most phylogenetically similar animals to humans i.e. the great apes, chimpanzees and orang-utans are rarely employed due to the costs involved in maintaining these animals and ethical concerns with their use (Locke et al, 2011). The availability of established data regarding the anatomy, physiology and biomechanics of the animal model and its extrapolative feasibility to human conditions and responses, are also desirable whilst developing an animal model. Historically, rodents such as mice and rats have been the animal of choice as a model for most human diseases. They are primarily suitable because of their ease of care and handling, high reproductive capacity, completed genome mapping that further facilitates genetic engineering to suit the disease to be studied and above all, a genetic similarity with humans. Due to their size, rats are more suitable for oral and periodontal research as their jaw size is bigger and more suitable for manipulations like extractions and implant placement. However, the real challenge is to determine a dosage of BPs that will induce BRONJ similar to that seen in humans, given possible variations in the absorption and re-release of BPs in animals due to genetic differences, as well as differences in the remodelling pattern of their bone. Depending on the animal model, significant variation has also been reported in the duration of BP administration required for the induction of a BRONJ lesion. However, as may be expected with smaller animals, all of these reports suggest that much less time is required than that reported in humans. For instance, in humans, the average duration for BRONJ occurrence in the case of treatment with the less potent non-nitrogen containing bisphosphonates such as etidronate is around 55 months, whereas in the case of highly potent nitrogen containing bisphosphonates including zoledronate acid (ZA), treatment duration of around 20 months (Palaska et al, 2009) is required. Except for a few studies in the canine model, most of the animals studies reviewed here were evaluated for BRONJ lesions within weeks after the commencement of drug therapy (Allen and Burr, 2008; Allen et al, 2010). This is an advantage experimentally, in terms of the total time required to assess various aspects of the disease in question, although it makes direct comparison with the disease in humans more difficult. To minimise such variations, the selection of the appropriate animal model, taking into account exposure variables such as exact dosage, duration and co-medications, is of paramount importance to achieve comparable levels of BP at the desired site to induce a BRONJ lesion. BPs are usually prescribed as a longterm use drug, ranging from daily use for oral drugs like alendronate to monthly or bi-annually for the more potent intravenous drugs including pamidronate Oral Diseases
and zoledronic acid. These are major factors that need to be considered when comparing animal models with humans given animals are usually exposed to BP’s for a much shorter duration and at different doses and intervals.
BRONJ animal models Various animals have been used to develop and study BRONJ etiopathogenesis. Both rodents (mice, Rice rats, Wistar rats, and most predominantly Sprague-Dawley rats) (Table 1), as well as large animals (beagle dogs and minipigs) (Table 2), have been used as models for BRONJ in short- and long-term studies involving both oral and intravenous BPs. However, a major limitation in these animal studies is that subjective symptoms such as pain, a common feature of BRONJ, cannot be obtained and recorded accurately. The diagnosis of BRONJ therefore has to be arrived at by clinical evaluation and histological assessment.
Small animal (Rodent) models Rice rats One of the earliest studies that demonstrated the necrotic effects of BPs on the mandible employed a rice rat model with subcutaneous administration of clodronate (Gotcher and Jee, 1981) (Table 1). This study was not aimed at developing an animal model per se as BRONJ as a clinical entity did not exist at that time. Rather, this experiment was conducted to study the pathogenesis of periodontal disease and the effects of BPs (then called diphosphonates) on the progression of the disease process in a rice rat model of experimentally induced periodontal bone loss. The effect of three BP dosing regimens (0.1, 1.0, 10 mg kg 1 day 1 of clodronate) on various parameters including ‘bone amount’, histological quantification of the different cell types present in the lesion, vascular and necrotic tissue spaces and the proliferative activity of fibroblasts were evaluated over 6, 12 and 18 weeks. The results reported an increased amount of bone in all animals treated with higher BP doses (1 and 10 mg kg 1 day 1) and for longer treatment periods (12 and 18 weeks). Reduced vascular spaces due to increased fibrosis were one of the significant findings that support the current etio-pathogenic hypothesis implying reduced vascularity as a contributing factor in BRONJ development. Further, alveolar bone trabeculea devoid of typical bone cells were reported to be protruding into the epithelia and oral cavity, which suggests the presence of ‘devitalized bone’, a clinical feature of BRONJ. Tibial bone changes were evaluated and compared with alveolar bone mass changes, and the response was found to be similar, that is, an anti-resorptive effect, but quantitatively higher than in alveolar bone. Although this model has the confounding factor of involving ligature-induced periodontal bone loss, this study was one of the first to describe clinical and histological features of BRONJ as it is known today, without any oral manipulation in the form of tooth extraction (Gotcher and Jee, 1981).
