Methodological Consideration of Various Intraosseous and Heterotopic Bone Grafts Implantation in Animal Models

Bone fracture healing is a complex process including inflammation, repair and remodeling. Bone grafts or substitute are widely accepted to treat impaired healing. Newly developed bone substitutes must undergo in vitro and in vivo testing before clinical application. All kinds of intraosseous and heterotopic implant models in small and/ or large animals are used for different bone substitutes. Orthotopic implant at models is used for evaluation of graft materials. Based on the statistic of literatures, we find small animals should be used first before large animal as osseous defect models. Rabbit and rat are the commonly chosen animals, while femur and calvaria are the most implanted anatomic sites. Critical size defect models are useful as bone defect model, but vary considerably between animals. Typical heterotopic ossification after implantation of bone substitute is found almost in all species of animals. No bone formation is found after subcutaneous implantation of bone granule grafts in small animal, but in large animal. In contrast, bone block grafts show a distinguished result of bone formation in small animals as well as in large animals. This article reviews currently animal bone defect models and anatomic implant site for bone graft, gives a recommendation for the future research. Journal of Tissue Science & Engineering J o u r n a l o f T iss ue S cience &ngine e r i n g

Following the development of nanotechnology, some new bioactive nanohydroxylapatite are widely applied in bone tissue engineering, whose composition and crystal structure is very close to natural bone [19]. For example, a fully synthetic nanocrystalline hydroxyapatite embedded in a matrix of silica gel (NanoBone ® ) has showed new bone formation after five weeks while demonstrating degradation of materials [20]. Currently, there is no ideal material bone graft that can provide all requirements of bone defect in different defect site. Nevertheless, they provide certain advantages, such as good store, ready availability and the prevention of a further operation for patient. Furthermore, all kinds of bone substitutes have also been applied as cellular scaffolding system [21]. Cell-based bone graft substitute have shown significant osteogenic potential in animal segmental bone defect [22].

Experimental Animal Choices for Evaluation of Bone Substitutes
New developed bone replacement material needs strictly test in vitro and in vivo due to cytotoxicity and physical and chemical characterise before clinical applications for human [23]. Before beginning animal experiment, biomaterial should be at first tested by cell culture in vitro for cytotoxicity and cytocompatibility. Basic demand of a bone substitute material for in vitro cell culture is that the bone graft supports cell attachment, proliferation and migration [24]. A variety of experimental animal species including mouse [25,26], rat [27,28], rabbit [29,30], dog [31], pig [32] and goat [33] have been used to test bone grafts. Although most animal models are in quadrupeds and may not model the load bearing conditions evident with bipedal humans. But bone structure, composition and biology of large animal (i.e., dog, sheep, pig and goat) are very similar to that of human bone, which demonstrated a very similar bone healing process compared with human bone remodelling [34]. On one hand, the use of bone defects in large animal models is highly recommended to clearly evaluate the capacity of the tissue-engineered bone substitute for its final clinical application. On the other hand, using large animal models has two major limitations, for example, high costs and the care for the respective animals. Therefore, 38% of the studies in bonehealing research already preferred the rat as the experimental animal, following the next rabbit 19%, mouse 15%, sheep 11%, dog 9%, goat 4% and other 4% [35]. That means above 70% experiment animal species are small animal (rat, rabbit and mouse). Therefore, Le Guehennec and co-workers suggested that small animals should be a prerequisite before preclinical implantation of bone substitutes in large animals [36]. Small animal approach using rats or mouse allows easily studying a reasonable number of animals at different time points. In a dorsal skinfold chamber model of mouse, the inflammatory and angiogenic host tissue response to biomaterial can be on-line given for evaluation of biocompatibility and vascularisation in vivo [37]. Selection of animal species should be decided by purpose of research, clear questions and experimental conditions [38]. Otherwise, due to various anatomic, biochemical, and gene expression of experiment animals, the same animal model used in different species may show conflicting results in bone healing studies [39].

