Here we present a surgical protocol in rabbits with the aim to assess bone substitution materials in terms of bone regeneration capacities. By using PEEK cylinders fixed onto rabbit skulls, osteoconduction, osteoinduction, osteogenesis and vasculogenesis induced by the materials may be evaluated either on live or euthanized animals.
The basic principle of the rabbit calvarial model is to grow new bone tissue vertically on top of the cortical part of the skull. This model allows assessment of bone substitution materials for oral and craniofacial bone regeneration in terms of bone growth and neovascularization support. Once animals are anesthetized and ventilated (endotracheal intubation), four cylinders made of polyether ether ketone (PEEK) are screwed onto the skull, on both sides of the median and coronal sutures. Five intramedullary holes are drilled within the bone area delimited by each cylinder, allowing influx of bone marrow cells. The material samples are placed into the cylinders which are then closed. Finally, the surgical site is sutured, and animals are awaken. Bone growth may be assessed on live animals by using microtomography. Once animals are euthanized, bone growth and neovascularization may be evaluated by using microtomography, immune-histology and immunofluorescence. As the evaluation of a material requires maximum standardization and calibration, the calvarial model appears ideal. Access is very easy, calibration and standardization are facilitated by the use of defined cylinders and four samples may be assessed simultaneously. Furthermore, live tomography may be used and ultimately a large decrease in animals to be euthanized may be anticipated.
The calvarial model of bone augmentation was developed in the 90’s with the aim to optimize the concept of guided bone regeneration (GBR) in the oral and craniofacial surgical domain. The basic principle of this model is to grow new bone tissue vertically on top of the cortical part of the skull. To do so, a reactor (e.g., titanium -dome, -cylinder or -cage) is fixed onto the skull to protect the bone regeneration conducted by a graft (e.g., hydrogel, bone substitute, etc.). With the aid of this model, titanium or ceramic cages1,2,3,4,5,6, GBR membranes7,8,9,10, osteogenic factors11,12,13,14,15,16,17, new bone substitutes12,16,17,18,19,20,21,22,23,24,25,26,27,28,29 or the mechanism of neovascularization during the bone regeneration process30 were assessed.
From a translational point of view, the calvarial model represents a one-wall defect that can be compared to a class IV defect in the jaw31. The aim is to grow new bone above a cortical area, without any lateral support from endogenous bone walls. The model is thus extremely stringent and assesses the real potential of vertical osteoconduction over the cortical part of the bone. If the model described herein is primarily dedicated to the assessment of osteoconduction in bone substitutes, osteogenesis and/or osteoinduction may be also assessed, as well as vasculogenesis1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30.
Essentially for ethical, practical and economic reasons, the calvarial model was developed in the rabbit in which the bone metabolism and structure are quite relevant when compared to human32. Of the 30 references cited above, 80% used the rabbit calvarial model1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,17,22,23,26,27,28,29,30,33, thus demonstrating the relevance of this animal model. In 2008, the Busenlechner group transferred the calvarial model to the pig, to allow the comparison of eight bone substitutes simultaneously20 (as compared to two bone substitutes with the rabbit). On the other hand, our group transferred the rabbit calvarial model to sheep. In brief, titanium domes were placed on sheep skulls to characterize the osteoconduction of a new 3D-printed bone substitute. These studies allowed us to develop and master the calvarial model and its analysis16,21.
The last three studies cited16,20,21, together with several other investigations12,17,18,19,22,23,24,26,27,28,29, confirmed the great potential of the calvarial model as a screening and characterization model. However, even though the results obtained were quite satisfactory, they also pointed out some limitations: (1) The use of titanium domes, which prevented X-ray diffusion and in turn live micro-CT use. These could not be removed before histological processing, forcing the researchers to embed the samples in poly(methyl methacrylate) resin (PMMA). The resulting analyses were therefore largely limited to topography. (2) High financial costs especially because of the cost of the animals, and costs related to the logistics, maintenance and the surgery of the animals. (3) Difficulties to obtain ethical approvals for large animals.
A recent study by Polo, et al.26 largely improved the model on the rabbit. Titanium domes were replaced by closable cylinders that could be filled with a constant volume of material. Four of these cylinders were placed on rabbit skulls. At completion, the cylinders could be removed so that biopsies were metal-free, introducing much more flexibility concerning sample processing. The rabbit calvarial model became attractive for simultaneous testing with lower costs, easy animal handling and facilitation of sample processing. Taking advantage of these recent developments, we have further improved the model by replacing titanium with PEEK to produce cylinders, thereby allowing X-ray diffusion and the use of microtomography on live animals.