Sample
76
92
60
30
NS
40
10
80
NS
Animal
Rice rats
SD rats female
SD rats female
SD rats
SD rats male
SD rats
Wistar rats female
Wistar rats male
c57bl/6j mice male
1
1
Dexamethazone (s.c) 1 mg kg 1 day 1 Zoledronic acid (s.c) 7.5 lgm kg 1 Zoledronic acid (s.c) 250 lgm kg 1 day Etidronate (s.c) 250 lgm kg 1 day
Zoledronic acid (i.v) 0.2 mg ml
Alendronate (oral) 200 lg kg 1 day 1 Dexamethazone (s.c) 1 mg kg 1 day 1
Zoledronic acid (i.p) 66 lg kg 1
Zoledronic acid (i.p) 0.1 mg kg 1 Pamidronate (i.p) 3 mg kg 1 Zoledronic acid (I.v inf) 20 lg kg
Clodronate (s.c) 0.1, 1.0, or 10.0 mg kg 1 day 1 Dexamethazone (s.c) 1 mg kg 1 Zoledronic acid (s.c) 7.5 lgm kg 1
Drug and dose
1
1
One dose at baseline and other after extraction at 3 weeks Thrice weekly; 3 weeks before and 12 weeks after ligature 14 days, after extraction Starting 2 days before extraction, 4 doses Once weekly, 2 weeks before and 3 weeks after extraction Daily for 7, 14, 21 days 1, 2 or 3 doses on every 7th day 7 days prior to, and 4 days after, extraction
Extraction of Left maxillary or mandibular molars
Daily for 7, 14, 21 days 1, 2 or 3 doses on every 7th Day Thrice weekly for 6 weeks or 8 weeks
Maxilla or Mandible Maxilla
Extraction of Right maxillary first molar
Maxilla
Extraction of Right Max first molar + 4 mm defect Extraction of Left maxillary or mandibular molar
Maxilla
Maxilla
Mandible, Anterior and Posterior; Femur Mandible
Mandible and Tibial metaphysis Maxilla or Mandible
Areas assessed
Extraction left maxillary 1st molar + Mucosal wound 2 mm
Extraction of mandibular right 1st molar and bone defect of 2 mm Ligature around crown of right first maxillary molar
None
None
Surgical Intervention
Daily for 6, 12, or 18 weeks
Frequency and Duration
BP and/or other drug used
Table 1 Small animal models of bisphosphonate related osteonecrosis of jaws (BRONJ)
No
No
Yes
Yes
No
Yes
No
Yes
No
‘Clinical’ lesion (BRONJ)
NS
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Histological necrosis
New bone formation and angiogenesis inhibited by ZA
Necrosis Score high in dexamethasone (DX) + ZA
Expansion of defect with central necrotic areas in test group
50% of ALN group had small sequestra, but only 30% of controls
ZA+ ligature induced larger areas of necrosis
ZA increased VEGF in serum of test animals
Reduced osteoblast and fibroblast numbers at the alveolar bone surface Necrosis was seen in 80% of maxilla and 60% of mandible sites Apoptosis seen in bone and marrow No effect on anterior mandible or femur
Other findings
(continued)
Kobayashi et al (2010)
Ali-Erdem et al (2011)
Biasotto et al (2010)
Abtahi et al (2012)
Aghaloo et al (2011)
Marino et al (2012)
Senel et al (2010)
Sonis et al (2009)
Gotcher and Jee (1981)
References
Animal models for BRONJ
D Sharma et al
749
Oral Diseases
Animal models for BRONJ
D Sharma et al
Oral Diseases
Bi et al (2010) 1
Zoledronic acid (i.p) 125 lg kg 1 Dexamethasone 5 mg kg Docetaxel 25 mg kg 1 NS C57BL6 mice
Drug and dose Sample
*started 3 weeks prior to tooth extraction, for six to twelve weeks. s.c, subcutaneous; I.p, Intaperitoneal; I.v, Intavenous; Inf, Infusion;, ZA, Zoledronic Acid; DEX, Dexamethazone; NS, not specified; SD, Sprague-Dawley.