Intraosseous Critical/Noncritical Defect and Heterotopic Implant Model
Various animal models and implantation sites are applied to test bone substitutes by means of different evaluations of parameters [40]. There is the fundamental distinction between intraosseous and heterotopic models for bone substitute materials. An intraosseous model allows analysis of osteoactivity and biodegradation of implanted materials in a bone environment [41]. The application of heterotopic implant model is gold standard for assessment of the biocompatibility as well as the osteogenic and osteoinductive potential of bone graft materials [42].
Using intraosseous model, bone graft materials are usually implanted in tibia [43], femur [44], ulna [45] and calvaria defect [46]. According to the size of the defect, in which the bone substitute will be implanted, they can be categorized as either critical or noncritical defects. A critical size defect is a segmental or drill whole defect with load or non-load mechanical stress, which does not heal spontaneously during the lifetime without pathological changes [47]. Frequently, segmental critical/noncritical size defect was chosen in animal models for testing functionality of bone substitute materials. Such defects are usually used to evaluate whether the bone graft material is in a position to bridge the defect [48]. The drill hole critical/noncritical defect is usually used to study a variety of bone defects and potential therapies designed to repair these bone defect, which is applied to test filling materials for bone defect [49].
By means of heterotopic implant model, the toxicity and carcinogenic effect of biomaterials could be also evaluated by direct contact with the subcutaneous and muscular tissue [50]. Cellularity and vascularity of the recipient bed are very important parameters for the osteoinductive capacity in the ectopic bone formation [51]. Subcutaneous and muscular implantation usually allows for an evaluation of the biocompatibility, angiogenesis, degradation and osteoinduction of the tested materials [52]. Using subcutaneous implant model in rat, dynamical changes are investigated after implantation of nanostructured hydroxyapatite until 12 days [53].
Lots of studies demonstrated that adipose tissue is a good source of adult stem cells with a multilineage differentiation potential, for example, bone, cartilage, skeletal muscle and other cell types [54]. Comparing to muscle tissue, adipose tissue is well-vascularized with a capillary density [55]. Hartman et al. reported that no significant difference existed between intramuscular and subcutaneous recipient sites after bone substitute implantation [56]. However, this method ultimately does not analyse the functionality of a bone substitute material, because subcutaneous tissue has different physiological structure and molecular biological reaction to implanted material compared to osseous environment [57]. Furthermore, new bone formation in heterotopic tissue does not provide information on the influence of biomechanical stress and dynamic load as intraosseus implantation [58].
In this review, we examined the literature relating to intraosseous and heterotopic model used large and small animal in the evaluation of bone substitute.

Bone Defect Model and Different Anatomical Sites for Implantation in Animals
Different species of experimental animals and defect models are applied in bone tissue engineering, but there have been little criteria for choice of animal and implantation site. Following rapid development of materials science, more and more new bone grafts are tested by means of experimental animals.
We have summarised literatures of the MEDLINE database during 1 January 2006 through 30 November 2015 for studies published about bone substitutes using bone defect models in different animals. In all experimental animals for bone defect models, 64% of all animals used were small animals. The small animals commonly chosen for bone defect models are rat (29%) and rabbit (29%). Mice are only used in about 6% of all cases, because of their difficult management ( Figure 1). Among big animals, sheep, dog and pig dominate at almost the same ratio of 10% to 13% (Figure 1). Bone grafts are made into different shapes in order to apply for different defect size and shape. Therefore, there is no established standard for anatomical implantation sites. However, independent from the animal employed in the experiments, long bones (49%) such as tibia (13%), radius (12%) and femur (24%) are commonly chosen as implantation region. Viateav states that the calvaria mimic the clinical bone defect environment best because of poor blood supply and presence of little bone marrow. Furthermore, calvarial bone is formed by intramembranous ossification rather than endochondral ossification [42]. Therefore, the calvarial defect model rate is as high as 25% of all defects used. Interestingly, bone grafts and implants currently applied in craniofacial surgery are increasing in number. In coherence with this tendency, mandibular defect model are found in literature in 14% of all cases. Other anatomic sites, such as the iliac, rib, ulna and spinal region, are used in about 12% of all cases ( Figure 2).

Critical Size Defects of Animal Interaosseous Model in Different Anatomical Sites
Multiple critical size defects of animal models have been performed to mimic bone defect environment for evaluation of bone grafts. In the following table, critical size defect in different anatomic sites of animals are summarized (Table 1).
A critical defect size is associated with the size of the animal and anatomical site of bone. Reproducible critical size defect model is very important for study design. However, numerous studies in vivo demonstrated that sizes of critical defects in the same species vary strongly from study to study [59]. In addition, researchers must consider that many systemic factors such as age, sex, endocrine and pharmacological status of animal strain could affect results [60].