In this article, we will describe the anesthesia and surgery processes and show examples of outputs that may be obtained using this protocol, i.e., (immuno-) histology, histomorphometry, live and ex vivo microtomography to evaluate the mechanisms of bone regeneration and quantify the new bone synthesis supported by bone substitute materials.
In line with Swiss legal requirements, the protocol was approved by an academic committee and supervised by the cantonal and federal veterinary agencies (authorizations n° GE/165/16 and GE/100/18).
1. Specific devices and animals
2. Surgery
3. Post-surgical treatment
The model described herein is dedicated to the assessment of osteoconduction in bone substitutes. Osteogenesis and-or osteoinduction of bone substitutes either (pre-)cellularized or loaded with bioactive molecules may be also assessed, as well as vasculogenesis1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30. A kinetic study may be used, from 3 days up to 3 months after surgery depending on the mechanisms and outputs to be analyzed. A classical timeline allowing descriptions at early and mid-time is: 2, 4, 6, 8 and 12 weeks. Note that a minimum of 6 samples per time point is mandatory to obtain significant results. Each sample to be tested needs to be placed at least one time in each position onto the skull per time point (random allocation). Finally, sham samples (e.g., cylinders filled with coagulated blood) have to be included in the protocol34.
Once the surgery is completed, bone growth may be monitored at different time points by using bone tomography on live animals. An example is shown in Figure 4A,B. Additional analysis requires animals to be sacrificed (lethal intravenous injection of 150 mg/kg pentobarbital (100 mg/mL). After euthanasia, samples are sectioned and cylinders are carefully removed (Figure 5). Biopsies are fixed with a solution of phosphate-buffered saline and 4% formaldehyde. Bone growth may then be assessed by using microtomography (Figure 4 C,D). Samples may also be processed for (immune-) histological staining. Histomorphometric analysis and specific stainings are then possible to complete the analysis more specifically (Figure 6).
Figure 1: Specifications of PEEK cylinders. Two holes (0.8 mm in diameter) were drilled on the lateral stabilizing tabs for screwing. The positions of the 5 intramedullary holes (0.8 mm in diameter) to be drilled on the skull within the area delimited by the cylinder are marked using red circles. Please click here to view a larger version of this figure.
Figure 2: Representative image of the rabbit skull and placement of the cylinders. Pictures showing the median and coronal sutures on the rabbit skull delineating the left-right parietal and frontal bones (A,B). Placement of the cylinders on both side of the sutures (C). Scale bars = 5 mm. Please click here to view a larger version of this figure.
Figure 3: Representative image of cylinders fixed, filled and caped. Picture showing four cylinders fixed onto the skull of a rabbit with titanium screws. Within the area delimited by each cylinder, 5 intramedullary holes (0.8 mm in diameter, ~1 mm in depth) were drilled under irrigation with a round bur to allow bone cell migration. Cylinders were filled with different bone substitute samples (calibrated volumes) before capping (one closed cylinder is shown only). Scale bar = 5 mm. Please click here to view a larger version of this figure.
Figure 4: Representative images from microtomographic (micro-CT) analysis. With the final goal to assess bone growth conducted by bone substitutes, 4 cylinders were fixed onto a rabbit skull with titanium screws and filled with bone substitute materials. (A) Live imaging: two-dimensional transverse scan (14 min, 99 kV/88 µA with a resolution of 20 μm) of a cylinder at 12 weeks. (B) three-dimensional (3D) reconstruction from live micro-CT analysis at 4 weeks (red circles: bone substitutes in cylinders; red arrow: control in which the cylinder is filled with coagulated blood). (C,D) After euthanasia (12 weeks), cylinders were removed before fixation and micro-CT analysis. (C) 2D transverse scan (57 min, 99 kV/88 µA with a resolution of 10 μm) of a cylinder and 3D-reconstruction of the total new bone in the cylinder (D). Bone substitute particles (red), new bone (green) and bone bed (yellow) are shown. Scale bars = 2 mm Please click here to view a larger version of this figure.
Figure 5: Representative images of a biopsy at 4 weeks. After euthanasia (4 weeks), samples were block-sectioned and cylinders were removed before fixation in 4% formalin, micro-CT analysis and histological processing. Scale bar = 5mm. Please click here to view a larger version of this figure.