ZA suppressed bone formation and caused necrosis of bone, along with inflammation and soft tissue disruption, in combination with Dex. Docetaxel can exacerbate the soft tissue defect Yes Yes Maxilla or Mandible Extraction of left maxillary or right mandibular first molar Twice weekly* Weekly* Weekly*
Frequency and Duration Animal
Table 1 (continued)
BP and/or other drug used
Surgical Intervention
Areas assessed
‘Clinical’ lesion (BRONJ)
Histological necrosis
Other findings
References
750
Mice Although mice are one of the most widely used laboratory animals, BRONJ studies in mice are infrequently reported. Kobayashi et al (2010) used male C57BL/6J mice, and either zoledronic acid (ZA, 250 lg kg 1 day 1) or etidronate (250 lg kg 1 day 1), given subcutaneously 7 days before the extraction of the right maxillary first molar, to develop a model of BRONJ. The BPs were continued for 4 days postextraction, and the mice were sacrificed on the fifth day. The authors reported that the typical feature of BRONJ, that is, the non-healing osseous defect, was not demonstrable in the test animals. They were however able to confirm an anti-angiogenic effect by quantifying reduced formation of CD31-positive blood vessels in the mice treated with ZA, which in turn affected bone formation. Etidronate however had no demonstrable negative effects on angiogenesis, healing or bone formation. However, Bi et al (2010), using the same mouse model, were able to demonstrate clinical BRONJ lesions when ZA was given in addition to dexamethasone (DX) docetaxel (a chemotherapy drug). They concluded that the co-medications exacerbated the clinical presentation of BRONJ with the combination of ZA, DX and docetaxel resulting in more necrotic bone and sequestra, reduced angiogenesis and bone remodelling, and bigger soft tissue defects at the extraction site. Wistar and Sprague-Dawley rats These animals are by far the most commonly used outbred albino rat species in oral and maxillofacial pathology research, due to their size that makes it easier to carry out dental procedures, such as extractions. A Wistar strain rat BRONJ model was developed by Biasotto et al (2010), who administered intravenous ZA (0.2 mg ml 1 once weekly) for 5 weeks following dental extraction and the creation of a 4 mm-diameter bone defect at the extraction site that was allowed to heal without sutures. Control rats had similar bone defects but without the ZA therapy. After 7 weeks healing, the test (ZA) rats showed a residual bone defect in the extraction site, whereas the control rats showed a normal healing process. Further, dissection of the ZA rat bone defect at 8 weeks showed minimal healing with an enlarged osseous defect surrounding central necrotic areas at the test site. Micro-CT scans revealed osteolytic areas around these bone defect. Microscopic examination confirmed the presence of mucosal ulceration and sequestration of bone with loss of osteocytes from the lacunae, all of which was suggestive of successful induction of BRONJ that was clinically and histologically similar to that seen in humans. However, one of the significant shortcomings of this study was that it did not evaluate the effect of BPs on the vascular and cellular components of the lesion. Furthermore, the model of ‘creating’ a bone defect, rather than only extracting a tooth, needs to be further evaluated as this is not generally representative of human treatment. Also, as discussed by the investigators, the BP dosage used was not commensurate with that administered clinically in the treatment of human diseases.
Animal models for BRONJ
Pautke et al (2012)
Huja et al (2011a,b)
Yes
ZA diminished the bone remodelling around healing implant 50% extraction sites showed impaired healing per animal; CT scan confirmed osteolytic areas; Histological necrosis in 29/30 sites in test animals Once a month, 4 Months Once weekly 6 weeks before and 10 weeks after extraction 1
1
10
Zoledronic acid (I.v Inf) 10 lgm kg Zoledronic acid (I.v Inf) 50 lgm kg 7
Beagle dogs Mini-pigs
Zoledronic acid (I.v Inf) 10 lgm kg 7 Beagle dogs
1
1
Zoledronic acid (I.v Inf) 6.7 lgm kg Alendronate (Oral) 0.2 mg kg 1 day 1 60 Beagle dogs female
s.c, subcutaneous; I.p, Intaperitoneal; I.v, Intavenous; Inf, Infusion; ZA, Zoledronic Acid; NS, not specified.