Heterotopic Model in Small and Large Animal Model
There are no standard criteria for intramusclular and subcutaneous implantations. Abdomen, back, limb and neck are popular regions for implant. In addition, the variety in different bone substitutes makes comparison difficult as each bone graft has specific weight, density and volume. The following table provides an overview of some publications about bone grafts implanted either subcutaneous or intramuscular ( Table 2).    Interestingly, our previous work showed granule grafts did not induce bone formation in small animals, but block grafts induced bone formation in small animals. Yuan et al confirmed that material-induced bone formation depends on the animal species and implanted material [61]. Likewise, various bone graft combining growth factor or stem cells shows different osteogenic properties in various heterotopic models [62][63][64][65].

Mechanics of heterotopic ossification is not completely clarified.
Kan et al demonstrated that morphogenetic protein (BMP) plays a crucial role in this case [66]. Heterotopic implantation induces hyperactivities of BMP receptors. Osteoinductive properties of BMP are reported first time by Urist et al [67,68]. Now recombinant BMPs (rBMPs) have been widely applied for treatment of spinal fusions, nonunion fractures, craniomaxillofacial and periodontal bone defects, and bone/tooth implant augmentation [69], as they are readily available.

Methodical analyse of bone graft after implantatation in animal
Bone grafts are made of a variety of components, structures, shapes and sizes; there are no established standard methods for analysing bone substitute in vivo. However, most often used methods are radiological evalution using x-rays, mikro CT, or high resolution CT as well as histochemical or immunohistochemical analysis. The most often used methods are visualized in Table 3.

Discussion
Ideal bone substitutes should support osteogenesis, osteointegration, osteoconduction and osteoinduction and contain or propagate the ingrowth of osteoblast or osteoprogenitor stem cells. However, only autologous bone graft has, as mentioned above, these properties. During the past 30 years different bone grafts have been developed to provide structural stability in order to overcome the clinical drawbacks associated with bone defects.
Currently, all kinds of bone grafts including natural bovine-derived hydroxyapatite and hydroxyapatite or silicon based bone substitute are widely applied in clinic. As synthetic material mimics' composition and structure of extracellular bone matrix and can be customized for specific application without the limit of amount supply, this is especially interesting for clinical application in orthopedic surgery [70]. Using 3D based imaging of defect positions, biomodels can be produced ahead of operation in reconstructive surgery with availability of different size and shape, thereby reducing operation time and patient trauma extremely and increasing the patient's safety [70][71][72][73][74][75][76][77][78][79][80][81][82][83][84][85][86][87][88]. All kinds of biomaterials combining BMP have shown successful effects in reconstructing long bones, spines and the facial skeleton defects .
Rational experimental design and careful selection of the animal model play a very important role in researching bone tissue engineering. No experiment model fulfils complete evaluation of new developed bone graft materials. A newly developed bone biomaterial should first be optimized in vitro [73,[93][94][95]104]. After that, various parameters are further evaluated through small animal models [106][107][108]. In the next step, large animal models are used to test the practicability functionality of graft materials before clinical application [86,105,[109][110][111][112][113][114][115]. For each specific question, a suitable animal model is needed. Each animal model has its own advantages and disadvantages. Intraosseous defect models are necessary to assess the possibility of clinical application for orthopaedic and craniofacial surgery. To evaluate biocompatibility, biodegradation and vascularisation, subcutaneous or intramuscular implantation models in mouse or rat are recommended [22,42,69,46,73,82,[84][85][86]88,90,93,94,96,[100][101][102][103]. But these models may not be suitable to answer questions of osteoinduction of granular bone grafts. A possible explanation for the difference between large and small animal in subcutaneous implantation of granule bone graft may be the increased micromovement in small animals [73,82,[84][85][86]88,90,93,94,96,100]. Small animals show constant and comparatively large movements of their body after implantation of granulated bone grafts compared to large animals [104,105]. Subcutaneous tissue in large animal provides relatively lesser movement of the subcutaneous tissue environment for implanted bone granules, which may be favourable for ectopic new bone formation. Another reason might be the increased stability of block scaffolds themselves compared to granules. Maybe granules tend to move easily with the environment, whereas block grafts remain stable and allow only micromovement on their surface. This might explain the increased bone formation on the surface around block scaffolds [74,87]. The micromovements might be too macro for osteocytes or stem cells to produce bone extracellular matrix in granules.
In conclusion, it is important for any animal study that many factors should be taken into account, when selecting the suitable species, the desire defect model and methods for evaluation.