Figure 6: Representative pictures of (immune-)histological sections. With the final goal to assess bone growth and neovascularization conducted by bone substitutes, 4 cylinders were fixed onto a rabbit skull with titanium screws and filled with bone substitutes. After euthanasia (12 weeks), cylinders were removed before fixation and histological processing. (A) Masson-Goldner staining (50x): bone substitute appears as mauve particles surrounded by new bone in green. (B) The slices were scanned and processed for digital extraction of bone substitute material so that the new bone (red) could be quantified easily. (C) Immunostaining of CD31 (arrows), a typical marker of endothelial cells and the neovascularization process. (D) Immunofluorescent staining (green) of a highly neovascularized zone in which some new capillaries highly express CD31 (arrow). Please click here to view a larger version of this figure.
The model described herein is simple and should be developed quite easily as long as all the steps are followed and the equipment is suitable. As the protocol described is a surgical method, all the steps appear critical and must be followed properly. It is critical to be trained for animal experiments, especially in rabbit handling and anesthesia. Do not hesitate to ask for professional anesthetist and veterinary help. It is critical to insist on the daily visual monitoring of animals before and after suture removal. Even if the skin from the skull is thick, abundant and loose, the cylinders’ fixation induces large tensions. If the sutures are removed too early, the wound might reopen and will require one more week of suturing and a new antibiotic treatment. In case of dehiscence out of the wound line, the surgeon will have to consider if suturing is possible or not. If not, sample rejection will have to be considered.
In addition to the critical steps described in the protocol section, the details below may be helpful in the proper implementation of this protocol. From a technical point of view, as the rabbits used are young and small, it is important to use a two-step intubation, as described in the protocol. The second (and final) tube is too large to be used in first intubation, and there is a real risk of “wrong way” that may be deleterious or even fatal.
Depending on the materials tested, it may be interesting to take some fresh blood sample from the ear IV line that is already placed. This could provide a good method to pre-infuse the bone substitute material with natural cells and growth factors. Freshly coagulated blood may also provide an ideal sham sample.
The way the intramedullary holes are drilled should be very helpful for future histological processing and histomorphometrical analysis. In effect, as long as the holes are (i) in the same location, (ii) the biopsies are similarly oriented and (iii) the cutting process is standardized (i.e., same thickness, same cut level), the fields that are evaluated are equivalent and their comparison is highly pertinent. As a matter of standardization, it may be of real interest to calibrate the volume-quantity of material to fill the cylinder, and to prepare it in advance, in a sterile way.
From a scientific point of view, depending on the products or hypothesis tested, it may be important to avoid placing the cylinders onto bone sutures. Some specific stem cells are present in these structures35, different from the mesenchymal bone stem cells that are involved in the mechanisms of intramembranous ossification, i.e., the bone regeneration process encountered in the orofacial area. Therefore, a real bias may occur in case of misplacement.
One of the major advantages of the calvarial model is the use of live microtomography by which the bone growth may be followed on a single animal as a longitudinal study. This strategy may largely reduce the number of euthanized animals and therefore respect the “3R rule”36. Depending on the microtomograph device used (resolution, space available for the animal), variations will occur in terms of strategy of anesthesia (IV, gas, etc.), as well as in the image resolution and the relevance of the resulting analysis.
We routinely use a microtomograph dedicated to animal experiments (e.g., Quantum GX) for routine checks. A typical experiment starts with a sedation, as described in step 2.3.1. Anesthesia is then maintained with 2% isoflurane in pure oxygen. This allows the animal to breathe calmly during a scanning time of about 14 min (99 kV/88 µA with a resolution of 20 μm); this is sufficient for checking basic parameters (engraftment, cylinder fixation, semiquantitative analysis of the bone growth, etc.). When looking for a precise qualitative and quantitative analysis, a very fine tuning of the scanning time, resolution, anesthesia and animal positioning will be needed.
Existing models developed for the assessement of osteoconduction, osteogenesis, osteoinduction as well as vasculogenesis, are numerous, especially in the orofacial field. Due to ethical and economic considerations, our purpose will be limited to rabbit or smaller animals. Apart from the skull, the locations in which the material may be tested are the mandibulla37,38,39,40,41,42,43,44,45,46,47, the diastema48 or the incisive sockets (after teeth extraction)49,50,51. Rats, mice, guinea pigs and rabbits may be used for each of the models described below. Briefly, a critical defect is drilled (or a socket is created by incisive extraction), filled with the material sample and then covered by a membrane. Apart from the obvious fact that only two samples can be tested in each of these locations (one sample per mandible side or per incisive socket), the surgical procedures are also largely more difficult and invasive. Accesses to the site are limited and if one is using rat or mice, the difficulty is further increased by the size of the animals. Finally, the critical size of the defects (i.e., dimensions that do not allow a spontaneous bone regeneration) is not well-defined and varies from one animal to another.