Yes
Maxilla and mandible Maxilla and Mandible
No
Yes ( 1%) Once a month, 4 Months
Monthly 3 or 6 Months Daily 3 or 6 Months
None
3rd premolar extraction and ipsilateral mini-implant Mini-implant placement Extraction of two premolars and one molar in both jaws on any one side
Mandible, maxilla, rib, and femur
No
No Right hemi-mandible Right ninth rib, Right tibia
NS
Huja et al (2011a,b)
Allen et al (2010)
Allen and Burr (2008)
Necrotic regions retained vasculature, that is, no evidence of avascular necrosis @ 3 months, ZA group had 95% lower intracortical BFR in mandible @ 6 months 99% BFR Rib showed similar effect Tibia was not affected Bone formation rate and turnover diminished in both extraction site and Implant site 25% and 33% necrosis in two groups NS No Mandible None 36 Beagle dogs female
Drug and dose
1
Alendronate (Oral) 0.2 or 1.0 mg kg
Sample Animal
day
1
3 years
Surgical Intervention Frequency and Duration
BP and/or other drug used
Table 2 Large animal models of bisphosphonate related osteonecrosis of jaws (BRONJ)
Areas assessed
‘Clinical’ lesion (BRONJ)
Histological necrosis
Other findings
References
D Sharma et al
The use of the Sprague-Dawley (SD) strain of rat in a BRONJ model mimicking human dosage and regimens was reported by Sonis et al (2009). These authors also included DX as a co-medication, as it is invariably used in the management of multiple myeloma patients. In this prospective controlled study, 92 rats were divided into eight test groups that received 1 mg kg 1 once daily dose of DX subcutaneously for either 1, 2 or 3 weeks followed by one, two or three doses of ZA at 7.5 lgm kg 1 subcutaneously on the last day of the scheduled DX regimen. The location of the oral lesion was also randomised to avoid the effects of anatomical and vascular variations by extracting all left mandibular molars in one half of the animals in each group and in the other half, all left maxillary molars were extracted. Rats were sacrificed at either 2 weeks or 4 weeks postextraction. Gross examination after sacrifice showed the presence of an open wound at the extraction site only in the rats with ZA + DX treatment, after both healing periods of 2 and 4 weeks. Histological features in the ZA + DX group showed ulcerative mucosa after 2 weeks, whereas only a few ZA-treated rats showed ulcerations after 4 weeks. An interesting observation was that the maxillary ulcerations were higher in frequency than mandibular lesions (80% vs 60%), when three cycles of ZA were administered along with DX. The effects of BP on bone vascularity were also studied and quantified, but no statistically significant difference between the groups was evident. PAS staining used to detect the presence of actinomyces, a group of microbes most commonly associated with BRONJ lesions, failed to confirm their presence. TUNEL staining confirmed the presence of apoptotic cells in the bone tissue, thus confirming necrosis of bone. Whilst the study was able to induce BRONJ, the drug dosage (too high for DX and too low for ZA) and the extraction of all molars instead of a single tooth extraction may be considered to be excessive when compared with other animal studies. A similar study reported by Ali-Erdem et al (2011) using a Wistar rat model and a ZA + DX drug regimen followed by extraction of left maxillary and mandibular molars showed that, although gross healing was similar in both control and ZA + DX groups, histological ‘necrosis scores’ were higher in ZA + DX-treated animals at the end of 28 days. Extractions were carried out in both arches, and the mandible showed slightly lower BRONJ incidence than the maxilla. Histological confirmation of reduced vascularisation was reported as significant at 2 and 4 weeks postextraction. In addition, Actinomyces bacteria commonly associated with BRONJ lesions were detected in 11 of 18 cases. Another group, who utilised a larger sample size (200 SD rats), was also able to induce clinical BRONJ lesions by administering pamidronate (3 mg kg 1 day 1) with DX (1 mg kg 1 day 1 for either 1, 2 or 3 weeks) in 35% of rats treated with the combination of drugs (L opezJornet et al, 2010). The pamidronate + DX group showed higher levels of inflammation and decreased vascularisation. In the control animals, pamidronate or DX alone failed to induce necrotic features clinically, and the inflammation was milder than in the drug combination group (Lopez-Jornet et al, 2010).
751
Oral Diseases
Animal models for BRONJ
D Sharma et al
752
To eliminate the possible contributing effect of co-medications and dental manipulation, Senel et al (2010) studied the effect of ZA (0.1 mg kg 1) and pamidronate (3 mg kg 1) injections three times a week for either 6 or 8 weeks in SD rats. No tooth extractions were performed, and the mandibles along with the femur were studied after sacrificing the animals 2 days after the drug therapy ended. Oral mucosal ulceration and bone exposure, a typical clinical feature of BRONJ, were conspicuously absent in all of the test groups although microscopically it was evident that inflammation along with necrosis of bone was present in 40% of the ZA group and 30% of the pamidronate group at 8 weeks. However, it is worth noting that even higher doses of ZA (0.1 mg kg 1), given more frequently than in other reported studies, failed to induce BRONJ lesions. Inflammation was the only noticeable effect, especially in the posterior mandible, which was postulated to be due to higher bone turnover in comparison with the anterior mandible. Another study that evaluated BRONJ induction in an SD rat model was reported by Marino et al (2012), who used two doses of ZA (20 lgm kg 1), one at baseline and the other postextraction, 3 weeks after initial ZA administration. A 2 mm defect was also created after the extraction, and the animals were sacrificed on either the day of extraction, or after 4 weeks or 8 weeks of healing. BRONJ lesions were demonstrable in 75% of the test (ZA) rats after 8 weeks healing. Histological examination was able to further confirm the presence of osteonecrosis in all test specimens at both time points. The effect of ZA on the circulating angiogenic cytokine, vascular endothelial growth factor (VEGF), was also evaluated in the serum. The levels of VEGF were reported to be increasing from baseline through the healing phase, with the highest levels in the test group at 8 weeks of healing. This was postulated to be due to a compensatory induction of an angiogenic response in the test group, in response to the delayed wound healing in these animals. One of the major limitations of this study was that no efforts were made to evaluate the effect of ZA on the vascularity of bone at the test sites. Perhaps the most recognised significant risk factor associated with BRONJ is surgical manipulation such as dental extractions or other procedures that expose alveolar bone to the oral cavity in patients on BP therapy. Recently, Abtahi et al (2012) incorporated this factor into their study by using a SD rat model where an extraction site with an additional mucosal wound anterior to the extraction was evaluated and compared with the non-treated contra-lateral side. All the rats underwent unilateral extractions, and test rats were started on 200 lg kg 1 of alendronate (ALN) for 14 days with or without DX (1 mg kg 1 once daily) for 4 days, starting 2 days prior to extraction. Lactate dehydrogenase histochemical staining was used to assess and compare the viability of the bone osteocytes at the extraction site to the non-extraction side. Gross clinical examination of the rats confirmed the presence of ulceration with exposed bone in all DX + ALN animals, in contrast to completely healed areas in control and ALN only treated animals. Histological examination showed the presence of large bone sequestra and ulcerated epithelium in
Oral Diseases
all DX + ALN-treated rats, compared with small sequestra in 50% of the ALN-treated rats and 30% of the control untreated rats, without any epithelial discontinuation in the two control groups. The extraction site sections stained for lactate dehydrogenase were not significantly different to those from the non-extraction side, apart from the area close to the necrotic part of the socket where necrotic osteocytes were reported. This study suggested that bone exposure is an essential factor for initiation of BRONJ. However, the inclusion of a soft tissue defect on the nonextraction side would have shown whether extraction is mandatory, or if a mere soft tissue defect could induce a BRONJ lesion. Another significant study investigating the effect of BPs on ligature-induced periodontal disease in SD rats was recently reported by Aghaloo et al (2011). ZA was administered three times per week (66 lg kg 1), commensurate to the drug dosage administered in cancer patients. After 3 weeks, periodontal disease was induced around the maxillary first molar using a wire ligature, and the rats were continued on ZA for 12 more weeks. Using micro-computed tomography, the authors were able to demonstrate bone loss associated with ligature-induced periodontal disease and that the bone loss was greater in the absence of ZA. Furthermore, radiographic features of BRONJ comparable with those found in human cases, including sequestration, periosteal bone deposition and alveolar bone expansion, were reported in the ZA-treated animals. Histological evaluation confirmed the presence of necrotic alveolar bone characterised by empty lacunae and the loss of overlying epithelium, whilst TUNEL assays confirmed that the percentage of apoptotic osteocytes was significantly higher in ligature sites of ZA-treated rats. Although this study used a higher drug dosage than others, periodontal disease was shown to significantly contribute to, as well as accentuate, the necrosis induced by BPs.