Besides these generalities, specific constraints may arise depending on the anatomical location subjected to the defect and to the subsequent analysis. As an example, in the mandibular model, teeth roots are present within the defect. This may interfere with the material and modify the bone regeneration process. The defect could be placed more distally on the ramus, but in that case, the bone would be very thin (two corticals affixed with an extremely thin medullary space) and the resulting defect volume may be too small45,46,47. Given these examples of limitations and depending on the model, large deviations may be anticipated in terms of material volume to be tested, quality and quantity of data collected.
As the evaluation of a material requires a maximum of standardization and calibration, the calvarial model appears ideal. Access is very easy, calibration and standardization are facilitated by the use of defined cylinders and 4 samples may be assessed simultaneously. Furthermore, live tomography may be used and finally a large decrease in animals to be euthanized may be anticipated.
The authors have nothing to disclose.
The authors are indebted to Geistlich AG (Wolhusen, CH) and the Osteology foundation (Lucerne, CH) (grant n°18-049) for their support, as well as Global D (Brignais, FR) for providing the screws. A particular thanks goes to Dr B. Schaefer from Geistlich. We are also grateful to Eliane Dubois and Claire Herrmann for their excellent histological processing and their precious advices. Finally, we warmly acknowledge Xavier Belin, Sylvie Roulet and the entire team of Pr Walid Habre, “experimental surgery Dpt” ,for their remarkable technical assistance.
Drugs | |||
Enrofloxacine Baytril 10% | Bayer | Antibiotic | |
Fentanyl | Bischel | For analgesia | |
Ketalar 50mg/ml | Pfizer | Ketamine for anesthesia | |
Lidohex | Bichsel | Lubricating gel for the eyes | |
Opsite | Smith and Nephew | 66004978 | Sprayable dressing |
Povidone iodine 10%, Betadine | Mundipharma | anti-infective agent | |
Propofol 2% | Braun | 3538710 | For anesthesia |
Rapidocain 2% | sintetica | Local anesthesia | |
Ringer-acetate | Fresenius Kabi | Volume compensation | |
Rompun 2% | Bayer | Xylazin for anesthesia | |
Sevoflurane 5% | Abbvie | For anesthesia | |
Sterile saline | Sintetica | ||
Temgesic | Reckitt Benckiser | Buprenorphine hydrochloride, analgesia | |
Thiopental Inresa | Ospediala | For anesthesia | |
Xylocaine 10% spray | Astra Zeneca | For intubation | |
Name | Company | Catalog Number | Comments |
Equipment | |||
Fresenius Vial pilot C | Imexmed | Infusion pump | |
Heated pad | Harvard Apparatus | ||
Suction dominant 50 | Medela | ||
Suction tubing Optimus | Promedical | 80342.2 | |
Surgical motor | Schick dental | Qube | Drilling of intramedullary holes |
Ventilation | Maquet Servo1 | ||
Name | Company | Catalog Number | Comments |
Material | |||
Cylinders and caps | Boutyplast | Customized | composition: PEEK (poly ether ether ketone) |
Manual self-retaining shaft | GlobalD | ACT1K | |
Mobile handle for self-retaining shaft | GlobalD | MTM | |
Self- drilling screws | GlobalD | VA1.2KL4 | cross-drive screws composed by Titanium grade5, ISO 5832-3 |
Name | Company | Catalog Number | Comments |
Surgical tray | |||
Endotracheal tube Shiley diameter 2,5mm | Covidien | 86233 | For intubation |
Endotracheal tube Shiley diameter 4,9mm | Covidien | 107-35G | For intubation |
Ethicon prolene 4-0 | Ehticon | 8581H | Non-resorbable suture |
Forceps | Marcel Blanc | BD027R | 145 mm |
Intubation catheter | Cook medical | Guide for intubation | |
Needlle holder | Marcel Blanc | BM008R | |
Needles BD Microlance3 | Becton Dickinson | 300300/304622 | 26G; 18G |
Periosteal | HU-Friedy | P9X | |
Round surgical burs | Patterson | 78000 | 0.8 mm in diameter, Drilling of intramedullary holes |
Scalpel | Swann-Morton | n°10 and n°15 | |
Scissors | Marcel Blanc | 00657 | 180 mm |
Syringes Omnifix | Braun | 4616057V | 5ml, 10ml and 50ml |
Venflon G22 | Braun | 42690985-01 | Vasofix safety for the ear iv line |