Large animal models Canines Canine models using beagles are considered appropriate for studying human skeletal disease processes, as these animals show intracortical mandibular bone remodelling and similar turnover rates to humans (Table 2). Long-term studies utilising the beagle dog BRONJ model have been carried out by Allen and Burr (2008) and Allen et al (2010). One of their earlier papers studied the effect of alendronate (either 0.2 mg kg 1 day 1 or 1 mg kg 1 day 1) on the beagle dog mandible over a period of 3 years. Although no spontaneous bone exposure was reported, histological necrosis was identifiable using basic fuchsin staining in 25% of the low dose group and 33% of dogs in the higher dose group (Allen and Burr, 2008). Confocal microscopy of the necrotic areas confirmed the absence of osteocytes and canalicular networks, although vasculature was evident. Intracortical bone formation, especially alveolar bone, was found to be significantly suppressed in response to both doses of alendronate, as evidenced by fewer osteons and a decreased rate of mineral apposition. The findings of this study did not appear to support the avasular necrosis hypothesis of BRONJ.
Animal models for BRONJ
D Sharma et al
When the same canine model was used to compare monthly intravenous ZA with daily oral alendronate, for either 3 or 6 months, no evidence of histological bone necrosis was demonstrable in any of the tested animals (Allen et al, 2010). The only changes were confined to a suppressed intracortical bone formation rate, as confirmed by fewer active osteons, and a lower mineral apposition rate at 3 months in the dogs treated with ZA. These changes were much more prominent after 6 months therapy in the ZA group. The extent of suppression of bone formation and mineral apposition rates was greater in ZA compared with alendronate, with both being greater than the vehicle. Rib and tibial bone were also evaluated for the effects of the test drugs and showed similar features of suppression. Neither of the above studies reported the induction of any ‘clinical lesions’ of BRONJ in the absence of surgical manipulation. Two recent papers utilising the dog model did use surgical manipulation along with ZA (0.1 mg kg 1 month 1) for 4 months, but failed to show signs of clinical necrosis at the extraction sites (Huja et al, 2011a,b). Osteocyte viability was assessed using a lactate dehydrogenase (LDH) assay, and bone activity was measured with standard static and dynamic bone histomorphometry. Histological examination revealed significantly larger necrotic lesions in the test group (more so in the maxilla) compared with controls. ZA reduced bone remodelling regardless of surgical manipulation in the maxilla and mandible (both alveolar and basal bone) as well as the rib, with minimal reduction noted in the femur. As this was a short-term study, longterm evaluation with this model is essential. Mini-pigs A mini-pig model was recently developed by Pautke et al (2012), wherein mini-pigs underwent bilateral extraction after 6 weekly infusions of 2 mg ZA. The infusions were subsequently continued for a further 10 weeks postextraction. All test (ZA) animals had unhealed extraction sockets compared with normal healing in control (untreated) animals. Computed tomographic confirmation of osteolytic lesions was also reported, and histological examination revealed osteonecrotic features such as empty lacunae, sequestration along with an absence of osteoblasts, and lack of bone remodelling in extraction sites in the ZA-treated minipigs. This study was able to demonstrate BRONJ induction with the combination of ZA and extraction, further confirming the fact that clinical lesions of BRONJ were inducible with surgical manipulation in BP-treated animals.
Conclusions There are significant challenges involved in simulating all of the conditions that are implicated in human BRONJ, including genetic and environmental factors, co-existing systemic diseases, concomitant medications along with the precise dosage for animal models. Further, another significant challenge in developing animal models for BRONJ is the recent modification of the human clinical diagnostic criteria to accommodate the ‘spontaneous’ clinical variant of BRONJ. The term ‘spontaneous’ is used to describe
BRONJ that develops without any history of oral surgical manipulation and does not manifest a typical clinical ‘bone exposure lesion’, but may present with other findings like non-periodontal disease-associated tooth mobility, fistula or sinus tract not associated with pulpal disease, bone or gingival swelling, pathological fracture and pain in relation to the maxillary sinus or tooth pain not attributable to any other causes (Fedele et al, 2010; Patel et al, 2012). Spontaneous BRONJ lesions may be studied in animal models wherein radiographic findings such as widening of the periodontal ligament space, alveolar bone loss unrelated to periodontal disease and/or lack of remodelled bone in the postextraction socket can be useful criteria in identifying such lesions. To date, the animal models developed for the purpose of studying the complex interdependent processes involved in the pathogenesis of BRONJ have had variable success in replicating the conditions that result in the development of clinical lesions that are representative of those found in humans. The administration of bisphosphonates alone does not appear to be sufficient, and it appears clear that surgical manipulation, generally involving tooth extraction, is necessary for successful induction of the classic ‘clinical’ lesion of BRONJ. However, it is unclear whether extraction of a tooth is an absolute prerequisite for BRONJ induction or if a soft tissue lesion alone is sufficient. It is also important to note that, even in the presence of a clinical lesion, the type of bisphosphonates used, as well as the doses and regimes administered, has a significant influence on the local and systemic characteristics of the induced condition. Therefore, thorough characterisation of the histological, radiological and systemic features of the condition induced in a given animal model is necessary to understand its applicability to the human condition. Whilst the inferences that are derived from these animal models are often difficult to apply directly to humans, a thorough understanding of the effects of disease modulating factor (s) through the use of animal models will play a significant role in terms of developing strategies to limit the debilitating effects of BRONJ, thereby improving preventive and treatment strategies in humans.
753
Author contributions Dr Sharma carried out the literature review and prepared the draft manuscript. Drs Hamlet, Petcu and Ivanovski contributed equally in the conseption of the review, as well as revision and preparation of the final manuscript.
References Abtahi J, Agholme F, Sandberg O, Aspenberg P (2012). Bisphosphonate-induced osteonecrosis of the jaw in a rat model arises first after the bone has become exposed. No primary necrosis in unexposed bone. J Oral Pathol Med 41: 494–499. Advisory Task Force on Bisphosphonate-Related Osteonecrosis of the Jaws, American Association of Oral and Maxillofacial Surgeons (2007). American association of oral and maxillofacial surgeons position paper on bisphosphonate-related osteonecrosis of the jaws. J Oral Maxillofac Surg 65: 369–376. Oral Diseases
Animal models for BRONJ
D Sharma et al
754
Aghaloo TL, Kang B, Sung EC et al (2011). Periodontal disease and bisphosphonates induce osteonecrosis of the jaws in the rat. J Bone Miner Res 26: 1871–1882. Ali-Erdem M, Burak-Cankaya A, Cemil-Isler S et al (2011). Extraction socket healing in rats treated with bisphosphonate: animal model for bisphosphonate related osteonecrosis of jaws in multiple myeloma patients. Med Oral Patol Oral Cir Bucal 16: e879–883. Allen MR, Burr DB (2008). Mandible matrix necrosis in beagle dogs after 3 years of daily oral bisphosphonate treatment. J Oral Maxillofac Surg 66: 987–994. Allen MR, Kubek DJ, Burr DB (2010). Cancer treatment dosing regimens of zoledronic acid result in near-complete suppression of mandible intracortical bone remodelling in beagle dogs. J Bone Miner Res 25: 98–105. Bi Y, Gao Y, Ehirchiou D et al (2010). Bisphosphonates cause osteonecrosis of the jaw-like disease in mice. Am J Pathol 177: 280–290. Biasotto M, Chiandussi S, Zacchigna S et al (2010). A novel animal model to study non-spontaneous bisphosphonates osteonecrosis of jaw. J Oral Pathol Med 39: 390–396. Drake MT, Clarke BL, Khosla S (2008). Bisphosphonates: mechanism of action and role in clinical practice. Mayo Clin Proc 83: 1032–1045. Fedele S, Porter SR, D’Aiuto F et al (2010). Non-exposed variant of bisphosphonate-associated osteonecrosis of the jaw: a case series. Am J Med 123: 1060–1064. Francis MR (1995). Oral bisphosphonates in the treatment of osteoporosis: a review. Curr Ther Res 56: 831–851. Gotcher JE, Jee WS (1981). The progress of the periodontal syndrome in the rice cat II. The effects of a diphosphonate on the periodontium. J Periodontal Res 16: 441–455. Huja SS, Mason A, Fenell CE et al (2011a). Effects of shortterm zoledronic acid treatment on bone remodelling and healing at surgical sites in the maxilla and mandible of aged dogs. J Oral Maxillofac Surg 69: 418–427. Huja SS, Kaya B, Mo X, D’Atri AM, Fernandez SA (2011b). Effect of zoledronic acid on bone healing subsequent to miniimplant insertion. Angle Orthod 81: 363–369. Khosla S, Burr D, Cauley J et al (2007). Bisphosphonate-associated osteonecrosis of the jaw: report of a task force of the American society for bone and mineral research. J Bone Miner Res 22: 1479–1491. Kobayashi Y, Hiraga T, Ueda A et al (2010). Zoledronic acid delays wound healing of the tooth extraction socket, inhibitsoral epithelial cell migration, and promotes proliferation and adhesion to hydroxyapatite of oral bacteria, without causing
Oral Diseases
osteonecrosis of the jaw, in mice. J Bone Miner Metab 28: 165–175. Locke DP, Hillier LW, Warren WC et al (2011). Comparative and demographic analysis of orang-utan genomes. Nature 469: 529–533. L opez-Jornet P, Camacho-Alonso F, Molina-Mi~ nano F, G omezGarc õa F, Vicente-Ortega V (2010). An experimental study of bisphosphonate-induced jaws osteonecrosis in Sprague-Dawley rats. J Oral Pathol Med 39: 697–702. Marino KL, Zakhary I, Abdelsayed RA et al (2012). Development of a rat model of bisphosphonate-related osteonecrosis of the jaw (BRONJ). J Oral Implantol 38: 511–518. Marx RE (2003). Pamidronate (Aredia) and zoledronate (Zometa) induced avascular necrosis of the jaws: a growing epidemic. J Oral Maxillofac Surg 61: 1115–1117. Otto S, Schreyer C, Hafner S et al (2012). Bisphosphonaterelated osteonecrosis of the jaws – Characteristics, risk factors, clinical features, localization and impact on oncological treatment. J Craniomaxillofac Surg 40: 303–309. Palaska PK, Cartsos V, Zavras AI (2009). Bisphosphonates and time to osteonecrosis development. Oncologist 14: 1154– 1166. Papapoulos SE (2008). Bisphosphonates: how do they work? Best Pract Res Clin Endocrinol Metab 22: 831–847. Patel S, Choyee S, Uyanne J et al (2012). Non-exposed bisphosphonate-related osteonecrosis of the jaw: a critical assessment of current definition, staging, and treatment guidelines. Oral Dis 18: 625–632. Pautke C, Kreutzer K, Weitz J et al (2012). Bisphosphonate related osteonecrosis of the jaw: a minipig large animal model. Bone 51: 592–599. Ruggiero SL (2011). Bisphosphonate-related osteonecrosis of the jaw: an overview. Ann N Y Acad Sci 1218: 38–46. Ruggiero SL, Dodson TB, Assael LA, Landesberg R, Marx RE, Mehrotra B, American Association of Oral and Maxillofacial Surgeons (2009). American association of oral and maxillofacial surgeons position paper on bisphosphonate-related osteonecrosis of the jaws-2009 update. J Oral Maxillofac Surg 67: 2–12. Senel FC, KadiogluDuman M, Muci E et al (2010). Jaw bone changes in rats after treatment with zoledronate and pamidronate. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 109: 385–391. Sonis ST, Watkins BA, Lyng GD, Lerman MA, Anderson KC (2009). Bony changes in the jaws of rats treated with zoledronic acid and dexamethasone before dental extractions mimic bisphosphonate-related osteonecrosis in cancer patients. Oral Oncol 45: 164–172.
